International Journal of Mathematical Combinatorics, vol. 2/2018

Florentin Smarandache

ISSN 1937 - 1055 VOLUME 2, INTERNATIONAL MATHEMATICAL JOURNAL 2018 OF COMBINATORICS EDITED BY THE MADIS OF CHINESE ACADEMY OF SCIENCES AND ACADEMY OF MATHEMATICAL COMBINATORICS & APPLICATIONS, USA June, 2018 Vol.2, 2018 ISSN 1937-1055 International Journal of Mathematical Combinatorics (www.mathcombin.com) Edited By The Madis of Chinese Academy of Sciences and Academy of Mathematical Combinatorics & Applications, USA June, 2018 Aims and Scope: The International J.Mathematical Combinatorics (ISSN 1937-1055) is a fully refereed international journal, sponsored by the MADIS of Chinese Academy of Sciences and published in USA quarterly comprising 110-160 pages approx. per volume, which publishes original research papers and survey articles in all aspects of Smarandache multi-spaces, Smarandache geometries, mathematical combinatorics, non-euclidean geometry and topology and their applications to other sciences. Topics in detail to be covered are: Smarandache multi-spaces with applications to other sciences, such as those of algebraic multi-systems, multi-metric spaces,· · · , etc.. Smarandache geometries; Topological graphs; Algebraic graphs; Random graphs; Combinatorial maps; Graph and map enumeration; Combinatorial designs; Combinatorial enumeration; Differential Geometry; Geometry on manifolds; Low Dimensional Topology; Differential Topology; Topology of Manifolds; Geometrical aspects of Mathematical Physics and Relations with Manifold Topology; Applications of Smarandache multi-spaces to theoretical physics; Applications of Combinatorics to mathematics and theoretical physics; Mathematical theory on gravitational fields; Mathematical theory on parallel universes; Other applications of Smarandache multi-space and combinatorics. Generally, papers on mathematics with its applications not including in above topics are also welcome. It is also available from the below international databases: Serials Group/Editorial Department of EBSCO Publishing 10 Estes St. Ipswich, MA 01938-2106, USA Tel.: (978) 356-6500, Ext. 2262 Fax: (978) 356-9371 http://www.ebsco.com/home/printsubs/priceproj.asp and Gale Directory of Publications and Broadcast Media, Gale, a part of Cengage Learning 27500 Drake Rd. Farmington Hills, MI 48331-3535, USA Tel.: (248) 699-4253, ext. 1326; 1-800-347-GALE Fax: (248) 699-8075 http://www.gale.com Indexing and Reviews: Mathematical Reviews (USA), Zentralblatt Math (Germany), Referativnyi Zhurnal (Russia), Mathematika (Russia), Directory of Open Access (DoAJ), International Statistical Institute (ISI), International Scientific Indexing (ISI, impact factor 1.743), Institute for Scientific Information (PA, USA), Library of Congress Subject Headings (USA). Subscription A subscription can be ordered by an email directly to Prof.Linfan Mao,PhD The Editor-in-Chief of International Journal of Mathematical Combinatorics Chinese Academy of Mathematics and System Science Beijing, 100190, P.R.China Email: maolinfan@163.com Price: US$48.00 Editorial Board (4th) Editor-in-Chief Linfan MAO Chinese Academy of Mathematics and System Science, P.R.China and Academy of Mathematical Combinatorics & Applications, USA Email: maolinfan@163.com Shaofei Du Capital Normal University, P.R.China Email: dushf@mail.cnu.edu.cn Deputy Editor-in-Chief Yuanqiu Huang Hunan Normal University, P.R.China Email: hyqq@public.cs.hn.cn Xiaodong Hu Chinese Academy of Mathematics and System Science, P.R.China Email: xdhu@amss.ac.cn Guohua Song Beijing University of Civil Engineering and H.Iseri Architecture, P.R.China Mansfield University, USA Email: songguohua@bucea.edu.cn Email: hiseri@mnsfld.edu Editors Arindam Bhattacharyya Jadavpur University, India Email: bhattachar1968@yahoo.co.in Said Broumi Hassan II University Mohammedia Hay El Baraka Ben M’sik Casablanca B.P.7951 Morocco Xueliang Li Nankai University, P.R.China Email: lxl@nankai.edu.cn Guodong Liu Huizhou University Email: lgd@hzu.edu.cn W.B.Vasantha Kandasamy Indian Institute of Technology, India Email: vasantha@iitm.ac.in Junliang Cai Beijing Normal University, P.R.China Email: caijunliang@bnu.edu.cn Ion Patrascu Fratii Buzesti National College Craiova Romania Yanxun Chang Beijing Jiaotong University, P.R.China Email: yxchang@center.njtu.edu.cn Han Ren East China Normal University, P.R.China Email: hren@math.ecnu.edu.cn Jingan Cui Ovidiu-Ilie Sandru Beijing University of Civil Engineering and Politechnica University of Bucharest Architecture, P.R.China Romania Email: cuijingan@bucea.edu.cn ii International Journal of Mathematical Combinatorics Mingyao Xu Peking University, P.R.China Email: xumy@math.pku.edu.cn Guiying Yan Chinese Academy of Mathematics and System Science, P.R.China Email: yanguiying@yahoo.com Y. Zhang Department of Computer Science Georgia State University, Atlanta, USA Famous Words: You can pay attention to the fact, in which case you’ll probably become a mathematician, or you can ignore it, in which case you’ll probably become a physicist. By Len Evans, a professor in Northwestern University. International J.Math. Combin. Vol.2(2018), 1-12 Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold Tamalika Dutta (Department of Mathematics, Jadavpur University, Kolkata-700032, India) Arindam Bhattacharyya (Department of Mathematics, Jadavpur University, Kolkata-700032, India) E-mail: tamalika.bagnan@gmail.com, bhattachar1968@yahoo.co.in Abstract: In this paper we have studied quasi conformal curvature tensor, Ricci tensor, projective curvature tensor, pseudo projective curvature tensor in Lorentzian β-Kenmotsu manifold admitting Ricci soliton and conformal Ricci soliton. Key Words: Trans-Sasakian manifold, β-Kenmotsu manifold, Lorentzian β-Kenmotsu manifold, Ricci soliton, conformal Ricci flow. AMS(2010): 53C25, 35K65, 53C44, 53D10, 53D15. §1. Introduction Hamilton started the study of Ricci flow [12] in 1982 and proved its existence. This concept was developed to answer Thurston’s geometric conjecture which says that each closed three manifold admits a geometric decomposition. Hamilton also [11]classified all compact manifolds with positive curvature operator in dimension four. Since then, the Ricci flow has become a powerful tool for the study of Riemannian manifolds, especially for those manifolds with positive curvature. Perelman also did an excellent work on Ricci flow [15], [16]. The Ricci flow equation is given by ∂g = −2S ∂t (1.1) on a compact Riemannian manifold M with Riemannian metric g. A solution to the Ricci flow is called a Ricci soliton if it moves only by a one-parameter group of diffeomorphism and scaling.Ramesh Sharma [18], M. M. Tripathi [19], Bejan, Crasmareanu [4]studied Ricci soliton in contact metric manifolds also. The Ricci soliton equation is given by £X g + 2S + 2λg = 0, (1.2) where £X is the Lie derivative, S is Ricci tensor, g is Riemannian metric, X is a vector field and λ is a scalar. The Ricci soliton is said to be shrinking, steady and expanding according as 1 The first author is supported by DST ’Inspire’ of India. Reference no: IF140748. September 10, 2017, Accepted May 8, 2018. 2 Received 2 Tamalika Dutta and Arindam Bhattacharyya λ is negative, zero and positive respectively. In 2005, A.E. Fischer [10] introduced the concept of conformal Ricci flow which is a variation of the classical Ricci flow equation. In classical Ricci flow equation the unit volume constraint plays an important role but in conformal Ricci flow equation scalar curvature R is considered as constraint. Since the conformal geometry plays an important role to constrain the scalar curvature and the equations are the vector field sum of a conformal flow equation and a Ricci flow equation, the resulting equations are named as the conformal Ricci flow equations. The conformal Ricci flow equation on M where M is considered as a smooth closed connected oriented n-manifold(n > 3), is defined by the equation [10] g ∂g + 2(S + ) = −pg ∂t n (1.3) and r = −1, where p is a scalar non-dynamical field(time dependent scalar field), r is the scalar curvature of the manifold and n is the dimension of manifold. In 2015, N. Basu and A. Bhattacharyya [3] introduced the notion of conformal Ricci soliton and the equation is as follows £X g + 2S = [2λ − (p + 2 )]g. n (1.4) The equation is the generalization of the Ricci soliton equation and it also satisfies the conformal Ricci flow equation. An almost contact metric structure (φ, ξ, η, g) on a manifold M is called a trans-Sasakian structure [14] if the product manifold belongs to the class W4 where W4 is a class of Hermitian manifolds which are closely related to locally conformal Kaehler manifolds [6]. A trans-Sasakian structure of type (0, 0), (0, β) and (α, 0) are cosymplectic [5], β−Kenmotsu [13], and α−Sasakian [13], respectively. §2. Preliminaries A differentiable manifold of dimension n is called Lorentzian Kenmotsu manifold [2] if it admits a (1, 1) tensor field φ, a covarient vector field ξ, a 1-form η and Lorentzian metric g which satisfy on M respectively such that φ2 X = X + η(X)ξ, g(X, ξ) = η(X), (2.1) η(ξ) = −1, η(φX) = 0, φξ = 0, (2.2) g(φX, φY ) = g(X, Y ) + η(X)η(Y ), (2.3) for all X, Y ∈ χ(M ). If Lorentzian Kenmotsu manifold M satisfies ∇X ξ = β[X − η(X)ξ], (∇X φ)Y = β(g(φX, Y )ξ − η(Y )φX), (2.4) 3 Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold (∇X η)Y = αg(φX, Y ), (2.5) where ∇ denotes the operator of covariant differentiation with respect to the Lorentzian metric g. Then the manifold M is called Lorentzian β−Kenmotsu manifold. Furthermore, on an Lorentzian β-Kenmotsu manifold M the following relations hold [1], [17]: η(R(X, Y )Z) = β 2 [g(X, Z)η(Y ) − g(Y, Z)η(X)], (2.6) R(ξ, X)Y = β 2 [η(Y )X − g(X, Y )ξ], (2.7) R(X, Y )ξ = β 2 [η(X)Y − η(Y )X], (2.8) S(X, ξ) = −(n − 1)β 2 η(X), (2.9) Qξ = −(n − 1)β 2 ξ, (2.10) S(ξ, ξ) = (n − 1)β 2 , (2.11) where β is some constant, R is the Riemannian curvature tensor, S is the Ricci tensor and Q is the Ricci operator given by S(X, Y ) = g(QX, Y ) for all X, Y ∈ χ(M ). Now from definition of Lie derivative we have (£ξ g)(X, Y ) = (∇ξ g)(X, Y ) + g(β[X − η(X)ξ], Y ) + g(X, β[Y − η(Y )ξ]) = 2βg(X, Y ) − 2βη(X)η(Y ). (2.12) Applying Ricci soliton equation (1.2) in (2.12) we get S(X, Y ) = = = 1 1 [−2λg(X, Y )] − [2βg(X, Y ) − 2βη(X)η(Y )] 2 2 −λg(X, Y ) − βg(X, Y ) + βη(X)η(Y ) Ág(X, Y ) + βη(X)η(Y ), (2.13) where Á = (−λ − β), which shows that the manifold is η-Einstein. Also QX = ÁX + βη(X)ξ, (2.14) S(X, ξ) = (Á + β)η(X) = Aη(X). (2.15) If we put X = Y = ei in (2.13) where {ei } is the orthonormal basis of the tangent space T M where T M is a tangent bundle of M and summing over i, we get R(g) = Án + β. Proposition 2.1 A Lorentzian β-Kenmotsu manifold admitting Ricci soliton is η-Einstein. 4 Tamalika Dutta and Arindam Bhattacharyya Again applying conformal Ricci soliton (1.4) in (2.12) we get S(X, Y ) = = 2 1 1 [2λ − (p + )]g(X, Y ) − [2βg(X, Y ) − 2βη(X)η(Y )] 2 n 2 B́g(X, Y ) + βη(X)η(Y ), (2.16) where 1 2 [2λ − (p + )] − β, 2 n which also shows that the manifold is η-Einstein. Also QX = B́X + βη(X)ξ, B́ = S(X, ξ) = (B́ + β)η(X) = Bη(X). (2.17) (2.18) (2.19) If we put X = Y = ei in (2.16) where {ei } is the orthonormal basis of the tangent space T M where T M is a tangent bundle of M and summing over i, we get r = B́n + β. For conformal Ricci soliton r(g) = −1. So −1 = B́n + β which gives B = n1 (−β − 1). Comparing the values of B from (2.17) with the above equation we get λ= 1 2 1 (β(n − 1) − 1) + (p + ) n 2 n Proposition 2.2 A Lorentzian β-Kenmotsu manifold admitting conformal Ricci soliton is η-Einstein and the value of the scalar λ= 1 1 2 (β(n − 1) − 1) + (p + ). n 2 n §3. Lorentzian β-Kenmotsu Manifold Admitting Ricci Soliton, Conformal Ricci Soliton and R(ξ, X).C̃ = 0 Let M be a n dimensional Lorentzian β-Kenmotsu manifold admitting Ricci soliton (g, V, λ). Quasi conformal curvature tensor C̃ on M is defined by C̃(X, Y )Z = aR(X, Y )Z + b[S(Y, Z)X − S(X, Z)Y + g(Y, Z)QX − g(X, Z)QY ] a r ][ + 2b][g(Y, Z)X − g(X, Z)Y ], (3.1) −[ 2n + 1 2n where r is scalar curvature. Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold 5 Putting Z = ξ in (3.1) we have C̃(X, Y )ξ = aR(X, Y )ξ + b[S(Y, ξ)X − S(X, ξ)Y + g(Y, ξ)QX − g(X, ξ)QY ] a r ][ + 2b][g(Y, ξ)X − g(X, ξ)Y ]. −[ 2n + 1 2n (3.2) Using (2.1), (2.8), (2.14), (2.15) in (3.2) we get C̃(X, Y )ξ = [−aβ 2 + Ab + Áb − [ r a ][ + 2b]](η(Y )X − η(X)Y ). 2n + 1 2n Let D = −aβ 2 + Ab + Áb − [ a r ][ + 2b], 2n + 1 2n so we have C̃(X, Y )ξ = D(η(Y )X − η(X)Y ). (3.3) Taking inner product with Z in (3.3) we get −η(C̃(X, Y )Z) = D[η(Y )g(X, Z) − η(X)g(Y, Z)]. (3.4) Now we consider that the Lorentzian β-Kenmotsu manifold M which admits Ricci soliton is quasi conformally semi symmetric i.e. R(ξ, X).C̃ = 0 holds in M , which implies R(ξ, X)(C̃(Y, Z)W ) − C̃(R(ξ, X)Y, Z)W − C̃(Y, R(ξ, X)Z)W − C̃(Y, Z)R(ξ, X)W = 0, (3.5) for all vector fields X, Y, Z, W on M . Using (2.7) in (3.5) and putting W = ξ we get η(C̃(Y, Z)ξ)X − g(X, C̃(Y, Z)ξ)ξ − η(Y )C̃(X, Z)ξ + g(X, Y )C̃(ξ, Z)ξ −η(Z)C̃(Y, X)ξ + g(X, Z)C̃(Y, ξ)ξ − η(ξ)C̃(Y, Z)X + g(X, ξ)C̃(Y, Z)ξ = 0. (3.6) Taking inner product with ξ in (3.6) and using (2.2), (3.3) we obtain g(X, C̃(Y, Z)ξ) + η(C̃(Y, Z)X) = 0. (3.7) Putting Z = ξ in (3.7) and using (3.3) we get −Dg(X, Y ) − Dη(X)η(Y ) + η(C̃(Y, Z)X) = 0. (3.8) Now from (3.1) we can write C̃(Y, ξ)X = aR(Y, ξ)X + b[S(ξ, X)Y − S(Y, X)ξ + g(ξ, X)QY − g(Y, X)Qξ] r a −[ ][ + 2b][g(ξ, X)Y − g(Y, X)ξ]. 2n + 1 2n (3.9) 6 Tamalika Dutta and Arindam Bhattacharyya Taking inner product with ξ and using (2.2), (2.7), (2.9), (2.10) in (3.9) we get η(C̃(Y, ξ)X) = aη(β 2 (g(X, Y )ξ − η(X)Y )) + b[Aη(X)η(Y ) + S(X, Y ) + η(X)(Áη(Y ) a r ][ + 2b][η(X)η(Y ) + g(X, Y )]. −βη(Y )) − g(X, Y )(−Á + β)] − [ 2n + 1 2n After a long simplification we have η(C̃(Y, ξ)X) = r a ][ + 2b]] 2n + 1 2n a r ][ + 2b]] + bS(X, Y ). +η(X)η(Y )[2Áb − aβ 2 − [ 2n + 1 2n g(X, Y )[Áb − bβ − aβ 2 − [ (3.10) Putting (3.10) in (3.5) we get ρg(X, Y ) + ση(X)η(Y ) = S(X, Y ), where ρ= (3.11) 1 r a [D + bβ + aβ 2 − Áb + [ ][ + 2b]] b 2n + 1 2n and σ= r a 1 [D + aβ 2 − 2Áb + [ ][ + 2b]]. b 2n + 1 2n So from (3.11) we conclude that the manifold becomes η-Einstein manifold. Thus we can write the following theorem: Theorem 3.1 If a Lorentzian β-Kenmotsu manifold admits Ricci soliton and is quasi conformally semi symmetric i.e. R(ξ, X).C̃ = 0, then the manifold is η-Einstein manifold where C̃ is quasi conformal curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. If a Lorentzian β-Kenmotsu manifold admits conformal Ricci soliton then after a brief calculation we can also establish that the manifold becomes η-Einstein, only the values of constants ρ, σ will be changed which would not hamper our main result. Hence we can state the following theorem: Theorem 3.2 A Lorentzian β-Kenmotsu manifold admitting conformal Ricci soliton and is quasi conformally semi symmetric i.e. R(ξ, X).C̃ = 0, then the manifold is η-Einstein manifold where C̃ is quasi conformal curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. §4. Lorentzian β-Kenmotsu Manifold Admitting Ricci Soliton, Conformal Ricci Soliton and R(ξ, X).S = 0 Let M be a n dimensional Lorentzian β-Kenmotsu manifold admitting Ricci soliton (g, V, λ). Now we consider that the tensor derivative of S by R(ξ, X) is zero i.e. R(ξ, X).S = 0. Then the 7 Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold Lorentzian β-Kenmotsu manifold admitting Ricci soliton is Ricci semi symmetric which implies S(R(ξ, X)Y, Z) + S(Y, R(ξ, X)Z) = 0. (4.1) Using (2.13) in (4.1) we get Ág(R(ξ, X)Y, Z) + βη(R(ξ, X)Y )η(Z) + Ág(Y, R(ξ, X)Z) + βη(Y )η(R(ξ, X)Z) = 0. (4.2) Using (2.7) in (4.2) we get Ág(β 2 [η(Y )X − g(X, Y )ξ], Z) + Ág(Y, β 2 [η(Z)X − g(X, Z)ξ]) + βη(β 2 [η(Y )X− g(X, Y )ξ])η(Z) + βη(Y )η(β 2 [η(Z)X − g(X, Z)ξ]) = 0. (4.3) Using (2.2) in (4.3) we have Áβ 2 η(Y )g(X, Z) − Áβ 2 η(Z)g(X, Y ) + Áβ 2 η(Z)g(X, Y ) − Áβ 2 η(Y )g(X, Z) +β 3 η(Y )η(X)η(Z) + β 3 g(X, Y )η(Z) + β 3 η(Y )η(X)η(Z) + β 3 g(X, Z)η(Y ) = 0. (4.4) Putting Z = ξ in (4.4) and using (2.2) we obtain g(X, Y ) = −η(X)η(Y ). Hence we can state the following theorem: Theorem 4.1 If a Lorentzian β-Kenmotsu manifold admits Ricci soliton and is Ricci semi symmetric i.e. R(ξ, X).S = 0, then g(X, Y ) = −η(X)η(Y ) where S is Ricci tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. If a Lorentzian β-Kenmotsu manifold admits conformal Ricci soliton then by similar calculation we can obtain the same result. Hence we can state the following theorem: Theorem 4.2 A Lorentzian β-Kenmotsu manifold admitting conformal Ricci soliton and is Ricci semi symmetric i.e. R(ξ, X).S = 0, then g(X, Y ) = −η(X)η(Y ) where S is Ricci tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. §5. Lorentzian β-Kenmotsu Manifold Admitting Ricci Soliton, Conformal Ricci Soliton and R(ξ, X).P = 0 Let M be a n dimensional Lorentzian β-Kenmotsu manifold admitting Ricci soliton (g, V, λ). The projective curvature tensor P on M is defined by P (X, Y )Z = R(X, Y )Z − 1 [S(Y, Z)X − S(X, Z)Y ]. 2n (5.1) Here we consider that the manifold is projectively semi symmetric i.e. R(ξ, X).P = 0 holds. 8 Tamalika Dutta and Arindam Bhattacharyya So R(ξ, X)(P (Y, Z)W ) − P (R(ξ, X)Y, Z)W − P (Y, R(ξ, X)Z)W − P (Y, Z)R(ξ, X)W = 0, (5.2) for all vector fields X, Y, Z, W on M . Using (2.7) and putting Z = ξ in (5.2) we have η(P (Y, ξ)W )X − g(X, P (Y, ξ)W )ξ − η(Y )P (X, ξ)W + g(X, Y )P (ξ, ξ)W −η(ξ)P (Y, X)W + g(X, ξ)P (Y, ξ)W − η(W )P (Y, ξ)X + g(X, W )P (Y, ξ)ξ = 0. (5.3) Now from (5.1) we can write P (X, ξ)Z = R(X, ξ)Z − 1 [S(ξ, Z)X − S(X, Z)ξ]. n−1 (5.4) Using (2.7), (2.15) in (5.4) we get P (X, ξ)Z = β 2 g(X, Z)ξ + 1 A S(X, Z)ξ + ( − β 2 )η(Z)X. n−1 n−1 (5.5) Putting (5.5) and W = ξ in (5.3) and after a long calculation we get A A 1 S(X, Y )ξ + ( + β 2 )η(X)Y − g(X, Y )ξ n−1 n−1 n−1 −( A + β 2 )η(Y )X = 0. n−1 (5.6) Taking inner product with ξ in (5.6) we obtain S(X, Y ) = −Ag(X, Y ), which clearly shows that the manifold in an Einstein manifold. Thus we can conclude the following theorem: Theorem 5.1 If a Lorentzian β-Kenmotsu manifold admits Ricci soliton and is projectively semi symmetric i.e. R(ξ, X).P = 0 holds, then the manifold is an Einstein manifold where P is projective curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. If a Lorentzian β-Kenmotsu manifold admits conformal Ricci soliton then using the same calculation we can obtain similar result, only the value of constant A will be changed which would not hamper our main result. Hence we can state the following theorem: Theorem 5.2 A Lorentzian β-Kenmotsu manifold admitting conformal Ricci soliton and is projectively semi symmetric i.e. R(ξ, X).P = 0 holds, then the manifold is an Einstein manifold where P is projective curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold 9 §6. Lorentzian β-Kenmotsu Manifold Admitting Ricci Soliton, Conformal Ricci Soliton and R(ξ, X).P̃ = 0 Let M be a n dimensional Lorentzian β-Kenmotsu manifold admitting Ricci soliton (g, V, λ). The pseudo projective curvature tensor P̃ on M is defined by P̃ (X, Y )Z = aR(X, Y )Z + b[S(Y, Z)X − S(X, Z)Y ] r a − [ + b][g(Y, Z)X − g(X, Z)Y ]. n n−1 (6.1) Here we consider that the manifold is pseudo projectively semi symmetric i.e. R(ξ, X).P̃ = 0 holds. So R(ξ, X)(P̃ (Y, Z)W ) − P̃ (R(ξ, X)Y, Z)W − P̃ (Y, R(ξ, X)Z)W − P̃ (Y, Z)R(ξ, X)W = 0, (6.2) for all vector fields X, Y, Z, W on M . Using (2.7) and putting W = ξ in (6.2) we have η(P̃ (Y, Z)ξ)X − g(X, P̃ (Y, Z)ξ)ξ − η(Y )P̃ (X, Z)ξ + g(X, Y )P̃ (ξ, Z)ξ −η(Z)P̃ (Y, X)ξ + g(X, Z)P̃ (Y, ξ)ξ − η(ξ)P̃ (Y, Z)X + η(X)P̃ (Y, Z)ξ = 0. (6.3) Now from (6.1) we can write P̃ (X, Y )ξ = aR(X, Y )ξ + b[S(Y, ξ)X − S(X, ξ)Y ] + a r [ + b][g(Y, ξ)X − g(X, ξ)Y ]. (6.4) n n−1 Using (2.1), (2.8), (2.15) in (6.4) and after a long calculation we get P̃ (X, Y )ξ = ϕ(η(X)Y − θ(Y )X), (6.5) a where ϕ = (aβ 2 − Ab − nr [ n−1 + b]). Using (6.5) and putting Z = ξ in (6.3) we obtain P̃ (Y, ξ)X + ϕη(X)Y − ϕg(X, Y )ξ = 0. (6.6) Taking inner product with ξ in (6.6) we get η(P̃ (Y, ξ)X) + ϕη(X)η(Y ) − ϕg(X, Y ) = 0. (6.7) Again from (6.1) we can write P̃ (X, ξ)Z = a(X, ξ)Z + b[S(ξ, Z)X − S(X, Z)ξ] + a r [ + b][g(ξ, Z)X − g(X, Z)ξ]. (6.8) n n−1 10 Tamalika Dutta and Arindam Bhattacharyya Using (2.1), (2.7), (2.15) in (6.8) we get P̃ (X, ξ)Z = aβ 2 [g(X, Z)ξ − η(Z)X] + b[Aη(Z)X − S(X, Z)ξ] a r + b][g(ξ, Z)X − g(X, Z)ξ]. + [ n n−1 (6.9) Taking inner product with ξ and replacing X by Y , Z by X in (6.9) we have η(P̃ (Y, ξ)X) = aβ 2 [−g(X, Y ) − η(X)η(Y )] + b[Aη(X)η(Y ) + S(X, Y )]+ r a [ + b][η(X)η(Y ) − g(X, Y )]. n n−1 (6.10) S(X, Y ) = T g(X, Y ) + U η(X)η(Y ), (6.11) Using (6.10) in (6.7) and after a brief simplification we obtain a a + b] − ϕ] and U = − 1b [ϕ + nr [ n−1 + b] + Ab − aβ 2 ]. where T = − 1b [−aβ 2 − nr [ n−1 From (6.11) we can conclude that the manifold is η-Einstein. Thus we have the following theorem: Theorem 6.1 If a Lorentzian β-Kenmotsu manifold admits Ricci soliton and is pseudo projectively semi symmetric i.e. R(ξ, X).P̃ = 0 holds, then the manifold is η Einstein manifold where P̃ is pseudo projective curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. If a Lorentzian β-Kenmotsu manifold admits conformal Ricci soliton then by following the same calculation we would obtain the same result, only the constant value of T and U will be changed. Hence we can state the following theorem: Theorem 6.2 A Lorentzian β-Kenmotsu manifold admitting conformal Ricci soliton and is pseudo projectively semi symmetric i.e. R(ξ, X).P̃ = 0 holds, then the manifold is η Einstein manifold where P̃ is pseudo projective curvature tensor and R(ξ, X) is derivation of tensor algebra of the tangent space of the manifold. §7. An Example of a 3-Dimensional Lorentzian β-Kenmotsu Manifold In this section we construct an example of a 3-dimensional Lorentzian β-kenmotsu manifold.To construct this, we consider the three dimensional manifold M = {(x, y, z) ∈ R3 : z 6= 0} where (x, y, z) are the standard coordinates in R3 . The vector fields e1 = e−z ∂ ∂ ∂ , e2 = e−z , e3 = e−z ∂x ∂y ∂z are linearly independent at each point of M. 11 Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold Let g be the Lorentzian metric defined by g(e1 , e1 ) = 1, g(e2 , e2 ) = 1, g(e3 , e3 ) = −1, g(e1 , e2 ) = g(e2 , e3 ) = g(e3 , e1 ) = 0. Let η be the 1-form which satisfies the relation η(e3 ) = −1. have Let φ be the (1, 1) tensor field defined by φ(e1 ) = −e2 , φ(e2 ) = −e1 , φ(e3 ) = 0. Then we φ2 (Z) = Z + η(Z)e3 , g(φZ, φW ) = g(Z, W ) + η(Z)η(W ) for any Z, W ∈ χ(M 3 ). Thus for e3 = ξ, (φ, ξ, η, g) defines an almost contact metric structure on M. Now, after calculating we have [e1 , e3 ] = e−z e1 , [e1 , e2 ] = 0, [e2 , e3 ] = e−z e2 . The Riemannian connection ∇ of the metric is given by the Koszul’s formula which is 2g(∇X Y, Z) = Xg(Y, Z) + Y g(Z, X) − Zg(X, Y ) −g(X, [Y, Z]) − g(Y, [X, Z]) + g(Z, [X, Y ]). (7.1) By Koszul’s formula we get ∇e1 e1 = e−z e3 , ∇e2 e1 = 0, ∇e3 e1 = 0, ∇e1 e2 = 0, ∇e2 e2 = ‘e−z e3 , ∇e3 e2 = 0, ∇e1 e3 = e−z e1 , ∇e2 e3 = e−z e2 , ∇e3 e3 = 0. From the above we have found that β = e−z and it can be easily shown that M 3 (φ, ξ, η, g) is a Lorentzian β-kenmotsu manifold. The results established in this note can be verified on this manifold. References [1] C.S. Bagewadi and E. Girish Kumar, Note on trans-Sasakian manifolds, Tensor.N.S., (65)(1) (2004), 80-88. [2] N.S. Basavarajappa, C.S. Bagewadi, D.G. Prakasha, Some results on :orentzian β- Kenmotsu manifolds, Annals of the University of Craiova, Math. Comp. Sci. Ser., (35) 2008, 7-14. 12 Tamalika Dutta and Arindam Bhattacharyya [3] Nirabhra Basu, Arindam Bhattacharyya, Conformal Ricci soliton in Kenmotsu manifold, Global Journal of Advanced Research on Classical and Modern Geometries, (2015), 15-21. [4] Bejan, Crasmareanu, Ricci solitons in manifolds with quasi-constant curvature, Publ. Math. Debrecen, (78)(1) (2011), 235-243. [5] D. E. Blair, Contact manifolds in Riemannian geometry, Lectures notes in Mathematics, Springer-Verlag, Berlin, (509) (1976), 146. [6] S. Dragomir and L. Ornea, Locallly conformal Kaehler geometry, Progress in Mathematics, 155, Birkhauser Boston, Inc., Boston, MA, 1998. [7] T. Dutta, N. Basu, A. Bhattacharyya, Some curvature identities on an almost conformal gradient shrinking RICCI soliton, Journal of Dynamical Systems and Geometric Theories, (13)(2) (2015), 163-178. [8] Tamalika Dutta, Arindam Bhattacharyya and Srabani Debnath, Conformal Ricci Soliton in Almost C(λ) Manifold, International J.Math. Combin., (3) (2016), 17-26. [9] T. Dutta, N. Basu, A. Bhattacharyya, Almost conformal Ricci solitons on 3-dimensional trans-Sasakian manifold, Hacettepe Journal of Mathematics and Statistics, (45)(5) (2016), 1379-1392. [10] A. E. Fischer, An introduction to conformal Ricci flow, Class.Quantum Grav., (21) (2004), S171-S218. [11] R. S. Hamilton, The Ricci flow on surfaces, Contemporary Mathematics, (71) (1988), 237261. [12] R. S. Hamilton, Three manifold with positive Ricci curvature, J.Differential Geom., (17)(2) (1982), 255-306. [13] K. Kenmotsu, A class of almost contact Riemannian manifolds, Tohoku Math. J., (24) (1972), 93-103. [14] A. J. Oubina, New classes of almost Contact metric structures, Publ.Math.Debrecen, (32) (1985),187-193. [15] G. Perelman, The entropy formula for the Ricci flow and its geometric applications, http://arXiv.org/abs/math/0211159, (2002) 1-39. [16] G. Perelman,Ricci flow with surgery on three manifolds, http://arXiv.org/abs/math/ 0303109, (2003) 1-22. [17] D.G. Prakasha, C.S. Bagewadi and N.S. Basavarajappa, On Lorentzian β-Kenmotsu manifolds, Int.Journal of Math., (19)(2) (2008), 919-927. [18] Ramesh Sharma, Almost Ricci solitons and K-contact geometry, Monatsh Math., (2014), 175:621-628. [19] M. M. Tripathi, Ricci solitons in contact metric manifolds, arXiv:0801,4222v1, [mathDG], (2008). International J.Math. Combin. Vol.2(2018), 13-23 Some Properties of Conformal β -Change H.S.Shukla and Neelam Mishra (Department of Mathematics and Statistics, D. D. U. Gorakhpur University, Gorakhpur (U.P.)-273009,India) E-mail: profhsshuklagkp@rediffmail.com, pneelammishra@gmail.com Abstract: We have considered the conformal β-change of the Finsler metric given by L(x, y) → L̄(x, y) = eσ(x) f (L(x, y), β(x, y)), where σ(x) is a function of x, β(x, y) = bi (x)y i is a 1-form on the underlying manifold M n ,and f (L(x, y), β(x, y)) is a homogeneous function of degree one in L and β.We have studied quasi-C-reducibility, C-reducibility and semi-C-reducibility of the Finsler space with this metric. We have also calculated V-curvature tensor and T-tensor of the space with this changed metric in terms of v-curvature tensor and T-tensor respectively of the space with the original metric. Key Words: Conformal change, β-change, Finsler space, quasi-C-reducibility, C- reducibility, semi-C-reducibility, V-curvature tensor, T-tensor. AMS(2010): 53B40, 53C60. §1. Introduction Let F n = (M n , L) be an n-dimensional Finsler space on the differentialble manifold M n equipped with the fundamental function L(x,y).B.N.Prasad and Bindu Kumari and C. Shibata [1,2] have studied the general case of β-change,that is, L∗ (x, y) = f (L, β),where f is positively homogeneous function of degree one in L and β, and β given by β(x, y) = bi (x)y i is a one- form on M n . The β-change of special Finsler spaces has been studied by H.S.Shukla, O.P.Pandey and Khageshwar Mandal [7]. The conformal theory of Finsler space was initiated by M.S. Knebelman [12] in 1929 and has been investigated in detail by many authors (Hashiguchi [8] ,Izumi[4,5] and Kitayama [9]). The conformal change is defined as L∗ (x, y) = eσ(x) L(x, y), where σ(x) is a function of position only and known as conformal factor. In 2008, Abed [15,16] introduced the change L̄(x, y) = eσ(x) L(x, y) + β(x, y), which he called a β-conformal change, and in 2009 and 2010,Nabil L.Youssef, S.H.Abed and S.G. Elgendi [13,14] introduced the transformation L̄(x, y) = f (eσ L, β), which is β-change of conformally changed Finsler metric L. They have not only established the relationships between some important tensors of (M n , L) and the corresponding tensors of (M n , L̄), but have also studied several properties of this change. 1 Received January 22, 2018, Accepted May 12, 2018. 