When running a frequency analysis (perturbation) for a cyclic symmetry model and NGRAPH=2 from a restart point, the frd file does not get the results for the second sector but shows NAN at the second sector nodal results. Running the same frequency step after the static step from one input gives meaningful results for the second sector. The calculated frequencies and the results for the 1st sector match for both.
Snippet from the input file with both static and frequency step:
*Tie, Name=cycsym, Position tolerance=0.05, Adjust=No, cyclic symmetry
csym_h, csym_l
*CYCLIC SYMMETRY MODEL, N=50, tie=cycsym, NGRAPH=2
0,0,0,1,0,0
*Step, Nlgeom, Inc=100
*Static
1.0, 1, 1E-05, 1.0
*Boundary, op=New
*Boundary
d_axial, 3, 3, 0
*Boundary
d_tang, 2, 2, 0
*Dload
Internal_Selection-1_Centrifugal_Load-1, CENTRIF, 112000, 0, 0, 0, 1, 0, 0
*Node file
RF, U
*El file
S, E, NOE
*Contact file
CDIS, CSTR
*end step
*Step, Perturbation
*restart,write,frequency=0
*Frequency
2
*select cyclic symmetry modes, nmin=4, nmax=4
*Node file
U
*el file
*contact file
*End step
Input for restart analysis:
*restart,read,step=1
*Step, Perturbation
*restart,write,frequency=0
*Frequency
2
*select cyclic symmetry modes, nmin=4, nmax=4
*Node file
U
*El file
*Contact file
*End step
Difference in frd files:
Static + frequency step:
restart + frequency step:
Any thoughts on that? I already checked that NGRPAH is read by restartread.f:
