semi rigid diaphragm

[hosiensg]

hi
dear fmk
how i can introduce semi rigid diaphragm for 3D model?

[fmk]

either model with plane stress quad elements (2 modelbuilders required and equalDOF between the 2 sets of nodes) or use a Penalty method to enforce the constraints (think of the alpha factor as providing in-plane stiffness) .. the Penalty is easier is just not as theoretical

[hosiensg]

i want to use from your quad elements suggestion
but i don't know you about "2 modelbuilders required and equalDOF between the 2 sets of nodes"
may explain more about it
what dof should use in equalDOF command?
may model is like: http://opensees.berkeley.edu/wiki/image ... d.Wsec.tcl + Brace element?

[hosiensg]

in this post: http://opensees.berkeley.edu/community/ ... ent#p49187
vesna said Quad element can be only used in 2D and it has 2 dofs per node. my model is 3D

[fmk]

create a set of node for the quad elements with model Basic -ndf 2 -ndf 2 & add the quad lemenets, create your 3d building model with another model Basic -ndm 3 -ndf 6 command, add nodes and elements and
when done use equalDOF to constrain the 2d nodes to 3d nodes at the boundaries

[fmk]

if you want evrything 3d, use a plate element .. but it adds lots of dof's if your mesh of floor is fine.

[hosiensg]

if i want everything 3d. i should use from "ElasticMembranePlateSection" element that its thickness is the thickness of floor and its node should constraint by my plane node
is it right? . because i use this method but the period of my structure has a lot of error when i compare it by ETABS model
another problem is the situation and number of nodes that created by meshing of plate. how i can know about this?
why there isn't good example in your site from this topic?

[fmk]

if the ETABS model is the same, i.e. plates and beams, the results will be the same if the models are the same. the # of nodes and stuff to use you need to determine for yourself as it will be model dependent, i suggest you start large and subdivide until the answer converges (it should converge if you have no sharp irregulaities) .. i trust you have taken some cours in finite element method were you are introduced to such concepts. and there is no good example because no-one has provided one and i am kept busy just answering posts, emails and keeping executables up and running!