14 H.S.Shukla and Neelam Mishra We have changed the order of combination of the above two changes in our paper [6], where we have applied β-change first and conformal change afterwards, i.e., L̄(x, y) = eσ(x) f (L(x, y), β(x, y)), (1.1) where σ(x) is a function of x, β(x, y) = bi (x)y i is a 1-form. We have called this change as conformal β-change of Finsler metric. In this paper we have investigated the condition under which a conformal β-change of Finsler metric leads a Douglas space into a Douglas space.We have also found the necessary and sufficient conditions for this change to be a projective change. In the present paper,we investigate some properties of conformal β-change. The Finsler space equipped with the metric L̄ given by (1.1) will be denoted by F¯n .Throughout the paper the quantities corresponding to F¯n will be denoted by putting bar on the top of them.We shall denote the partial derivatives with respect to xi and y i by ∂i and ∂˙i respectively. The Fundamental quantities of F n are given by 2 L = hij + li lj , li = ∂˙i L. gij = ∂˙i ∂˙j 2 Homogeneity of f gives Lf1 + βf2 = f, (1.2) where subscripts 1 and 2 denote the partial derivatives with respect to L and β respectively. Differentiating above equations with respect to L and β respectively, we get Lf12 + βf22 = 0 and Lf11 + βf21 = 0. (1.3) f11 /β 2 = (−f12 )/Lβ = f22 /L2 , (1.4) f11 = β 2 ω, f12 = −Lβω, f22 = L2 ω, (1.5) Hence we have which gives where Weierstrass function ω is positively homogeneous of degree -3 in L and β. Therefore Lω1 + βω2 + 3ω = 0, (1.6) where ω1 and ω2 are positively homogeneous of degree -4 in L and β. Throughout the paper we frequently use the above equations without quoting them. Also we have assumed that f is not linear function of L and β so that ω 6= 0. The concept of concurrent vector field has been given by Matsumoto and K. Eguchi [11] and S. Tachibana [17], which is defind as follows: The vector field bi is said to be a concurrent vector field if bi|j = −gij bi |j = 0, (1.7) where small and long solidus denote the h- and v-covariant derivatives respectively. It has been Some Properties of Conformal β -Change 15 proved by Matsumoto that bi and its contravariant components bi are functions of coordinates alone. Therefore from the second equation of (1.7),we have Cijk bi = 0. The aim of this paper is to study some special Finsler spaces arising from conformal β-change of Finsler metric,viz., quasi-C-reducible, C-reducible and semi-C-reducible Finsler spaces. Further, we shall obtain v-curvature tensor and T-tensor of this space and connect them with v-curvature tensor and T-tensor respectively of the original space. §2. Metric Tensor and Angular Metric Tensor of F̄ n Differentiating equation (1.1) with respect to y i we have l̄i = eσ (f1 li + f2 bi ). (2.1) Differentiating (2.1) with respect to y j , we get h̄ij = e2σ where mi = bi − f f1 hij + f L2 ωmi mj , L (2.2) β Li . L From (2.1) and (2.2) we get the following relation between metric tensors of F n and F̄ n : ḡij = e2σ f f1 pβ gij − li lj + (f L2 ω + f22 )bi bj + p(bi lj + bj li ) , L L (2.3) where p = f1 f2 − f βLω. The contravariant components ḡ ij of the metric tensor of F̄ n , obtainable from ḡ ij ḡjk = δki , are as follows: L4 ω i j fβ L ij pL3 pL2 i j i j ij −2σ j i g + 3 − ∆f2 l l − ḡ = e b b − 2 (l b + l b ) , (2.4) f f1 f f1 t L 2 f f1 t f f1 t where li = g ij lj , b2 = bi bi , bi = g ij bj , g ij is the reciprocal tensor of gij of F n , and t = f1 + L3 ω∆, ∆ = b2 − β2 . L2 f li + f2 mi , (b) ∂˙i f1 = −eσ βLωmi , L σ 2 ˙ (c)∂i f2 = e L ωmi , (d) ∂˙i p = −βqLmi , 3ω (e) ∂˙i ω = − li + ω2 mi , (f )∂˙i b2 = −2C..i , L 2β (g) ∂˙i ∆ = −2C..i − 2 mi , L (a) ∂˙i f = eσ (2.5) (2.6) 16 H.S.Shukla and Neelam Mishra 3q (a) ∂˙i q = − li , (b) ∂˙i t = −2L3 ωC..i + [L3 ∆ω2 − 3βLω]mi , L 3q (c) ∂˙i q = − li + (4f2 ω2 + 3ω 2 L2 + f ω22 )mi . L (2.7) §3. Cartan’s C-Tensor and C-Vectors of F̄ n Cartan’s covariant C-tensor Cijk of F n is defined by C̄ijk = 1˙ ˙ ˙ 2 ∂i ∂j ∂k L = ∂˙k gij 4 and Cartan’s C-vectors are defined as follows: i jk Ci = Cijk g jk , C i = Cjk g . (3.1) We shall write C 2 = C i Ci . Under the conformal β-chang (1.1) we get the following relation between Cartan’s C-tensors of F n and F̄ n : p qL2 2σ f f1 (3.2) Cijk + (hij mk + hjk mi + hki mj ) + mi mj mk . C̄ijk = e L 2L 2 We have (a) mi li = 0, β2 = ∆ = b i mi , L2 (c) gij mi = hij mi = mj . (b) mi bi = b2 − (3.3) From (2.1), (2.3), (2.4) and (3.2), we get h C̄ij = p qL3 (hij mh + hhj mi + hhi mj ) + mj mk mh 2f f1 2f f1 pL∆ 2pL + qL4 ∆ L mj mk n h , − C.jk nh − 2 hjk nh − ft 2f f1 t 2f 2 f1 t h Cij + (3.4) where nh = f L2 ωbh + plh and hij = g il hlj , C.ij = Crij br , C..i = Crji br bj and so on. Proposition 3.1 Let F̄ n = (M n , L̄) be an n-dimensional Finsler space obtained from the conformal β-change of the Finsler space F n = (M n , L), then the normalized supporting element l̄i , angular metric tensor h̄ij , fundamental metric tensor ḡij and (h)hv-torsion tensor C̄ijk of F̄ n are given by (2.1), (2.2), (2.3) and (3.2), respectively. F n From (2.4),(3.1),(3.2) and (3.4) we get the following relations between the C-vectors of of and F̄ n and their magnitudes C̄i = Ci − L3 ωCi.. + µmi , (3.5) 17 Some Properties of Conformal β -Change where µ= p(n + 1) 3pL3 ω∆ qL3 ∆(1 − L3 ω∆) − + ; 2f f1 2f f1 2f f1 C̄ i = where e−2σ L i C + M i, f f1 µe−2σ L i L4 ω i m − C − Ci − e2σ L3 ωCi.. + µ∆ M = f f1 f f1 .. i and C̄ 2 = (3.6) L3 ω i L b + yi f f1 ft e−2σ 2 C + λ, p (3.7) where λ = 2µe−2σ L e−2σ L 3 − L ω∆ µ2 ∆ + C. f f1 f f1 − (1 + 2µ∆) L3 ω + 1 − 3µ + e2σ L2 ωf f1 C. L3 ωC... +L3 ωC..r e4σ Lωf 2 f12 Ci.. − µ∆ L3 ωbr − e2σ L2 ωf f1 C..r − 2C r . §4. Special Cases of F̄ n In this section, following Matsumoto [10], we shall investigate special cases of F̄ n which is conformally β-changed Finsler space obtained from F n . Definition 4.1 A Finsler space (M n , L) with dimension n ≥ 3 is said to be quasi-C-reducible if the Cartan tensor Cijk satisfies Cijk = Qij Ck + Qjk Ci + Qki Cj , (4.1) where Qij is a symmetric indicatory tensor. The equation (3.2) can be put as C̄ijk = e2σ f f1 1 Cijk + π(ijk) L 6 3p , hij + qL2 mi mj mk L where π(ijk) represents cyclic permutation and sum over the indices i, j and k. Putting the value of mk from equation (3.5) in the above equation, we get C̄ijk = e 2σ f f1 1 Cijk + π(ijk) L 6µ 3p hij + qL2 mi mj )(C̄k − Ck + L3 ωCk.. L . 18 H.S.Shukla and Neelam Mishra Rearranging this equation, we get C̄ijk = f f1 1 3p Cijk + π(ijk) hij + qL2 mi mj C̄k L 6µ L 1 3p + π(ijk) . hij + qL2 mi mj L3 ωCk.. − Ck 6µ L e2σ Further rearrangment of this equations gives C̄ijk = π(ijk) (H̄ij C̄k ) + Uijk , where H̄ij = 3p e2σ 6µ {( L hij Uijk = e 2σ (4.2) + qL2 mi mj ), and f f1 1 Cijk + π(ijk) L 6µ 3p hij + qL2 mi mj L 3 L ωCk.. − Ck (4.3) Since H̄ij is a symmetric and indicatory tensor,therefore from equation (4.2) we have the following theorem. Theorem 4.1 Conformally β-changed Finsler space F̄ n is quasi-C-reducible iff the tensor Uijk of equation (4.3) vanishes identically. We obtain a generalized form of Matsumoto’s result [10] as a corollary of the above theorem. Corollary 4.1 If F n is Reimannian space, then the conformally β-changed Finsler space F̄ n is always a quasi-C-reducible Finsler space. Definition 4.2 A Finsler space (M n , L) of dimension n ≥ 3 is called C-reducible if the Cartan tensor Cijk is written in the form Cijk = 1 (hij Ck + hki Cj + hjk Ci ). n+1 (4.4) 1 Define the tensor Gijk = Cijk − (n+1) (hij Ck + hki Cj + hjk Ci ). It is clear that Gijk is symmetric and indicatory. Moreover, Gijk vanishes iff F n is C-reducible. Proposition 4.1 Under the conformal β-change(1.1), the tensor Ḡijk associated with the space F̄ n has the form f f1 Ḡijk = e2σ Gijk + Vijk (4.5) L where Vijk = 1 π(ijk) {(e2σ (n + 1)(α1 hij + α2 mi mj )mk + e2σ ωL2 mi mj Ck (n + 1) +e2σ L2 ω(f f1 hij + L3 ωmi mj )Ck.. }, α1 = e2σ p µf f1 e2σ − , 2L L(n + 1) α2 = (4.6) e2σ qL2 µe2σ ωL2 − . 6 (n + 1) Some Properties of Conformal β -Change 19 From (4.5) we have the following theorem. Theorem 4.2 Conformally β-changed Finsler space F̄ n is C-reducible iff F n is C-reducible and the tensor Vijk given by (4.6) vanishes identically. Definition 4.3 A Finsler space (M n , L) of dimension n ≥ 3 is called semi-C-reducible if the Cartan tensor Cijk is expressible in the form: Cijk = s r (hij Ck + hki Cj + hjk Ci ) + 2 Ci Cj Ck , n+1 C (4.7) where r and s are scalar functions such that r + s = 1. Using equations (2.2), (3.5) and (3.7) in equation (3.2), we have C̄ijk = e2σ ∆L(f1 q − 3pω) f f1 p (h̄ij C̄k + h̄ki C̄j + h̄jk C̄i ) + Cijk + C̄ C̄ C̄ i j k . L 2µf f1 2f f1 µtC̄ 2 If we put p(n + 1) ′ ∆L(f1 q − 3pω) ,s = , 2µf f1 2f f1 µt r′ = we find that r′ + s′ = 1 and s′ r′ f f1 Cijk + (h̄ij C̄k + h̄ki C̄j + h̄jk C̄i ) + 2 C̄i C̄j C̄k . C̄ijk = e2σ L n+1 C̄ (4.8) From equation (4.8) we infer that F̄ n is semi-C- reducible iff Cijk = 0, i.e. iff F n is a Reimannian space. Thus we have the following theorem. Theorem 4.3 Conformally β-changed Finsler space F̄ n is semi-C-reducible iff F n is a Riemannian space. §5. v-Curvature Tensor of F̄ n The v-curvature tensor [10] of Finsler space with fundamental function L is given by r r Shijk = Cijr Chk − Cikr Chj Therefore the v-curvature tensor of conformally β-changed Finsler space F̄ n will be given by r r . − C̄ikr C̄hj S̄hijk = C̄ijr C̄hk From equations (3.2)and(3.4), we have r C̄ijr C̄hk f f1 p r Cijr Chk + (Cijk mh + Cijh mk + Cihk mj L 2L f f1 L 2 ω pf1 (C.ij hhk + Chk hij ) − C.ij C.hk +Chjk mi ) + 2Lt t = e2σ (5.1) 20 H.S.Shukla and Neelam Mishra p2 ∆ L2 (qf1 − 2pω) hhk hij + (C.ij mk mh + C.hk mi mj ) 4f Lt 2t p(p + L3 q∆) p2 (hij mh mk + (hij mh mk + hhk mi mj ) + 4Lf t 4Lf f1 +hhk mi mj + hhj mi mk + hhi mj mk + hjk mi mh + hik mh mj ) L2 (2pqt + (qf1 − 2pω)(2p + L3 q∆)) mi mj mh mk . + 4f f1 t + (5.2) We get the following relation between v-curvature tensors of (M n , L) and (M n , L̄): S̄hijk = e2σ f f1 Shijk + dhj dik − dhk dij + Ehk Eij − Ehj Eik , L (5.3) where P =L s 1/2 t , Q= dij = P C.ij − Qhij + Rmi mj , (5.4) Eij = Shij + T mi mj , (5.5) L (2ωp − f1 q) p L (qf1 − ωp) pg √ ,R = √ , T = √ √ ,S = . 2L2 f ω 2f1 f ω 2L2 st 2 st Proposition 5.1 The relation between v-curvature tensors of F n and F̄ n is given by (5.3). When bi in β is a concurrent vector field,then C.ij = 0. Therefore the value of v-curvature tensor of F̄ n as given by (5.3) is reduced to the extent that dij = Rmi mj − Qhij . §6. The T-Tensor Thijk The T-tensor of F n is defined in [3] by Thijk = LChij |k +Chij lk + Chik lj + Chjk li + Cijk lh , (6.1) r r r Chij |k = ∂˙k Chij − Crij Chk − Chrj Cik − Chir Cjk . (6.2) where In this section we compute the T-tensor of F̄ n , which is given by T̄hijk = L̄C̄hij¯|k + C̄hij l̄k + C̄hik ¯lj + C̄hjk l̄i + C̄ijk l̄h , (6.3) r r r . − C̄hir C̄jk − C̄hrj C̄ik C̄hij¯|k = ∂˙k C̄hij − C̄rij C̄hk (6.4) where The derivatives of mi and hij with respect to y k are given by 1 β ∂˙k mi = − 2 hik − (li mk ), L L 1 ∂˙k hij = 2Cijk − (li hjk + lj hki ) L (6.5) Some Properties of Conformal β -Change 21 From equations (3.2)and (6.5), we have ∂˙k C̄hij f f1 p ∂k Chij + (Cijk mh + Cijh mk + Cihk mj + Chjk mi ) L L pβ p − 3 (hij hhk + hhj hik + hih hjk ) + (hjk lh mi + hhk lj mi 2L 2L2 +hhk li mj + hik lh mj + hjk li mh + hjk lh mi + hij lh mk + hhj li mk βq (hij mh mk +hik lj mk + hij lk mh + hjh lk mi + hhi lk mj ) − 2 +hhk mi mj + hhj mi mk + hhi mj mk + hjk mi mh + hik mh mj ) qL − (li mj mh mk + lj mi mh mk + lh mi mj mk + hk mi mj mh ) 2 L2 (4f2 ω2 + 3L2 ω 2 + f ω22 )mh mi mj mk . + 2 = e2σ (6.6) Using equations (6.5) and (5.2) in equation(6.4), we get C̄hij¯|k f f1 e2σ p Chij |k − (Cijk mh + Cijh mk + Cihk mj + Chjk mi ) L 2L 2f βt βq L2 p∆ 2σ 2σ −pe (hij hhk + hhj hik + hih hjk ) − e + 3 3 4f L t 4f L t 2 p2 f1 + pqf1 L3 ∆ + 3p2 (hij mh mk + hhk mi mj + hhj mi mk + hhi mj mk + 4Lf f1t e2σ p +hjk mi mh + hik mh mj ) − [lh (hjk mi + hij mk 2L2 +hik mj ) + lj (hhk mi + hik mkh + hih mk ) + li (hhk mj + hjk mh = e2σ e2σ qL (li mj mh mk 2 pf1 e2σ (C.ij hhk +lj mi mh mk + lh mi mj mk + hk mi mj mh ) − 2Lt e2σ f f1 L2 ω +C.hj hik + C.hk hij + C.ik hh + C.hi hjk + C.jk hhi ) + (C.ij C.hk t 2σ 2 e L (qf1 − 2pω) (C.ij mk mh +C.hj C.ik + C.hi C.jk ) − 2t +C.hk mi mj + C.hj mi mk + C.ik mj mh 2 2 2 2σ L (4f2 ω2 + 3L ω + f ω22 ) +C.hi mj mk + C.jk mh mi ) + e 2 2 3 3L (2pqt + (qf1 − 2pω)(2p + L q∆) − mi mj mh mk . (6.7) 4f f1 t +hhj mk ) + lk (hij mh + hjh mi + hhi mj )] − Using equations (2.1), (3.2) and (6.6) in equation (6.3), we get the following relation 22 H.S.Shukla and Neelam Mishra between T-tensors of Finsler spaces F n and F̄ n : T̄hijk f (f1 f2 + f βLω) f 2 f1 Thijk + (Cijk mh + Cijh mk + Cihk mj L2 2L f 2 f1 L 2 ω +Chjk mi ) + (C.ij C.hk + C.hj C.ik + C.hi C.jk ) t pf1 − (C.ij hhk + C.hj hik + C.hk hij + C.ik hh + C.hi hjk + C.jk hhi ) 2Lt f L2 (qf1 − 2pω) (C.ij mk mh + C.hk mi mj + C.hj mi mk − 2t p(2f βt + L2 p∆) +C.ik mj mh + C.hi mj mk + C.jk mh mi ) − (hij hhk 4L3 t 2 βqf pf2 p f1 + pqf1 L3 ∆ + 3p2 + − +hhj hik + hih hjk ) − 4Lf1t 2 L (hij mh mk + hhk mi mj + hhj mi mk + hhi mj mk + hjk mi mh 2 L (4f2 ω2 + 3L2 ω 2 + f ω22 ) +hik mh mj ) + + 2L2 f2 q 2 3L2 (2pqt + (qf1 − 2pω)(2p + L3 q∆) − mi mj mh mk . 4f1 t = e3σ (6.8) Proposition 6.1 The relation between T-tensors of F n and F̄ n is given by (6.7). If bi is a concurrent vector field in F n , then C.ij = 0. Therefore from(6.8), we have T̄hijk f 2 f1 p(2f βt + L2 p∆) T − (hij hhk + hhj hik + hih hjk ) hijk L2 4L3 t 2 p f1 + pqf1 L3 ∆ + 3p2 t βqf pf2 − (hij mh mk + hhk mi mj + − 4Lf1 t 2 L +hhj mi mk + hhi mj mk + hjk mi mh + hik mh mj ) L2 (4f2 ω2 + 3L2 ω 2 + f ω22 ) 3L2 (qf1 − 2pω)(2p + L3 q∆) + 2L2 f2 q + + 2 4Lf f1t 2 3L 2pqt mi mj mh mk . − 4Lf f1 t = e 3σ (6.9) If bi is a concurrent vector field in F n , with vanishing T-tensor then T-tensor of F n is given by T̄hijk p(2f βt + L2 p∆) (hij hhk + hhj hik + hih hjk ) 4L3 t 2 p f1 + pqf1 L3 ∆ + 3p2 t βqf pf2 − (hij mh mk + − 4Lf1 t 2 L +hhk mi mj + hhj mi mk + hhi mj mk + hjk mi mh + hik mh mj ) 2 L (4f2 ω2 + 3L2 ω 2 + f ω22 ) 3L2 2pqt + − 2 4Lf f1t 2 3 3L (qf1 − 2pω)(2p + L q∆) 2 + + 2L f2 q mi mj mh mk . 4Lf f1t = e 3σ − (6.10) Some Properties of Conformal β -Change 23 Acknowledgement The work contained in this research paper is part of Major Research Project “Certain Investigations in Finsler Geometry” financed by the U.G.C., New Delhi. References [1] B.N.Prasad and Bindu Kumari, The β-change of Finsler metric and imbedding classes of their tangent spaces, Tensor N. S., 74, (2013),48-59. [2] C. Shibata, On invariant tensors of β-change of Finsler metric, J. Math.Kyoto Univ., 24(1984),163-188. [3] F.Ikeda, On the tensor Tijkl of Finsler spaces, Tensor N. S., 33(1979),203-209. [4] H.Izumi, Conformal transformations of Finsler spaces I, Tensor N.S., 31(1977),33-41. [5] H.Izumi, Conformal transformations of Finsler spaces II. An h-conformally flat Finsler space, Tensor N.S., 33(1980),337-369. [6] H.S.Shukla and Neelam Mishra, Conformal β-changes of Finsler metric, J.Int.Acad.Phys.Sci., 21(1)(2017), 19-30. [7] H.S.Shukla, O.P.pandey and Khageshwar Mandal, The β-change of Special Finsler Spaces, International J.Math.Combin., 1(2017), 78-87. [8] M.Hashiguchi, On conformal transformation of Finsler metric, J.Math.Kyoto Univ.,16(1976),2550. [9] M.Kitayama, Geometry of Transformations of Finsler metrics, Hokkaido University of Education,Kushiro Campus, Japan, 2000. [10] M.Mastsumoto, Foundations of Finsler Geometry and Special Finsler Spaces, Kaiseisha Press, Saikawa,Otsu, 520 Japan, 1986. [11] M. Matsumoto and K. Eguchi, Finsler space admitting a concurrent vector field, Tensor N.S., 28(1974), 239-249. [12] M.S.Knebelman, Conformal geometry of generalized metric spaces, Proc.Nat.Acad. Sci., USA.,15(1929),376-379. [13] N.L.Youssef, S.H.Abed and S.G.Elgendi, Generalized β-conformal change of Finsler metrics, Int.J.Geom.Meth.Mod.Phys.,7,4(2010).ArXiv No.:math.DG/0906.5369. [14] N.L.Youssef,S.H.Abed and S.G.Elgendi, Generalized β-conformal change and special Finsler spaces, Int.J.Geom.Meth.Mod.Phys., 8,6(2010). ArXiv No.:math.DG/1004.5478v3. [15] S.H.Abed, Conformal β-changes in Finsler spaces, Proc.Math.Phys.Soc.Egypt, 86(2008),7989.ArXiv No.:math.DG/0602404. [16] S.H.Abed, Cartan connections associated with a β-conformal change in Finsler geometry, Tensor N.S., 70(2008),146-158. ArXiv No.:math.DG/0701491. [17] S.Tachibana, On Finsler spaces which admit a concurrent vector field, Tensor N. S., 1 (1950),1-5. International J.Math. Combin. Vol.2(2018), 24-32 Equitable Coloring on Triple Star Graph Families K.Praveena (Department of Computer Science, Dr.G.R. Damodaran College of Science, Coimbatore-641014, Tamilnadu, India) M.Venkatachalam (Department of Mathematics, Kongunadu Arts and Science College, Coimbatore-641029, Tamilnadu, India) E-mail: praveena4ks@gmail.com, venkatmaths@gmail.com Abstract: An equitable k-coloring of a graph G is a proper k-coloring of G such that the sizes of any two color class differ by at most one. In this paper we investigate the equitable chromatic number for the Central graph, Middle graph, Total graph and Line graph of Triple star graph K1,n,n,n denoted by C(K1,n,n,n ), M (K1,n,n,n ), T (K1,n,n,n ) and L(K1,n,n,n ) respectively. Key Words: Equitable coloring, Smarandachely equitable k-coloring, triple star graph, central graph, middle graph, total graph and line graph. AMS(2010): 05C15, 05C78. §1. Introduction A graph consist of a vertex set V (G) and an edge set E(G). All Graphs in this paper are finite, loopless and without multiple edges. We refer the reader [8] for terminology in graph theory. Graph coloring is an important research problem [7, 10]. A proper k-coloring of a graph is a labelling f : V (G) → {1, 2, · · · , k} such that the adjacent vertices have different labels. The labels are colors and the vertices with same color form a color class. The chromatic number of a graph G, written as χ(G) is the least k such that G has a proper k-coloring. Equitable colorings naturally arise in some scheduling, partitioning and load balancing problems [11,12]. In 1973, Meyer [4] introduced first the notion of equitable colorability. In 1998, Lih [5] surveyed the progress on the equitable coloring of graphs. We say that a graph G = (V, E) is equitably k-colorable if and only if its vertex set can be partitioned into independent sets {V1 , V2 , · · · , Vk } ⊂ V such that ||Vi | − |Vj || ≤ 1 holds for every pair (i, j). The smallest integer k for which G is equitable k-colorable is known as the equitable chromatic number [1,3] of G and denoted by χ= (G). On the other hand, if V can be partitioned into independent sets {V1 , V2 , · · · , Vk } ⊂ V with ||Vi | − |Vj || ≥ 1 holds for every pair (i, j), such a k-coloring is called a Smarandachely equitable k-coloring. In this paper, we find the equitable chromatic number χ= (G) for central, line, middle and total graphs of triple star graph. 1 Received September 14, 2017, Accepted May 15, 2018. Equitable Coloring on Triple Star Graph Families 25 §2. Preliminaries For a given graph G = (V, E) we do a operation on G, by subdividing each edge exactly once and joining all the non adjacent vertices of G. The graph obtained by this process is called central graph of G [1] and is denoted by C(G). The line graph [6] of a graph G, denoted by L(G) is a graph whose vertices are the edges of G and if u, v ∈ E(G) then uv ∈ E(L(G)) if u and v share a vertex in G. Let G be a graph with vertex set V (G) and edge set E(G). The middle graph [2] of G denoted by M (G) is defined as follows. The vertex set of M (G) is V (G) ∪ E(G) in which two vertices x, y are adjacent in M (G) if the following condition hold: (1) x, y ∈ E(G) and x, y are adjacent in G; (2) x ∈ V (G), y ∈ E(G) and they are incident in G. Let G be a graph with vertex set V (G) and edge set E(G). The total graph [1,2] of G is denoted by T (G) and is defined as follows. The vertex set of T (G) is V (G) ∪ E(G). Two vertices x, y in the vertex set of T (G) is adjacent in T (G), if one of the following holds: (1) x, y are in V (G) and x is adjacent to y in G; (2) x, y are in E(G) and x, y are adjacent in G; (3) x is in V (G), y is in E(G) and x, y are adjacent in G. Triple star K1,n,n,n [9] is a tree obtained from the double star [2] K1,n,n by adding a new pendant edge of the existing n pendant vertices. It has 3n + 1 vertices and 3n edges. §3. Equitable Coloring on Central Graph of Triple Star Graph Algorithm 1. Input: The number ‘n’ of k1,n,n,n ; Output: Assigning equitable colouring for the vetices in C(K1,n,n,n ). begin for i = 1 to n { V1 = {ei } C(ei ) = i; V2 = {ai }; C(ai ) = i; } V3 = {v}; C(v) = n + 1; 26 K.Praveena and M.Venkatachalam for i = 2 to n { V4 = {vi }; C(vi ) = i − 1; V5 = {wi }; C(wi ) = i − 1; V6 = {ui }; C(ui ) = i − 1; } C(v1 ) = n; C(w1 ) = n; C(u1 ) = n; for i = 1 to 5 { V7 = {si }; C(si ) = n + 1; } for i = 6 to n { V8 = {si }; C(si ) = i; } V = V1 ∪ V2 ∪ V3 ∪ V4 ∪ V5 ∪ V6 ∪ V7 ∪ V8 ; end Theorem 3.1 For any triple star graph K1,n,n,n the equitable chromatic number χ= [C(K1,n,n,n )] = n + 1. Proof Let {vi : 1 ≤ i ≤ n}, {wi : 1 ≤ i ≤ n} and {ui : 1 ≤ i ≤ n} be the vertices in K1,n,n,n . The vertex v is adjacent to the vertices vi (1 ≤ i ≤ n). The vertices vi (1 ≤ i ≤ n) is adjacent to the vertices wi (1 ≤ i ≤ n) and the vertices wi (1 ≤ i ≤ n) is adjacent to the vertices ui (1 ≤ i ≤ n). By the definition of central graph on K1,n,n,n , let the edges vvi , vi wi and wi ui (1 ≤ i ≤ n) of K1,n,n,n be subdivided by the vertices ei , ai , si (1 ≤ i ≤ n) respectively. Equitable Coloring on Triple Star Graph Families 27 Clearly, V [C(K1,n,n,n )] [ [ = {v} {vi : 1 ≤ i ≤ n} {wi : 1 ≤ i ≤ n} [ [ {ui : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} [ {ai : 1 ≤ i ≤ n} ∪ {si : 1 ≤ i ≤ n} The vertices v and ui (1 ≤ i ≤ n) induces a clique of order n + 1 (say kn+1 ) in [C[K1,n,n,n ]]. Therefore χ= [C(K1,n,n,n )] ≥ n + 1 Now consider the vertex set V [C(K1,n,n,n )] and the color class C = {c1 , c2 , c3 , · · · cn+1 }. Assign an equitable coloring to C(K1,n,n,n ) by Algorithm 1. Therefore χ= [C(K1,n,n,n )] ≤ n + 1. An easy check shows that ||vi | − |vj || ≤ 1. Hence χ= [C(K1,n,n,n )] = n + 1. §4. Equitable Coloring on Line graph of Triple Star Graph Algorithm 2. Input: Output: The number ‘n’ of K1,n,n,n ; Assigning equitable coloring for the vertices in L(K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; V2 = {si }; C(si ) = i; } for i = 2 to n { V3 = {ai }; C(ai ) = i − 1; } 2 28 K.Praveena and M.Venkatachalam C(a1 ) = n; V = V1 ∪ V2 ∪ V3 ; end Theorem 4.1 For any triple star graph K1,n,n,n the equitable chromatic number, χ= [L(K1,n,n,n )] = n. Proof Let {vi : 1 ≤ i ≤ n}, {wi : 1 ≤ i ≤ n} and {ui : 1 ≤ i ≤ n} be the vertices in K1,n,n,n . The vertex v is adjacent to the vertices vi (1 ≤ i ≤ n) with edges ei (1 ≤ i ≤ n). The vertices vi (1 ≤ i ≤ n) is adjacent to the vertices wi (1 ≤ i ≤ n) with edges ai (1 ≤ i ≤ n). The vertices wi (1 ≤ i ≤ n) is adjacent to the vertices ui (1 ≤ i ≤ n) with edges si (1 ≤ i ≤ n). By the definition of line graph on K1,n,n,n the edges ei , ai , si (1 ≤ i ≤ n) of K1,n,n,n are the vertices of L(K1,n,n,n ). Clearly V [L(K1,n,n,n )] = {ei : 1 ≤ i ≤ n} [ [ {ai : 1 ≤ i ≤ n} {si : 1 ≤ i ≤ n} The vertices ei (1 ≤ i ≤ n) induces a clique of order n (say Kn ) in L(K1,n,n,n ). Therefore χ= [L(K1,n,n,n )] ≥ n. Now consider the vertex set V [L(K1,n,n,n )] and the color class C = {c1 , c2 , · · · cn }. Assign an equitable coloring to L(K1,n,n,n ) by Algorithm 2. Therefore χ= [L(K1,n,n,n )] ≤ n. An easy check shows that ||vi | − |vj || ≤ 1. Hence χ= [L(K1,n,n,n )] = n. §5. Equitable Coloring on Middle and Total Graphs of Triple Star Graph Algorithm 3. Input: Output: The number ‘n′ of K1,n,n,n ; Assigning equitable coloring for the vertices in M (K1,n,n,n ) and T (K1,n,n,n ). 2 Equitable Coloring on Triple Star Graph Families begin for i = 1 to n { V1 = {ei }; C(ei ) = i; V2 = {si }; C(si ) = i; } V3 = {v}; C(v) = n + 1; for i = 2 to n { V4 = {vi }; C(vi ) = i − 1; } C(v1 ) = n; for i = 3 to n { V5 = {ai }; C(ai ) = i − 2; } C(a1 ) = n + 1; C(a2 ) = n + 1; for i = 4 to n { V6 = {wi } C(wi ) = i − 3; } C(w1 ) = n − 1; C(w2 ) = n; 29 30 K.Praveena and M.Venkatachalam C(w3 ) = n + 1; for i = 1 to n { V7 = {ui }; C(ui ) = i + 1; } V = V1 ∪ V2 ∪ V3 ∪ V4 ∪ V5 ∪ V6 ∪ V7 end Theorem 5.1 For any triple star graph K1,n,n,n the equitable chromatic number, χ= [M (K1,n,n,n )] = n + 1, n ≥ 4. Proof Let V (K1,n,n,n ) = {v} S S {vi : 1 ≤ i ≤ n}∪ {wi : 1 ≤ i ≤ n} {ui : 1 ≤ i ≤ n}. By the definition of middle graph on K1,n,n,n each edge vvi , vi wi and wi ui (1 ≤ i ≤ n) in K1,n,n,n are subdivided by the vertices ei , wi , si (1 ≤ i ≤ n) respectively. Clearly V [M (K1,n,n,n )] [ [ = {v} {vi : 1 ≤ i ≤ n} {wi : 1 ≤ i ≤ n} [ [ {ui : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} [ [ {ai : 1 ≤ i ≤ n} {si : 1 ≤ i ≤ n} The vertices v and ei (1 ≤ i ≤ n) induces a clique of order n+ 1 (say kn+1 ) in [M (K1,n,n,n )]. Therefore χ= [M (K1,n,n,n )] ≥ n + 1. Now consider the vertex set V [M (K1,n,n,n )] and the color class C = {c1 , c2 , · · · cn+1 }. Assign an equitable coloring to M (K1,n,n,n ) by Algorithm 3. Therefore χ= M [(K1,n,n,n )] ≤ n + 1, ||vi | − |vj || ≤ 1. Hence χ= [M (K1,n,n,n )] = n + 1 ∀n ≥ 4. 2 Theorem 5.2 For any triple star graph K1,n,n,n the equitable chromatic number, χ= [T (K1,n,n,n )] = n + 1, n ≥ 4. S S S Proof Let V (K1,n,n,n ) = {v} {vi : 1 ≤ i ≤ n} {wi : 1 ≤ i ≤ n} {ui : 1 ≤ i ≤ n} and S E(K1,n,n,n ) = {ei : 1 ≤ i ≤ n} {ai : 1 ≤ i ≤ n} ∪ {si : 1 ≤ i ≤ n}. Equitable Coloring on Triple Star Graph Families 31 By the definition of Total graph, the edge vvi , vi wi and wi ui (1 ≤ i ≤ n) of K1,n,n,n be subdivided by the vertices ei , ai and si (1 ≤ i ≤ n) respectively. Clearly V [T (K1,n,n,n )] = [ [ {v} {vi : 1 ≤ i ≤ n} {wi : 1 ≤ i ≤ n} [ [ {ui : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} [ [ {ai : 1 ≤ i ≤ n} {si : 1 ≤ i ≤ n}. The vertices v and ei (1 ≤ i ≤ n) induces a clique of order n + 1 (say kn+1 ) in T (K1,n,n,n ). Therefore χ= [T (K1,n,n,n )] ≥ n + 1, n ≥ 4. Now consider the vertex set V (T (K1,n,n,n)) and the color class C = {c1 , c2 , · · · , cn+1 }. Assign an equitable coloring to T (K1,n,n,n ) by Algorithm 3. Therefore χ= [T (K1,n,n,n)] ≤ n + 1, n ≥ 4, ||vi | − |vj || ≤ 1. Hence χ= [T (K1,n,n,n)] = n + 1, ∀n ≥ 4. 2 References [1] Akbar Ali. M.M, Kaliraj.K and Vernold Vivin.J, On equitable coloring of central graphs and total graphs, Electronic Notes in Discrete Mathematics, 33,(2009),1-6. [2] Venkatachalam .M, Vernold Vivin.J, Akbar Ali. M.M, Star Coloring and Equitable coloring on Star and double star graph families, Mathematical and Computational Models, Recent trends, ICMCM 2009, P.S.G College of Technology, Coimbatore, Narosa Publishing House, New Delhi, India, December (2009), 286-291. [3] B.L.Chen and K.W.Lih , Equitable coloring of trees, J.Combin. Theory Ser.B, 61 (1994),8387. [4] W. Meyer, Equitable coloring, Amer. Math. Monthly, 80 (1973) 920-922. [5] K.-W. Lih, The equitable coloring of graphs, in: D.-Z. Du, P.M. Pardalos (Eds.), in: Handbook of Combinatorial Optimization, Vol.3, Kluwer Academic Publishers, 1998, pp. 543-566. [6] Frank Harary, Graph theory, Narosa Publishing Home, 1969. [7] Li J.W., Zhang Z.F., Chen X.E., Sun Y.R., A Note on adjacent strong edge coloring of K(n,m), Acte Mathematicae Application Sinica, 22(2): 273-276 (2006). [8] D.B. West, Introduction to Graph Theory (Second ed.), Prentice-Hall, Upper Saddle River, NJ, 2001. [9] Akhlak Mansuri, On harmonious chromatic number of triple star graph, Journal of Hyperstructures, 5(1) (2016),26-32. [10] Chen X.E., Zhang Z.F., AVDTC number of generalized Halin graphs with maximum degree at least 6, Acte Mathematicae Application Sinica, 24(1): 55-58 (2008) 32 K.Praveena and M.Venkatachalam [11] B. F. Smith, P. E. Bjorstad and W. D. Gropp, Domain Decomposition; Parallel Multilevel Methods for Elliptic Partial Differential Equations, Cambridge University Press, Cambridge, 224 p. (1996). [12] A. Tucker, Perfect graphs and an application to optimizing municipal services, SIAM Review, 15 (1973), 585-590. International J.Math. Combin. Vol.2(2018), 33-43 On the Tangent Vector Fields of Striction Curves Along the Involute and Bertrandian Frenet Ruled Surfaces Şeyda Kılıçoğlu (Faculty of Education, Department of Mathematics, Başkent University, Ankara, Turkey) Süleyman Şenyurt and Abdussamet Çalışkan (Faculty of Arts and Sciences, Department of Mathematics, Ordu University, Ordu, Turkey) E-mail: seyda@baskent.edu.tr,senyurtsuleyman@hotmail.com, abdussamet65@gmail.com Abstract: In this paper we consider nine special ruled surfaces associated to an involute of a curve α and its Bertrand mate α∗∗ with k1 6= 0. They are called as involute Frenet ruled and Bertrandian Frenet ruled surfaces, because of their generators which are the Frenet vector fields of curve α. First we give the striction curves of all Frenet ruled surfaces. Then the striction curves of involute and Bertrandian Frenet ruled surfaces are given in terms of the Frenet apparatus of the curve α. Some results are given on the striction curves of involute and Bertrand Frenet ruled surfaces based on the tangent vector fields in E3 . Key Words: Frenet ruled surface, involute Frenet ruled surface, Bertrandian Frenet ruled surface, evolute-involute curve, Bertrand curve pair, striction curves. AMS(2010): 53A04, 53A05. §1. Introduction A ruled surface can always be described (at least locally) as the set of points swept by a moving straight line. A ruled surface is one which can be generated by the motion of a straight line in Euclidean 3 − space [2]. Choosing a directrix on the surface, i.e. a smooth unit speed curve α (s) orthogonal to the straight lines and then choosing v(s) to be Dunit vectors along E ′ the curve in the direction of the lines, the velocity vector αs and v satisfy α , v = 0 where αs = α′ . The fundamental forms of the B − scroll with null directrix and Cartan frame in the Minkowskian 3 − space are examined in [5]. The properties of some ruled surfaces are also examined in E3 [6] , [7] ,[9] and [11]. A striction point on a ruled surface ϕ(s, v) = α(s) + v.e(s) is the foot of the common normal between two consecutive generators (or ruling). To illustrate the current situation, we bring here the famous example of L. K. Graves [3], so called the B − scroll. The special ruled surfaces B − scroll over null curves with null rulings in 3dimensional Lorentzian space form has been introduced by L. K. Graves. The Gauss map of B-scrolls has been examined in [1]. Deriving a curve based on an other curve is one of the main subjects in geometry. Involute-evolute curves and Bertrand curves are of these kinds. An involute of a given curve is well-known concept in Euclidean 3 − space. We can say that evolute 1 Received November 24, 2017, Accepted May 16, 2018. 