Exaust manifold tuning

[Martin K]

OK - this is supposed to be the beginning of something very important - whatever engine you are running.

If you are fitting any engine into a Stratos replica, then you need to consider the exhaust manifold/system. You might be offered one that someone else has made, but you may want to get one made for the job. So, why not get it made right? Or at the very least, understand how close to the ideal the system that you have purchased is?

I HOPE that by proposing the theory that I know, I will incite engineers who know better (through experience) to respond with some useful guidance (or criticism!) and we will all end up learning something. What do you think?.

[Sando]

Hi Martin I'd be interested your theories and everyone else's too.
There are a lot of good Engineering brains on this site and the more I can learn from them the better.
PS I think any exhaust would be better then the original Hawk short primary Beta manifold I've got! :eek:
Can't weigh up whether to tune the balls off the Vx or save my pennies for a 24valver to get the same if not more out as standard.....anyway thats for a different thread. I keep hanging my nose over a TIG welder anyway so either way I'll probably make up my own zorst.

Lets hear your thoughts.
Rob

[chris.richard]

If it's going to be a long essay, would it be better in the newsletter?

[strat6v]

Just to add my little bit.

When discussing the throttle body topic, the question of manifold primaries and there lengths were raised. Did i understand things right when it was mentioned that a certain length equates to max power at a given rpm? and if so what if a variable butterfly was added at the end of the merge collector like the ones fitted to a bike exhaust?. Would it give the effect of a variable length header?

If a large amount of mods were made to an engine, then aftermarket management would probably be needed. Perhaps one of these could run the exhaust valve from an auxillary output.

[Martin K]

Just to add my little bit.

When discussing the throttle body topic, the question of manifold primaries and there lengths were raised. Did i understand things right when it was mentioned that a certain length equates to max power at a given rpm? and if so what if a variable butterfly was added at the end of the merge collector like the ones fitted to a bike exhaust?. Would it give the effect of a variable length header?

If a large amount of mods were made to an engine, then aftermarket management would probably be needed. Perhaps one of these could run the exhaust valve from an auxillary output.

I think we will keep this simple! Most engines seem to be able to manage without a variable length exhaust!

[Martin K]

If it's going to be a long essay, would it be better in the newsletter?

I don't know enough about the site to be able to answer that. But most people come into the forum to compare notes, so to speak, don't they?

[guy mayers]

Can't weigh up whether to tune the balls off the Vx or save my pennies for a 24valver to get the same if not more out as standard.....
Lets hear your thoughts.
Rob


From someone who has already changed from the VX go the V6 route! It'll save you money in the long run! You should be able to buy a couple of engines for the price of tuning the VX to similar power levels. The V6, with it's electronics and injection will be more reliable. It will also be more fuel efficient and will ad more value to the car than the VX does.
Guy

[chris.richard]

I don't know enough about the site to be able to answer that. But most people come into the forum to compare notes, so to speak, don't they?

For a discussion, the forum's better. Once "the answer" is arrived at, an article in the newsletter is the best way of spreading the information to club members.

[Martin K]

It might be thought that tuning the exhaust system will only have an effect on highly tuned engines, running at high revs and with long duration cams having long valve overlap periods. This is not true. A tuned manifold will improve cylinder filling on any engine. What is also true is that a badly designed manifold will cause a tremendous LOSS of power on an engine running at high revs and with long duration cams and large valve overlap periods!

The main function of the exhaust manifold and system is to allow the escaping gases to leave the cylinder 'clean' so that the next charge of air/fuel mixture is as pure as possible - uncontaminated by any exhaust gas.
As with inlet tuning there are two things that come into play - inertia tuning and pressure wave, or 'pulse', tuning.

Inertia tuning first. Inertia is caused by the volume of gas leaving the cylinder having a mass and a speed. A correctly tuned manifold will use this kinetic energy to extract more exhaust gas from the cylinder and, during the overlap period, draw in some fresh charge too. The advantage of beginning to fill the cylinder with a fresh charge of air/fuel should not be overlooked as it has two functions. 1) to begin the induction process early by getting the 'static' column of air in the inlet tract moving towards the cylinder before the piston starts sucking it in and 2) the fresh air/fuel mixture helps clean the cylinder of it's remaining combustion by-products. To achieve this we need a negative pressure in the cylinder when the inlet valve opens, which is while the piston is still coming up to top dead center on the exhaust stroke. Seems to be impossible, but it isn't.

When the exhaust valve opens before bottom dead centre on the power stroke, the gasses are still expanding from combustion. This causes the gas to rush out of the cylinder towards the low pressure in the manifold.
The mass is pretty much a constant, so long as it is kept 'together' and not allowed to disperse. The speed varies with two prime things 1) cross sectional area of the exhaust piping and 2) temperature.

If the cross sectional area increases, the speed will slow. If the temperature decreases, the speed will slow. Any slowing of the gas speed will reduce the kinetic energy in the gas and this reduces scavenging. We can't increase the mass but we can take care of the speed. This is where pipe diameter, length and insulating the manifold comes into play - to keep the heat in and therefore the speed up.

As the mass of gas flows down the pipe it leaves behind it a low pressure area and it is this low pressure area that sucks or scavenges the remaining gasses in the cylinder and also begins to pull in the fresh charge when the inlet valve opens. However, as soon as the gas hits a larger cross sectional area of pipe the gas speed will slow and the resulting 'pull' will reduce. Changes in cross sectional area where the gas speed will slow are typically where the maniford joins the head (if the primary exhaust pipes are too big - disaster!) and where the primary joins two or more pipes of the same or larger diameter as in Y junction (typical in a 4 into 2 into 1 exhaust) or any number if pipes at a collector (3 into 1 for a V6 or 4 into 1 four cylinder or V8).

So the first considerations are that the pipe diameter should be correct and the length of the primary should be sufficient to maintain the gas speed for long enough to maintain the low pressure area behind it until the exhaust valve is closed. It might seem that if you keep the primaries as long as possible you can't go wrong, however the longer the pipe the greater the flow resistance and flow resistance will create it's own back pressure, thus counter act the advantage slowing down the gas flow and reducing, even eliminating, the scavenging effect. So primary length is critical for inertial tuning.

Pressure wave tuning is the other thing that can help extract the exhaust gas and increase the air/fuel mixture in the cylinder. Above we were talking about the actual speed of the gas flow through the primary pipe. Now we are talking about a different speed - the speed of sound. As the exhaust valve opens a shock wave is created which propogates from the valve seat and down the primary pipe. This wave travels at the speed of sound - far faster than the exhaust gas itself. When the wave reaches a any junction in the pipe it will be reflected back down the pipe but as a negative wave. As this wave carries with it a negative pressure, if we can get the primary length right it will arrive back at the exhaust valve around TDC and so it will help scavenge the remaining exhaust gas and pull in some fresh charge.

