Can exhaust heat be used to reduce automotive drag?

[boneh3ad]

I am completely clear on what you are proposing and have been for some time. My last post essentially proves that a car does not produce anywhere near enough hot gas to create a bubble of warm air around it and so any gas expelled forward will just remain in the boundary layer anyway. That would not modify the effective shape of the car and so C_D would remain the same and the density used in calculating drag would remain the same free stream density as it is without any forward exhaust.

Now, supposing that the engine actually did make enough hot exhaust to generate a bubble of any meaningful size that would actually change C_D (by changing the effective shape of the car seen by the free stream), that could possibly be advantageous purely from an aerodynamics standpoint. The bubble shape would have to be just right, though, as a car typically has a lower drag coefficient than does as sphere. However, even if this was done, you still have to push all of that incoming air out of the way to make space for your bubble, so the engine is now working extra hard to eject a stream of exhaust forward with enough force to create a bubble out in front of the car. Even doing this a foot in front of a car would require quite a bit of forward momentum in the exhaust. Now your forward exhaust, which requires more engine power to generate, is also acting as a forward-acting thrust. You've now double penalized yourself.

I also have no doubt that on a busy highway, the temperature will rise above ambient by a few degrees due to passing cars. Of course, this already happens, so it's not that much of an improvement.

Finally, if your car is driving into a warmer pocket of air, the lower density means your car engine is going to run less efficiently since it's taking in less air per volume for combustion.

 

[jack action]

The question raised by Q_Goest could be answered easier in another form:

Do cars driving in cold climates have more drag than cars in hotter climates?

I don't have a verifiable answer to that question, but I never heard of anything about significant differences from region-to-region or season-to-season comparisons.

 

[boneh3ad]

The question raised by Q_Goest could be answered easier in another form:

Do cars driving in cold climates have more drag than cars in hotter climates?

I don't have a verifiable answer to that question, but I never heard of anything about significant differences from region-to-region or season-to-season comparisons.

They do, but their engines also run more efficiently. Cars have less drag in Denver than the do in Houston due to altitude, too. These effects aren't gigantic, but they're finite.

 

[Q_Goest]

Not sure why there's so much confusion here.

However, even if this was done, you still have to push all of that incoming air out of the way to make space for your bubble, so the engine is now working extra hard to eject a stream of exhaust forward with enough force to create a bubble out in front of the car. Even doing this a foot in front of a car would require quite a bit of forward momentum in the exhaust. Now your forward exhaust, which requires more engine power to generate, is also acting as a forward-acting thrust. You've now double penalized yourself.

Certainly it could be done, but you didn't answer my question. It seems to me that the suggestion is akin to mounting a jet engine on the wrong side of the car: heating and expanding air is what they do!

Does this help?

When I say the hot air (coming from out of the exhaust and also out of the radiator for example) mixes with the air, I meant that it simply disperses to heat the surrounding air just as it would for any car driving down a road. I'm imagining an exhaust pipe sticking 10 or 20 feet in front of the car with a header on it to disperse the hot exhaust as the car drives through it for example. The radiator could also be stuck way up ahead of the car like that. It really doesn't matter. The point would be only to try and imagine dispersing the amount of heat into the atmosphere ahead of the car somehow. How that's done isn't important. Then, assuming a constant specific heat, the temperature of that air will vary linearly depending on how much cool air mixes with hot air.
We might change the original question just slightly to help clarify what I’m thinking. Imagine for example, we have a single car driving down a straight road with another car behind it and ask if the car behind it is going through air that is any warmer than the one ahead of it.

 

[boneh3ad]

Not sure why there's so much confusion here.Does this help?

And we have repeatedly said that the amount of heating as a result is so minuscule as to barely register and doesn't address the fact that outside of the heated region, you are sill pushing full-density air out of the way. Further, you would need to disperse that heated air so far ahead of the car that you would either need a large amount of forward thrust to get it far enough ahead toaster or you would need a 15 for pole sticking out of the front of your car to dispense the gas. Otherwise the effect will be confined to the boundary layer, which has already been addressed several times.

 

[cjl]

Consider that my car with a 2.3 L engine running at 3000 rpm at 70 mph down the highway is going to pass about 0.15 kg/s of air through it (2.3 L * 1/1000 m³/L * 1.29 kg/m³ * 3000 rpm *1/60 min/sec).

Not that it really makes any difference to the final conclusion, but unless your 2.3L engine is a two-stroke, this number is a factor of 2 too high. Each cylinder takes in air on every other revolution, so a 2.3L engine only pulls in 1.15L per revolution.

