DNS code works for turbulent but blow up for laminar, why?

hnemati · October 25, 2013, 20:26

Dear all
I have a DNS code in cylindrical coordinate. The code is written to simulate supercritical flows. The code works fine for Re=5300, but while I decrease the Re=1700, it blow up. I think the problem is the enthalpy solver. Is there anybody help me?
thanks

ztdep · October 25, 2013, 21:05

maybe compressible incompressible problem, they use totally different algorithms.

FMDenaro · October 26, 2013, 05:22

Depending on the discrete scheme, the lower the Re number is, the lower the dt should be fixed to fulfill the stability constraint

cfdnewbie · October 27, 2013, 17:38

Just to rule something out: How do you start the simulation, what is your initial condition? If you re-use a initial condition from a high Re for example, that could lead to a blow-up.

hnemati · October 28, 2013, 04:17

Quote:

Originally Posted by cfdnewbie

Just to rule something out: How do you start the simulation, what is your initial condition? If you re-use a initial condition from a high Re for example, that could lead to a blow-up.

Thanks all for replying.
No I am using initial condition for laminar flow, no perturbation.
I think I need to reduce dt in my simulation becase of CFL condition

FMDenaro · October 28, 2013, 04:23

Quote:

Originally Posted by hnemati

Thanks all for replying.
No I am using initial condition for laminar flow, no perturbation.
I think I need to reduce dt in my simulation becase of CFL condition

what about your time integration method? More probabily, the stability constraint you violated is related to the diffusive part rather than the convective part of the discretization

hnemati · October 28, 2013, 04:29

Quote:

Originally Posted by FMDenaro

what about your time integration method? More probabily, the stability constraint you violated is related to the diffusive part rather than the convective part of the discretization

I dont think so, becase I am using two sub-step for stabilazion. that is superstable for turbulent.

calf.Z · November 3, 2018, 00:15

Hello SAM，
I am also dealing with supercritical fluids problem by using DNS.Have you done a good simulation？

October 25, 2013, 20:26	DNS code works for turbulent but blow up for laminar, why?	#1
hnemati Member SAM Join Date: Apr 2012 Posts: 74 Rep Power: 14	Dear all I have a DNS code in cylindrical coordinate. The code is written to simulate supercritical flows. The code works fine for Re=5300, but while I decrease the Re=1700, it blow up. I think the problem is the enthalpy solver. Is there anybody help me? thanks

October 25, 2013, 21:05		#2
ztdep Senior Member p ding Join Date: Mar 2009 Posts: 427 Rep Power: 19	maybe compressible incompressible problem, they use totally different algorithms. hnemati likes this.

October 26, 2013, 05:22		#3
FMDenaro Senior Member Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,849 Rep Power: 73	Depending on the discrete scheme, the lower the Re number is, the lower the dt should be fixed to fulfill the stability constraint hnemati likes this.

October 27, 2013, 17:38		#4
cfdnewbie Senior Member cfdnewbie Join Date: Mar 2010 Posts: 557 Rep Power: 20	Just to rule something out: How do you start the simulation, what is your initial condition? If you re-use a initial condition from a high Re for example, that could lead to a blow-up. hnemati likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Free DNS code ?	Dr. Jones	Main CFD Forum	8	November 7, 2000 20:44
Open source CFD code development, possible?	Dr. Yazid Bindar	Main CFD Forum	27	July 18, 2000 01:18
State of the art in CFD technology	Juan Carlos GARCIA SALAS	Main CFD Forum	39	November 1, 1999 15:34
3D DNS code in C with large arrays.	S.Sri Govind	Main CFD Forum	3	July 23, 1999 14:09
What kind of Cmmercial CFD code you feel well?	Lans	Main CFD Forum	13	October 27, 1998 11:20

November 3, 2018, 00:15		#8
calf.Z Senior Member Jianrui Zeng Join Date: May 2018 Location: China Posts: 157 Rep Power: 8	Hello SAM， I am also dealing with supercritical fluids problem by using DNS.Have you done a good simulation？