Non-reflective boundary conditions

nsi · October 18, 2017, 22:04

Dear CFD friends!

I need to apply an NRBC to my compressible inviscid Euler eqn.s solver (shock involved problem). I tried some classical methods introduced by

Poinsot & Lele's paper
Kevin Thompson's paper

but both were subjected to numerical stability issues. I then walked through Giles's paper but could hardly understand either his philosophy or how to apply it to my solver. For example, he said the c's, i.e., {c1, c2, c3, c4} is the amplitude of the characteristic waves, but who knows how much the c's equal...

Any suggestion about other types of NRBCs?

wb

praveen · October 19, 2017, 08:11

Boundary conditions can be tricky and it is difficult to give general answer. Can you describe the physical setup and what exactly you tried so far ?

nsi · October 22, 2017, 19:35

Praveen,

Attached is a cartoon of my problem set up.

I have been trying to impose a robust NRBC at both ends.

praveen · October 22, 2017, 22:43

If the ends are closed, then it is like a solid wall bc. If the ends are open to the atmosphere, and if shocks exit through them, then it will be difficult to give any rational bc. When shock is away from the ends, you have some external atmospheric state and you can use characteristics/riemann invariants approach. When shock is near the exit, the concept external state itself is meaningless because the shock would have changed the state on the other side of the boundary.

Have you tried a simple neumann bc, i.e., set the state in a ghost cell equal to the cell at the boundary.

sbaffini · October 23, 2017, 05:09

On most boundaries I tipically use the full convective scheme with ghost cells.

For your problem, a shock tube I guess, I confirm that the approach suggested by praveen (analogous to a full supersonic outlet) can work.

FMDenaro · October 23, 2017, 07:45

Quote:

Originally Posted by nsi

Praveen,

Attached is a cartoon of my problem set up.

I have been trying to impose a robust NRBC at both ends.

Have a look here for a 1D analysis and BC.s https://books.google.it/books/about/...AJ&redir_esc=y

Of course, if you are simulationg a 2D flow, it will develop reflections in the closed case.

AshwaniAssam · October 24, 2017, 03:59

You may go through the Blazek book of CFD, it is a comprehensive and more direct from implementation point of view.

monika_1387 · May 22, 2019, 10:51

Hi, Did you figured out how to put this NRBC BC for nozzle flow at outlet BC? Because I am trying to put NRBC BC but thats not working so trying for GILES if that works.

Can you please suggest if that worked for you? I am working on nozzle flow (supersonic at outlet end and trying to put non reflective BC at outlet.

Quote:

Originally Posted by nsi

Dear CFD friends!

I need to apply an NRBC to my compressible inviscid Euler eqn.s solver (shock involved problem). I tried some classical methods introduced by

Poinsot & Lele's paper
Kevin Thompson's paper

but both were subjected to numerical stability issues. I then walked through Giles's paper but could hardly understand either his philosophy or how to apply it to my solver. For example, he said the c's, i.e., {c1, c2, c3, c4} is the amplitude of the characteristic waves, but who knows how much the c's equal...

Any suggestion about other types of NRBCs?

wb

LuckyTran · May 22, 2019, 10:54

Quote:

Originally Posted by monika_1387

Can you please suggest if that worked for you? I am working on nozzle flow (supersonic at outlet end and trying to put non reflective BC at outlet.

A supersonic outlet shouldn't need a NRBC.

nsi · May 22, 2019, 11:17

Hi, Monika!

I would say go check Jiri Blazek's book Computational Fluid Dynamics: Principles and Applications, read the boundary condition chapter, and you will get what you need. It has a detailed description of the application of all types of B.C.s such as super/subsonic inlet, super/subsonic outlet, etc.

Best,

WB

Quote:

Originally Posted by monika_1387

Hi, Did you figured out how to put this NRBC BC for nozzle flow at outlet BC? Because I am trying to put NRBC BC but thats not working so trying for GILES if that works.

Can you please suggest if that worked for you? I am working on nozzle flow (supersonic at outlet end and trying to put non reflective BC at outlet.

monika_1387 · May 22, 2019, 12:08

Thank you for directing me. I ran the simulation and seeing reflected pressure waves from outlet boundary bouncing back to inlet side.. So because of that reason I believe non reflecting boundary will work. Still Let me look into this CFD book.

Quote:

Originally Posted by nsi

Hi, Monika!

I would say go check Jiri Blazek's book Computational Fluid Dynamics: Principles and Applications, read the boundary condition chapter, and you will get what you need. It has a detailed description of the application of all types of B.C.s such as super/subsonic inlet, super/subsonic outlet, etc.

Best,

WB

FMDenaro · May 22, 2019, 12:14

Quote:

Originally Posted by monika_1387

Hi, Did you figured out how to put this NRBC BC for nozzle flow at outlet BC? Because I am trying to put NRBC BC but thats not working so trying for GILES if that works.

Can you please suggest if that worked for you? I am working on nozzle flow (supersonic at outlet end and trying to put non reflective BC at outlet.

