Turbos use the heat of the exhaust system to turn the impeller producing boost for the engine. Many cast turbo manifolds designed for optimal flow work tremendously well and obviously have more metal to them than typical headers. Would it be worth it to explore ways to use thicker walled exhaust tubing or even double walled for the hot side piping? Coatings and wrapping help but can more be done to make the system even more efficient? What benefits could be expected from a thickwalled well flowing hotside system? Discuss please.

 

Actually, from a heat transfer perspective, you want a thin wall tube to retain heat in the exhaust gas. I know this is not intuitive at first glance but consider the following: Metal is a good conductor of heat. Air is a poor conductor of heat (a good insulator). The thicker the metal wall, the more efficient the heat transfer out of the gas and into the surrounding metal.

Now imagine a metal tube with a real nice insulator around it, like air, a thermal coating, or header wrap. A header wrap made of metallics would only increase heat transfer out of the exhaust gases, when compared to the for mentioned materials, because it has a higher heat transfer coefficient.

Another approach. Think about an uncoated cast iron manifold and how incredibly hot it gets on the outside surface compared to coated header. It's measured surface temp would be greater than a thin walled tube because the heat transfer is greater, and has transfered MORE heat out of the gas. This is not a good thing for retaining turbo energy.

So yeah, it does seem like an infinitely thick header would insulate the gas, but its the exact opposite, the high heat transfer coefficient of the metal transfers heat quicker and more efficiently out of the gas. Just think of that huge mass of metal hogging all of the heat energy.

-Scott.

 

At warmup, thicker metal (more mass) would mean more "hogging energy" as the metal warms, which may or may not have a significant impact on exhaust gas temperature. But once it's at operating temperature, under steady state conditions, a thicker wall will indeed be a better insulator. The properties of the air surrounding are identical in either case, so that's a moot point. Infinitely thick would be infinitely insulating, infinitely thin would be infinitely conductive.

 

Um a thickwall and a thinwall pipe will both heat up to the same temperature, and in a matter of minutes or seconds. Once heated up there's no difference.

The best thing you can do, is use a ceramic thermal barrier coating like HPC. This will greatly slow down heat transfer from the pipe to the surrounding air.

 

?

Infinitelly thick would never heat up and reach equalibrium, it would therefore continuously sink heat.

Very thin pipe would heat up very quickly and therefore sink heat only until it has become heat saturated.

Once heat saturated, a thin and thick wall pipe would dissipate heat into the air at basically the same rate because their surface areas are roughly the same.

Cast manifolds are thick because it would be hard to cast a super thin one. Cast is more brittle than mild steel and has different thermal expansion and cooling properties. It would be more prone to cracking if cast super thin.
Exhaust manifolds also carry the (more turbulant) heat of 4 cylinders as opposed to the thinner walled header which carries the heat of just one cylinder in an orderly flow.

 

Other way around. A thick pipe will draw more heat from the exhaust gas which you don't want. A thin pipe with a nice isulator like air around it is going to keep heat in the gas better.

Your using the exhaust energy to heat up the mass of metal. A larger mass will 'sink' more heat from the gas.

Me thinks I need to blow the dust off the ol' thermo book

-Scott.

 

I agree with the concept here of 'continuously heat sink'. Regardless of what the surface temp ends up at - you are going to end up with a cooler gas in this scenario than if you would have had air (insulator) at relatively close distance to the exhaust gas (ie, thin wall).

-Scott.

 

Excellent replies everyone. I was actually thinking of it more along the lines of once the metal got up to operating temp, it would then insulate the exhaust gas. Thin walled makes more sense though as the thicker metal would need a constant supply of heat to maintain its temperature. The thin metal would also need a constant supply of heat but it would be much less.

So, if air is used as an insulator, how do you think a double walled pipe would fair? This design would have an air space between the walls. Nothing drastic just a small amount. Say 1-5/8" primary tube with a 1-7/8" or 2" outer tube. We could even get crazy and coat the inner tube in and out and then inside of the outer tube before sliding it over top the inner tube.

 

I honestly believe your temperature gains would be negligible over just using a ceramic HPC style coating.
Time would be better spent optimizing flow efficiency to the turbo rather than trying to nuke it to gain exhaust gas velocity.

 

Quote:

Excellent replies everyone. I was actually thinking of it more along the lines of once the metal got up to operating temp, it would then insulate the exhaust gas. Thin walled makes more sense though as the thicker metal would need a constant supply of heat to maintain its temperature. The thin metal would also need a constant supply of heat but it would be much less.

That's essentially it but be careful thinking steady state/heatsoak of the metal. Heat is constantly being lost to the surroundings through the pipe. Heat 'flows' through a material from hot to cold. Metal has a higher thermal conductivity than air, ceramic, header wrap and would tend to simply dissipate your exhaust heat. The more metal you are heating, the more energy it takes because it becomes better at flowing your heat out of the pipe. Best to get a good insulator on there asap.

Quote:

So, if air is used as an insulator, how do you think a double walled pipe would fair? This design would have an air space between the walls. Nothing drastic just a small amount. Say 1-5/8" primary tube with a 1-7/8" or 2" outer tube. We could even get crazy and coat the inner tube in and out and then inside of the outer tube before sliding it over top the inner tube.

