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

You are not logged in.

#1 2013-12-09 17:39:01

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

[SOLVED]error while writing a mode shape in a med file

hello

this code, trying to calculate frequency with hysteretic damping

ASSEMBLAGE( 
  MODELE=model,CHAM_MATER=material,CARA_ELEM=elemcar,CHARGE=ground,
  NUME_DDL=CO('NUMEDDL'),
  MATR_ASSE=(
	_F(MATRICE=CO('rigidity'),OPTION='RIGI_MECA',),
	_F(MATRICE=CO('masse'),OPTION='MASS_MECA',),
	_F(MATRICE=CO('damp'),OPTION='RIGI_MECA_HYST',),
  ),
);

modes=MODE_ITER_SIMULT(
  MATR_RIGI=damp,
  MATR_MASS=masse,
  CALC_FREQ=_F(
	OPTION='CENTRE',
	FREQ=2,
	NMAX_FREQ=8,
  ),
);

IMPR_RESU(
  FORMAT='MED',UNITE=80,
  RESU=_F(RESULTAT=modes,NOM_CHAM='DEPL',),
);

raises the following error when trying to print the mode shape in a med file

   !-------------------------------------------------------!
   ! <EXCEPTION> <DVP_1>                                   !
   !                                                       !
   !    Erreur de programmation : condition non respectée. !
   !                                                       !
   !                                                       !
   !                                                       !
   ! Il y a probablement une erreur dans la programmation. !
   ! Veuillez contacter votre assistance technique.        !
   !-------------------------------------------------------!

jean pierre aubry

Last edited by jeanpierreaubry (2013-12-18 14:30:26)

Offline

#2 2013-12-16 00:02:07

Thomas DE SOZA
Guru
From: EDF
Registered: 2007-11-23
Posts: 3,066

Re: [SOLVED]error while writing a mode shape in a med file

Hi Jean-Pierre,

Please provide the files necessary to reproduce this bug.

TdS

Offline

#3 2013-12-16 07:42:29

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

hello Thomas

here are the files
simple test case, you know it!

jean pierre aubry


Attachments:
modes-damp-error.zip, Size: 2.99 KiB, Downloads: 237

Offline

#4 2013-12-16 21:55:18

CLF
Member
From: Sweden
Registered: 2009-01-27
Posts: 112
Website

Re: [SOLVED]error while writing a mode shape in a med file

Hi

I believe that the solver has a problem with hysteretic damping. Take a look at document R5.01.01.

If memory serves me correctly, you can use Rayleigh and possibly, viscous damping, but not hysteretic.

Sincerely
Claes

Last edited by CLF (2013-12-16 21:55:50)

Offline

#5 2013-12-17 07:35:52

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

hello

CLF wrote:

Hi
I believe that the solver has a problem with hysteretic damping. Take a look at document R5.01.01.

first
which part of  R5.01.01 are you talking off i cannot see such a statement in R5.01.01

second
does not look like a solver problem
frequency and damping are calculated with values in good agreement with the undamped structure values

besides that what parameter do you advise me to use for a Rayleigh analysis of a steel structure including some damping discrete element
AMOR_ALPHA = a_alpha ??
AMOR_BETA = a_beta ??


jean pierre aubry

Offline

#6 2013-12-17 18:53:35

Dav
Member
Registered: 2011-06-22
Posts: 53

Re: [SOLVED]error while writing a mode shape in a med file

Hi Jean-Pierre,

I don't understand very well how do you use you NUME_DDL function.
For me a calculation of a NUME_DDL is based over matrice elem.
So i would use CALC_MATR_ELEM in order to calculate the elementary rigidity matrice afterward i will reuse this function to calculate
the elementary rigi_meca_amor matrice from elementary rigi_meca matrice.

Next i would calculate NUME_DDL from this elementary rigi_meca_amor matrice.
But maybe it's just a formulation way.

Does you first calculation without hysteretic properties work?

Regards,

Dav

Offline

#7 2013-12-18 00:59:54

CLF
Member
From: Sweden
Registered: 2009-01-27
Posts: 112
Website

Re: [SOLVED]error while writing a mode shape in a med file

Hi Jean Pierre

Apologies, my mistake - I misread Symmetric Complex for Asymmetric Complexe.