Both methods of primary pipe tuning - inertia and pressure wave - are based on the time it takes for something to happen. Unfortunately, the time available between exhaust valve opening, inlet valve opening and the exhaust valve finally closing again is variable with engine speed, whereas the speed of the gas is more or less constant and the speed of the pressure wave is more or less constant too, so we have a conflict which results in our tuning working best at only one engine speed. However, it isn't true to say that it only an advantage at one engine speed as we have the opportunity to create a low pressure area in the cylinder anywhere between the inlet valve opening and the end of the overlap period. Anywhere in this period will give us some advantage.

However, the advantage to be had by getting the inlet charge moving early in the induction stroke is greater at higher engine speed. As engine speed rises the actual time the piston is travelling down the cylinderl, drawing in the fresh charge, becomes less and less. However, the time taken for the charge to overcome it's inertia and 'get moving' remains the same, so the consequence is less efficient cylinder filling and a fall in torque as rpm rises.

For competition purposes, therefore, the best use of manifold tuning is to improve the cylinder filling at higher rpm - above where peak torque occurs as 'standard'. This gives the greatest increase in BHP and increases the useable power range of the engine.

For other engine uses, a different engine speed may be preferred. To keep the same characteristics as the 'standard' engine, aiming for maximum scavenging at the existing peak torque engine speed would seem a good idea. This should give a similar torque curve to the standard engine in terms of shape, but with an overall increase. If an improvement in torque below the standard peak torque rpm is required then the tuning can be based around that lower rpm.

It is in this area that the most debate will be had, I am sure. Once you have picked your 'target' engine speed, the rest is down to proven formula based on valve timing, engine speed and exhaust gas temperature. But it can be seen that different lengths will suit different folks!

I hope I haven't bored you to death - and I hope this post is in the right place.

I look forward to your comments before boring you more with the maths!

Martin

[Sando]

Well I'm still interested martin ! :)
Rob

[chris.richard]

Sums? I'm game!

[Martin K]

Before we do our sums we need to know about the speed of sound in the exhaust system and that means we need to know something about the exhaust gas temperature.

Although the exhaust gas enters the manifolds at 800 degrees Celsius or more, it cools as it travels through the port and down the exhaust manifold 1) by losing heat through conduction and 2) because it is expanding from a high pressure area into a low pressure area (Boyles law and all that stuff). Temperature in the exhaust will also rise with load of course, so it cannot be taken as a constant. Under hard acceleration with a full throttle things get hotter!

What this will mean ultimately is that whatever length we choose for the primaries at, say, 400 degrees mean temperature will be too short at 500 degrees mean temperature. This will have the effect of raising the RPM at which the tuning peak occurs. In one trial design tuned for 5000 rpm at 400 degrees the tuning peak went to 5350 rpm at 500 degrees.

Mean temperature between the exhaust port and the tailpipe will be in the order of 400 degrees. However, as we are only dealing with the first metre or so of the exhaust system, we will assume a higher mean temperature of 550 degrees Celsius rising to 700 degrees at full power under load. This gives us a speed of sound in the range of 570 to 620 metres per second.

If you have ever wondered why tuned inlet tracts are so much shorter than tuned exhausts - the above gives the explanation. The speed of sound in the exhaust can be nearly twice the speed it is in the inlet due to temperature, so exhaust tuned lengths are nearly twice as long.

We also need to know when the exhaust valve leaves its seat and when the inlet valve opens. As an example, the 24 valve Alfa V6 on standard cams these two values are 46 degrees BBDC and 14 degrees BTDC respectively. As we want the maximum negative pressure in the cylinder at the time the inlet valve opens this gives us a crank rotation of 46 degrees + 180 degrees - 14 degrees, or 212 degrees, by which time we want the pressure wave in the exhaust manifold to return to it's maximum negative value, which on a pressure wave occurs 270 degrees after it begins to propagate. This means we need the leading edge of the negative wave to arrive back at the exhaust port at 2/3rds of 212 degrees after the exhaust valve opens, which is 141 crank degrees after the exhaust valve opened. We also need to choose an engine speed at which this will occur, as we need to know the time in which this must happen. Once we have chosen the engine speed we have an equation that will give us the length. If we were tuning for the full wave length (which we are not):

2 x L(ength)=(Crank angle x speed of sound) / ((360/60) x engine RPM)

But we are tuning for only the positive half of the wave, which will be reflected back as a negative wave, so we can halve the length of the primary by dividing by two again, giving

4 x L = (Crank angle x speed of sound) / (6 x engine RPM)

So:

L= (Crank angle x speed of sound) / (6 x engine RPM x 4)

Or:

L=(Crank angle x speed of sound) / (engine RPM x 24)

Therefore, at 5000 rpm:

L (in metres) = (141 x 570)/(24 x 5000)

My calculator gives me .67 metres which is 26.4"

At a mean exhaust gas temperature of 700 degrees giving a speed of sound of 620 metres/second this would increase the primary length required to .73 metres or 28.7". Keeping the same previous length of .67 metres will shift the scavenging peak to about 5450 RPM at the higher gas temperature. This is not such a bad effect to create, given that when driving in a flat out style it is higher BHP at high revs that we want, and better cylinder filling above peak torque will deliver that.

The above length certainly suits my engine. As the standard torque curve is almost flat between 5000 and 6000 RPM, the exhaust is tuned to add power within this range whatever the exhaust gas temperature.

With different cam timing or different engine speeds you can see what the difference is if you play with the numbers. For example, a typical 300-degree race cam has valve timing of exhaust opening at 82 BBDC and inlet opening at 42 BTDC. As the exhaust valve does not close until 38 degrees after top dead centre it would be undesirable to have the pressure in the exhaust manifold going positive before then, so to prevent this we probably want to delay the reflected wave so that maximum negative pressure arrives at the exhaust port no more than, say, 30 degrees before TDC – 12 degrees after the inlet has begun to open. We do not want to delay it too long, however, as there will be a lag between the low pressure appearing in the cylinder and the fresh charge being drawn in (due to the inertia of the air in the inlet tract) so the advantage of opening the inlet valve early will be negated).

This means for our wave to reflect back as negative peak at 30 degrees BTDC we are looking at a crank timing of 2/3rds (82+180-30) - about 155 degrees. So, either a) for a peak at 5000 rpm the primaries will need to be longer (.741 metres at 550 degrees Celsius or .805 metres at 700 degrees Celsius) or b) the peak scavenging will occur at higher RPM. However, as the race cam will move the peak torque up the rev range anyway, this may suit us. As the power band will now extend to around 8500 RPM rather than 7000 rpm and we probably won't be using the revs below 5500 rpm for any serious driving, so we will need to improve scavenging at around 6000 rpm.

The good news is that the same primary length of 67 centimetres on the standard cams giving a peak at 5000 rpm will give peak torque with the race cam at 5960 rpm.

