Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Dynamic Analysis of The Stress and Strain State of The Spur Gear Pair

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

D.DIMITRIJEVIĆ, V.NIKOLIĆ, I.ATANASOVSKA.

: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR 20
UDK: 621.83:519.673:539.319
COSATI: 13-09, 20-11

Dynamic Analysis of the Stress and Strain State of the


Spur Gear Pair

Dejan Dimitrijević, PhD (Eng)1)


Vera Nikolić, PhD(Eng)2)
Ivana Atanasovska, PhD (Eng)3)

A spur gear pair dynamic model for the gear dynamic contact loading, dynamic contact stress state and dynamic
contact strain state analysis is presented. A dynamic model of the gear set with two degrees of freedom is used. The
transmission is analyzed using the nonlinear finite elements method where a novel approach for interpreting the
results of the stress and strain state using stress and/or strain tensor invariants is developed. For a more general
approach, the software for the finite element analysis of the gear set as a whole is developed, using the open source
finite elements framework CODE-ASTER/SALOME.

Key words: stress and strain state, dynamic load, stress and strain state analysis, spur gears, deformation, stress and
strain tensor, finite element method.

Introduction Finite element model of the spur gear pair

S TARTING with one gear pair, a meshing action is


observed using no additional assumptions concerning
the gear geometry or the meshing action boundary and load
The gears analyzed in this paper are the ones from the
deep drilling machine gear set [2,3,4]. The gear geometries
are generated using the software developed by the author [9,
10,11]
conditions. A numeric experiment is developed simply by and adapted for the SALOME platform. The obtained
using the boundary conditions on the gear-shaft connection, geometries are transferred to SALOME mesh generation
which are assumed to equal zero. The software developed software and discretized using four node tetrahedral
for the purpose of dynamic analysis of the gear-boxes can elements there (Fig.1).
be used to obtain the following results from such a model:
a) forces that arise in gear contact during the gear meshing
action as a result of the incremental quasi-static series of
analysis, for each meshing action time step,
b) the comprehensive modal analysis for the gear pair, and
the
c) forces obtained in a) will be used further as an input for
the nonlinear dynamic analysis. For the purpose of this
paper, it is assumed that the obtained force is a harmonic
function and in the form:
Fn = F0 sin ωt (1)
where F0 is the resultant force for the observed meshing
action time step.
d) Finally, the dynamic analysis of the gear meshing action Figure 1. Gear set model
is done[1]. The contact load as inner dynamic load is
The gear geometry parameters are given in Table 1. The
taken as an input value obtained in a) and approximated
mesh parameters of the geared system are:
using the harmonic function (as it is described in c)).
The developed software can be modified for the purpose − Number of nodes: 29777
of calculation of different types of geared systems. It is − Number of elements: 160491
optional which of the mentioned steps will be of the highest − Number of elements of gear 1 (smaller gear): 33370
priority, and depends, of course, on the purpose of the − Number of elements of gear 2: 81037
concrete analysis. That is not of interest here, and therefore The displacement boundary conditions (Fig.2) are set on
will not be discussed. the gear-shaft connections nodes. The displacement

1)
Sophien str. 93, 91052 Erlangen, GERMANY
2)
Mechanical Engineering Faculty, Sestre Janić 6, 34000 Kragujevac, SERBIA
3)
Institut IMS a.d., Bulevar Vojvode Mišića 43, 11000 Belgrade, SERBIA
D.DIMITRIJEVIĆ, V.NIKOLIĆ, I. ATANASOVSKA.: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR 21

