Femap and Patran

[ThomasH]

Hi everybody

First of all, this is not a tread asking if Femap is better than Patran or viceversa. That question probably has no easy answer.

Instead, I might become involved in a project where both softwares will be used depending och who does what. The question is can Femap and Patran be made to "talk" to each other efficiently?

Obviously we can communicate through Nastran bulk data but how about for example Patran neutral files. Can we make the information not available in the Nastran file (like groups, colors etc) available or is this probably a fruitless effort.

I couple of year ago I was involved in a project where models were moved from one to the other but we did it once. And it was a lot of manual "fixing" involved. Hence, I am very aware of the problems that can exist. So, I doubt that this is an efficient work method. Doing some of it in Patran and some of it in Femap and moving the models back and forth on a daily basis.

But I may be to pessimistic and I would value your opinions or experiences.

Thomas

 

[BlasMolero]

Dear Thomas,
There are three direct interfaces between FEMAP and MSC.Patran. You can write a FEMAP model to the MSC.Patran format for analysis, read an existing MSC.Patran model, or read analysis results for post-processing. The model interfaces use the MSC.Patran neutral file format.

A powerful feature of FEMAP V11.1 that helps a lot to open legacy models from Patran in FEMAP is "SURFACE FROM MESH" command. You can add cutouts, stiffener locations, whatever is required and re-mesh, as geometry & mesh is fully associated, really powerful:

Also, another great feature of FEMAP V11.1 is "MODEL MERGE" command that allows to copy entities from one model to another with complete control over inclusion and renumbering of:
[ul]
[li]FE Entity types[/li]
[li]Boundary conditions[/li]
[li]Materials[/li]
[li]Properties[/li]
[li]Groups[/li]
[li]Layers[/li]
[li]Analysis set[/li]
[li]Model orientation and transformation control[/li]
[li]Can also be used to transfer frequently used Materials, Properties, Layups etc. from existing models to new models[/li]

Also "ATTACH RESULTS" command is really powerful in FEMAP, you can share between FEMAP & PATRAN the nastran results *.op2 files without “internalizing” the contents of each output file *.op2 into the FEMAP database. This is especially helpful when you have very large output file such as transient response, frequency response, and nonlinear analysis.
Best regards,
Blas.

~~~~~~~~~~~~~~~~~~~~~~
Blas Molero Hidalgo
Ingeniero Industrial
Director

IBERISA
48011 BILBAO (SPAIN)
WEB:

http://www.iberisa.com

Blog de FEMAP & NX Nastran:

http://iberisa.wordpress.com/

 

[rb1957]

i'd ask the question "why?" ... both Patran and FeMap are pre- and post-processors around an FE engine (like Nastran) ... ie they both do the same job, so why have one talk to the other ?

Quando Omni Flunkus Moritati

 

[ThomasH]

Hi again

The problem is as follows.

We have a project that needs an analyst 100% but we will probably have two guys involved. One with a lot of experience in Femap, the other with a lot in Patran.

Say that they work 50/50, will 50+50 equal 100% or will the model conversion take a significant amount of time. My guess is that it will take some time but I can't put any percentage of that part. Personally, I don't beleive in the idea.

Another idea from somebody was to just buy another licence of Patran and temporarily convert the Femap guy. My opinion was that it was a bad idea. I don't think you just swap softwares and become 100% productive overnight.

I just want to make sure that the expectations are reasonable. Otherwise we will have two very unhappy analysts in the end.

Regards

Thomas

 

[rb1957]

splitting the job won't work out 50/50 ... more like 60/60 or maybe 80/80 ... ie there's a cost to having more people and different s/ware.

converting one analyst to the other s/ware should be a non-starter (since it sounds you've a very short fuse on this). either would (should?) be able to pick up the other s/ware, but they'll be much slower untill they get up to speed (using the s/ware for some hundreds of hours)

Quando Omni Flunkus Moritati

 

[ThomasH]

rb1957:

I agree, it won't be 50/50. If it ends up being 60/60 or 80/80, I don't know. Answering that question is the problem I am trying to avoid. But there is definetely some cost involved with this solution. And I want that cost visible now. Not as a surprise in a few months.

Anyway, thanks for your input.

Thomas

 

[rb1957]

dividing it 50/50 makes the cost invisible.

maybe budget 50/50 and have a 50% contingency ?

maybe budget "if one analyst then X hrs; if 2 then X+Y"

then you can discuss (argue?, or just plain cuss ??) over the extra cost

Quando Omni Flunkus Moritati

 

[GregLocock]

Sounds like somebody needs to read "the mythical man month". Having said that we do have a core modelling group, who spit out meshes, and then the analysts fine tune the meshes and run them.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?