I am soon going to fit some fast road/rally 280 degree cams with timing of 31 – 69 –69 –31. With these cams the same primary length will effectively be tuned for 5600 rpm with 700 degree Celsius mean exhaust gas temperature due to the exhaust valve opening earlier before bottom dead centre.

If anyone is interested I can let you have an Excel spreadsheet that will do all the maths for you.

Bye

Martin

You will find some great charts that demonstrate the effect of this method of tuning, albeit on a four cylinder aero engine, here.

http://www.cafefoundation.org/aprs/EPG%20PART%20IV.pdf

And here:

http://www.aircraftspruce.com/catalog/eppages/powerflow.php

A manifold designed using these principles adds over 23 hp to an engine that produces 133 hp on standard manifolds. Quite incredible!

[Martin K]

Quite a few people seem to have read this thread, but no-one has added to it. it would be great to have some feedback!

Thanks

Martin

[CorseChris]

Feedback....

...this is fascinating stuff Martin. Ideal material for an in-depth newsletter article, no? :)

Seriously interesting thread, though I have nothing technical to add as it's gone way beyond my area of knowledge on this subject. I'll just watch and learn if I may!

I do have access to the brain of some similarly smart people as yourself though, so if any questions do arise, let me know and I'll see if I can enlist the interest of an ex-Cosworth Racing induction/exhaust tuning specialist...

[chris.richard]

Presumably wrapping the headers to keep the heat in raises the internal temperature and the speed of sound inside the headers, so changing the rpm of the maximum effect. Am I right? if so, you'd need to take that into account in the calculations, or is wrapping more about reducing external temperatures than raising internal ones?

[Martin K]

Yes, insulation does keep the heat in, keeping the gas speed itself up and speed of sound in the gas will be higher too. I did mention this above and the range of temperatures and therefore gas speeds givem allow for wrapping the headers. By my calculations the difference in RPM between wrapped and unwrapped will be a few hundred RPM at most - say 300-400 - and this effect is only seen during hard driving at higher RPM, so allowing heat to build up in the exhaust manifolding pipework. Materials can also have an effect. Stainless steel is a less efficient conductor than mild steel so a stainless manifold will run hotter exhaust gas temperatures than a mild steel one of the same design. For 100% accuracy the only true way to know for sure is to measure the EGT at two points - close to th head and close to the collector and then take an average.

The main reason for insulating is to keep the energy in the exhaust gas. The added benefit is that there is some reduction in heat in the engine bay.

[Martin K]

Just a word of warning! Don't use the same formula for a turbo manifold.

Turbo manifolds have an important role to play if you are to get the best performance from the turbo, and different tuning priorities come into play. Turbo engines run with very little valve overlap and also they have a positive pressure on the inlet side which will force the charge into the cylinder as soon as the inlet valve is opened - so long as the cylinder pressure is less than the boost pressure which we can safely say it will be.

So scavenging is not the priority on turbo systems, but exhaust flow and in particular pressure waves and their reflections are very important. The timing of these pulses need to be somewhat different for a turbo engine, so pipe lengths will be different also.

Martin

[chris.richard]

Just back from a medical conference in Manchester. One of the topics that came up has uncanny parallels to this.

The human arterial system makes use of similar pressure wave reflections. Inside us though, the pipe divides into more smaller tubes: 1-2-4 etc, rather than an exhaust's 4-1 or 4-2-1. There are pressure wave reflections off the dividing points in the same way. The pressure waves come from the aortic heart valve opening as the blood starts to flow from the ventricle into the aorta (the first biggest artery in the chest). Really a single cylinder engine I suppose. One difference is that the artery walls are elastic, and dampen and modify the reflections so that they raise the pressure at the heart valve just after it shuts, increasing the driving pressure of blood into the heart's own blood supply, (which flows mainly when the heart is relaxing - sort of on the "suck" or cylinder filling part of the cycle).
In disease, as the arteries harden and lose the elasticity, the reflected pressure wave arrives back at the heart quicker, impeding the emptying of the ventricle, and also losing the boost to heart blood flow in the filling cycle.

Physiology lesson over.

No, I couldn't interest them in exhaust manifold design either! :rolleyes:

[Martin K]

Fascinating, Chris.

(Now thinking about the potential for flexible exhaust manifolds.............)

Ah! Expansion chambers - they have a similar effect!

[strat6v]

Back to valves in the pipes too!

[Martin K]

True - maybe you can get a cardiologist to come up with a design!

[Sando]

Hi Martin
Great stuff. Any formulas to also include the diameter of the primaries? As gas flow increases with temperature, what is the relationship with size of tubing?
I know the velocity of fluids are effected by this and always thought that exhaust maniforlds too needed to be neither too narrow or two wide in relation to the length of the primaries. The 24v example is good as this is what I'll be aiming for.
Thanks
Rob

[Martin K]

The size of the primaries should be no greater than the curtain area of the exhaust valve - i.e. the circumference of the exaust port - measured at the smallest diameter on the valve seat - multiplied by the valve lift. For two exhaust vales in a 4 valve head, obviously the curtain area of both must be added together. This enables the exhaust gas to maintain the speed in the manifold primaries that it has as it leaves the cylinder. As the gas will be cooling and slowing down a slightly smaller pipe can be used. If you work out the diameter of the ideal pipe you want and available 'stock' tubing comes in one size that is too big but the next size down is too small, go for the smaller one. Where the manifold joins the head, you need to engineer a perfect match between the exhaust port and the manifold sizes so that the gas flow is maintained without any disturbance due to a sudden step change in cross sectional area.

Obviously cooling is increased by the surface area that the exhaust gas comes into contact with. Smaller diameter pipes have less surface area so more heat will be kept in the exhaust gas.

Because not all two, three, four primaries are flowing gas at the same time, the diameter of the second and subsequent pipes in the manifold, where two or more pipes join, don't have to as big as you might imagine. If the gasses suddenly meet a much larger cross sectional area, they will slow down - something we don't want. So the next pipe should certainly have an area no bigger than sum of the areas of the pipes being joined, and generally considerably smaller than that is better.

Between 0.5 and 0.85 of the sum of the areas of the incoming pipes is typical. So, for example, 3 1.625" primaries can be joined into 1 pipe between 2.0" and 2.5" diameter pipe and you will be 'in range'. The larger diameters may produce less back pressure, the smaller diameters will give better scavenging as a result of exhaust gasses from one cylinder helping produce a lower pressure at the exhaust valve of other cylinders joined to the same collector. Similarly, 3 1.75" pipes could be joined into a pipe between 2.125" and 2.75".

If you have checked out the paper written on aircraft engine exhaust tuning, they put 4 1.75" pipes into a 2.25" pipe. The area of the 2.25" pipe is only .41 of the area of the four pipes being joined, and yet there were no back pressure problems found with the 4 into 1 system. However, going this small conflicts with other papers I have read on the subject which all recommend an area greater than 0.5 of the total input area.