34 Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan and involute are methods of deriving a new curve based on a given curve. The involute of a curve is called sometimes evolvent and evolvents play a part in the construction of gears. The evolute is the locus of the centers of osculating circles of the given planar curve [12]. Let α and α∗ be the curves in Euclidean 3−space. The tangent lines to a curve α generate a surface called the tangent surface of α. If a curve α∗ is an involute of α, then by definition α is an evolute of α∗ . Hence if we are given a curve α, then its evolutes are the curves whose tangent lines intersect α orthogonally. By using a similar method we produce a new ruled surface based on an other ruled surface. The differential geometric elements of the involute D̃ scroll are examined in [10]. It is well-known that if a curve is differentiable in an open interval at each point then a set of three mutually orthogonal unit vectors can be constructed. We say the set of these vectors are called Frenet frame or moving frame vectors. The rates of these frame vectors along the curve define curvatures of the curve. The set whose elements are frame vectors and curvatures of a curve α is called Frenet-Serret apparatus of the curve. Let Frenet vector fields of α be V1 (s) , V2 (s) , V3 (s) and let first and second curvatures of the curve α(s) be k1 (s) and k2 (s) , respectively. Then the quantities {V1 , V2 , V3 , k1 , k2 } are called the Frenet-Serret apparatus of the curves. If a rigid object moves along a regular curve described parametrically by α(s). then we know that this object has its own intrinsic coordinate system. The Frenet formulae are also well known as      V1 0 k1 0 V̇1         V̇2  =  −k1 0 k2    V2     V3 0 −k2 0 V̇3 E D where curvature functions are defined by k1 (s) = kV1 (s)k, k2 (s) = − V2 , V̇3 . Let unit speed regular curve α : I → E3 and α∗ : I → E3 be given. If the tangent at the point α(s) to the curve α passes through the tangent at the point α∗ (s) to the curve α∗ then the curve α∗ is called the involute of the curve α, for ∀s ∈ I provided that hV1 , V1∗ i = 0. We can write α∗ (s) = α (s) + (c − s)V1 (s) (1.1) the distance between corresponding points of the involute curve in E3 is ([4],[8]) d α(s), α∗ (s) = |c − s|, c = constant, ∀s ∈ I. Theorem 1.1([4],[8]) The Frenet vectors of the involute α∗ , based on its evolute curve α are   V1∗ = V2 ,           V ∗ = −k1 V1 + k2 V3 1 2 (k12 + k22 ) 2         k V + k1 V3   V3∗ = 2 1 1 .  (k12 + k22 ) 2 (1.2) Tangent Vector Fields Of Striction Curves 35 The first and the second curvatures of involute α∗ are ′ p 2 k1 ′ ′ −k 2 2 2 k2 k1 + k2 k k1 − k k2 , k2∗ = 2 2 1 2 = , k1∗ = λk1 λk1 (k1 + k2 ) λk1 (k12 + k22 ) (1.3) where (σ − s)k1 > 0, k1 6= 0. Let α : I → E3 and α∗∗ : I → E3 be two C 2 − class differentiable unit speed curves and let V1 (s), V2 (s), V3 (s) and V1∗∗ (s), V2∗∗ (s), V3∗∗ (s) be the Frenet frames of the curves α and α∗∗ , respectively. If the principal normal vector V2 of the curve α is linearly dependent on the principal normal vector V2∗∗ of the curve α∗∗ , then the pair (α, α∗∗ ) is called a Bertrand curve pair [4], [8]. Also α∗∗ is called a Bertrand mate. If the curve α∗∗ is a Bertrand mate of α then we may write α∗∗ (s) = α (s) + λV2 (s) (1.4) If the curve α∗∗ is Bertrand mate α (s) then we have hV1∗∗ (s) , V1 (s)i = cos θ = constant. Theorem 1.2([4],[8]) The distance between corresponding points of the Bertrand curve pair in E3 is constant. Theorem 1.3([4]) If the second curvature k2 (s) 6= 0 along a curve α(s) then α(s) is called a Bertrand curve provided that nonzero real numbers λ and β λk1 + βk2 = 1 hold along the curve α(s) where s ∈ I. It follows that a circular helix is a Bertrand curve. Theorem 1.4([4]) Let α : I → E3 and α∗∗ : I → E3 be two C 2 − class differentiable unit speed curves and let the quantities {V1 , V2 , V3 , k1 , k2 } and {V1∗∗ , V2∗∗ , V3∗∗ , k1∗∗ , k2∗∗ } be Frenet-Serret apparatus of the curves α and its Bertrand mate α∗∗ respectively, then  βV1 + λV3   , V1∗∗ = p    λ2 + β 2        (1.5) V2∗∗ = V2 ,         −λV1 + βV3  ∗∗   ; λk2 > 0 V3 = p 2 λ + β2 The first and the second curvatures of the offset curve α∗∗ are given by  k1 − λ k12 + k22  βk − λk  k1∗∗ = 2 1 2 2 = , (λ + β ) k2 (λ2 + β 2 ) k22 1   . k2∗∗ = 2 (λ + β 2 ) k2 (1.6) 36 Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan Due to this theorem, we can write βk1 − λk2 = m =⇒ k ∗∗ ′ 2 k1∗∗ = k2∗∗ 1 1 = = , ∗∗ k1 βk1 − λk2 m 1 −m′ ds p =⇒ ∗∗ = p · 2 2 2 ds m k2 λ + β k2 λ2 + β 2 A differentiable one-parameter family of (straight) lines {α(u), X(u)} is a correspondence that assigns to each u ∈ I a point α(u) ∈ R3 and a vector X(u) ∈ R3 , X(u) 6= 0, so that both α(u) and X(u) depend differentiable on u. For each u ∈ I, the line L which passes through α(u) and is parallel to X(u) is called the line of the family at u. Given a one-parameter family of lines {α(u), X(u)} the parameterized surface ϕ(u, v) = α(u) + v.X(u) where u ∈ I and v ∈ R (1.7) is called the ruled surface generated by the family {α(u), X(u)}. The lines L are called the rulings and the curve α(u) is called an anchor of the surface ϕ, [2]. Theorem 1.5([2]) The striction point on a ruled surface ϕ(u, v) = α(u) + v.X(u) is the foot of the common normal between two consecutive generators (or ruling). The set of striction points defines the striction curve given by c(u) = α(u) − hα′u , Xu′ i .X(u) hXu′ , Xu′ i (1.8) where Xu′ = DT X(u). §2. On the Tangent Vector Fields of Striction Curves Along the Involute and Bertrandian Frenet Ruled Surfaces Definition 2.1 In the Euclidean 3 − space, let α(s) be the arc length curve. The equations    ϕ1 (s, u1 ) = α (s) + u1 V1 (s) ϕ2 (s, u2 ) = α (s) + u2 V2 (s)    ϕ3 (s, u3 ) = α (s) + u3 V3 (s) (2.1) are the parametrization of the ruled surface which is called V1 − scroll ( tangent ruled surface), V2 − scroll (normal ruled surface) and V3 − scroll (binormal ruled surface) respectively in [6]. Theorem 2.1([6]) The striction curves of Frenet ruled surfaces are given by the following 37 Tangent Vector Fields Of Striction Curves matrix  c1 − α   0     c2 − α  =  0    c3 − α 0 0  0 V1      0    V2  . V3 0 k1 k22 +k22 0 Theorem 2.2 The tangent vector fields T1 , T2 and T3 belonging to striction curves of Frenet ruled surface is given by  T1      = [T ] =  T 2     T3 or  T1 1 0 k22 η kc′2 (s)k ( kη1 ) kc′2 (s)k 0  0 0 1  1     T2  =  a    T3 1 where a= k22 , η kc′2 (s)k 0 ′ k1 k2 η kc′2 (s)k  0 V1  V1       V2  .  V3      c    V2  V3 0 0 b ′ b= ( kη1 ) 2 and η = k12 + k22 . , c = η kc1′ k(s) k2 k kc′2 (s)k Proof It is easy to give this matrix because we have already got the following equalities T1 (s) = T3 (s) = α′ (s) = V1 . Since c2 (s) = α(s) + k12 c′2 (s) = T2 (s) = k1 V2 , where k12 + k22 = η 6= 0, hence we have + k22 ′ k22 k1 k2 k1 V2 + V1 + V3 , η η η c′2 (s) ηk 2 V1 + (k ′ η − k1 η ′ ) V2 + ηk2 k1 V3 . = 2 1 1 ′ kc2 (s)k 2 2 4 ′ 3 ′ η k2 + (k1 ηk1 η ) 2 2.1 Involute Frenet Ruled Surfaces In this subsection, first we give the tangent, normal and binormal Frenet ruled surfaces of the involute-evolute curves. Further we write their parametric equations in terms of the Frenet apparatus of the involute-evolute curves. Hence they are called involute Frenet ruled surfaces as in the following way. 38 Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan Definition 2.2([6]) In the Euclidean 3-space, let α(s) be the arc length curve. The equations ϕ∗1 (s, v1 ) = ϕ∗2 ϕ∗3 (s, v2 ) = (s, v3 ) = α∗ (s) + v1 V1∗ (s) = α (s) + (σ − s)V1 (s) + v1 V2 (s) , ∗ α (s) + ∗ α (s) + v2 V2∗ v3 V3∗ (s) = α (s) + (σ − s)V1 (s) + v2 (s) = α (s) + (σ − s)V1 (s) + v3 −k1 V1 + k2 V3 1 (k12 + k22 ) 2 k2 V1 + k1 V3 1 (k12 + k22 ) 2 ! , ! are the parametrization of the ruled surfaces which are called involute tangent ruled surface, involute normal ruled surface and involute binormal ruled surface, respectively. We can deduce from Theorem 2.1 striction curves of the involute Frenet ruled surfaces are given by the following matrix  c∗1 − α∗   0     c∗ − α∗  =  0    2 ∗ ∗ c3 − α 0 0 0 k1∗ k1∗2 +k2∗2  V1∗     ∗  0    V2  . V3∗ 0 0 It is easy to give the following matrix for the striction curves of four Frenet ruled surfaces along the involute curve α∗ . c∗1 (s) = c∗3 (s) = α∗ (s) , k∗ α∗ (s) + ∗2 1 ∗2 V2∗ (s) . k1 + k2 c∗2 (s) = Also we can write explicit equations of the striction curves on involute Frenet ruled surfaces in terms of Frenet apparatus of an evolute curve α. Theorem 2.3 The equations of the striction curves on involute Frenet ruled surfaces in terms of Frenet apparatus of an evolute curve α are given by  c∗1 − α       c∗ − α  =     2  c∗3 − α σ−s 0 2 k 0 (σ − s) 1 − k2 +k2 1(1+m) ( 1 2) σ−s 0 0 (σ − s)k1 k2 (k12 + k22 ) (1 + m) 0  V1       V2  .  V3 Theorem 2.4 The tangent vector fields T1 ∗ , T2 ∗ , T3 ∗ of striction curves belonging to an involute Frenet ruled surface in terms of Frenet apparatus by themselves are given by  T1 ∗    V1 ∗  1 0 0    ∗   ∗ [T ∗ ] =   T2  =  a T3 ∗ 1 b∗   ∗   c∗    V2  . V3 ∗ 0 0 39 Tangent Vector Fields Of Striction Curves a∗ = η∗ ∗ ′ k1 ∗ ′ η∗ k2 k2∗ 2 k1∗ k2∗ ∗ ∗ ∗ ∗2 ∗2 ∗ , µ = , b = , c = , η = k + k . 2 1 ′ ′ ′ ∗ ∗ ∗ ∗ k1∗ c2 (s) c2 (s) η c2 (s) 2.2 Bertrandian Frenet ruled surfaces In this subsection, first we give the tangent, normal and binormal Frenet ruled surfaces of the Bertrand mate α∗∗ . Further we write their parametric equations in terms of the Frenet apparatus of the Bertrand curve α. Hence they are called Bertrandian Frenet ruled surfaces as in the following way. Definition 2.3([6]) In the Euclidean 3 − space, let α(s) be the arc length curve. The equations ϕ∗∗ 1 (s, w1 ) ϕ∗∗ 2 (s, w2 ) ϕ∗∗ 3 (s, w3 ) βV1 + λV3 = α∗∗ (s) + w1 V1∗∗ (s) = α + λV2 + w1 p , λ2 + β 2 = α∗∗ (s) + w2 V2∗∗ (s) = α + (λ + w2 ) V2 , = α∗∗ (s) + w3 V3∗∗ (s) = α + λV2 + w3 −λV1 + βV3 p λ2 + β 2 (2.2) ! , are the parametrization of the ruled surfaces which are called Bertrandian tangent ruled surface, Bertrandian normal ruled surface and Bertrandian binormal ruled surface, respectively. We can also deduce from Theorem 2.1 the striction curves of Bertrand Frenet ruled surfaces are given by the following matrix  ∗∗ c∗∗ 1 −α   0     c∗∗ − α∗∗  =  0    2 ∗∗ ∗∗ c3 − α 0 0 k1∗∗ k1∗∗2 +k2∗∗2 0 0  V1∗∗     ∗∗  0    V2  . V3∗∗ 0 It is easy to give the following matrix for the striction curves belonging to Bertrand Frenet ruled surfaces ∗∗ ∗∗ c∗∗ 1 (s) = c3 (s) = α (s) k1∗∗ ∗∗ V ∗∗ (s) c∗∗ 2 (s) = α (s) + ∗∗2 k1 + k2∗∗2 2 Theorem 2.5 The equations of the striction curves on Bertrandian Frenet ruled surfaces in terms of Frenet apparatus of curve α  c∗∗ 1 −α   0     c∗∗ − α  =  0    2  c∗∗ 3 −α 0 λ λ+ m(λ2 +β 2 )k2 (m2 +1) λ 0     V1   0  V .  2  V3 0 40 Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan Proof Since the equations of the striction curves on Bertrandian Frenet ruled surfaces in terms of Frenet apparatus of curve α are ∗∗ c∗∗ 3 (s) = α (s) = α (s) + λV2 (s) c∗∗ 1 (s) = the first and the second curvatures of the curve α∗∗ are given by k1∗∗ = k2∗∗ = (λ2 1 1 . Also k2 k2∗∗ = 2 and 2 + β ) k2 (λ + β 2 ) βk1 − λk2 and (λ2 + β 2 ) k2   2 2 ∗∗ λ + β k k 2 1 ∗∗  V2 . c∗∗ V ∗∗ (s) = α + λ + 2 (s) = α (s) + ∗∗2 k1 + k2∗∗2 2 (βk1 − λk2 )2 + 1 Theorem 2.6 The tangent vector fields T1∗∗ , T2∗∗ and to Bertrandian Frenet ruled surface are given by  T1∗∗   1 0     T ∗∗  =  a∗∗  2   T3∗∗ 1 b∗∗ 0  0 2 T3∗∗ of striction curves belonging V1∗∗     ∗∗  c∗∗    V2  V3∗∗ 0 where a∗∗ = k2∗∗ 2 ′ ∗∗ η c∗∗ 2 (s) , b∗∗ = ′ k1∗∗ ∗∗ η ′ c∗∗ 2 (s) , c∗∗ = k1∗∗ k2∗∗ ′ c∗∗ 2 (s) η ∗∗ and η ∗∗ = k1∗∗ 2 + k2∗∗ 2 . Theorem 2.7 The product of tangent vector fields T1 ∗ , T2 ∗ , T3 ∗ and tangent vector fields T1 ∗∗ , T2 ∗∗ , T3 ∗∗ of striction curves on an involute and Bertrandian Frenet ruled surface respectively, are given by  0  T [T ∗ ] [T ∗∗ ] = A   B 0 Ab∗∗ a∗∗ B + b∗∗ a∗ A + c∗∗ C b∗∗ A 0   B   0 where the coefficients are A= q (λ2 + β 2 )(k1 2 + k2 2 ) , B = b∗ (−βk1 + λk2 ) + c∗ , C = b∗ + c∗ (−λk2 + βk1 ). Proof Let [T ∗ ] = [A∗ ] [V ∗ ] and [T ∗∗ ] = [A∗∗ ] [V ∗∗ ] be given. By using the properties of a matrix following result can be obtained: 41 Tangent Vector Fields Of Striction Curves T [T ∗ ] [T ∗∗ ] T = [A∗ ] [V ∗ ] ([A∗∗ ] [V ∗∗ ]) T T = [A∗ ] [V ∗ ] [V ∗∗ ] [A∗∗ ]  1 0 0  V1 ∗   1 0    ∗∗ b∗∗ V2 ∗    a 1 0 V3 ∗ 0 T   V1 ∗∗ V1 ∗ 0    ∗∗    V2 ∗  b∗   V2   V3 ∗∗ V3 ∗ 0  0 b∗∗ A 0   ∗∗ ∗∗ ∗ ∗∗  = A  B a B + b a A + c C B . 0 b∗∗ A 0  ∗ =   a 1  1  ∗ =   a 1  b∗   c∗   0  0   c∗   0 0  V1 ∗∗ T   ∗∗   c∗∗    V2  V3 ∗∗ 0 T 1 0 0  a∗∗ b∗∗ c∗∗   1 0 0 2 As a result of Theorem 2.1 we can write that in the Euclidean 3 − space, the position of the unit tangent vector field T1∗ , T2∗ , T3∗ and T1∗∗ , T2∗∗ , T3∗∗ of striction curves belonging to ruled ∗∗ ∗∗ ∗ ∗∗ surfaces ϕ∗1 , ϕ∗2 , ϕ∗3 and ϕ∗∗ 1 , ϕ2 , ϕ3 respectively, along the curve α and α , can be expressed by the following equations  hT1∗ , T1∗∗ i hT1∗ , T2∗∗ i hT1∗ , T3∗∗ i    T ∗ ∗∗ ∗ ∗∗ ∗ ∗∗  [T ∗ ] [T ∗∗ ] =   hT2 , T1 i hT2 , T2 i hT2 , T3 i  , hT3∗ , T1∗∗ i hT3∗ , T2∗∗ i hT3∗ , T3∗∗ i T here [T ∗∗ ] is the transpose matrix of [T ∗∗ ] . Hence we may write that, there are four tangent vector fields on striction curves which are perpendicular to each other, for the involute and Bertrandian Frenet ruled surfaces given above. Since hT1∗ , T1∗∗ i = hT1∗ , T3∗∗ i = hT3∗ , T1∗∗ i = hT3∗ , T3∗∗ i = 0, it is trivial. Theorem 2.8 (i) The tangent vector fields of striction curves on an involute tangent and Bertrandian normal ruled surfaces are perpendicular under the condition " (βk1 − λk2 )(λ2 + β 2 )k2 (βk1 − λk2 )2 + 1 #′ = 0, λ2 = −β 2 or k1 2 = −k2 2 . (ii) The tangent vector fields of striction curves on an involute binormal and Bertrandian normal ruled surfaces are perpendicular under the condition " (βk1 − λk2 )(λ2 + β 2 )k2 (βk1 − λk2 )2 + 1 #′ = 0, λ2 = −β 2 or k1 2 = −k2 2 . 42 Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan Proof (i) Since hT1∗ , T2∗∗ i = b∗∗ A and hT1∗ , T2∗∗ i = 0 " " 2 2 (βk1 − λk2 )(λ + β )k2 (βk1 − λk2 )2 + 1 (βk1 − λk2 )(λ2 + β 2 )k2 (βk1 − λk2 )2 + 1 #′ b∗∗ A = 0 (λ2 + β 2 )(k1 2 + k2 2 ) = 0 #′ q = 0 or q (λ2 + β 2 )(k1 2 + k2 2 ) = 0, this completes the proof. 2 (ii) Since hT1∗ , T2∗∗ i = hT3∗ , T2∗∗ i = b∗∗ A, the proof is trivial. Theorem 2.9 (i) The tangent vector fields of striction curves on an involute normal and Bertrandian tangent ruled surfaces are perpendicular under the condition 3 −βk1 + λk2 = k22 ( kk12 )′ (k12 + k22 ) 2 i h 2 (k12 + k22 )3 + k24 ( kk12 )′ 5 λk1 (k12 +k22 ) 2 k (k12 +k22 )3 +k24 ( k1 )′ 2 2 !′ . (ii) The tangent vector fields of striction curves on an involute normal and Bertrandian binormal ruled surfaces are perpendicular under the condition 3 −βk1 + λk2 = k22 ( kk12 )′ (k12 + k22 ) 2 h i 2 (k12 + k22 )3 + k24 ( kk12 )′ 5 λk1 (k12 +k22 ) 2 k (k12 +k22 )3 +k24 ( k1 )′ 2 2 !′ . Proof (i) Since hT2∗ , T1∗∗ i = B = b∗ (−βk1 + λk2 ) + c∗ and hT2∗ , T1∗∗ i = 0 B = b∗ (−βk1 + λk2 ) + c∗ βk1 − λk2 + k22 ( kk12 )′ (k12 + h i 2 (k12 + k22 )3 + k24 ( kk12 )′ 5 λk1 (k12 +k22 ) 2 k (k12 +k22 )3 +k24 ( k1 )′ −βk1 + λk2 = this completes the proof. = 0 3 k22 ) 2 2 3 k22 ( kk12 )′ (k12 + k22 ) 2 h i 2 (k12 + k22 )3 + k24 ( kk12 )′ 2 = 0 !′ 5 λk1 (k12 +k22 ) 2 2 k (k12 +k22 )3 +k24 ( k1 )′ 2 !′ , (ii) Since hT2∗ , T1∗∗ i = hT2∗ , T3∗∗ i = B = b∗ (−βk1 + λk2 ) + c∗ , the proof is trivial. 2 Corollary 2.1 The inner product between tangent vector fields of striction curves on an involute Tangent Vector Fields Of Striction Curves 43 normal and Bertrandian normal ruled surfaces of the (α∗ , α∗∗ ) is hT2∗ , T2∗∗ i = a∗∗ B + b∗∗ a∗ A + c∗∗ C. References [1] Alias L.J., Ferrandez A., Lucas P. and Merono M. A., On the Gauss map of B-scrolls, Tsukuba J. Math., 22, 371-377, 1998. [2] Do Carmo, M. P., Differential Geometry of Curves and Surfaces, Prentice-Hall, ISBN 013-212589-7, 1976. [3] Graves L.K., Codimension one isometric immersions between Lorenzt spaces, Trans. Amer. Math. Soc., 252, 367-392, 1979. [4] Hacısalihoğlu H.H., Diferensiyel Geometri, Cilt 1, Ínönü Üniversitesi Yayinlari, Malatya, 1994. [5] Kılıçoğlu Ş, Hacısalihoğlu H.H. and Şenyurt S., On the fundamental forms of the B-scroll with null directrix and Cartan frame in Minkowskian 3-space, Applied Mathematical Sciences, doi.org/10.12988/ams.2015.53230, 9(80), 3957 - 3965, 2015. [6] Kılıçoğlu Ş, Şenyurt, S. and Hacısalihoğlu H.H., On the striction curves of Involute and Bertrandian Frenet ruled surfaces in E3 , Applied Mathematical Sciences, 9(142), 7081 7094, 2015, http://dx.doi.org/10.12988/ams.2015.59606. [7] Kılıçoğlu Ş, Şenyurt, S. and Çalışkan, A., On the striction curves of Involutive Frenet ruled surfaces in E3 , Konuralp Journal of Mathematics, 4(2), 282-289, 2016. [8] Lipschutz M.M., Differential Geometry, Schaum’s Outlines. [9] Şenyurt, S. and Çalışkan, A., A new approach on the striction curves along Bertrandian Darboux Frenet ruled surface, —it AIP Conference Proceedings, 1726, 020032 (2016), doi: 10.1063/1.4945858. [10] Şenyurt, S. and Kılıçoğlu Ş., On the differential geometric elements of the involute D̃ scroll, Adv. Appl. Cliff ord Algebras, 2015 Springer Basel,doi:10.1007/s00006-015-0535-z. [11] Şenyurt, S., On involute B-scroll a New View, University of Ordu, Journal of Science and Technology, 4(1), 59-74, 2014. [12] Springerlink, Encyclopaedia of Mathematics, Springer-Verlag, Berlin Heidelberg, New York, 2002. International J.Math. Combin. Vol.2(2018), 44-66 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 B. Basavanagoud and Chitra E. (Department of Mathematics, Karnatak University, Dharwad - 580 003, Karnataka, India) E-mail: b.basavanagoud@gmail.com, chitrachandra79.cecc@gmail.com Abstract: For a graph G, the first, second and third leap Zagreb indices are the sum of squares of 2-distance degree of vertices of G; the sum of product of 2-distance degree of end vertices of edges in G and the sum of product of 1-distance degree and 2-distance degrees of vertices of G, respectively. In this paper, we obtain the expressions for these three leap Zagreb indices of generalized xyz point line transformation graphs T xyz (G) when z = 1. Key Words: Distance, degree, diameter, Zagreb index, leap Zagreb index, reformulated Zagreb index. AMS(2010): 05C90, 05C35, 05C12, 05C07. §1. Introduction Let G = (V, E) be a simple graph of order n and size m. The k-distance degree of a vertex v ∈ V (G), denoted by dk (v/G) = |Nk (v/G)| where Nk (v/G) = {u ∈ V (G) : d(u, v) = k} [17] in which d(u, v) is the distance between the vertices u and v in G that is the length of the shortest path joining u and v in G. The degree of a vertex v in a graph G is the number of edges incident to it in G and is denoted by dG (v). Here N1 (v/G) is nothing but NG (v) and d1 (v/G) is same as dG (v). If u and v are two adjacent vertices of G, then the edge connecting them will be denoted by uv. The degree of an edge e = uv in G, denoted by d1 (e/G) (or dG (e)), is defined by d1 (e/G) = d1 (u/G) + d1 (v/G) − 2. The complement of a graph G is denoted by G whose vertex set is V (G) and two vertices of − m edges. G are adjacent if and only if they are nonadjacent in G. G has n vertices and n(n−1) 2 The line graph L(G) of a graph G with vertex set as the edge set of G and two vertices of L(G) are adjacent whenever the corresponding edges in G have a vertex incident in common. The complement of line graph L(G) or jump graph J(G) of a graph G is a graph with vertex set as the edge set of G and two vertices of J(G) are adjacent whenever the corresponding edges in G have no vertex incident in common. The subdivision graph S(G) of a graph G whose vertex set S is V (G) E(G) where two vertices are adjacent if and only if one is a vertex of G and other is 1 Supported by the University Grants Commission (UGC), New Delhi, through UGC-SAP DRS-III for 2016-2021: F.510/3/DRS-III/2016(SAP-I) and the DST INSPIRE Fellowship 2017: No.DST/INSPIRE Fellowship/[IF170465]. 2 Received February 16, 2017, Accepted May 18, 2018. On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 45 an edge of G incident with it. The partial complement of subdivision graph S(G) of a graph G S whose vertex set is V (G) E(G) where two vertices are adjacent if and only if one is a vertex of G and the other is an edge of G non incident with it. We follow [11] and [13] for unexplained graph theoretic terminologies and notations. The first and second Zagreb indices [9] of a graph G are defined as follows: dG (v)2 P M1 (G) = and M2 (G) = v∈V (G) P dG (u)dG (v), uv∈E(G) respectively. These are widely studied degree based topological indices due to their applications in chemistry. For details see the papers [5, 7, 8, 10, 18]. The first Zagreb index [15] can also be expressed as P M1 (G) = [dG (u) + dG (v)] uv∈E(G) Ashrafi et al. [1] defined the first and second Zagreb coindices as M1 (G) = P [dG (u) + dG (v)] P and M2 (G) = uv6∈E(G) [dG (u)dG (v)], uv6∈E(G) respectively. In 2004, Milićević et al. [14] reformulated the Zagreb indices in terms of edge-degrees instead of vertex-degrees. The first and second reformulated Zagreb indices are defined, respectively, as EM1 (G) = P dG (e)2 and EM2 (G) = P [dG (e)dG (f )] e∼f e∈E(G) In [12], Hosamani and Trinajstić defined the first and second reformulated Zagreb coindices respectively as EM1 (G) = X [dG (e) + dG (f )], e∼f EM2 (G) = X [dG (e) + dG (f )]. e≁f In 2017, Naji et al. [16] introduced the leap Zagreb indices. For a graph G, the first, second, and third leap Zagreb indices [16] are denoted and defined respectively as: LM1 (G) = X d2 (v/G)2 , v∈V (G) LM2 (G) = X d2 (u/G)d2 (v/G), uv∈E(G) LM3 (G) = X d1 (v/G)d2 (v/G). v∈V (G) Throughout this paper, in our results we write the notations d1 (v) and d1 (e) respectively for degree of a vertex v and degree of an edge e of a graph. 46 B. Basavanagoud and Chitra E. §2. Generalized xyz-Point-Line Transformation Graph T xyz (G) The procedure of obtaining a new graph from a given graph by using incidence (or nonincidence) relation between vertex and an edge and an adjacency (or nonadjacency) relation between two vertices or two edges of a graph is known as graph transformation and the graph obtained by doing so is called a transformation graph. For a graph G = (V, E), let G0 be the graph with V (G0 ) = V (G) and with no edges, G1 the complete graph with V (G1 ) = V (G), G+ = G, and G− = G. Let G denotes the set of simple graphs. The following graph operations depending on x, y, z ∈ {0, 1, +, −} induce functions T xyz : G → G. These operations are introduced by Deng et al. in [6]. They called these resulting graphs as xyz-transformations of G, denoted by T xyz (G) = Gxyz and studied the Laplacian characteristic polynomials and some other Laplacian parameters of xyz-transformations of an r-regular graph G. In [2], Wu Bayoindureng et al. introduced the total transformation graphs and studied the basic properties of total transformation graphs. Motivated by this, Basavanagoud [3] studied the basic properties of the xyz-transformation graphs by calling them xyz-point-line transformation graphs by changing the notion of xyztransformations of a graph G as T xyz (G) to avoid confusion between parent graph G and its xyz-transformations. Definition 2.1([6]) Given a graph G with vertex set V (G) and edge set E(G) and three variables x, y, z ∈ {0, 1, +, −}, the xyz-point-line transformation graph T xyz (G) of G is the graph with vertex set V (T xyz (G)) = V (G) ∪ E(G) and the edge set E(T xyz (G)) = E((G)x ) ∪ E((L(G))y ) ∪ E(W ) where W = S(G) if z = +, W = S(G) if z = −, W is the graph with V (W ) = V (G) ∪ E(G) and with no edges if z = 0 and W is the complete bipartite graph with parts V (G) and E(G) if z = 1. Since there are 64 distinct 3 - permutations of {0, 1, +, −}. Thus obtained 64 kinds of generalized xyz-point-line transformation graphs. There are 16 different graphs for each case when z = 0, z = 1, z = +, z = −. In this paper, we consider the xyz-point-line transformation graphs T xyz (G) when z = 1. Example 2.1 Let G = K2 · K3 be a graph. Then G0 be the graph with V (G0 ) = V (G) and with no edges, G1 the complete graph with V (G1 ) = V (G), G+ = G, and G− = G which are depicted in the following Figure 1. Figure 1 T The self-explanatory examples of the path P4 and its xyz-point-line transformation graphs (P4 ) are depicted in Figure 2. xy1 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 Figure 2 47 48 B. Basavanagoud and Chitra E. §3. Leap Zagreb Indices of T xy1 (G) Theorem 3.1([3]) Let G be a graph of order n and size m. Then (1) |V (T xyz (G))| = n + m; (2) |E(T xyz (G))| = |E(Gx )| + |E(L(G)y )| + |E(W )|, where   0 if x = 0.      n if x = 1. 2 |E(Gx )|=   m if x = +.     n − m if x = −.  2  0 if y = 0.      m if y = 1. 2 |E(L(G)y )|= 1  −m + M1  if y = +.  2   m+1 1  − 2 M1 if y = −.  2  0 if z = 0.      mn if z = 1. |E(W )|=   m if z = +.     m(n − 2) if z = −. The following Propositions are useful for calculating d2 (T xy1 (G)) in Observation 3.4. Proposition 3.2([4]) Let G be a graph of order n and size m. Let v be a vertex of G. Then   m      n+m−1 dT xy1 (G) (v) =   m + dG (v)     n + m − 1 − dG (v) if x = 0, y ∈ {0, 1, +, −} if x = 1, y ∈ {0, 1, +, −} if x = +, y ∈ {0, 1, +, −} if x = −, y ∈ {0, 1, +, −} Proposition 3.3([4]) Let G be a graph of order n and size m. Let e be an edge of G. Then   n      n+m−1 dT xy1 (G) (e) =   n + dG (e)     n + m − 1 − dG (e) if y = 0, x ∈ {0, 1, +, −} if y = 1, x ∈ {0, 1, +, −} if y = +, x ∈ {0, 1, +, −} if y = −, x ∈ {0, 1, +, −} Observation 3.4 Let G be a connected (n, m) graph. Then   (n − 1) if v ∈ V (G) (1) d2 (v/T 001 )(G)=  (m − 1) if v = e ∈ E(G) On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1   0 if v ∈ V (G) (2) d2 (v/T 101 )(G)=  (m − 1) if v = e ∈ E(G)   n − 1 − d (v/G) if v ∈ V (G) 1 (3) d2 (v/T +01 )(G)=  (m − 1) if v = e ∈ E(G)   d (v/G) if v ∈ V (G) 1 (4) d2 (v/T −01 )(G)=  (m − 1) if v = e ∈ E(G)   n − 1 if v ∈ V (G) (5) d2 (v/T 011 )(G)=  0 if v = e ∈ E(G)   0 if v ∈ V (G) (6) d2 (v/T 111 )(G)=  0 if v = e ∈ E(G)   n − 1 − d (v/G) if v ∈ V (G) 1 (7) d2 (v/T +11 )(G)=  0 if v = e ∈ E(G)   d (v/G) if v ∈ V (G) 1 (8) d2 (v/T −11 )(G)=  0 if v = e ∈ E(G)   n−1 if v ∈ V (G) (9) d2 (v/T 0+1 )(G)=  m − 1 − d1 (e/G) if v = e ∈ E(G)   0 if v ∈ V (G) (10) d2 (v/T 1+1 )(G)=  m − 1 − d1 (e/G) if v = e ∈ E(G)   n − 1 − d (v/G) if v ∈ V (G) 1 (11) d2 (v/T ++1 )(G)=  m − 1 − d1 (e/G) if v = e ∈ E(G)   d (v/G) if v ∈ V (G) 1 (12) d2 (v/T −+1 )(G)=  m − 1 − d1 (e/G) if v = e ∈ E(G)   n−1 if v ∈ V (G) (13) d2 (v/T 0−1 )(G)=  d1 (e/G) if v = e ∈ E(G)   0 if v ∈ V (G) (14) d2 (v/T 1−1 )(G)=  d1 (e/G) if v = e ∈ E(G)   n − 1 − d (v/G) if v ∈ V (G) 1 (15) d2 (v/T +−1 )(G)=  d1 (e/G) if v = e ∈ E(G)   d (v/G) if v ∈ V (G) 1 (16) d2 (v/T +−1 )(G)=  d1 (e/G) if v = e ∈ E(G) 49 The above Observation 3.4 is useful for computing leap Zagreb indices of transformation graphs T xy1 (G) in the forthcoming theorems. Theorem 3.5 Let G be (n, m) graph. Then 50 B. Basavanagoud and Chitra E. (1) LM1 (T 001 (G)) = n(n − 1)2 + m(m − 1)2 ; (2) LM2 (T 001 (G)) = mn(m − 1)(n − 1); (3) LM3 (T 001 (G)) = mn(m + n − 2). Proof The graph T 001 (G) has n + m vertices and mn edges, refer Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4 we get the following. LM1 (T 001 (G)) X = d2 (v/T 001 (G))2 v∈V (T 001 (G)) X = d2 (v/T 001 (G))2 + X uv∈E(T 001 (G)) X = uv∈E(S(G)) + X 2 2 n(n − 1) + m(m − 1) . = = d2 (e/T 001 (G))2 e∈E(G) v∈V (G) LM2 (T 001 (G)) X d2 (u/T 001 (G)) d2 (v/T 001 (G)) d2 (u/T 001 (G)) d2 (v/T 001 (G)) uv∈E(S(G)) d2 (u/T 001 (G)) d2 (v/T 001 (G)) = (n − 1)(m − 1)2m + (n − 1)(m − 1)(mn − 2m) = mn(n − 1)(m − 1). LM3 (T 001 (G)) X = v∈V (T 001 (G)) = X v∈V (G) + d1 (v/T 001 (G)) d2 (v/T 001 (G)) d1 (v/T 001 (G)) d2 (v/T 001 (G)) X e∈E(G) d1 (e/T 001 (G)) d2 (e/T 001 (G)) = mn(n − 1) + mn(m − 1) = mn(m + n − 2). 2 Theorem 3.6 Let G be (n, m) graph. Then (1) LM1 (T 101 (G)) = m(m − 1)2 ; (2) LM2 (T 101 (G)) = 0; (3) LM3 (T 101 (G)) = mn(m − 1). Proof Notice that the graph T 101 (G) has n + m vertices and mn + n(n−1) edges by 2 Theorem 3.1. According to the definitions of first, second and third leap Zagreb indices along 51 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 with Propositions 3.2, 3.3 and Observation 3.4, calculation shows the following. LM1 (T 101 (G)) X = d2 (v/T 101 (G))2 v∈V (T 101 (G)) X = d2 (v/T 101 (G))2 + v∈V (G) = LM2 (T 101 (G)) m(m − 1) . X = X = uv∈E(G) X + X + X uv∈E(S(G)) X v∈V (T 101 (G)) = d2 (u/T 101 (G)) d2 (v/T 101 (G)) d2 (u/T 101(G)) d2 (v/T 101 (G)) uv∈E(S(G)) + d2 (u/T 101 (G)) d2 (v/T 101 (G)) uv ∈E(G) / = d2 (e/T 101 (G))2 e∈E(G) 2 uv∈E(T 101 (G)) LM3 (T 101 (G)) X d2 (u/T 101(G)) d2 (v/T 101 (G)) d2 (u/T 101(G)) d2 (v/T 101 (G)) = 0. d1 (v/T 101 (G)) d2 (v/T 101 (G)) X d1 (v/T 101 (G)) d2 (v/T 101 (G)) v∈V (G) + X d1 (e/T 101 (G)) d2 (e/T 101 (G)) = mn(m − 1). 