It will make a good senior project, but sadly not a commercial product for intercooling. For a hint to why, look at the BTU/minute heat rejection requirements from the intercooler. Get that from the compressor maps, coupled with the engine airflow requirements at redline. You will calculate air mass flow, and intercooler inlet temperature. Pick a reasonable intercooler outlet temperature and you will get the heat flux.

Peltier devices are great for moderate delta-T's with low heat fluxes (like cooling laser diodes and infrared detectors) but they don't do squat in high heat flux applications.

Good thinking though, don't stop.....just scale it down, like 2 orders of magnitude. How about a blown 16cc engine?

Regards,
BigMoose

 

good idea. although, i honestly think it would not work. first of all, you would need some way of getting the heat out of the hot side of the tec, probably some form of water cooling. second, you would need A TON of tec's to cool the average-sized intercooler. and third, the air passing through the intercooler already hardly has enough time to get rid of most its heat, by transfering it into the intercooler. the intercooler acts as a heat sink. even if the actual intercooler was colder that only means it has the potential to absorb a larger quantity of heat.

worth trying though....
i think peltiers should stick to cooling cpu's

 

all good thoughts. after speaking with some experts in this field and crunching some numbers it seems that the power used by the modules will be greater than the additional power from the engine. also, the fact that the weight of the system(heat sinks, batteries, numerous modules) would be a lot heavier than a nitrous or water cooling system. doh.

 

good stuff to give the brain something to chew on for a while.

 

You're right - I did a MASSIVE in-depth design project on something like this, and it was just for the purpose of cooling a multiple laser diode system, but it ended up getting just absurd - once you needed more than a few TECs, the power requirements (and the ENORMOUS heatsink and fans required to cool the TECs) made it just wayyyy beyond practical.

 

I was thinking about this last year. TECs require LOTS of power, and most of them are 24v if I remember correctly, so you have to calculate how many watts of heat they will dissipate at 12-14v. Also, any TEC that does a decent job of cooling is going to need either a gonzo heatsink on it, or could be water cooled.

The thing I would wonder about is how well could the hot *** water from an engine cool it. I imagine it would cool it a little - if you've ever messed with a Peltier you'd know that if you run one without a heatsink or some kind of cooling on the hot side that the guts will melt because they get so hot.

Last, if you could power and cool the elements sufficiently, I bet it would work ok since the only thing the elements are cooling is hot air.

 

carb cleaner cools that aluminum really fast and does it really well
you can get that intercooler clean too....

 

Good work, now I'll give you another idea to ponder. Lets base this all on a quarter mile run.... 'cause it won't work for a road course or any extended period of time for that matter. Lets start with Liquid Nitrogen (LN2), and subcool that inlet manifold down. Take the expanding nitrogen gas and drive an expansion turbine (which may be the turbine on a turbo).

Do you have enough energy to drive the compressor for forced induction just using intercooler heat? When you find out you need a little more heat, how about letting that expanding nitrogen get friendly with the exhaust after it has been thru an intercooler and before we send it back thru the turbine. Want fast spool up? Liquid N2 is your friend.


There is a cycle you can play with and couple to the Otto heat engine cycle. You can do all sorts of heat recovery and power boost. Always start with LN2 at pressure add heat and go from there. Have fun with it. It will give your thermodynamics a true workout. Maybee a senior project there?

Regards,
BigMoose

 

I was planning on doing this to my 89 turbo2(rx-7), but one dilemma....these thermoelectric modules aka peltier elements consume major amount of electricity. To cover the Intercooler...it'll take so much power that even 2-3 batteries won't hold up. Also these peltier elements are pretty expensive.

Its use to water cool CPUs in computers are feasible because you don't need that big of an element.

Jay Kim

EDIT: aquamist system would be more sound, both financially and performance wise.



[Modified by AllMotorMonster, 12:44 PM 11/19/2002]

 

hahaha..