Hi all,

I was wondering if anybody could advise me on creating 2d model in DesignModeler? I am new to CFD and want to be able to create a 2D simulation of an aerofoil to compare with test data, however I am having problems creating the model!

I have created a coordinates file for the NACA 0012 aerofoil I want to model, and am inputting this into designmodeler using Create => Points. I am then using the lines from points tool to create the top and bottom sides of the aerofoil, and then creating a surface using Concept => Surfaces from lines. I then create a flow region in a similar manner, and cut the aerofoil from the flow region using BodyOperation.

If I try to proceed to CFX mesh from this however, it tells me that CFX mesh only works on Solid Body Volumes. (I have selected 2d simulation in the options part of workbench)

I guess I am doing something very basically wrong. I thought I should be creating sketches in designModeler, however, I cant create a sketch from the points created in the coordinates file! If anybody has any advice or useful information for me it would be very much appreciated. Once simulated, I hope to calculate drag and lift forces of the aerofoil to compare with test data, so if anybody has any information on how to do this I would also be very grateful!!!

Sorry for the long post!

Thank you in advance!

If you have access to FLUENT you could try that (also). http://instruct1.cit.cornell.edu/cou...foil/index.htm kind regards Michael Bo

Unfortunately I only have access to CFX...

You have to extrude one layer of you surface (giving a volume with one layer one layer of cells) since CFX is a fully 3D solver and not very suitable for 2D cases.

Thats great thanks- do you know how to extrude a line body, for instance though? The points created for the aerofoil (Create => Points) don't seem to be able to be selected when creating a sketch. (I assume what I need to do is create a sketch of the aerofoil and then extrude the sketch an arbitrary distance, say 5mm (The aerofoil is 150mm cord length).

In DesignModeler I can only use the "Lines from points" function to draw lines from the points, and then create a surface from the lines. I can't, however, extrude the "line body" or "surface body" to create a 3D aerofoil.

Is it possible to make the "sketch" function use the points I have created, or is there some other way to create a 3D solid volume using the surface from lines function?

Hope that makes sense!!!

As always, any advice is appreciated!

Thanks again.

Just forget about DesignModeler. I suggest to use any convinient CAD program (Rhino for example). Create geometry using it and then import to DM. And why 2D model? If you use CFX, than use 3D model.

Please refer to "NACA 2412 simulation in CFX" at this forum. You are most likely wasting time with your approach. Robin's comments are very specific about it.

Andrew

Thanks- hadn't seen that post. The reason for the CFD analysis is to compare CFd results with results I am taking in a wind tunnel, for an aerofoil over a range of 180deg AOA, to simulate an aerofoil being used in a VAWT.

I presume that the best method is to use a 3D, rather than 2D model, that as closely matches the set up in the wind tunnel as possible (ie tunnel walls etc)?

I have access to Solidworks- am i right in thinking that I can import models I create in Solidworks into CFX-mesh?

All the more, you should use 3D model. First, I advice you to read theory about CFD. You are partially right^ you will be able to import your solidworks model to DesighModeler and then use CFX-mesh or ICEM

why only partially right? presumably the results wont differ, no matter what CAD package I use to create the model, provided the geometry is the same? I've tried importing solidworks geometry as a STEP file and it seems to work okay...

Thanks,

Andrew

The reason a 3D model was suggested is that the flow in the wind tunnel is probably not 2D. So 2D is an approximation to the real thing. CFX is a fully 3D solver but you run 2D problems by creating an extruded mesh 1 element thick. The element depth should be approximately equal to the minimum of the streamwise or cross stream element size.

It will be very difficult to get any near correct results beyond stall up to 180 degree angle of attack. You might get up to stall and just beyond stall, but that's about it. Convergence will be tough at angles just beyond stall due to the highly transient nature of the flow. You may have to run transient calculations with high mesh resolution and average the results over time, which is expensive. Try lower angles of attack first and see how it goes.

Read the postings from Robin as well, they are relevant to what you are trying to do.

Neale