Topic: compare the solution accuracy of CCT Element in OOFEM with ANSYS

Hello bp,
Recently, I compared the solution of OOOFEM with ANSYS.  I used the CCT Element in OOFEM and shell63 element ANSYS.  The material properties, structural geometry, boundary conditions and the element behaviors are exactly the same.  However, the solution of OOFEM is about only 10% of the solution abtained by ANSYS. 
My model is a 6m long, 4m wide, 0.5m thick plate. The two long ends are clamped and a 2000N vertically down force applies in the middle of the plate.

Here is the solution of node 2358 in ANSYS, the coordinate of node 2358 is (2.5500,2.0028), and the force applies on this node.
    NODE      UX          UY          UZ          USUM 
2358   0.0000      0.0000    -0.19361E-06 0.19361E-06
    NODE      ROTX        ROTY        ROTZ        RSUM 
2358 -0.91102E-10-0.38963E-08  0.0000     0.38973E-08

Here is the solution of node 2358 in OOFEM:
Node        2358:
  dof 1   d -2.44355595e-007
  dof 2   d -1.30415359e-010
  dof 3   d -3.75831204e-009


Here is the apdl program in ANSYS.
finish
/clear
/filname,plate3,1
/title,plate3
/units,si
*set,long,5
*set,width,4
*set,meshlength,0.1
/prep7
et,1,shell63
r,1,0.5
keyopt,1,1,2
!only bending effect are considered
mp,ex,1,2e11
mp,prxy,1,0.16
rectng,,long,,width
esize,meshlength
Asel,S,AREA,,1
Aatt,1,1,1
mshape,1,2d
Amesh,ALL
/solu
nsel,all
nsel,s,loc,x,0
d,all,all,0
nsel,all
nsel,s,loc,x,5
d,all,all,0
nsel,all
nsel,s,node,,2358
f,all,fz,-2000
allsel
solve
save
finish
/post1
nsel,all
nsel,s,node,,2358
prnsol,u,comp
prnsol,rot,comp

I also provide the data file I made for the OOFEM below. So could you please check it for me? I have checked it by myself several times, but I can't fingure out where the fault is.
Thank you.

2

Re: compare the solution accuracy of CCT Element in OOFEM with ANSYS

Hi,

quite strange. I have made a simple calculation by hand:
clamped beam, point load in the middle, displacement in the middle is w=(F*L^3)/(192.0*E*I).
I count in kN, m, KPa:
I=(1/12)*4*0.5^3 = 0.0416666
w = (2.0*6^3)/(192.0*2.e8*0.0416666) = 2.7e-7 [m]

which is approximately the value obtained using oofem. Of course, the shear was not taken into account, but it will be negligible.

Borek.