Professional Documents
Culture Documents
BRACED FRAMES
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 01/12/19. Copyright ASCE. For personal use only; all rights reserved.
INTRODUCTION
quake type excitation depends upon the stiffness, the relationship be-
tween stiffness and maximum drift is not linear. Consider, e.g., the com-
monly used power law approximation to the pseudo-acceleration response
spectrum, S„, given by
Sa = C0T-* (1)
in which c0 = some appropriate constant; T = the period of vibration of
the structure; and p = the exponent which characterizes the spectrum.
[The Uniform Building Code, e.g., recommends a value of P = 1/2 (15)].
For a structure of mass m, the maximum displacement, Amax is related
to the stiffness, K, approximately as
Amax = ~ = / ( m ) K - a . (2)
0)
H •£.
A t
h
/ .1
-\\ ML'\D h—>-—^1
^/°75
V&^OS)
1 \
0.4 0.6
FIG. 3.—-Variations of Stiffness for Dif- FIG. 4.~-Variations of Stiffness for Dif-
ferent Aspect Ratios, Diagonal Bracing ferent Aspect Ratios, Split-K Bracing
(h/Ic = 0.2S, h/A^L2 = 0.001, EIb/GA'bL2 (Same Parameter Ratios as Fig. 3)
= 0.01)
342
/ h
/ 11
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 01/12/19. Copyright ASCE. For personal use only; all rights reserved.
\ f*—— L ——-|
1 0
\ w -
^v \ °25
a^_Nv\,0-10
, , ,
0.0 0.2 0.4 0.6
the shear area) and the topological parameters e/L and h/L. For sim-
plicity in the present analysis, the axial and shear stiffnesses of the col-
umns are assumed to be infinite. The effectiveness of bracing can be
seen by considering how the stiffness varies as the location of the brace,
measured by e/L, is changed. Thus, in each of the cases shown in Figs.
3-6, all of the member stiffness parameters and a selected structure as-
pect ratio, h/L, are held fixed while e/L varies from 0 to 1. For compar-
ison, each case has been normalized by dividing the braced stiffness by
the stiffness of the corresponding moment resisting frame (e/L = 1).
The variation of stiffness with the eccentricity ratio in a diagonally braced
frame is shown in Fig. 3 for aspect ratios of 1.0, 0.75, and 0.50. It is
evident that for values of e/L > 0.5 little benefit is gained by bracing.
For e/L < 0.5, substantial increases in stiffness can be achieved. This
effect is particularly noticeable for relatively narrower bays (e.g., h/L =
1). While the stiffness is greatest at small values of e/L, it will be seen
in the next section that e/L must be kept sufficiently large in order that
the members can meet ductility demands at extreme overloads.
For wide bays it is often impractical to use a single diagonal brace
because their length may become excessive. For these situations, the split-
K framing system [or some other similar system, such as the frames
depicted in Fig. 1(c) and 1(d)] may then be more appropriate. The effect
of eccentricity on stiffness for this case is shown in Fig. 4, again for a
variety of frame aspect ratios. It should be noted that these aspect ratios
measure a different quantity than previously. It is believed that this def-
inition provides a meaningful comparison between the two bracing ar-
rangements since, for equal values of e/L and h/L, the bracing angle is
the same in the two cases. It is apparent that this bracing scheme is more
sensitive to the eccentricity ratio than the frames braced with a single
diagonal. Here, the benefits of bracing for small e/L are greater than
before, but these benefits rapidly disappear-with increasing e/L.
As shown in Fig. 5, when the columns are large in relation to the
343
INELASTIC BEHAVIOR
1. The section strength properties, denoted by Vp*, Mp*, and Mp, axe
defined by Eqs. 7-9, in the following section. A consistent set of units
has been used.
It is instructive to note from this study the sequential formation of
345
has not been considered in this analysis. The effect of this interaction
would be the earlier formation of the plastic hinges at the column bases.
Since the inelastic deformations can be almost entirely attributed to
shear for active links with the properties given in Table 1, it is appro-
priate to define member ductility, |x,„, as
7 GA'
Pm = — = -77- 7 (3)
7y Vu
in which 7 = the average shear distortion; a n d GA' = t h e elastic cross
sectional shear stiffness. V„ = t h e ultimate shear capacity, which m a y
be less than V£ d u e to the presence of b e n d i n g moment, a n d can be
determined for a specific problem using Eq. 10. The lateral deflection at
the t o p of t h e structure, A, can also be normalized, t h e resulting quantity
often being associated with the term structure ductility, |x s .