 

[boneh3ad]

Not that it really makes any difference to the final conclusion, but unless your 2.3L engine is a two-stroke, this number is a factor of 2 too high. Each cylinder takes in air on every other revolution, so a 2.3L engine only pulls in 1.15L per revolution.

Good catch. That will teach me for posting while jet lagged.

 

[David Morgan]

This turned into a fun thread!

To clarify my original position a bit, I wasn't really imagining dispensing the actual exhaust gases to the front of the car. My original thought was somewhere along the lines of creating a front bumper out of the exhaust header.

Especially in turbo charged engines, the exhaust piping near the turbo and the turbo housing itself can glow red to orange hot under WOT conditions... the color change of the steel would indicate a surface temperature somewhere around 500C to 800C.

While certainly unrealistic to sustain such temperatures, as an academic exercise, what effect (if any) would an 8cm diameter by 1m steel cylinder at a sustained temperature of 500C to 800C have if it was the first part of the vehicle to contact the oncoming air stream?

 

[Q_Goest]

Hi David. As an acedemic excersize, it's an interesting question. As a practical application, it's highly impractical because you need to put the heat way out in front of your car as I'd suggested earlier. It won't work to create a thin, hot boundary layer next to your car. That doesn't change drag at all. The point of reducing drag by heating the air is to reduce the density of the air you need to disturb when moving through it.

 

[jack action]

While certainly unrealistic to sustain such temperatures, as an academic exercise, what effect (if any) would an 8cm diameter by 1m steel cylinder at a sustained temperature of 500C to 800C have if it was the first part of the vehicle to contact the oncoming air stream?

This is the realistic answer to that question:

In reality, injecting any meaningful amount of hot air (or air of any temperature) into the boundary layer will dramatically alter the stability characteristics of the boundary layer and likely lead very rapidly to transition to turbulence if the flow was not already turbulent at that point. In that case, the viscous drag will immediately increase tenfold, not with the square root of the temperature, due to the effects of a turbulent boundary layer. That has been my point all along. If instead the boundary layer was already turbulent at the injection point (so somewhere farther down the hood), then the effect would be much, much less.

 

[boneh3ad]

This is the realistic answer to that question:

Given that he expanded his idea, I can expand on my answer. Generally speaking, heating up a subsonic boundary layer, as is the case of having a hot bumper, destabilizes it. The Tollmien-Schlicting waves that are usually the ultimate cause of transition to turbulence on a relatively flat surface like a car hood become more unstable when the wall is hot. This would tend to favor earlier transition with an increase in drag.

On the other hand, if you manage to heat the boundary layer up early like that without causing substantial growth (for example, by doing so in the region before T-S waves become unstable in the first place), then you may get the opposite effect. In that case you would create a hot boundary layer and before the instability takes hold, it enters a much cooler region. In that case, the wall would look like a cold wall to the boundary layer and would tend to stabilize it and quite possible delay transition and reduce drag.

How far it would delay transition (if at all, as I've never seen anyone publish any data on this so it's just a mental exercise at this point) would be the key question. If it's only delayed for a very short distance downstream, the effect on drag would be negligible.

 

[ScooterGuy]

Hi, all.

Ok, rather than adding the hot air to the front of the vehicle to reduce air density, what if we added the air to the wake to increase base pressure?

On page 228 of Wolf Heinrich Hucho's book Aerodynamics of Road Vehicles (4th Edition), he states that using an active means of wake abatement such as base bleeding has shown to increase base pressure (the vacuum behind a vehicle in the wake) by as much as 30%. He also states that using another means, actively blowing air into the wake using a blower (which was used by F1 racers and was subsequently banned, IIRC) dramatically lowered wake drag and more than offset the power requirement of the blower(s).

I'm building a new frame and aerodynamic body for my bike (a high-fuel-efficiency project), and propose to combine the two active means of wake abatement above by using the engine exhaust to power a Coanda nozzle (much like you'd find in the Dyson bladeless fan). The Coanda nozzle will pull air from the engine compartment to cool engine components, and there will be a venturi nozzle at the front of the bike to feed cool outside air into the engine compartment from the stagnation point at the front of the bike. Thus, I reduce the pressure drag at the front by draining off that stagnation point air, I provide cooling air to the engine compartment (the radiators will be cooled separately by two cross-over ducts which will also be used for cross-wind negation), and I increase the base pressure at the wake, thereby reducing drag. All with no extra power required.

What do you think? Would this be feasible?