Cancel
Showing results for 
Search instead for 
Did you mean: 

USER FATAL MESSAGE 5423 (MPYAD) - ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES

Solution Partner Experimenter Solution Partner Experimenter
Solution Partner Experimenter

I have a question about Simcenter 3D.

I have a problém sol 103. I need to use a pre-load screw connection and flexible RBE3 element with DOF'S removed. The solution ends with an error „USER FATAL MESSAGE 5423 (MPYAD)“.

 

I attach a sample example with an error, it contains two calculations:

  • 103_preload_NOT_RBE3-FIX …. works, error-free calculation
  • 103_preload_WITH_RBE3-FIX… Error 5423 (MPYAD)

 

Both tasks should give the same result. Can you advise where there may be a problem with RBE3?

 

Download here:

https://cloud.techsim.cz:8443/owncloud/index.php/s/K78XkrTaR4qQmOF

 

Thx

 

Simcenter 3D with ERROR 5423.png

4 REPLIES

Re: USER FATAL MESSAGE 5423 (MPYAD) - ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES

Siemens Phenom Siemens Phenom
Siemens Phenom

Answers are in the .f06 file:

 

 *** USER WARNING MESSAGE 2050 (ESCHEL)
     AUTOMPC PROCESSING ENCOUNTERED          2 SMALL PIVOTS WHICH WERE NOT SMALL ENOUGH TO MAKE THE EQUATION REDUNDANT
     AND THEREFORE MAY MAKE THE CONSTRAINT MATRIX ILL-CONDITIONED RESULTING IN LARGE NUMERICAL ROUNDOFF ERRORS.
     SETTING THE PARAMETER AMPCZ TO 1.0E-4 OR LARGER WILL RESULT IN THESE EQUATIONS BEING ELIMINATED (REDUNDANT).
     THIS SITUATION OCCURS MOST FREQUENTLY WHEN EXCESSIVE DOFS ON GRIDS OF RIGID ELEMENTS ARE RESTRAINED.

 

and:

 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  1 ILLEGALLY DEFINED IN SETS   UM   US                     
 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  2 ILLEGALLY DEFINED IN SETS   UM   US                     
 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  3 ILLEGALLY DEFINED IN SETS   UM   US                     
 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  4 ILLEGALLY DEFINED IN SETS   UM   US                     
 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  5 ILLEGALLY DEFINED IN SETS   UM   US                     
 *** USER FATAL MESSAGE 2101 (GP4)
     GRID POINT        65918 COMPONENT  6 ILLEGALLY DEFINED IN SETS   UM   US                     

 

If your intent is to constrain all of the independent DOF in the RBE3 in all 6 DOF, you should be using a RBE2 instead. In an RBE2, the first grid is independent and all other grids are dependent. See the attached update. I copied struna_sim1-103_preload_with_rbe3-fix.dat to struna_sim1-103_preload_with_rbe2-fix.dat and hand edited the RBE3 to change it to RBE2.

Re: USER FATAL MESSAGE 5423 (MPYAD) - ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES

Solution Partner Experimenter Solution Partner Experimenter
Solution Partner Experimenter
Thank you for your effort to answer.
Of course, I changed the AMPCZ parameter (to 1.0E-4 or larger). But without result.

I need to connect the face to the center point without adding stiffness (RBE2 adds rigidity). Can somehow set up RBE3? Any parameter? Or use another element? Which does not add stiffness.

Re: USER FATAL MESSAGE 5423 (MPYAD) - ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES

Siemens Phenom Siemens Phenom
Siemens Phenom

In your original post, you provided two variants of the model:

  • One with SPC's defined on all 6 DOF of every independent RBE3 grid.
  • One with a SPC on all 6 DOF of the dependent RBE3 grid.

You stated that you expected "Both tasks should give the same result"

 

Defining a SPC on every dof of every independent grid is adding infinite stiffness to the model at the SPC DOF. The dependent RBE3 grid will have zero displacement since all of the independent DOF are fixed.

 

Replacing the RBE3 with a RBE2 and constraining all 6 dof of the RBE2 independent grid will yield the same results. You cannot expect to fix all DOF and not add stiffness. Those two requirements are diametrically opposed.

Re: USER FATAL MESSAGE 5423 (MPYAD) - ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES

Solution Partner Experimenter Solution Partner Experimenter
Solution Partner Experimenter

My problem is that. I solve the problem originally solved in Ansys, I am now dealing with the continuation of the Simcenter. For the calculation, it is necessary to have the same boundary conditions. In ansys, the Remote Point - Deformable (that is, the RBE3 type) and node fix (ux = uy = uz = 0), as shown in the appendix.

 

Can the same boundary condition in Simcenter be made? I assumed that the RBE3 could do this.

Ansys boundary conditions_Remote point - deformable (RBE3).png