boundary conditions are set to be zero in all degrees of stiffness/ dumping functions. The solution for the
freedom. differential equations given in (2) is not possible without
The load boundary conditions are generated on the explicitly defined mesh stiffness and dumping functions,
meshing teeth on the edges that are opposite to the contact because of the fact that the equations are non-linear
surfaces of each tooth. The form of the load function is functions of time and the solution can be obtained only
taken according to (1) with the amplitude obtained from the using numeric methods.
nonlinear static analysis. This analysis helps to determine The differential equation of the model in Fig3 are:
the form of the amplitude force distribution in the gear
x1 + k (t )( x&1 − x&2 ) + c(t )( x1 − x2 ) = Fn (t )
m1&&
mesh. The detailed procedure of how the force is taken
from nonlinear static analysis will not be discussed in detail (2)
here. The force distribution on other transmission elements x2 − k (t )( x&1 − x&2 ) − c(t )( x1 − x2 ) = − Fn (t )
m2 &&
is calculated in the same way.
where:
All other meshing conditions are not taken into
consideration. The following assumptions are made: − m1 (i=1,2) are equivalent masses of transmission ele-
− It is assumed that the contact is without friction ments,
− There is no sliding between the teeth during the meshing − c ( t ) is the gear mesh stiffness function depending on t,
action. − k ( t ) is the gear mesh dumping function depending on t.
− The transmission is functioning in a constant temperature
field. All other thermodynamic effects can be neglected. Shaft stiffness/dumping outside the gears on shaft 2 are
not considered.
− The proportional damping is used according to
C = 2β M + α K
& Stress and strain state analysis using stress and strain
− The continuum discretization is done using the tetrahedral tensor invariants
finite elements with the linear approximation of the finite The calculation of the component stresses and strains is
element geometry.
done by the nonlinear Finite Element method using the
already described procedure.
Table 1. Constructive transmission parameters
Parameter Gear 1 Gear 2
Number of teeth 38 65
Modul (m) 0.003 0.003
Pitch circle diameter (m) 0.114 0.195
Addendum circle diameter (m) 0.120 0.201
Base circle diameter (m) 0.1068 0.1878
Measurement teeth number (m) 5 6
Standard angle (degrees) 20 20
Axial distance (m) 0.129 0.129005

Figure 3. A simplified dynamic model of the gear set in Fig.1

Starting from the known component stresses as the


components of tensor N, it is possible to obtain stress tensor
invariants in every node after the finite elements analysis is
done. Since:
σ x −σ0 τ xy τ xz
f (σ 0 ) = N − σ 0 I = τ xy σ y −σ0 τ yz = 0 (3)
τ xz τ yz σ z −σ0
yields:

−(σ 03 − L1σ 02 + L2σ 0 − L3 ) = 0 (4)


Figure 2. Displacement and force boundary conditions
where
L1 = f (σ 1 , σ 2 , σ 3 ) = σ 1 + σ 2 + σ 3 = tr (G ) , (5)
Analytic model of the spur gear transmission
The gear set model in Fig.1 can be simplified using the
transmission model in Fig.3 for dynamic analysis. This L2 = f (σ 1 , σ 2 , σ 3 ) = σ 1σ 2 + σ 2σ 3 + σ 1σ 3
(6)
class of dynamic models can be used for the analysis of the = 1 ( tr (G 2 ) − tr (G ) 2 )
oscillation parameters of the transmission, only if meshing 2
functions of the stiffness and dumping are known. As this is
not the case, it is necessary to experimentally obtain the L3 = f (σ 1 , σ 2 , σ 3 ) = σ 1σ 2σ 3 = det(G ) , (7)
form of the mesh stiffness of the geared mesh, or to use
other means like finite elements method to obtain unknown and
22 D.DIMITRIJEVIĆ, V.NIKOLIĆ, I.ATANASOVSKA.: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR

⎛ σ1 ⎞ The same discussion can be applied for the strains.


