[ameyer]

Another newbie hits the pavement: Hello cfd-community

I searched the forum (and outside) but couldn't get a thourough answer:

I'm working on an aerodynamic structure that is driven by a fan and produces lift (very vague, but i wouln't like to go into further details about it). Now I'm hoping to optimize this a bit through cfd. Therefore I want to compare the lifting force of several slightly modified versions against each other.

To make the lift-values comparable, i need to ensure that the flow through my models is driven by an equal amount of power - like i did in my real-world experiments by controlling the input power to the fan.

In my first attempt I simply used a constant-velocity inlet, but looking at the inlet pressure values in paraview revealed that there is a huge spread in average inlet pressure amongst the models. At constant velocity this would mean different power-levels for a (hypothetical) fan driving the flow.

My first (real newbie) question:

Is it ok to look at pressure values directly at an inlet-patch to calculate power, or should this be done a bit downstream on internal cells?



I already realized that the TotalPressure BC would be more in my favour, but what I really need is some kind of "constant pumping power" inlet.

So here's my second question:

Is there an "open-foam-way" to achieve this? I don't ask you to solve this problem for me, just a hint if I'm simply overseeing something.

I had a look at the various derived BCs in the code - none of them seemed to fit directly. Maybe the fan-patch could be used for it with an apropriate function for the pressure-drop.

[ameyer]

As it turned out, I already gave the answer myself, just in case somebody stumbles across teh same problem:

The solution for me was to use the pressureFan bc for p on the inlet while using pressureInletVelocity for U. To achieve what i called 'constant-power-inlet' I wrote a simple script to calculate a fan-curve for a given input-power and inlet-area using the following formula:

input_power = fan_pressure_drop_total * rate_of_volume

where:

fan_pressure_drop_total = fan_pressure_drop_static + fan_pressure_drop_dynamic

(fan_pressure_drop_dynamic = 0.5 * density * square_of(U))


In this nomenclature the fan-curve would be:

fan_pressure_drop_static = f (rate_of_volume)

leading to:

input_power = ( f (rate_of_volume) + fan_pressure_drop_dynamic ) * rate_of_volume


which finally can be resolved for:

f (rate_of_volume) = (input_power / rate_of_flow) - 0.5 * density * square_of(rate_of_flow / A)

(with A beeing inlet-area)

As p in icoFoam is not simply pressure but (pressure / density) the pressure-values from that formula (or the fomula itself) would have to be divided by density.