7 Replies Latest reply on Feb 6, 2013 8:49 AM by Bill McEachern

# How to simulate high temperature heat pipes?

In Flow Simulation there is an option for heat pipes, but it does not have a limiting operating temperature range.  Real heat pipes only work as heat pipes within a certain temperature range.  I am specifically interested in 600K water based SS316 heat pipes for space applications.  So my question, is it possible to model high temperature heat pipes in Flow Simulation using the built in heat pipe system?  The current heat pipe system will work if I force steady state situations, but I need a startup and cool down.  I have tried using a combination of solids that act as heat pipes and the inherent heat pipe function, but it only gives me mildly reliable results.  My system looks like this:

• ###### Re: How to simulate high temperature heat pipes?

If you understand the transient behavoir of your heat pipes thermal properties you can use a highly non-linear heat capacity and conductivity curves for the heat pipe material to transfer the energy as per the real phase change device. For example when the heat pipe is frozen it does not move heat very well so the materail conductivity is low as the temperature increases the conductivity increases. you need to have the heat capacity tuned so that it can absorb the latenet heat to sublimate the water or working fluid  and then maybe its heat capacity settles out to a constant value. I have't thought about for more than a few minutes but I know for sure that the program is very robust with very highly non-linear heat capacity and conductivity curves. Seems to me you should be able to get it done with out using the heat pipe model but you do need to understand how the thing behaves.

• ###### Re: How to simulate high temperature heat pipes?

Adam -

do you know how it will behave and want to see how it affects the rest of the system?  Or do you want to determine how it behaves?

Before Flow added heat pipes I had to define heat absorption/rejection for a heat pipe system using local temperatures and equations.  If you have data on the heat pipe you can induce the same thermal features manually by surface/volume sources and eliminate the limitations.

• ###### Re: How to simulate high temperature heat pipes?

Let me add some more data.  I am trying to heat up the hot side of the system shown from 298K to 600K using a steady heat source.  I do not know the exact details on the heat pipes just that they are possible (waiting until it is published by a NL).  I could use a heat sink and source but I do not belive it lets you use direct temperature dependence, and I need it to be temerature dependant.  If I use a material approximation for the heat pipe using Cp and k then it will conduct out to the second cylinder block, but when it reaches it the "HP" acts like a hot rod sitting on a cold plane.  I need it to act like a hollow tub with heat being emitted almost uniformly on the "cold" side of its surface.  Is there a way to make a source and sink temperature dependant?  Are if then statements incorporated into solidworks?  They would be really useful if they were.  The eventual goal is analysis of transient scenarios for the entire system (loss of one HP etc.).

• ###### Re: How to simulate high temperature heat pipes?

I would have to try to go back to those models, but I believe I setup global goals and then using them defined dependent sources.  The hot surface had a negative wattage source based on the difference in temperature between the two goals, and the cold surface had a positive wattage source based on the temperature difference.  This was the simplest analogy of a heat pipe but good enough for my needs.

Is that the type of temperature dependence you need?  I believe you should be able to define any typical relation in there as far as watts per deg C delta, etc.

• ###### Re: How to simulate high temperature heat pipes?

Not exactly, I think...  I need the bulk mass of the heat pipe to increase in temperature to 600K on one side, and at that temperature I need it to transfer any incident heat as it is designed to (nearly isothermal and isotropic) to its opposite side.  The source on the cold side should be roughly the same as the sink on the hot side.  Because I do not plan to exceed the operating temperatures and am staying within their operational power levels, the sink and source need to act like a switch turning on and off at the temperature bounds.  I see what you meen by the goal dependance though, I'll look into that.

I could have a heat pipe in the hot and cold side cylinders with a rod possessing specialized k and Cp values equivalent to the heat pipe's mass and conductance between them.  This might allow me to simulate the time necessary to heat up the HP's while still preserving the behavior.  Think that might work?

• ###### Re: How to simulate high temperature heat pipes?

You may be able to make a fluid subdomain inside the heat pipe on each side and then link the goals as volume sources - this would include the transient effects of warming the water in the pipe.  Or pick a solid that has properties similar to water (or create a "solid" with the same properties as water) and apply the volume source to that.

I'm not sure I get your 600K switch comment.  So until the one side hits 600K nothing happens?

• ###### Re: How to simulate high temperature heat pipes?

With all due respect to Chris you can not model the heat pipe with full physical fidelity so having a fluid sub domain seems pointless. Just model the heat pipe with a material that has a discontinuity in k till it hits say 598k and them ramp it up to the operating value at say 600k. You will need to adjust Cp so the right amount of energy gets absorbed to allow the temp to increase the way you want. Essentially heat pipes, at least in my experience, suck the big one when they are "frozen" k equals the metals conductivity, then as they start working they behave as awesome conductors - like 100,000 W/mk - a couple of degrees of delta across the whole thing. So you need a temperature depenent material property that mimics this behavior, tuned to your specific device. Well, that is what i would do. good luck.