G=⎜ σ2 ⎟, (8) Starting from the secular equation for obtaining main
⎜ σ ⎟ dilatations in the form:
⎝ 3⎠
ε x − ε0 γ xy γ xz
is a principal stress tensor in the stress principal directions
coordinate system at an arbitrary body point. γ xy ε y − ε 0 γ yz = 0 (14)
Using the secular equation in its determinant form (3) or γ xz γ yz ε z − ε0
its algebraic form (4) it is possible to obtain the principal
yields:
stresses as functions of stress tensor invariants L1, L2 i L3,
using relations-expressions (5), (6) and (7) as a system of −(ε 03 − J1ε 02 + J 2ε 0 − J 3 ) = 0 (15)
three nonlinear algebraic equations with unknown principal
stresses. That yields the functional dependency between the where J1, J2, i J3 are known invariants of the strain tensor
principal stress and the stress invariants in the form: obtained from known strain components of the geared
σ 0 = f ( L1 , L2 , L3 ) . transmission in every arbitrary point. The solutions of the
When the component stresses of the tensor are known, it is secular equation (15) are the principal strains given. For the
also possible to obtain three sets of three cosines of the main principal directions, the principal strains build the strain
stress directions. The principal stress directions are directions tensor in the following matrix:
that define the coordinate system of principal directions,
⎛ ε1 ⎞
where there are no shear stresses in the defined planes. S=⎜ ε2 ⎟ (16)
When the principal stresses are different, σ 1 ≠ σ 2 ≠ σ 3 , ⎜ ε ⎟
⎝ 3⎠
the maximum values of the shear stresses are found in three
normal planes, where the principal axis 1 and one of these where the strain tensor invariants are given in the form:
planes are collinear, and the angles between the other
principal planes and the mentioned plane are 45 degrees. J1 = f (ε1 , ε 2 , ε 3 ) = ε1 + ε 2 + ε 3 = tr (S) (17)
The maximum values of the shear stresses are obtained in
the following form: J 2 = f (ε1 , ε 2 , ε 3 ) = ε1ε 2 + ε 2ε 3 + ε1ε 3
σ2 − σ3 σ − σ1 σ −σ2
τ (I) = ± , τ (II) = ± 2 , τ (III) = ± 1 . (9)
2 2 2 = 1 ( tr (S 2 ) − tr (S) 2 ) (18)
There are no shear stresses in the principal planes. On 2
the other hand, in the maximal shear stress planes, there are
not only extreme values of shear stresses, but also normal J 3 = f (ε1 , ε 2 , ε 3 ) = ε1ε 2ε 3 = det(S) (19)
stress components with values given in the form:
Analogous to (11), (12) and (13) for the principal
σ2 + σ3 σ1 + σ 3 σ1 + σ 2 stresses, it is possible to obtain the equation of the principal
σ (I) = , σ (II) = , σ (III) = . (10)
2 2 2 strains by solving the system of equations (17, 18, and 19):
As already stated, the principal directions can be The principal strain directions can be obtained in the
obtained from the component stresses using the following following form:
relations: α0 β0
=
α0 β0 γ0 γ xy γ xz ε x − ε 0 γ xz
= = = C0 (11) −
K 31 K 32 K 33 ε y − ε 0 γ yz γ xy γ yz
or
γ0
α0 β0 = = C0 (20)
= ε x − ε0 γ xy
τ xy τ xz σ x − σ 0 τ xz γ xy ε y − ε0

σ y − σ 0 τ yz τ xy τ yz
if:
γ0 α 02 + β 02 + γ 02 = 1
= = C0 (12) (21)
σ x −σ0 τ xy
τ xy σ y −σ0 Depending on the value of the principal invariants J1, J2
and J3, the properties of the principal strains (relative
the system is defined if condition (13) is satisfied: deformations) is every body line element at every point are
α 02 + β 02 + γ 02 = 1 (13) for example:
a) if J3 > 0, the roots (principal strains) of (7) are positive
The relations (11) and (12) should be taken for s =1,2,3. and different and they are extensions,
They give the equations for three main stress directions b) if J3 < 0, there are negative roots (principal strains), are
represented with three direction cosines, with total of nine copressions and
direction cosines. c) if J3 = 0, on of the roots (principal strains) equals zero.
Depending on the value of the principal invariants, the
properties of the main stresses are defined:
a) if L3 > 0, the roots (principal stresses) given in (8) are Results
positive and different. Using the above procedures, the results are obtained for
b) if L3 < 0, there are negative roots (principal stresses), and the gear set in Fig.1, using the boundary conditions
c) if L3 = 0, one of the roots (principal stresses) is equal to presented in Fig.2 using the gear set parameters given in
zero. Table 1.
D.DIMITRIJEVIĆ, V.NIKOLIĆ, I. ATANASOVSKA.: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR 23

Figure 4. Gear set load distribution in the time step t = 0.02

The gear set load distribution along the line of contact


for the gear pair given in Fig.4, and distribution of loads on
the transmission elements for t = 0.02 is given in Fig.5 as Figure 7. Stress distribution along the line of contact (t = 0 – t = 0,02)
obtained by the Von Mises theory
an output from the nonlinear dynamic analysis of the geared
transmission. The additional results for the strain distribu-
tion along the line of contact is given in Fig.8, and the Von
Misses stress distribution along the line of contact in each
time step (Fig.7) together with the distribution of the Von
Misses stress on the transmission elements for t = 0.02 is
presented in Fig.6.
The given results are the input results for the calculations
of principal invariants in each node for each element of the
presented gear set. The distribution of the principal stresses
in the zone of contact is presented in Figures10 and 11 as a
result of the software developed by the author that can be
used for the detailed transmission analysis. An example of
the principal stress and strain direction calculation for the
selected nodes in contact is presented on Fig.9 and
corresponding Tables: Table 2, and Table 3.

