Viscous flow with euler boundary condition

osoucy · January 23, 2014, 10:30

Hi,

Suppose I need to solve for the flow field around two spheres that are about 2 to 5 diameters apart. I'm not interested in calculating accurate aerodynamic coefficients or pressure distributions. I only need a moderately accurate representation of the flow velocity on a macro level. The final flow field velocities will interpolated on a coarse grid (cell sizes of about 5% of the spheres diameters).

Also, I would only be interested in computing turbulence effects and flow separation near one of the sphere. Since, I'm trying to minimize computational cost, would it make sense to impose navier-stokes boundary condition on one sphere, to impose euler boundary condition on the other one and to solve the flow with a turbulence model? Would it even converge?

Any help would be appreciated.

Regards,

jjalonso · May 12, 2015, 16:46

Hi there,

It depends on what you mean by "moderately accurate flow at the macro level". If you do not care about the wake of one of the cylinders, you could very well impose an inviscid boundary condition there and a viscous / no-slip on the other cylinder, thus saving he boundary-layer mesh on one of the cylinders. The calculation will run more quickly (and it should converge) but you will be missing some of the physics. Then again, if you do not care about the physics in that portion of the domain, things should be fine.

Best,

Juan

January 23, 2014, 10:30	Viscous flow with euler boundary condition	#1
osoucy New Member Olivier Soucy Join Date: Oct 2012 Location: Montreal, Canada Posts: 6 Rep Power: 13	Hi, Suppose I need to solve for the flow field around two spheres that are about 2 to 5 diameters apart. I'm not interested in calculating accurate aerodynamic coefficients or pressure distributions. I only need a moderately accurate representation of the flow velocity on a macro level. The final flow field velocities will interpolated on a coarse grid (cell sizes of about 5% of the spheres diameters). Also, I would only be interested in computing turbulence effects and flow separation near one of the sphere. Since, I'm trying to minimize computational cost, would it make sense to impose navier-stokes boundary condition on one sphere, to impose euler boundary condition on the other one and to solve the flow with a turbulence model? Would it even converge? Any help would be appreciated. Regards,

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Question about heat transfer coefficient setting for CFX	Anna Tian	CFX	1	June 16, 2013 06:28
Mass flow rate boundary condition for continuity equation for oscillating flow	p07ip705	Main CFD Forum	0	February 28, 2013 01:33
Airfoil boundary condition	Frank	Main CFD Forum	1	April 21, 2008 18:36
How can apply mass flow rate boundary condition?	Sima	Phoenics	1	December 1, 2007 18:55
Boundary Conditions	Jan Ramboer	Main CFD Forum	11	August 16, 1999 08:59

May 12, 2015, 16:46		#2
jjalonso New Member Juan J. Alonso Join Date: Jan 2013 Posts: 4 Rep Power: 13	Hi there, It depends on what you mean by "moderately accurate flow at the macro level". If you do not care about the wake of one of the cylinders, you could very well impose an inviscid boundary condition there and a viscous / no-slip on the other cylinder, thus saving he boundary-layer mesh on one of the cylinders. The calculation will run more quickly (and it should converge) but you will be missing some of the physics. Then again, if you do not care about the physics in that portion of the domain, things should be fine. Best, Juan