Community
CFD Forum
Welcome to Autodesk’s CFD Forums. Share your knowledge, ask questions, and explore popular CFD topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

trouble changing scenario to compressable flow

11 REPLIES 11
Reply
Message 1 of 12
paul175
348 Views, 11 Replies

trouble changing scenario to compressable flow

as an incompressible ... things worked and seemed to make sense,

but the air was moving at mach 2.6 ... hence the switch to compressible

 

fluid is air ... 1 inlet 9 outlets. 

boundary conditions were:

inlet: piecewise linear temperature and pressure over 2 milliseconds

outlet: 0 pressure

 

changes for compressibility:

set air to 'variable'

select 'compressible' in solver dialog

boundary conditions are:

inlet: total temperature and total pressure and mass flow

outlet: pressure 0, and 'unknown'

 

simulation gets to the second time step and crashes ...

 

I read the paper on interior compresibles and didn't catch my mistake.

What did miss ?

 

/Steven

 

11 REPLIES 11
Message 2 of 12
Royce_adsk
in reply to: paul175

Hi Paul,

 

It seems to me that you have overconstrained your model a bit with the compressible setup.

 

Review the suggest model setup guidelines here:

http://wikihelp.autodesk.com/Simulation_CFD/enu/2014/Help/0360-Learning360/0400-Simulati400/0435-Com...

 

My suggestion based on what you wrote would be to remove the mass flow rate and the unknown boundary conditions and give it another shot.  That would make the model setup jive with the 1st bullet point model setup in the above link.



Royce.Abel
Technical Support Manager

Message 3 of 12
paul175
in reply to: Royce_adsk

if I remove the mass fow rate, the summary begins with "0 inlets" and the rest is crap ...

Message 4 of 12
Royce_adsk
in reply to: paul175

Hi Paul,

 

The summary files labels ALL pressure boundary conditions are 'Outlets' even though they very well could be inlets.  It is just book post-processing keeping and not important.

 

What does 'rest is crap' in regards to?

 

-Royce



Royce.Abel
Technical Support Manager

Message 5 of 12
paul175
in reply to: Royce_adsk

I am referring to the 'Messages' tab in the simulator, where it does differentiate between inputs and outputs.

Despite inititial conditions for the air volume of pressure=0, velocity=(0,0,0) and temp=72 ... I am seeing pressures (psi) in the millions, velocities (in/sec) in the bilions, and temperatures of absolute zero -- -459F .

the simulation seems very confused ... or is it me ?


Since the first message I have replaced the total pressure constant with a piecewise linear from 13601 to 0(psi); total temperature constant with a piecewise linear from 2092 to 22 (degrees C); and the mass flow constant with a piecewise linear from 1.75 to 0 (kg/s). The simulator still crashes in the first few iterations ... and reports bizarre values.

Message 6 of 12
paul175
in reply to: paul175

maybe its my computations of the new boundary conditions...

 

input temp = 3361F, pressure = 8000psi, velocity = 2970 ft/s
so speed of sound = 3190 ft/s ... and mach number = 0.9311
from nasa I got a gamma of 1.3028622041 for 3361F and 8000psi
I get a total temp of 2092.295673C and a total pressure of 13600.77257psi
mass flow is 1.75 grams in 2 milliseconds -- or .875 kg/s

by 2 milliseconds temp and pressure are back to ambient and flow==0

 

I'm setting these as piecewise linear boundary conditions. And yet ...

Message 7 of 12
paul175
in reply to: paul175

this help ?

Message 8 of 12
Royce_adsk
in reply to: paul175

I was just going to ask for that. I'll check it out later today.

Thanks!


Royce.Abel
Technical Support Manager

Message 9 of 12
Royce_adsk
in reply to: Royce_adsk

Can we make this 1/4 symmetric?

Also, could we move the projectile to be placed outside of the fluid domain in CAD.  We will use the modtion initial position settings to place it in the barrel.  this will allow us to turn on mesh enhancement inside of the barrel.

 



Royce.Abel
Technical Support Manager

Message 10 of 12
Royce_adsk
in reply to: Royce_adsk

Just to give you an idea of how I am troubleshooting this problem. I am first turning OFF the transient motion portion of the setup and will work on getting the problem solving the compressible flow transiently to determine the needed mesh and solver configurations for this to work.

 

First Stab.png

 



Royce.Abel
Technical Support Manager

Message 11 of 12
paul175
in reply to: Royce_adsk

I don't think so. The incompressible runs show significant differences in flow among quadrants centered at 12:00 3:00 6:00 and 9:00 -- looks to be heat related.
Message 12 of 12
Royce_adsk
in reply to: paul175

Hi Paul,

 

I backed up a bit and built a 2D model so we could perform some quick testing.  This model is fairly unstable due to the explosive nature especially in the first few iterations.

 

With that in mind I decided to work my way up in solver setting complexity to determine what setup would be required.  The only way I could get the model to be stable and give what looks to be a fairly reasonable solution required settings that does not support our Motion package.  With that in mind, I would think you could study the silencer design fairly well without even including the bullet as part of the simulation.  Trying to run this sort of model with motion would be an extermaly long runtime.  

 

Check out the attached 2D model that you can run out on your end.



Royce.Abel
Technical Support Manager

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report