Solidworks Flow Simulation: Recovery of Forces on a Body

Simulation Fluid Mechanics
,
1

Hi all

 

I'm trying to simulate the air penetration of a simplified profile portion with Solidworks Flow Simulation.

I try to recover the forces applied to the last in an air flow at 20m/s. This then allows me to calculate the SCx of the shape.

I'm getting surprising results... The file (Solidworks 2014) is in PJ.

The geometry tested:

 

I use the assistant to set up an air flow at 20m/s according to Z:

 

Balance sheet:

 

How can we explain this effort of 1.45N following X?

Since the simulation is symmetrical, it should not exist. The problem remains if I refine the mesh.

Any idea of what is the problem?

Thank you in advance for your help.


cobra_simplifie.sldprt
2

Good evening

 

You have attached screenshots that are almost illegible.

You have to put all your images in a ZIP file.

For your value in X you have to look at how your object is fixed in relation to the flow (angle or not) because given your shape you could perhaps have an intrados effect.

Kind regards

 

 

3

Thank you for your feedback.

Weird for the screenshots, at home they are visible.

I put them back in the PC.

The object is well positioned in the axis of the flow, no problem on that side.

 


pb_solidworks_flow_simulation.zip
4

Good evening

I also meant totally constrained!

What type of connection did you put in place to fix your room?

What is the size of the area where you should make your simulation (the size of the box in which the flow will take place) and especially how is your part centered in this rectangular parallelepiped.

Kind regards

5

Thank you for your feedback!

 

>>>I also meant totally constrained!

Your answer challenges me, I thought that the geometry remained fixed during the fluid simulation. Is this not the case?

>>>What type of connection did you put in to fix your room?

None for the moment. 

>>>What is the size of the area where to do your simulation?

This is the area automatically created by Solidworks, see dimensions below and in PJ.

I just tested by multiplying each dimension of this area by 2: I still have a parasitic force of 1N according to X.

 


dimensions.zip
6

There is always an X-factor in a calculation, you have just given us the value! Bvo! #meilleurréponse

7

Good evening @ eric9.benoit

How many nozzles did you put in? I have the impression that you have only put a few.

Can you take a screenshot of the nozzle flows from the top view.

Kind regards

8

Good evening everyone

Your problem aroused my curiosity, so I tested your model !

I think the problem is the networking. I tried with more than 1 million stitches (116,000 in your calculation) and there, the force following X is 0.16 N

With 2 million meshes, X = 0.08 N

In my experience, meshing is very important with flow, you should not hesitate to do several calculations, more and more meshed, until the results stabilize.

And in relation to Zozo_mp's  remark, in flow it doesn't matter how fixed the part is (free, fixed or with constraints). it  has no impact.

1 Like
9

Hello @ Nibblerg

Thank you for the information on the fixing of the part and especially on the fineness of the mesh.
I have the impression that the discretization is not the same between simulation and flowsimulation.

Have a good week

Kind regards

10

>>>Can you take a screenshot of the nozzle flows from the top view.

No problem, the flow is very homogeneous.

 

>>>I think the problem is the networking. I tried with more than 1 million stitches (116,000 in your calculation) and there, the force following X is 0.16 N. With 2 million meshes, X = 0.08 N

Great, thanks for this info, I'll try it!

Thank you so much to both of you!

1 Like
11

I tested, indeed by strongly refining the mesh, the following force X decreases. Thank you!!!!

On the other hand, by setting the mesh to the finest, I only have about 500,000 meshes.

How do you specify an even finer mesh?

12

For a finer mesh:

The table you show me is the "global mesh", it will act on the entire field of calculation, In it, you can uncheck "automatic parameters" and you will be able to manually define the number of meshes in each XYZ direction

Then, to go even further, you can create a "local mesh", by selecting your room for example. And there, you will be able to finely mesh the fluid cells close to the contour of your part.

There are really a lot of parameter options for the mesh, hard to explain them all! 

1 Like