Strangers graph values of the shears forces and bendings moments. 2

[rodrigomanera]

Good night,
I´m analying a building with the SAP2000 v12. The structure is compose for frames and shells like a structure of a simple building.
But I can´t undestand, why the both elements show strangers graph values of the shears forces and bendings moments, when I share the shell and frame elements and connecte the both elements with “break at intersection point” or if assign “insertion point” and “end offsets” for the frame.
These graphs looks like a “sow teeth” and there are differents values from the same joints.
This picture is an example of this case.
There is something that makes me in doubt. When I want to show a bending moment for a single shell the graph is ok, on other hand, if I start moving up or moving down the shells into my structure the graph isn´t ok.
I´d like to know what could I do to change this!?

 

[kabitchovitch]

please send an SDB file

 

[rodrigomanera]

This is the archive ".sdb" that you asked me kabitchovitch....Notice the strangers graph values of the shears forces and bendings moments, the differents values from the same joints and when I want to show a bending moment for a single shell the graph is ok, on other hand, if I start moving up or moving down the shells into my structure the graph isn´t ok......thanks for your help.

 

[kabitchovitch]

hello rodmanera
thank you for your send
I had cheked your model , i had made some changes , i think that you would not devied your area and your frames ( beams ) this is the things that had given you a stranger diagram of the moments .
i had changed your model , then i had mesched internaly your slab ( assign area automatic area mesh) 20x20 in a smal square 0.5m x 0.5m and i had jointed all the beams , this that means there is not devided beam when we select a beam we will have only one beam selected .
finaly i had made a diaphragm at each level.
i had used sap2000 v 14.
PS : it is better to use internal mesh than deviding a slab , beacause when we devid a slab , the beams also are devided into small beams , this is what had given you a stranger diagrams .
sorry for my english.

 

[rodrigomanera]

Hello Kabitchovitch....thanks for your attention. I did the way you told me, and now it´s ok. But I have two questions for you: The first is: I use "divide areas" when I´m working with just "shell elements" and when I have a shell and frame elements I have to use the "automatic area mesh"???....the second questions is: When I did the way you taught me, I can´t see how I divided the shell with the "automatic area mesh". There´s a way that I can see correctly my shell divided??....

 

[kabitchovitch]

hello Rodmanera
when you will do "automatic area mesh" , you can't see the mesh on the shel until the analysis had been done , then if you see that your shell was mesh as you wanted that is good , if not , you will lock the model and remesh the shell .
for exemple : you had meshed the shell by 6X4 and after analysis you see that isn't the way you that you want, you must lock the model and you will remsh the shell by 4x6 .
I wish you to understant if not ask me again .

 

[IJR]

kabitchovitch has probably solved your problem.

I would like to add the following general outlook based on statics.

In any model where 2 way behavior is modelled(shells+frames or intersecting frames) sawtooths will show up, especially when the model is coarse

This is because the elements perpendicular to the one you are studying will take moments in terms of torsion. Similar for shear. For frame models, then release torsions and moments accordingly. For shell + frame models, use finer mesh and appropriate shell thickness and properties.

respects
IJR