My plasticity analysis seems to get caught in an endless loop of repeated
bisections in which the new time increment is the same as before, thereby
repeating the exact same solution over and over. I am getting a message
about the incremental plastic strain computed in the iteration being larger
than the criterion, which then forces another bisection. There is also some
reference to re-specifying the criterion with the CUTCONTROL command.
What should I do to improve my convergence?
This "infinite bisection loop" scenario does not occur often, but we have seen it occur when there is a singularity point from a load or boundary condition or some other poor modeling technique which is then compounded by the algorithm that ensures that the plastic strain developed in that increment is not too large. The maximum plastic strain increment permitted is set with the CUTCONTROL,PLSLIMIT command. The default value is 0.15 (i.e., 15%). If this limit is increased, you can often gain the needed convergence, but there is always a risk that the solution will be inaccurate if the increment is too large. Therefore, Defect 25090 was filed to improve the bisection algorithm by forcing a smaller "new time increment" when the calculated new time increment is too close to the previous time increment that failed to converge. Defect 25090 was resolved by adding this expanded capability to the bisection algorithm in the UP20040621 build of ANSYS 9.0. A test case showing the new capability follows. fini /clear /title, Modified Version of Test Set Problem c0040.dat /prep7 MP,PRXY,,0.3 et,1,93 et,2,43 et,3,143 et,4,181 et,5,91,,1 r,1,.1 r,2,1 $rmore $rmore,,,.1 r,3,10 $rmore $rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 rmore,,,.01 mp,ex,1,210000 tb,biso,1 tbdata,1,1000 n,11 n,12,1 n,13,1,2 n,14, ,2 ngen,6,10,11,14,1,,,2 type,1 $real,1 $e,11,12,13,14 type,2 $real,1 $e,21,22,23,24 type,3 $real,1 $e,31,32,33,34 type,4 $real,1 $e,41,42,43,44 type,5 $real,2 $e,51,52,53,54 type,5 $real,3 $e,61,62,63,64 emid fini /solu nsel,s,loc,x d,all,ux d,all,roty d,all,rotz nsel,r,loc,y d,all,all nsel,s,loc,x,1 !d,all,roty,10 d,all,roty,20 nsel,all !outpr,,1 cutcontrol,plslimit,0.05 ! force bisection to test new algorithm ... solve save fini /eof = = = = = = = = = = = = = = = = = = = = = = = = = == = = = = = = = = = = = = = = = = = = = = = = = ANSYS 8.1 "infinite" loop of exact same time step size: ======================================= *** LOAD STEP 1 SUBSTEP 1 COMPLETED. CUM ITER = 4 *** TIME = 0.502354E-01 TIME INC = 0.502354E-01 *** MAX PLASTIC STRAIN STEP = 0.4547E-01 CRITERION = 0.5000E-01 *** AUTO STEP TIME: NEXT TIME INC = 0.50235E-01 UNCHANGED FORCE CONVERGENCE VALUE = 9835. CRITERION= 50.18 MOMENT CONVERGENCE VALUE = 723.4 CRITERION= 3.691 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 1.005 FORCE CONVERGENCE VALUE = 0.2107E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 1.680 TIME= 16:08:25 The calculated reference FORCE CONVERGENCE VALUE = 2.509738066E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.291 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4135 FORCE CONVERGENCE VALUE = 0.1143E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 1.800 TIME= 16:08:25 The calculatedreference FORCE CONVERGENCE VALUE = 1.535734252E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1389E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 1.800 TIME= 16:08:25 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 7 *** BEGIN BISECTION NUMBER 1 NEW TIME INCREMENT= 0.49983E-01 FORCE CONVERGENCE VALUE = 9785. CRITERION= 49.92 MOMENT CONVERGENCE VALUE = 719.8 CRITERION= 3.673 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.9997 FORCE CONVERGENCE VALUE = 0.2100E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 1.960 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 2.499463588E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.290 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4113 FORCE CONVERGENCE VALUE = 0.1135E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 2.050 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 1.526108984E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1384E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 2.050 TIME= 16:08:26 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED.CUM ITER = 9 *** BEGIN BISECTION NUMBER 2 NEW TIME INCREMENT= 0.49983E-01 FORCE CONVERGENCE VALUE = 9785. CRITERION= 49.92 MOMENT CONVERGENCE VALUE = 719.8 CRITERION= 3.673 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.9997 FORCE CONVERGENCE VALUE = 0.2100E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 2.200 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 2.49848524E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.290 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4113 FORCE CONVERGENCE VALUE = 0.1136E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 2.290 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 1.52837624E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1384E-01 CRITERION= 0.5943E-01 <<< CONVERGED ***NOTE *** CP= 2.290 TIME= 16:08:26 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 11 *** BEGIN BISECTION NUMBER 3 NEW TIME INCREMENT= 0.49983E-01 FORCE CONVERGENCE VALUE = 9785. CRITERION= 49.92 MOMENT CONVERGENCE VALUE = 719.8 CRITERION= 3.673 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.9997 FORCE CONVERGENCE VALUE = 0.2100E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 2.440 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 2.499014762E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.290 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4113 FORCE CONVERGENCE VALUE = 0.1140E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 2.520 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 1.528175966E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1384E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 2.520 TIME= 16:08:26 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 13 *** BEGIN BISECTION NUMBER 4 NEW TIME INCREMENT= 0.49983E-01 FORCE CONVERGENCE VALUE = 9785. CRITERION= 49.92 MOMENT CONVERGENCE VALUE = 719.8 CRITERION= 3.673 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.9997FORCE CONVERGENCE VALUE = 0.2099E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 2.670 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 2.499271489E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.290 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4113 FORCE CONVERGENCE VALUE = 0.1134E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 2.750 TIME= 16:08:26 The calculated reference FORCE CONVERGENCE VALUE = 1.525069123E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1384E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 2.750 TIME= 16:08:26 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps oruse the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 15 *** BEGIN BISECTION NUMBER 5 NEW TIME INCREMENT= 0.49983E-01 FORCE CONVERGENCE VALUE = 9785. CRITERION= 49.92 MOMENT CONVERGENCE VALUE = 719.8 CRITERION= 3.673 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.9997 FORCE CONVERGENCE VALUE = 0.2100E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 2.900 TIME= 16:08:27 The calculated reference FORCE CONVERGENCE VALUE = 2.498798155E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.290 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4113 FORCE CONVERGENCE VALUE = 0.1134E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 2.980 TIME= 16:08:27 The calculated reference FORCE CONVERGENCE VALUE = 1.52442257E-10 is less than 1E-2 a threshold zero, Avalue of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1384E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 2.980 TIME= 16:08:27 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 17 *** BEGIN BISECTION NUMBER 6 NEW TIME INCREMENT= 0.49983E-01 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = ANSYS 9.0 forced bisection to prevent "infinite" loop: ===================================== *** LOAD STEP 1 SUBSTEP 1 COMPLETED. CUM ITER = 4 *** TIME = 0.502354E-01 TIME INC = 0.502354E-01 *** MAX PLASTIC STRAIN STEP = 0.4547E-01 CRITERION = 0.5000E-01 *** AUTO STEP TIME: NEXT TIME INC = 0.50235E-01 UNCHANGED FORCE CONVERGENCE VALUE = 9835. CRITERION= 50.18 MOMENT CONVERGENCE VALUE = 723.4 CRITERION= 3.691 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 1.005 FORCE CONVERGENCE VALUE = 0.2107E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 1.730 TIME= 16:08:58 The calculated reference FORCE CONVERGENCE VALUE = 2.509738066E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.291 CRITERION= 0.6539E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.4135 FORCE CONVERGENCE VALUE = 0.1143E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING *** CP= 1.820 TIME= 16:08:58 The calculated reference FORCE CONVERGENCE VALUE = 1.535734252E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1389E-01 CRITERION= 0.5943E-01 <<< CONVERGED *** NOTE *** CP= 1.820 TIME= 16:08:58 The incremental plastic strain computed in this iteration is larger than the criterion of 5% leading to bisection. You may try incrementing the load more slowly by increasing the number of substeps or use the CUTCONTROL command to re-specify this criterion. *** LOAD STEP 1 SUBSTEP 2 NOT COMPLETED. CUM ITER = 7 *** BEGIN BISECTION NUMBER 1 NEW TIME INCREMENT= 0.44207E-01 FORCE CONVERGENCE VALUE = 8654. CRITERION= 44.16 MOMENT CONVERGENCE VALUE = 636.6 CRITERION= 3.248 EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.8841 FORCE CONVERGENCE VALUE = 0.1927E-09 CRITERION= 0.5206E-04 <<< CONVERGED *** WARNING *** CP= 1.960 TIME= 16:08:58 The calculated reference FORCE CONVERGENCE VALUE = 2.269815309E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 1.263 CRITERION= 0.6524E-01 EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.3630 FORCE CONVERGENCE VALUE = 0.1018E-09 CRITERION= 0.5312E-04 <<< CONVERGED *** WARNING ***CP= 2.040 TIME= 16:08:58 The calculated reference FORCE CONVERGENCE VALUE = 1.372784325E-10 is less than 1E-2 a threshold zero, A value of 1E-2 or specified MINREF is used. Check results carefully. MOMENT CONVERGENCE VALUE = 0.1257E-01 CRITERION= 0.5943E-01 <<< CONVERGED >>> SOLUTION CONVERGED AFTER EQUILIBRIUM ITERATION 2 *** LOAD STEP 1 SUBSTEP 2 COMPLETED. CUM ITER = 8 *** TIME = 0.944425E-01 TIME INC = 0.442071E-01 *** MAX PLASTIC STRAIN STEP = 0.4422E-01 CRITERION = 0.5000E-01 *** AUTO STEP TIME: NEXT TIME INC = 0.44207E-01 UNCHANGED = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = |
||
|