Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2019-10-17 20:14:47

astercode
Member
Registered: 2019-06-15
Posts: 26

Rigid body modes with TRI_DIAG and MODE_RIGIDE

Hi everyone,

I am having a structure consisting only of beams. I would like to perform an eigenmodes analysis similar to EIGRL of Nastran.

So I tried to use TRI_DIAG with MODE_RIGIDE on.

modes = CALC_MODES(CALC_FREQ=_F(NMAX_FREQ=20),
                   MATR_MASS=masse,
                   MATR_RIGI=rigidity,
                   NORM_MODE=_F(NORME='MASS_GENE'),
                   OPTION='PLUS_PETITE',
                   SOLVEUR_MODAL=_F(METHODE='TRI_DIAG',
                                    MODE_RIGIDE='OUI',
                                    NMAX_ITER_ORTHO=5,
                                    NMAX_ITER_QR=30,
                                    PREC_LANCZOS=1e-08,
                                    PREC_ORTHO=1e-12),
                   TYPE_RESU='DYNAMIQUE',
                   VERI_MODE=_F(STOP_ERREUR='NON'))

I have the following questions:

1. Are the lines above going to give modes normalized by mass?
2. When using the boundary condition (fixed at some nodes) as CHARGE
in the ASSEMBLAGE I get 0 rigid modes and then all the proper modes.
However when removing the CHARGE and leaving the model free, in the message
file it says that it found 6 rigid modes, but immediatelly after the
solution stops and prints a big error message asking many questions
and complaining for singular matrix.

Any idea what goes wrong? I tried many different things but didn't manage
to get the solution complete and containing the 6 rigid body modes.

Thanks

Last edited by astercode (2019-10-17 20:15:16)

Offline

#2 2019-10-20 18:40:42

astercode
Member
Registered: 2019-06-15
Posts: 26

Re: Rigid body modes with TRI_DIAG and MODE_RIGIDE

Any help, anyone?

Offline

#3 2019-10-21 14:55:00

dezsit
Member
Registered: 2012-06-27
Posts: 51
Website

Re: Rigid body modes with TRI_DIAG and MODE_RIGIDE

Hello, could you post your full message file please?
BR
dezsit

Offline

#4 2019-10-22 21:58:43

astercode
Member
Registered: 2019-06-15
Posts: 26

Re: Rigid body modes with TRI_DIAG and MODE_RIGIDE

Here it is. If I remove the charge I get six rigid modes but solution never ends due to singularity.


Attachments:
upload_modal.comm, Size: 2.01 KiB, Downloads: 9

Offline

#5 2019-10-23 08:01:04

dezsit
Member
Registered: 2012-06-27
Posts: 51
Website

Re: Rigid body modes with TRI_DIAG and MODE_RIGIDE

Hello,

Actually we would need to see the message file, with the error message, but you can also try this:

TRI_DIAG with OPTION='BANDE' and minimum frequency slightly below zero to handle all numerically zero values, it usually calculates the free body modes, and other eigenmodes in the given freq. range

= CALC_MODES(CALC_FREQ=_F(FREQ=(-1.0, 6000.0),
                                NMAX_ITER_SHIFT=5),
                   MATR_MASS=M,
                   MATR_RIGI=K,
                   OPTION='BANDE',
                   SOLVEUR_MODAL=_F(METHODE='TRI_DIAG',
                                    MODE_RIGIDE='OUI'),
                   VERI_MODE=_F(STOP_ERREUR='NON'))

If does not work, please upload the message file, or the complete case to reproduce the error.

BR.
dezsit

Offline

#6 2019-10-29 20:50:21

astercode
Member
Registered: 2019-06-15
Posts: 26

Re: Rigid body modes with TRI_DIAG and MODE_RIGIDE

Option Band did the trick and rigid body modes reported correctly.
Thanks!!!

Offline