CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   Newbie Problems (http://www.cfd-online.com/Forums/ansys-meshing/85460-newbie-problems.html)

 West February 25, 2011 22:34

Newbie Problems

Hello guys, my first time here, so I'm sorry if my questions are to stupid..
I have to run a simulation to find the drag coefficient of various types of tennis balls considering the fluffs and the lines they have.. something like this:

I tried to make the geometry on ANSYS Design Modeler, but all i could make was a sphere. What i have to do to make those lines on the sphere? And can make the fluff on this program, or i have to use another one. I hope someone can help me. Thanks.

 PSYMN February 26, 2011 07:51

Surface roughness...

Making the lines will be an interesting challenge for any CAD tool... Sorry, no idea.

But for the fluff, I think you want to use a surface roughness boundary condition (rather than actually modeling the fluff). Separate the surface of your smooth sphere into "fluf" and "lines" boundary conditions and set different surface roughness so the solver handles the boundary conditions differently.

Keep us posted, I am interested to hear your final solution to this tricky CFD problem.

 West February 26, 2011 15:47

Thanks for the attention in first place.
I understood, but to set the roughness like you said i have to delimitate the lines on the sphere, don't I? And I don't know how to do so..
And which command I use to set the roughness of the surface?
Sorry if it's too stupid.
Thanks.

 PSYMN February 26, 2011 18:54

Right, you would first need to draw your boundaries on the sphere so you can break it up into separate boundary conditions... Sorry, I don't know how to do that.

Once you have them as separate parts, you can mesh and export to the solver. Then in the solver you set those as viscous walls and adjust the setting for surface roughness (under boundary conditions => Edit).

 West April 26, 2011 20:03

I made the geometry using Solidworks:

http://img96.imageshack.us/i/semttulowyq.jpg/

but now i have another problem, i can't import it to the ANSYS Design Modeler. I get an error that says:
Attach Failed"

Can somebody help me?

http://img96.imageshack.us/i/semttulowyq.jpg/

 PSYMN April 27, 2011 06:04

But I can tell you that CAD interfaces require royalties to the various CAD vendors and are therefore usually sold separately. It appears to be saying that you do not have a Solidworks CAD interface. (or perhaps your license is just used up or not running).

If you agree that you don't have a Solidworks CAD interface, you could look into getting one or output in a different format (such as ACIS (SAT)).

Simon

 West June 18, 2011 13:55

Hello again..

Got help from a friend and I did simulate it like I wanted. But, now I have to calculate the drag coefficient of the ball, and I'm having some problems doing that. Can anybody help me? Thanks..

 PSYMN June 18, 2011 20:05

That depends on your solver, some have "Drag" listed... In others you can just sum the force in a particular direction (such as X) over the entire surface...

 West August 18, 2011 19:42

Quote:
 Originally Posted by PSYMN (Post 312592) That depends on your solver, some have "Drag" listed... In others you can just sum the force in a particular direction (such as X) over the entire surface...
Humm, I don't know if I'm doing it right, cause i'm getting the ball's drag coefficient 0.21, and that's half the expected..

this is my expression on ANSYS, is it right?

Cd = -2*(force_z()@Ball + force_z()@Lines)/(1.225[kg/m^3]*(1[m/s])^2*(0.003201928904[m^2]))

 PSYMN August 18, 2011 23:40

It is late and I don't want to dig out a book to really confirm the expression, but it looks ok (assuming that "Ball" and "Lines" are the two surface components of the tennis ball and you are just summing up the forces)...

However, It could be a number of things.

Perhaps your ball really does have less drag due to the fuzzy surface you are modeling, or maybe the ridges break up the flow?

Perhaps you don't have these surface treatments right? Or perhaps other bocos or properties are wrong?

Or perhaps your units are not right?

Have you tried modeling it as a smooth sphere (just change the wall condition), does that give you what you expect? Have you checked the expression in a book or asked about it on the solver forum?

 West August 19, 2011 00:24

Thankyou for the fast reply at first..
I did chech the expression and its correct, as the units..
I'll try to do what you said to the smooth ball, but what i'm trying right now is to raise the flow velocity, which is 1m/s right now.. i guess the reynolds number is too low to apply that formula, and maybe raising the velocity it will help.. can u please detail what u mean by this:

"Perhaps you don't have these surface treatments right? Or perhaps other bocos or properties are wrong?"

the only thing i did to the surface was apply the condition "Rough Wall" and set the Roughness Height (or something like that) to 0.65mm (which i estimated as the height of the fluff on the tennis ball) i don't know if there was any other thing to do... so..

thank u for your help, again

 PSYMN August 20, 2011 23:51

New territory

Right, I just meant that maybe with that surface roughness, you get that Cd...

Perhaps it is not the right way to simulate the fuzz accurately. Perhaps it is not really intended (or calibrated) for the fuzzy nature of a tennis ball and would be better suited to a more granular solid surface roughness? I am not sure, but I just mean that a Cd being off doesn't mean your expression is wrong, there could be many factors. Maybe your Cd expectation is wrong? It is hard to tell.

Trying a smooth ball (more predictable and backed up with a lot of experimental and simulation data) will give you some confidence in your base level and then you can move forward. you could also check if anyone else has done any work with fuzz and see if they say what sort of boco they used to simulate it... It would be great if someone has already done some physical testing to calibrate the CFD... Or maybe that is what you will end up doing... (Adjust the surface roughness until it meets experiment over a range of velocities and conditions and then write a paper on that...)

 West August 20, 2011 23:54

Thanks man, I'l try what you said.

 PSYMN August 21, 2011 00:03

One other thought...

Any time you are dealing with a ball in a flow, it is likely an unsteady solution...

If you are doing a steady state solve, you can pretty much bet that you will get a different answer every time you run the solver... This also makes it very difficult on convergence, etc.

Make sure that you are not just trying to solve this with a steady state solution (as you might for an airfoil).

Simon