[chris.richard]

Smaller diameter pipes have less surface area so more heat will be kept in the exhaust gas.
BUT, smaller pipes have a larger surface area/volume ratio, so are better at losing heat - which is why radiators of all types have multiple small pipes, not one big one. :confused:


If you have checked out the paper written on aircraft engine exhaust tuning, they put 4 1.75" pipes into a 2.25" pipe. The area of the 2.25" pipe is only .41 of the area of the four pipes being joined, and yet there were no back pressure problems found with the 4 into 1 system.

Possibly something to do with aircraft having stub exhausts and no silencers?


Any idea if the tortuosity of the pipes has a significant bearing? Some systems have multiple tight curves which can't help.
Is flow laminar or turbulent in the headers?

[Martin K]

You are correct, but we are comparing one big pipe with one small pipe - not lots of small pipes. Micro bore central heating has lower heat losses through the pipework than 15 mm pipework. I agree that if we wanted to lose more heat we would have lots of small pipes. The smaller pipes have a lower surface area which creates less drag than a larger surface area.

The flow should be considered to be mainly laminar when it leaves the exhaust port, and that is the way we should aim to keep it. However, the length of the pipes is, in performance terms, more important than the tightness of bends as any back pressure caused by bends will have a marginal effect compared to the tuning. As bends tend to be unavoidable they should be kept to as large a radius as possible. Most important is that the tubes are not kinked in anyway when being formed into a bend, so as to affect cross sectional area. A change in shape, to so say rectangular for example, is acceptable so long as the change is progressive and the cross sectional area remains the same.

I think that in the aircraft paper I mentioned the engine speed is low (< 3000 rpm) so that the combined volume and mass of gas to be carried per second through the collector is small compared to an engine running at higher rpm. I have come across no mention of collector size compared to rpm in any engineering books or papers on the net, but it makes some sense. Any input on this topic from other engineers would be welcome, of course.

[chris.richard]

Tight curves can induce turbulence in previously laminar flow by breakaway and vortices in the inside of the bend.

[colin artus]

For primary diameters, from A G Bell 'Performance Tuning In Theory and Practice';

Primary Pipe Diameter in inches = ((scid x 16.38)/ (ppl + 3)x 25))) x 2.1
(scid is single cylinder displacement in cubic inches)

As for collectors I can find no reliable formulas although I did find a 'ballpark' formula that really only applies to drag race engines where the end of the collector is also the end of the exhaust. Bear in mind that it is the degree of change in sectional area that determines the amplitude of any reflected wave. On a drag engine where the collector terminates in an open pipe this is effectively infinite and the reflection will have nearly the same amplitude as the original wave. But on most cars the remainder of the exhaust and its silencer kills this through length and energy absorbtion leaving the primary/collector boundary as the generator of reflection. Thus the formula below for collector length is really redundant.


Collector diameter (inches) = 1.9 x ppd (1.6) might be more appropriate)
Collector length = .5 x ppl

Colin

[Martin K]

Tight curves can induce turbulence in previously laminar flow by breakaway and vortices in the inside of the bend.

Yes, and something to be avoided. If you need to have a 90 degree tight 'elbow' turn, where the inside radius is less than the diameter of the pipe, it is better for lamina flow to have two 45 degree 'elbow' turns (i.e. cut and welded pipes) than to have one tight bend. However, such an arrangement gives an increase in cross sectional area at the elbow, because cutting a pipe at 45 degrees produces an elipse of larger area than a perpendicular cut.

So, to keep a constant cross sectional area and to maintain good lamina flow, it is better to have longer bends with the radius of the inner bend ideally no less than twice the pipe diameter.

Going back to pipe diameters again, I have looked closely at the Abarth designed 4-2-1 exhaust for the 037. The primaries are 39 mm ID, the secondaries are 45 mm ID and the collector is 52 mm ID.

Analysing these numbers you can see that the secondaries are about 1.15 the diameter of the primaries and 1.33 times the area. The collector is similarly 1.15 times the diameter of the secondaries and 1.33 the area of the secondaries.

So the secondaries are 2/3rds (0.66666) the area of the two pipes going into them (a third for each pipe?) and the collector is 2/3rds (0.66666) the area of the two pipes going into it (again, a third for each pipe?).

The collector area is therefore 2/3rds of 2/3rds of the TOTAL primary area (0.444444) or 1.77777 times the area of 1 primary, which means the collector must be 1.33 times the diameter of the primaries - which we know it is!. Going back to the aircraft paper I think it now makes sense, although the collector area being only 0.41 of the total primary area still seems a little small.

Utilising Abarth brains and research, with a 3 into one system using the same 39 mm ID primaries we would expect the area of the collector to be .444444 times the total primary area, which happens to be 3/4 of a 4 into 1 collector area. (As there are 3 exhaust flows rather than 4 this makes absolute sense). This is 1.77777 x 0.75 which is 1.33333, or 4/3rds, the area of one primary. To get this collector area the diameter needs to be 1.15 the diameter of the primary, which is 45 mm ID - the same as the secondary in the 4-2-1 system. This makes my previous collector sizing a little over-large for a 3 into 1 system.

The Abarth 4:2:1 system had the following tube lengths, which by my calculations would return negative pressure waves at around TDC in the rpm ranges shown:

Primaries: 415 mm 7000 - 9000 rpm
Secondaries: 300 mm 3700 - 5700 rpm
Collector: 150 mm 2900 - 4900 rpm

The 4:2:1 arrangement was probably selected over a 4:1 system mainly to give a boost to torque in the low to mid range (on top of that provided by the supercharger) from 3000 rpm to 5600 rpm and to aid scavenging at high revs too. (La Lancia shows peak torque at 5000 rpm and peak bhp at 7000 - 8000 depending on the version).

[Arthur]

Well,

I feel quite humble.
Almost humble enough not to point with a snigger to 29 inches being a pretty fair estimate for a primary length (I confess here, this was pointed out by Mr Emerald himself in CCC days, as being the primary length of every aftermarket exhaust he'd seen that worked, as opposed to the theoretical sizes supposedly describing said exhausts). And I defy anyone to work to fractions with 2-inch tube, hand grinder, and MIG set.
Oh yes, if you aim to have a go, buy a lot of 2mm of 3mm sqare section soft iron rod in metre lengths. The idea is you twist 2 lengths together, then when you bend them in the pattern you want to subsequently make in tube, they don't spring about - they hold the shape really well. And that's as good as the help gets; after that you're on your own. And when you compromise, compromise on the long side - all this will do is drop the rev range for best scavenging a hundred revs or so, where going too short may raise it to a point you never actually use.......