Figure 8. Strain distribution along the line of contact (t = 0 - t = 0.02)

Figure 9. Nodes used for calculating an example of the stress and strain
state (left) and example of the graphic representation of the principal stress
analysis results in the chosen nodes (right)

Figure 5. Stress distribution along the line of contact (t = 0 - t = 0.02)

Figure 6. Stress distribution for t = 0.02 by the Von Mises theory Figure 10. Contact strain tensor invariants for t = 0.02
24 D.DIMITRIJEVIĆ, V.NIKOLIĆ, I.ATANASOVSKA.: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR

Table 3. Component stresses and strains in node: 1180, principal


  stress,strain invariants, principal stresses and principal strains and
appropriate principal directions, in t=0.02
Node: 1180
σx (Pa) σy (Pa) σz (Pa)
-4,364E+08 -1,265E+08 -6,712+07
τxy (Pa) τxz (Pa) τyz (Pa)
1,218E+08 1,394+07 1,782E+08
α β γ
1,57 1,59 2,37E-02
1,49 1,55 8,27E-02
1,53 1,29 2,79E-01

L1 L2 L3
-6,31E+08 -4,66E+16 3,04E+27
σ1 (Pa) σ2 (Pa) σ3 (Pa)
Figure 11. Contact stress tensor invariants for t = 0.02
-6,92E+09 -4,24E+08 1,03E+09
εx (µm) εx (µm) εx (µm)
Conclusions 180,126 -11,681 48,446
γxy (µm) γxz (µm) γyz (µm)
The Finite Element Method is used in this paper to 150,750 17,263 220,616
develop a real model of the geared set using the self α β γ
developed software by the author and using an open source 1,61 1,79 2,26E-01
concept of the FEM engineering analysis. The calculated 1,07 6,66E-01 1,98
model is very flexible due to the methodology used, and 1,15 5,03-01 1,83E+00
because of the CODE_ASTER/SALOME [12] engine. J1 J2 J3
Since the conditions in which the dynamic analysis is -1,20E+02 -6,40E+04 -3,34E+06
done are quasistatic, it is to say that the suggested method ε1 (µm) ε2 (µm) ε3 (µm)
-3,38E+02 6,39E+01 1,54E+02
approximates very well the real working conditions of the
transmission. On the other hand, as every dynamic
calculation, it is exhaustive and time consuming. Literatura
A novel method for the results analysis together with the
[1] PARKER,G.R., VIJAYAKAR,M.S., IMAJO,T. (2000): Non-Linear
developed software is used for the analysis of the principal Dynamic Response of a Spur Gear Pair: Modelling and Experimental
stress and strain state of the geared transmission using Comparisons, Journal of Sound and vibration, 237(3), 435-455,
principal stress and strain invariants as calculation doi:10.1006/jsvi.2000.3067
parameters for the examination of strain and stress state in [2] DIMITRIJEVIĆ,D. (1997): Dinamička Analiza složenih mašinskih
each point of the body gear. sistema, Magistarska teza, Mašinski fakultet, Univerzitet u
Kragujevcu, 1-109
The software allows a generalized analysis of every
[3] DIMITRIJEVIĆ,D., NIKOLIĆ,V. (1998a): Automatic finite elements
system for which this kind of analysis can be useful. mesh generation with eigenfrequencies and mode shapes calculations
Table 2. Component stresses and strains in node: 3954, principal stress, for vecihle gears, MVM-Volume 24, Number 2, Kragujevac, June
strain invariants, principal stresses and principal strains and appropriate 1998.
principal directions, for t=0.02 [4] DIMITRIJEVIĆ,D., NIKOLIĆ,V. (1998b): Eigenfrequencies
Node: 3954 analysis for the deep drilling machine gear set, The Scientific Journal
FACTA UNIVERSATES, Series: Mechanical Engineering, l, 5.
σx (Pa) σy (Pa) σz (Pa)
3,704E+08 -1,257E+08 4,178E+07 [5] HEDRIH (Stevanović),K. (2003): A Trigger of Coupled
τxy (Pa) τxz (Pa) τyz (Pa) Singularities, Meccanica 38: 623–642, 2003.
-1,123E+06 4,400E+06 -4,650E+06 [6] HEDRIH (Stevanović), K. (2005): Nonlinear dynamics of a heavy
material particle along a circle which rotates and optimal control,
α β γ Springer, 2005
1,57 1,57 5,10E-03 [7] HEDRIH (Stevanović),K. (2006): Transversal vibrations of double-
1,56 1,57 8,37E-03 plate systems, Acta Mechanica Sinica (2006) 22:487–501
1,57 1,57 1,72E-03 [8] HEDRIH (Stevanović),K. (2006): Transversal vibrations of the
axially moving sandwich belts, Arch Appl Mech (2007) 77: pp.523–
L1 L2 L3 539
2,87E+09 -3,53E+16 -3,37E+27 [9] NIKOLIĆ,V., DIMITRIJEVIĆ,D. (1997a): On application of
σ1 (Pa) σ2 (Pa) σ3 (Pa) numerical methods in the studies of toothed gears, International
Symposium "Machines and Mechanisms", Belgrade, September, 2-5.
-1,39E+09 6,72E+08 3,59E+09 1997.
εx (µm) εx (µm) εx (µm) [10] NIKOLIĆ,V. MAKSIMOVIĆ,K., MAKSIMOVIĆ,S.: Fatigue life
188,363 -118,736 -15,062 prediction of gear teeth under general load spectrum, Int. Conference
γxy (µm) γxz (µm) γyz (µm) POWER TRANSMISSIONS`03, Varna, Bugarska, 2003.
-1,390 5,448 -5,757 [11] NIKOLIĆ,V., DIMITRIJEVIĆ,D. (1998): Rešavanje kontaktnih
problema teorije elastičnosti primenoin metode konačnih elemenata,
α β γ Naučno-stručni skup Istraživanje i razvoj mašinskih elemenata i
1,57 1,46 1,12E-01 sistema IRMES'98, Beograd, 10. i 11. septembar 1998, str.163-168
1,02 5,50E-01 1,57
1,52 1,57 4,96E-02 [12] DIMITRIJEVIĆ,D. NIKOLIĆ,V. (2004): Dynаmic modells of the
geаred systems- overview of the newly developed methods аnd
J1 J2 J3 procedures, IRMES'04, Kragujevac, Zbornik radova, str. 495 ... 500
5,46E+01 -2,30E+04 -1,53E+04
[13] Code Aster/Salome platforms (http://www.code-aster.com and
ε1 (µm) ε2 (µm) ε3 (µm) http://www.salome-platform.org.
-1,27E+02 6,63E-01 1,81E+02 Received: 04.09.2007.
D.DIMITRIJEVIĆ, V.NIKOLIĆ, I.ATANASOVSKA.: DYNAMIC ANALYSIS OF THE STRESS AND STRAIN STATE OF THE SPUR GEAR PAIR 25

