Non-Uniform Heat Flux on Heat Pipes

[MrRogers1987]

I am working on designing a thermal control system for an array of telescopes that need to be kept at very cold temperatures, and within a very tight tolerance. Constant conductance heat pipes are being used to bus waste heat from the telescope detectors to a deep space radiator. Each detector has an interface pad on the evaporator of the heat pipe loop.

A recent requirement change now has it so that 1 telescope in the array needs to be kept much colder than the rest. Notionally the plan is to use a TEC on the single telescope that needs to be kept colder. It will interface with the same heat pipe evaporator as the other units, and it's hot-side reject temperature will be similar to the temperature the other units needs to be maintained at. So the desire is to have the evaporator operating at the same temperature, but more heat will be input and a larger radiator will be used at the condenser to compensate.

I am not very familiar with detailed heat pipe design other than understanding the underlying principles of the two-phase capillary flow. I am concerned there could potentially be an issue though, since the heat flux into the heat pipe evaporator will now be highly non-uniform down its length. Instead of having ~1W coming in at each interface there will now be close to 4W at one of the interfaces, and ~1 at the rest. Will this lead to any undesirable temperature gradients within the evaporator, or irregular flow rates? Or potentially some other issues that would bring operation out of ideal conditions?

 

[davefitz]

Regardless of any opinion or any results from a computer model, it would seem to be standard operating procedure to test a prototype under a wide range of conditions and proof test the design concept. Opinions and models are not exactly perfect at predicting all possible physical actions.

The weaknesses of heat pipes may be corrosion loss of wall thickness, explosive failure due to chemical reactions ( for synthetic coolants) ,and structural weaknesses due to the need to maintain a minimum wall thickness. The 2 phase conditions ensures no thermal gradients provided DNB departure from nucleate boiling ( or dryout) does not occur. The heat pipe internal structure likely should be designed to allow a liquid recharge rate of 4 times average at the location in question to ensure dryout or DNB does not occur.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[IRstuff]

Changing the radiator is equivalent to reducing the thermal resistivity; presumably, you'd need to do the same with the heat pipes, since you've added to the overall heat load, and that'll be reflected in a higher deltaT across the heat pipe, which will required increasing the heat transfer capacity of the heat pipe(s)

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[MrRogers1987]

Thanks for the quick replies.

@davefitz
Agreed 100% that some EDU will need to be built, and significant testing done at both proto-type and flight model level. At this stage my goal is primarily to determine if the approach we are pursuing is even feasible/realistic. At which point we will engage with a specialized heat pipe vendor such as ACT to begin more detailed design efforts.

Just to make sure I am reading you correctly: it sounds like my concerns were not wholly unfounded, but that there are design features which can be implemented as mitigations. Is that a fair summary?

Would having this localized change in the heat pipe internal wick structure create significant manufacturing hurdles though? I assume these pipes are typically extruded, so having a different cross section for an interior section seems like it would be cumbersome to fabricate.

@IRstuff
Yes, agreed the heat pipe design will need to be different to accommodate the added heat load. My goal is to resize the radiator to result in a similar dT across the heat pipes as we were predicting prior to the added heat load. My concern is over the the fact that I will now have essentially a more localize heat flux, as opposed to a fairly uniform one. I wasn't sure if this invalidated any of your typical/ideal heat pipe operation assumptions? Or if, like davefitz alluded to, there might need to be local design features/changes implemented to account for the higher heat flux in one location?

 

[IRstuff]

A graphic or figure would help; I've actually no sure idea what your layout looks like

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[3DDave]

If the heat sources are in series then having the high heat input anywhere but the most distant from the radiator seems like it would short-circuit more distant heat inputs. If they share fluid parallel to the radiator as a common, shared fluid volume, what directs a proportional amount of coolant back to each channel?

Does this work even if they are all identical?