Point one - big gains from exhaust pipe size and length for the aircraft engine. Yeah but no but yeah but they generally run at constant speed, where you can do stuff like this. You can also do silly things with static engines and loads of room - like 32-foot sytems with incredible scavenging - but you can't get 32 foot under a car (well, under a Viper, maybe).
Point 2 - try "Scientific design of Exhaust and Intake systems", Phillip H Smith and John C Morrison, ISBN number 0-8376-0309-9
This does some very specific stuff, and some general stuff - its the general stuff that is most illuminating. The primary aim of any performance exhaust (and intake) system is to NOT BUGGER UP ANYTHING ON THE WAY IN AND OUT. If you achieve that, you will have achieved more than most people, including a number who have spent an arm and a leg on an aftermarket system.
No point me banging on here; read the book; its complicated enough to put you to sleep, but simple enough to mark up the things that work and the things that emphatically don't.

Meanwhile, don't stop with this thread, I'm enjoying it!

Arthur.

[Martin K]

Well,

I feel quite humble.
Almost humble enough not to point with a snigger to 29 inches being a pretty fair estimate for a primary length (I confess here, this was pointed out by Mr Emerald himself in CCC days, as being the primary length of every aftermarket exhaust he'd seen that worked, as opposed to the theoretical sizes supposedly describing said exhausts). And I defy anyone to work to fractions with 2-inch tube, hand grinder, and MIG set.
Oh yes, if you aim to have a go, buy a lot of 2mm of 3mm sqare section soft iron rod in metre lengths. The idea is you twist 2 lengths together, then when you bend them in the pattern you want to subsequently make in tube, they don't spring about - they hold the shape really well. And that's as good as the help gets; after that you're on your own. And when you compromise, compromise on the long side - all this will do is drop the rev range for best scavenging a hundred revs or so, where going too short may raise it to a point you never actually use.......

Point one - big gains from exhaust pipe size and length for the aircraft engine. Yeah but no but yeah but they generally run at constant speed, where you can do stuff like this. You can also do silly things with static engines and loads of room - like 32-foot sytems with incredible scavenging - but you can't get 32 foot under a car (well, under a Viper, maybe).
Point 2 - try "Scientific design of Exhaust and Intake systems", Phillip H Smith and John C Morrison, ISBN number 0-8376-0309-9
This does some very specific stuff, and some general stuff - its the general stuff that is most illuminating. The primary aim of any performance exhaust (and intake) system is to NOT BUGGER UP ANYTHING ON THE WAY IN AND OUT. If you achieve that, you will have achieved more than most people, including a number who have spent an arm and a leg on an aftermarket system.
No point me banging on here; read the book; its complicated enough to put you to sleep, but simple enough to mark up the things that work and the things that emphatically don't.

Meanwhile, don't stop with this thread, I'm enjoying it!

Arthur.

No need to be humble Arthur - the 037 used primaries that were 415 mm or 16.3" long.....

To be honest, I have read many published articles by 'Mr Emerald' and I have read too much that was completely inaccurate to give the man much credability in my view. I remember one article when he wrote complete b***s about cam timing. Back to exhaust manifolds. Can all the engines he came across be 12 valve engines with the same cam timing? I doubt it. Did do any comparative dyno testing? I doubt it.

But to be honest, I hope all the people I meet in competition in the future will have listened to you an Mr Emerald and not tested the theories - then I can keep any unfair advantage I manage to gain to myself!

There are many old engine tuning 'theories' that have been carried over from the days when engines had only two valves per cylinder and didn't rev much, but sadly these don't all carry their validity forward into the modern age of multi valve engines. One of these supposed trueisms is that 'peak bhp appears at about twice the engine speed where peak torque is generated'. I have read this many, many times. In the days when peak power was delivered between 6500 and 7000 on most 'performance' 2 valve engines, tuning the exhaust for peak torque at three thousand and something rpm made sense. But most 4 valve engines, certainly those without variable valve timing, don't make peak torque as such low revs (the reasons why are for another thread I think!) so the days of the 29" pipe are fading into the distance, I fear.

Largely engine tuners who cut their teeth on 2 valve per cylinder engines believe that multi valve engines perform in much the same way. They don't, and the engine tuners who know this are the ones to talk to. But I digress.

As you point out, Arthur, 29 inches doesn't work for all engine speeds, does it? If you have an engine like the 24 valve Alfa V6 you might agree it would be pointless tuning the exhaust for 3500 rpm where peak torque is at 5000 rpm. The difference in primary length required is about 9". With such a difference in length it is easy to work with reasonable tolerances - I am competent enough to cut a pipe to within 2 mm of the required length - an accuracy of ± < 0.5%. Even if I can't, I know a man who can. Incidently, I am curious; What engine are you running that requires 2" primaries? 820 cc per cylinder - a 6500 cc V8 perhaps?

If you need a one-off exhaust and have to get one made, it is as easy and no more expensive to have one made to your own design which will work better than a badly designed one. One business made an exhaust sysetm for me where the primaries were about 8 - 10 inches long - and they told me they knew what they were doing! I haven't used them since, but they are still producing 'aftermarket systems' that people are buying!

But I guess to be on the safe side all the Formula One teams, like you Arthur, probably stick to a 29 inch system - so as not to bugger anything up that their engine builders have achieved.

Going back to the case of the Lycoming, 29" primaries would have been too short though, wouldn't they? Incidently, piston engined aircraft rarely have exhaust systems running the length of the fuselage, yet without a tuned manifold the exhaust system would have to be about 1.7 metres long.

It should also be remembered that the overall length of an exhaust system has an effect on the performance of the manifold; where the exhaust system is long (as in the case of front engined cars) a poorly designed manifold may not be so obvious if the exhaust system beyond it helps with extraction. In a short system, as in the Stratos, the manifolding is far more critical, as there is no long pipework to aid extraction and compensate for poor manifold performance.

Now can I find a thread where I can hide from Arthur..........................:)

p.s Thanks for the reference to the book. I must get a copy. Or is it worth it? Does the forward say "Don't bother reading this book, either a) go to the pub and listen to what your mates tell you to do or b) just get N pieces of 2" pipe, cut them to 29" long and weld them together"

[David May]

Phillip H Smith's book was written in 1954 (mine has a £2 10sh price tag!) but some things just don't change. Many aircraft engines were running 4-valves 20 years earlier than that.

[Martin K]

Phillip H Smith's book was written in 1954 (mine has a £2 10sh price tag!) but some things just don't change. Many aircraft engines were running 4-valves 20 years earlier than that.

So do I take it you are another 29" man, David? I haven't read the book as you know, but with respect to current relevence, maybe you would consider the following review by an independent third parties:

FROM THE PUBLISHER
This is one of the best-selling technical books in its field. For years, engineers, engine designers, high-performance tuners and racers have depended on this book to help maximize their engines' potential. Dr. John C. Morrison is one of the foremost authorities on the analysis of the induction and exhaust processes of high-speed engines. Together with Philip Smith, he gives a thorough explanation of the physics that govern the behavior of gases as they pass through an engine, and the theories and practical research methods used in designing more efficient induction manifolds and exhaust systems, for both competition and street use. An outstanding, authoritative book.

