Apago PDF Enhancer
517
each section, these ratios have been determined at 11 different
points, and the orientation of the principal axes has been indicated
at each point.†
It is clear that s
max
does not exceed s
m
in either of the two
sections considered in Fig. 8.8 and that, if it does exceed s
m
else-
where, it will be in sections close to the load P, where s
m
is small
compared to t
m
.‡ But, for sections close to the load P, Saint-Venant’s
principle does not apply, Eqs. (8.3) and (8.4) cease to be valid, except
in the very unlikely case of a load distributed parabolically over the
end section (cf. Sec. 6.5), and more advanced methods of analysis
taking into account the effect of stress concentrations should be
used. We thus conclude that, for beams of rectangular cross section,
and within the scope of the theory presented in this text, the maxi-
mum normal stress can be obtained from Eq. (8.1).
In Fig. 8.8 the directions of the principal axes were determined
at 11 points in each of the two sections considered. If this analysis
were extended to a larger number of sections and a larger number
of points in each section, it would be possible to draw two orthogonal
systems of curves on the side of the beam (Fig. 8.9). One system
would consist of curves tangent to the principal axes corresponding
to s
max
and the other of curves tangent to the principal axes corre-
sponding to s
min
. The curves obtained in this manner are known as
the stress trajectories. A trajectory of the first group (solid lines)
defines at each of its points the direction of the largest tensile stress,
while a trajectory of the second group (dashed lines) defines the
direction of the largest compressive stress.§
The conclusion we have reached for beams of rectangular cross
section, that the maximum normal stress in the beam can be obtained
from Eq. (8.1), remains valid for many beams of nonrectangular cross
section. However, when the width of the cross section varies in such
a way that large shearing stresses t
xy
will occur at points close to the
surface of the beam, where s
x
is also large, a value of the principal
stress s
max
larger than s
m
may result at such points. One should be
particularly aware of this possibility when selecting W-beams or
S-beams, and calculate the principal stress s
max
at the junctions b and
d of the web with the flanges of the beam (Fig. 8.10). This is done
by determining s
x
and t
xy
at that point from Eqs. (8.1) and (8.2),
respectively, and using either of the methods of analysis of Chap. 7
to obtain s
max
(see Sample Prob. 8.1). An alternative procedure, used
in design to select an acceptable section, consists of using for t
xy
the
maximum value of the shearing stress in the section, t
max
5 VyA
web
,
given by Eq. (6.11) of Sec. 6.4. This leads to a slightly larger, and thus
conservative, value of the principal stress s
max
at the junction of the
web with the flanges of the beam (see Sample Prob. 8.2).
8.2 Principal Stresses in a Beam
†See Prob. 8.C2, which refers to a program that can be written to obtain the results shown
in Fig. 8.8.
‡As will be verified in Prob. 8.C2, s
max
exceeds s
m
if x # 0.544c.
§A brittle material, such as concrete, will fail in tension along planes that are perpendicular
to the tensile-stress trajectories. Thus, to be effective, steel reinforcing bars should be
placed so that they intersect these planes. On the other hand, stiffeners attached to the
web of a plate girder will be effective in preventing buckling only if they intersect planes
perpendicular to the compressive-stress trajectories.
Tensile
Compressive
P
Fig. 8.9 Stress trajectories.
a
b
c
d
e
Fig. 8.10 Key stress
analysis locations
in I-shaped beams.
bee80288_ch08_512-547.indd Page 517 10/30/10 4:57:04 PM user-f499bee80288_ch08_512-547.indd Page 517 10/30/10 4:57:04 PM user-f499/Users/user-f499/Desktop/Temp Work/Don't Delete Job/MHDQ251:Beer:201/ch08/Users/user-f499/Desktop/Temp Work/Don't Delete Job/MHDQ251:Beer:201/ch