Table 1 in section 3.1 and the text above does indeed state that you should be able to run the case with hysteretic damping if you use the QZ or SORENSEN methods.

As an anecdote. In the past, when running complex mode analysis using another FE solver, there was sometimes the problem when damping was significant. The support guys then stated Complex Mode solving to be a sport for gentlemen with time and resource to spare. Admittedly - it is fun.

About Rayleigh Damping. You have to calculate a best fit for the Alpha and Beta values to the physical damping you want as it cannot give you constant damping.

NAFEMS dynamics primer (quoted from memory) states something along the lines that Rayleigh Damping is a purely mathematical construct with no physical damping mechanism to mimic. Furthermore, it ceases to be reliable when damping becomes significant, say, at 10% of critical and higher.

As the saying goes - it is damping or the damn-d thing.

Sincerely
Claes

Last edited by CLF (2013-12-18 01:00:50)

Offline

#8 2013-12-18 07:33:53

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

hello

thanks to Dav and CLF

but i must insist again

Code_Aster does calculate the frequenies and the damping, and i believe them to be true, on a simple academic case
but it raises an ERROR while trying to PRINT on the MED file

jean pierre aubry

Offline

#9 2013-12-18 09:11:32

Dav
Member
Registered: 2011-06-22
Posts: 53

Re: [SOLVED]error while writing a mode shape in a med file

Hi,

Maybe stupid, but have you try to specify the model over you want to display the deplacements ?

Regards ,

Dav

Offline

#10 2013-12-18 09:41:00

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

not exactly stupid but, keyword MODELE is not allowed in IMPR_RESU since version 11.2.14, or around that

besides that i can print the med results for the undamped calculation

moreover an attempt to display the damped modes in STANLEY/Gmsh fails as well

and printing in ASCII in .resu file is ok

Offline

#11 2013-12-18 12:18:56

Dav
Member
Registered: 2011-06-22
Posts: 53

Re: [SOLVED]error while writing a mode shape in a med file

In your .resu file, the results are real or complex?

Your modes should be complex and you cannot display them like this.
Have you tried to specify the part of your modes you want to display?
In IMPR_RESU function, there is a logical key name "PARTIE", choose real or imaginary.

Regards,

Dav

Offline

#12 2013-12-18 14:29:24

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

Dav wrote:

In your .resu file, the results are real or complex?v

real

Dav wrote:

In IMPR_RESU function, there is a logical key name "PARTIE", choose real or imaginary.

ok that is it,
so obvious whan you think about it, afterwards!

thanks for the help

jean pierre aubry

Offline

#13 2013-12-18 14:46:57

Dav
Member
Registered: 2011-06-22
Posts: 53

Re: [SOLVED]error while writing a mode shape in a med file

No problem!

I don't know if it has changed with new version but i have already had this problem and the mess file returned me a message which say something like:
" Impossible to display complex results.
In IMPR_RESULT, specify the part of the result you want to display"

But sometimes this function didn't automatically appear in EFFICAS (Unfortunately hahaha)

Well, fine if it is ok!

Regards,

Dav

Offline

#14 2013-12-18 15:10:15

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,616

Re: [SOLVED]error while writing a mode shape in a med file

Dav wrote:

But sometimes this function didn't automatically appear in EFFICAS (Unfortunately hahaha)

i never use it!

Offline

#15 2018-04-28 08:38:07

praful
Member
From: Chengdu, China
Registered: 2016-01-07
Posts: 46

Re: [SOLVED]error while writing a mode shape in a med file

I was trying from the files in the forum. And,I am facing the same problem. Without output file, the program runs smoothly, but as I add the output file, it gives an error. Although, it calculates the frequency
IMPR_RESU(
  FORMAT='MED',
  RESU=_F(
    NOM_CHAM=('DEPL', ),
    PARTIE='REEL',
    RESULTAT=modes
  ),
  UNITE=80
)


Attachments:
message.rar, Size: 6.83 KiB, Downloads: 131

Prafulla Malla

Offline