CUSTOMER REVIEWS:

Robert, an exhaust system engineer., December 20, 2004, 2 out of 5 stars
Technical, but general.

The information in this book is state of the art - for 1972. Also, it's primary focus is racing cars, so don't expect to learn much of use in today's street cars with four-valve engines and catalytic converters.

In desperate need of updating!, July 26, 2003
Reviewer: "dragonht" (IL United States)
Very outdated. I regret not having read the reviews here prior to purchasing this book. The last time this book was updated was 1972! There is nothing in the way of forced induction. Turbochargers are not mentioned and there is a single reference to superchargers on page 17! "Fuel Injector" is mentioned once. I am definitely sending this one back. Not the first time I have been disappointed with Bentley Publishing. FYI, 1972 was the year Pong came out. As I said, the book desperately needs to be updated.


Stuck in the '60's, November 23, 1999
Reviewer: Louis W. Ott (Portland, Oregon)
I read the book cover to cover. My impression is that this book contains exhaust and intake design information that was current in the 50's and '60s. Very little information for modern 4 or 5 valve per cylinder engines, and fuel injection tuned intake systems. No information on practical design of V8 performance exhaust system for street emmissions legal exhaust. Needs updated.

[David May]

Actually I'm a 29cm man these days (Q4 3-into-1 primaries!)
Although the book is dated (mine is actually a 1968 revision) the scientific data stands as true today as ever (and I would have thought ideal for the USA where they still have orgasms over push-rods) although there are few 'instant formulae' for backyard welders.
There is a nice picture of the Dino V6 plumbing but no clues as to how to get such long frontpipes in the back of a Stratos!
I don't remember ever having had a car where there was enough room for an ideal system!

[strat6v]

Had a look at the 'burns stainless' website. It has a bit of info and they also do a computer model of an ideal header for 'your' engine spec.How much that costs i don't know.

What happens to the header theories for a supercharged engine?

John.

[Martin K]

What happens to the header theories for a supercharged engine?

John.

As long as you are not talking about exhaust driven turbocharging (and I know you are not!), no difference at all. Of course you will have a positive pressure behind the inlet valve when it opens, but increasing the differential in pressure by having a negative pressure in the cylinder at inlet valve opening will provide a benefit to flow, so tuned lengths remain the same. You may want to tune for a different engine speed, however, as the supercharger is likely to shift peak torque lower down the rev range.

On the exhaust stroke up until inlet valve opening of course there is no difference from a normally aspirated engine except that there is more gas to get rid off due to more fuel and air being burnt. If bigger exhaust valves and/or higher lift exhaust cams are being used to handle the extra gas flow then the 'curtain area' would be greater and so correspondingly larger diameter primaries would probably be required.

Or MAYBE Colin's quoted formula for primary pipe diameter could be modified to allow for the fact that the cylinder is effectively larger in capacity as a result of the boost pressure???

Primary diameter = ((SCDI x dr x 16.38)/(ppl+3) x 25) x 2.1)

Where SCDI= Capacity of one cylinder in Cubic Inches
dr is the density ratio, NOT the boost pressure
and ppl is primary pipe length in inches.

With good intercooling and a 70% efficient (screw) compressor dr would be about 1.7 at 1 bar boost.

[strat6v]

Your confusing me!! (doesn't take much) Math was never my strong point.

I will have to have new manifolds if i go the supercharger route. I plan on fitting the compressor next to the block, on the front of the engine so the primaries will have to project forward and round back under. Can your formula give me a diameter and length for these?

If i can motivate myself on monday, i'll put the 24v back in and mock up for the modifications. I think i'll be ok for space to fit the manifold but It will be a squeeze for the alternator and compressor drive. Hopefully i can still use the standard belt tensioner.

Have you looked on the burns website? Those merge collectors are works of art!!

John.

[Martin K]

I've seen the Burns website now - they do make some fantastic looking gear.

[colin artus]

Those Burns merge collectors aren't as hard to make as they make out- there's a fairly simple trick to their construction which a friend of mine figured out so he could make his own. You still need to be a dab hand with the Tig gear though!

Colin

[Martin K]

Those Burns merge collectors aren't as hard to make as they make out- there's a fairly simple trick to their construction which a friend of mine figured out so he could make his own. You still need to be a dab hand with the Tig gear though!

Colin

We may be quizzing you on that, Colin!!

[strat6v]

Most of the aftermarket stainless exhaust people that i have contacted won't or can't fabricate them. Apart from Burns do any of you know who might?

[Arthur]

Martin,

I'm having fun here!
2 inches - a bit rash, I accept. My own are, I think, about 1.6 or so.
And I agree heartily with pretty much all you say - the point with the aircraft engine is that it's a constant speed engine, so you can play all sorts of games in very narrow windows, and that the aircraft will pretty much fly at max torque rpm for fuel efficiency, rather than at max BHP rpm, although before you b.....k me again, I guess those two speeds will be pretty close in practice. (by clever design, natch).

The book is pretty much dated now - not much to do with multi-valves and high engine speeds. But the difference between then and now is not the theory or the maths - it's the quantity of data that electronic sensors can give compared to the mechanical crap that used to be used. The point I keep alluding to here is that regrettably the maths is not the whole story - or rather, it is, but it is so much more complex than a simple pipe length calculated from organ-stop theory, or gas speed and wave frequency. (and all the pipe length data changes if you slap a shady silencer box on the end, and I'd say 90% of the aftermarket cans are trash...)
There's a whole lot more going on, and I accept that it is now much better documented, such that a much better guess is possible "off the shelf". But it's still a guess, is my point, and I for one don't have this info at my fingertips. The guys who do, tend to do it for a living, and are reluctant to throw away trade secrets.

I still maintain that if you have to do it yourself, there's less informed help out there than you may imagine (until you've done it, then you can't move for the buggers), and even now with all the gear available off-the-shelf you can still come an expensive cropper cos a lot of it doesn't do what it claims. How then do you, as a punter with the money and a lot of hope, sort the trash from the good stuff? Cos you can bet your life the salesman won't care........
So yes, if you have a man who knows, or if you know yourself, or if you have access to parts that really work, then all the stuff I post here will be worthless. But if you are going it alone, then all I'm really trying to say is "don't make it worse, and here are gudelines as to how to achieve that". And a reassurance that if you achieve that, you'll have done more than most people manage.
Typical here is, I suppose, the perennial 4-1 against 4-2-1 argument (substitute cylinder numbers as appropriate). Point here is that the 4-2-1 implies resonant tuning. This is a bugger to get right, and is the basis for all the complicated sums. 4-1 maintains "independence" of exhaust tubes, and while in selected applications it won't give the whack that the 4-2-1 will, it works pretty damn well across the board provided the primaries are long enough. It's only when the paranoic look at a claimed few BHP difference somewhere in the rev range for a different system that the arguments start about absolute numbers, valve timings, relative pipe lengths - all cobblers. Or at least, all cobblers unless you are a dedicated circuit racer where those last few BHP at the top end mean win or 3rd.