2 e∈E(G) Theorem 3.7 Let G be (n, m) graph. Then (1) LM1 (T +01 (G)) = n(n − 1)2 + m(m − 1)2 + M1 (G) − 4m(n − 1); (2) LM2 (T +01 (G)) = M2 (G) − (n − 1)M1 (G) + m[(n − 1)2 + (m − 1)(n2 − n − 2m)]; (3) LM3 (T +01 (G)) = m[n(n + m) − 2(m + 1)] − M1 (G). Proof By Theorem 3.1, we know that the graph T +01 (G) has n + m vertices and m(n + 1) edges. By using the definitions of first, second and third leap Zagreb indices and applying 52 B. Basavanagoud and Chitra E. Propositions 3.2, 3.3 and Observation 3.4 we get the following. LM1 (T +01 (G)) X = d2 (v/T +01 (G))2 v∈V (T +01 (G)) X = d2 (v/T +01 (G))2 + = 2 v∈V (G) (n − 1 − d1 (v/G)) + X = v∈V (G) X e∈E(G) (m − 1)2 X (m − 1)2 (n − 1)2 + d1 (v/G)2 − 2(n − 1)d1 (v/G) + e∈E(G) n(n − 1)2 + m(m − 1)2 + M1 (G) − 4m(n − 1). = LM2 (T +01 (G)) d2 (e/T +01 (G))2 e∈E(G) v∈V (G) X X X = uv∈E(T +01 (G)) X = uv∈E(G) + d2 (u/T +01 (G)) d2 (v/T +01 (G)) d2 (u/T +01 (G)) d2 (v/T +01 (G)) X uv∈E(S(G)) X + d2 (u/T +01 (G)) d2 (v/T +01 (G)) [d2 (u/T +01(G))][d2 (v/T +01 (G))] uv∈E(S(G)) X = uv∈E(G) + (n − 1)2 − (n − 1)(d1 (u/G) + d1 (v/G)) + d1 (u/G) · d1 (v/G) X (m − 1)(n − 1 − d1 (u/(G)) uv∈E(S(G)) X + (m − 1)(n − 1 − d1 (u/(G)) uv∈E(S(G)) = M2 (G) − (n − 1)M1 (G) + m[(n − 1)2 + (m − 1)(n2 − n − 2m)]. LM3 (T +01 (G)) X = v∈V (T +01 (G)) = d1 (v/T +01 (G)) d2 (v/T +01 (G)) X d1 (v/T +01 (G)) d2 (v/T +01 (G)) v∈V (G) + X d1 (e/T +01(G)) d2 (e/T +01 (G)) e∈E(G) = X v∈V (G) = [(m + d1 (v/G))(n − 1 − d1 (v/G))] + m[n(n + m) − 2(m + 1)] − M1 (G). Theorem 3.8 Let G be (n, m) graph. Then (1) LM1 (T −01 (G)) = M1 (G) + m(m − 1)2 ; X e∈E(G) n(m − 1) 2 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 53 (2) LM2 (T −01 (G)) = M2 (G) + 2m2 (m − 1); (3) LM3 (T −01 (G)) = m[n(m + 1) + 2(m − 1)] − M1 (G). Proof We know the graph T −01 (G) has n + m vertices and (n − 1)( n2 + m) edges, refer Theorem 3.1. By definitions of the first, second and third leap Zagreb indices and applying Propositions 3.2, 3.3 and Observation 3.4 we have the following. LM1 (T −01 (G)) X = v∈V X = d2 (v/T −01 (G))2 (T −01 (G)) d2 (v/T −01 (G))2 + LM2 (T −01 (G)) 2 M1 (G) + m(m − 1) . X = uv∈E(T −01 (G)) X = uv ∈E(G) / + X X + uv∈E(S(G)) X d2 (u/T −01 (G)) d2 (v/T −01 (G)) d2 (u/T −01(G)) d2 (v/T −01 (G)) uv∈E(S(G)) = d2 (u/T −01(G)) d2 (v/T −01 (G)) d2 (u/T −01(G)) d2 (v/T −01 (G)) [d1 (u/G)] [d1 (v/G)] + + X uv∈E(S(G)) LM3 (T −01 (G)) X v∈V (T −01 (G)) X v∈V (G) + (m − 1)d1 (u/G) d1 (v/T −01 (G)) d2 (v/T −01 (G)) d1 (v/T −01 (G)) d2 (v/T −01 (G)) X d1 (e/T −01(G)) d2 (e/T −01 (G)) e∈E(G) = (m − 1)d1 (u/G) M2 (G) + 2m2 (m − 1). = = X uv∈E(S(G)) uv ∈E(G) / = d2 (e/T −01 (G))2 e∈E(G) v∈V (G) = X m[n(m + 1) + 2(m − 1)] − M1 (G). 2 Theorem 3.9 Let G be (n, m) graph. Then (1) LM1 (T 011 (G)) = n(n − 1)2 ; (2) LM2 (T 011 (G)) = 0; (3) LM3 (T 011 (G)) = mn(n − 1). Proof We are easily know that the graph T 011 (G) has n + m vertices and m( m−1 2 + n) 54 B. Basavanagoud and Chitra E. edges by Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4 we know the following. LM1 (T 011 (G)) X = d2 (v/T 011 (G))2 v∈V (T 011 (G)) X = d2 (v/T 011 (G))2 + LM2 (T 011 (G)) 2 n(n − 1) . X = uv∈E(T 011 (G)) X = uv∈E(L(G)) X + X + uv∈E(S(G)) X + uv∈E(S(G)) X = v∈V (T 011 (G)) = X v∈V (G) + d2 (u/T 011 (G)) d2 (v/T 011 (G)) d2 (u/T 011 (G)) d2 (v/T 011 (G)) uv ∈E(L(G)) / LM3 (T 011 (G)) d2 (e/T 011 (G))2 e∈E(G) v∈V (G) = X d2 (u/T 011 (G)) d2 (v/T 011 (G)) d2 (u/T 011(G)) d2 (v/T 011 (G)) d2 (u/T 011(G)) d2 (v/T 011 (G)) = 0. d1 (v/T 011 (G)) d2 (v/T 011 (G)) d1 (v/T 011 (G)) d2 (v/T 011 (G)) X e∈E(G) d1 (e/T 011 (G)) d2 (e/T 011 (G)) = mn(n − 1). 2 Theorem 3.10 Let G be (n, m) graph. Then LM1 (T 111 (G)) = LM2 (T 111 (G)) = LM3 (T 111 (G)) = 0. Proof Notice that the graph T 111 (G) has n + m vertices and n(n−1) + m(m−1) + mn edges 2 2 by Theorem 3.1. By definitions of the first, second and third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4, we get similarly the desired result as the proof of 2 above theorems. Theorem 3.11 Let G be (n, m) graph. Then (1) LM1 (T +11 (G)) = (n − 1)(n2 − n − 4m) + M1 (G); (2) LM2 (T +11 (G)) = m(n − 1)2 − (n − 1)M1 (G) + M2 (G); (3) LM3 (T +11 (G)) = m[(n − 1)(n + 2) − 2m] − M1 (G). Proof Clearly, the graph T +11 (G) has n + m vertices and m(m+1) 2 + mn edges by Theorem On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 55 3.1. By definitions of the first, second and the third leap Zagreb indices, we get the following by applying Propositions 3.2, 3.3 and Observation 3.4. LM1 (T +11 (G)) X = d2 (v/T +11 (G))2 v∈V (T +11 (G)) X = d2 (v/T +11 (G))2 + v∈V (G) X = X d2 (e/T +11(G))2 e∈E(G) [(n − 1)2 + d1 (v/G)2 − 2(n − 1)d1 (v/G)] v∈V (G) = LM2 (T +11 (G)) (n − 1)(n2 − n − 4m) + M1 (G). X = uv∈E(T +11 (G)) X = uv∈E(G) + d2 (u/T +11 (G)) d2 (v/T +11 (G)) X uv∈E(L(G)) X + uv ∈E(L(G)) / X + uv∈E(S(G)) X + uv∈E(S(G)) = LM3 (T +11 (G)) v∈V (T +11 (G)) = X v∈V (G) + d2 (u/T +11 (G)) d2 (v/T +11 (G)) d2 (u/T +11 (G)) d2 (v/T +11 (G)) d2 (u/T +11(G)) d2 (v/T +11 (G)) d2 (u/T +11(G)) d2 (v/T +11 (G)) m(n − 1)2 − (n − 1)M1 (G) + M2 (G). X = d2 (u/T +11 (G)) d2 (v/T +11 (G)) d1 (v/T +11 (G)) d2 (v/T +11 (G)) d1 (v/T +11 (G)) d2 (v/T +11 (G)) X e∈E(G) d1 (e/T +11 (G)) d2 (e/T +11 (G)) = m[(n − 1)(n + 2) − 2m] − M1 (G). 2 Theorem 3.12 Let G be (n, m) graph. Then (1) LM1 (T −11 (G)) = M1 (G); (2) LM2 (T −11 (G)) = M2 (G); (3) LM3 (T −11 (G)) = 2m(n + m − 1) − M1 (G). + m(m−3) + mn edges, Proof Obviously, the graph T −11 (G) has n + m vertices and n(n−1) 2 2 refer Theorem 3.1. Similarly, by definitions of the first, second and the third leap Zagreb indices 56 B. Basavanagoud and Chitra E. along with Propositions 3.2, 3.3 and Observation 3.4 we know the following. LM1 (T −11 (G)) X = d2 (v/T −11 (G))2 v∈V (T −11 (G)) X = d2 (v/T −11 (G))2 + LM2 (T −11 (G)) M1 (G). X = uv∈E(T −11 (G)) X = uv ∈E(G) / + X X uv ∈E(L(G)) / X + d2 (u/T −11(G)) d2 (v/T −11 (G)) d2 (u/T −11(G)) d2 (v/T −11 (G)) uv∈E(L(G)) + d2 (e/T −11 (G))2 e∈E(G) v∈V (G) = X d2 (u/T −11(G)) d2 (v/T −11 (G)) d2 (u/T −11(G)) d2 (v/T −11 (G)) [d2 (u/T −11(G))][d2 (v/T −11 (G))] uv∈E(S(G)) X + [d2 (u/T −11(G))][d2 (v/T −11 (G))] uv∈E(S(G)) = M2 (G). LM3 (T −11 (G)) X = v∈V (T −11 (G)) = X v∈V (G) + d1 (v/T −11 (G)) d2 (v/T −11 (G)) d1 (v/T −11 (G)) d2 (v/T −11 (G)) X d1 (e/T −11(G)) d2 (e/T −11 (G)) e∈E(G) = 2m(n + m − 1) − M1 (G). 2 Theorem 3.13 Let G be (n, m) graph. Then (1) LM1 (T 0+1 (G)) = n(n − 1)2 + m(m − 1)(m + 3) − 2(m − 1)M1 (G) + EM1 (G); 2 (2) LM2 (T 0+1 (G)) = [ (m−1) − n(n − 1)]M1 (G) − (m − 1)EM1 (G) + EM2 (G) 2 + m(m − 1)[n(n − 1) − (m − 1)] + 2mn(n − 1); (3) LM3 (T 0+1 (G)) = (m + n − 1)M1 (G) − EM1 (G) + m[n(n + m) − 2(m − 1)]. Proof Notice that the graph T 0+1 (G) has n + m vertices and m(n − 1) + M12(G) edges by Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices we get the On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 57 following by applying Propositions 3.2, 3.3 and Observation 3.4. LM1 (T 0+1 (G)) X = d2 (v/T 0+1 (G))2 v∈V (T 0+1 (G)) X = d2 (v/T 0+1 (G))2 + = 2 (n − 1) + v∈V (G) = LM2 (T 0+1 (G)) 2 X uv∈E(T 0+1 (G)) X uv∈E(L(G)) X + X + uv∈E(S(G)) X uv∈E(L(G)) X + X + d2 (u/T 0+1(G)) d2 (v/T 0+1 (G)) d2 (u/T 0+1(G)) d2 (v/T 0+1 (G)) d2 (u/T 0+1(G)) d2 (v/T 0+1 (G)) (m − 1)2 − (m − 1)(d1 (u/G) + d1 (v/G)) + d1 (u/G) · d1 (v/G) uv∈E(S(G)) uv∈E(S(G)) = e∈E(G) [(m − 1)2 + d1 (e/G)2 − 2(m − 1)d1 (e/G)] d2 (u/T 0+1(G)) d2 (v/T 0+1 (G)) uv∈E(S(G)) = X n(n − 1) + m(m − 1)(m + 3) − 2(m − 1)M1 (G) + EM1 (G). = = d2 (e/T 0+1(G))2 e∈E(G) v∈V (G) X X [(n − 1)(m − 1 − d1 (v/G))] [(n − 1)(m − 1 − d1 (v/G))] (m − 1)2 [ − n(n − 1)]M1 (G) − (m − 1)EM1 (G) + EM2 (G) 2 +m(m − 1)[n(n − 1) − (m − 1)] + 2mn(n − 1). LM3 (T 0+1 (G)) X = v∈V (T 0+1 (G)) = d1 (v/T 0+1 (G)) d2 (v/T 0+1 (G)) X d1 (v/T 0+1 (G)) d2 (v/T 0+1 (G)) v∈V (G) + X d1 (e/T 0+1 (G)) d2 (e/T 0+1 (G)) e∈E(G) = X v∈V (G) = [m(n − 1)] + X e∈E(G) [(n + d1 (e/G))(m − 1 − d1 (e/G))] (m + n − 1)M1 (G) − EM1 (G) + m[n(n + m) − 2(m − 1)]. Theorem 3.14 Let G be (n, m) graph. Then (1) LM1 (T 1+1 (G)) = m(m − 1)(m + 3) − 2(m − 1)M1 (G) + EM1 (G); (2) LM2 (T 1+1 (G)) = (m−1)2 M1 (G) 2 − (m − 1)EM1 (G) + EM2 (G) − m(m − 1)2 ; 2 58 B. Basavanagoud and Chitra E. (3) LM3 (T 1+1 (G)) = (m − n − 1)M1 (G) − EM1 (G) + m[n(m + 1) − 2(m − 1)]. Proof Clearly, the graph T 1+1 (G) has n + m vertices and (n − 1)( n2 + m) + M12(G) edges by Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices we therefore get the following by Propositions 3.2, 3.3 and Observation 3.4. LM1 (T 1+1 (G)) X = d2 (v/T 1+1 (G))2 v∈V (T 1+1 (G)) X = d2 (v/T 1+1 (G))2 + X d2 (e/T 1+1 (G))2 e∈E(G) v∈V (G) X (m − 1)2 + d1 (e/G)2 − 2(m − 1)d1 (e/G) = e∈E(G) = m(m − 1)(m + 3) − 2(m − 1)M1 (G) + EM1 (G). LM2 (T 1+1 (G)) X = uv∈E(T 1+1 (G)) X = uv∈E(G) + X d2 (u/T 1+1 (G)) d2 (v/T 1+1 (G)) uv ∈E(G) / X + d2 (u/T 1+1(G)) d2 (v/T 1+1 (G)) uv∈E(L(G)) X + uv∈E(S(G)) X + uv∈E(S(G)) = X uv∈E(L(G)) = d2 (u/T 1+1(G)) d2 (v/T 1+1 (G)) d2 (u/T 1+1 (G)) d2 (v/T 1+1 (G)) d2 (u/T 1+1(G)) d2 (v/T 1+1 (G)) d2 (u/T 1+1(G)) d2 (v/T 1+1 (G)) (m − 1)2 − (m − 1)(d1 (u/G) + d1 (v/G)) + d1 (u/G) · d1 (v/G) (m − 1)2 M1 (G) − (m − 1)EM1 (G) + EM2 (G) − m(m − 1)2 . 2 LM3 (T 1+1 (G)) X = v∈V (T 1+1 (G)) = X v∈V (G) + d1 (v/T 1+1 (G)) d2 (v/T 1+1 (G)) d1 (v/T 1+1 (G)) d2 (v/T 1+1 (G)) X e∈E(G) = d1 (e/T 1+1 (G)) d2 (e/T 1+1 (G)) (m − n − 1)M1 (G) − EM1 (G) + m[n(m + 1) − 2(m − 1)]. Theorem 3.15 Let G be (n, m) graph. Then (1) LM1 (T ++1 (G)) = (n − 1)[n(n − 1) − 4m] + m(m − 1)(m + 3) 2 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 59 − (2m − 3)M1 (G) + EM1 (G); (2) LM2 (T ++1 (G)) = h (m−1)2 2 − (n − 1)(n + 1) M1 (G) + M2 (G) − (m − 1)EM1 (G) +EM2 (G) + m[n(2n − 3) − m(3m − 4) + mn(n − 1)] P + d2 (u/G)d2 (v/G) u∈V (G),v∈E(G),u∼v P d2 (u/G)d2 (v/G); + u∈V (G),v∈E(G),u≁v (3) LM3 (T ++1 (G)) = mn(m + n − 2) + (m − n − 2)M1 (G) − EM1 (G). Proof Clearly, the graph T ++1 (G) has n+m vertices and mn+ M12(G) edges by Theorem 3.1. Now by definitions of the first, second and the third leap Zagreb indices, applying Propositions 3.2, 3.3 and Observation 3.4 we have the following. LM1 (T ++1 (G)) X = d2 (v/T ++1 (G))2 v∈V (T ++1 (G)) X = d2 (v/T ++1 (G))2 + = 2 v∈V (G) d2 (e/T ++1 (G))2 e∈E(G) v∈V (G) X X [n − 1 − d1 (v/G)] + X [m − 1 − d1 (e/G)]2 e∈E(G) = (n − 1)[n(n − 1) − 4m] + m(m − 1)(m + 3) − (2m − 3)M1 (G) +EM1 (G). LM2 (T ++1 (G)) X = uv∈E(T ++1 (G)) X = uv∈E(G) + X X uv∈E(S(G)) + X uv∈E(S(G)) = d2 (u/T ++1 (G)) d2 (v/T ++1 (G)) d2 (u/T ++1 (G)) d2 (v/T ++1 (G)) uv∈E(L(G)) + d2 (u/T ++1 (G)) d2 (v/T ++1 (G)) d2 (u/T ++1(G)) d2 (v/T ++1 (G)) d2 (u/T ++1(G)) d2 (v/T ++1 (G)) (m − 1)2 − (n − 1)(n + 1) M1 (G) + M2 (G) − (m − 1)EM1 (G) 2 +EM2 (G) + m[n(2n − 3) − m(3m − 4) + mn(n − 1)] X d2 (u/G)d2 (v/G) + u∈V (G),v∈E(G),u∼v + X u∈V (G),v∈E(G),u≁v d2 (u/G)d2 (v/G). 60 B. Basavanagoud and Chitra E. LM3 (T ++1 (G)) X = v∈V (T ++1 (G)) X = v∈V (G) d1 (v/T ++1 (G)) d2 (v/T ++1 (G)) X + d1 (v/T ++1 (G)) d2 (v/T ++1 (G)) e∈E(G) d1 (e/T ++1 (G)) d2 (e/T ++1 (G)) mn(m + n − 2) + (m − n − 2)M1 (G) − EM1 (G). = 2 Theorem 3.16 Let G be (n, m) graph. Then (1) LM1 (T −+1 (G)) = m(m − 1)(m + 3) − (2m − 3)M1 (G) + EM1 (G); 2 (2) LM2 (T −+1 (G)) ="(m−1) M1 (G) + M2 (G) − (m − 1)EM1 (G) + EM2 (G) + m(m − 1)(m + # 1) 2 P − P d2 (u/G)d2 (v/G) + u∈V (G),v∈E(G),u∼v d2 (u/G)d2 (v/G) ; u∈V (G),v∈E(G),u≁v (3) LM3 (T −+1 (G)) = (m − n − 2)M1 (G) − EM1 (G) + mn(m + 3). Proof Notice that the graph T −+1 (G) has n + m vertices and n(n−1) + m(n − 2) + M12(G) 2 edges, refer Theorem 3.1. We are easily get the following by definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4. LM1 (T −+1 (G)) X = d2 (v/T −+1 (G))2 v∈V (T −+1 (G)) X = LM2 (T −+1 (G)) m(m − 1)(m + 3) − (2m − 3)M1 (G) + EM1 (G). X = uv∈E(T −+1 (G)) X = uv∈E(G) / + uv∈E(L(G)) + X uv∈E(S(G)) + X uv∈E(S(G)) = d2 (u/T −+1 (G)) d2 (u/T −+1 (G)) X d2 (e/T −+1 (G))2 e∈E(G) v∈V (G) = X d2 (v/T −+1 (G))2 + d2 (v/T −+1 (G)) d2 (v/T −+1 (G)) d2 (u/T −+1 (G)) d2 (u/T −+1 (G)) d2 (u/T −+1 (G)) d2 (v/T −+1 (G)) d2 (v/T −+1 (G)) d2 (v/T −+1 (G)) (m − 1)2 M1 (G) + M2 (G) − (m − 1)EM1 (G) 2 +EM2 (G) + m(m − 1)(m + 1)  X − d2 (u/G)d2 (v/G) + u∈V (G),v∈E(G),u∼v X u∈V (G),v∈E(G),u≁v  d2 (u/G)d2 (v/G) . On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 LM3 (T −+1 (G)) X = d1 (v/T −+1 (G)) d2 (v/T −+1 (G)) v∈V (T −+1 (G)) = 61 X d1 (v/T −+1 (G)) d2 (v/T −+1 (G)) v∈V (G) X d1 (e/T −+1(G)) d2 (e/T −+1 (G)) + e∈E(G) = (m − n − 2)M1 (G) − EM1 (G) + mn(m + 3). 2 Theorem 3.17 Let G be (n, m) graph. Then (1) LM1 (T 0−1 (G)) = n(n − 1)2 + EM1 (G); (2) LM2 (T 0−1 (G)) = EM2 (G) + n(n − 1)M1 (G) − 2mn(n − 1); (3) LM3 (T 0−1 (G)) = (n + m − 1)M1 (G) − EM1 (G) + m(n2 − 3n − 2m + 2). M1 (G) Proof Notice that the graph T 0−1 (G) has n + m vertices and m( m+1 edges, 2 + n) − 2 refer Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4 we get the following. LM1 (T 0−1 (G)) X = d2 (v/T 0−1 (G))2 v∈V (T 0−1 (G)) = X d2 (v/T 0−1 (G))2 + LM2 (T 0−1 (G)) 2 n(n − 1) + EM1 (G). X = uv∈E(T 0−1 (G)) X = uv ∈E(L(G)) / + X X uv∈E(S(G)) X = uv ∈E(L(G)) / + X d2 (u/T 0−1 (G)) d2 (v/T 0−1 (G)) d2 (u/T 0−1(G)) d2 (v/T 0−1 (G)) d2 (u/T 0−1(G)) d2 (v/T 0−1 (G)) [d1 (u/G) · d1 (v/G)] + uv∈E(S(G)) = d2 (u/T 0−1 (G)) d2 (v/T 0−1 (G)) uv∈E(S(G)) + d2 (e/T 0−1 (G))2 e∈E(G) v∈V (G) = X X uv∈E(S(G)) (n − 1)d1 (v/G) EM2 (G) + n(n − 1)M1 (G) − 2mn(n − 1). (n − 1)d1 (v/G) 62 B. Basavanagoud and Chitra E. LM3 (T 0−1 (G)) X = v∈V (T 0−1 (G)) = d1 (v/T 0−1 (G)) d2 (v/T 0−1 (G)) X d1 (v/T 0−1 (G)) d2 (v/T 0−1 (G)) v∈V (G) + X d1 (e/T 0−1 (G)) d2 (e/T 0−1 (G)) e∈E(G) = (n + m − 1)M1 (G) − EM1 (G) + m(n2 − 3n − 2m + 2). 2 Theorem 3.18 Let G be (n, m) graph. Then (1) LM1 (T 1−1 (G)) = EM1 (G); (2) LM2 (T 1−1 (G)) = EM2 (G); (3) LM3 (T 1−1 (G)) = (n + m − 1)M1 (G) − EM1 (G) − 2m(n + m − 1). M1 (G) Proof Clearly, the graph T 1−1 (G) has n + m vertices and n(n−1) + m( m+1 2 2 + n) − 2 edges by Theorem 3.1. Whence, by definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4 we get the following. LM1 (T 1−1 (G)) = v∈V = X d2 (v/T 1−1 (G))2 (T 1−1 (G)) X d2 (v/T 1−1 (G))2 + LM2 (T 1−1 (G)) EM1 (G). X = uv∈E(T 1−1 (G)) X = uv∈E(G) + X X X uv∈E(S(G)) + d2 (u/T 1−1(G)) d2 (v/T 1−1 (G)) d2 (u/T 1−1 (G)) d2 (v/T 1−1 (G)) uv ∈E(L(G)) / + d2 (u/T 1−1(G)) d2 (v/T 1−1 (G)) uv ∈E(G) / + d2 (e/T 1−1 (G))2 e∈E(G) v∈V (G) = X X uv∈E(S(G)) = EM2 (G). d2 (u/T 1−1(G)) d2 (v/T 1−1 (G)) d2 (u/T 1−1 (G)) d2 (v/T 1−1 (G)) d2 (u/T 1−1 (G)) d2 (v/T 1−1 (G)) On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 LM3 (T 1−1 (G)) X = v∈V (T 1−1 (G)) 63 d1 (v/T 1−1 (G)) d2 (v/T 1−1 (G)) X d1 (v/T 1−1 (G)) d2 (v/T 1−1 (G)) = v∈V (G) X d1 (e/T 1−1(G)) d2 (e/T 1−1 (G)) + e∈E(G) (n + m − 1)M1 (G) − EM1 (G) − 2m(n + m − 1). = 2 Theorem 3.19 Let G be (n, m) graph. Then (1) LM1 (T +−1 (G)) = M1 (G) + EM1 (G) + (n − 1)[n(n − 1) − 4m]; (2) LM2 (T +−1 (G)) ="(n − 1)2 M1 (G) + M2 (G) + EM2 (G) − m(n − 1)(n + 1) P − P d2 (u/G)d2 (v/G) + u∈V (G),v∈E(G),u∼v # d2 (u/G)d2 (v/G) ; u∈V (G),v∈E(G),u≁v (3) LM3 (T +−1 (G)) = (n + m − 2)M1 (G) − EM1 (G) + m[(n − 1)(n + 2) − 2(2m + n − 1)]. M1 (G) Proof Clearly, the graph T +−1 (G) has n + m vertices and m( m+3 edges by 2 + n) − 2 Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices along with Propositions 3.2, 3.3 and Observation 3.4 we therefore get the following. LM1 (T +−1 (G)) = v∈V X X = d2 (v/T +−1 (G))2 (T +−1 (G)) LM2 (T +−1 (G)) M1 (G) + EM1 (G) + (n − 1)[n(n − 1) − 4m]. X = uv∈E(T +−1 (G)) X = uv∈E(G) + uv ∈E(L(G)) / + X uv∈E(S(G)) + X uv∈E(S(G)) = d2 (u/T +−1 (G)) d2 (u/T +−1 (G)) X d2 (e/T +−1 (G))2 e∈E(G) v∈V (G) = X d2 (v/T +−1 (G))2 + d2 (v/T +−1 (G)) d2 (v/T +−1 (G)) d2 (u/T +−1 (G)) d2 (u/T +−1 (G)) d2 (u/T +−1 (G)) d2 (v/T +−1 (G)) d2 (v/T +−1 (G)) d2 (v/T +−1 (G)) (n − 1)2 M1 (G) + M2 (G) + EM2 (G) − m(n − 1)(n + 1)  X X − d2 (u/G)d2 (v/G) + u∈V (G),v∈E(G),u∼v u∈V (G),v∈E(G),u≁v  d2 (u/G)d2 (v/G) . 64 B. Basavanagoud and Chitra E. LM3 (T +−1 (G)) X = v∈V (T +−1 (G)) X = v∈V (G) + d1 (v/T +−1 (G)) d2 (v/T +−1 (G)) d1 (v/T +−1 (G)) d2 (v/T +−1 (G)) X e∈E(G) d1 (e/T +−1 (G)) d2 (e/T +−1 (G)) = (n + m − 2)M1 (G) − EM1 (G) + m[(n − 1)(n + 2) − 2(2m + n − 1)]. 2 Theorem 3.20 Let G be (n, m) graph. Then (1) LM1 (T −−1 (G)) = M1 (G) + EM1 (G); P (2) LM2 (T −−1 (G)) = M1 (G) + EM2 (G) + d2 (u/G)d2 (v/G) u∈V (G),v∈E(G),u∼v P d2 (u/G)d2 (v/G); + u∈V (G),v∈E(G),u≁v (3) LM3 (T −−1 (G)) = (n + m − 2)M1 (G) − EM1 (G). M1 (G) Proof Notice that the graph T −−1 (G) has n+ m vertices and n(n−1) + m( m−1 2 2 + n)− 2 edges by Theorem 3.1. By definitions of the first, second and the third leap Zagreb indices, Propositions 3.2, 3.3 and Observation 3.4, we are easily get the following. LM1 (T −−1 (G)) = v∈V = X d2 (v/T −−1 (G))2 (T −−1 (G)) X d2 (v/T −−1 (G))2 + = M1 (G) + EM1 (G). X = uv∈E(T −−1 (G)) X = uv ∈E(G) / + X X uv∈E(S(G)) + d2 (u/T −−1 (G)) d2 (v/T −−1 (G)) d2 (u/T −−1(G)) d2 (v/T −−1 (G)) uv ∈E(L(G)) / + d2 (e/T −−1(G))2 e∈E(G) v∈V (G) LM2 (T −−1 (G)) X X uv∈E(S(G)) d2 (u/T −−1(G)) d2 (v/T −−1 (G)) d2 (u/T −−1 (G)) d2 (v/T −−1 (G)) d2 (u/T −−1 (G)) d2 (v/T −−1 (G)) = M1 (G) + EM2 (G) + X u∈V (G),v∈E(G),u∼v + X u∈V (G),v∈E(G),u≁v d2 (u/G)d2 (v/G). d2 (u/G)d2 (v/G) On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 LM3 (T −−1 (G)) X = v∈V (T −−1 (G)) = 65 d1 (v/T −−1 (G)) d2 (v/T −−1 (G)) X d1 (v/T −−1 (G)) d2 (v/T −−1 (G)) v∈V (G) + X d1 (e/T −−1(G)) d2 (e/T −−1(G)) e∈E(G) = (n + m − 2)M1 (G) − EM1 (G). 2 References [1] A. R. Ashrafi, T. Došlić, A. Hamzeh, The Zagreb coindices of graph operations, Discrete Appl. Math., 158(15), 1571–1578, 2010. [2] W. Baoyindureng, M. Jixiang, Basic properties of total transformation graphs, J. Math. Study, 34(2), 109–116, 2001. [3] B. Basavanagoud, Basic properties of generalized xyz-Point-Line transformation graphs, J. Inf. Optim. Sci., 39(2), 561–580, 2018, DOI: 10.1080/02522667.2017.1395147. [4] B. Basavanagoud, C. S. Gali, Computing first and second Zagreb indices of generalized xyz-Point-Line transformation graphs, J. Global Research Math. Arch., 5(4), 100–122, 2018. [5] C. M. Da fonseca, D. Stevanović, Further properties of the second Zagreb index, MATCH Commun. Math. Comput. Chem., 72, 655–668, 2014. [6] A. Deng, A. Kelmans, J. Meng, Laplacian Spectra of regular graph transformations, Discrete Appl. Math., 161, 118–133, 2013. [7] B. Furtula, I. Gutman, M. Dehmer, On structure-sensitivity of degree-based topological indices, Appl. Math. Comput., 219, 8973–8978, 2013. [8] M. Goubko, T. Réti, Note on minimizing degree-based topological indices of trees with given number of pendent vertices, MATCH Commun. Math. Comput. Chem., 72, 633– 639, 2014. [9] I. Gutman, N. Trinajstić, Graph theory and molecular orbitals, Total π-electron energy of alternate hydrocarbons, Chem. Phys. Lett., 17, 535–538, 1972. [10] I. Gutman, K. C. Das, The first Zagreb index 30 years after, MATCH Commun. Math. Comput. Chem., 50, 83–92, 2004. [11] F. Harary, Graph Theory, Addison-Wesely, Reading Mass, 1969. [12] S. M. Hosamani, N. Trinajstić, On reformulated Zagreb coindices, Research Gate, 2015-0508 T 09:07:00 UTC. [13] V. R. Kulli, College Graph Theory, Vishwa International Publications, Gulbarga, India, 2012. [14] A. Milićević, S. Nikolić, N. Trinajstić, On reformulated Zagreb indices, Mol. Divers., 8(4), 393–399, 2004. [15] S. Nikolić, G. Kovačević, A. Milićević, N. Trinajstić, The Zagreb indices 30 years after, Croat. Chem. Acta., 76(2), 113–124, 2003. 66 B. Basavanagoud and Chitra E. [16] A. M. Naji, N. D. Soner, Ivan Gutman, On Leap Zagreb indices of graphs, Commun. Comb. Optim., 2(2), 99–117, 2017. [17] N. D. Soner, A. M. Naji, The k-distance neighbourhood polynomial of a graph, Int. J. Math. Comput. Sci. WASET Conference Proceedings, San Francico, USA, Sep 26-27, 3(9) Part XV, 2359–2364, 2016. [18] G. Su, L. Xiong, L. Xu, The Nordhaus-Gaddum-type inequalities for the Zagreb index and coindex of graphs, Appl. Math. Lett., 25(11), 1701–1707, 2012. International J.Math. Combin. Vol.2(2018), 67-79 A Generalization on Product Degree Distance of Strong Product of Graphs K.Pattabiraman (Department of Mathematics, Annamalai University, Annamalainagar 608 002, India) E-mail: pramank@gmail.com Abstract: In this paper, the exact formulae for the generalized product degree distance, reciprocal product degree distance and product degree distance of strong product of a connected graph and the complete multipartite graph with partite sets of sizes m0 , m1 , · · · , mr−1 are obtained. Key Words: Reciprocal product degree distance, product degree distance, strong product. AMS(2010): 05C12, 05C76 §1. Introduction All the graphs considered in this paper are simple and connected. For vertices u, v ∈ V (G), the distance between u and v in G, denoted by dG (u, v), is the length of a shortest (u, v)-path in G and let dG (v) be the degree of a vertex v ∈ V (G). The strong product of graphs G and H, denoted by G ⊠ H, is the graph with vertex set V (G) × V (H) = {(u, v) : u ∈ V (G), v ∈ V (H)} and (u, x)(v, y) is an edge whenever (i) u = v and xy ∈ E(H), or (ii) uv ∈ E(G) and x = y, or (iii) uv ∈ E(G) and xy ∈ E(H). A topological index of a graph is a real number related to the graph; it does not depend on labeling or pictorial representation of a graph. In theoretical chemistry, molecular structure descriptors (also called topological indices) are used for modeling physicochemical, pharmacologic, toxicologic, biological and other properties of chemical compounds [12]. There exist several types of such indices, especially those based on vertex and edge distances. One of the most intensively studied topological indices is the Wiener index. Let G be a connected graph. Then W iener index of G is defined as W (G) = 1 2 X dG (u, v) u, v ∈ V (G) with the summation going over all pairs of distinct vertices of G. This definition can be further 1 Received September 9, 2017, Accepted May 20, 2018. 68 K.Pattabiraman generalized in the following way: Wλ (G) = 1 2 X dλG (u, v), u, v ∈ V (G) where dλG (u, v) = (dG (u, v))λ and λ is a real number [13, 14]. If λ = −1, then W−1 (G) = H(G), where H(G) is Harary index of G. In the chemical literature also W 12 [29] as well as the general case Wλ were examined [10, 15]. Dobrynin and Kochetova [6] and Gutman [11] independently proposed a vertex-degreeweighted version of Wiener index called degree distance, which is defined for a connected graph G as 1 X (dG (u) + dG (v))dG (u, v), DD(G) = 2 u,v∈V (G) where dG (u) is the degree of the vertex u in G. Similarly, the product degree distance or Gutman index of a connected graph G is defined as DD∗ (G) = 1 2 X dG (u)dG (v)dG (u, v). u,v∈V (G) The additively weighted Harary index(HA ) or reciprocal degree distance(RDD) is defined in [3] as 1 X (dG (u) + dG (v)) HA (G) = RDD(G) = . 2 dG (u, v) u,v∈V (G) Similarly, Su et al. [28] introduce the reciprocal product degree distance of graphs, which can be seen as a product-degree-weight version of Harary index RDD∗ (G) = 1 2 X u,v∈V (G) dG (u)dG (v) . dG (u, v) In [16], Hamzeh et al. recently introduced generalized degree distance of graphs. Hua and Zhang [18] have obtained lower and upper bounds for the reciprocal degree distance of graph in terms of other graph invariants. Pattabiraman et al. [22, 23] have obtained the reciprocal degree distance of join, tensor product, strong product and wreath product of two connected graphs in terms of other graph invariants. The chemical applications and mathematical properties of the reciprocal degree distance are well studied in [3, 20, 27]. The generalized degree distance, denoted by Hλ (G), is defined as Hλ (G) = 1 2 X (dG (u) + dG (v))dλG (u, v), u,v∈V (G) where λ is a real number. If λ = 1, then Hλ (G) = DD(G) and if λ = −1, then Hλ (G) = A Generalization on Product Degree Distance of Strong Product of Graphs 69 RDD(G). Similarly, generalized product degree distance, denoted by Hλ∗ (G), is defined as Hλ∗ (G) = 1 2 X dG (u)dG (v)dλG (u, v). u,v∈V (G) If λ = 1, then Hλ∗ (G) = DD∗ (G) and if λ = −1, then Hλ∗ (G) = RDD∗ (G). Therefore the study of the above topological indices are important and we try to obtain the results related to these indices. The generalized degree distance of unicyclic and bicyclic graphs are studied by Hamzeh et al. [16, 17]. Also they are given the generalized degree distance of Cartesian product, join, symmetric difference, composition and disjunction of two graphs. The generalized degree distance and generalized product degree distance of some classes of graphs are obtained in [24, 25, 26]. In this paper, the exact formulae for the generalized product degree distance, reciprocal product degree distance and product degree distance of strong product G ⊠ Km0 , m1 , ··· , mr−1 , where Km0 , m1 , ··· , mr−1 is the complete multipartite graph with partite sets of sizes m0 , m1 , · · · , mr−1 are obtained. The first Zagreb index is defined as M1 (G) = X dG (u)2 u∈V (G) and the second Zagreb index is defined as M2 (G) = X dG (u)dG (v). uv∈E(G) In fact, one can rewrite the first Zagreb index as M1 (G) = X (dG (u) + dG (v)). uv∈E(G) The Zagreb indices were found to be successful in chemical and physico-chemical applications, especially in QSPR/QSAR studies, see [8, 9]. For S ⊆ V (G), hSi denotes the subgraph of G induced by S. For two subsets S, T ⊂ V (G), not necessarily disjoint, by dG (S, T ), we mean the sum of the distances in G from each vertex P dG (s, t). of S to every vertex of T, that is, dG (S, T ) = s ∈ S, t ∈ T §2. Generalized Product Degree Distance of Strong Product of Graphs In this section, we obtain the Generalized product degree distance of G ⊠ Km0 , m1 , ··· , mr−1 . Let G be a simple connected graph with V (G) = {v0 , v1 , · · · , vn−1 } and let Km0 , m1 , ··· , mr−1 , r ≥ 2, be the complete multiparite graph with partite sets V0 , V1 , · · · , Vr−1 and let |Vi | = mi , 0 ≤ i ≤ r − 1. In the graph G ⊠ Km0 , m1 , ··· , mr−1 , let Bij = vi × Vj , vi ∈ V (G) and 0 ≤ j ≤ r − 1. 70 K.Pattabiraman For our convenience, the vertex set of G ⊠ Km0 , m1 , ··· , mr−1 is written as V (G) × V (Km0 , m1 , ..., mr−1 ) = Let B = {Bij }i = 0,1,··· , n−1 . Let Xi = j = 0,1,··· , r−1 r−1 S r−1 n−1 [ Bij . i=0 j =0 Bij and Yj = n−1 S Bij ; we call Xi and Yj as layer i=0 j =0 and column of G ⊠ Km0 , m1 , ..., mr−1 , respectively. If we denote V (Bij ) = {xi1 , xi2 , · · · , ximj } and V (Bkp ) = {xk1 , xk2 , · · · , xk mp }, then xiℓ and xkℓ , 1 ≤ ℓ ≤ j, are called the corresponding S vertices of Bij and Bkp . Further, if vi vk ∈ E(G), then the induced subgraph hBij Bkp i of G ⊠ Km0 , m1 , ··· , mr−1 is isomorphic to K|Vj ||Vp | or, mp independent edges joining the corresponding vertices of Bij and Bkj according as j 6= p or j = p, respectively. The following remark is follows from the structure of the graph Km0 , m1 , ··· , mr−1 . Remark 2.1 Let n0 and q be the number of vertices and edges of Km0 , m1 , ··· , mr−1 . Then the sums r−1 X mj mp = 2q, j, p = 0 j 6= p r−1 X m2j = n20 − 2q, j=0 r−1 X j, p = 0 j 6= p m2j mp = n0 q − 3t r−1 X = r−1 X j=0 mj m2p , j, p = 0 j 6= p m3j = n30 − 3n0 q + 3t j=0 and r−1 X m4j = n40 − 4n20 q + 2q 2 + 4n0 t − 4τ, ′ where t and τ are the number of triangles and K4s in Km0 , m1 , ··· , mr−1 . The proof of the following lemma follows easily from the properties and structure of G ⊠ Km0 , m1 , ··· , mr−1 . Lemma 2.2 Let G be a connected graph and let Bij , Bkp ∈ B of the graph G′ = G ⊠ Km0 , m1 , ··· , mr−1 , where r ≥ 2. Then (i) If vi vk ∈ E(G) and xit ∈ Bij , xkℓ ∈ Bkj , then  1, if t = ℓ, dG′ (xit , xkℓ ) = 2, if t = 6 ℓ, A Generalization on Product Degree Distance of Strong Product of Graphs 71 and if xit ∈ Bij , xkℓ ∈ Bkp , j 6= p, then dG′ (xit , xkℓ ) = 1. (ii) If vi vk ∈ / E(G), then for any two vertices xit ∈ Bij , xkℓ ∈ Bkp , dG′ (xit , xkℓ ) = dG (vi , vk ). (iii) For any two distinct vertices in Bij , their distance is 2. The proof of the following lemma follows easily from Lemma 2.2, which is used in the proof of the main theorems of this section. Lemma 2.3 Let G be a connected graph and let Bij , Bkp ∈ B of the graph G′ = G ⊠ Km0 , m1 , ..., mr−1 , where r ≥ 2. (i) If vi vk ∈ E(G), then  m m , if j 6= p, j p dH (B , B ) = ′ ij kp G  mj (mj +1) , if j = p, 2 (ii) If vi vk ∈ / E(G), then dH G′ (Bij , Bkp ) =   mj mp dG (vi ,vk ) , 2  mj , dG (vi ,vk ) if j 6= p, if j = p.  m m , if j 6= p, j p (iii) dH (B , B ) = ′ ij ip G  mj (mj −1) , if j = p. 2 Lemma 2.4 Let G be a connected graph and let Bij in G′ = G ⊠ Km0 , m1 , ··· , mr−1 . Then the degree of a vertex (vi , uj ) ∈ Bij in G′ is dG′ ((vi , uj )) = dG (vi ) + (n0 − mj ) + dG (vi )(n0 − mj ), where n0 = r−1 P mj . j=0 Now we obtain the generalized product degree distance of G ⊠ Km0 , m1 , ··· , mr−1 . Theorem 2.5 Let G be a connected graph with n vertices and m edges. Then Hλ∗ (G ⊠ Km0 , m1 , ··· , mr−1 ) n = (4q 2 + n20 + 4n0 q)Hλ∗ (G) + 4q 2 Wλ (G) + (4q 2 + 2n0 q)Hλ (G) + (4q 2 − n0 q − 3t) 2 i M1 (G) h 2 4n0 q − 2q 2 + 4n0 t + 9t + 7n0 q − n0 − 3n20 − 2n30 + 8τ + 2 h i +m 3n0 q + 2n0 t − 2q 2 − 3t − 4q + 4τ h i +2λ M1 (G)(2q 2 − 2n0 t − 6t − 2q − 4τ ) + m(2q 2 − 2n0 t − n0 q − 3t − 4τ ) i +(2λ − 1)M2 (G) 2q 2 − 2n0 t − 3n30 + 10n0 q + n20 − 18t − 6q − n0 − 4τ . 