1G check with Superelement CMS in MSC Nastran v2018.2

[cookiemonster69]

Hi everyone, I'm a student trying to get familiar with Superelement creation and assembly in MSC Nastran v2018.2.

I created a dummy FEM of 2 ducts attached to a valve. The purpose is to make a comparison between the complete FEM and a version where the Valve acts as a DMIG model using Component Mode Synthesis (CMS) and subject it to different checks like Freefree, Fixfix, 1G, etc.

For the creation run I am using sol103, I am also using these commands on the Case Control section:
PARAM AUTOQSET YES
EXTSEOUT(STIF,MASS,ASMBULK,EXTBULK,EXTID=100,DMIGPCH)

And this command on the Bulk Data section:
ASET1 123456 6001 6002 $ 6001 and 6002 being my FEM's interface "external" grid points

This way, I obtain a .pch+.asm files to include on the assembly run. Since I am trying to do a 1G check, I am using a GRAV load on the residual+DMIG but when checking the .f06 file the OLOAD Resultant matrix comes empty (zeros) and this warning message appears:

*** USER WARNING MESSAGE 2070 (MTRXIN)
REQUESTED DMIG MATRIX KAAX IS UNDEFINED.
User Information:
A DMIG matrix was selected in the Case Control but was not present
in the Bulk Data.
*** USER WARNING MESSAGE 2070 (MTRXIN)
REQUESTED DMIG MATRIX MAAX IS UNDEFINED.
User Information:
A DMIG matrix was selected in the Case Control but was not present
in the Bulk Data.

This is how I'm importing the DMIG model onto the assembly run:
SOL 101
GEOMCHECK SUMMARY
CEND
K2GG=KAAX
M2GG=MAAX
TITLE = 1G VALVE DMIG
ECHO = NONE
EQUILIBRIUM=YES
WEIGHTCHECK = YES
MEFFMASS(ALL)
SPCFORCES=ALL
OLOAD=ALL
SUBCASE 1
SUBTITLE= 1G X-DIRECTION
SPC = 2
LOAD = 1
SUBCASE 2
SUBTITLE= 1G Y-DIRECTION
SPC = 2
LOAD = 2
SUBCASE 3
SUBTITLE= 1G Z-DIRECTION
SPC = 2
LOAD = 3
BEGIN BULK
PARAM POST -1
PARAM WTMASS .00259
PARAM AUTOSPC NO
PARAM GRDPNT 0
PARAM K6ROT 1.
PARAM COUPMASS 1
PARAM SNORM 20.0
PARAM PRTMAXIM YES
PLOTEL 70640 6002 6001 $ interface nodes
$=======2=======3=======4=======5=======6=======7=======8=======9=======
SPC 2 50000 123456 0. 50006 123456 0.
$
GRAV 1 0 386.1 1. 0. 0.
GRAV 2 0 386.1 0. 1. 0.
GRAV 3 0 386.1 0. 0. 1.
$
INCLUDE 'DUCTS.bdf'
INCLUDE 'MAT1_COORDS.bdf'
INCLUDE 'valve_dmig.asm'
INCLUDE 'valve_dmig.pch'
ENDDATA

Is there a reason why the run is not detecting the DMIG matrixes KAAX and MAAX?

 

[cookiemonster69]

Update: the 1G check was completed. The use of P2X=PAX fixed the zero OLOAD resultant matrix. The warning message is still appearing but the 1G check is successful now.

 

[dmapguru]

In the current configuration, your K2GG, M2GG commands apply to all superelements, i.e. superelement 100 (the external one) and superelement 0 (the residual structure). If you look closer at the f06 file, you will see that the warning messages apply only to the residual structure because MSC Nastran cannot find DMIGs with the names you gave as they are included in the BEGIN SUPER=100 block (I am inferring there is no warning message for superelement 100). There is a way to avoid these warning messages, but as long as you did not want to include the DMIGs (with the names given on K2GG and M2GG) in the residual structure, you can safely ignore the warnings.

I am a little curious about why the addition of the P2G=PAX command "fixed the zero OLOAD resultant matrix" as you put it, from which I am inferring that the OLOAD is not non-zero. The EXTSEOUT command only requests stiffness and mass reduction, so there probably is no PAX matrix in the pch file from the external superelement creation run. Can you post the file you are using to create the external superelement (i.e. the one with the EXTSEOUT command), and I can run it, along with the residual run you already gave us, to see what is going on?

DG