Deflections of Partially Composite Continuous
Deflections of Partially Composite Continuous
Deflections of Partially Composite Continuous
Received 25 February 1997; received in revised form 1 October 1997; accepted 6 January 1998
Abstract
A simple method to evaluate the deflections at midpoints of the various spans of continuous
composite beams having partial shear connection is developed. The method can be used irres-
pective of the type of loading and the configuration, i.e. the number and lengths of spans, of
the beam. In this method the actual continuous beam is represented by a series of single span
beams for which the deflections are already determined by using a linear partial interaction
theory. The redundant moments at the internal supports of the beam are assumed to be the same
for both the full and partial interaction cases. 1999 Elsevier Science Ltd. All rights reserved.
Notation
* Corresponding author.
0143-974X/99/$ - see front matter 1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 3 - 9 7 4 X ( 9 8 ) 0 0 0 0 1 - 7
292 N. Abdulrazzaq Jasim, A. Atalla / Journal of Constructional Steel Research 49 (1999) 291–301
1. Introduction
Current practice [1] permits composite beams to be designed with partial shear
connection. Slipping of the concrete slab with respect to the steel beam is thus inevi-
table. Thus the method of analysis used must be able to take account of the strength
and stiffness of the shear connectors.
Recently, extensive work [2–5] has been carried out concerning the deflections of
composite beams with partial shear connection. A method for predicting the deflec-
tions at midspans of simply supported beams is developed [2]. This method is derived
from an exact solution of the governing differential equations that are obtained from
linear partial interaction theory. The results are presented in the form of a design
chart which gives the deflections of partially composite beams as ratios of the corre-
sponding values for the equivalent fully composite ones. This chart can be used
irrespective of the type of loading, so it is suitable for everyday design. The same
method of analysis has been extended to cover the continuous composite beams [3].
Different types of loading and different configurations, i.e. number and lengths of
spans, of continuous beams are considered. The deflections at midpoints of the criti-
cal spans are evaluated and design charts in the same manner as for simply supported
beams are constructed. A general conclusion is drawn that continuous beams can be
classified into groups for each of which one design chart may be used.
Although the classification mentioned above can be made, it remains that for the
unusual cases of continuous beams which are not included in this classification the
governing differential equations need to be solved afresh. This constitutes the only
restriction on the previous method. It is the purpose of this paper to provide a gen-
eralized method for determining the deflections of continuous composite beams with
partial shear connection. In this new method the principle of superposition is utilized
so that the deflection can be evaluated irrespective of the type of loading and the
configuration of the continuous beam. Design charts are also constructed to simplify
the calculations required.
In this theory [6] the composite beam is assumed to be composed of two elastic
materials, each of which possesses the same elastic modulus in tension and com-
N. Abdulrazzaq Jasim, A. Atalla / Journal of Constructional Steel Research 49 (1999) 291–301 293
pression. It is assumed that the shear connectors are equally spaced along the beam
and have equal moduli, and the connection is smeared along the beam. The load–
slip characteristic of the connectors is taken as linear. No separation between the
concrete slab and steel beam is considered.
Using this theory, the deflection, y, of composite beam is given [3] by the differen-
tial equation
d2y M a d2F
⫽ ⫹ (1)
dx2 E2I C1E2Im dx2
where F is the solution of the differential equation
d2F
⫺ C1F ⫽ C2M (2)
dx2
and
C1 ⫽
kn
pE2 冉
Am 1 ⫹
a2
ImAm 冊
kna
C2 ⫽
pE2Im
Im ⫽ I1t ⫹ I2
Am ⫽ 1/A1t ⫹ 1/A2
The application of Eqs. (1) and (2) to simply supported beams, subjected to different
types of loading, leads to the equation [2]
n1 ⫽ 1 ⫹
3(C ⫺ 1)
K 2 冉
1
1 ⫺ tanh(K)
K 冊 (3)
where
n1 ⫽ yp/yf
C ⫽ I/Im
K ⫽ √C1L/2
Eq. (3) has been used [2] to construct a chart in terms of the deflection ratio n1 and
the factor K2 and is given for various values of the factor C.
In the case of continuous composite beams, Eqs. (1) and (2) are also applied to
different configurations of continuous beams subjected to different types of loading
[3]. Different equations, of a form similar to Eq. (3), depending on the configuration
and type of loading are obtained.
For the purpose of the present paper, the solutions of Eqs. (1) and (2) are needed
for the case of a propped cantilever (Fig. 1). The bending moment, M, is determined
at a distance x and introduced into Eq. (2). The solution, F, is found by direct inte-
gration and using the boundary conditions
294 N. Abdulrazzaq Jasim, A. Atalla / Journal of Constructional Steel Research 49 (1999) 291–301
x
L
F ⫽ 0 at x ⫽ 0,
and
dF/dx ⫽ 0 at x ⫽ L (slip ⫽ 0).
