I have an ANSYS structural model in which a single "mass node" (MASS21 element) is connected to the nodes on the surface of a mesh of solid elements using CERIG. Constraints are applied to the mass node, and gravity (uniform acceleration) acts on the structure. Using PRRSOL at releases 7.0 and 7.1, the net reaction of the mass node is only that due to the large mass itself. The contribution of force from the weight of the rest of the structure CERIG'd to the mass node is not included in the total reported by PRRSOL. Is this a bug?
Yes, this is defect 13655. The code was fixed at release 8.0. Mention of this defect was supposed to be included in the incompatibilities section under 8.0 Release notes, but it did note make it into the online documentation.
1) At release 7.0/7.1, OUTPR,RSOL printed out nodal forces in individual constraint equations (in addition to the reaction of the constrained node). The net sum of the reaction plus contributions from each CE was correct. However, PRRSOL did not report the additional forces from individual CE's and so was incorrect.
2) Attached is a test model. Input file test01.inp may be run in batch mode at Release7.1 by double clicking (in Windows Explorer) on batch file test01_71.bat. A batch run at 8.1 may be produced by runnng the test01_81.bat. This is a SOLID45 cantilever beam (total mass=0.0936 kg) with base CERIG'd to a single mass node (mass=1000kg). Local acceleration is 9.81 m/s^2. I interrogate reactions and nodal forces 4 ways:
a) Inspect printout in test01_71.out and test01_81.out created by OUTP,RSOL,LAST
b) PRRS in /POST1 (saved to files prrs_71.out and prrs_81.out)
c) NFORCE in /POST1 with only the beam base nodes selected (saved to files nfor_beam_71.out and nfor_beam_81.out).
d) NFORCE in /POST1 with only the mass node selected (saved to files nfor_bigmass_71.out and nfor_bigmass_81.out).
Net z component results are as follows:
a) Printout (OUTP,RSOL,LAST): 9810, CE reaction listing: 0.9182 9810.92
b) PRRS in /POST1: 9810.0 9810.9
c) NFORC (beam base nodes): -0.9182 -0.9182
d) NFORC (big mass node): -9810.0 -9810.0
So, at Release 7.1, the printed reaction solution (items "a" above) was "broken down" into the reaction acting on the mass element alone and an additional reaction contributed by the constraint equations. At Release 8.1, this format has apparently been superceded by a simpler "summary" of the net reaction (the mass node reaction PLUS that contributed by the beam and transmitted to the mass node through the CE's).
Note that nodal forces (NFORC) are only those that the selected elements exert on selected nodes. CE's are not considered in the report provided by NFORC. This explains why item (d) is not equal in magnitude to item (b). PRRSOL considers forces from constraint equations while NFORC/FSUM do not.