Euler solution vs viscid solution of flow in CFD??

[Aerodynamicist#1]

Hi,

Can anyone explain to me what advantage if any there is in running Fluent CFD software using the euler (inviscid) form of the Navier Stokes equations as opposed to the viscid form. What kind of error could you expect to see and what difference in solution time would be observed? Also can anyone explain in simple terms the choice of discretisation of the flow equations?

Thanks

[charlyfierros]

It depends in what are you wanting to model. For example, you can not use inviscid fluid to model deattached flow, like a formula one car. The difference in computing time between euler and navier is quite large.
best wishes
carlos.

[syguy]

Comparing Euler to Reynolds-Averaged Navier-Stokes

Advantage of Euler

• Runtime on same grid 40-50% faster (assuming 2 equation turbulence model such as k-epsilon)
• Don't need to resolve boundary layer - saving memory and reducing runtime further - could be another 25% saving dependent on boundary layer resolution required (wall functions vs low Reynolds number turbulence model)
• For attached flows reasonably accurate lift (downforce) prediction and pressure drag prediction

Disadvantage of Euler

• Can't predict separation except from sharp edges
• Doesn't predict skin friction (~25% of the total drag on a simple 2D wing)

I would guess the discretization schemes are specific to Fluent and its guidelines - best to consult the manual.

[Aerodynamicist#1]

Thanks guys that helps make it quite a bit clearer for me! Your replies are very much appreciated.