Exhaust manifold (vs. headers)

[Rex Chan]

I was going to ask fb this question, but decided FSAE.com was more appropriate:

Most FSAE teams running 4 cylinders use headers/extractors, either a 4-1 or 4-2-1. Is there a really obvious reason that I'm missing why I've almost never seen a team run a manifold (like on a production car)?

There seem to be a lot of benefits in cost/packaging/mass/simplicity/etc. The main drawback is power, but less power = less fuel used. And in FSAE-A, singles (that weigh more than a good Euro 4 cylinder team) have enough power to go very fast.

We plan to do the analysis to see what power loss can be expected, but I seem to remember when I asked for numbers in 2011, it was in the order of 5-10kW loss.

[NickFavazzo]

The power gains from running a nice exhaust system over a bad one are usually from being less restrictive and simply tuning for that. So running a "manifold" may reduce the power but not necessarily the fuel use too.

[Series One]

Efficiency aside, I would have thought difficulty of manufacture would be one reason why cast (I assume you mean cast) manifolds are not very popular. It just isn't worth the time and effort required for a one off car.

[Racer-X]

As far as performance is concerned I would never do it because it would be hard to tune and the runners would be unequal length.

Then there is the fact that the common FSAE bike motors don't come with them stock. So that would mean I would have to cast one. Rather than casting something I would rather make a "shorty" header if space was an issue.

A header could always be designed to perform like a manifold if you wanted...

[Charlie]

You are making a very flawed assumption that more power always equals more fuel used.

Lots of performance gains are from efficiency which will result in a better power-to-fuel used ratio.

I would suspect that a properly tuned exhaust will increase efficiency at a rate that could result in better BSFC, which means you can generate the same power for less fuel.

Making a very efficient package, then tuning for fuel economy, will always give a better result in fuel than restricting power in an inefficient manner.

[Jon Burford]

Rex, I can't see it's worth the effort. I would never deliberately make something worse, it will probably reduce power and not necessarily fuel consumption I should not think, Surely it's overall efficiency you are shooting for?

the manifold you have looks fine to me!!

[Warpspeed]

The vehicle manufacturers are mainly interested in cost, packaging, robustness and ease of assembly.
And a crappy heavy tortuous cast iron manifold is often the result to save a few pennies.

There are definite weight and flow advantages to tube headers, and power too.

Fuel economy on the exhaust side is mainly about exhaust pumping losses.
And as fuel economy is usually much more of an issue at part throttle anyway, you can get by with a horrible exhaust manifold reasonably well.

But if you want to go really fast flat out, that is a very different issue.

[Moreboost]

how is a cast manifold going to be lighter then stainless tube?

[RenM]

Besides Packaging there isn't a single reason why to go for a manifold.

First: For a Prototype its really easy to build your own exhaust headers from steel tubes and it wont cost or weigh a lot.

But of course the main reason is engine tuning:

A low exhaust back pressure is desirable, but its anyways really hard to build an exhaust header that creates a lot of back pressure on its own. (the muffler can however)

Most important are the runner lengths.
The difference between a good exhaust header and a bad one can be massive. Tuning your valve timings and intake and exhaust runner lengths can lead to a huge increase in volumetric efficiency. We saw a difference of 10% at one speed between a badly tuned exhaust runner and a good one.
And don't forget you are also influencing your trapping ratio with the pressure waves at the intake and exhaust valves during valve overlap. A high pressure ratio (high intake pres, low exhaust pressure) can lead to very high overflow losses, where fresh gas is lost to the exhaust including the fuel. This leads to a very bad bsfc and increased fuel consumption!

With a manifold you wont have equal runner lengths for each cylinder, so your cylinder filling will vary a lot, leading to different air/fuel ratios and optimal ignition advance. You can compensate for that if you are adjusting the values for each cylinder at each speed (an overall factor wont be sufficient as the effect changes with the engine speed), but thats going to take a lot of time....

In the end: More power and a better bsfc will always be beneficial for your competition points. If you want to reduce power to reduce fuel consumption go for lean a/f ratio at WOT but dont reduce power by making your gas exchange worse in every respect.

[Rex Chan]

Thanks for all the replies!

Sorry for the confusion - when I was talking about a manifold, I meant the geometry, not making it by casting. I was thinking of 4 very short runners into a log manifold (sort of a reverse of our intake), then one exhaust tube/pipe coming out.

FSAE-A is all about economy (not efficiency, yet), so a manifold is a way to restrict power.

Charlie: I not sure I understand what you're saying about efficiency - doesn't all energy for power come from fuel burned? To be efficient, we'd run at Lambda > 1.00, to ensure complete combustion. We're already running LA=1.00 at WOT for enduro, and hopefully the guys will try to find the lean limit this year. After that, isn't the only to cut fuel used to reduce amount of air entering the cylinder by making the airflow path less efficient?

edit: based on all the above replies, it makes more sense to try to push SAE-A to move to an FSG style efficiency points score for the Fuel event. I quite like the sound and apparent power (only 49kW SAE corrected on the SS engine dyno) of past MUR/Melbourne engines, so would be a shame to give it up (though Monash went to a single in some parts due to this).