_ A _ K
Ay H„
in which K = the elastic lateral frame stiffness, condensed to the loaded
degree of freedom; a n d H„ = t h e frame limit load.
The state of deformation w h e n t h e columns form plastic hinges, [event
4 in Fig. 7(a)] corresponds to a structure ductility of |xs = 7.4. At that
level of structure deformation, t h e m e m b e r ductilities were found to be:
|x„ = 72, |x„ = 55, a n d ^l, = 22, in which |xf„ represents t h e ductility of
the member which first w e n t inelastic at event i. Two conclusions can
be made from this example: First, the ratio of a member ductility mea-
sure to structure ductility measure can be quite high. This p h e n o m e n o n
is typical of the eccentrically braced frame a n d occurs mostly d u e to frame
geometry. It is important to note that the definition of ductility em-
ployed applies only to active links which deform predominantly in shear.
Comparing ductility magnitudes with those, e.g., in a m o m e n t resisting
frame can be misleading because the m o d e of deformation in t h e t w o
cases is significantly different, a n d thus so are t h e definitions of ductil-
ity. A second conclusion is that the ductility d e m a n d s can vary greatly
throughout a structure. In general, some active links are likely to be
located in regions that are more critical than others a n d m u s t be detailed
accordingly.
Additional insight into t h e possible behavior of an eccentrically braced
frame is gained by considering t h e redistribution of forces d u e to inelas-
tic action in the frame. The distribution of bending m o m e n t s for t h e
three story frame is s h o w n in Fig. 7(b) for t h e three stages of loading
corresponding to events 1, 3, a n d 4 (the distribution for event 2 is quite
similar to that of event 1). O n e can see that in this particular example
a significant change in the w a y the lateral loads are resisted takes place
as the inelasticity progresses. Of particular interest is the fact that col-
u m n s originally in double curvature bending change to single curvature
bending d u e to the redistributions. Such a p h e n o m e n o n should be con-
o46
L
(a) f 1 (b) (c)
sidered in the design of the columns. The often used practice of de-
signing for factored elastic moments could be unconservative in certain
cases.
It is recognized that the example presented here may be an extreme
case; the structural drift at event 4 may well be in excess of certain code
limitations. Furthermore, it should be possible to proportion an eccentri-
cally braced frame such that a more favorable redistribution of bending
moments will occur under a specified loading pattern. This direction will
not be pursued herein. The example presented points out the impor-
tance of incorporating inelastic analyses into the design process. An elas-
tic analysis may not be sufficient to insure the integrity of a structural
system which takes advantage of inelastic activity. This comment applies
equally to moment resisting frames.
The elasto-plastic analysis does give an indication of how the member
ductilities vary throughout the structure. One is left, however, with some
question as to how these ductility demands correspond with those de-
rived from an inelastic dynamic analysis or another static analysis using
a different pattern of loads. Additional static analyses on the structure
being considered here have shown that the distribution of ductility de-
mand is significantly influenced by the form of loading imposed.
For preliminary design purposes, it is often sufficient to know only
the order of magnitude of the member ductility demands. Such an es-
timate can be obtained from a rigid-plastic analysis of the structure. Ki-
nematieally admissible fields of deformation for three frames having rigid-
plastic members are shown in Fig. 8. In any one of these frames, all of
the member ductility demands are equal under the rigid-plastic as-
sumption. Because member deformations are forced to be uniform
throughout these structures, one would expect that the member ductility
demands found in this manner would, in some sense, represent an av-
erage of those obtained from an elasto-plastic analysis. Nevertheless, the
rigid-plastic analysis can provide some guidelines for the selection of the
links.
To obtain an idea of how the structure deformation relates to the
member deformation under the rigid-plastic assumption, it is sufficient
to consider the simple frame. The collapse mechanism of this frame is
shown in Fig. 9. Geometrical considerations lead to the following rela-
347
61 = ye (5)
For shear links (small e/L) 7 measures the average shear strain, while
for bending links (large e/L) it measures the plastic hinge rotations. When
e is much smaller than L, the ratio of member deformations to structure
deformations is quite large. Using Eq. 5 and the definitions of ductility,
given by Eqs. 3 and 4, one can compute an approximate rigid-plastic
member ductility which corresponds to event 4 of the elasto-plastic anal-
ysis of the three story frame example. The two approaches are compared
assuming an equal drift at the top of the structure. The angular drift 9
is then the tip deflection divided by the height of the structure. The
value of 7 from Eq. 5 is then used in Eq. 3 to obtain a value of |A„, = 55
for each of the links of the three story frame. One should note that this
value is intermediate to the values obtained from the elasto-plastic analysis.
