There are no BCs applied to the reference point (REF node and ROT node) of the rigid body in the original model so I left it unconstrained in Abaqus too. But there are no RBMs since the bottom part is fixed and the top one is tied to it.
I agree. First to chek are reaction forces but this is a nonlinear problem. There is a moment and the disk is rotating. The more it rotates , the less contact area. This problem seems to be selected from a
Challenge collection dude 
.
noo it has not to be challenging
why it is nonlinear? because of the contact definition? contact is tied. This definition also makes problem nonlinear in calculix?
I think that if there is a way to see reaction force(tangential) results with cylindirical coordinates, the problem will be solved
Anyone has an opinion about it?
Yes, you can even see this when submitting your analysis:
*WARNING reading *STATIC: a nonlinear analysis is requested
but no time increment nor step is specified
the defaults (1,1) are usedStatic analysis was selected
Newton-Raphson iterative procedure is active
You won’t see this if you replace tied contact with tie constraint. Of course, tied contact is not as highly nonlinear as regular contact.
Why would that help ? The forces you get are already normal and tangential to the contact surface. Cylindrical coordinates would orient them like in the case of a cylindrical surface with no inclination (e.g. surface of the hole in your model).
1 Like
I have made some refinement and differences are not so dramatic.
Ccx 5.3476E+03 against 5.4801E+03
Ccx 5.0736E+03 against 5.2061E+03
1 Like
I would appreciate it if you could share the input file. The results are quite close.
When the optistruct and abqus results were the same but the calculix results were different, I thought I had made a mistake during post-processing. I guess I couldn’t use CF, CFS, CFN properly.
Disla found very close results, my mistake will probably show up there
As I suggested, some refinement was necessary. The mesh was very coarse. If you have the geometry, you can try with different mesh densities (and maybe element types) yourself. It can be an interesting comparison. Apparently, Abaqus can handle such a coarse mesh better than CalculiX.
I would expect a difference in the results due to the mesh, but I wouldn’t expect a 2x difference(since resultants are the same with both optistruct and abaqus). I think there are touches of disla somewhere else. I hope he will send the model
Not really. Mecway needs custom cards to set up PRESSURE-OVERCLOSURE tied but I have try to respect all your parameters.Nonlinear aproach according to your contact behaviour.
*STEP,NLGEOM=YES,INC=100,AMPLITUDE=STEP
*STATIC,SOLVER=PARDISO
0.1,1,0,0.2
*CLOAD,AMPLITUDE=Az_380_1
380,3,1
You can try with my mesh and we can compare with Prepomax.
I am confused. You only sent the mesh in the model. Other than that, I think all boundary conditions are the same. For example, remote point coordinates are x=100,y=200,z=300 and 274N force is applied in the z direction.
Here are the prepomax results. Nothing is changed or i had a mistake in somewhere?
I also upload to model which has your mesh.
https://file.io/5qhMcXFPN5ce
That seems a Prepomax problem. N and S Forces has three components and Prepomax is only providing one.
Check directly the .dat file.
1 Like
The .dat file shows the same:
statistics for slave set INTERNAL_SELECTION-1_SOLID_PART-2_TO_SOLID_PART-1_SLAVE, master set INTERNAL_SELECTION-1_SOLID_PART-2_TO_SOLID_PART-1_MASTER and time 0.1000000E+01
total surface force (fx,fy,fz) and moment about the origin (mx,my,mz)
-1.420358E-09 5.600911E-10 2.740000E+02 5.472503E+04 -2.736495E+04 -2.328908E+00
center of gravity and mean normal
-3.607826E-07 7.777775E-01 -1.803783E-04 1.124352E-17 7.071068E-01 -1.568284E-16
moment about the center of gravity(mx,my,mz)
5.451192E+04 -2.736495E+04 -2.328908E+00
area, normal force (+ = tension) and shear force (size)
8.996832E+01 3.960013E-10 2.740000E+02
1 Like
Seems like only CF has been requested. Is that possible?. This is my dat file as a response to .