Let's put some numbers on this. My exhaust, put together by Mark of TubeTorque in Macclesfied, has unequal length primaries, 3-1 both banks, only meeting up at the silencer, after transisting unequal secondaries. The exhaust is covered at huge expense by insulating cloth from maniflold to final collector. Not for gas speed; to keep underbody temps reasonable. I have an inlet stub under a Pipercross foam element at the inlet. Otherwise it's bog standard 12V Cloverleaf.
It's been rollered at 209 BHP at around 5600 rpm, and 217 ft/lb at 4800 rpm, on standard unleaded. It pulls about 80% torque from about 1800 rpm to 6000. This is somewhat in excess of the standard claims for the 12V Cloverleaf. With all due reservations about rollers noted, it is still above expectation. Doesn't mean I'm a clever sod, it just means I got the sums wrong in the right direction, as per my own advice, and didn't bugger anything up on the way. It's a really tractable roadie, which is what I use it as, with occasional track days as I can afford.
The Strat was never intended as a viable circuit car (hats off to the "stretched" version, which is no mean motor, and to Mr Rutter, who defies that above observation on a regular basis!!!!), so there's not much point going the max BHP route. Better by far to go for "area under the torque curve" and get a motor that will pull tree stumps from any revs, at the expense of the big power numbers or "maximums" anywhere. And the sheer quantity of cash said search can consume leaves me weak and shaking at the thought!

Right,
enough for a Sunday morning!
All the best, Arthur.

[Martin K]

Thanks for that, Arthur.

The point about aircraft engines running at single engine speeds is noted. What they found about the 4-2-1 system led them to rule it out for their purposes - a single engine speed - but of course in automotive applications, particularly for road and rally use, the better spread of torque is worth the sacrifice in top end power because you get a much more useful engine. However, their is no real alternative to the 3 into 1 design for a V6.

The point about what happens after the manifold is noted too, BUT it has no effect on the PULSE tuning of the PRIMARIES.

Although tuning for a single frequency/engine speed is all you can do with fixed length equal primaries, the tuned frequency is only the centre frequency of the operation of the tuning. Either side of that engine speed the negative pulse will arrive earlier, at lower engine speed, or later at higher revs. As the overlap period is several degrees with even the mildest cams, negative pressure in the cylinder during that period will assist cylinder filling.

Also important to remember is that as long as the cylinder doesn't go positive before the exhaust valve closes then the manifold is doing a good job. On the 24 V QV/Cloverleaf manifold, from about 2000 rpm until 9000 rpm the first positive pulse from the exhaust will arrive in the cylinder after BDC and before TDC. That not only damages the cylinder filling it increases pumping loads. Yet the engine performs pretty well in terms of torque and power produced. You can see, therefore, that any increase in the length of the primaries will be of benefit.

However, as the 24 valve engine, without radical modifications, will not allow the engine to perform well below about 4500 rpm tuning the exhaust to try and improve significantly torque below that level will be wasted effort and compromise what extra torque could be produced in the rev range the engine has been designed to work at.

Tuning for where the EXISTING peak torque is will produce exactly the result you have achieved and quite rightly claim as the most desirable - viz to increase the area under the torque curve to a maximum. So tuning for 5600 rpm (peak torue on the standard 24 V QV engine will mean that the cylinder will see a negative pulse up until the exhaust valve closes from 4600 rpm upwards. Even at 3600 rpm the cylinder will not see a positive pulse at IVO. So a manifold designed in this way will improve the torque of the engine from 3600 rpm up to the rev limiter. Of course increasing the primary length will also improve performance at all engine speeds because of improved exhaust gas inertia.

I agree that with your 12 valve engines it is possible to get much better torqe at lower rpm than with the 24 V engine, but that is why there is a fundamentally different requirement in the exhaust system of the 24 V engine, which produces peak power at 6900 rpm in standard tune.

Indeed I would refer to your own words of guidance, which in summary say be careful that any exhaust doesn't subtract from the good work the engine designers did. To this end, tuning the exhaust for anything OTHER than their torque curve would go against your own argument. So I will stick with 5600 rpm as the correct engine speed to tune the primaries for and be happy with what I get in the way of power improvement as a result.

Another thing that is important to remember is that if you really do spend most of your driving life at part throttle you will have a negative pressure in the inlet port most of the time, so the best you can hope for from your new manifold is a slight improvement in fuel efficiency and you won't be anywhere near the peak torque that the engine can produce at any engine speed. It is only with a wide open throttle - 'WOT' - I love that tla! - that you will realise the benefit of your exhaust manifold tuning. And if you never get near the red line (or limiter) then there IS no point in tuning the engine to perform at it's maximum power output.

In a nutshell, Arthur, as long a you have a 12 valve engine, yours will always be bigger than mine.....................:)

[Arthur]

Martin,

At last we appear to converge on a combination of maths and common dog.
Just out of interest, I whipped out me other one yesterday (exhaust/inlet tuning book) from some US citizen doing big V8's on a Haynes title. (rubbish, by the way).
With reference to Morrison and Phipps, I note the respected US citizen claimed a couple of things. One, that the exhaust velocity in one tube will drag a vacuum (and so aid cylinder filling) in the joined tube. A-la Morrison, absolutely not, this does not happen, and demonstrably doesn't happen. The only help you get is from the reflected negative pulse (think wave here). Which is dead handy at popping backwards around the seemingly tightest bends, or closely angled joins, and is as capable of presenting an unwanted positive pulse as it is a useful negative one.

The other one made me think, particularly about rev ranges. He pulled up a Volumetric Efficiency curve for a well modded big banger motor. This now did something like 6500 rpm. And? Three peaks, at 2000, 4000, and 6000 rpm. Not max torque speeds, and not max power speeds, but peaks in volumetric efficiency.
If you hark back to older, slower motors, and a lot of diesel engines, you can then say that with a max % VE at say 3000 rpm, then its not too bad a guess to say your max bhp is going to be close to 2 x engine speed of max torque.
If you substitute VE for torque, and don't say bhp at 2 x speed, but "at harmonics of max VE revs", you actually come up with "top of the head" numbers that work pretty well.
In modern motors, since valve opening periods tend to be longer these days, to get the power up, and allow free revving, this will tend to push the first VE peak up the speed range, and you may never see the first harmonic, never mind the second, to be able to judge the power. And then, of course, the flow restrictions take over and further modify the whole thing.

Just one of those things - but because what you see in practice may be wildly different from a 1972, 2500 rpm test bed engine, don't for a moment think that the theory is wrong. Just more complex than first appeared.

