F1 Exhaust System

[hecti]

Hey guys,
I was just wondering about an exhaust related question the other day.
I heard some talk that mclaren were running butterfly valves in their exhaust systems during the late 90s early 2000s in the merc v10s. Is this true? I know that the noise they made was special and was caused by something in the exhaust, but where they butterfly vales diverting exhaust gases into different exhaust sections?
thanks

[FatFangio]

The larger sensor is a wide band oxygen sensor. Unlike a 'narrow band' exhaust gas oxygen sensor, these are capable of measuring mixtures other than stoichiometric (chemically ideal). They will give reasonably reliable results from 10:1 to 20:1.

These sensors are the LSU4 made by Bosch and have been used in some road car applications.

They can be used for closed loop control, but F1 teams may choose to use them for monitoring only because a sensor failure, sensor response time and accuracy, exhaust gas leak, wiring problem, etc. could cause the mixture to go too rich (==slow) or too lean (==kaboom).

There is a page on how these sensors work here: http://www.megamanual.com/PWC/LSU4.htm

FatFangio

[adam2007]

just listened to schumacher 2001 3.0 v10 on baby what a sound

[928S]

Yes those sensors are lambda and EGT, lambda is just another name for oxygen sensor, one thing in the article that I would query and I can get an answer next time I speak to them is that the material fatigues due to stress. Now I agree that Inconel does creep, you should hear it when you weld these pipes, it is a bit creepy. However it is very tough and I would have thought the vibration would be worse as they are not supported.

I am going to use a refabricated Nascar inconel exhaust on my road car and I spoke to the manufacturer who makes F1 exhausts for various teams and at least in my case (albeit a supported exhaust) they said not a worry in the world and my pipes are only 0.7 mm thick. They do go down in pipe thickness to 0.6 mm but that is not a big difference. I also agree with the previous answer about Ti, it just can't take the heat as well as inconel although I am told it flows better when welding than the inco which has a rather stodgy weld pool. Inconel btw is more expensive than the Ti also.

So I really doubt that the lower team throw their exhausts out after every race, they are quite expensive headers. The other little tid bit is that they often use the inco in coal fired power stations and it lasts decades and nothing is better than it. Other materials need periodic replacements.

[flynfrog]

Yes those sensors are lambda and EGT, lambda is just another name for oxygen sensor, one thing in the article that I would query and I can get an answer next time I speak to them is that the material fatigues due to stress. Now I agree that Inconel does creep, you should hear it when you weld these pipes, it is a bit creepy. However it is very tough and I would have thought the vibration would be worse as they are not supported.

I am going to use a refabricated Nascar inconel exhaust on my road car and I spoke to the manufacturer who makes F1 exhausts for various teams and at least in my case (albeit a supported exhaust) they said not a worry in the world and my pipes are only 0.7 mm thick. They do go down in pipe thickness to 0.6 mm but that is not a big difference. I also agree with the previous answer about Ti, it just can't take the heat as well as inconel although I am told it flows better when welding than the inco which has a rather stodgy weld pool. Inconel btw is more expensive than the Ti also.

So I really doubt that the lower team throw their exhausts out after every race, they are quite expensive headers. The other little tid bit is that they often use the inco in coal fired power stations and it lasts decades and nothing is better than it. Other materials need periodic replacements.

Are you sure that your road car is going to get them hot enough. Inco needs to re-anneal from time to time while in use or it will start to work harden and crack.

[ringo]

I know this sounds like blasphemy becuase of the implications on the car's volumetric efficiency and power, but can there be a case where the packaging and and thus aerodynamic benefits of a certain manifold design, (for example log manifold ) actual give it precedence on an F1 car over an equal length manifold?

There would be a noticeable loss in power, but how much power really?
The log manifold has some things going for it if we think about cooling flow through the packaging etc. Yes there are issues with tuning and pressure waves etc.

After seeing renault and their front exhausts you cant help but think engine power can be sacrificed for other benefits.



these do take up quite some space. To be clear I'm not endorsing the log manifold to be used any time soon, just going out on a limb on this one.

There must be a way to package an equal length manifold more efficiently, even with the use of oval sections for aero benefits.

The log, or something close to it would be the desperation move. Especially for an engine with power to spare like the Mercedes.

[riff_raff]

ringo,

A modern, stepped F1 header design produces lots of additional torque at its design point. It's more than worth the additional weight and space it occupies.

With regards to a "log" manifold design, it would have one serious drawback, in that there would likely be unacceptable flow interferences and short circuiting between adjacent cylinders with the long duration valve timings that an F1 engine uses. With 4 exhausts discharging into the log manifold at 180deg intervals, and an exhaust period of maybe 320(?) crank degrees per cylinder, there would always be two cylinders interfering with each other. The header avoids this problem by isolating the flows with individual primary tubes.

Of course, another alternative is using individual primaries and no collector. It would not be as efficient as a merged header, but it sure looks cool on Top Fuel and Funny Car drag racers.



riff_raff

[PlatinumZealot]

This might sound crazy, but I feel that the flow out of those top fuel engines is so high, that there is a giant "suction" after each exhaust pulse so fast and so powerful that the air imediately surrounding the opening of those 4 pipes is the collector itself! In other words the pulses are so strong and fast that you don't need a collector pipe for the pulse effects to reach the other pipes. Sounds plausible?

[buzzmatrix]

.....At peak revs, a formula engine will blast out exhaust gases 95,000 times a minute. To scavenge maximum power the exhaust pipes need to be as short as possible. Unfortunately , to help generate maximum torque and responsiveness at lower revs, longer splender pipes are called for. As F1 regulations don’t permit variable-geometry exhausts, the answer lies in the best possible compromise......