*Contact print, Totals=Yes, Master=Surface-1, Slave=Surface-2
CF,CFN,CFS
statistics for slave set SURFACE-2, master set SURFACE-1 and time 0.1000000E+01
total surface force (fx,fy,fz) and moment about the origin (mx,my,mz)
9.204445E-02 -1.733186E-01 2.740011E+02 5.473326E+01 -2.737018E+01 -3.587583E-04
center of gravity and mean normal
-3.863084E-08 7.777922E-04 1.838451E-08 4.005903E-06 7.071063E-01 -3.078518E-05
moment about the center of gravity(mx,my,mz)
5.452015E+01 -2.737019E+01 -2.871736E-04
area, normal force (+ = tension) and shear force (size)
8.995838E-05 -1.309895E-01 **2.740012E+02**
statistics for slave set SURFACE-2, master set SURFACE-1 and time 0.1000000E+01
total surface force (fx,fy,fz) and moment about the origin (mx,my,mz)
-1.456484E+00 -9.446124E-01 5.347611E+03 3.771550E+01 1.288453E-06 1.191307E-02
center of gravity and mean normal
-3.863084E-08 7.777922E-04 1.838451E-08 4.005903E-06 7.071063E-01 -3.078518E-05
moment about the center of gravity(mx,my,mz)
3.355617E+01 -2.052674E-04 1.078019E-02
area, normal force (+ = tension) and shear force (size)
8.995838E-05 -8.325744E-01 **5.347611E+03**
statistics for slave set SURFACE-2, master set SURFACE-1 and time 0.1000000E+01
total surface force (fx,fy,fz) and moment about the origin (mx,my,mz)
1.548529E+00 7.712938E-01 -5.073610E+03 1.701776E+01 -2.737018E+01 -1.227183E-02
center of gravity and mean normal
-3.863084E-08 7.777922E-04 1.838451E-08 4.005903E-06 7.071063E-01 -3.078518E-05
moment about the center of gravity(mx,my,mz)
2.096397E+01 -2.736998E+01 -1.106736E-02
area, normal force (+ = tension) and shear force (size)
8.995838E-05 7.015849E-01 **5.073610E+03**
1 Like
Right, only CF is requested in the OP’s file because PrePoMax supports only this. But it’s strange how CalculiX outputs this showing 3 times the same table but with correct normal force only in one table and correct shear force in another table. Also, why is normal force labeled as shear force in the second table ?
1 Like
I also do not understand well what TOTAL Normal or TOTAL Shear force means in a curved Surface.
I guess that’s the sum of all the normal components of each surface. But that’s not a Normal Force. It has sense for a flat surface but here, It has lost it’s meaning ¿isn’t it?.
1 Like
In Abaqus, it’s defined this way:
CFN
Total force due to contact pressure (CFNn, n = 1, 2, 3).CFNM
Magnitude of total force due to contact pressure.CFS
Total force due to frictional stress (CFSn, n = 1, 2, 3).CFSM
Magnitude of total force due to frictional stress.CFT
Total force due to contact pressure and frictional stress (CFTn, n = 1, 2, 3).CFTM
Magnitude of total force due to contact pressure and frictional stress.
So CFN and CFS in CalculiX should correspond to CFNM and CFSM, respectively. And the results I got from Abaqus before confirm this:
CONTACT OUTPUT FOR SECONDARY SURFACE SURFACE-2 AND MAIN SURFACE SURFACE-1
FOOT- CFNM CFN1 CFN2 CFN3 CFSM CFS1 CFS2 CFS3
NOTE
5408. 0.2589 1.743 -5408. 5134. -0.2589 -1.743 5134.
1 Like
Yep.
And the Overall resultant of forces is mantained except from a very small un-balance in x and y component.
| x | y | z | |
|---|---|---|---|
| CFN | 1.55E+00 | 7.71E-01 | -5.07E+03 |
| CFS | -1.46E+00 | -9.45E-01 | 5.35E+03 |
| CFN+CFS | 9.20E-02 | -1.73E-01 | 2.74E+02 |
1 Like