Why won't the input below with COMBI214 converge or give correct answers?/PREP7ET,1,MASS21,0,0,2 R,1,1.0ET,2,COMBI214 ! ELEMENT TYPE # 3: BEARING ELEMENTKEYOPT,2,2,1 ! YZ PLANEKEYOPT,2,3,0 ! ELEMENT IS SYMMETRICR,2,10,10,10.0,10.0,, ! REAL CONSTANTS ( K11,K22,K12,K21) NO DAMPING/COM, NODESN,1, 0.0, 0.0, 0.0N,2, 0.0, 1.0, 0.0N,3, 0.0, 0.0, 1.0/COM, ELEMENTS/COM, MASSTYPE,1REAL,1E,1/COM, BEARINGSTYPE,2REAL,2E,2,1,3ALLSEL,ALLFINI/SOLUD,1,UZD,ALL,UX ! NO TRACTION D,2,ALL ! SECOND NODES OF BEARINGD,3,ALL ! SECOND NODES OF BEARINGOUTPR,ALL,ALLANTYPE,TRANS !NONLINEAR TRANSIENT DYNAMIC ANALYSIS!NLGEOM,ON ! LARGE DEFLECTION OPTIONNROPT,FULLKUSE,-1NEQIT = 30TIME,1.0F,1,FY,1.0KBC,0NSUBST,10,10,10LSWRITE,1 ! WRITE LOAD STEP FILE 1TIME,10.0F,1,FY,0KBC,1NSUBST,90,90,90LSWRITE,2 ! WRITE LOAD STEP FILE 2FINISH/SOLUOUTRES,ALL,ALLOUTPR,ALL,ALLLSSOLVE,1,2,1 ! READ IN LOAD STEP FILES 1 and 2FINISH

 This is defect 63544. The problem was caused by node orientation. Typically, the i and k nodes are at a negative offset to the node on the centerline. In this model, they have a positive offset, and that was not accounted for correctly in the code. If nodes 2 and 3 are given a negative y and z location, the problem disappears. The defect has been fixed in 12.0.