72 K.Pattabiraman Proof Let G′ = G ⊠ Km0 , m1 , ..., mr−1 . Clearly, Hλ∗ (G′ ) 1 2 = 1 2 = + X dG′ (Bij )dG′ (Bkp )dλG′ (Bij , Bkp ) Bij , Bkp ∈ B n−1 r−1 X X dG′ (Bij )dG′ (Bip )dλG′ (Bij , Bip ) i = 0 j, p = 0 j 6= p n−1 X r−1 X dG′ (Bij )dG′ (Bkj )dλG′ (Bij , Bkj ) i, k = 0 j = 0 i 6= k + n−1 X r−1 X dG′ (Bij )dG′ (Bkp )dλG′ (Bij , Bkp ) i, k = 0 j, p = 0 i 6= k j 6= p + r−1 n−1 XX ! dG′ (Bij )dG′ (Bij )dλG′ (Bij , Bij ) i=0 j =0 . (2.1) We shall obtain the sums of (2.1) are separately. First we calculate A1 = n−1 P r−1 P i = 0 j, p = 0 j 6= p By Lemma 2.4, we have T1′ dG′ (Bij )dG′ (Bip )dλG′ (Bij , Bip ). For that first we find T1′ . = dG′ (Bij )dG′ (Bip ) = dG (vi )(n0 − mj + 1) + (n0 − mj ) dG (vi )(n0 − mp + 1) + (n0 − mp ) = (n0 + 1)2 − (n0 + 1)mj − (n0 + 1)mp + mj mp d2G (vi ) + 2n0 (n0 + 1) − (2n0 + 1)mj − (2n0 + 1)mp + 2mj mp dG (vi ) + n20 − n0 mp − n0 mj + mj mp . From Lemma 2.3, we have dλG′ (Bij , Bip ) = mj mp . Thus T1′ dλG′ (Bij , Bip ) = T1′ mj mp = (n0 + 1)2 mj mp − (n0 + 1)m2j mp − (n0 + 1)mj m2p + m2j m2p d2G (vi ) + 2n0 (n0 + 1)mj mp − (2n0 + 1)m2j mp − (2n0 + 1)mj m2p + 2m2j m2p dG (vi ) + n20 mj mp − n0 m2j mp − n0 mj m2p + m2j m2p . A Generalization on Product Degree Distance of Strong Product of Graphs 73 By Remark 2.1, we have T1 r−1 X = T1′ dλG′ (Bij , Bip ) j, p = 0 j 6= p = 2q 2 + 2qn0 + 2n0 t + 2q + 4τ + 6t d2G (vi ) + 2qn0 + 4n0 t − 4q 2 + 6t + 8τ dG (vi ) + 2n0 t + 2q 2 + 4τ . From the definition of the first Zagreb index, we have A1 = n−1 X T1 i=0 = Next we obtain A2 = 2q 2 + 2qn0 + 2n0 t + 2q + 4τ + 6t M1 (G) +2m 2qn0 + 4n0 t − 4q 2 + 6t + 8τ +n 2n0 t + 2q 2 + 4τ . n−1 P r−1 P i, k = 0 j = 0 i 6= k By Lemma 2.4, we have T2′ = = = dG′ (Bij )dG′ (Bkj )dλG′ (Bij , Bkj ). For that first we find T2′ . dG′ (Bij )dG′ (Bkj ) dG (vi )(n0 − mj + 1) + (n0 − mj ) dG (vk )(n0 − mj + 1) + (n0 − mj ) (n0 − mj + 1)2 dG (vi )dG (vk ) + (n0 − mj )(n0 − mj + 1)(dG (vi ) + dG (vk )) +(n0 − mj )2 . Thus A2 = r−1 n−1 X X T2′ dλG′ (Bij , Bkj ) j = 0 i, k = 0 i 6= k = r−1 X j =0 n−1 X i, k = 0 i 6= k vi vk ∈E(G) T2′ dλG′ (Bij , Bkj ) + r−1 X j =0 n−1 X i, k = 0 i 6= k vi vk ∈E(G) / T2′ dλG′ (Bij , Bkj ) 74 K.Pattabiraman By Lemma 2.3, we have A2 r−1 X = = = n−1 X j=0 i, k = 0 i 6= k vi vk ∈E(G) r−1 X n−1 X j=0 T2′ j=0 i, k = 0 i 6= k vi vk ∈E(G) r−1 X n−1 X j=0 r−1 X T2′ 1 − 2λ + 2λ mj mj + i, k = 0 i 6= k vi vk ∈E(G) n−1 X T2′ m2j dλG (vi , vk ), i, k = 0 i 6= k vi vk ∈E(G) / r−1 X 1 − 2λ + 2λ mj mj + m2j − m2j + j =0 n−1 X T2′ m2j dλG (vi , vk ) i, k = 0 i 6= k vi vk ∈E(G) / r−1 n−1 X X T2′ m2j dλG (vi , vk ) T2′ (2λ − 1) m2j − mj + j = 0 i, k = 0 i 6= k = S1 + S2 , (2.2) where S1 and S2 are the sums of the terms of the above expression, in order. Now we calculate S1 . For that first we find the following. h (2λ − 1)T2′ m2j − mj = (2λ − 1) m4j − (2n0 + 3)m3j + (n20 + 4n0 + 3)m2j −(n0 + 1)2 mj dG (vi )dG (vk ) + m4j − (2n0 + 2)m3j + (n20 + 3n0 + 1)m2j − (n20 + n0 )mj (dG (vi ) + dG (vk )) i + m4j − (2n0 + 1)m3j + (n20 + 2n0 )m2j − n2o mj . By Remark 2.1, we have T2′′ r−1 X = (2λ − 1)T2′ m2j − mj j =0 = h (2λ − 1) 2q 2 − 2n0 t − 4τ − 3n30 + 10n0 q − 18t + n20 − 6q − n0 dG (vi )dG (vk ) + 2q 2 − 4τ − 2n0 t − 6t − 2q (dG (vi ) + dG (vk )) i + 2q 2 − 4τ − 2n0 t − n0 q − 3t . Hence S1 = n−1 X T2′′ i, k = 0 i 6= k vi vk ∈E(G) = h (2λ − 1) 2q 2 − 2n0 t − 4τ − 3n30 + 10n0 q − 18t + n20 − 6q − n0 2M2 (G) + 2q 2 − 4τ − 2n0 t − 6t − 2q 2M1 (G) i +2m 2q 2 − 4τ − 2n0 t − n0 q − 3t . A Generalization on Product Degree Distance of Strong Product of Graphs 75 Next we calculate S2 . For that we need the following. T2′ m2j m4j − (2n0 + 2)m3j + (n0 + 1)2 m2j dG (vi )dG (vk ) + m4j − (2n0 + 1)m3j + (n20 + n0 )m2j (dG (vi ) + dG (vk )) + m4j − 2n0 m3j + n20 m2j . = By Remark 2.1, we have T2 = r−1 X T2′ m2j j =0 = 2q 2 − 4τ − 2n0 t − 6t + 2n0 q − 2q + n20 dG (vi )dG (vk ) + 2q 2 − 4τ − 2n0 t − 3t + n0 q (dG (vi ) + dG (vk )) + 2q 2 − 4τ − 2n0 t . From the definitions of Hλ∗ ,Hλ and Wλ , we obtain S2 n−1 X = T2 dλG (vi , vk ) i, k = 0 i 6= k = 2 2q 2 − 4τ − 2n0 t − 6t + 2n0 q − 2q + n20 Hλ∗ (G) +2 2q 2 − 4τ − 2n0 t − 3t + n0 q Hλ (G) +2 2q 2 − 4τ − 2n0 t Wλ (G). Now we calculate A3 = n−1 P r−1 P i, k = 0 j, p = 0 i 6= k j 6= p pute T3′ . By Lemma 2.4, we have T3′ dG′ (Bij )dG′ (Bkp )dλG′ (Bij , Bkp ). For that first we com- = dG′ (Bij )dG′ (Bkp ) = dG (vi )(n0 − mj + 1) + (n0 − mj ) dG (vk )(n0 − mp + 1) + (n0 − mp ) = dG (vi )dG (vk )(n0 − mj + 1)(n0 − mp + 1) + dG (vi )(n0 − mj + 1)(n0 − mp ) +dG (vk )(n0 − mp + 1)(n0 − mj ) + (n0 − mj )(n0 − mp ). 76 K.Pattabiraman Since the distance between Bij and Bkp is mj mp dλG (vi , vk ). Thus T3′ mj mp = dG (vi )dG (vk ) (n20 + 2n0 + 1)mj mp − (n0 + 1)m2j mp − (n0 + 1)mj m2p + m2j m2p +dG (vi ) (n20 + n0 )mj mp − (n0 + 1)mj m2p − n0 m2j mp + m2j m2p +dG (vk ) (n20 + n0 )mj mp − n0 mj m2p − (n0 + 1)m2j mp + m2j m2p + n20 mj mp − n0 mj m2p − n0 m2j mp + m2j m2p . By Remark 2.1, we obtain T3 = r−1 X j, p = 0, j 6= p T3′ mj mp = dG (vi )dG (vk ) 2n0 q + 2n0 t + 2q + 2q 2 + 6t + 4τ +(dG (vi ) + dG (vk )) qn0 + 2n0 t + 3t + 2q 2 + 4τ + 2n0 t + 2q 2 + 4τ . Hence A3 = n−1 X i, k = 0 i 6= k T3 dλG (vi , vk ) = 2Hλ∗ (G) 2n0 q + 2n0 t + 2q + 2q 2 + 6t + 4τ +2Hλ (G) qn0 + 2n0 t + 3t + 2q 2 + 4τ +2Wλ (G) 2n0 t + 2q 2 + 4τ . Finally, we obtain A4 = n−1 P r−1 P i=0 j =0 T4′ . By Lemma 2.4, we have T4′ dG′ (Bij )dG′ (Bij )dλG′ (Bij , Bij ). For that first we calculate = dG′ (Bij )dG′ (Bij ) 2 = dG (vi )(n0 − mj + 1) + (n0 − mj ) = d2G (vi )(n0 − mj + 1)2 + 2dG (vi )(n0 − mj )(n0 − mj + 1) + (n0 − mj )2 . From Lemma 2.3, the distance between (Bij and (Bij is mj (mj − 1). Thus T4′ mj (mj − 1) = d2G (vi ) m4j − (2n0 + 3)m3j + ((n0 + 1)2 + 2)m2j − (n0 + 1)2 mj +2dG (vi ) m4j − (2n0 + 2)m3j + (n20 + 3n0 + 1)m2j − (n20 + n0 )mj + m4j − (2n0 + 1)m3j + (n20 + 2n0 )m2j − n20 mj . A Generalization on Product Degree Distance of Strong Product of Graphs 77 By Remark 2.1, we obtain T4 = r−1 X j =0 T4′ mj (mj − 1) = d2G (vi ) 4n20 q − 2n30 − 3n20 − 2n0 t + 5n0 q − 9t − 6q − n0 − 4τ +2dG (vi ) 2q 2 − 2n0 t − 2q − 6t − 4τ + 2q 2 − 2n0 t − n0 q − 3t − 4τ . Hence A4 = n−1 X T4 dλG′ (Bij , Bij ) i=0 = M1 (G) 4n20 q − 2n30 − 3n20 − 2n0 t + 5n0 q − 9t − 6q − n0 − 4τ +4m 2q 2 − 2n0 t − 2q − 6t − 4τ +n 2q 2 − 2n0 t − n0 q − 3t − 4τ . Adding A1 ,S1 ,S2 ,A3 and A4 we get the required result. 2 If we set λ = 1 in Theorem 2.5, we obtain the product degree distance of G⊠Km0 , m1 , ··· , mr−1 . Theorem 2.6 Let G be a connected graph with n vertices and m edges. Then DD∗ (G ⊠ Km0 , m1 , ··· , mr−1 ) = (4q 2 + n20 + 4n0 q)DD∗ (G) + 4q 2 W (G) n +(4q 2 + 2n0 q)DD(G) + (4q 2 − n0 q − 3t) 2 i M1 (G) h 2 2 4n0 q + 6q − 4n0 t − 15t + 7n0 q − n0 − 3n20 − 2n30 − 8τ + h2 i +m n0 q − 2n0 t + 2q 2 − 9t − 4q − 4τ h i +M2 (G) 2q 2 − 2n0 t − 3n30 + 10n0 q + n20 − 18t − 6q − n0 − 4τ for r ≥ 2. Setting λ = −1 in Theorem 2.5, we obtain the reciprocal product degree distance of G ⊠ Km0 , m1 , ··· , mr−1 . Theorem 2.7 Let G be a connected graph with n vertices and m edges. Then RDD∗ (G ⊠ Km0 , m1 , ··· , mr−1 ) = (4q 2 + n20 + 4n0 q)RDD∗ (G) + 4q 2 H(G) n +(4q 2 + 2n0 q)RDD(G) + (4q 2 − n0 q − 3t) 2 78 K.Pattabiraman i M1 (G) h 2 4n0 q + 2n0 t + 3t + 7n0 q − n0 − 3n20 − 2n30 − 2q + 4τ 2 i h 5n q 9t 0 +m + n0 t − q 2 − − 4q + 2τ 2 2 i M2 (G) h 2 2q − 2n0 t − 3n30 + 10n0 q + n20 − 18t − 6q − n0 − 4τ − 2 + for r ≥ 2. References [1] A.R. Ashrafi, T. Doslic and A. Hamzeha, The Zagreb coindices of graph operations, Discrete Appl. Math., 158 (2010) 1571-1578. [2] N. Alon, E. Lubetzky, Independent set in tensor graph powers, J. Graph Theory, 54 (2007) 73-87. [3] Y. Alizadeh, A. Iranmanesh, T. Doslic, Additively weighted Harary index of some composite graphs, Discrete Math., 313 (2013) 26-34. [4] A.M. Assaf, Modified group divisible designs, Ars Combin., 29 (1990) 13-20. [5] B. Bresar, W. Imrich, S. Klavžar, B. Zmazek, Hypercubes as direct products, SIAM J. Discrete Math., 18 (2005) 778-786. [6] A.A. Dobrynin, A.A. Kochetova, Degree distance of a graph: a degree analogue of the Wiener index, J. Chem. Inf. Comput. Sci., 34 (1994) 1082-1086. [7] S. Chen , W. Liu , Extremal modified Schultz index of bicyclic graphs, MATCH Commun. Math. Comput. Chem., 64(2010)767-782. [8] J. Devillers, A.T. Balaban, Eds., Topological Indices and Related Descriptors in QSAR and QSPR, Gordon and Breach, Amsterdam, The Netherlands, 1999. [9] M.V. Diudea(Ed.), QSPR/QRAR Studies by Molecular Descriptors, Nova, Huntington (2001). [10] B. Furtula, I.Gutman, Z. Tomovic, A. Vesel, I. Pesek, Wiener-type topological indices of phenylenes, Indian J. Chem., 41A(2002) 1767-1772. [11] I. Gutman, Selected properties of the Schultz molecular topological index, J. Chem. Inf. Comput. Sci., 34 (1994) 1087-1089. [12] I. Gutman, O.E. Polansky, Mathematical Concepts in Organic Chemistry, Springer-Verlag, Berlin, 1986. [13] I. Gutman, A property of the Wiener number and its modifications, Indian J. Chem., 36A(1997) 128-132. [14] I. Gutman, A.A. Dobrynin, S. Klavzar. L. Pavlovic, Wiener-type invariants of trees and their relation, Bull. Inst. Combin. Appl., 40(2004)23-30. [15] I. Gutman, D.Vidovic and L. Popovic, Graph representation of organic molecules. Cayley’s plerograms vs. his kenograms, J.Chem. Soc. Faraday Trans., 94 (1998) 857-860. [16] A. Hamzeh, A. Iranmanesh, S. Hossein-Zadeh , M.V. Diudea, Generalized degree distance of trees, unicyclic and bicyclic graphs, Studia Ubb Chemia, LVII, 4(2012) 73-85. [17] A. Hamzeh, A. Iranmanesh, S. Hossein-Zadeh, Some results on generalized degree distance, Open J. Discrete Math., 3 (2013) 143-150. A Generalization on Product Degree Distance of Strong Product of Graphs 79 [18] H. Hua, S. Zhang, On the reciprocal degree distance of graphs, Discrete Appl. Math., 160 (2012) 1152-1163. [19] W. Imrich, S. Klavžar, Product graphs: Structure and Recognition, John Wiley, New York (2000). [20] S.C. Li, X. Meng, Four edge-grafting theorems on the reciprocal degree distance of graphs and their applications, J. Comb. Optim., 30 (2015) 468-488. [21] A. Mamut, E. Vumar, Vertex vulnerability parameters of Kronecker products of complete graphs, Inform. Process. Lett., 106 (2008) 258-262. [22] K. Pattabiraman, M. Vijayaragavan, Reciprocal degree distance of some graph operations, Trans. Comb., 2(2013) 13-24. [23] K. Pattabiraman, M. Vijayaragavan, Reciprocal degree distance of product graphs, Discrete Appl. Math., 179(2014) 201-213. [24] K. Pattabiraman, Generalization on product degree distance of tensor product of graphs, J. Appl. Math. & Inform., 34(2016) 341- 354. [25] K. Pattabiraman, P. Kandan, Generalized degree distance of strong product of graphs, Iran. J. Math. Sci. & Inform., 10 (2015) 87-98. [26] K. Pattabiraman, P. Kandan, Generalization of the degree distance of the tensor product of graphs, Aus. J. Comb., 62(2015) 211-227. [27] G. F. Su, L.M. Xiong, X.F. Su, X.L. Chen, Some results on the reciprocal sum-degree distance of graphs, J. Comb. Optim., 30(2015) 435-446. [28] G. Su, I. Gutman, L. Xiong, L. Xu,Reciprocal product degree distance of graphs, Manuscript. [29] H. Y. Zhu, D.J. Klenin, I. Lukovits, Extensions of the Wiener number, J. Chem. Inf. Comput. Sci., 36 (1996) 420-428. International J.Math. Combin. Vol.2(2018), 80-86 Semifull Line (Block) Signed Graphs V. Lokesha1 , P. S. Hemavathi1,2 and S. Vijay3 1. Department of Studies in Mathematics, Vijayanagara Sri Krishnadevaraya University, Ballari-583 105, India 2. Department of Mathematics, Siddaganga Institute of Technology, B.H.Road, Tumkur-572 103, India 3. Dept. of Mathematics, Government Science College, Hassan-573 201, India E-mail: hemavathisuresh@gmail.com Abstract: In this paper we introduced the new notions semifull signed graph and semifull line (block) signed graph of a signed graph and its properties are obtained. Also, we obtained the structural characterizations of these notions. Further, we presented some switching equivalent characterizations. Key Words: Signed graphs, balance, switching, semifull signed graph, semifull line (block) signed graph, negation of a signed graph, semifull Smarandachely graph. AMS(2010): 05C22. §1. Introduction For all terminology and notation in graph theory we refer the reader to consult any one of the standard text-books by Chartrand and Zhang [2], Harary [3] and West [12]. If B = {u1 , u2 , · · · , ur , r ≥ 2} is a block of a graph Γ, then we say that vertex u1 and block B are incident with each other, as are u2 and B and so on. If two blocks B1 and B2 of G are incident with a common cut vertex, then they are adjacent blocks. If B = {e1 , e2 , · · · , es , s ≥ 1} is a block of a graph Γ, then we say that an edge e1 and block B are incident with each other, as are e2 and B and so on. This concept was introduced by Kulli [7]. The vertices, edges and blocks of a graph are called its members. The line graph L(Γ) of a graph Γ is the graph whose vertex set is the set of edges of Γ in which two vertices are adjacent if the corresponding edges are adjacent (see [3]). The semifull graph SF (Γ) of a graph Γ is the graph whose vertex set is the union of vertices, edges and blocks of Γ in which two vertices are adjacent if the corresponding members of Γ are adjacent or one corresponds to a vertex and the other to an edge incident with it or one corresponds to a block B of Γ and the other to a vertex v of Γ and v is in B. In fact, this notion was introduced by Kulli [8]. Generally, for a subset B ′ ⊂ B, a semifull Smarandachely S S graph SSF (Γ) of a graph Γ on B ′ is the graph with V (SSF (Γ)) = V (Γ) E(Γ) B ′ , and two vertices are adjacent in SSF (Γ) if the corresponding members of Γ are adjacent or one corresponds to a vertex and the other to an edge incident with it or one corresponds to a block 1 Received November 17, 2017, Accepted May 22, 2018. Semifull Line (Block) Signed Graphs 81 B ′ of Γ and the other to a vertex v of Γ with v ∈ B ′ . Clearly, SSF (Γ) = SF(Γ) if B ′ = B. In [9], the author introduced the new notions called “semifull line graphs and semifull block graphs” as follows: The semifull line graph SF L(Γ) of a graph Γ is the graph whose vertex set is the union of the set of vertices, edges and blocks of Γ in which two vertices are adjacent SF L(Γ) if the corresponding vertices and edges of Γ are adjacent or one corresponds to a vertex of Γ and other to an edge incident with it or one corresponds to a block B of Γ and other to a vertex v of Γ and v is in B. The semifull block graph SF B(Γ) of a graph Γ is the graph whose vertex set is the union of the set of vertices, edges and blocks of Γ in which two vertices are adjacent in SF B(Γ) if the corresponding vertices and blocks of Γ are adjacent or one corresponds to a vertex of Γ and other to an edge incident with it or one corresponds to a block B of Γ and other to a vertex v of Γ and v is in B. A signed graph is an ordered pair Σ = (Γ, σ), where Γ = (V, E) is a graph called underlying graph of Σ and σ : E → {+, −} is a function. We say that a signed graph is connected if its underlying graph is connected. A signed graph Σ = (Γ, σ) is balanced, if every cycle in Σ has an even number of negative edges (See [4]). Equivalently, a signed graph is balanced if product of signs of the edges on every cycle of Σ is positive. Signed graphs Σ1 and Σ2 are isomorphic, written Σ1 ∼ = Σ2 , if there is an isomorphism between their underlying graphs that preserves the signs of edges. The theory of balance goes back to Heider [6] who asserted that a social system is balanced if there is no tension and that unbalanced social structures exhibit a tension resulting in a tendency to change in the direction of balance. Since this first work of Heider, the notion of balance has been extensively studied by many mathematicians and psychologists. In 1956, Cartwright and Harary [4] provided a mathematical model for balance through graphs. A marking of Σ is a function ζ : V (Γ) → {+, −}. Given a signed graph Σ one can easily define a marking ζ of Σ as follows: For any vertex v ∈ V (Σ), ζ(v) = Y σ(uv), uv∈E(Σ) the marking ζ of Σ is called canonical marking of Σ. The following are the fundamental results about balance, the second being a more advanced form of the first. Note that in a bipartition of a set, V = V1 ∪ V2 , the disjoint subsets may be empty. Theorem 1.1 A signed graph Σ is balanced if and only if either of the following equivalent conditions is satisfied: (1)(Harary [4]) Its vertex set has a bipartition V = V1 ∪ V2 such that every positive edge joins vertices in V1 or in V2 , and every negative edge joins a vertex in V1 and a vertex in V2 . (2)(Sampathkumar [10]) There exists a marking µ of its vertices such that each edge uv in Γ satisfies σ(uv) = ζ(u)ζ(v). Let Σ = (Γ, σ) be a signed graph. Complement of Σ is a signed graph Σ = (Γ, σ ′ ), where 82 V. Lokesha, P. S. Hemavathi and S. Vijay for any edge e = uv ∈ Γ, σ ′ (uv) = ζ(u)ζ(v). Clearly, Σ as defined here is a balanced signed graph due to Theorem 1.1. A switching function for Σ is a function ζ : V → {+, −}. The switched signature is ζ σ (e) := ζ(v)σ(e)ζ(w), where e has end points v, w. The switched signed graph is Σζ := (Σ|σ ζ ). We say that Σ switched by ζ. Note that Σζ = Σ−ζ (see [1]). If X ⊆ V , switching Σ by X (or simply switching X) means reversing the sign of every edge in the cutset E(X, X c ). The switched signed graph is ΣX . This is the same as Σζ where ζ(v) := − if and only if v ∈ X. Switching by ζ or X is the same operation with different c notation. Note that ΣX = ΣX . Signed graphs Σ1 and Σ2 are switching equivalent, written Σ1 ∼ Σ2 if they have the same underlying graph and there exists a switching function ζ such that Σζ1 ∼ = Σ2 . The equivalence class of Σ, [Σ] := {Σ′ : Σ′ ∼ Σ} is called the its switching class. Similarly, Σ1 and Σ2 are switching isomorphic, written Σ1 ∼ = Σ2 , if Σ1 is isomorphic to a switching of Σ2 . The equivalence class of Σ is called its switching isomorphism class. Two signed graphs Σ1 = (Γ1 , σ1 ) and Σ2 = (Γ2 , σ2 ) are said to be weakly isomorphic (see [11]) or cycle isomorphic (see [13]) if there exists an isomorphism φ : Γ1 → Γ2 such that the sign of every cycle Z in Σ1 equals to the sign of φ(Z) in Σ2 . The following result is well known (see [13]): Theorem 1.2(T. Zaslavsky, [13]) Two signed graphs Σ1 and Σ2 with the same underlying graph are switching equivalent if and only if they are cycle isomorphic. §2. Semifull Line Signed Graphs Motivated by the existing definition of complement of a signed graph, we now extend the notion called semifull line graphs to realm of signed graphs: the semifull line signed graph SF L(Σ) of a signed graph Σ = (Γ, σ) as a signed graph SF L(Σ) = (SF L(Γ), σ ′ ), where for any edge e1 e2 in SF L(Γ), σ ′ (e1 e2 ) = σ(e1 )σ(e2 ). Further, a signed graph Σ = (Γ, σ) is called semifull line signed graph, if Σ ∼ = SF L(Σ′ ) for some signed graph Σ′ . The following result indicates the limitations of the notion of semifull line signed graphs as introduced above, since the entire class of unbalanced signed graphs is forbidden to be semifull line signed graphs. Theorem 2.1 For any signed graph Σ = (Γ, σ), its semifull line signed graph SF L(Σ) is balanced. ′ Proof Let σ denote the signing of SF L(Σ) and let the signing σ of Σ be treated as a marking of the vertices of SF L(Σ). Then by definition of SF L(Σ), we see that σ ′ (e1 e2 ) = σ(e1 )σ(e2 ), for every edge e1 e2 of SF L(Σ) and hence, by Theorem 1, the result follows. 2 For any positive integer k, the k th iterated semifull line signed graph, SF Lk (Σ) of Σ is Semifull Line (Block) Signed Graphs 83 defined as follows: SF L0 (Σ) = Σ, SF Lk (Σ) = SF L(SF Lk−1 (Σ)) Corollary 2.2 For any signed graph Σ = (Γ, σ) and for any positive integer k, SF Lk (Σ) is balanced. Proposition 2.3 For any two signed graphs Σ1 and Σ2 with the same underlying graph, their semifull line signed graphs are switching equivalent. Proof Suppose Σ1 = (Γ, σ) and Σ2 = (Γ′ , σ ′ ) be two signed graphs with Γ ∼ = Γ′ . By Theorem 2.1, SF L(Σ1 ) and SF L(Σ2 ) are balanced and hence, the result follows from Theorem 1.2. 2 The semifull signed graph SF(Σ) of a signed graph Σ = (Γ, σ) as a signed graph SF(Σ) = (SF (Γ), σ ′ ), where for any edge e1 e2 in SF(Γ), σ ′ (e1 e2 ) = σ(e1 )σ(e2 ). Further, a signed graph Σ = (Γ, σ) is called semifull signed graph, if Σ ∼ = SF(Σ′ ) for some signed graph Σ′ . The following result indicates the limitations of the notion of semifull signed graphs as introduced above, since the entire class of unbalanced signed graphs is forbidden to be semifull signed graphs. Theorem 2.4 For any signed graph Σ = (Γ, σ), its semifull signed graph SF(Σ) is balanced. ′ Proof Let σ denote the signing of SF(Σ) and let the signing σ of Σ be treated as a marking of the vertices of SF (Σ). Then by definition of SF(Σ), we see that σ ′ (e1 e2 ) = σ(e1 )σ(e2 ), for 2 every edge e1 e2 of SF (Σ) and hence, by Theorem 1, the result follows. For any positive integer k, the k th iterated semifull line signed graph, SF k (Σ) of Σ is defined as follows: SF 0 (Σ) = Σ, SF k (Σ) = SF(SF k−1 (Σ)) Corollary 2.5 For any signed graph Σ = (Γ, σ) and for any positive integer k, SF k (Σ) is balanced. Proposition 2.6 For any two signed graphs Σ1 and Σ2 with the same underlying graph, their semifull signed graphs are switching equivalent. Proof Suppose Σ1 = (Γ, σ) and Σ2 = (Γ′ , σ ′ ) be two signed graphs with Γ ∼ = Γ′ . By Theorem 2.4, SF(Σ1 ) and SF(Σ2 ) are balanced and hence, the result follows from Theorem 1.2. 2 In [9], the author characterizes graphs such that semifull line graphs and semifull graphs are isomorphic. Theorem 2.7 Let Γ be a nontrivial connected graph. The graphs SF L(Γ) and SF(Γ) are isomorphic if and only if Γ is a block. 84 V. Lokesha, P. S. Hemavathi and S. Vijay In view of the above result, we have the following result that characterizes the family of signed graphs satisfies SF L(Σ) ∼ SF(Σ). Theorem 2.8 For any signed graph Σ = (Γ, σ), SF L(Σ) ∼ SF(Σ) if and only if Γ is a block. Proof Suppose that SF L(Σ) ∼ SF (Σ). Then clearly, SF L(Γ) ∼ = SF (Γ). Hence by Theorem 2.7, Γ is a block. Conversely, suppose that Σ is a signed graph whose underlying graph is a block. Then by Theorem 2.7, SF L(Γ) and SF (Γ) are isomorphic. Since for any signed graph Σ, both SF L(Σ) 2 and SF (Σ) are balanced, the result follows by Theorem 1.2. The following result characterize signed graphs which are semifull line signed graphs. Theorem 2.9 A signed graph Σ = (Γ, σ) is a semifull line signed graph if and only if Σ is balanced signed graph and its underlying graph Γ is a semifull line graph. Proof Suppose that Σ is balanced and Γ is a semifull line graph. Then there exists a graph Γ such that SF L(Γ′ ) ∼ = Γ. Since Σ is balanced, by Theorem 1.1, there exists a marking ζ of Γ such that each edge uv in Σ satisfies σ(uv) = ζ(u)ζ(v). Now consider the signed graph Σ′ = (Γ′ , σ ′ ), where for any edge e in Γ′ , σ ′ (e) is the marking of the corresponding vertex in Γ. Then clearly, SF L(Σ′ ) ∼ = Σ. Hence Σ is a semifull line signed graph. ′ Conversely, suppose that Σ = (Γ, σ) is a semifull line signed graph. Then there exists a signed graph Σ′ = (Γ′ , σ ′ ) such that SF L(Σ′ ) ∼ = Σ. Hence, Γ is the semiful line graph of Γ′ and by Theorem 2.1, Σ is balanced. 2 In view of the above result, we can easily characterize signed graphs which are semifull signed graphs. The notion of negation η(Σ) of a given signed graph Σ defined in [5] as follows: η(Σ) has the same underlying graph as that of Σ with the sign of each edge opposite to that given to it in Σ. However, this definition does not say anything about what to do with nonadjacent pairs of vertices in Σ while applying the unary operator η(.) of taking the negation of Σ. For a signed graph Σ = (Γ, σ), the SF L(Σ) (SF (Σ)) is balanced. We now examine, the conditions under which negation η(Σ) of SF L(Σ) (SF(Σ)) is balanced. Theorem 2.10 Let Σ = (Γ, σ) be a signed graph. If SF L(Γ) (SF(Γ)) is bipartite then η(SF L(Σ)) (η(SF (Σ))) is balanced. Proof Since SF L(Σ) (SF (Σ)) is balanced, if each cycle C in SF L(Σ) (SF(Σ)) contains even number of negative edges. Also, since SF L(Γ) (SF (Γ)) is bipartite, all cycles have even length; thus, the number of positive edges on any cycle C in SF L(Σ) (SF (Σ)) is also even. Hence η(SF L(Σ)) (η(SF (Σ))) is balanced. 2 Semifull Line (Block) Signed Graphs 85 §3. Semifull Block Signed Graphs Motivated by the existing definition of complement of a signed graph, we now extend the notion called semifull block graphs to realm of signed graphs: the semifull block signed graph SF B(Σ) of a signed graph Σ = (Γ, σ) as a signed graph SF B(Σ) = (SF B(Γ), σ ′ ), where for any edge e1 e2 in SF B(Γ), σ ′ (e1 e2 ) = σ(e1 )σ(e2 ). Further, a signed graph Σ = (Γ, σ) is called semifull block signed graph, if Σ ∼ = SF L(Σ′ ) for some signed graph Σ′ . The following result indicates the limitations of the notion of semifull block signed graphs as introduced above, since the entire class of unbalanced signed graphs is forbidden to be semifull block signed graphs. Theorem 3.1 For any signed graph Σ = (Γ, σ), its semifull block signed graph SF B(Σ) is balanced. ′ Proof Let σ denote the signing of SF B(Σ) and let the signing σ of Σ be treated as a marking of the vertices of SF B(Σ). Then by definition of SF B(Σ), we see that σ ′ (e1 e2 ) = 2 σ(e1 )σ(e2 ), for every edge e1 e2 of SF B(Σ) and hence, by Theorem 1, the result follows. For any positive integer k, the k th iterated semifull block signed graph, SF Bk (Σ) of Σ is defined as follows: SF B0 (Σ) = Σ, SF B k (Σ) = SF B(SF B k−1 (Σ)) Corollary 3.2 For any signed graph Σ = (Γ, σ) and for any positive integer k, SF B k (Σ) is balanced. Proposition 3.3 —it For any two signed graphs Σ1 and Σ2 with the same underlying graph, their semifull block signed graphs are switching equivalent. Proof Suppose Σ1 = (Γ, σ) and Σ2 = (Γ′ , σ ′ ) be two signed graphs with Γ ∼ = Γ′ . By Theorem 3.1, SF B(Σ1 ) and SF B(Σ2 ) are balanced and hence, the result follows from Theorem 1.2. 2 In [9], the author characterizes graphs such that semifull block graphs and semifull graphs are isomorphic. Theorem 3.4 Let Γ be a nontrivial connected graph. The graphs SF B(Γ) and SF(Γ) are isomorphic if and only if Γ is P2 . In view of the above result, we have the following result that characterizes the family of signed graphs satisfies SF B(Σ) ∼ SF(Σ). Theorem 3.5 For any signed graph Σ = (Γ, σ), SF B(Σ) ∼ SF(Σ) if and only if Γ is P2 . Proof Suppose that SF B(Σ) ∼ SF (Σ). Then clearly, SF B(Γ) ∼ = SF (Γ). Hence by Theorem 16, Γ is P2 . Conversely, suppose that Σ is a signed graph whose underlying graph is P2 . Then by Theorem 16, SF B(Γ) and SF(Γ) are isomorphic. Since for any signed graph Σ, both SF B(Σ) and SF (Σ) are balanced, the result follows by Theorem 2. 2 86 V. Lokesha, P. S. Hemavathi and S. Vijay In view of the Theorem 2.9, we can easily characterize signed graphs which are semifull block signed graphs. Acknowledgement The authors are thankful to the anonymous referee for valuable suggestions and comments for the improvement of the paper. References [1] R. P. Abelson and M. J. Rosenberg, Symoblic psychologic:A model of attitudinal cognition, Behav. Sci., 3 (1958), 1-13. [2] G.T. Chartrand and P. Zhang, An Introduction to Graph Theory, Walter Rudin Series in Advanced Mathematics, Mc- Graw Hill Companies Inc., New York (2005). [3] F. Harary, Graph Theory, Addison-Wesley Publ. Comp., Massachusetts, Reading (1969). [4] F. Harary, On the notion of balance of a signed graph, Michigan Math. J., 2 (1953), 143-146. [5] F. Harary, Structural duality, Behav. Sci., 2(4) (1957), 255-265. [6] F. Heider, Attitudes and Cognitive Organisation, Journal of Psychology, 21 (1946), 107112. [7] V. R. Kulli, The semitotal block graph and the total block graph of a graph, Indian J. Pure and Appl.Math., 7 (1976), 625-630. [8] V.R. Kulli, The semifull graph of a graph, Annals of Pure and Applied Mathematics, 10(1) (2015), 99-104. [9] V. R. Kulli, On semifull line graphs and semifull block graphs, J. Comp. & Math. Sci., 6(7) (2015), 388-394. [10] E. Sampathkumar, Point signed and line signed graphs, Nat. Acad. Sci. Letters, 7(3) (1984), 91-93. [11] T. Sozánsky, Enueration of weak isomorphism classes of signed graphs, J. Graph Theory, 4(2)(1980), 127-144. [12] D.B. West, Introduction to Graph Theory, Prentice-Hall of India Pvt. Ltd., 1999. [13] T. Zaslavsky, Signed graphs, Discrete Appl. Math., 4(1) (1982), 47-74. International J.Math. Combin. Vol.2(2018), 87-96 Accurate Independent Domination in Graphs B.Basavanagoud (Department of Mathematics, Karnatak University, Dharwad-580 003, India) Sujata Timmanaikar (Department of Mathematics, Government Engineering College, Haveri-581 110, India) E-mail: b.basavanagoud@gmail.com, sujata123rk@gmail.com Abstract: A dominating set D of a graph G = (V, E) is an independent dominating set, if the induced subgraph hDi has no edges. An independent dominating set D of G is an accurate independent dominating set if V − D has no independent dominating set of cardinality |D|. The accurate independent domination number ia (G) of G is the minimum cardinality of an accurate independent dominating set of G. In this paper, we initiate a study of this new parameter and obtain some results concerning this parameter. Key Words: Domination, independent domination number, accurate independent domination number, Smarandache H-dominating set. AMS(2010): 05C69. §1. Introduction All graphs considered here are finite, nontrivial, undirected with no loops and multiple edges. For graph theoretic terminology we refer to Harary [1]. Let G = (V, E) be a graph with |V | = p and |E| = q. Let ∆(G)(δ(G)) denote the maximum (minimum) degree and ⌈x⌉(⌊x⌋) the least (greatest) integer greater(less) than or equal to x. The neighborhood of a vertex u is the set N (u) consisting of all vertices v which are adjacent with u. The closed neighborhood is N [u] = N (u) ∪ {u}. A set of vertices in G is independent if no two of them are adjacent. The largest number of vertices in such a set is called the vertex independence number of G and is denoted by βo (G). For any set S of vertices of G, the induced subgraph hSi is maximal subgraph of G with vertex set S. The corona of two graphs G1 and G2 is the graph G = G1 ◦ G2 formed from one copy of G1 and |V (G1 )| copies of G2 where the ith vertex of G1 is adjacent to every vertex in the ith copy of G2 . A wounded spider is the graph formed by subdividing at most n − 1 of the edges of a star K1,n f or n ≥ 0. Let Ω(G) be the set of all pendant vertices of G, that is the set of vertices of degree 1. A vertex v is called a support vertex if v is neighbor of a pendant vertex and dG (v) > 1. Denote by X(G) the set of all support vertices in G, M (G) be the set 1 Supported by University Grant Commission(UGC), New Delhi, India through UGC-SAP- DRS-III, 20162021: F.510/3/DRS-III/2016 (SAP-I). 