Limit total pressure
 

Hello,

I'd like to know if there is a way to limit total pressure in Fluent 14.5.

In my simulations, after converging with different methods, I always observe total pressure values that are higher than the upstream total pressure. It's not much, and these excesses are only localized close to the outer region of the boundary layer, but the uncertainty related to these phenomena is higher than the precision that i would like to have on my results.

These simulations are performed on static adiabatic walls, no energy exchange with the air.

Thanks in advance,

Marcus

At 'solution controls', you have 'limits'. I think that is what you are looking for (or at least very close to it).

Unfortunately with "Solution Limits" I can only limit the static pressure...

I need to apply the limiting conditions on the total pressure, while letting pressure and velocity vary according to the conservation equations.

Then I guess what you really want is not to limit the pressure, but improve the accuracy of the simulation, isn't it?
In that case I would focus more on mesh refinement in that region, or improved numerical schemes, or something in that direction.

What you wanted to do, is decrease the freedom for the solver (not allowing certain values of the total pressure), to get what you expect. I think it is better to increase the freedom for the solver (giving more nodes on which the pressure can vary), to get the solution to the equations.

Hello pakk,

exactly, I would like to see what happens if I prevent mechanical energy from being created in the flow by adding a constraint on the total pressure.

I have already tried with refinement (both manual and automatic, with the adaptation options), and with different numerical methods, but the results are always the same.

Unfortunately the size of the mesh is already considerable, and the only option I have to refine even more is performing simulations on a portion of the geometry. The problem is that the geometry is not symmetric, and I will eventually need to simulate the whole body... so my idea was to perform two sets of simulations :

1) on a portion of the geometry with an extremely refined mesh
2) on a portion of the geometry with the mesh I can actually use on the whole body, with the additional constraint on the total pressure

if the results obtained with the second method are acceptable, when compared to the ones obtained with the ideal strategy, I will try to use it to calculate the solution around the whole body.

Are you certain that the extra mechanical energy is a numerical error? Could that energy be coming from turbulence or any other models?

For simple inviscid potential flow, total energy can never increase along a streamline, but I don't think that's true for real viscous flows. Consider that your result could be accurate.

Hello Kokemoor,

thanks for the suggestion. I have indeed thought of it, but I haven't managed to find any reference that could help me understand the phenomenon from a physical point of view.

it's just that I don't see how total pressure could increase in a viscous flow... wouldn't it mean that heat should convert into mechanical energy? and wouldn't that violate the second law of thermodynamics? If you have any reference about this topic, could you please indicate them to me?

What viscosity can allow here is not the transfer of heat into mechanical energy, but the transfer of mechanical energy from one streamline to another. The total energy leaving your system should definitely be less than what's entering (unless you have a moving wall or something like that injecting energy into the system), but some part of the flow may transfer energy to another, leading to local total energies greater than the inlet energy.

Hello Kokemoor,

I hadn't thought about it that way... I'm really interested about the subject, and I would like to read more about it.

The thing is, while those phenomena are always present for every model and mesh that I tried, their distribution and intensity vary, so that I cannot still be sure that the results represent the physical reality. some previous observations, criteria, and validated results would be really helpful.