Hi everybody,
I have an old question :
if I have a frame+walls building(4 stories) Can I define wall considering 4 nonlinearBeam-Column elements (along its height) and nonLinearBeam-Column elements(2:one on right and one on left) constrained to column's node by rigidLink Command?
Or Do I have to define it as shell element...but in this way do I have to consider Plate Fiber Section to consider it as nonLinear element?
I'm confused...please tell me if it's possible the first solution!
Thanks
wall
Moderators: silvia, selimgunay, Moderators
modeling all using beams
It is quite wrong to model a wall using beam elements. I just finished teaching mechanics of materials and the first thing we cover in going over beams is that the theory covers slender beams, while there is a set of additional terms (also approximate) that take care of shear deformations for beams that have h/l<1/5 or so...
walls are typically h/l~ 1 or more so even the beam with shear deformations is not good enough...
That is where the plain strain (solid) element comes into play...
Boris Jeremic
walls are typically h/l~ 1 or more so even the beam with shear deformations is not good enough...
That is where the plain strain (solid) element comes into play...
Boris Jeremic
Boris, you are right if the wall's height to length ratio are low you shouldn't use a beam element. However, if you are looking at a system where the main lateral resistance is the wall, then I wouldargue that if the wall is four stories tall then the h/l is probably > 3, then it will be flexure dominated and a beam element works ok. I have comparison to test data to support this.
However you should aggregate on the shear stiffness to the section, because it still has a significant impact. But this can be done with the section aggregator command. We have done this and shown very good simulation of the force-displacement response response for single rectangular walls.
However you should aggregate on the shear stiffness to the section, because it still has a significant impact. But this can be done with the section aggregator command. We have done this and shown very good simulation of the force-displacement response response for single rectangular walls.
short beams
Sure if it is 4 storry high it might be only marginally better case. Even for a 4 story high wall, there are beams (plates) on each level that bring in normal forces and bending moments and shear forces. So yes the tall wall would have a main (probably) mode of bending along the height (say a very tal cantiliver) but there would be local influences (major) from structural components on each level. The vertical stresses in fibers from bending moments and from shear forces for seismic loading would be way...
The only case were ling beam would work (even if 3/1 is not really a long beam) is if connection with beams and plates at each level do not take moments (or take very little), but only shear and normal forces. However I do not remember seing many building details where that is the case...
Look at some of the papers by folks from Delft (de Borst, Feenstra...) on what is really going in plain concrete elements...
Boris
The only case were ling beam would work (even if 3/1 is not really a long beam) is if connection with beams and plates at each level do not take moments (or take very little), but only shear and normal forces. However I do not remember seing many building details where that is the case...
Look at some of the papers by folks from Delft (de Borst, Feenstra...) on what is really going in plain concrete elements...
Boris