Heatpipe / vapor exchange intercooler?

[Splinter]

Looking back at my nigh-over computer tweaking days, there were three primary ways to cool hot chips

1. Air-Air heatsinks - Cheap, easy to install, low cooling capacity
2. Water/radiator blocks - Expensive, difficult to install, med-high cooling capacity (depending on setup)
3. Heatpipe heatsinks - Expensive, east to install, high cooling capacity

After searching around on google, I couldnt see any mention of using heatpipe systems in an intercooler, but the more I think about it the better idea it seems to me.

Water is heavy. Really heavy. It also takes up bunches of space. So while an AWIC can beat the pants off an AAIC, its no where near as simple

So why not take a bunch of heatpipes, jam the cold ends in the path of the hot turbo end, and connect the hot end to a big radiator in the front of the car?

It'd probably be expensive, but I'd imagine it'd work great, with minimum enginebay clutter and very low lag

[Subtle]

What kind of intercooling is used by the Subaru rally team?

[Splinter]

It would appear to be air-air

But rally cars dont tend to stop at traffic lights often

[Subtle]

Riiigghhht!

The tmic has much more of a heat soak problem than does a fmic. It seems that the main source of exogenous heat to the fmic when the car is stopped would be from the pavement on a hot day.

The latter doesn't happen often in Victoria.

[93forestpearl]

Heatpipe heatsinks? Huh? I'm not shure exactly what you mean. If there was a more efficient, better way to cool the air charge, I would think the Subaru World Rally Team would be on top of it.

I tend to think there is a reason a $750,000 rally car uses an air to air intercooler.

[Splinter]

Heatpipes use a low-pressure refridgerant to move heat around through vaporization/condensation.

My GeForce 5900 Ultra (extremely high wattage video card) is cooled completely passively using two of these.

[93forestpearl]

Hmm. Interesting idea. How fast can they cool when the temp is raised in a fraction of a second. I know/think integrated circuits don't climb in temp as fast as the air chrage coming out of a turbo.

[Splinter]

They should react instantly. Or however long the heat takes to propogate through the material of the pipe wall.

[93forestpearl]

I'm curious to learn more, but I'm having trouble grasping the overall picture from the small pic you posted. Sound interesting though. Maybe someday someone will make a system with that technology that is much more efficient that air-to-air IC's.

Time for bed though; I've got 9:00 class(its 1:30 here).

[Splinter]

A. Heat is absorbed in the evaporating section.
B. Fluid boils to vapor phase.
C. Heat is released from the upper part of cylinder to the environment; vapor condenses to liquid phase.
D. Liquid returns by gravity to the lower part of cylinder (evaporating section).



Thats a very simple example.

More advanced units use a material mesh on the inside surface of the pipe to 'wick' the fluid against gravity, so the heat source doesnt have to be below the heatsink/radiator

[vrg3]

A few thoughts:

A heat pipe must be mostly vertical, since it relies on gravity to recirculate the refrigerant. This creates packaging concerns.

I don't how how much power your NV35 draws exactly, but it's probably somewhere around 120 watts, right? Consider that the stock AWIC radiator has a rated cooling capacity of about five and a half kilowatts.

[Splinter]

So instead of 2 heatpipes, it'd need 50.

Hmmm...

[Legacy777]

Even if you get the size and mounting orientation, I don't think the heat pipe is going to react as quickly as you think.

If you can find time data on how long a complete cycle would take, it may help.

[professor]

heat pipes MOVE heat, they don't get rid of it. So the use would have to be to move the heat from somewhere it is, to somewhere you want it where you can transfer that heat. At that point you still have the same problems and solutions with regards to getting rid of the heat (transferring it to the atmosphere).

similar to heat pipes is to use blocks of very fast-conducting materials like copper and copper-beryllium alloys, one end of which is placed in the hot medium, and the other, in the tranfer medium either air or water.

the usual use of either strategy is to extract heat from areas where nothing else will fit. They are used extensively in the cores of plastic injection molds, where the heat is moved out to water lines where it is carried away to water / water exchangers normally. Note that if you have rom for water lines in the cores, you don't need the heat pipes. The same would go for intercoolers where there is plenty of room for the water normally.

[rallysam]

Seems pretty similar to an AWIC in terms of advantages/disadvantages. The only difference is that you choose a fluid that happens to change state when it's heated, but from the outside you're not going to see a functionally different system. Because there's no compressor in this system, this system works just like an AWIC in that it just picks up a little energy from the hot air and loses a bit to the ambient air. If you actually used this for a car instead of a computer, I'm not sure why you wouldn't circulate the magic fluid around with a pump (just like they do in an AWIC system). Obviously this system is going to be more complicated and difficult than doing the same thing with water.


Pro: AWIC and this magic system both give you more freedom to position the radiator further away from the throttle body to avoid heat soak.

Con: AWIC and this magic systm both add an extra loop and an extra heat exchange step that make them less efficient than air-air systems..... unless you're vikash and have a cooler full of ice water in the rear seat circulating through the system .