Analiza dinamičkog stanja napona i stanja deformacija u


spregnutom zupčastom paru

U radu je predstavljen rezultat istraživanja stanja napona i stanja deformacije dinamčki spregnutog zupčastog para
koji se koristi za analizu unutrašnjeg dinamičkog opterećenja, vremenski-promenljivog stanja napona i stanja
deformacija. Korišćen je model dinamike spregnutih zupčanika sa dva stepena slobode kretanja. Analiza stanja
napona i stanja deformacije zupčastog para je urađena korišćenjem novog pristupa za prikaz stanja napona i stanja
deformacija u zavisnosti od invarijanti tenzora stanja napona, odnosno tenzora stanja deformacija u pojedinim
karakterističnim tačkama zupčanika. Na bazi teorije date u radu, razvijen je software za analizu napona i
deformacije primenom metode konačnih elemenata koji se bazira na korišćenju softvera CODE-ASTER/SALOME.

Ključne reči: naponsko stanje, dinamičko naprezanje, analiza napona, zupčasti prenosnik, deformacija, tenzor
napona, metoda konačnih elemenata..

Analyse dynamique de l’état de tension et déformation chez une pair


d’engrenages couplés
Le résultat des essais sur l’état de tension et les déformations d’une pair d’engrenages dynamiquement couplés, utilisé
pour l’analyse de la charge dynamique interne, de l’état de tension variable dans le temps et l’état de déformation, est
présenté dans cet article. On a utilisé le modèle dynamique d’engrenages couplés à deux degrés de liberté des
mouvements.L’analyse de l’état de tension et de déformation de pair d’engrenages est faite à l’aide da la nouvelle
approche de présentation de l’état de tension et déformations dans les points d’engrenages. A partir de la théorie
exposée dans ce travail, on a développé un logiciel pour analyser la tension et les déformations au moyen de la
méthode des éléments finis, basée sur l’utilisation du logiciel CODE-ASTER/SALOME.

Mots clés: état de tension, tension dynamique, analyse de tension, engrenage, déformation, tenseur des contraintes,
méthode des éléments finis.

You might also like