Ah ha, now I see what your saying about a double wall. All my previous comments were related to simply wall thickness. This is the same concept as a double pane window in your house and I think it would work to some extent. The hot exhaust is like the interior of your house. In the Winter, the heat in your home tries to move to the colder outside ambient through a medium (glass in this case) which has a thermal conductivity associated with it. Adding an insulator like air between two sheets of glass significant reduces the heat transfer to the second pane, because air is a pretty good insulator.

So heat is flowing through the cross section of the first pane and then boom - hits the insulator which greatly reduces the transfer. The air sloooowly transfers heat to the second pane where it is then lost to the atmosphere, mostly via convection. I don't see why the same concept would not work to insulate the exhaust.

Good idea. Biggest drawback might be difficulty to fabricate on a tube header, large increase in overall pipe OD, etc.



-Scott.

 

Check the ol' heat transfer book too - delta T is proportional to d (distance)...

 

There's no reason this wouldn't work, though I can't see it being practical. Multiple panes of glass work to create a layer of stagnant air, reducing convective transfer, while still being transparent. Everywhere else in your house's walls, doors, etc. there is insulation - no double paned siding or drywall. Insulation is engineered to be thermally insulating; draw a parallel here to something we use in the automotive world - header wrap.

Here's another practical example of heat transfer principles - when you try to keep your body warm in a cold environment, do you put on thinner or thicker clothing?
Thicker, to reduce heat transfer, maintaining a great dT between you and the environment.

 

Darnit Evil. I promise to revisit heat loss in pipe this weekend. Looking briefly at Fourier, it does seem that enlarging outside radius with everything else constant reduces heat transfer rate. However, does not take into account increased surface area and convective heat transfer which would increase.

Prior to (potentially) looking like a fool - I'll stop now

-Scott.

 

I didn't take that into account either, but I think we're talking about small enough changes in surface area that it is negligible.

Now really stir things up - should we consider radiant transfer from the outside of the pipe? What's a typical exhaust gas temp? Too bad my heat transfer bible is at work... do you have any emissivity data/curves in your possession?

 

Hey Guys,

First, I want to apologize for conveying some concepts/analogies that were not entirely correct. It's amazing how much you lose an a few years!

Anyhow, from a conduction perspective, Evil has got it right. Even a material with 'high' thermal conductivity can insulate! For a radial pipe, conduction heat transfer depends on a coefficient (material property), temperature difference across the medium, and distance across the medium. The larger the distance - the more resistance there is to heat flow - the less heat transfer to the surface (Heat must get to the surface before it can be dissipated). In a plane wall, the gradient can be expressed by dT/dx. For a radial system it's expressed dT/ln(r2/r1). Where r2/r1 is ouside pipe radius over inside pipe radius. It still means that the thicker the medium, the less heat will transfer - But in a flat wall the gradient is linear - in a pipe it is logarithmic. That's kinda cool, for a pipe you can increase and increase outside diameter but heat transfer decreases with diminishing returns. You can do this in excel to see for yourself or I can email someone if they are interested.

Let's take a step back and look at what is happening from the exhaust gas all the way to the atm. Below is a link to a solution comparing outside dimension of two pipes. The only thing that changes is the pipe thickness - you can see that convection from the pipe outside surface to the atm is the single biggest reason you lose efficiency in a thicker pipe. (see, I told you I would study this weekend)


http://i296.photobucket.com/albums/m...atTransfer.jpg


Summary of what you see in the attached:

CONVECTION (gas to inner wall)
First we have a high speed exhaust gas racing though a pipe. The gas will transfer heat to the pipe wall via convection. I'm sure there is a better way to compute this (looking at turbulent internal flow), but it doesn't matter for our comparison on outside pipe dimension - because this transfer remains the same for both scenarios (R1A = R1B)

.....R1, R2, R3, in the attached are resistance to heat flow BTW. So big resistance means small heat transfer.

CONDUCTION (inner wall to outer wall)
Like stated earlier more thickness makes it harder to transfer heat to the outside surface. Make the outside dimension 30feet and you could certainly hold you hand on the outside - not much transfer to the surface.

CONVECTION (outer wall to free-stream atm)
Dependant on outside surface area and properties of the air flowing over the pipe. All of this assumes 'natural convection'. Air velocity is small and only occurs from heating the air close to the pipe, which rises, colder air sinks, and you end up with some mixing of air.

Side Note: What happens under the hood at 100mph? Probably significant ram pressure to flow a lot of air over your hot-side piping. 'Forced Convection' should be used to see what affect high speed air has on removing heat......the faster the air over the pipe, the more convective heat transfer.

RADIATION (outer wall to surroundings)
Another way heat can leave a surface (the other was conv.) None of the attached takes into account Irradiation, Reflection, etc. You'll see surface energy emission is dependant on a surface property of the material (epsilon), temp differential between surface and surroundings (in Kelvins) and outside radius. As outside radius increases, so does emission power......but, the thicker the pipe wall gets, the lower the surface temperature to the fourth power.....so larger pipe seems to maybe have less radiation than a thinner one at the same gas temp. They are close in the example and not significantly different.

Side Note 2: Stefan-Boltzmann Constant is sigma

That should pretty much do it. You can see that there is 7.3% more heat loss in the thicker pipe. I would conclude that you want the thinnest possible pipe thickness while maintaining structural integrity to put MORE heat into your turbo!

Thanks guys,
Scott.