applying CFD loads to FEA mesh 2

[rb1957]

we use CFD and a CFD model/mesh to determine the applied airploads.

then we have our FEA and it's mesh.

To apply the CFD loads we have in the past run the CFD model with the FEA mesh and of course gotten different answers.

I was thinking that it should be "easy" to interpolate an FEA grid location onto the CFD mesh (like the 4 nearest grids) (or maybe from the four nearest grids that box in the FEA grid determine the pressure surface locally), and so calculate the pressure at each grid.
This won't be "perfect" but it should be better …

thoughts ?

another day in paradise, or is paradise one day closer ?

 

[DeiviG]

Hi rb, what software are you using?, I'm solving right now a multiphysics problem with ansys it's not in aerodynamics, but we transfer CFD loads to FEM model. Using Ansys there are 2 methods to solve those kind of problems. 1 way-direction, which means you solve the CFD problem first, and then transfer your loads (either thermal or pressure) to the mechanical problem, from my perspective this method is simpler compared to 2 way-coupling, which is the second method. In this last method, you solve CFD and FEM simultaneously and those solvers are communicating with each other all the time.

Now, considering the mesh, as long as you guarantee a good y+ and wall mesh sizing, your results should be fair enough.

 

[rb1957]

I'm using NASTRAN to solve the structure, and someone else is using a CFD-optimised model to obtain the airloads. The point is I don't want to use the same model for both …
1) the CFD model will have a lot of detail where I (the structure's guy) don't want it, and
2) the structures mesh doesn't have detail where CFD wants it.

On another project we ran the FEA mesh through the CFD analysis and got very different results (you'd've thunk it !?), that had to be "hand waved" (conservatively apply the highest load factor to all loads directions).

another day in paradise, or is paradise one day closer ?

 

[BlasMolero]

Hello!,
Using FEMAP with Simcenter Nastran you can import your CFD loads and apply to your structural mesh using interpolation methods, the mesh don't need to be the same type.
The command MODEL > LOAD > MAP OUTPUT FROM MODEL... allows “mapping” of output data from 2-D and 3-D elements in one model (Source) onto another (Target). The two models can have completely different meshes and FEMAP gives offers a number of options and several differ­ent “mapping approaches” used to “map” the output.

Also you can run the DATA SURFACE EDITOR with plenty of options to import loads. “Data Surfaces” allow you to apply vari­able loading conditions using a number of predefined methods. Each method has a unique dialog box with all of the required inputs. FEMAP uses the information from a specific dialog box and creates a table of values which is saved with the model and can then be applied to any number of loading conditions:

Best regards,
Blas.

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

IBERISA
48004 BILBAO (SPAIN)
WEB:

http://www.iberisa.com

Blog de FEMAP & NX Nastran:

http://iberisa.wordpress.com/

 

[rb1957]

thx, just what I was looking for ! was thinking "data surface" should do something for me … but didn't take it past a random thought.

another day in paradise, or is paradise one day closer ?

 

[IceBreakerSours]

I have nothing to add to the thread but I am curious: Are you mapping a transient spatial field from a CFD job into your FEA job? Or, is it just the end point you are more interested in?

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[rb1957]

I don't understand the question ...
someone runs CFD and gives me airloads (pressures), steady state.

another day in paradise, or is paradise one day closer ?

 

[Michal Galuszka]

Hi!
Please take a look on CFD pressure mapping tool. It’s a stand-alone app I have written in C#, it might be useful to solve your issue.

https://github.com/galuszkm/CFD_Mapper

Currently only LS-Dyna and MSC Marc are supported but I can easily extend it to other solver. If you need some support feel free to contact me.

Best regards,
Michael

 

[rb1957]

thx but the help (and YT) has shown me how to do this in FeMap ... dead simple.

another day in paradise, or is paradise one day closer ?

 

[TG_eng]

rb1957, how are you given the CFD airloads? Are they in a CSV file as X Y Z coordinates and pressure value?

Thanks in advance

 

[rb1957]

yes, pressures at CFD points. It seemed like a reasonable thing for the pre-processor to do ... to understand two meshes on the same surface, and to interpolate from one mesh to the other. It seems QED (Quite Easily Done).

another day in paradise, or is paradise one day closer ?

 

[TG_eng]

I wrote an API awhile back that automated reading CFD data from a csv file. The purpose was to create multiple data surfaces from the x y z points and multiple columns of pressure values. Then it created the corresponding load sets with a surface pressure load of 1 using the data surface input so when solving or exporting the analyses, the load would expand to the elements attached and interpolate based on the data surface. The API depended on having geometry to apply the loads to.

I shortened the API to just read the the CFD data and create the corresponding data surfaces. I have attached the files if you are interested. And I can always expand on this if I have a specific direction.

 

[Burner2k]

[hx but the help (and YT) has shown me how to do this in FeMap ... dead simple.]
rb1957, if possible, could you post links to the YT videos which helped you to accomplish it in Femap?

 

[rb1957]

sure, see pic ...

another day in paradise, or is paradise one day closer ?

 

[Burner2k]

Thanks...I managed to find the same video as well. Good one...too bad AFAIK, Patran does not come with such easy tools.