For the three frames shown in Fig. 8, the moment resisting frame (a)
places the smallest demand on the active links, for a given value of 6,
according to Eq. 5, since 0 = 7 for this configuration. As was pointed
out earlier, it is advantageous to make the link as small as possible to
maximize the elastic stiffness of the frame. On the other hand, Eq. 5
suggests that the shorter the link is, the higher will be its ductility de-
mand. The imperative condition that a link's ductility demand be less
than the ductility, which it can supply, will limit the shortness of the
link.
It should be recognized that a structure employing short links may
sustain considerable floor damage in a severe earthquake. The regions
of potential damage are encircled by dashed lines in Fig. 8. An estimate
of the floor deflection, 8, for a given 8 for the simple structure is given
by
8 = 6(L - e) (6)
It is evident that the optimal eccentrically braced frame represents a
maximization of the structural stiffness under the constraint that the
member ductility demands must be physically realizable. The next sec-
tion deals with the problem of determining member capacity and gives
348
MEMBER CAPACITY
Mp = cTyZ (9)
in which Ty = the yield stress in pure shear; a y = the yield stress in pure
tension (a y = V5 Ty for the von Mises yield criterion); d, b, tw, and tf
= the section depth, flange width, web thickness, and flange thickness
respectively; and Z = the plastic section modulus. If one neglects the
small increase in shear capacity for points to the left of A, the curve can
be approximated as
)^B
FIG. 11,—Typical Shear Link Details for (A) Strong and (B) Weak Axis Column
Connections
CONCLUSIONS
ACKNOWLEDGMENTS
The writers thank James O. Malley a n d Kazuhiko Kasai for their con-
tributions to the current research on eccentrically braced frames. Also,
the support of the National Science Foundation, Grant N o . CEE 81-07217
is gratefully acknowledged. The opinions expressed in this paper are
those of the writers and d o not necessarily reflect the views of NSF.
APPENDIX.—REFERENCES
1. Clough, R. W., and Penzien, J., Dynamics of Structures, McGraw Hill Book
Co., New York, N.Y., 1975.
2. Drucker, D. C , "Effect of Shear on the Plastic Bending of Beams," Journal
of Applied Mechanics, Vol. 23(4), 1956, p. 509.
3. "Eccentric Bracing is Key to Seismic Resistance," Engineering News Record,
Oct. 25, 1979, pp. 32-33.
4. Hjelmstad, K. D., and Popov, E. P., "Cyclic Behavior and Design of Link
Beams," Journal of Structural Engineering, ASCE, in press.
5. Hodge, P. G., Plastic Analysis of Structures, McGraw Hill Book Co., New York,
N.Y., 1959.
6. Libby, J. R., "Eccentrically Braced Frames Construction—A Case History,"
AISC Engineering Journal, Vol. 18, No. 4, 1981, pp. 149-153.
7. Malley, J. O., and Popov, E. P., "Design of Shear Links in Eccentrically Braced
Frames," Journal of Structural Engineering, ASCE, in press.
8. Neal, B. G., "Effect of Shear Force on the Fully Plastic Moment of an I-Beam,"
Journal of Mechanics and Engineering Science, Vol. 3, No. 3, 1961, p. 258.
9. Popov, E. P., and Bertero, V. V., "Seismic Analysis of Some Steel Building
Frames," Journal of the Engineering Mechanics Division, ASCE, Vol. 106, No.
EMI, Feb., 1980, pp. 75-92.
10. Popov, E. P., and Hjelmstad, K. D., "Cyclic Web Buckling," Proceedings,
Structural Stability Research Council, 1981, p. 57.
11. Popov, E. P., and Manheim, D. N., "Eccentric Bracing of Steel Frames in
Seismic Design," Transactions, 6th International Conference on Structural
Mechanics in Reactor Technology, Paris, France, Vol. K(b), K 13/8, Aug.,
1981.
12. Roeder, C. W., and Popov, E. P., "Eccentrically Braced Steel Frames for
Earthquakes," Journal of the Structures Division, ASCE, Vol. 104, No. ST3, Mar.,
1978.
13. Roeder, C. W., and Popov, E. P., "Design of an Eccentrically Braced Steel
Frame," AISC Engineering Journal, Third Quarter, 1978.
14. Roeder, C. W., and Popov, E. P., "Inelastic Behavior of Eccentrically Braced
Steel Frames Under Cyclic Loadings," EERC Report 77-18, Univ. of Califor-
nia, Berkeley, Calif., Aug., 1977.
15. Uniform Building Code, 1982 ed., International Conference of Building Offi-
cials, Whittier, Calif., 1982.
353