The solution F is then introduced into Eq. (1). Integrating and applying the bound-
ary conditions
y ⫽ 0 at x ⫽ 0,
y ⫽ 0 at x ⫽ L,
and
dy/dx ⫽ 0 at x ⫽ L
the deflection at midspan, yp, of a partially composite beam is given by
1 6(C ⫺ 1)
n2 ⫽ (13 ⫺ 11CN) ⫹ (4)
2 K4
冋再
2 cosh (K) ⫹ K 1 ⫹ 冉 CN
4 冊 冎
sinh(K) /cosh (2K) ⫹ K 2 1 ⫺
CN
2 冉
⫺2 冊 册
where
n2 ⫽ yp/yf
CN ⫽ G1/G2
3(C ⫺ 1)
G2 ⫽ 1 ⫹ {2K ⫺ tanh(2K)}
8K 3
G1 ⫽ 1 ⫹
2K 4 再
3(C ⫺ 1) 1 ⫹ K·sinh(2K)
cosh (2K)
⫺1 冎
Eq. (4), as Eq. (3), can also be proved to be independent of the type of loading,
N. Abdulrazzaq Jasim, A. Atalla / Journal of Constructional Steel Research 49 (1999) 291–301 295
so it may be used for any type of loading. This equation is given as a chart in Fig.
2 for various values of C and K2.
3. Proposed method
The proposed method is based on the assumption that the redundant moments at
the internal supports have the same values for both the partial and full interaction
cases. Such an assumption can be justified since it has been found [7] that the per-
centage decrease or increase, depending on the position along the beam, in the redun-
dant moments for the case of partial interaction is within 0–8% of the values for the
case of full interaction. Therefore, the procedure in the proposed method starts with
the determination of the moments at the internal supports, assuming the beam as
fully composite. Any of the methods available in the structural analysis texts [8] can
be used for this purpose. These moments are then used for the partially composite
beam.
Suppose, as a general case, that the deflection at the midpoint of an internal span
C = 4.0
4
C = 3.5
C = 3.0
3
yp / yf
C = 2.5
C = 2.0
2
C = 1.5
1
−3 −2 −1 0 1 2 3 4
log10k
of a partially composite beam is required (Fig. 3(a)). Separation of this span necessi-
tates the application of the predetermined end moments Mi and Mj as shown in Fig.
3(b). This specified span is then replaced by three single span beams, one simply
supported and two propped cantilever beams (Fig. 3(c)–(e)), each with an appropriate
ratio of the total load, w. The fixed end moments of the propped cantilevers provide
the end moments Mi and Mj of the actual span. The deflection at midpoint, yp, of
the actual span can be obtained by
yp ⫽ y1p ⫹ y2p ⫹ y3p (5)
where y1p is the deflection at midspan of the partially composite simply supported
beam subjected to a load of (tw) (Fig. 3(c)), y2p is the deflection at midspan of the
partially composite propped cantilever with left fixed end subjected to a load of (uw)
(Fig. 3(d)) and y3p is the deflection at midspan of the partially composite propped
cantilever with right fixed end subjected to a load of (vw) (Fig. 3(e)). The total load,
w, on the actual span is given by direct summation of the individual components on
the equivalent three single span beams, i.e.
w ⫽ tw ⫹ uw ⫹ vw (6)
where t, u, and v are the ratios of the loads on the single span beams to the actual
total load. This means that
t⫹u⫹v⫽1 (7)
To determine these ratios the fixed end moments in Fig. 3(d) and (e) are first writ-
ten as
w = general load
i j
L
(a)
t.w
(c)
L
w +
Mi Mj u.w
(d)
L
L i2 +
(b) v.w
(e)
L
j3
Mi2 ⫽ uM1w
Mj3 ⫽ vM2w
where M1w and M2w are the fixed end moments of propped cantilevers of Fig. 3(d)
and (e), respectively, assuming them to be subjected to the total load, w. However,
these fixed end moments give the end moments Mi and Mj of the actual span. Hence
Mi ⫽ uM1w
Mj ⫽ vM2w
from which
u ⫽ Mi/M1w
v ⫽ Mj /M2w
Substituting these into Eq. (7) yields
t ⫽ 1 ⫺ (Mi/M1w ⫹ Mj /M2w)
The method of Ref. [3] states that Eqs. (1) and (2) can be solved for various cases
of continuous beams having different configurations and types of loading. With this
method it is necessary for many charts to be constructed, each of which belongs to
a group of the cases considered.
In contrast to this method, the proposed method needs only two charts, one for
the simply supported beam and the other for the propped cantilever. These charts
may be used irrespective of the type of loading. The charts alongside Eq. (11) can
be used to calculate the deflections of partially composite continuous beams irrespec-
tive of the configuration and type of loading.
The results obtained by the proposed method are compared with those obtained
using the method presented in Ref. [3] for different continuous beams. The results
are given in Table 1. This table shows that the results of the two methods are in
general the same. The slight differences that appear in some cases are attributed to
Table 1
Comparison of results of proposed method with those of Ref [3]
yp/yf
K2 Ref. [3] Eq. (11) Ref. [3] Eq. (11) Ref. [3] Eq. (11) C
the assumption of the same distribution of bending moments for both the partial and
full interaction cases.
For a given partially composite continuous beam and given type of loading the
procedure of solution can be summarized as follows.
1. The moments at the internal supports, M1,…, Mi, Mj,…, are calculated assuming
the beam to be fully composite.
2. The span at which the deflection is required is specified.
3. The deflection, yf, at midpoint of this span is calculated assuming the beam to
be fully composite.
4. Assuming a fully composite case the deflections yw2 and yw3 at the midspan and
the fixed end moments M1w and M2w of the propped cantilevers having the same
length and subjected to the same load as the specified span of continuous beam
are evaluated.
5. For a fully composite case the deflection, yw1, at the midspan of the simply sup-
ported beam having the same length and subjected to the same load as for the
specified span is evaluated.
6. The parameters t, u and v are determined (u ⫽ Mi/M1w, v ⫽ Mj/M2w, t ⫽ 1 ⫺ u
⫺ v).
7. The ratio R is determined.
8. Depending on the properties of the given partially composite continuous beam,
the factors C and K are calculated.
9. From the charts of the simply supported beam and the propped cantilever the
ratios n1 and n2 are determined.
10. Using Eq. (11) the deflection, yp, at the midpoint of the specified span of the
partially composite continuous beam can be calculated.
An example to illustrate this procedure is given in Appendix A.
6. Conclusions
Appendix A
Suppose that the deflection, yp, at midpoint of the middle span of the partially
composite continuous beam shown in Fig. 4(c) is required. Also, assume the beam
to have properties giving C ⫽ 3.0 and K ⫽ 10.
Using the method of consistent deformations [8] and considering the beam as fully
composite, the moments at the four supports are given as
M1 ⫽ M4 ⫽ 0
3
M2 ⫽ M3 ⫽ WL
40
and the deflection at midpoint of the middle span as
11 W ⫻ L3
yf ⫽
960 E2I
For a simply supported beam of length L and subjected to a concentrated load W at
midspan, the deflection under the load is
W ⫻ L3
yw1 ⫽
48E2I
Similarly, for a propped cantilever of length L and load W at midspan, the deflection
under the load is
7W ⫻ L3
yw2 ⫽ yw3 ⫽
768E2 ⫻ I
and the fixed end moment is
w w
x x
L L L L
(a) (b)
w
L/2 L/2
x 4
1 2 3 4
L L L
(c)
3WL
M1w ⫽ M2w ⫽
16
Thus,
u⫽ 冉 冊冒冉 冊
3
40
WL
3
16
WL ⫽ 0.4, v ⫽ 0.4, t ⫽ 0.2
and
冉
R ⫽ 2 ⫻ 0.4 ⫻
7W ⫻ L3
768E2I 冊冒冉 0.2 ⫻
W ⫻ L3
48E2I 冊
⫽ 1.75
From charts, or Eqs. (3) and (4), and for C ⫽ 3.0 and K2 ⫽ 100,
n1 ⫽ 1.05
n2 ⫽ 1.12
Using Eq. (11), the required deflection, yp, is given as a ratio of yf by
yp 1.05 ⫹ 1.12 ⫻ 1.75
⫽ ⫽ 1.09
yf 1 ⫹ 1.75
This value is the same as that obtained from equation (15) of Ref. [3].
References
[1] British Standards Institution (1976) Draft Standard for the Use of Structural Steel in Buildings, Part
3: Composite Construction. British Standards Institution, London, 147 pp.
[2] Jasim NA, Mohamed ALI, Anis A. Deflections of composite beams with partial shear connection.
The Structural Engineer 1997;75(4):58–61.
[3] Jasim NA. Computation of deflections for continuous composite beams with partial interaction, Proc.
Inst. of Civil Engineers, Structures and Buildings 1997;122:347–54.
[4] Jasim, N. A., Mohamed, Ali & Anis, A., The effect of partial interaction on the deflection due to
shrinkage of steel–concrete composite beams, The Structural Engineer (submitted).
[5] Jasim NA. Deflections of partially composite beams with linear connector density. Journal of Con-
structional Research, 1999;49:241–54.
[6] Johnson, R. P., Composite Structures of Steel and Concrete, 1, Crosby Lockwood Staples, London,
1975, 210 pp.
[7] Jasim, N. A. & Abdulamer, A., The behaviour of continuous composite beams with partial shear
connection, The Fourth Basrah Engineering Conference, 1997, 202–217.
[8] Sack, R. L., Structural Analysis, Chap. 11, McGraw-Hill, New York, pp. 339–381.