Total Pressure problem in a duct

[physixlover]

Hello
Please help me, a physical relation which my friend is facing, I'm unable to get my head around..
I created a 2D S-duct, the contours are as follows
1.Static pressure
2.Velocity contours
3. Total pressure gradient graph

intake Velocity 20m/s
Guage pressure is taken as '0' pascals

there is flow separation and reversal at + cruvature
created a rake through the duct to find out the pressure gradient

but the total pressure should be constant in duct, only static pressure changes but i was unable to find out what physical problem is causing such as change as shown in graph

Thanks in Advance

 

[physixlover]

sorry for got to attach geometry, rake view

 

[boneh3ad]

Inviscid or viscous?

 

[physixlover]

viscous, Laminar.

Thanks

 

[boneh3ad]

In a viscous flow, total pressure won't be conserved, so it shouldn't have constant total pressure.

I also notice you have a negative static pressure, which makes no sense. That isn't physically possible.

 

[bigfooted]

there is probably an offset used in the CFD code for static pressure, like 101325 Pa. There are good numerical reasons for doing this. 0 Pa would then actually be 101325 Pa.

Also, the Casimir effect can cause negative pressures, but that's quantum physics, so naturally nothing works as it's supposed to.

 

[boneh3ad]

The Casimir effect still doesn't cause negative fluid pressures. For starters, it is a quantum electrodynamic effect that occurs canonically in a vacuum. That isn't relevant in the least.

 

[Aero51]

Yeah as boneh3ad said total pressure isn't conserved in viscus flows. Did you do any hand calculations as a preliminary check of your results? I don't have much CFD experience personally, but it seems strange that the flow separates only at your first curve and not the second.

 

[Mithrandir13]

ever heard "Pressure Recovery"...

a duct with 98% Pressure recovery is nice...

an "S" duct with a 96% recovery (a length-diameter ratio of say 4'ish) is good too...

if all is perfect... you will loose total pressure from boundary layer effects...

incorporate diffusion and see how messy it gets! :O