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

You are not logged in.

#1 2018-06-19 11:05:34

ingladisa
Member
Registered: 2017-12-31
Posts: 62

First impressions after six months of Aster learning and practicing

I remember very first impact with Salome-Aster, at the beginning of this year:
It is french, but with english language... challenging, but ok
It is quite complicated, ok, I try to follow tutorials...
At this point it is "mandatory" to cite very useful Cyprien Rusu (FEAforAll) videotutorial, I think fundamental to understand basis of Salome-Meca logic. Aster has a lot of documentation and training material, but it is necessary to understand it's "language", its logic. It is useful also to have a previous significant experience in FEA with codes at level of NASTRAN or ANSYS and a good experience in meshing and modelling. Code-Aster is at this same level, but with the advantages of an OpenSource.
Code-Aster is an advanced code, Salome is a good user interface, very good being an OpenSource, but not at level of last User Interfaces available for ANSYS or similar software, i.e. APEX for MSC NASTRAN, it has not the same level of "automation" in meshing and model definition, but in any case we are comparing an OpenSource with very very expensive software.

After first learning months, I began to appreciate Code_Aster capabilities: a very nice management of stresses and efforts in elements, very precise results. Code_Aster demands an input .comm file perfect, it is much less tolerant than nastran for example, but when it runs it is a swiss clock.
I apreciate management of the mesh, of its elements from 0d to 3d, it provides very easy management of composite models with plates and solids, for example.
Also the use and results accuracy of RBE3 I found a bit better than in other codes.
The only true defect I found, that I wrote in this forum few time ago, is that solid linear mesh displacement results seems to provide a result not so precise when compared with other software and with hand calculation (where possible). This aspect wold need a further deep investigation...

At his point I am vey satisfied of starting learning cod_Aster, I start to use it for my jobs where possible, and I am thinking to use it for some research studies I have in mind. At the end, the best way to learn is to apply it.

Last edited by ingladisa (2018-06-19 11:17:38)

Offline

#2 2018-06-19 14:42:48

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

Re: First impressions after six months of Aster learning and practicing

Hello,

Thank you very much for this feedback. This is the kind of feedback that encourages us, the development team, to improve further the software.

Here are a few comments below.

ingladisa wrote:

I think fundamental to understand basis of Salome-Meca logic. Aster has a lot of documentation and training material, but it is necessary to understand it's "language", its logic.

This sentence really captures what code_aster is about. It has indeed a logic and it comes with its own programming language (the Aster command language) on top of Python that needs to be understood so that the learning curve gets easier and easier.
The upcoming next generation version of code_aster that will add a fully python compatible module to that mix will improve further the user experience on that aspect.

ingladisa wrote:

It is useful also to have a previous significant experience in FEA with codes at level of NASTRAN or ANSYS and a good experience in meshing and modelling.

This is clearly an advantage currently. We do think though that things like AsterStudy will be changing that observation in the future helping newcomers to numerical simulation get into code_aster.

ingladisa wrote:

Code-Aster is at this same level, but with the advantages of an OpenSource.
Code-Aster is an advanced code, Salome is a good user interface, very good being an OpenSource, but not at level of last User Interfaces available for ANSYS or similar software, i.e. APEX for MSC NASTRAN, it has not the same level of "automation" in meshing and model definition, but in any case we are comparing an OpenSource with very very expensive software.

This is also our reading currently. We're working hard to bridge the gaps.

ingladisa wrote:

The only true defect I found, that I wrote in this forum few time ago, is that solid linear mesh displacement results seems to provide a result not so precise when compared with other software and with hand calculation (where possible). This aspect wold need a further deep investigation...

We may have missed that thread. Could you reference it here so that we may have a look?

ingladisa wrote:

At his point I am vey satisfied of starting learning cod_Aster, I start to use it for my jobs where possible, and I am thinking to use it for some research studies I have in mind. At the end, the best way to learn is to apply it.

Good news!

Kind regards
TdS

Offline

#3 2018-06-19 15:08:08

ingladisa
Member
Registered: 2017-12-31
Posts: 62

Re: First impressions after six months of Aster learning and practicing

Offline

#4 2018-06-25 21:46:36

ingladisa
Member
Registered: 2017-12-31
Posts: 62

Re: First impressions after six months of Aster learning and practicing

Today I pulled the neck up to Code_Aster:
The model: four floors masonry building, walls defined by hesaedric solid mesh, floor planes by 2D square elements. Mesh element side dimension 10cm (!).
input frequency range 0.3 to 8 hz, method: Lanczos  (default Sorensen method did not converge)
Computation mode: parallel MUMPS with 6 CPU and 12 Gb (i have a total of 16 Gb on i7), for a total of almost 20 minutes wit cpu fans at full throttle... All ok at the end, after some unsuccessful runs with Sorensen (faster but with quite less convergence capability in this application).

Last edited by ingladisa (2018-06-25 21:57:40)

Offline

#5 2021-08-18 06:44:21

sihaqqi
Member
Registered: 2018-05-11
Posts: 14

Re: First impressions after six months of Aster learning and practicing

what about its usage for people who do not have knowledge of French language. It is easy to go through all warning sand errors in Ansys but French translation from google for these issues can lead to wrong conclusions as the translation is a machine translation.

Is it possible to get all messages and error warnings in English in Code Aster?

Last edited by sihaqqi (2021-08-18 06:45:16)

Offline