CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLOW-3D (http://www.cfd-online.com/Forums/flow-3d/)
-   -   Hull modelling? (http://www.cfd-online.com/Forums/flow-3d/27358-hull-modelling.html)

Adam Beamish March 5, 2009 16:45

Hull modelling?
 
Hello,

I'm a mechanical engineering student attempting to model several trim vane designs for a military amphibious vehicle. I'm familiar with NX 5, Gambit, Fluent, and the SolidWorks suite, but can't find a lot of resources for FLOW-3d.

Are there any tutorials or resources available for this type of problem? Could someone assist me in setting up a simulation?

Regards,

Adam C. Beamish Mechanical Engineering IV

HC March 6, 2009 12:11

Re: Hull modelling?
 
You may find some information at http://www.flow3d.com/apps/mar/app_mar_hull.html http://www.flow3d.com/resources/news...ring08_03.html

In FLOW3D, you can do some simple designs by using the primitives and the quadratic function, and if the designs are complicated, you might need to turn to CAD software to create the objects, export them into the STL format and then import them into FLOW3D under the meshing&geometry tab.

Then setup an appropriate mesh to resolve the objects in FLOW3D, turn on models such as turbulence and moving object, and apply the correct boundary conditions (like fixed velocity, fixed fluid height, wave boundary etc) around the mesh.

Adam Beamish March 6, 2009 14:37

Re: Hull modelling?
 
Thanks for the advice, HC.

I've already got the geometry modeled and exported, but I don't know how to ascribe weight or density to it.

Additionally, I'm not sure how to set the boundary conditions such that the control volume STARTS with water in it (i.e. isn't filling)

The earth race modeling is almost EXACTLY what I want to do, only with the waves on a much smaller scale. If you have some more detailed instructions, I'd REALLY appreciate it.

HC March 6, 2009 16:21

Re: Hull modelling?
 
You need to expand the moving component's tree and the "Type of Moving object" sub-tree in the meshing&geometry tab. There you can specify the motion freedom and uniform density or total mass with inertial tensor. If the object has uniform density specified, the inertial tensor will be calculated by the software automatically. For more information on how to calculate inertial tensor, please go to http://en.wikipedia.org/wiki/Moment_of_inertia

To set the initial water in the domain, you can go to the Initial tab, and specify the fluid height in Z direction, which is the value of the elevation of the water surface.

About the wave boundary condition, you can click the top menu "Help > Content" and search with the keyword "Wave Boundary". The sides may use symmetry boundary and the downstream will have an outflow boundary. Also, you may need expand the domain length and width in order to minimize the reflection.

Adam Beamish March 6, 2009 16:41

Re: Hull modelling?
 
Thanks again.

How can I make my hull display in the preprocess? Right now all I see is the fluid filling the control volume, and my component isn't even in the object list.

For some reason, even though I set the initial fluid height to 0, when I preprocess the simulation, the fluid height doesn't seem to change. Is there a way to initialize (like in fluent) to force the program to rebuild ALL data?

Thank you so much for your help.

HC March 6, 2009 16:45

Re: Hull modelling?
 
Did you include the "Solid volume" when you render the results in 3D?

What is the Zmin value of your mesh? Do you have any other fluid regions created in the Initial tab?

HC March 6, 2009 16:48

Re: Hull modelling?
 
After you set the fluid height to 0 and preprocessed the setup, did you reload the result?

Adam Beamish March 6, 2009 16:59

Re: Hull modelling?
 
Yes. I always reload the result.

No, I only have one fluid region.

Zmin value of mesh is -10 000.

I assume the initial height is a percentage of the Z domain? In messing around with it, it seems that 0.5 corresponds to half of the control volume being filled.

I just turned on solid volume in the iso-surface overlay options, but the result is very strange.

I've uploaded some screenshots to http://flickr.com/photos/fleamo/sets/72157614811025473/

Maybe you could take a look?

Thanks.

Adam Beamish March 6, 2009 17:02

Re: Hull modelling?
 
Also, I've given my solid some velocity properties. I've set it to prescribed motion, and 1.0 in the Y translation velocity, but the solid doesn't seem to move at all when I run the simulation.

Adam Beamish March 6, 2009 17:05

Re: Hull modelling?
 
Could I perhaps send you the simulation file via email?

HC March 6, 2009 17:14

Re: Hull modelling?
 
From the screenshots you uploaded, it seems like the cell size is too big to resolve the moving object.

The reason why you didn't see the motion of the object was probably due to the finish time. As it seems like you have a big domain and big cell size, if the finish time is small, the distance the object traveled would not be that obvious compared to the domain/cell size.

You can try a much smaller domain and also reduce the cell size, to see if you are going to observe the same thing.

FYI, Flow Science provides general training class this April.

Adam Beamish March 9, 2009 11:38

Re: Hull modelling?
 
Thanks for the info. Unfortunately, if I don't have this whole thing sorted out by April, I'll be in trouble.

Any thoughts on why only a part of the solid body was visible, seemingly floating above the fluid surface?

I DID manage to get the body to move (by increasing initial velocity substantially). However, it doesn't seem to interact with the fluid in any way.

I was expecting waves and disturbance, but it was completely still.


All times are GMT -4. The time now is 12:04.