2 Received January 11, 2018, Accepted May 25, 2018. 88 B. Basavanagoud and Sujata Timmanaikar of vertices which are adjacent to support vertex and J(G) be the set of vertices which are not adjacent to a support vertex. The diameter diam(G) of a connected graph G is the maximum distance between two vertices of G, that is diam(G) = maxu,v∈V (G) dG (u, v). A set B ⊆ V is a 2-packing if for each pair of vertices u, v ∈ B, NG [u] ∩ NG [v] = φ A proper coloring of a graph G = (V (G), E(G)) is a function from the vertices of the graph to a set of colors such that any two adjacent vertices have different colors. The chromatic number χ(G) is the minimum number of colors needed in a proper coloring of a graph. A dominator coloring of a graph G is a proper coloring in which each vertex of the graph dominates every vertex of some color class. The dominator chromatic number χd (G) is the minimum number of color classes in a dominator coloring of a graph G. This concept was introduced by R. Gera at.al [3]. A set D of vertices in a graph G = (V, E) is a dominating set of G, if every vertex in V − D is adjacent to some vertex in D. The domination number γ(G) of G is the minimum cardinality of a dominating set. For a comprehensive survey of domination in graphs, see [4, 5, 7]. Generally, if hDi ≃ H, such a dominating set D is called a Smarandache H-dominating set. A dominating set D of a graph G = (V, E) is an independent dominating set, if the induced subgraph hDi has no edges, i.e., a Smarandache H-dominating set with E(H) = ∅. The independent domination number i(G) is the minimum cardinality of an independent dominating set. A dominating set D of G = (V, E) is an accurate dominating set if V −D has no dominating set of cardinality |D|. The accurate domination number γa (G) of G is the minimum cardinality of an accurate dominating set. This concept was introduced by Kulli and Kattimani [6, 9]. An independent dominating set D of G is an accurate independent dominating set if V − D has no independent dominating set of cardinality |D|. The accurate independent domination number ia (G) of G is the minimum cardinality of an accurate independent dominating set of G. This concept was introduced by Kulli [8]. For example, we consider the graph G in Figure 1. The accurate independent dominating sets are {1, 2, 6, 7} and {1, 3, 6, 7}. Therefore ia (G) = 4. 2b 1b 6 b 7 b G: b 3 b b 4 5 Figure 1 §2. Results Observation 2.1 1. Every accurate independent dominating set is independent and dominating. Hence it is a minimal dominating set. Accurate Independent Domination in Graphs 89 2. Every minimal accurate independent dominating set is a maximal independent dominating set. Proposition 2.1 For any nontrivial connected graph G, γ(G) ≤ ia (G). Proof Clearly, every accurate independent dominating set of G is a dominating set of G. 2 Thus result holds. Proposition 2.2 If G contains an isolated vertex, then every accurate dominating set is an accurate independent dominating set. Now we obtain the exact values of ia (G) for some standard class of graphs. Proposition 2.3 For graphs Pp , Wp and Km,n , there are (1) ia (Pp ) = ⌈p/3⌉ if p ≥ 3; (2) ia (Wp ) = 1 if p ≥ 5; (3) ia (Km,n ) = m f or 1 ≤ m < n. Theorem 2.1 For any graph G, ia (G) ≤ p − γ(G). Proof Let D be a minimal dominating set of G. Then there exist at least one accurate independent dominating set in (V − D) and by proposition 2.1, ia (G) ≤ |V | − |D| ≤ p − γ(G). Notice that the path P4 achieves this bound. 2 Theorem 2.2 For any graph G, ⌈p/ △ +1⌉ ≤ ia (G) ≤ ⌊p △ / △ +1⌋ and these bounds are sharp. Proof It is known that p/ △ +1 ≤ γ(G) and by proposition 2.1, we see that the lower bound holds. By Theorem 2.1, ia (G) ≤ p − γ(G), ≤ p − p/ △ +1 ≤ p △ / △ +1. Notice that the path Pp , p ≥ 3 achieves the lower bound. This completes the proof. Proposition 2.4 If G = Km1 ,m2 ,m3 ,··· ,mr , r ≥ 3, then ia (G) = m1 if m1 < m2 < m3 · · · < mr . 2 90 B. Basavanagoud and Sujata Timmanaikar Theorem 2.3 For any graph G without isolated vertices γa (G) ≤ ia (G) if G 6= Km1 ,m2 ,m3 ,··· ,mr , r ≥ 3. Furthermore, the equality holds if G = Pp (p 6= 4, p ≥ 3), Wp (p ≥ 5) or Km,n f or 1 ≤ m < n. Proof Since we have γ(G) ≤ γa (G) and by Proposition 2.1,γa (G) ≤ ia (G). Let γa (G) ≤ ia (G). If G = Km1 ,m2 ,m3 ,··· ,mr , r ≥ 3 then by Proposition 2.4, ia (G) = m1 if m1 < m2 < m3 · · · < mr and also accurate domination number is ⌊p/2⌋ + 1 i.e., γa (G) = ⌊p/2⌋ + 1 > m1 = ia (G), a contradiction. 2 Corollary 2.1 For any graph G, ia (G) = γ(G) if diam(G) = 2. Proposition 2.5 For any graph G without isolated vertices i(G) ≤ ia (G). Furthermore, the equality holds if G = Pp (p ≥ 3), Wp (p ≥ 5) or Km,n f or 1 ≤ m < n. Proof Every accurate independent dominating set is a independent dominating set. Thus 2 result holds. Definition 2.1 The double star Sn,m is the graph obtained by joining the centers of two stars K1,n and K1,m with an edge. Proposition 2.6 For any graph G, ia (G) ≤ βo (G). Furthermore, the equality holds if G = Sn,m . Proof Since every minimal accurate independent dominating set is an maximal independent dominating set. Thus result holds. 2 Theorem 2.4 For any graph G, ia (G) ≤ p − α0 (G). Proof Let S be a vertex cover of G. Then V − S is an accurate independent dominating set. Then ia (G) ≤ |V − S| ≤ p − α0 (G). 2 Corollary 2.2 Fr any graph G, ia (G) ≤ p − β0 (G) + 2. Theorem 2.5 If G is any nontrivial connected graph containing exactly one vertex of degree △ (G) = p − 1, then γ(G) = ia (G) = 1. Proof Let G be any nontrivial connected graph containing exactly one vertex v of degree deg(v) = p − 1. Let D be a minimal dominating set of G containing vertex of degree deg(v) = p = 1. Then D is a minimum dominating set of G i.e., |D| = γ(G) = 1. (1) Also V − D has no dominating set of same cardinality |D|. Therefore, |D| = ia (G). Hence, by (1) and (2) γ(G) = ia (G) = 1. (2) 2 Accurate Independent Domination in Graphs 91 Theorem 2.6 If G is a connected graph with p vertices then ia (G) = p/2 if and only if G = H ◦ K1 ,where H is any nontrivial connected graph. Proof Let D be any minimal accurate independent dominating set with |D| = p/2. If G 6= H ◦ K1 then there exist at least one vertex vi ∈ V (G) which is neither a pendant vertex nor a support vertex. Then there exist a minimal accurate independent dominating set D′ containing vi such that |D′ | ≤ |D| − {vi } ≤ p/2 − {vi } ≤ p/2 − 1, which is a contradiction to minimality of D. Conversely, let l be the set of all pendant vertices in G = H ◦ K1 such that |l| = p/2. If G = H ◦ K1 , then there exist a minimal accurate independent dominating set D ⊆ V (G) containing all pendant vertices of G. Hence |D| = |l| = p/2. 2 Now we characterize the trees for which ia (T ) = p − ∆(T ). Theorem 2.7 For any tree T , ia (T ) = p − ∆(T ) if and only if T is a wounded spider and T 6= K1 , K1,1 . Proof Suppose T is wounded spider. Then it is easy to verify that ia (T ) = p − ∆(T ). Conversely, suppose T is a tree with ia (T ) = p − ∆(T ). Let v be a vertex of maximum degree ∆(T ) and u be a vertex in N (v) which has degree 1. If T − N [v] = φ then T is the star K1,n , n ≥ 2. Thus T is a double wounded spider. Assume now there is at least one vertex in T − N [v]. Let S be a maximal independent set of hT − N [v]i. Then either S ∪ {v} or S ∪ {u} is an accurate independent dominating set of T. Thus p = ia (T ) + ∆(T ) ≤ |S| + 1 + ∆(T ) ≤ p. This implies that V − N (v) is an accurate independent dominating set. Furthermore, N (v) is also an accurate independent dominating set. The connectivity of T implies that each vertex in V − N [v] must be adjacent to at least one vertex in N (v). Moreover if any vertex in V − N [v] is adjacent to two or more vertices in N (v), then a cycle is formed. Hence each vertex in V − N [v] is adjacent to exactly one vertex in N (v). To show that ∆(T ) + 1 vertices are necessary to dominate T, there must be at least one vertex in N (v) which are not adjacent to any vertex in V − N [v] and each vertex in N (v) has either 0 or 1 neighbors in V − N [v]. Thus T is a wounded spider. 2 Proposition 2.7 If G is a path Pp , p ≥ 3 then γ(Pp ) = ia (Pp ). We characterize the class of trees with equal domination and accurate independent domination number in the next section. §3. Characterization of (γ, ia )-Trees For any graph theoretical parameter λ and µ, we define G to be (λ, µ)-graph if λ(G) = 92 B. Basavanagoud and Sujata Timmanaikar µ(G). Here we provide a constructive characterization of (γ, ia )-trees. To characterize (γ, ia )-trees we introduce family τ1 of trees T = Tk that can be obtained as follows. If k is a positive integer, then Tk+1 can be obtained recursively from Tk by the following operation. Operation O Attach a path P3 (x,y,z) and an edge mx, where m is a support vertex of a tree T. τ = {T /obtained from P5 by finite sequence of operations of O} b b b b b b b b b b b b b b Tree T belonging to family τ1 Observation 3.1 If T ∈ τ, then 1. ia (T ) = ⌈p + 1/3⌉; 2. X(T ) is a minimal dominating set as well as a minimal accurate independent dominating set of T ; 3. hV − Di is totally disconnected. Corollary 3.1 If tree T with p ≥ 5 belongs to the family τ then γ(T ) = |X(T )| and ia (T ) = |X(T )|. Lemma 3.1 If a tree T belongs to the family τ then T is a (γ, ia ) − tree. Proof If T = Pp , p ≥ 3 then from proposition 2.7 T is a (γ, ia )−tree. Now if T = Pp , p ≥ 3 then we proceed by induction on the number of operations n(T ) required to construct the tree T . If n(T ) = 0 then T ∈ P5 by proposition 2.7 T is a (γ, ia )-tree. Assume now that T is a tree belonging to the family τ with n(T ) = k, for some positive integer k and each tree T ′ ∈ τ with n(T ′ ) < k and with V (T ′ ) ≥ 5 is a (γ, ia )-tree in which X(T ′ ) is a minimal accurate independent dominating set of T ′ . Then T can be obtained from a tree T ′ belonging to τ by operation O where m ∈ V (T ′ ) − (M (T ′ ) − Ω(T ′ )) and we add Accurate Independent Domination in Graphs 93 path (x, y, z) and the edge mx. Then z is a pendant vertex in T and y is a support vertex and x ∈ M (T ). Thus S(T ) = X(T ′ ) ∪ {y} is a minimal accurate independent dominating set of T. Therefore ia (T ) ≥ |X(T )| = |X(T ′ )| + 1. Hence we conclude that ia (T ) = ia (T ′ ) + 1. By the induction hypothesis and by observation 3.1(2) ia (T ′ ) = γ(T ′ ) = |X(T ′ )|. In this way ia (T ) = |X(T )| and in particular ia (T ) = γ(T ). 2 Lemma 3.2 If T is a (γ, ia ) − tree, then T belongs to the family τ . Proof If T is a path Pp , p ≥ 3 then by proposition 2.7 T is a (γ, ia ) − tree. It is easy to verify that the statement is true for all trees T with diameter less than or equal to 4. Hence we may assume that diam(T ) ≥ 4. Let T be rooted at a support vertex m of a longest path P. Let P be a m − z path and let y be the neighbor of z. Further, let x be a vertex belongs to M (T ). Let T be a (γ, ia )-tree. Now we proceed by induction on number of vertices |V (T )| of a (γ, ia )tree. Let T be a (γ, ia )-tree and assume that the result holds good for all trees on V (T ) − 1 vertices. By observation 3.1(2) since T is (γ, ia )-tree it contains minimal accurate independent dominating set D that contains all support vertices of a tree. In particular {m, y} ⊂ D and the vertices x and z are independent in hV − Di. Let T ′ = T − (x, y, z). Then D − {y}is dominating set of T ′ and so γ(T ′ ) ≤ γ(T ) − 1. Any dominating set can be extended to a minimal accurate independent dominating set of T by adding to it the vertices (x, y, z) and so ia (T ) ≤ ia (T ′ ) + 1. Hence, ia (T ′ ) ≤ γ(T ′ ) ≤ γ(T ) + 1 ≤ ia (T ) − 1 ≤ ia (T ′ ). Consequently, we must have equality throughout this inequality chain. In particular ia (T ′ ) = γ(T ′ ) and ia (T ) = ia (T ′ ) + 1. By inductive hypothesis any minimal accurate independent dominating set of a tree T ′ can be extended to minimal accurate 2 independent dominating set of a tree T by operation O. Thus T ∈ τ. As an immediate consequence of lemmas 3.1 and 3.2, we have the following characterization of trees with equal domination and accurate independent domination number. Theorem 3.1 Let T be a tree. Then ia (T ) = γ(T ) if and only if T ∈ τ. §4. Accurate Independent Domination of Some Graph Families In this section accurate independent domination of fan graph,double fan graph, helm graph and gear graph are considered. We also obtain the corresponding relation between other dominating parameters and dominator coloring of the above graph families. Definition 4.1 A fan graph, denoted by Fn can be constructed by joining n copies of the cycle graph C3 with a common vertex. Observation 4.1 Let Fn be a fan. Then, 1. Fn is a planar undirected graph with 2n + 1 vertices and 3n edges; 2. Fn has exactly one vertex with ∆(Fn ) = p − 1; 3. Diam(Fn ) = 2. 94 B. Basavanagoud and Sujata Timmanaikar Theorem 4.1([2]) For a fan graph Fn , n ≥ 2, χd (Fn ) = 3. Proposition 4.1 For a fan graph Fn , n ≥ 2, ia (Fn ) = 1. Proof By Observation 4.1(2) and Theorem 2.5 result holds. 2 Proposition 4.2 For a fan graph Fn , n ≥ 2, ia (Fn ) < χd (Fn ). Proof By Proposition 4.1 and Theorem 4.1, we know that χd (Fn ) = 3. This implies that ia (Fn ) < χd (Fn ). 2 Definition 4.2 A double fan graph, denoted by F2,n isomorphic to Pn + 2K1 . Observation 4.2 1. F2,n is a planar undirected graph with (n + 2) vertices and (3n − 1) edges; 2. Diam(G) = 2. Theorem 4.2([2]) For a double fan graph F2,n , n ≥ 2, χd (F2,n ) = 3. Theorem 4.3 For a double fan graph F2,n , n ≥ 2, ia (F2,2 ) = 2, ia (F2,3 ) = 1, ia (F2,5 ) = 3 and ia (F2,n ) = 2 if n ≥ 7. Proof Our proof is divided into cases following. Case 1. If n = 2 and n ≥ 7, then F2,n , n ≥ 2 has only one accurate independent dominating set D of |D| = 2. Hence, ia (F2,n ) = 2. Case 2. If n = 3, thenF2,3 has exactly one vertex of ∆(G) = p − 1. Then by Theorem 2.5, ia (F2,n ) = 1. Case 3. If n=5 and D be a independent dominating set of G with |D| = 2, then (V − D) also has an independent dominating set of cardinality 2. Hence D is not accurate. Let D1 be a independent dominating set with |D1 | = 3, then V − D1 has no independent dominating set of cardinality 3. Then D1 is accurate. Hence, ia (F2,n ) = 3. Case 4. If n=4 and 6, there does not exist accurate independent dominating set. 2 Proposition 4.3 For a double fan graph F2,n , n ≥ 7, γ(F2,n ) = i(F2,n ) = γa (F2,n ) = ia (F2,n ) = 2 . Proof Let F2,n , n ≥ 7 be a Double fan graph. Then 2k1 forms a minimal dominating set of F2,n such that γ(F2,n ) = 2. Since this dominating set is independent and in (V − D) there is no independent dominating set of cardinality 2 it is both independent and accurate independent dominating set. Also it is accurate dominating set. Hence, γ(F2,n ) = i(F2,n ) = γa (F2,n ) = ia (F2,n ) = 2. 2 Accurate Independent Domination in Graphs 95 Proposition 4.4 For Double fan graph F2,n , n ≥ 7 ia (F2,n ) ≤ χd (F2,n ). Proof The proof follows by Theorems 4.2 and 4.3. 2 Definition 4.3([1]) For n ≥ 4,the wheel Wn is defined to be the graph Wn = Cn−1 + K1 . Also it is defined as W1,n = Cn + K1 . Definition 4.4 A helm Hn is the graph obtained from W1,n by attaching a pendant edge at each vertex of the n-cycle. Observation 4.3 A helm Hn is a planar undirected graph with (2n+1) vertices and 3n edges. Theorem 4.4([2]) For Helm graph Hn , n ≥ 3, χd (Hn ) = n + 1. Proposition 4.5 For a helm graph Hn , n ≥ 3, ia (Hn ) = n. Proof Let Hn , n ≥ 3 be a helm graph. Then there exist a minimal independent dominating set D with |D| = n and (V − D) has no independent dominating set of cardinality n. Hence D 2 is accurate. Therefore ia (Hn ) = n. Proposition 4.6 For a helm graph Hn , n ≥ 3 γ(Hn ) = i(Hn ) = γa (Hn ) = ia (Hn ) = n. Proposition 4.7 For a helm graph Hn , n ≥ 3 ia (Hn ) = χd (Hn ) − 1. Proof Applying Proposition 4.5 , ia (Hn ) = n = n + 1 − 1 = χd (Hn ) − 1 by Theorem 4.4, 2 χd (Hn ) = n + 1. Hence the proof. Definition 4.5 A gear graph Gn also known as a bipartite wheel graph, is a wheel graph W1,n with a vertex added between each pair of adjacent vertices of the outer cycle. Observation 4.4 A gear graph Gn is a planar undirected graph with 2n + 1 vertices and 3n edges. Theorem 4.5([2]) For a gear graph Gn , n ≥ 3, χd (Gn ) = ⌈2n/3⌉ + 2. 96 B. Basavanagoud and Sujata Timmanaikar Theorem 4.6 For a gear graph Gn , n ≥ 3 , ia (Gn ) = n. Proof It is clear from the definition of gear graph Gn is obtained from wheel graph W1,n with a vertex added between each pair of adjacent vertices of the outer cycle of wheel graph W1,n . These n vertices forms an independent dominating set in Gn such that (V − D) has no independent dominating set of cardinality n. Therefore, the set D with cardinality n is accurate independent dominating set of Gn . Therefore ia (Gn ) = n. 2 Corollary 4.1 For any gear graph Gn , n ≥ 3, γ(Gn ) = i(Gn ) = n − 1. Proposition 4.8 For a gear graph Gn , n ≥ 3, ia (Gn ) = γa (Gn ). Proposition 4.9 For a graph Gn , n ≥ 3 ia (Gn ) = γ(Gn ) + 1 = i(Gn ) + 1. Proof Applying Theorem 4.6 and Corollary 4.1, we know that ia (Gn ) = n = n − 1 + 1 = 2 γ(Gn ) + 1 = i(Gn ) + 1. References [1] F. Harary, Graph Theory, Addison-Wesley, Reading, Mass, (1969). [2] K.Kavitha and N.G.David, Dominator coloring of some classes of graphs, International Journal of Mathematical Archive, 3(11), 2012, 3954-3957. [3] R.Gera, S. Horton, C. Rasmussen, Dominator Colorings and Safe Clique Partitions, Congressus Numerantium , (2006). [4] T. W. Haynes, S. T. Hedetniemi and P. J. Slater, Fundamentals of Domination in Graphs, Marcel Dekker, Inc., New York, (1998). [5] T. W. Haynes, S. T. Hedetniemi and P. J. Slater, Domination in Graphs- Advanced Topics, Marcel Dekker, Inc., New York, (1998). [6] V. R. Kulli and M.B.Kattimani, The Accurate Domination Number of a Graph, Technical Report 2000:01, Dept.Mathematics, Gulbarga University, Gulbarga, India (2000). [7] V. R. Kulli, Theory of Domination in Graphs, Vishwa International Publications, Gulbarga, India (2010). [8] V. R. Kulli, Advances in Domination Theory-I, Vishwa International Publications, Gulbarga, India (2012). [9] V. R. Kulli and M.B.Kattimani, Accurate Domination in Graphs, In V.R.Kulli, ed., Advances in Domination Theory-I, Vishwa International Publications, Gulbarga, India (2012) 1-8. International J.Math. Combin. Vol.2(2018), 97-113 On r-Dynamic Coloring of the Triple Star Graph Families T.Deepa and M. Venkatachalam (Department of Mathematics, Kongunadu Arts and Science College, Coimbatore - 641 029,Tamilnadu, India) E-mail: deepathangavelu88@gmail.com, venkatmaths@gmail.com Abstract: An r-dynamic coloring of a graph G is a proper coloring c of the vertices such that |c(N (v))| ≥ min {r, d(v)}, for each v ∈ V (G). The r-dynamic chromatic number of a graph G is the minimum k such that G has an r-dynamic coloring with k colors. In this paper we investigate the r-dynamic chromatic number of the central graph, middle graph, total graph and line graph of the triple star graph K1,n,n,n denoted by C(K1,n,n,n ), M (K1,n,n,n ), T (K1,n,n,n ) and L(K1,n,n,n ) respectively. Key Words: Smarandachely r-dynamic coloring, r-dynamic coloring, triple star graph, central graph, middle graph, total graph and line graph. AMS(2010): 05C15. §1. Introduction Graphs in this paper are simple and finite. For undefined terminologies and notations see [5, 17]. Thus for a graph G, δ(G), ∆(G) and χ(G) denote the minimum degree, maximum degree and chromatic number of G respectively. When the context is clear we write, δ, ∆ and χ for brevity. For v ∈ V (G), let N (v) denote the set of vertices adjacent to v in G and d(v) = |N (v)|. The r-dynamic chromatic number was first introduced by Montgomery [14]. An r-dynamic coloring of a graph G is a map c from V (G) to the set of colors such that (i) if uv ∈ E(G), then c(u) 6= c(v) and (ii) for each vertex v ∈ V (G), |c(N (v))| ≥ min {r, d(v)}, where N (v) denotes the set of vertices adjacent to v, d(v) its degree and r is a positive integer. Generally, for a subgraph G′ ≺ G and a coloring c on G if |c(N (v))| ≥ min {r, d(v)} for v ∈ V (G \ G′ ) but |c(N (v))| ≤ min {r, d(v)} for u ∈ V (G′ ), such a r coloring is called a Smarandachely r-dynamic coloring on G. Clearly, if G′ = ∅, a Smarandachely r-dynamic coloring is nothing else but the r-dynamic coloring. The first condition characterizes proper colorings, the adjacency condition and second condition is double-adjacency condition. The r-dynamic chromatic number of a graph G, written χr (G), is the minimum k such that G has an r-dynamic proper k-coloring. The 1-dynamic chromatic number of a graph G is equal to its chromatic number. The 2-dynamic chromatic number of a graph has been studied under the name dynamic chromatic number denoted by χd (G) [1-4, 8]. By simple observation, we can show that χr (G) ≤ χr+1 (G), however χr+1 (G) − χr (G) can 1 Received September 9, 2017, Accepted May 26, 2018. 98 T.Deepa and M. Venkatachalam be arbitrarily large, for example χ(P etersen) = 2, χd (P etersen) = 3, but χ3 (P etersen) = 10. Thus, finding an exact values of χr (G) is not trivially easy. There are many upper bounds and lower bounds for χd (G) in terms of graph parameters. For example, for a graph G with ∆(G) ≥ 3, Lai et al. [8] proved that χd (G) ≤ ∆(G) + 1. An upper bound for the dynamic chromatic number of a d-regular graph G in terms of χ(G) and the independence number of G, α(G), was introduced in [7]. In fact, it was proved that χd (G) ≤ χ(G) + 2log2 α(G) + 3. Taherkhani gave in [15] an upper bound for χ2 (G) in terms of the chromatic number, the maximum degree ∆ and the minimum degree δ. i.e., χ2 (G)−χ(G) ≤ (∆e)/δlog 2e ∆2 + 1 . Li et al. proved in [10] that the computational complexity of χd (G) for a 3-regular graph is an NP-complete problem. Furthermore, Li and Zhou [9] showed that to determine whether there exists a 3-dynamic coloring, for a claw free graph with the maximum degree 3, is NP-complete. N.Mohanapriya et al. [11, 12] studied the dynamic chromatic number for various graph families. Also, it was proven in [13] that the r− dynamic chromatic number of line graph of a helm graph Hn is   n − 1,      n + 1,     n + 2, χr (L(Hn )) = n + 3,        n + 4,   n + 5, δ ≤ r ≤ n − 2, r = n − 1, r = n and n ≡ 1 mod 3, r = n and n 6≡ 1 mod 3, r = n + 1 = ∆, n ≥ 6 and 2n − 2 ≡ 0 mod 5, r = n + 1 = ∆, n ≥ 6 and 2n − 2 6≡ 0 mod 5. In this paper, we study χr (G), the r- dynamic chromatic number of the middle, central, total and line graphs of the triple star graphs are discussed. §2. Preliminaries Let G be a graph with vertex set V (G) and edge set E(G). The middle graph [6] of G, denoted by M (G) is defined as follows. The vertex set of M (G) is V (G) ∪ E(G). Two vertices x, y of M (G) are adjacent in M (G) in case one of the following holds: (i) x, y are in E(G) and x, y are adjacent in G. (ii) x is in V (G), y is in E(G), and x, y are incident in G. The central graph [16] C(G) of a graph G is obtained from G by adding an extra vertex on each edge of G, and then joining each pair of vertices of the original graph which were previously non-adjacent. Let G be a graph with vertex set V (G) and edge set E(G). The total graph [6, 16] of G, denoted by T (G) is defined in the following way. The vertex set of T (G) is V (G) ∪ E(G). Two vertices x, y of T (G) are adjacent in T (G) in case one of the following holds: (i) x, y are in V (G) and x is adjacent to y in G. (ii) x, y are in E(G) and x, y are adjacent in G. (iii) x is in V (G), y is in E(G), and x, y are incident in G. The line graph [13] of G denoted by L(G) is the graph with vertices are the edges of G On r-Dynamic Coloring of the Triple Star Graph Families 99 with two vertices of L(G) adjacent whenever the corresponding edges of G are adjacent. Theorem 2.1 For any triple star graph K1,n,n,n , the r-dynamic chromatic number χr (C(K1,n,n,n )) =   2n + 1, r = 1     3n + 1, 2 ≤ r ≤ ∆ − 1 4n + 1, r ≥ ∆ Proof First we apply the definition of central graph on K1,n,n,n . Let the edge vvi , vi wi and wi ui be subdivided by the vertices ei (1 ≤ i ≤ n), e′i (1 ≤ i ≤ n) and e′′i (1 ≤ i ≤ n) in K1,n,n,n . S S S Clearly V (C(K1,n,n,n )) = {v} {vi : 1 ≤ i ≤ n} {wi : 1 ≤ i ≤ n} {ui : 1 ≤ i ≤ n} S S ′ S ′′ {ei : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n}. The vertices vi (1 ≤ i ≤ n) induce a clique of order n (say Kn ) and the vertices v, ui (1 ≤ i ≤ n) induce a clique of order n + 1 (say Kn+1 ) in C(K1,n,n,n ) respectively. Thus, we have χr (C(K1,n,n,n )) ≥ n + 1. Case 1. r = 1. Consider the color class C1 = {c1 , c2 , c3 , · · · , c(2n+1) } and assign the r-dynamic coloring to C(K1,n,n,n ) by Algorithm 2.1.1. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (C(K1,n,n,n )) = 2n + 1. Case 2. 2 ≤ r ≤ ∆ − 1. Consider the color class C2 = {c1 , c2 , c3 , · · · , c(3n+1) } and assign the r-dynamic coloring to C(K1,n,n,n ) by Algorithm 2.1.2. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (C(K1,n,n,n )) = 3n + 1. Case 3. r ≥ ∆. Consider the color class C3 = {c1 , c2 , c3 , · · · , c(4n+1) } and assign the r-dynamic coloring to C(K1,n,n,n ) by Algorithm 2.1.3. Thus, an easy check shows that the r− adjacency condition is 2 fulfilled. Hence χr (C(K1,n,n,n )) = 4n + 1. Algorithm 2.1.1 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in C(K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; 100 T.Deepa and M. Venkatachalam for i = 1 to n { V3 = {vi }; C(vi ) = n + i + 1; } for i = 1 to n { V4 = {e′i }; C(e′i ) = n + 1; } for i = 1 to n { V5 = {wi }; C(wi ) = i; } for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; } for i = 1 to n { V7 = {ui }; C(ui ) = i; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.1.2 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in C(K1,n,n,n ). begin for i = 1 to n { V1 = {ui }; C(ui ) = i; } for i = 1 to n { V2 = {e′′i }; C(e′′i ) = n + 1; } On r-Dynamic Coloring of the Triple Star Graph Families for i = 1 to n { V3 = {wi }; C(wi ) = n + i + 1; } for i = 1 to n { V4 = {e′i }; C(e′i ) = i; } for i = 1 to n { V5 = {vi }; C(vi ) = 2n + i + 1; } for i = 1 to n − 1 { V6 = {ei }; C(ei ) = 2n + i + 2; } C(en ) = 2n + 2; V7 = {v}; C(v) = n + 1; S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.1.3 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in C(K1,n,n,n ). begin for i = 1 to n { V1 = {ui }; C(ui ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n { V3 = {wi }; C(wi ) = n + i + 1; } 101 102 T.Deepa and M. Venkatachalam for i = 1 to n { V4 = {vi }; C(vi ) = 2n + i + 1; } for i = 1 to n { V5 = {ei }; C(ei ) = 3n + i + 1; } for i = 1 to n { V6 = {e′i }; C(e′i ) = i; } for i = 1 to n { V7 = {e′′i }; C(e′′i ) = 3n + 2; } V = V1 end S V2 S V3 S V4 S V5 S V6 S V7 ; Theorem 2.2 For any triple star graph K1,n,n,n , the r-dynamic chromatic number χr (M (K1,n,n,n )) =    n + 1, 1 ≤ r ≤ n n + 2, r = n + 1    n + 3, r ≥ ∆ Proof By definition of middle graph, each edge vvi , vi wi and wi ui be subdivided by the vertices ei (1 ≤ i ≤ n) , e′i (1 ≤ i ≤ n) and e′′i (1 ≤ i ≤ n) in K1,n,n,n and the vertices v, ei induce S a clique of order n + 1(say Kn+1 ) in M (K1,n,n,n ). i.e.,V (M (K1,n,n,n )) = {v} {vi : 1 ≤ i ≤ S S S S S n} {wi : 1 ≤ i ≤ n} {ui : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} {e′i : 1 ≤ i ≤ n} {e′′i : 1 ≤ i ≤ n}. Thus we have χr (M (K1,n,n,n )) ≥ n + 1. Case 1. 1 ≤ r ≤ n. Consider the color class C1 = {c1 , c2 , c3 , · · · , c(n+1) } and assign the r-dynamic coloring to M (K1,n,n,n ) by Algorithm 2.2.1. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (M (K1,n,n,n )) = n + 1, for 1 ≤ r ≤ n. Case 2. r = n + 1. On r-Dynamic Coloring of the Triple Star Graph Families 103 Consider the color class C2 = {c1 , c2 , c3 , · · · , c(n+1) , c(n+2) } and assign the r-dynamic coloring to M (K1,n,n,n ) by Algorithm 2.2.2. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (M (K1,n,n,n )) = n + 2, for r = n + 1 . Case 3. r = ∆. Consider the color class C3 = {c1 , c2 , c3 , · · · , cn , c(n+1) , c(n+2) , c(n+3) } and assign the rdynamic coloring to M (K1,n,n,n ) by Algorithm 2.2.3. