Linear Displacement Loading

[nallia]

Hello!

I investigate a 2-dimensional problem with 4 DOF. The first two DOF are displacement in x1 and x2-direction (u1, u2). The considered area is a unit square.

I want to apply a linear displacement pattern in x2-direction to the top edge of the square with the nodal values at (0,1) u2 = 0.2 and at (1,1) u2 = -0.2.
I tried implementing this using CSURface but it did not work. I checked my results and the corner values of displacement did not match the displacement I defined in CSURface.

Thanks for your help!

Code: [Select]

! Linear surface loading
CSURface
CART DISP, 2
LINEar
1 1.0 1.0 -0.2
2 0.0 1.0 0.2


[JStorm]

I think you are looking for the edge-features, e.g. EBOUndary and EDISplacement, which helps you to select all nodes in a plane where one coordinate is constant.

[nallia]

Is it possible to define a linear displacement pattern using EDISplacement? I only managed to implement constant displacement to an edge using EDIS. An example for a linear displaccement pattern would be very helpful to me.

[JStorm]

Sorry, I missed the point that you need linear changing displacements.

[Prof. R.L. Taylor]

Currently, CSURf does not have an option to generate any  linear or nonlinear distribution.  If you have an edge with nodes that are equally numbered you can use the DISP option as

DISP
  n1 ninc  d1
  n2         d2

The displacement for dof = 1 should vary linearly between the nodes.  Otherwise you could write your own user module.

[Prof. R.L. Taylor]

What version of feap are you using.  I trested ver8.4 and ver8.5 on the attacched and CSURf does do linear displacements.

[Prof. S. Govindjee]

I think the problem is that you have written

Code: [Select]

CART DISP 2it should just be

Code: [Select]

DISP 2

[nallia]

Thank you very much for your reply! I use ver. 8.5.

I tried the input form for CSURF you suggested but the boundary condition was not applied. I checked this in the results.

Maybe there is a conflict with the other boundary conditions? My input form is by now not executed at all.

[Prof. S. Govindjee]

make sure your csurf points generate a right-hand rule outward normal to the surface that you are trying to identify (same for when you apply the boundary codes).  -- note I haven't looked yet at your file.

[Prof. R.L. Taylor]

1. I have corrected the csurf by removing the CART before the DISP

2. You should NEVER do a batch solution with many steps prior to carefully testing the mesh on a small mesh with one or two steps.

3. It is best to use an interactive test by plotting the mesh, boundary conditions, any loads, etc.

4. After that look at the output file Otest and look at all the data feap input and echoed.  In your case you would see that the CSURf labeled what you input as FORCES not DISPLacements.

Doing these few steps will save you time and help you understand what might be missing.

Be sure your user element is working now before doing the big problem.  Good luck.