Problem applying Pressure drop in Flotran.
Problem applying Pressure drop in Flotran.
(OP)
Hi Everyone,
I am working on a 2-d Heat transfer problem ,in which there is flow over a flat plate of finite thickness kept in reatangular computational domain .I need to vary the upper and lower boundaries of the computational domain for the flow over a flat plate keeping the pressure drop constant across the inlet(left boundary) and outlet(Right) boundaries for all cases and also apply a constant inlet velocity for all cases.
The only boundary condition at the outlet -gradient of velocity is zero.
How do I apply a constant pressure drop across the inlet and outlet and also giving a constant velocity at the inlet??Can anyone help me with this??
regards,
Tom
I am working on a 2-d Heat transfer problem ,in which there is flow over a flat plate of finite thickness kept in reatangular computational domain .I need to vary the upper and lower boundaries of the computational domain for the flow over a flat plate keeping the pressure drop constant across the inlet(left boundary) and outlet(Right) boundaries for all cases and also apply a constant inlet velocity for all cases.
The only boundary condition at the outlet -gradient of velocity is zero.
How do I apply a constant pressure drop across the inlet and outlet and also giving a constant velocity at the inlet??Can anyone help me with this??
regards,
Tom





RE: Problem applying Pressure drop in Flotran.
if I understood well, the conditions needed are as follows:
- P1-P2=k1, where P1 is pressure @ inlet and P2 is pressure @ outlet
- v1=k2
- dv2/dL=0
From my understanding, these conditions are redundant:
- if you specify v1, there is no need to specify P1 altogether (=the problem is not pressure-driven); P2 will serve as a reference (=in Flotran pressures are relative ones), then P1 will be an output quantity and so will be P1-P2. If you have v1 already calculated "by hand" for a given P1-P2, then you should model your domain exactly as you did for the hand-made calcs (=null-roughness? laminar or turbulent model? etc...): you will see that P1-P2 obtained by the CFD for the imposed v1 will match what you expect (within a small error);
- if you specify P1 and P2, it is not correct to specify v1: the problem is pressure-driven. v1 will be an output quantity, provided that your domain is bounded by some wall somewhere (wind also is a bounded problem, where the dimensions of the domain are extremely large!), otherwise it is not really "physical" that a pseudo-infinite domain has a "natural" pressure difference...
To summarize:
1- if your plate is enclosed in a region bounded by walls, you can drive your problem either by velocity or by pressure (remember: geometrical properties + velocity condition + pressure condition are NOT an independent triplet of conditions: if you choose two, the third is dependent; for obvious reasons, a CFD can not change geo to accomodate imposed velocity + pressure)
2- if your plate is in an "external" region (=surrounded by a pseudo-infinite environment), the only choice from my point of view is to drive the problem by velocity.
Regards