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (M (K1,n,n,n )) = n + 3, for r ≥ ∆. 2 Algorithm 2.2.1 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in M (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n { V3 = {vi }; C(vi ) = n + 1; } for i = 1 to n − 1 { V4 = {e′i }; C(e′i ) = i + 1; } C(e′n ) = 1; for i = 1 to n − 2 { V5 = {wi }; C(wi ) = i + 2; } C(wn−1 ) = 1; C(wn ) = 2; for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; 104 T.Deepa and M. Venkatachalam } for i = 1 to n { V7 = {ui }; C(ui ) = i; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.2.2 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in M (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n { V3 = {vi }; C(vi ) = n + 2; } for i = 1 to n { V4 = {e′i }; C(e′i ) = n + 1; } for i = 1 to n − 1 { V5 = {wi }; C(wi ) = i + 1; } C(wn ) = 1; for i = 1 to n − 2 { V6 = {e′′i }; C(e′′i ) = i + 2; } C(e′′n−1 ) = 1; On r-Dynamic Coloring of the Triple Star Graph Families C(e′′n ) = 2; for i = 1 to n { V7 = {ui }; C(ui ) = n + 1; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.2.3 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in M (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n { V3 = {vi }; C(vi ) = n + 2; } for i = 1 to n { V4 = {e′i }; C(e′i ) = n + 3; } for i = 1 to n { V5 = {wi }; C(wi ) = n + 1; } for i = 1 to n − 1 { V6 = {e′′i }; C(e′′i ) = i + 1; } C(e′′n ) = 1; for i = 1 to n 105 106 T.Deepa and M. Venkatachalam { V7 = {ui }; C(ui ) = n + 2; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Theorem 2.3 For any triple star graph K1,n,n,n , the r-dynamic chromatic number,   n + 1, 1 ≤ r ≤ n     r + 1, n + 1 ≤ r ≤ ∆ − 2 χr (T (K1,n,n,n )) =   2n, r = ∆−1     2n + 1, r ≥ ∆ Proof By definition of total graph, each edge vvi , vi wi and wi ui be subdivided by the vertices ei (1 ≤ i ≤ n) , e′i (1 ≤ i ≤ n) and e′′i (1 ≤ i ≤ n) in K1,n,n,n and the vertices v, ei induce S a clique of order n + 1(say Kn+1 ) in T (K1,n,n,n ). i.e.,V (T (K1,n,n,n)) = {v} {vi : 1 ≤ i ≤ S S S S S n} {wi : 1 ≤ i ≤ n} {ui : 1 ≤ i ≤ n} {ei : 1 ≤ i ≤ n} {e′i : 1 ≤ i ≤ n} {e′′i : 1 ≤ i ≤ n}. Thus, we have χr (T (K1,n,n,n)) ≥ n + 1. Case 1. 1 ≤ r ≤ n. Consider the color class C1 = {c1 , c2 , c3 , · · · , c(n+1) } and assign the r-dynamic coloring to T (K1,n,n,n) by Algorithm 2.3.1. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (T (K1,n,n,n )) = n + 1, for 1 ≤ r ≤ n. Case 2. n + 1 ≤ r ≤ ∆ − 2. Consider the color class C2 = {c1 , c2 , c3 , · · · , c(2n−1) } and assign the r-dynamic coloring to T (K1,n,n,n) by Algorithm 2.3.2. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (T (K1,n,n,n )) = r + 1, for n + 1 ≤ r ≤ ∆ − 2. Case 3. r = ∆ − 1. Consider the color class C3 = {c1 , c2 , c3 , · · · , c2n } if r = ∆ − 1 and assign the r-dynamic coloring to T (K1,n,n,n) by Algorithm 2.3.3. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (T (K1,n,n,n )) = 2n for r = ∆ − 1. Case 4. r = ∆. Consider the color class C4 = {c1 , c2 , c3 , · · · , c2n+1 } if r = ∆ and assign the r-dynamic coloring to T (K1,n,n,n) by Algorithm 2.3.4. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (T (K1,n,n,n )) = 2n + 1 for r ≥ ∆. 2 Algorithm 2.3.1 Input: The number ”n” of K1,n,n,n . On r-Dynamic Coloring of the Triple Star Graph Families Output: Assigning r-dynamic coloring for the vertices in T (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n − 3 { V3 = {vi }; C(vi ) = i + 3; } C(vn−2 ) = 1; C(vn−1 ) = 2; C(vn ) = 3; for i = 1 to n − 2 { V4 = {e′i }; C(e′i ) = i + 2; } C(e′n−1 ) = 1; C(e′n ) = 2; for i = 1 to n − 1 { V5 = {wi }; C(wi ) = i + 1; } C(wn ) = 1; for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; } for i = 1 to n { V7 = {ui }; C(ui ) = i; } 107 108 T.Deepa and M. Venkatachalam V = V1 end S V2 S V3 S V4 S V5 S V6 S V7 ; Algorithm 2.3.2 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in T (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n − 2 { V3 = {vi }; C(vi ) = r + 1; } C(vn−1 ) = n + 2; C(vn ) = n + 3; for i = 1 to n − 3 { V4 = {e′i }; C(e′i ) = n + i + 2; } C(e′n−2 ) = n + 2; C(e′n−1 ) = n + 3; C(e′n ) = n + 4; for i = 1 to n − 1 { V5 = {wi }; C(wi ) = i + 1; } C(wn ) = 1; for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; } for i = 1 to n On r-Dynamic Coloring of the Triple Star Graph Families { V7 = {ui }; C(ui ) = i; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.3.3 Input: The number ”n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in T (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n − 1 { V3 = {vi }; C(vi ) = n + i + 1; } C(vn ) = n + 2; for i = 1 to n − 2 { V4 = {e′i }; C(e′i ) = n + i + 2; } C(e′n−1 ) = n + 2; C(e′n ) = n + 3; for i = 1 to n − 1 { V5 = {wi }; C(wi ) = i + 1; } C(wn ) = 1; for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; } 109 110 T.Deepa and M. Venkatachalam for i = 1 to n { V7 = {ui }; C(ui ) = i; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Algorithm 2.3.4 Input: The number “n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in T (K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } V2 = {v}; C(v) = n + 1; for i = 1 to n { V3 = {vi }; C(vi ) = n + i + 1; } for i = 1 to n − 1 { V4 = {e′i }; C(e′i ) = n + i + 2; } C(e′n ) = n + 2; for i = 1 to n − 1 { V5 = {wi }; C(wi ) = i + 1; } C(wn ) = 1; for i = 1 to n { V6 = {e′′i }; C(e′′i ) = n + 1; } for i = 1 to n On r-Dynamic Coloring of the Triple Star Graph Families 111 { V7 = {ui }; C(ui ) = i; } S S S S S S V = V1 V2 V3 V4 V5 V6 V7 ; end Theorem 2.4 For any triple star graph K1,n,n,n , the r-dynamic chromatic number, χr (L(K1,n,n,n )) =   n, 1 ≤ r ≤ n − 1 n + 1, r ≥ ∆ Proof First we apply the definition of line graph on K1,n,n,n . By the definition of line graph, each edge of K1,n,n,n taken to be as vertex in L(K1,n,n,n ).The vertices e1 , e2 ,· · · , en induce a clique of order n in L(K1,n,n,n). i.e., V (L(K1,n,n,n )) = E(K1,n,n,n ) = {ei : 1 ≤ i ≤ S S n} {e′i : 1 ≤ i ≤ n} {e′′i : 1 ≤ i ≤ n}. Thus, we have χr (L(K1,n,n,n )) ≥ n. Case 1. 1 ≤ r ≤ ∆ − 1. Now consider the vertex set V (L(K1,n,n,n )) and color class C1 = {c1 , c2 , · · · , cn }, assign r dynamic coloring to L(K1,n,n,n ) by Algorithm 2.4.1. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (L(K1,n,n,n )) = n, for 1 ≤ r ≤ ∆ − 1. Case 2. r ≥ ∆. Now consider the vertex set V (L(K1,n,n )) and color class C2 = {c1 , c2 , · · · , cn , cn+1 }, assign r dynamic coloring to L(K1,n,n,n ) by Algorithm 2.4.2. Thus, an easy check shows that the r− adjacency condition is fulfilled. Hence, χr (L(K1,n,n,n )) = n + 1 for r ≥ ∆. 2 Algorithm 2.4.1 Input: The number “n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in L(K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } for i = 1 to n − 1 { V2 = {e′i }; C(e′i ) = i + 1; } C(e′n ) = 1; 112 T.Deepa and M. Venkatachalam for i = 1 to n − 2 { V3 = {e′′i }; C(e′′i ) = i + 2; } C(e′′n−1 ) = 1; C(e′′n ) = 2; S S V = V1 V2 V3 ; end Algorithm 2.4.2 Input: The number “n” of K1,n,n,n . Output: Assigning r-dynamic coloring for the vertices in L(K1,n,n,n ). begin for i = 1 to n { V1 = {ei }; C(ei ) = i; } for i = 1 to n { V2 = {e′i }; C(e′i ) = n + 1; } for i = 1 to n − 1 { V3 = {e′′i }; C(e′′i ) = i + 1; } C(e′′n ) = 1; S S V = V1 V2 V3 ; end References [1] A. Ahadi, S. Akbari, A. Dehghana, M. Ghanbari, On the difference between chromatic number and dynamic chromatic number of graphs, Discrete Math. 312 (2012), 2579–2583. [2] S. Akbari, M. Ghanbari, S. Jahanbakam, On the dynamic chromatic number of graphs, Contemp. Math. (Amer. Math. Soc.), 531 (2010), 11–18. [3] S. Akbari, M. Ghanbari, S. Jahanbekam, On the list dynamic coloring of graphs, Discrete Appl. Math. 157 (2009), 3005–3007 On r-Dynamic Coloring of the Triple Star Graph Families 113 [4] M. Alishahi, Dynamic chromatic number of regular graphs, Discrete Appl. Math. 160 (2012), 2098–2103. [5] J. A. Bondy, U. S. R. Murty, Graph Theory, Springer, 2008. [6] Danuta Michalak, On middle and total graphs with coarseness number equal 1, Springer Verlag, Graph Theory, Lagow Proceedings, Berlin Heidelberg, New York, Tokyo, (1981), 139–150. [7] A. Dehghan, A. Ahadi, Upper bounds for the 2-hued chromatic number of graphs in terms of the independence number, Discrete Appl. Math. 160(15) (2012), 2142–2146. [8] H. J. Lai, B. Montgomery, H. Poon, Upper bounds of dynamic chromatic number, Ars Combin. 68 (2003), 193–201. [9] X. Li, W. Zhou, The 2nd-order conditional 3-coloring of claw-free graphs, Theoret. Comput. Sci. 396 (2008), 151–157. [10] X. Li, X. Yao, W. Zhou, H. Broersma, Complexity of conditional colorability of graphs, Appl. Math. Lett. 22 (2009), 320–324. [11] N. Mohanapriya, J. Vernold Vivin and M. Venkatachalam, δ- dynamic chromatic number of helm graph families, Cogent Mathematics, 3(2016), No. 1178411. [12] N. Mohanapriya, J. Vernold Vivin and M. Venkatachalam, On dynamic coloring of Fan graphs, Int J of Pure Appl Math, 106(2016), 169-174. [13] N. Mohanapriya, Ph.D thesis, A Study on Dynamic Coloring of Graphs, Bharathiar University, (2017), Coimbatore, India. [14] B. Montgomery, Dynamic Coloring of Graphs, ProQuest LLC, Ann Arbor, MI, (2001), Ph.D Thesis, West Virginia University. [15] A. Taherkhani, r-Dynamic chromatic number of graphs, Discrete Appl. Math., 201(2016), 222–227. [16] J. Vernold Vivin, Ph.D Thesis, Harmonious Coloring of Total Graphs, n−Leaf, Central Graphs and Circumdetic Graphs, Bharathiar University, (2007), Coimbatore, India. [17] A. T. White, Graphs, Groups and Surfaces, American Elsevier, New York, 1973. International J.Math. Combin. Vol.2(2018), 114-121 (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder V.Ramachandran (Department of Mathematics, Mannar Thirumalai Naicker College, Madurai, Tamil Nadu, India) E-mail: me.ram111@gmail.com Abstract: A (p, q)-graph G is said to be (1, N )-arithmetic labelling if there is a function φ from the vertex set V (G) to {0, 1, N, (N + 1), 2N, (2N + 1), · · · , N (q − 1), N (q − 1) + 1} so that the values obtained as the sums of the labelling assigned to their end vertices, can be arranged in the arithmetic progression {1, N + 1, 2N + 1, · · · , N (q − 1) + 1}. In this paper we prove that ladder and subdivision of ladder are (1, N )-arithmetic labelling for every positive integer N > 1. Key Words: Ladder, subdivision of ladder, one modulo N graceful, Smarandache k modulo N graceful. AMS(2010): 05C78. §1. Introduction V.Ramachandran and C. Sekar [8, 9] introduced one modulo N graceful where N is any positive integer. In the case N = 2, the labelling is odd graceful and in the case N = 1 the labelling is graceful. A graph G with q edges is said to be one modulo N graceful (where N is a positive integer) if there is a function φ from the vertex set of G to {0, 1, N, (N + 1), 2N, (2N + 1), · · · , N (q − 1), N (q − 1) + 1} in such a way that (i) φ is 1 − 1 (ii) φ induces a bijection φ∗ from the edge set of G to {1, N + 1, 2N + 1, · · · , N (q − 1) + 1} where φ∗ (uv) = |φ(u) − φ(v)|. Generally, a graph G with q edges is called to be Smarandache k modulo N graceful if one replacing N by kN in the definition of one modulo N graceful graph. Clearly, a graph G is Smarandache k modulo N graceful if and only if it is one modulo kN graceful by definition. B. D. Acharya and S. M. Hegde [2] introduced (k, d)- arithmetic graphs. A (p, q)- graph G is said to be (k, d)- arithmetic if its vertices can be assigned distinct nonnegative integers so that the values of the edges, obtained as the sums of the numbers assigned to their end vertices, can be arranged in the arithmetic progression k, k + d, k + 2d, . . . , k + (q − 1)d. Joseph A. Gallian [4] surveyed numerous graph labelling methods. V.Ramachandran and C. Sekar [10] introduced (1, N )-Arithmetic labelling. We proved that stars, paths, complete bipartite graph Km,n , highly irregular graph Hi (m, m) and cycle C4k are (1, N )-Arithmetic labelling,C4k+2 is not (1, N )-Arithmetic labelling. We also proved that no graph G containing an odd cycle is (1, N )-arithmetic labelling for every positive integer 1 Received August 24, 2017, Accepted May 28, 2018. (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder 115 N . A (p, q)-graph G is said to be (1, N )-Arithmetic labelling if there is a function φ : V (G) → {0, 1, N, (N + 1), 2N, (2N + 1), · · · , N (q − 1), N (q − 1) + 1}. In this situation the induced mapping φ∗ to the edges is given by φ∗ (uv)=φ(u) + φ(v). If the values of φ(u) + φ(v) are 1, N + 1, 2N + 1, . . . , N (q − 1) + 1 all distinct, then we call the labelling of vertices as (1, N )- Arithmetic labelling. In case if the induced mapping φ∗ is defined as φ∗ (uv)=|φ(u) − φ(v)| and if the resulting edge labels are are distinct and equal to {1, N + 1, 2N + 1, · · · , N (q − 1) + 1}. We call it as one modulo N graceful. In this paper we prove that Ladder and Subdivision of Ladder are (1, N )-Arithmetic labelling for every positive integer N > 1. §2. Main Results Definition 2.1 A graph G with q edges is said to be one modulo N graceful (where N is a positive integer) if there is a function φ from the vertex set of G to {0, 1, N, (N + 1), 2N, (2N + 1), · · · , N (q − 1), N (q − 1) + 1} in such a way that (i) φ is 1 − 1 and (ii) φ induces a bijection φ∗ from the edge set of G to {1, N + 1, 2N + 1, · · · , N (q − 1) + 1} where φ∗ (uv) = |φ(u) − φ(v)|. Definition 2.2 A (p, q)-graph G is said to be (1, N )-Arithmetic labelling if there is a function φ from the vertex set V (G) to {0, 1, N, (N + 1), 2N, (2N + 1), · · · , N (q − 1), N (q − 1) + 1} so that the values obtained as the sums of the labelling assigned to their end vertices, can be arranged in the arithmetic progression {1, N + 1, 2N + 1, · · · , N (q − 1) + 1}. Definition 2.3 A (p, q)- graph G is said to be (k, d)- arithmetic if its vertices can be assigned distinct nonnegative integers so that the values of the edges, obtained as the sums of the numbers assigned to their end vertices, can be arranged in the arithmetic progression k, k + d, k + 2d, · · · , k + (q − 1)d. Definition 2.4([7]) Let G be a graph with p vertices and q edges. A graph H is said to be a subdivision of G if H is obtained from G by subdividing every edge of G exactly once. H is denoted by S(G). Definition 2.5 The ladder graph Ln is defined by Ln = Pn × K2 where Pn is a path with × denotes the cartesian product. Ln has 2n vertices and 3n − 2 edges. Theorem 2.6 For every positive integer n, ladder Ln is (1, N )-Arithmetic labelling, for every positive integer N > 1. Proof Let u1 , u2 , · · · , un and v1 , v2 , · · · , vn be the vertices of Ln , respectively, and let ui vi+1 , i = 1, 2, · · · , n − 1. vi ui+1 , i = 1, 2, · · · , n − 1 and ui vi , i = 1, 2, · · · , n be the edges of Ln . The ladder graph Ln is defined by Ln = Pn ×K2 where Pn is a path with × denotes the cartesian product. Then the ladder Ln has 2nvertices and 3n − 2 edges as shown in figures following. Define φ(ui ) = N (i − 1) for i = 1, 2, 3, · · · , n, φ(vi ) = 2N (i − 1) + 1 for i = 1, 2, 3, · · · , n. 116 V.Ramachandran u1 v1 v2 u2 u3 v3 v4 u4 vn−1 un−1 vn un Figure 1 Ladder Ln where n is odd u1 v1 v2 u2 u3 v3 v4 u4 un−1 vn−1 un vn Figure 2 Ladder Ln where n is even From the definition of φ it is clear that {φ(ui ), i = 1, 2, · · · , n} [ {φ(vi ), i = 1, 2, · · · , n} = {0, N, 2N, . . . , N (n − 1)} ∪ {1, 2N + 1, 4N + 1, · · · , 2N (n − 1) + 1} It is clear that the vertices have distinct labels. Therefore φ is 1 − 1. We compute the edge labels as follows: for i = 1, 2, · · · , n, φ∗ (vi ui ) =| φ(vi ) + φ(ui ) | = 3N (i − 1) + 1; for i = 1, 2, · · · , n − 1, φ∗ (vi+1 ui ) =| φ(vi+1 ) + φ(ui ) | = N (3i − 1) + 1, φ∗ (vi ui+1 ) =| φ(vi ) + φ(ui+1 ) | = N (3i − 2) + 1. This shows that the edges have the distinct labels {1, N + 1, 2N + 1, · · · , N (q − 1) + 1}, where q = 3n − 2. Hence Ln is (1, N )-Arithmetic labelling for every positive integer N > 1. 2 Example 2.7. A (1, 5)-Arithmetic labelling of L6 is shown in Figure 3. (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder 0 1 11 5 10 21 31 15 20 41 51 25 117 Figure 3 Example 2.8 A (1, 2)-Arithmetic labelling of L7 is shown in Figure 4. 0 1 5 2 4 9 13 6 8 17 21 10 12 25 Figure 4 Theorem 2.9 A subdivision of ladder Ln is (1, N )-Arithmetic labelling for every positive integer N > 1. Proof Let G = Ln . The ladder graph Ln is defined by Ln = Pn × K2 where Pn is a path with × denotes the cartesian product. Ln has 2n vertices and 3n − 2 edges. A graph H is said to be a subdivision of G if H is obtained from G by subdividing every edge of G exactly once. H is denoted by S(G). Then the subdivision of ladder Ln has 5n − 2 vertices and 6n − 4 edges as shown in Figure 5. Let H = S(Ln ). 118 V.Ramachandran u1 u12 u2 u23 u3 11 u34 u4 u45 u5 u56 u6 un−1 un w1 w2 w3 w4 w5 v1 v12 v2 v23 v3 v34 v4 v45 v5 v56 w6 wn−1 un−1,n vn−1,n wn v6 vn−1 vn Figure 5 Subdivision of ladder Ln Define the following functions: η : N → N by and γ : N → N by   N (2i − 1) η(i) =  2N (i − 1)   2N (i − 1) γ(i) =  N (2i − 1) if i is even if i is odd if i is even if i is odd Define φ : V → {0, 1, 2, · · · , q} by φ(ui ) = η(i), i = 1, 2, · · · , n φ(vi ) = γ(i), i = 1, 2, · · · , n. Define   1 + (i − 1)4N φ(ui,i+1 ) =  (4i − 1)N + 1 if i is odd if i is even. For i = 1, 2, · · · , n − 2, define φ(vi,i+1 ) = φ(ui+1,i+2 ) − 4N, φ(vn−1,n ) = φ(un−2,n−1 ) + 4N, 119 (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder   1 + (4i − 3)N φ(wi ) =  4N n − 4N + 1 if i = 1, 2, · · · , n − 1 if i = n . It is clear that the vertices have distinct labels. Therefore φ is 1 − 1. We compute the edge labels as follows: φ∗ (wn un ) =| φ(wn ) + φ(un ) | = 6N n − 6N + 1, φ∗ (wn vn ) =| φ(wn ) + φ(vn ) | = 6N n − 5N + 1,   6N n − 12N + 1 φ∗ (vn−1,n vn−1 ) =| φ(vn−1,n ) + φ(vn−1 ) | =  6N n − 8N + 1 if n is odd if n is even. ,   6N n − 9N + 1 if n is odd φ∗ (vn−1,n vn ) =| φ(vn−1,n ) + φ(vn ) | =  6N n − 7N + 1 if n is even. For i = 1, 2, · · · , n − 1,   N (6i − 4) + 1 if i is even φ∗ (wi ui ) =| φ(wi ) + φ(ui ) | =  N (6i − 5) + 1 if i is odd.   N (6i − 5) + 1 if i is even φ∗ (wi vi ) =| φ(wi ) + φ(vi ) | =  N (6i − 4) + 1 if i is odd. For i = 1, 2, · · · , n − 1,   N (6i − 2) + 1 φ∗ (ui,i+1 ui ) =| φ(ui,i+1 ) + φ(ui ) | =  N (6i − 6) + 1 if i is even if i is odd.   N (6i − 3) + 1 if i is odd φ∗ (ui,i+1 ui+1 ) =| φ(ui,i+1 ) + φ(ui+1 ) | =  N (6i − 1) + 1 if i is even. For i = 1, 2, · · · , n − 2,   N (6i − 6) + 1 φ∗ (vi,i+1 vi ) =| φ(vi,i+1 ) + φ(vi ) | =  N (6i − 2) + 1 if i is even if i is odd.   N (6i − 3) + 1 φ∗ (vi,i+1 vi+1 ) =| φ(vi,i+1 ) + φ(vi+1 ) | =  N (6i − 1) + 1 if i is even if i is odd. This shows that the edges have distinct labels {1, N + 1, 2N + 1, · · · , N (q − 1) + 1} with 2 q = 6n − 4. Hence S(Ln ) is (1, N )-Arithmetic labelling for every positive integer N > 1. Example 2.10 A (1, 3)-Arithmetic labelling of S(L5 ) is shown in Figure 6. 120 V.Ramachandran 4 0 3 1 10 16 9 6 22 13 28 12 15 25 34 40 21 18 37 46 24 27 49 Figure 6 Example 2.11 A (1, 10)-Arithmetic labelling of S(L6 ) is shown in Figure 7. 0 11 1 30 51 91 40 41 50 131 70 151 31 20 71 81 10 111 60 171 121 80 90 161 191 110 201 Figure 7 100 (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder 121 References [1] B. D. Acharya, On d-sequential graphs, J. Math. Phys. Sci., 17 (1983), 21-35. [2] B. D. Acharya and S. M. Hegde, Arithmetic graphs, Journal of Graph Theory, 14 (1990) 275-299. [3] B. D. Acharya and S. M. Hegde, On certain vertex valuations of a graph I, Indian J. Pure Appl. Math., 22 (1991) 553-560. [4] Joseph A. Gallian, A Dynamic Survey of Graph Labeling, The Electronic Journal of Combinatorics, #DS6 (2016). [5] S.W.Golomb, How to Number a Graph in Graph theory and Computing, R.C. Read, ed., Academic Press, New york (1972)23-27. [6] R. B. Gnanajothi, Topics in Graph theory, Ph.D. thesis, Madurai Kamaraj University, 1991. [7] KM. Kathiresan, Subdivisions of ladders are graceful, Indian J. Pure Appl. Math., (1992), 21–23. [8] V. Ramachandran, C. Sekar, One modulo N gracefullness of arbitrary supersubdivisions of graphs, International J. Math. Combin., Vol.2 (2014), 36-46. [9] V. Ramachandran, C. Sekar, One modulo N gracefulness of supersubdivision of ladder, Journal of Discrete Mathematical Sciences and Cryptography, Vol.18 (3) (2015), 265-274. [10] V. Ramachandran, C. Sekar, (1, N )-arithmetic graphs, International Journal of Computers and Applications, Vol.38 (1) (2016) 55-59. [11] C. Sekar, Studies in Graph Theory, Ph.D. thesis, Madurai Kamaraj University, 2002. International J.Math. Combin. Vol.2(2018), 122-128 3-Difference Cordial Labeling of Corona Related Graphs R.Ponraj Department of Mathematics, Sri Paramakalyani College, Alwarkurichi-627412, India M.Maria Adaickalam Department of Mathematics Manonmaniam Sundaranar University, Abishekapatti, Tirunelveli-627012,Tamilnadu India E-mail: ponrajmaths@gmail.com, mariaadaickalam@gmail.com Abstract: Let G be a (p, q) graph. Let f : V (G) → {1, 2, · · · , k} be a map where k is an integer 2 ≤ k ≤ p. For each edge uv, assign the label |f (u) − f (v)|. f is called k-difference cordial labeling of G if |vf (i) − vf (j)| ≤ 1 and |ef (0) − ef (1)| ≤ 1 where vf (x) denotes the number of vertices labelled with x, ef (1) and ef (0) respectively denote the number of edges labelled with 1 and not labelled with 1. A graph with a k-difference cordial labeling is called a k-difference cordial graph. In this paper we investigate 3-difference cordial labeling behavior of DTn ⊙ K1 DTn ⊙ 2K1 ,DTn ⊙ K2 and some more graphs. Key Words: Difference cordial labeling, Smarandachely k-difference cordial labeling, path, complete graph, triangular snake, corona. AMS(2010): 05C78. §1. Introduction All Graphs in this paper are finite ,undirect and simple. Let G1 , G2 respectively be (p1 , q1 ), (p2 , q2 ) graphs. The corona of G1 with G2 , G1 ⊙ G2 is the graph obtained by taking one copy of G1 and p1 copies of G2 and joining the ith vertex of G1 with an edge to every vertex in the ith copy of G2 . Ponraj et al. [3], has been introduced the concept of k-difference cordial labeling of graphs and studied the 3-difference cordial labeling behavior of of some graphs. In [4,5,6,7] they investigate the 3-difference cordial labeling behavior of path, cycle, complete graph, complete (t) bipartite graph, star, bistar, comb, double comb, quadrilateral snake, C4 , S(K1,n ), S(Bn,n ) and carona of some graphs with double alternate triangular snake double alternate quadrilateral snake . In this paper we examine the 3-difference cordial labeling behavior of DTn ⊙ K1 DTn ⊙ 2K1 ,DTn ⊙ K2 etc. Terms are not defined here follows from Harary [2]. §2. k-Difference Cordial Labeling Definition 2.1 Let G be a (p, q) graph and let f : V (G) → {1, 2, · · · , k} be a map. For 1 Received November 26, 2017, Accepted May 29, 2018. 3-Difference Cordial Labeling of Corona Related Graphs 123 each edge uv, assign the label |f (u) − f (v)|. f is called a k-difference cordial labeling of G if |vf (i) − vf (j)| ≤ 1 and |ef (0) − ef (1)| ≤ 1 where vf (x) denotes the number of vertices labelled with x, ef (1) and ef (0) respectively denote the number of edges labelled with 1 and not labelled with 1. A graph with a k-difference cordial labeling is called a k-difference cordial graph. On the other hand, if |vf (i) − vf (j)| ≥ 1 or |ef (0) − ef (1)| ≥ 1, such a labeling is called a Smarandachely k-difference cordial labeling of G. A double triangular snake DTn consists of two triangular snakes that have a common path. That is a double triangular snake is obtained from a path u1 u2 · · · un by joining ui and ui+1 to two new vertices vi (1 ≤ i ≤ n − 1) and wi (1 ≤ i ≤ n − 1). First we investigate the 3-difference cordial labeling behavior of DTn ⊙ K1 . Theorem 2.1 DTn ⊙ K1 is 3-difference cordial. S S Proof Let V (DTn ⊙ K1 ) = V (DTn ) {xi : 1 ≤ i ≤ n} {vi′ , wi′ : 1 ≤ i ≤ n − 1} and S S E(DTn ⊙ K1 ) = E(DTn ) {ui xi : 1 ≤ i ≤ n} {vi vi′ , wi wi′ : 1 ≤ i ≤ n − 1}. Case 1. n is even. First we consider the path vertices ui . Assign the label 1 to all the path vertices ui (1 ≤ i ≤ n). Then assign the label 2 to the path vertices v1 , v3 , v5 , · · · and assign the label 1 to the path vertices v2 , v4 , v6 , · · · . Now we consider the vertices wi . Assign the label 2 to all the vertices wi (1 ≤ i ≤ n − 1). Next we move to the vertices vi′ and wi′ . Assign the label 2 ′ to the vertices v2i+1 for all the values of i = 0, 1, 2, 3, · · · and assign the label 1 to the vertices v2i for i = 1, 2, 3, · · · . Next we assign the label 1 to the vertex w1′ and assign the label 3 to the vertices w2′ , w3′ , w4′ , · · · Finally assign the label 3 to all the vertices of xi (1 ≤ i ≤ n). The vertex condition and the edge conditions are vf (1) = vf (2) = 6n−3 and vf (3) = 6n−2 and 3 3 ef (0) = 4n − 4 and ef (1) = 4n − 3. Case 2. n is odd. Assign the label to the path vertices ui (1 ≤ i ≤ n), vi (1 ≤ i ≤ n − 1), wi (1 ≤ i ≤ n − 1), (1 ≤ i ≤ n − 1), xi (1 ≤ i ≤ n) as in case 1. Then assign the label 3 to all the vertices (1 ≤ i ≤ n − 1). Since ef (0) = 4n − 3, ef (1) = 4n − 4 and vf (1) = vf (3) = 2n − 1 and 2 vf (2) = 2n − 2, DTn ⊙ K1 is 3-difference cordial. vi′ wi′ Next investigation about DTn ⊙ 2K1 . Theorem 2.2 DTn ⊙ 2K1 is 3-difference cordial. S S Proof Let V (DTn ⊙ 2K1) = V (DTn ) {xi , yi : 1 ≤ i ≤ n} {vi′ , vi′′ , wi′ , wi′′ : 1 ≤ i ≤ n − 1} S S and E(DTn ⊙ 2K1 ) = E(DTn ) {ui xi , ui yi : 1 ≤ i ≤ n} {vi vi′ , vi vi′′ , wi wi′ , wi wi′′ : 1 ≤ i ≤ n − 1}. Case 1. n is even. Consider the path vertices ui . Assign the label 1 to the path vertex u1 . Now we assign the labels 1,1,2,2 to the vertices u2 , u3 , u4 , u5 respectively. Then we assign the labels 1,1,2,2 to 124 R.Ponraj and M.Maria Adaickalam the next four vertices u6 , u7 , u8 , u9 respectively. Proceeding like this we assign the label to the next four vertices and so on. If all the vertices are labeled then we stop the process. Otherwise there are some non labeled vertices are exist. If the number of non labeled vertices are less than or equal to 3 then assign the labels 1,1,2 to the non labeled vertices. If it is two then assign the label 1,1 to the non labeled vertices. If only one non labeled vertex is exist then assign the label 1 only. Next we consider the label vi . Assign the label 2 to the vertex v1 . Then we assign the label 2 to the vertices v2 , v4 , v6 , · · · and assign the label 3 to the vertices v3 , v5 , v7 , · · · . Next we move to the vertices xi and yi . Assign the label 2 to the vertices x1 and x2 and we assign the label 3 to the vertices y1 and y2 . Now we assign the label 1 to the vertices x4i+1 and x4i for all the values of i = 1, 2, 3, · · · . Then we assign the label 1 to the vertices x4i+3 for i = 0, 1, 2, 3, · · · . Next we assign the label 2 to the vertices x4i+2 for all the values of i = 1, 2, 3, · · · . Now we assign the label 3 to the vertices y4i+3 for i=0,1,2,3,... For all the values of i = 1, 2, 3, · · · . assign the label 3 to the vertices y4i+1 and y4i+2 . Then we assign the label 2 to the vertices y4i for i = 1, 2, 3, · · · . Now we consider the vertices vi′ and vi′′ . For all ′ ′ the values of i=1,2,3... assign the label 1 to the vertices v4i+1 , v4i+2 . Assign the label 1 to the ′ ′ vertices v4i for i = 1, 2, 3, · · · . Then we assign the label 2 to the vertices v4i+3 for all the values ′′ ′′ ′′ of i = 0, 1, 2, 3, · · · . Consider the vertices vi . Assign the label 3 to the path vertex v4i+1 , v4i+2 ′′ ′′ for all the values of i = 0, 1, 2, 3, · · · . Next we assign the label 2 to the vertices v4i and v4i+3 for i = 1, 2, 3, · · · . Now we assign the label 3 to the vertices wi (1 ≤ i ≤ n − 1). Next we move to the vertices wi′ and wi′′ . Assign the label 1 to all the vertices of wi′ (1 ≤ i ≤ n − 1) and we assign the label 2 to all the vertices of wi′′ (1 ≤ i ≤ n − 1). Since vf (1) = vf (2) = vf (3) = 3n − 2 and ef (1) = 11n−8 , this labeling is 3-difference cordial labeling. and ef (0) = 11n−10 2 2 Case 2. n is odd. First we consider the path vertices ui . Assign the label 1,1,2,2 to the first four path vertices u1 , u2 , u3 , u4 respectively. Then we assign the labels 1,1,2,2 to the next four vertices u5 , u6 , u7 , u8 respectively. Continuing like this assign the label to the next four vertices and so on. If all the vertices are labeled then we stop the process. Otherwise there are some on labeled vertices are exist. If the number of non labeled vertices are less than or equal to 3 then assign the labels 1,1,2 to the non labeled vertices. If it is 2 assign the labels 1,1 to the non labeled vertices. If only one non labeled vertex exist then assign the label 1 to that vertex. Consider the vertices vi . Assign the label 2 to the vertices v1 , v3 , v5 , · · · and we assign the label 3 to the vertices v2 , v4 , v6 , · · · . Next we move to the vertices wi . Assign the label to the vertices wi (1 ≤ i ≤ n−) as in case 1. Now we consider the vertices xi and yi . Assign the label 2 to the vertices x4i+1 for all the values of i = 0, 1, 2, 3, · · · . For all the values of i=0,1,2,3,... assign the label 1 to the vertices x4i+2 and x4i+3 . Then we assign the label 1 to the vertices x4i for all the values of i = 1, 2, 3, · · · . Next we assign the label 3 to the vertices y4i+1 and y4i+2 for all the values of i = 0, 1, 2, 3, · · · and we assign the label 3 to the vertices y4i for i = 1, 2, 3, · · · . Then we assign the label 2 to the vertices y4i+3 for all values i = 0, 1, 2, 3, · · · . Next we move to the vertices vi′ and vi′′ . For all the values of i = 0, 1, 2, 3, · · · assign the label 1 to the vertices ′ ′ ′ v4i+1 and v4i+3 . Now we assign the label 1 to the vertices v4i for i = 1, 2, 3, · · · . Next we assign ′ the label 2 to the vertices v4i+2 for i = 01, 2, 3, · · · . Consider the vertices vi′ . Assign the label 3-Difference Cordial Labeling of Corona Related Graphs 125 ′′ ′′ 3 to the vertices v4i+1 and v4i+2 for all the values of i = 0, 1, 2, 3, · · · and we assign the label 1 to the vertices v4i for i = 1, 2, 3, · · · . For the values of i = 0, 1, 2, 3, · · · assign the label 2 to the vertices v4i+3 . Finally we consider the vertices wi′ and wi′′ . Assign the label to the vertices wi′ (1 ≤ i ≤ n − 1) and wi′′ (1 ≤ i ≤ n − 1) as in case 1. The vertex and edge condition are vf (1) = vf (2) = vf (3) = 3n − 2 and ef (0) = ef (1) = 11n−9 . 