Hi everybody,

I have to perform a thermal and mechanical analysis with Ansys and I need to take the HTC e Tbulk temperatures BCs from a CFD analysis with CFX.

I applied Twall fix value condition, in CFX, on the walls of my model and I used k-omega model of turbolency. The y+ value of the grid on the walls is very small (<1).

When I analyse the results of CFX analysis, I see that the values of HTC and Heat flux on the walls are very high and they are not very credible (more than 10^5 W/m^2K).

I tried to read the CFX's manual and I found how CFX calculate the HTC value....but I don't found a solution of my problem.

How can I resolve this incident?

Regards

Daniel

Hi,

HTC in the literature is referenced to a temperature at infinity, whereas by default CFX's HTC is referenced to a temperature close to the wall. The numbers are not comparable.

If you have already done the run you should get the wall heat flux from CFX and calculate the HTC from a relevant reference temperature. If you are doing more simulations you can set an expert parameter to explicitly use a defined reference temperature.

Glenn Horrocks

Hi Glenn, thanks for your respone.

I have read how CFX consider the value of HTC, but I didn't understand wich temperature I have to use as reference temperature (where, in the domain far from walls, I have to take it?) . If the wall heat flux is very high, how can I obtain a good HTC value?

I tried to use the correlations of convective heat transfer for calculate Nusselt numbers and then the HTC values, but I'don't know if it is a good work.

Maybe I can set the espert parameter "tbulk for HTC" with a global average temperature?

Thanks

Daniel

Hi,

The reference temperature for HTC is usually taken as either the ambient temperature or inlet temperature.

If you want to define the HTC reference temperature in CFX you have to use the tbulk for HTC expert parameter.

Glenn Horrocks

HTC value is an invented value. As soon as the temperature you chose is a good representation of the heat transfer phenomena it is ok. For a blade for example Lakshminarayana suggests to use the recovery temperature.

Trecovery = Tinfinite*(1+r*(gam-1)/2 * Mach^2)

Tinfinite is the temperature outside the thermal boundary layer. The Mach number is again considered outside the thermal B.L.

If the speed is slow you can consider the static temperature outside the B.L. by itself.

If you have zone with very different pattern of flows (e.g. pipes with different branches and very local feature) you can also consider to compute local values of T as representative of the heat transfer phenomena.

Max