[Ahmed A. Serageldin]

I am trying to model the internal convection through very long pipe in transient manner with total length of 204 m and diameter of 27 mm. it takes very long time as the total number of element about 9 million. so, my question is it possible to calculate the heat transfer coefficient profile from the 9 million element model and adjust that for coarse mesh model of 700000 element?
I possible how I can do that?, I already tried many time but the option in Fluent for convection need also the free stream temperature which I didnot need to defined.

[lcarasik]

Originally Posted by Ahmed A. Serageldin

I am trying to model the internal convection through very long pipe in transient manner with total length of 204 m and diameter of 27 mm. it takes very long time as the total number of element about 9 million. so, my question is it possible to calculate the heat transfer coefficient profile from the 9 million element model and adjust that for coarse mesh model of 700000 element?
I possible how I can do that?, I already tried many time but the option in Fluent for convection need also the free stream temperature which I didnot need to defined.

204 m is approximately 7555 hydraulic diameters which you likely only need 40 diameters at most to get a thermally fully developed flow. After which, that's all you'd need to measured the HTC.

Do you really need CFD for this problem? If it is just pipe flow, you can use the equations found in a standard heat transfer textbook.

[Ahmed A. Serageldin]

Originally Posted by lcarasik

204 m is approximately 7555 hydraulic diameters which you likely only need 40 diameters at most to get a thermally fully developed flow. After which, that's all you'd need to measured the HTC.

Do you really need CFD for this problem? If it is just pipe flow, you can use the equations found in a standard heat transfer textbook.

Thanks for your kindly reply, actually this pipe is a part of the whole domain, but it is the most important part as the other parts are pure conduction heat transfer problem, so what i need to do is to calculate the heat transfer coefficient inside this tube by using very fine mesh and then using this profile for coarse mesh model to decrease the total time of simulation.

[lcarasik]

Originally Posted by Ahmed A. Serageldin

Thanks for your kindly reply, actually this pipe is a part of the whole domain, but it is the most important part as the other parts are pure conduction heat transfer problem, so what i need to do is to calculate the heat transfer coefficient inside this tube by using very fine mesh and then using this profile for coarse mesh model to decrease the total time of simulation.

Is there anything particularly special about this pipe? Like is the flow non-newtonian? Elsewise, it doesn't sound like actually simulating the pipe is valuable as opposed to using a standard heat transfer correlation.

[Ahmed A. Serageldin]

Originally Posted by lcarasik

Is there anything particularly special about this pipe? Like is the flow non-newtonian? Elsewise, it doesn't sound like actually simulating the pipe is valuable as opposed to using a standard heat transfer correlation.

No, there is no special about this pipe. But I asked about is it possible to use the heat transfer coefficient profile from previous simulation in current simulation.

[LuckyTran]

When you say coarser mesh, you mean a coarse mesh of a different domain right? Not the same pipe? Else, why would you impose the solution of a fine grid onto a coarse grid.

Originally Posted by Ahmed A. Serageldin

No, there is no special about this pipe. But I asked about is it possible to use the heat transfer coefficient profile from previous simulation in current simulation.

Can you? Of course. Any simulation that you do has no idea where the boundary conditions come from. It's not omnipotent.

But I have to ask the same question, do you really need CFD for this problem? If it's laminar then the Nusselt number is already known exactly pretty much. If it's turbulent, can't you just get the htc from Dittus-Boelter or something?