Peltier Intercooler Theory and Discussion
 

I was thinking about intercoolers and trying to apply some of my knowledge of cooling as it applies to the world of PC overclocking.

I have designed several water cooled PC's and currently run a 2 gig processor at about 2.5 gig.

There are lots of ways to cool a PC chip, as described on this link

http://www.ocforums.com/showthread.php?t=287576

However the most interesting way that might apply to cooling air as it goes into the engine is a Peltier.

http://www.peltier-info.com/4fins.jpg

The way this works is you have two different types of metals and you charge them electrically, and one of them gets really hot, and the other gets really cold. The electricity transfers heat from one metal to the other. In PC applications you then apply cooling to the hot side, and apply the cold side to the processor. However as you can see in the above picture, people have used them to cool tubes before.

A similar technology is used in heating water at the destination in lue of using a traditional water heater. Many Hotels use this technology and you can put one in your house for about 2000 bucks.

My thought was that if you could build a hexagonal metal pipe, preferably made from copper, mabey with some fins on the inside to create more surface area, and then you applied peltiers to the outside and ran them off of the alternator, as they run off of DC power, and typically use about 100-200 watts each, you could cool the copper to about -100 degrees. The outside would get rather hot, but the air passing through the front grill would cool it which would in turn allow more cooling of the inside of the intercooler.

A relatively small intercooler could in theory produce a very decent amount of cooling.

The issue is the amount of power required to run the peltiers depending on how many you use and how big they are. Also there would be an issue with how quickly the air would pass through the intercooler and thus not how as much time to transfer heat to the copper.

I had also thought that you could possibly apply this before it hits the compressor/blower and thus cool the air before it gets compressed. This should allow your supercharger to be more effecient.

One might take it a step further and cool the air again as it leaves the supercharger.

Thoughts?

interesting idea (or rather an application for an idea)...not sure how fesable (spelling?) it is for a car though...haha and i'm sure not gonna be the test for it ;). always nice to see people coming up with interesting solutions to common problems:D

Apparently I'm not the first to think of this idea.

http://www.autospeed.com/cms/A_2088/article.html

This guy however is an idiot. He did it all wrong. He first started out with only two peltiers trying to cool a gallon of WATER. LOL

Next he followed up that problem with bench testing it with no air flow over the heatsinks.

He then upgraded his heat sinks, but mounted inproperly to the peltier, so that heat transfer was restricted to a very small surface area.

He then finds a commercially made example of what he is trying to build, and discounts it without paying the 37 bucks to try it out.

He did have a good idea in using the car's body as a heat sink, for the peltiers. I'll keep digging.

I think this idea would be incredibly good on something involving water (like a radiator or water-intercooler), but falls short when trying to cool air, as air doesn't have nearly the same thermo properties as metals or liquid. You'd basically be trying to cool air that was moving with velocity, over a very small section of "area" and wouldn't really effect the overall volume at all. Aka only the outside air closest to the traveling stream would be cooled, the central volume would be uneffected.

Water on the other hand has excellent thermal absorbtion and would readily cool from the peltiers. (If volume was at a respectiable level)

hey another overclocker :) fancy meeting you here, Im currently running at A64 3200 939@ 260x10 2,2,2,5 using bh-5 and a rambooster :p interesting theory with the pelts...........Id like to see a intercooler filled with frionor some type of accelerated coolant and sandwich layered piping with the same between the inner and outer layer :p

I actually thought about using this stuff to make a real CAI a few months ago. Then just completely forgot about it. My only concern was condensation and amperage draw.

I think the volume of air and the velocity of its movement from outside the car to combustion would be too much when pressing the gas hard and it wouldn't have time to cool it off quickly enough to be so effective. Then you have the problem of idle when not too much air is moving and it could fool/confuse the ECU in its operation as to how the car should run. One last consideration is the air could get so cold at idle while the engine is so hot, it my do some damage, maybe :confused: not sure.


I am not an expert as you can see from my response.

Assuming the efficiency of the engine is about 30% the radiator has to get rid of the other 60% or something. So if your engine generates 238HP, this would be way over 400HP or 300kW cooling power that the radiator needs to generate.
An Intercooler might be a third of the size of the radiator and therefore have a cooling power of maybe 100kW.

As far as I know a Peltier has an efficiency of about 4% (well if its run as an electric generator), so if you want to cool it with a 200W Peltier you might really have about 10W of cooling power. (Even if I was off by 1000% you still couldn't cool much.)

Take a hairdryer and blow on a piece of copper and measure the temperature. Are you still getting -100 degrees? And a hairdryer does never pump as much air as a compressor (not even a leaf blower does).

Peltier are much more effecient than 4% They are more like 60-70% efficient.

However you are transferring heat, not generating it. The heat energy exists already in the air, you are simply pulling it out to put it somewhere else. So the Peltier is moving the energy, not converting it.

Atleast I believe thats what it does.

I was thinking about a 2 foot length of piping, that was essentially a peltier itself in that it was made out of the metals that make the peltier work, coated in ceramics. This tube would be placed in a location where alot of Air would move accross it when the car was running.

As for the rough idle of the engine, the MAF sensor would handle all of that. The car would simply think you were moving through colder or warmer air as the case may be.

i bought a few of these to experiment with last year.

in a nut shell, they take too much current to run, produce more heat than you can get rid of on the hot side, and don't cool very well on the cold side.

but at least i got a few coffee heaters now :)

Won't it be easier to setup an ethanol (or methanol) injection kit to cool the intake temp. I remember from the few things I read about Peltier is that the current drawing is pretty high.