CFD non axisymmetric design external pods etc.....

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Hi,

You need to look at some general purpose, unstructured CFD codes like Fluent, Star, OpenFoam, etc that can model any arbitrary geometry. These work well for automotive and industrial problems. Not sure how well they work for the Mach number regime you want.

Elmer is also an open-source general CFD code. I've taken a look at OpenFoam and Elmer, but got a bit too complex for my taste. :) I'd be very interested in hearing if you have success running CFD on the design.

Cheers,
Sampo N.
 
The pic's for your job look fantastic! I use to run Fluent at ATK before I retired. It is a nice code. I am not sure that you have to go that route to analyze your project. You might want to check with Andy Woerner, the previous owner of What's Up Hobbie's. He built and flew a US Space Shuttle scale model that was very impressive. I think that he might still be making this model for paying customers.

Not only do you have the trouble of running Fluent for this model, but you need to grid it first. For a 3-D problem you could easily being looking at a grid that has 5 million grid points, a very formidable problem. I found that running Fluent on the company machine was difficult around the 5 million grid number. I would do everything possible to cut down on the size of the problem. For example, on a problem like this, if you don't have any weird wind or other boundary condition, you could get by with 180 degree symmetry. This would cut down on the size of the problem by 1/2. One time I was able to go between fins on a 4-fin rocket (45 degree symmetry) and I was able to cut down the size of the problem by 1/8. The problem was still around 6 or 9 million grid points and this was a steady-state solution. I had to run it twice, once for subsonic and once for supersonic. I did include the plumes of the rocket motor firing. I doubt if Andy Woerner did all this. It is a fun problem, if you have the CFD code, the computer and computer time, and your own time.
 
Last edited:
Not had chance to get into fluent yet, my knowledge of cfd is equal to nill. I have trouble working the phone never mind anything this complex....but persistance may pay off ...or the laptop will get a battering. Just getting it up and running on the lappy is a mission in itself........anyways put some more pics up so enjoy....IMAG0150.jpg11.jpgIMAG0320.jpgIMAG0329.jpgIMAG0330.jpg
 
Most impressive! :eyepop:

So ... how many motors are going to be used to put these things into the sky?

Greg
 
Did you get through your CFD work yet? In a sense I am more curious about that then the rocket itself. What CFD program allows you to send a Google Sketch Up file into it?
 
The pic's for your job look fantastic! I use to run Fluent at ATK before I retired. It is a nice code. I am not sure that you have to go that route to analyze your project. You might want to check with Andy Woerner, the previous owner of What's Up Hobbie's. He built and flew a US Space Shuttle scale model that was very impressive. I think that he might still be making this model for paying customers.

Not only do you have the trouble of running Fluent for this model, but you need to grid it first. For a 3-D problem you could easily being looking at a grid that has 5 million grid points, a very formidable problem. I found that running Fluent on the company machine was difficult around the 5 million grid number. I would do everything possible to cut down on the size of the problem. For example, on a problem like this, if you don't have any weird wind or other boundary condition, you could get by with 180 degree symmetry. This would cut down on the size of the problem by 1/2. One time I was able to go between fins on a 4-fin rocket (45 degree symmetry) and I was able to cut down the size of the problem by 1/8. The problem was still around 6 or 9 million grid points and this was a steady-state solution. I had to run it twice, once for subsonic and once for supersonic. I did include the plumes of the rocket motor firing. I doubt if Andy Woerner did all this. It is a fun problem, if you have the CFD code, the computer and computer time, and your own time.

Are there any packages that allow a linear (source/sink/vortex) model to be computed like this? Something similar possibly to AVL but tailored more towards rockets? I'm sure that with enough tinkering, you may even be able to get a good result from AVL. Granted, linear models only work at small AOA/sideslip, but if you go beyond that, then I think you've got bigger problems. Such a linear model would be trivially fast on even a slow modern computer.
 
I have tried to use Khamsin in Google Sketch-up but it's mind numbingly difficult to figure out how it works. There are no vids and the user manual is nil.....I you're a cfd expert and have worked in high end engineering all your life and grown up with CFD you prob going to be ok.....if your a sound engineer with an interest in aviation....well ...your pretty much ******. :) I need a morons guide to cfd.....all these software systems and great and I'm not knocking them, but you really really need to know what your doing with them. I don't learn very well unless someone else shows me on a one to one basis....I need an expert to sit me down for 6 months and show me what it all means. You can't learn this stuff in a few days...if at all. Unless I wish to pay someone to show me how to use this software....'it aint happening'
 
although I have a friend over in Australia who is running this through other software.....hoping for some results......
 
I cannot believe what I see here... just FANTASTIC... !!!

More please!!!!
 
Glad to see progress made! This is some killer work!


Later!

--Coop
 
Ok...The Google sketch up did not work. 2 problems. The skin wasn't built on a mesh nor was it a solid object......consequently it turned into and I quote "an srb sail". Although this did make me laugh it highlights just how much time and effort one can waste on cfd stuff when you are making stuff up as you go along. If I'm learning anything with this stuff it's that you need guidance when things gets complicated. The extra little tit on the main nose cone apparently played havoc with the boundary layer and was recommended to go....A boat tail on the main centre tank also apparently would massively reduce drag.....so boat tail ok no problem but the tit stays! I think the Russians had their radar mounted there so...........The Russians where defo onto something with the angled srb nose cones though. It massively reduces turbulent airflow around main stack.....There is a little bit between the srbs but not to much to cause a major concern.....more update soon......The entire stack is being built in solid works so well find out how that pans out......over
 
Back
Top