Hallo,
danke für die Antwort. Bei mir scheint es so zu sein, dass von Anfang an die Verschiebungskonvergenz mit herangezogen wird, ohne die Kraftkonvergenz abzuwarten. Hab mal den Teil aus der solve.out anghängt.
DFORCE CONVERGENCE VALUE = 0.3600E+07 CRITERION= 0.1837E+05
SPARSE MATRIX DIRECT SOLVER.
Number of equations = 2240227, Maximum wavefront = 64482
Memory allocated for solver = 4714.310 MB
Memory required for in-core = 15576.120 MB
Optimal memory required for out-of-core = 619.832 MB
Minimum memory required for out-of-core = 432.618 MB
*** NOTE *** CP = 400.812 TIME= 15:32:14
The Sparse Matrix solver is currently running in the optimal
out-of-core memory mode. This memory mode achieves a good balance of
memory usage and hard drive usage in order to obtain an efficient
solution time.
*** WARNING *** CP = 400.812 TIME= 15:32:14
Pivoting has been activated for the Sparse Matrix Solver. The memory
required may greatly exceed predicted amount. In that event rerun
with increased memory using -m option (refer to the Basic Analysis
Procedures Guide, Ch. 19).
curEqn= 366393 totEqn= 2240227 Job CP sec= 695.797
Factor Done= 25% Factor Wall sec= 97.736 rate= 7211.6 Mflops
curEqn= 858651 totEqn= 2240227 Job CP sec= 986.703
Factor Done= 50% Factor Wall sec= 208.928 rate= 6716.6 Mflops
curEqn= 1440207 totEqn= 2240227 Job CP sec= 1301.391
Factor Done= 70% Factor Wall sec= 330.060 rate= 5897.8 Mflops
curEqn= 1525146 totEqn= 2240227 Job CP sec= 1363.016
Factor Done= 76% Factor Wall sec= 353.539 rate= 5989.5 Mflops
curEqn= 2240227 totEqn= 2240227 Job CP sec= 1659.141
Factor Done= 100% Factor Wall sec= 466.841 rate= 6005.5 Mflops
curEqn= 2240227 totEqn= 2240227 Job CP sec= 1659.141
Factor Done= 100% Factor Wall sec= 0.000 rate= 0.0 Mflops
Sparse solver maximum pivot= 1.101715243E+19 at node 261971 UZ.
Sparse solver minimum pivot= -3.66668565E+12 at node 403312 UX.
Sparse solver minimum pivot in absolute value= 713.576619 at node
321812 UZ.
DISP CONVERGENCE VALUE = 819.7 CRITERION= 0.3059
EQUIL ITER 1 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 6.117
LINE SEARCH PARAMETER = 1.000 SCALED MAX DOF INC = 6.117
FORCE CONVERGENCE VALUE = 0.6204E+06 CRITERION= 0.1875E+05
*** WARNING *** CP = 3170.562 TIME= 16:16:40
The rotational degree of freedom of pilot node 755479 is not well
constrained which can cause rigid body motion.
*** WARNING *** CP = 3170.562 TIME= 16:16:40
The rotational degree of freedom of pilot node 755480 is not well
constrained which can cause rigid body motion.
*** WARNING *** CP = 3170.562 TIME= 16:16:40
The rotational degree of freedom of pilot node 755481 is not well
constrained which can cause rigid body motion.
DISP CONVERGENCE VALUE = 181.6 CRITERION= 0.3362
EQUIL ITER 2 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= 0.6061
LINE SEARCH PARAMETER = 0.9892 SCALED MAX DOF INC = 0.5996
FORCE CONVERGENCE VALUE = 0.6275E+05 CRITERION= 0.1914E+05
DISP CONVERGENCE VALUE = 12.19 CRITERION= 0.3381
EQUIL ITER 3 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= -0.1133
LINE SEARCH PARAMETER = 1.000 SCALED MAX DOF INC = -0.1133
FORCE CONVERGENCE VALUE = 8012. CRITERION= 0.1953E+05 <<< CONVERGED
DISP CONVERGENCE VALUE = 2.843 CRITERION= 0.3381
EQUIL ITER 4 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC= -0.2063E-01
LINE SEARCH PARAMETER = 0.9734 SCALED MAX DOF INC = -0.2008E-01
FORCE CONVERGENCE VALUE = 726.6 CRITERION= 0.1992E+05 <<< CONVERGED
DISP CONVERGENCE VALUE = 0.1848 CRITERION= 0.3381 <<< CONVERGED
EQUIL ITER 5 COMPLETED. NEW TRIANG MATRIX. MAX DOF INC = -0.1533E-01
>>> SOLUTION CONVERGED AFTER EQUILIBRIUM ITERATION 5
Ist zwar hier, für den ersten Step, nur eine Iteration mehr, aber im zweiten wirkt sich das extrem zeitraubender aus. Ich hab auch schon versucht das "Program Controlled" auf "Remove zu ändern. Leider kein Erfolg, der Solver meint er muss das Verschiebungskriterium trotz dem benutzen. Wenn er das unbedingt brauch, dann vertrau ich ihm mal.
Noch eine kleine Zusatzfrage. In dem Auszug sind 3 Warnings, dass Pilot Nodes nicht richtig constrained sind. Bei diesen Knoten handelt es sich um Punktmassen. Wie können dies nicht richtig constrained sein? Hab die doch in WB an die Struktur über Flächen angebunden.
Gruß
Christian
Eine Antwort auf diesen Beitrag verfassen (mit Zitat/Zitat des Beitrags) IP