I was wondering why should the exhaust pipes be long for higher torque and responsiveness ?

[Edis]

Silicon nitride (ceramic) turbines are very rare, even in racing applications.

For the turbine housing there are several options. Up to roughly 700 degC grey iron can be used. Up to 900 degC nodular iron, Ni-resist, Si-Mo alloyed iron are possebilities. Above 900 degC heat resistant casting steel and stainless casting steel such as HK30 are options. Garrett racing turbochargers usually use thinwall HK30 castings.

This is not my impression, both Toyota and Nissan are using ceramic turbines on their top-models. Another popular application is ceramic coating of the turbine housing, in order to contain the thermal energy.

Ceramic turbines are rare in both production and racing, but that does not mean there are exceptions. The IHI turbocharger used by the Honda RA168E F1 engine had silicon nitride turbines. Nissan have used it in some of their production engines, first introduced in the mid eighties - marketed as a worlds first by Nissan it reduced the inertia of the rotor by about 35%. These units were supplied by Garrett and saw production rates of about 2000 units per month. Toyota also used ceramic turbines on some domestic engines like the JZ with a claimed inertia reduction of 60% (export version had conventional metal turbines). Like the Nissan turbochargers the ceramic turbines had a tendency to fail during high turbocharger speed and temperature, so for tuning they were usually replaced. Mitsubishi have also used turbines made of titanium aluminide on a few production cars, but this is in even more insignificant numbers. Neither solution have had any significant market share compared to the industry standard nickel based superalloy.

As for ceramic coatings, I've never seen a real ceramic coating applied to a turbocharger turbine housing. And no, ceramic powder containing paints don't count. Garrett do however use carbon cloth/metal foil heat insulation around some of their racing turbochargers. This is mainly to keep the temperature down in the engine bay.

[Edis]

.....At peak revs, a formula engine will blast out exhaust gases 95,000 times a minute. To scavenge maximum power the exhaust pipes need to be as short as possible. Unfortunately , to help generate maximum torque and responsiveness at lower revs, longer splender pipes are called for. As F1 regulations don’t permit variable-geometry exhausts, the answer lies in the best possible compromise......

I was wondering why should the exhaust pipes be long for higher torque and responsiveness ?

When the exhaust valve opens a pressure pulse will form in the exhaust port. This pressure pulse will travel downstream the exhaust pipe, roughly with the speed of sound, where it will meet the area increase in the collector. This will cause a reflective pulse to form, but this pulse will have a negative pressure instead of a positive. This pulse travels upsteam the exhaust pipe and correctly tuned, it will arrive when the exhaust valve is about to close. So with the correct length, diameter and area ratios the exhaust system will decrease the pressure in the exhaust port at the correct time which improve the gas exchange and increase volumetric efficiency.

Generally, longer pipes means that the exhaust system is tuned to work well at low engine speeds, while higher speeds require a shorter length.

[marekk]

This might sound crazy, but I feel that the flow out of those top fuel engines is so high, that there is a giant "suction" after each exhaust pulse so fast and so powerful that the air imediately surrounding the opening of those 4 pipes is the collector itself! In other words the pulses are so strong and fast that you don't need a collector pipe for the pulse effects to reach the other pipes. Sounds plausible?

Only if the speed of exhaust gases is near it's mach number, and shock waves start to form, changing local pressures, me think.
No idea if it's the case, though.

[Edis]

This might sound crazy, but I feel that the flow out of those top fuel engines is so high, that there is a giant "suction" after each exhaust pulse so fast and so powerful that the air imediately surrounding the opening of those 4 pipes is the collector itself! In other words the pulses are so strong and fast that you don't need a collector pipe for the pulse effects to reach the other pipes. Sounds plausible?

A collector is simply an area increase, releasing directly to the atmosphere will have the same effect. But when releasing to the atmosphere each cylinder is on its own so to say. This will cost low speed torque, but for a Top Fuel engine this is not a problem since they don't use low engine speeds.

[Sayshina]

I thought you were refering to an accumulated raft of data that allowed computer simulation of exhaust gas dynamics for the design and manufacture of tuned exhaust systems?

Mathematical models don't use anything but governing equations that you can program in. They require no 'data' to build.

Just like a suspension system can be modelled by a mass-spring-damper system. It doesn't matter what suspension type you are using they all obey the same laws of physics.

This is exactly the same for air flow.

But it's not just air flow. There's a lot more going on inside there. I'm reminded of the engine designer who was talking about the intake side, thought they had good sims and knew what was going on, then discovered that in reality fuel was being injected into 1 intake, being spit out, and injested and used by another. This was a guy with decades of experience, and he seemed genuinely surprised.

As to why Titanium might be used more on bikes than cars, I'd point out that in most cases bike motors and chassis tend to be finished to a higher level than cars. They're on display. The public seems to prefer the looks of Ti over Inconel. Add to that the lower heat rejection of a bike motor, for various reasons (much greater drivability needed, requirement to not cook the rider), and the normally greater cooling available (most bike exausts are at least partially in free stream) and it seems to be a done deal.

On the straight manifold idea, when studying for a vessel engineering license the book said straight vs. tuned manifold were exactly equal in overall efficiency. You trade a general reduction in efficiency across the board for a much bigger increase in efficieny at a couple of specific engine speeds.

Not that you can read too much into that, I'm talking about V16 2 stroke deisels running at max. speeds I'm pretty sure an F1 engine can't even idle at.

[Jersey Tom]

Mathematical models don't use anything but governing equations that you can program in. They require no 'data' to build.

Disagree.