Bow Shock In Capsule and Canopy

  • 41 Views
  • Last Post 25 May 2020
tllarose posted this 19 May 2020

Hello,

I am using the fluent solver to trying and resolve a simplified case of a deployed parachute at Mach 1.85. Main issue is that the bow shock and wakes are not resolving properly, picture attached. I am using a pressure inlet where the gauge pressure is the dynamic pressure of the system and the operating pressure is the air pressure at my desired altitude. I could use some help on this setup. Also wondering where there is some good literature about the implications of playing with the turbulent viscosity ratios/numbers beyond the users manual.

Would it also be good to set my fluid density to ideal gas, and the solver to density based? This is my current set-up. Also using SA turbulence model and energy equation is on.

Mesh is about 9Mil elements, probably could be resolved better around geometry edges but when I use sphere of influence it blows up the element number, any suggestions there? Also thinking of making a model of just the capsule and trying to create the bow shock there at a far lower element count. So suggestions for how to fully properly resolve a 3D mesh for supersonic flow would be great.

Again my main issue right now is that the pressure profile is not correct at all, also getting these messages in the fluent console each iteration and my residuals blow up to e+12. Also confused why the first message references a time step but I am just solving 1000 iterations frozen in time. Assuming it is something about the solver giving the problem some sort of time domain/parameter so that it can resolve the simulation?

 time step reduced in 131 cells due to excessive temperature change

 absolute pressure limited to 1.000000e+00 in 84 cells on zone 4 

 absolute pressure limited to 5.000000e+10 in 364 cells on zone 4 

 temperature limited to 1.000000e+00 in 85 cells on zone 4 

 temperature limited to 5.000000e+03 in 2358 cells on zone 4 

 

Thanks for reading,

Tim LaRose

 

Order By: Standard | Newest | Votes
Kalyan Goparaju posted this 19 May 2020

Hello Tim, 

Since you know the velocity, why not use the velocity-inlet instead of pressure-inlet? Your operating pressure can still be the atmospheric pressure at the given altitude.

In compressible flows, the fluid density is indeed determined according to idea gas law. So, yes, please set density to ideal gas law. As for the solver, the pressure-based scheme should be adept enough for simulating this case. As for the turbulence model, I would recommend starting with the default k-omega SST model. A 9 million mesh is a reasonable start. I would recommend making the above changes first instead of changing the mesh. 

Thanks, 

Kalyan

 

tllarose posted this 19 May 2020

Hi there Kalyan,

Firstly, I meant to say pressure-far-field for the input. Many places online say this is the way to go for compressible flow, in addition to a density based solver. They also suggest that velocity inlets are only for in compressible flow. Could you please shed some light on why these other settings you're suggesting still allow for an accurate representation of flow.

Also, great that I can set op pressure to atmospheric at that altitude, but then where will the solver incorporate the dynamic pressure of the system? Does this just get solved for when you used pressure based solver and velocity input?

Do you have any knowledge of if I should set dynamic pressure or another pressure value as gauge pressure if I use a pressure-far-field inlet BC?

Still looking to give your approach a spin.

Thanks again,

Tim LaRose

Kalyan Goparaju posted this 20 May 2020

Hello Tim, 

For Compressible Flows, Pressure far-field as initial condition should suffice. Please continue using it. As for dynamic pressure, it is calculated based on velocity and density, you don't have to specify this value. 

Please look at this section of the User Manual for some tips on how to model compressible flows

https://ansyshelp.ansys.com/account/secured?returnurl=/Views/Secured/corp/v201/en/flu_ug/flu_ug_sec_compressible.html

Thanks, 

Kalyan

tllarose posted this 21 May 2020

Hi there Kalyan,

Running the sim now with gauge pressure at 0 for the BCs, operating pressure is still atmospheric. This does make the most sense, ran with a non-zero gauge before because the solver gave me a warning that I shouldn’t have a zero gauge.

Could you please speak to why the solver doesn’t like a zero gauge? What is happening with these BCs in relation to the problem as a whole?

Also, why use SST? What is simple about it’s deployment/use in the solver? Using SA right now as it is considered to be a good low equation option for my supersonic conditions according to the parachute modeling literature.

Thanks again, Tim

Kalyan Goparaju posted this 21 May 2020

Hello Tim, 

I am guessing that when you ran the earlier case, you had both operating pressure and gauge pressure set to 0 (by mistake perhaps) and hence Fluent displayed the error message. 

SA model is actually quite good for aerospace applications and has been observed to perform well for wall-bounded flows and in adverse pressure gradients. However, It is known to yield poor results for free shear layer flows. For your case though, I think it should be fine. Here is a link that has the general description of the different turbulence models including their pros and cons. Hope this helps. 

https://ansyshelp.ansys.com/account/secured?returnurl=/Views/Secured/corp/v201/en/flu_ug/flu_ug_sec_turb_rans.html%23flu_ug_sec_turb_rans_trans

Thanks, 

Kalyan

tllarose posted this 21 May 2020

Kalyan,

For the pressure, I had operatingn pressure set to atmospheric and gauge pressure on inlet and outlet BCs set to 0. The solver gave a pre-calculation warning. Could you please speak to why this is.

Thanks, Tim

Kalyan Goparaju posted this 21 May 2020

Hi Tim, 

Can  you please post an image of the warning you see? Also, try setting operating pressure to 0 and then set the gauge pressures to absolute total pressure at the inlet, and absolute atmospheric pressure at the outlet. 

Thanks, 

Kalyan

rahkumar posted this 21 May 2020

Hello Tim, 

Another suggestion... You can also try simulating using Density based method with Explicit formulation and AUSM scheme. 

Regards, Rahul

tllarose posted this 25 May 2020

Hi Y'all,

Image of my error is attached. Had some success setting OP pressure to my atmospheric value and gauge pressures to zero. I have updated the geometry and am about to run the solver again with these settings. Kalyan I am a little confused because I thought we were saying beforehand that velocity at the inlet will create a dynamic pressure in the model so if I wanted to switch it up wouldn't I just set op pressure to zero and the inlet and outlet gauge pressure to atmospheric?

Rahul I appreciate the suggestion, what makes you say those approaches may be valid here?

Will post pictures of my new geometry once I run this solver again.

Thanks,

Tim

tllarose posted this 25 May 2020

 ... also running my new sim with the explict formulation and AUSM scheme selected

Kalyan Goparaju posted this 25 May 2020

Hello Tim, 

That was just a suggestion in case setting setting operating pressure to my atmospheric value and gauge pressures to zero didn't work.

Thanks, 

Kalyan

Close