All the best,
Arthur.

[Martin K]

Of course positive pressure waves appear and can be reflected back to the port. Previously I mentioned this and it is part of the tuning rational. It cannot be avoided at all engine speeds but it can be avoided at critical engine speeds with the correct design. The low pressure created behind the the flow of gas from one cylinder certainly DOES appear as a low pressure at any other cylinder connected to it - as does the high pressure at the front of the column of gas. Simply putting your hand over your tail pipe will reveal the puffs of high pressure, the low pressure between each puff you will have to accept, as it is not so easily verified with your hand!

It is for this reason that in a basic 4-2-1 exhaust system tuned for kinetic scavenging but not tuned for pressure wave scavenging, a pair of cylinders which are going to cause undesirable interference to each other (i.e. presenting a high pressure in it's companions' primary while the exhaust valve is open) are not connected together. Hence 1 and 4, 2 and 3 are connected together rather than 1 and 2, 3 and 4.

You don't say much about the setup of the engine with 3 VE peaks - but there could be a number of reasons for it. One peak will undoubtedly be the VE peak associated with the cam timing - this, along with the basic engine parameters such as bore, stroke and port/valve size, is what sets the fundamental peak VE without any other influences. Second will come the inlet tract tuning - which can be designed to have one of its' peaks (usually the first) coinciding with the 'natural' VE peak, or shifted to occur at lower or higher RPM at will (within physical size limitations). Finally you have the exhaust tuning. This can add one or more peaks depending on the layout. A correctly designed 4-1 pressure wave tuned system will only have one peak within the rev range you speak of. A 4-2-1 will have at least two peaks and maybe 3 within the rev range, depending on design.

I will not get into any debate in this thread about un-equal length primaries - I think that it is best left to the manufacturers of such systems to justify their provenence and if users of such systems are happy with them, so be it. However, maybe this V8 used unequal length primaries.

And, of course, it may have used a manifold tuned for 2000 rpm which would have harmonics at multiples of this engine speed. However, such a design would inevitably produce positive pressure at the exhaust valve during the overlap period at engine speeds between these multiples, i.e at 3000 and 5000 rpm - which is obvioulsy undesirable. Also, with primaries needing to be in the order of 1.5 metres or more to tune to 2000 rpm, I don't think such a system would be practical in many automotive engine installations - it maybe practical on a boat or a truck, but not in a small cars.

If we have converged I am happy that it is on what I predicted at the beginning of this thread. The maths and physics are pre-determined, but agreeing on the engine speed(s) at which to apply the maths and physics would be the principal area of debate.

[colin artus]

Most of the aftermarket stainless exhaust people that i have contacted won't or can't fabricate them. Apart from Burns do any of you know who might?

I asked my friend about quoting a price: £110 for a 3:1 and £125 for a 4:1 in stainless (304 or 316) with the proviso that primary diameter was a readily available tube size ie 1.5 inch. If there was enough interest for quantity he could reduce the price. A transition or flange would add a little extra.

Colin

[Martin K]

I asked my friend about quoting a price: £110 for a 3:1 and £125 for a 4:1 in stainless (304 or 316) with the proviso that primary diameter was a readily available tube size ie 1.5 inch. If there was enough interest for quantity he could reduce the price. A transition or flange would add a little extra.

Colin

I take it Colin that the price quoted is for the merge part of the collector only?

i.e. any takers would still have to have the primaries made, including head mating flanges, and also have manufacture the 'secondary' pipe beyond the collector itself?

Does his design include the pulse converter/venturi seen on the Burns units?

Thanks

Martin

[colin artus]

Yes the price is just for the merge part of the collector; I got him to price them on the same basis that Burns do, so as to make a comparison easier. The additon of a flred transition etc would extra.
My friend can make primaries, flanges and the rest but these need to be made on the car, which is a different kettle of fish.

[Martin K]

I found this advert and picture:

High performance racing exhaust 2.5 - 3.0 - 3.2 V6
Voor de 2.5 - 3.0 - 3.2 V6 24V fase III motoren voorzien van standaard vier katalysatoren levert Madeno Racing een high performance sportspruitstuk-kit, waarbij de eerste twee pre-kats komen te vervallen. De sportspruitstukken kunnen worden gemonteerd in combinatie met de originele middelste katalysatoren, ofwel optioneel met sport metaalkatalysator.
De maximale vermogenstoename bedraagt maar liefst + 15 pk @ 7000 rpm, terwijl bij 5800 rpm al ruim 12 pk meer vermogen en 16 Nm koppel beschikbaar is. Naast de ruime vermogens- en koppeltoename produceren de V6 krachtbronnen uitgerust met een sportspruitstuk-kit een uitermate geraffineerde racy Alfa-sound. Can anyone translate Dutch into English? I know that 1 Pk is just less than 1 HP so 15 PK = 14.79 HP. This is a useful improvement! 16Nm is 11.8 ft/lb - also a useful gain.

[pimms]

I will try to translate it for you, not sure if all the technical stuff is correctly translated.
For the 2.5, 3.0 and 3.2 V6 with standard four "cats" (don't know this one: Katalisator ?) Madeno Racing offers a high performance sportsmanifold-kit. The first 2 pre cats are left out, it can be mounted with the other 2 original cats or as an option with a sports metal cat. Power increase is +15 hp at 7000 rpm ( already 12 hp at 5800 ) and 16 Nm torque. They produce a nice racy Alfa sound.

Gr. Pim

[chris.richard]

For the 2.5 - 3.0 - 3.2 V6 24V phases III engines Madeno Racing provide four catalysts high a performance sportspruitstuk sportspruitstuk-kit with standard provide, where the first two pre-kats comes expire. The sportspruitstukken in combination with the original middle catalysts, or optionally with sport metal catalyst are able be assembled. The maximum capacity increase amounts to no less than + 15 pk @ 7000 rpm, whereas at 5800 rpm already wide 12 pk more capacity and 16 Nm cross-belt are available. Beside the wide capacities - and cross-belt increase produces the V6 strength sources equipped with a sportspruitstuk sportspruitstuk-kit extremely refined racy Alfa-sound.

:D :D :confused: :confused:

[strat6v]

Supersprint had an advert in autoitalia a short while back. It looks very much like the same kit.

[pimms]

nice to hear someone else knows some Duch Chris, just to complete your translation, the word "spruitstuk" means manifold in English.

[Martin K]

I think that Chris's translation may be literal but not in context. Cross-belt must be torque and 'capacity increase' must mean power increase, surely?

From what I have read on another site I believe that the Supersprint item is different from thie one above.

[rutthenut]

Chris had uses the 'Babelfish' translation web site, which is a useful site to remember if you want to get an idea of some foreign text.

http://babelfish.altavista.com/

For even more 'entertaining' results, try a two-way translation from, say, English to Dutch, then translate that Dutch back to English. Or maybe don't bother!