[Warpspeed]

Rex, if economy is the absolute priority, headers still win out through lower total back pressure, although in practice the difference may be more theoretical than practical over what you are proposing.

Biggest gains will come from a much higher compression ratio and some really creative valve timing.

I suggest you research "Atkinson Cycle" and "Miller cycle" engines.
These usually always run a supercharger for best results, but you can do something very similar with a normally aspirated engine.

The trick is to close the inlet valve very late, when the piston is a long way up the bore.
This reduces the dynamic compression ratio because compression only begins after the inlet valve has completely closed.

This enables you to run a crazy high compression ratio with a very small combustion chamber without exceeding the temperatures and pressures that could reach the detonation threshold.

Once it fires, you can get a very high expansion ratio, especially if you keep the exhaust valve opening very late.
This won't get you any more power, but it can sure will help the BSFC and thermal efficiency if that is your main goal.

A much larger exhaust valve can also help reduce exhaust pumping losses which also helps.

[Boffin]

Rex

Be very careful restricting flow through your exhaust. I will guess that this will make you add fuel.

I say this from experience though and our exhaust efforts 09 to 10.
09 ran the 08 exhaust, which was a 1" primary (port is 1 3/8". It ran a tapered nozzle about 2" long to change size). It also coupled incorrect pairs (1&3, 2&4 i believe) in a 4-2-1 exhaust.

In 2010 we stepped up to a 1.5" primary (1.5" straight to the 1 3/8" port. There is a reason for doing this) and couple correct pairs together (1&4, 2&3), still keeping the 4-2-1. The intake and engine internals were unchanged from 09.

By doing this we reduced fuelling by an average of 27% (I have the data to prove this), while increasing torque and thus power everywhere.
Torque was 8Nm up to 75nm @ 7300rpm shifted 300rpm down (09 peak was 7600rpm).
Power increase 12-14kw, up to 68kw @ 11300rpm, shifted up 2900 rpm (09 peak was 8400rpm).

If you want to reduce fuel usage with a 4 cylinder car, choking flow to do it not a good way.
A very smart way to do it is to look towards Sophia and the way they use their engine and a lesser degree to Swinburne’s petrol car and look not at the engine but at the drivetrain.

[RenM]

If you are already running lean and still want to cut power reduce the rev range further. It will reduce mechanical and throttle losses.

If you have no problem with knocking increase your water temperature to reduce in cylinder heat transfer losses.

However at some point you will have to question your car concept...

[RenM]

Originally posted by Warpspeed:
This enables you to run a crazy high compression ratio with a very small combustion chamber without exceeding the temperatures and pressures that could reach the detonation threshold.



I reckon they are still running the 600 cbr, so extremly high compression ratios (>15) will geometrically not be possible due to the very small stroke. Everything beneath can possibly be run with e85 without knocking.

[Rex Chan]

RenM - I agree, a fast 4 cylinder is prob never going to do well at fuel economy (though UWA did it in 2011, by running in low RPM/low power range).

yes, we are running CBR600RR 03-06, with E85, so have had no knock issues (at stock 12:1 CR).

[Mbirt]

I'm with RenM. Instead of reducing volumetric efficiency, why not just shift it to lower revs to downspeed the engine? You could go beyond changing duct lengths and try concepts like closing the intake ports down with epoxy for increased charge inertia where you need it at low speed. I saw tiny diameter intake runners on at least Toledo and Stuttgart this year.

I have seen a healthy reduction in POT BSFC using atkinson cycle operation on another engine I'm working with, but the reduction in low-mid range torque will not be welcomed in an autocross car.

[Warpspeed]

Originally posted by Mbirt:

I have seen a healthy reduction in POT BSFC using atkinson cycle operation on another engine I'm working with, but the reduction in low-mid range torque will not be welcomed in an autocross car.

Atkinson cycle is a fairly radical approach to improve thermal efficiency at the expense of power output. No argument about that.

But as it involves valve timing and compression ratio changes, it does point the way towards a possible compromise situation.

As suggested earlier, a smaller bore longer stroke engine might be a better candidate for this type of approach.

[Mbirt]

Please, someone. Do it: http://www.ski-doo.com/technol...ogies/4-strokes.aspx (http://www.ski-doo.com/technologies/engine-technologies/4-strokes.aspx)

[Jon Burford]

Originally posted by Mbirt:
Please, someone. Do it: http://www.ski-doo.com/technol...ogies/4-strokes.aspx (http://www.ski-doo.com/technologies/engine-technologies/4-strokes.aspx)

Yes please.