Brainiacs: Thermocouples on Internal Combustion Engines?

[Josh]

So I was thinking (I know, that can be dangerous), one of the biggest drawbacks to the internal combustion engines is that they are not very efficient - way too much heat is produced, which is of course lost energy that could be used to propel the car. So then I thought, isn't there a way to easily capture that energy and convert it to - at very least - act as a alternator and recharge the car's battery?

So I did some Googling (again, this can be dangerous) and found out that a cheap and effective way to do just that is a thermocouple, and that I am not the first one to think of this idea. There are a couple of papers out from the 1990's which explored this exact idea, and a bunch of patents how to do it.

My question is, why isn't this done? If a thermocouple could produce enough energy to keep the car's battery charged, it would alleviate the need for an alternator and increase gas mileage. Thoughts?

[quexos]

Originally Posted by Josh

My question is, why isn't this done? If a thermocouple could produce enough energy to keep the car's battery charged, it would alleviate the need for an alternator and increase gas mileage. Thoughts?

Obviously because whoever comes with something efficient, the oil industry will buy the patents to make sure it does not come out so that gas keeps selling at the pumps.
Why do you think those revolutionary engines on water, air, or half a gallon of gas for 60 miles or so, have been invented so many years ago and never came out to the market ?

[Septimus Prime]

I thought it was because petrol is still more power efficient.

[Mr. Hanky]

Just so we get this cleared up early in this topic, an IC engine isn't inefficient because it creates a great deal of heat. Actually, creating heat is the answer that leads to efficiency. The problem is being able to harness that heat once it has been generated. The IC engine encounters limitations to potential efficiency because it has a finite tolerance to high temperatures (which comes from the heat, of course). The higher the temp, the shorter the lifespan of the components. If you could construct the engine such that it could withstand very high temperatures w/o degrading, it would then be able to harness the energy for optimal efficiency. Short of that, the only available option we have right now is to skim off the readily available energy (heat-induced volumetric expansion inside the cylinders) and dissipate all "surplus" heat to stabilize to a nominal operational temperature (as opposed to letting it operate at very high temperatures). The dissipation is the part that makes it inefficient.

As for thermocouples, I'm not familiar with how they work to convert heat to electricity. However, I imagine to convert a considerable amount of heat, it would take some beefy devices. Most likely, it is not a lossless process, itself, as well (so you would still only get partial recoup, which would impact potential efficiencies, once again). We would no longer be talking about little stick-on devices to do the job. Therein would lie serious obstacles as far as packaging and cost. I would certainly be curious to see a working prototype that demonstrates this technique, though.

[DiverSpear]

Because a thermocouple is used for measuring temperature. It does create a voltage, 2 dissimilar metals react to temperature creating a voltage. The problem is that resultant voltage is milli-volts DC or mVDC.

[Josh]

The reaction may be small, but if you can control the reaction and perform it millions on times, you should be able to get enough voltage to recharge a 12v car battery. Theoretically. One thing I haven't been able to find is how much volt/amp/current a thermocouple generates, and if the generation is dependant on the size of the thermocouple (is running one huge thermocouple equal to running millions of micro thermocouples in series?).

Toyota is planning to add a solar panel to the room of their new Prius, which will be used to partially run the AC compressor. Solar panels can't generate much in the way of power, but if a thermocouple attached to an engine block could produce more, why not give that a try?

[Blu-Dog]

Originally Posted by Josh

The reaction may be small, but if you can control the reaction and perform it millions on times, you should be able to get enough voltage to recharge a 12v car battery. Theoretically. One thing I haven't been able to find is how much volt/amp/current a thermocouple generates, and if the generation is dependant on the size of the thermocouple (is running one huge thermocouple equal to running millions of micro thermocouples in series?).

Thermocouples are very slick. They're simply two dissimilar metals, joined at an alloy point, and application of heat stimulates the flow of electrons from one metal to the other. The result is a very small, but measurable, production of electric current.

Two problems, though.

1. Current is produced in a very inefficient way. Even if you increase the surface are of the two metals, you don't get a linear increase in output. In fact, you get a loss of output; the larger area requires more heat to generate power, and heat creates resistance. Most thermocouples are extremely small for this reason.

2. The eutectic (the melting point of the two metals where they alloy) goes down as efficiency goes up. The lower the eutectic, the higher the output, so more heat doesn't create more power; it burns out the thermocouple. Metering devices for really high temperatures use thermocouples, but the output us much lower, and meters must be more sensitive to detect the lower currents. A home oven has cheap thermocouples to operate a gas valve; a stee furnace has a much more expensive setup for measurement.

Anyway, as a concept, it works; as a practical method of power generation, there are huge hurdles to overcome. It's not impossible, but it's not an on the shelf, unused means of getting things done.

[Josh]

I knew someone whould know the answer to this question Thanks for the info!