First of all, what about the Mach number at inflow? At the supersonic outflow the informations travels from the interior and you can use the characteristic conditions but your problem could be due to some wrong condition at the inflow. Of course a bug in the code can be also a possible cause.

monika_1387 · May 22, 2019, 12:41

Its Minlet=0.2911 (subsonic) and outlet as supersonic BC.

Quote:

Originally Posted by FMDenaro

First of all, what about the Mach number at inflow? At the supersonic outflow the informations travels from the interior and you can use the characteristic conditions but your problem could be due to some wrong condition at the inflow. Of course a bug in the code can be also a possible cause.

FMDenaro · May 22, 2019, 12:46

Quote:

Originally Posted by monika_1387

Its Minlet=0.2911 (subsonic) and outlet as supersonic BC.

Are you aware of the difference of the BCs. setting between supersonic and subsonic conditions? What about the variable you are letting from the interior?

monika_1387 · May 22, 2019, 12:53

I am a new user of SU2 so learning it. But here is my attached.cfg file you can look once for BCs which I give. I am trying for NRBC, GILES or Riemann to avoid these reflections at outlet boundary. See the attached .cfg file for reference. Appreciate your comments.

Quote:

Originally Posted by FMDenaro

Are you aware of the difference of the BCs. setting between supersonic and subsonic conditions? What about the variable you are letting from the interior?

FMDenaro · May 22, 2019, 13:56

Quote:

Originally Posted by monika_1387

I am a new user of SU2 so learning it. But here is my attached.cfg file you can look once for BCs which I give. I am trying for NRBC, GILES or Riemann to avoid these reflections at outlet boundary. See the attached .cfg file for reference. Appreciate your comments.

Why don't you post the question in the SU2 section?

monika_1387 · May 22, 2019, 14:15

Sure. will do.

Quote:

Originally Posted by FMDenaro

Why don't you post the question in the SU2 section?

October 18, 2017, 22:04	Non-reflective boundary conditions	#1
nsi New Member Join Date: Oct 2017 Posts: 7 Rep Power: 8	Dear CFD friends! I need to apply an NRBC to my compressible inviscid Euler eqn.s solver (shock involved problem). I tried some classical methods introduced by Poinsot & Lele's paper Kevin Thompson's paper but both were subjected to numerical stability issues. I then walked through Giles's paper but could hardly understand either his philosophy or how to apply it to my solver. For example, he said the c's, i.e., {c1, c2, c3, c4} is the amplitude of the characteristic waves, but who knows how much the c's equal... Any suggestion about other types of NRBCs? wb

October 19, 2017, 08:11		#2
praveen Super Moderator Praveen. C Join Date: Mar 2009 Location: Bangalore Posts: 342 Blog Entries: 6 Rep Power: 18	Boundary conditions can be tricky and it is difficult to give general answer. Can you describe the physical setup and what exactly you tried so far ? __________________ http://cpraveen.github.io http://twitter.com/cfdlab http://github.com/cpraveen

October 22, 2017, 22:43		#4
praveen Super Moderator Praveen. C Join Date: Mar 2009 Location: Bangalore Posts: 342 Blog Entries: 6 Rep Power: 18	If the ends are closed, then it is like a solid wall bc. If the ends are open to the atmosphere, and if shocks exit through them, then it will be difficult to give any rational bc. When shock is away from the ends, you have some external atmospheric state and you can use characteristics/riemann invariants approach. When shock is near the exit, the concept external state itself is meaningless because the shock would have changed the state on the other side of the boundary. Have you tried a simple neumann bc, i.e., set the state in a ghost cell equal to the cell at the boundary. FMDenaro likes this. __________________ http://cpraveen.github.io http://twitter.com/cfdlab http://github.com/cpraveen

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Domain Imbalance	HMR	CFX	5	October 10, 2016 05:57
Multiphase flow - incorrect velocity on inlet	Mike_Tom	CFX	6	September 29, 2016 01:27
Error - Solar absorber - Solar Thermal Radiation	MichaelK	CFX	12	September 1, 2016 05:15
Low Mixing time Problem	Mavier	CFX	5	April 29, 2013 00:00
reflective boundary conditions on generalized coor	buaalzr	Main CFD Forum	3	January 2, 2008 12:43

October 23, 2017, 05:09		#5
sbaffini Senior Member Paolo Lampitella Join Date: Mar 2009 Location: Italy Posts: 2,151 Blog Entries: 29 Rep Power: 39	On most boundaries I tipically use the full convective scheme with ghost cells. For your problem, a shock tube I guess, I confirm that the approach suggested by praveen (analogous to a full supersonic outlet) can work.

October 24, 2017, 03:59		#7
AshwaniAssam Senior Member Ashwani Join Date: Sep 2013 Location: Hyderabad Posts: 154 Rep Power: 12	You may go through the Blazek book of CFD, it is a comprehensive and more direct from implementation point of view.