2 2 We now investigate the graph DTn ⊙ K2 . Theorem 2.3 DTn ⊙ K2 is 3-difference cordial. S S Proof Let V (DTn ⊙ K2 ) = V (DTn ) {xi , yi : 1 ≤ i ≤ n} {vi′ , vi′′ , wi′ , wi′′ : 1 ≤ i ≤ n − 1} S S and E(DTn ⊙K2 ) = E(DTn ) {ui xi , ui yi , xi yi : 1 ≤ i ≤ n} {vi vi′ , vi vi′′ , vi′ vi′′ , wi wi′ , wi wi′′ , wi′ wi′′ : 1 ≤ i ≤ n − 1}. Case 1. n is even. Consider the path vertices ui . Assign the label 1 to the path vertices u1 , u2 , u3 , · · · . Then we assign the labels 2 to the vertices v1 , v2 , v3 , · · · . Next we assign the labels 3 to the vertices w1 , w2 , w3 , w4 . Now we consider the vertices vi′ and vi′′ . Assign the label 2 to the vertex v1′ . Then we assign the label 1 to the vertices v2′ , v3′ , v4′ , v6′ , · · · . Now we assign the label 3 to the vertices v1′′ , v2′′ , v3′′ , v4′′ , · · · . Next we move to the vertices wi′ and wi′′ . Assign the label 1 to the vertex w1′ . Then we assign the label 1 to the vertices w2′ , w4′ , w6′ , · · · and assign the label 2 to the vertices w3′ , w5′ , w7′ , · · · . Assign the label 2 to the vertices w1′′ , w2′′ , w3′′ , w4′′ , · · · . Finally we move to the vertices xi and yi . Assign the label 1 to the vertices x1 , x3 , x5 , · · · and we assign the label 2 to the vertices x2 , x4 , x6 , · · · then we assign the label 3 to the vertices y1 , y2 , y3 , · · · . Clearly in this case the vertex and edge condition is given in vf (1) = vf (2) = vf (3) = 3n − 2 and ef (0) = 7n − 5 and ef (1) = 7n − 6. Case 2. n is odd. Assign the label to the vertices ui (1 ≤ i ≤ n), vi (1 ≤ i ≤ n − 1) and wi (1 ≤ i ≤ n − 1) as in case 1. Consider the vertices vi′ and vi′′ . Assign the label 1 to the vertices v1′ , v2′ , v3′ , v4′ , · · · . Then assign the label to the vertices vi′′ (1 ≤ i ≤ n − 1) as in case 1. Now we move to the vertices wi′ and wi′′ . Assign the label 1 to the vertices w1′ , w3′ , w5′ , · · · and we assign the label 3 to the vertices w2′ , w4′ , w6′ , · · · . Next we assign the label to the vertices wi′′ (1 ≤ i ≤ n − 1) as in case 1. Now we consider the vertices xi and yi . Assign the label 2 to the vertices x1 , x3 , x5 , · · · and we assign the label 1 to the vertices x2 , x4 , x6 , · · · . Then we assign the label to the vertices yi (1 ≤ i ≤ n) as in case 1. Since vf (1) = vf (2) = vf (3) = 3n − 2 and ef (0) = 7n − 6 and ef (1) = 7n − 5, this labeling is 3-difference cordial labeling. 2 A double quadrilateral snake DQn consists of two quadrilateral snakes that have a common S path. Let V (DQn ) = {ui : 1 ≤ i ≤ n} {vi , wi , xi, yi : 1 ≤ i ≤ n − 1} and E(DQn ) = {ui ui+1 , vi wi , xi yi , wi ui+1 , yi ui+1 : 1 ≤ i ≤ n − 1}. Now we investigate the graphs DQn ⊙ K1 ,DQn ⊙ 2K1 and DQn ⊙ K2 . Theorem 2.4 DQn ⊙ K1 is 3-difference cordial. 126 R.Ponraj and M.Maria Adaickalam S S Proof Let V (DQn ⊙ K1 ) = V (DQn ) {u′i : 1 ≤ i ≤ n} {vi′ , wi′ , x′i , yi′ : 1 ≤ i ≤ n − 1} S S and E(DQn ⊙ K1 ) = E(DQn ) {ui u′i : 1 ≤ i ≤ n} {ui vi′ , wi wi′ , xi x′i , yi , yi′ : 1 ≤ i ≤ n − 1}. Assign the label 1 to the path vertex u1 . Next we assign the labels 1,1,2 to the vertices u2 , u3 , u4 respectively. Then we assign the labels 1,1,2 to the next three path vertices u5 , u6 , u7 respectively. Proceeding like this we assign the label to the next three vertices and so on. If all the vertices are labeled then we stop the process. Otherwise there are some non labeled vertices are exist. If the number of non labeled vertices are less than or equal to 2 then assign the labels 1,1 to the non labeled vertices. If only one non labeled vertex exist then assign the label 1 only. Now we consider the vertices vi and wi . Assign the label 2 to the vertices v3i+1 and v3i+2 for all the values of i=0,1,2,3... For all the vales of i = 1, 2, 3, · · · assign the label 1 to the vertices v3i . Then we assign the label 3 to the vertices wi (1 ≤ i ≤ n). Next we move to the vertices xi and yi . Assign the labels 2,3 to the vertices x1 and y1 respectively. Then we assign the label 2 to the vertices x2 , x5 , x8 , · · · . Now we assign the label 1 to the vertices x3 , x6 , x9 , · · · and the vertices x4 , x7 , x10 , · · · . Assign the label 3 to the vertices y1 , y2 , y3 , · · · . We consider the vertices u′i . Assign the labels 2,3 to the vertices u′1 and u′2 respectively. Now we assign the label 1 to the vertices u′3 , u′6 , u′9 , · · · and we assign the label 3 to the vertices u′4 , u′7 , u′10 , · · · . Then we assign the label 2 to the vertices u′5 , u′8 , u′11 , · · · . Next we move to the vertices vi′ and wi′ . Assign the the label 3 to the vertex w1′ . Now assign the label 1 to all the vertices of vi′ (1 ≤ i ≤ n − 1) and we assign the label 2 to the vertices w2′ , w3′ , w4′ , · · · . We consider the vertices x′i and yi′ . Assign the label 2,1 to the vertices x′1 and y1′ respectively. Also we assign the label 2 to the vertices x′2 , x′5 , x′8 , · · · and the vertices x′3 , x′6 , x′9 , · · · . Then we assign the label 1 to the vertices x′4 , x′7 , x′10 , · · · . Next we assign the label 3 to the vertices y2′ , y3′ , y4′ , ... The vertex condition is ef (0) = 6n − 5 and ef (1) = 6n − 6. Also the edge condition is given in Table 1 following. 2 Nature of n n ≡ 0 (mod 3) n ≡ 1 (mod 3) n ≡ 2 (mod 3) vf (1) 10n−9 3 10n−10 3 10n−8 3 vf (2) 10n−6 3 10n−7 3 10n−8 3 vf (3) 10n−9 3 10n−7 3 10n−8 3 Table 1 Theorem 2.5 DQn ⊙ 2K1 is 3-difference cordial. S S Proof Let V (DQn ⊙ 2K1 ) = V (DQn ) {u′i , u′′i : 1 ≤ i ≤ n} {vi′ , vi′′ , wi′ , wi′′ , x′i , x′′i , yi′ , yi′′ : S S 1 ≤ i ≤ n−1} and E(DQn ⊙2K1 ) = E(DQn ) {ui u′i , ui u′′i : 1 ≤ i ≤ n} {vi vi′ , vi vi′′ , wi wi′ , wi wi′′ , xi x′i , xi x′′i , yi yi′′ , yi yi′′ : 1 ≤ i ≤ n − 1}. First we consider the path vertices ui . Assign the label 1 to the path vertices u1 , u3 , u5 , · · · and we assign the label 2 to the path vertices u2 , u4 , u6 , · · · . Clearly the last vertex un received the label 2 or 1 according as n ≡ 0mod 2 or n ≡ 1 (mod 2). Next we move to the vertices vi and wi . Assign the label 1 to all the vertices of vi (1 ≤ i ≤ n) and we assign the label 3 to the vertices w1 , w2 , w3 , ... Then we assign the label to the vertices xi (1 ≤ i ≤ n − 1) is same as assign the label to the vertices vi (1 ≤ i ≤ n − 1) and we assign the label to the vertices yi (1 ≤ i ≤ n − 1) is same as assign the label to the vertices wi (1 ≤ i ≤ n − 1). Next we move to the vertices u′i and vi′′ . Assign the label 2 to the vertices 3-Difference Cordial Labeling of Corona Related Graphs 127 ′′ v1′ , v2′ , v3′ , · · · then we assign the label 3 to the vertex v1′′ . Assign the label 3 to the vertices v2i for ′′ all the values of i = 1, 2, 3, · · · and we assign the label 2 to the vertices v2i+1 for i = 1, 2, 3, · · · . Next we consider the vertices wi′ and wi′′ . Assign the label 1 to the vertices w1′ , w2′ , w3′ , · · · and we assign the label 3 to the vertices w1′′ , w2′′ , w3′′ , · · · . Next we move to the vertices x′i and x′′i . Assign the label 1 to all the vertices of x′i (1 ≤ i ≤ n − 1) and we assign the label 2 to all the vertices of x′′i (1 ≤ i ≤ n − 1). Now we assign the label 2 to the vertices y1′ , y2′ , y3′ , · · · and we assign the label 3 to the vertices y1′′ , y2′′ , y3′′ , · · · . Finally we move to the vertices u′i and u′′i . Assign the label 2 to the vertices u′1 , u′3 , u′5 , · · · and we assign the label 1 to the vertices u′2 , u′4 , u′6 , · · · . Next we assign the label 2 to the vertices u′′1 , u′′3 , u′′5 , · · · and we assign the label 3 to the vertices u′′2 , u′′4 , u′′6 , · · · . The vertex condition is vf (1) = vf (2) = vf (3) = 15n−12 . Also 3 2 the edge condition is given in Table 2. Values of n n ≡ 0 (mod 2) n ≡ 1 (mod 2) ef (0) 17n−16 2 17n−15 2 ef (1) 17n−14 2 17n−15 2 Table 2 Theorem 2.6 DQn ⊙ K2 is 3-difference cordial. S S Proof Let V (DQn ⊙ K2 ) = V (DQn ) {u′i , u′′i : 1 ≤ i ≤ n} {vi′ , vi′′ , wi′ , wi′′ , x′i , x′′i , yi′ , yi′′ : S S 1 ≤ i ≤ n − 1} and E(DQn ⊙ K2 ) = E(DQn ) {ui u′i , ui u′′i , u′i u′′i : 1 ≤ i ≤ n} {vi vi′ , vi vi′′ , vi′ vi′′ , wi wi′ , wi wi′′ , wi′ wi′′ , xi x′i , xi x′′i , x′i x′′i , yi yi′′ , yi yi′′ , yi′ yi′′ : 1 ≤ i ≤ n − 1}. First we consider the path vertices ui . Assign the label 1 to the vertex u1 . Then we assign the label 1 to the vertices u2 , u4 , u6 , · · · and we assign the label 2 to the path vertices u1 , u3 , u5 , · · · . Note that in this case the last vertex un received the label 1or 2 according as n ≡ 0 (mod 2) or n ≡ 1 (mod 2). Next we move to the vertices vi and wi . Assign the label 2 to the vertex v1 . Then we assign the label 3 to all the vertices of wi (1 ≤ i ≤ n − 1). Assign the label 1 to the vertices v2 , v3 , v4 , ... We consider the vertices xi and yi . Assign the label to the vertices xi (1 ≤ i ≤ n − 1) is same as assign the label to the vertices vi (1 ≤ i ≤ n − 1) and assign the label to the vertices yi (1 ≤ i ≤ n − 1) is same as assign the label to the vertices wi (1 ≤ i ≤ n − 1). Next we move to the vertices vi′ and vi′′ . Assign the label 2 to the vertices v1′ , v2′ , v3′ , · · · and assign the label 3 to the vertices v1′′ , v2′′ , v3′′ , · · · . Consider the vertices x′i and x′′i . Assign the label 1 to all the vertices of x′i (1 ≤ i ≤ n − 1). Assign the label 2 to the vertex x′′1 . Then we assign the label 3 to the vertices x′′2 , x′′3 , x′′4 , · · · . Now we assign the label 2 to all the vertices of wi′ (1 ≤ i ≤ n − 1) and assign the label 3 to all the vertices of wi′′ (1 ≤ i ≤ n − 1). Now we move to the vertices yi′ and yi′′ . Assign the label 1 to the vertices y1′ , y2′ , y3′ , · · · and we assign the label 2 to the vertices y1′′ , y2′′ , y3′′ , · · · . Next we move to the vertices u′i and u′′i . Assign the label 1,3 to the vertices u′1 and u′′1 respectively. Assign the label 1 to the vertices u′2i for all the values of i = 1, 2, 3, · · · and assign the label 2 to the vertices u2i+1 for i = 1, 2, 3, · · · then we assign the label 2 to the vertices u′′2 , u′′3 , u′′4 , · · · . The vertex and edge conditions are vf (1) = vf (2) = vf (3) = 15n − 12 3 128 R.Ponraj and M.Maria Adaickalam and ef (0) = 11n − 9, ef (1) = 11n − 10. 2 References [1] J.A.Gallian, A Dynamic survey of graph labeling, The Electronic Journal of Combinatorics, 19 (2016) #Ds6. [2] F.Harary, Graph Theory, Addision wesley, New Delhi (1969). [3] R.Ponraj, M.Maria Adaickalam and R.Kala, k-difference cordial labeling of graphs, International Journal of Mathematical Combinatorics, 2(2016), 121-131. [4] R.Ponraj, M.Maria Adaickalam, 3-difference cordial labeling of some union of graphs, Palestine Journal of Mathematics, 6(1)(2017), 202-210. [5] R.Ponraj, M.Maria Adaickalam, 3-difference cordial labeling of cycle related graphs, Journal of Algorithms and Computation, 47(2016), 1-10. [6] R.Ponraj, M.Maria Adaickalam, 3-difference cordiality of some graphs, Palestine Journal of Mathematics, 2(2017), 141-148. [7] R.Ponraj, M.Maria Adaickalam, R.Kala, 3-difference cordiality of corona of double alternate snake graphs, Bulletin of the International Mathematical Virtual Institute , 8(2018), 245-258. International J.Math. Combin. Vol.2(2018), 129-133 Graph Operations on Zero-Divisor Graph of Posets N.Hosseinzadeh (Department of Mathematics, Islamic Azad University, Dezful Branch, Dezful, Iran) E-mail: narges.hosseinzadeh@gmail.com Abstract: We know that some large graphs can be constructed from some smaller graphs by using graphs operations. Many properties of such large graphs are closely related to those of the corresponding smaller ones. In this paper we investigate some operations of zero-divisor graph of posets. Key Words: Poset, zero-divisor graph, graph operation. AMS(2010): 06A11,05C25. §1. Introduction In [2], Beck, for the first time, studied zero-divisor graphs of the commutative rings. Later, D. F. Anderson and Livingston investigated nonzero zero-divisor graphs of the rings (see [1]). Some researchers also studied the zero-divisor graph of the commutative rings. Subsequently, others extended the study to the commutative semigroups with zero. These can be seen in [3, 5, 7, 8]. Assume (P, ≤) is a poset (i.e., P is a partially ordered set) with the least element 0. For every x, y ∈ P , defined of L(x, y) = {z ∈ P |z ≤ x and z ≤ y}. x is a zero-divisor element of P if l(x, y) = 0, for some 0 6= y ∈ P . Γ(P ) is the zero-divisor graph of poset P , where the its vertex set consists of nonzero zero-divisors elements of P and x is adjacent to y if only if L(x, y) = {0}. In this paper, P denotes a poset with the least element 0 and Z(P ) is nonzero zero-divisor elements of P . The zero-divisor graph is undirected graph with vertices Z(S) such that for every distinct x, y ∈ Z(S), x and y are adjacent if only if L(x, y) = {0}. Throughout this paper, G always denotes a zero-divisor graph which is a simple graph (i.e., undirected graph without loops and multiple) and the set vertices of G show V (G) and the set edges of G denotes E(G). The degree of vertex x is the number of edges of G intersecting x. N (x), which is the set of vertices adjacent to vertex x, is called the neighborhood of vertex x. If n is a (finite or infinite) natural number, then an n-partite graph is a graph, which is a set of vertices that can be partitioned into subsets, each of which edges connects vertices of two different sets. A complete n- partite graph is a n- partite graph such that every vertex is adjacent to the vertices which are in a different part. A graph H is a subgraph of G if V (H) ⊆ V (G) and E(H) ⊆ E(G). H is called an induced subgraph of G if for every x, y ∈ V (H), {x, y} ∈ E(G). A subgraph H of G is called a clique if H is a complete graph. The clique number ω(G) of G is the least upper 1 Received November 17, 2017, Accepted May 30, 2018. 130 N.Hosseinzadeh bound of the cliques sizes of G. Many large graphs can be constructed by expanding small graphs, thus it is important to know which properties of small graphs can be transfered to the expanded ones, for example Wang in [6] proved that the lexicographof vertex transitive graphs is also vertex transitive as well as the lexicographic product of edge transitive graphs. Specapan in [9] found the fewest number of vertices for Cartesian product of two graphs whose removal from the graph results in a disconnected or trivial graph. Motivated by these, we consider five kinds of graph products as the expander graphs which is described below and we can verify if regard the product of them can be regarded as a Cayley graph of the semigroup which is made by their product underlying semigroup and if the answer is positive does it inherit Col-Aut-vertex property of from the precedents. Let Γ = (V, E) be a simple graph, where V is the set of vertices and E is the set of edges of G. An edge joins the vertex u to the vertex v is denoted by (u, v). In [10], the authors described the following definition: Definition 1.1 A simple graph G is called a compact graph if G does not contain isolated vertices and for each pair x and y of non-adjacent vertices of G, there exists a vertex z with N (x) ∪ N (y) ⊆ N (z). Definition 1.2 Let Γ1 = (V1 , E1 ) and Γ2 = (V2 , E2 ) be two graphs. Γ = (V, E), the product of them is a graph with vertex set V = V1 × V2 , and two vertices (u1 , u2 ) is adjacent to (v1 , v2 ) in Γ if one of the relevant conditions happen depending on the product. (1) Cartesian product. u1 is adjacent to v1 in Γ1 and u2 = v2 or u1 = v1 and u2 is adjacent to v2 in Γ2 ; (2) Tensor product. u1 is adjacent to v1 in Γ1 and u2 is adjacent to v2 in Γ2 ; (3) Strong product. u1 is adjacent to v1 in Γ1 and u2 = v2 or u1 = v1 and u2 is adjacent to v2 in Γ2 or u1 is adjacent to v1 in Γ1 and u2 is adjacent to v2 in Γ2 ; (4) Lexicographic. u1 is adjacent to v1 in Γ1 or u1 = v1 and u2 is adjacent to v2 in Γ2 ; (5) Co-normal product. u1 is adjacent to v1 in Γ1 or u2 is adjacent to v2 in Γ2 ; (6) Modular product. u1 is adjacent to v1 in Γ1 and u2 is adjacent to v2 in Γ2 or u1 is not adjacent to v1 in Γ1 and u2 is not also adjacent to v2 in Γ2 . §2. Preliminary Notes In this section, we recall some lemmas and definitions. from Dancheny Lu and Tongsue We in [10], the authors described following definition. Definition 2.1 A simple graph G is called a compact graph if G does not contain isolated vertices and for each pair x and y of non-adjacent vertices of G, there exists a vertex z with N (x) ∪ N (y) ⊆ N (z). It has been showed the following theorem in [10]. Theorem 2.2 A simple graph G is the zero-divisor graph of a poset if and only if G is a Graph Operations on Zero-Divisor Graph of Posets 131 compact graph. §3. Cartesian Product Through this section, we assume that P and Q are two posets with the least element 0. Assume G and H are in zero-divisors graphs of P and Q, respectively. N (x) and N (a) are neighborhoods in G and H , respectively, where x ∈ V (P ) and a ∈ V (Q). Theorem 3.1 Let Γ be the Cartesian product of two zero-divisor graph of G and H. Then N (x, a) = (N (x) × {a}) ∪ ({x} × N (a)), for any (x, r) ∈ V (G × H). Proof Let (s, r) ∈ N (x, a). Therefore, (s, r) is adjacent to (x, a). Thus, s is adjacent to x in G and r = a or s = x and r is adjacent to a in H. Hence, s ∈ N (x) and r = a or s = x and r ∈ N (a). It can be concluded that to N (x, a) = (N (x) × {a}) ∪ ({x} × N (a)). 2 Theorem 3.2 Let G and H be two compact graphs. Then Γ the cartesian product of them is not a compact graph. Proof Let (x, a) and (y, b) be two arbitrary vertices not being adjacent of the graph Γ, where (x, a) 6= (y, b). Therefore, x and y are not adjacent in G or a 6= b in H and x 6= y in G or a, b are not adjacent in H. Assume that there exists (z, c) ∈ V (Γ) such that N (x, a)∪N (y, b) ⊆ N (z, c). That is, (N (x) × {a}) ∪ ({x} × N (a)) ∪ (N (y) × {b}) ∪ ({y} × N (b)) ⊆ N (z) × {c}) ∪ ({z} × N (c)). Assume that (m, a), (n, a) ∈ (N (x) × {a}) such that (m, a) ∈ N (z) × {c} and (n, a) ∈ {z} × N (c). Then, m ∈ N (z), a = c, a ∈ N (c). Hence, ac = 0 and c2 = 0. That is a contradiction. Therefore, N (x) × {a} has intersection only one of N (z) × {c} and {z} × N (c). Similary, we get this subject for ({x} × N (a)), (N (y) × {b}) and ({y} × N (b)). Now, suppose N (x) × {a} ⊆ N (z) × {c}( i.e., a = c, N (x) ⊆ N (z)). If {x} × N (a) ⊆ N (z) × {c}, we have N (a) = {c}. Hence, ac = 0. On the other hand a = c, then c2 = 0. That is a contradiction. Therefore, {x} × N (a) ⊆ {z} × N (c), that is x = z and N (a) ⊆ N (c). Then, N (x) = N (z) and N (a) = N (c). Suppose N (y) × {b, } ⊆ N (z) × {c}. Thus, N (y) ⊆ N (z) = N (x), b = c. Hence, a = b = c, N (a) = N (b) = N (c). If {y} × N (b) ⊆ N (z) × {c}, y ∈ N (z) and N (b) = c. Then, bc = c2 = 0. That is a contract. Therefore, {y} × N (b) ⊆ {c} × N (z). We get y = z and N (b) ⊆ N (c). It leads to a = b = c and 2 x = y = z. That is a contradiction. Corollary 3.2 Let G and H be two compact graphs of two poset. Then, the cartesian product of them is not a graph of a poset. Proof Referring to the theorem above and [10], it is clear. 2 132 N.Hosseinzadeh §4. Tensor Product Through this section, we assume that G and H are two zero-divisor graphs of poset P and Q with the least element 0, respectively. Theorem 4.1 Γ is the tensor product of the graphs G and H. Then N (x, a) = (N (x) × N (a)) for any (x, a) ∈ V (G × H). Proof Assume (s, r) ∈ N (x, a). Then, (s, r) is adjacent to (x, a). By Definition 1.2, s and x are adjacent and r and a are adjacent too. Therefore, s ∈ N (x) and r ∈ N (a). It leads to N (x, a) = N (x) × N (a). 2 §5. Strong Product Through this section, we assume that H and K are two zero-divisor graphs of poset P and Q with the least element 0, respectively. By Definition 1.2, Theorems 3.1 and 4.1, we conclude the following theorems. Theorem 5.1 Γ the strong product of two zero-divisors graphs G and H of posets. Then, if runs for any (x, a) ∈ V (Γ), N (x, a) = (N (x) × {a}) ∪ ({x} × N (a)) ∪ (N (x) × N (a)) Proof By Definition 1.2, for any (r, s) ∈ N (x, a), where (x, a) ∈ V (Γ), r is adjacent to x inG and s = a or r = x and s is adjacent to a in H or r is adjacent to x in G and s is adjacent to aiH. Therefore, N (x, a) = (N (x) × {a}) ∪ ({x} × N (a)) ∪ (N (x) × N (a)). 2 §6. Co-normal Product Theorem 6.1 Γ is the co-normal product of two graphs G and H of two the posets of P and Q, respectively. Then for any (x, a) ∈ V (Γ), N (x, a) = (N (x) × V (H)) ∪ (V (H) × N (a)). Proof By Definition 1.2, if ((s, r) is adjacent to (x, a), s and x are adjacent in G or r, a are 2 adjacent in H. Thus, N (x, a) = (N (x) × V (H)) ∪ (V (H) × N (a)). Theorem 6.2 If Γ is the co-normal product of two compact graphs G and H, then Γ is a compact graph. Proof Let (x, a) and (y, b) not be in Γ and (x, a) 6= (y, b). By referring the virtue of Definition 1.2, we get x andy are not adjacent in G and a and b are not adjacent in H. Then there exist z ∈ G and s ∈ H such that N (x) ∪ N (y) ⊆ N (z) and N (a) ∪ N (b) ⊆ N (c). Hence, N (x, a) ∪ N (y, b) = ⊆ (N (x) × N (a)) ∪ (N (y) × N (b) (N (z) × N (c)) ∪ (N (z) × N (c)) = N (z) × N (c) Now, we get the following corollary. 2 Graph Operations on Zero-Divisor Graph of Posets 133 Corollary 6.3 The co-product of two zero-divisor graphs of posets is a zero-divisor graph of a poset. Proof By the above theorem and [10], it is clear. 2 §7. Lexicographic Product Theorem 7.1 Γ the lexicographic product of two zero-divisor graphs G and H of the two posets P and Q, respectively. Then, N (x, a) = (N (x) × V (H)) ∪ ({x} × N (a)), for any (x, a) ∈ V (Γ). Proof By Definition 1.2, assume (s, r) ∈ N (x, a). Therefore, s and x are adjacent in G or s = x in G and r and a are adjacent in H. Therefore, N (x, a) = (N (x) × V (H)) ∪ ({x} × N (a)), for any (x, a) ∈ V (γ). 2 §8. Modular Product Theorem 8.1 Γ the Modular product of two zero-divisor graphs G and H of the two posets P and Q respectively. Then, N (x, a) = (N (x) × N (a)) ∪ (N c (x) × N c (a)), for any (x, a) ∈ V (Γ). Proof By Definition 1.2, assume (s, r) ∈ N (x, a). Therefore, s and x are adjacent in G while r and a are adjacent in H or s and x are not adjacent in G whereas r and a are not adjacent in H. Thus, N (x, a) = (N (x) × N (a)) ∪ (N c (x) × N c (a)). 2 References [1] D. F. Anderson, P. S. Livingston,The zero-Divisor Graph of a Commutative Ring, J. Algebra, 159 (1991), 500-514. [2] I. Beck, Coloring of Commutative Rings, J. Algebra, 116 (1988), 208-226. [3] F. R. DeMeyer, L. DeMeyer, Zero-Divisor Graphs of Semigroups, J. Algebra, 283 (2005), 190-198. [4] Dancheny Lu and Tongsue We, The zero-divisor graphs of posets and application to semigroups, Graphs and Combinatorics, (2010) 26, 793-804. [5] F. R. DeMeyer, T. McKenzie and K. Schneider, The Zero-Divisor Graphs of a Commutative Semigroup, Semigroup Forum, 65 (2002), 206-214. [6] F. Li, W. Wang, Z. Xu and H. Zhao, Some results on the lexicographic product of vertextransitive graphs, Applied Math. Letters, 24(2011), 1924-1926. [7] S. E. Wright, Lengths of paths and cyclies in zero-divisor graphs and digraphs of semigroups, Comm. Algebra, 35 (2007), 1987-1991. [8] T. Wu and F. Cheng, The structure of zero-divisor semigroup with graph Kn ◦ K2 , Semigroup Forum, 76(2008), 330-340. [9] S. Spacapan, Connectivity of Cartesian product of graphs, Applied Mathematica Letters, Vol. 21, Issue 7, (2008), 682-685. [10] Dancheny Lu and Tongsue We, The zero-divisor graphs of posets and application to semigroups, Graphs and Combinatorics, (2010) 26, 793-804. 134 International Journal of Mathematical Combinatorics The tragedy of the world is that those who are imaginative have but slight experience, and those who are experienced have feeble imaginations. By Alfred North Whitehead, A British philosopher and mathematician Author Information Submission: Papers only in electronic form are considered for possible publication. Papers prepared in formats, viz., .tex, .dvi, .pdf, or.ps may be submitted electronically to one member of the Editorial Board for consideration in the International Journal of Mathematical Combinatorics (ISSN 1937-1055). An effort is made to publish a paper duly recommended by a referee within a period of 3 − 4 months. Articles received are immediately put the referees/members of the Editorial Board for their opinion who generally pass on the same in six week’s time or less. In case of clear recommendation for publication, the paper is accommodated in an issue to appear next. Each submitted paper is not returned, hence we advise the authors to keep a copy of their submitted papers for further processing. Abstract: Authors are requested to provide an abstract of not more than 250 words, latest Mathematics Subject Classification of the American Mathematical Society, Keywords and phrases. Statements of Lemmas, Propositions and Theorems should be set in italics and references should be arranged in alphabetical order by the surname of the first author in the following style: Books [4]Linfan Mao, Combinatorial Geometry with Applications to Field Theory, InfoQuest Press, 2009. [12]W.S.Massey, Algebraic topology: an introduction, Springer-Verlag, New York 1977. Research papers [6]Linfan Mao, Mathematics on non-mathematics - A combinatorial contribution, International J.Math. Combin., Vol.3(2014), 1-34. [9]Kavita Srivastava, On singular H-closed extensions, Proc. Amer. Math. Soc. (to appear). Figures: Figures should be drawn by TEXCAD in text directly, or as EPS file. In addition, all figures and tables should be numbered and the appropriate space reserved in the text, with the insertion point clearly indicated. Copyright: It is assumed that the submitted manuscript has not been published and will not be simultaneously submitted or published elsewhere. By submitting a manuscript, the authors agree that the copyright for their articles is transferred to the publisher, if and when, the paper is accepted for publication. The publisher cannot take the responsibility of any loss of manuscript. Therefore, authors are requested to maintain a copy at their end. Proofs: One set of galley proofs of a paper will be sent to the author submitting the paper, unless requested otherwise, without the original manuscript, for corrections after the paper is accepted for publication on the basis of the recommendation of referees. Corrections should be restricted to typesetting errors. Authors are advised to check their proofs very carefully before return. June 2018 Contents Ricci Soliton and Conformal Ricci Soliton in Lorentzian β-Kenmotsu Manifold By Tamalika Dutta and Arindam Bhattacharyya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01 Some Properties of Conformal β -Change By H.S.Shukla and Neelam Mishra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Equitable Coloring on Triple Star Graph Families By K.Praveena and M.Venkatachalam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 On the Tangent Vector Fields of Striction Curves Along the Involute and Bertrandian Frenet Ruled Surfaces By Şeyda Kılıçoğlu, Süleyman Şenyurt and Abdussamet Çalışkan . . . . . . . . . . . . . . . . . . . . . . . . 33 On the Leap Zagreb Indices of Generalized xyz-Point-Line Transformation Graphs T xyz (G) when z = 1 By B. Basavanagoud and Chitra E. . . . . . . . . . . . . . . . . . . . . . . 44 A Generalization on Product Degree Distance of Strong Product of Graphs By K.Pattabiraman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Semifull Line (Block) Signed Graphs By V. Lokesha, P. S. Hemavathi and S. Vijay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Accurate Independent Domination in Graphs By B. Basavanagoud and Sujata Timmanaikar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 On r-Dynamic Coloring of the Triple Star Graph Families By T.Deepa and M. Venkatachalam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 (1,N)-Arithmetic Labelling of Ladder and Subdivision of Ladder By V.Ramachandran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 3-Difference Cordial Labeling of Corona Related Graphs By R.Ponraj and M.Maria Adaickalam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Graph Operations on Zero-Divisor Graph of Posets By N.Hosseinzadeh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 An International Journal on Mathematical Combinatorics

RELATED PAPERS

RELATED TOPICS

Log In

International Journal of Mathematical Combinatorics, vol. 2/2018

International Journal of Mathematical Combinatorics, vol. 2/2018

Related Papers

RELATED PAPERS

RELATED TOPICS