When you think about it, with a crankshaft that has the pistons moving in pairs, and camshaft and timing that ensures that one of the pair fires each time they reach round about TDC, there is a possiblity of using the pressure waves in the exhaust pipes, if THEY are paired. Each pressure wave, formed as the exhaust valve opens, is followed down the exhaust pipe by a wave of negative pressure. This double wave will affect the flow in pipes that join the first. If the primary pipe length is right, then the negative part of the wave will reach the pipe from the other cylinder of the pair, just as that exhaust valve opens. Thus the flow of exhaust gases is much easier from the cylinder, less remains, and more new charge gas can get in for the next stroke. For this 'Scavenging' or 'Extractor' effect, the cylinders should be linked 6-1, 2-4, 3-5. The three secondaries can then be linked after a similar, but less critical length. to the tail pipe. The effect works best in a certain rev. range, when the valve opening coincides best with the arrival of the negative wave.

A lot of work has been done on exhaust design, see "Performance Tuning in theory and Practice" by A.Graham Bell. For instance:
Primary pipe length = ((850 x (180+degrees exhaust valve opens before TDC))/RPM) - 3 (in mms) This gives about 15". See Bell's book for other calculations from theory of Pipe ID, and tailpipe dimensions. From practice, Gareth Thomas recommended 17" for the Primaries, so T&P don't disagree.

The extractor effect cannot occur if the cylinders are linked in threes!
The problem with 6-3-1 is that it is more difficult to fit the three secondary pipes between the sump and the chassis rail than just two secondaries. Most after market tubular exhaust manifolds are 6-2 for this reason, but 6-3-1's are available (TriumphTune/Moss), aren't any more expensive and are worth the hassle.

Coop, you takes your choice!

Colin
1972 British racing green TR6

6-2-1 and 6-3-1 manifolds(headers) are available from Cambridge motorsport in the UK.(www.cambridgemotorsport.com) Also checkout Phoenix performance exhausts at http://www.quiller.zx3.net/Quiller/Parts/exhausts%20stainless.htm
These are not cheap but excellent products, not sure if they will fit GT6.

Cheers

Graeme

YES I designed that back in 1982.
(though a fat lot of people ever give me credit for it!!) Ever since I proved it worked, instead of that dreadful thing sold at the time by SAH, there has been a rush of people ready to take credit for it, from exhaust makers who don't know anything about anything, to unscrupulous traders trying to live off other people's brains!
I get them every day!

It all smacks of the usual money over brains argument.
I spent the money and time working around the many Triumph engined problems....eg. the truly dreadful exhaust port breathing and the inlet temperature optimisations needed, and got a fat lot of thanks for that too! But there again, I didn't sell the head work I did, and power is in the head!

However,
I'm not at all happy about the "abortions" sold in the name of my nice design.
This system works by primary resonance, but in an interference (ie wave front addition) mode.

The primary resonance is ideally placed when the primary is about 19" long and about 42mm pipe diameter but a 38mm pipe will raise the frequency by about 10% so a change of about 250rpm.

However because the copies sold of this manifold are so dreadful and the primary pipes so unequal, the majority of these conversions end up with really significant flat spots all up the range....s you have to compensate by making a real trick engine usually.

(resonance becomes a factor at 2500rpm, a large increase in breathing efficiency takes place at this frequency, but if the pipes are made like that, this effect will be spread out all over the place between 2200-2800rpm which is MOST undesirable! It can be shown that as the pipe enters 1st harmonic (around 5000rpm) this will usually be the place where the system is at peak efficiency so you get peak power (in practice with a TR6 about 5500rpm)

Of course people will get out their pocket calculator and say this is all wrong, but you have to take into account the extra length added by the exhaust tract in the head too, as well as the fact it's a non crossflow design and each individual car has huge cooling problems....below....

Pipe resonance is enormously affected by temperature, so the ideal manifold will be so configured as to be cooled by the speed coefficient (cooling air from driving) as to "come on song" as it gets hot. The resonance of an exhaust system rises VERY substantially as temperature rises.....this is critical to engine performance.
It's called the "virtuous curve".

Lastly but not least this design of manifold is NOT AT ALL like anything done before for these cars, the majority of realisations I have seen have dreadful secondary lengths, the manifold can be said to run into back pressure between 6800-7500rpm, WELL BEYOND the rev range of most Triumph road engines, and inaccessible for a TR6...

So that puts paid to the argument about "it's only for cars with wild cams or race engines", this is simply a total failure to understand what is going on..

My manifold is OK in it's original design for those engines, but some optimisations were made on the racing engine (which was after all doing 100bhp/L@7900rpm), so much as to junk it altogether for a full race engine and develop a much better model.

This model was never shown or put in print for the reasons above....I didn't want other people taking credit for my hard work, by copying it....heck they even illegally copied my Tuning Book!!!!!

G Thomas

Best wishes,

Jeff
Temecula, California
'68 Spitfire Mk3 aka "Mrs. Jones" FD 21032 L

I eventually ended up with the Pheonix one which, while nicely enough made suffers from most of the deficiencies mentioned. I was particularly disappointed in the variations (and shortness) in the primary lengths. Tis also pretty pricey. I did speak to the Pheonix guys about this but they reckon it's impossible to make something that is both faithful to the original design and fits all the cars the marketing men want it to. Well maybe, but it didn't fit my car with carbs! I guess I should have been more energetic and made my own.

Still having said that it does seem to work to some extent as the car goes pretty well now, although as I have never run it with carbs, I don't know how much is due to the FI and how much the manifold. In combination with a Canley GT silencer it certainly makes a good sound.

For those curious to know what it looks like I attach a pic. The messy bit at the top is my homemade FI manifold.

Cheers

Nick

Nick Jones
Somerset UK

www.tengaston.plus.com

On our trick XJS exhaust (which actually uses the original 'orrible cast iron manifolds) we get about a 15bhp gain to 235bhp, on an otherwise bog stn theoretical 220bhp motor, from adding a strategically placed silencer, and removing the original restrictive centre ones.
This gives MUCH lower exhaust noise, as well as huge boost in torque from 1800rpm. You can read about it here,-
http://www.jagclub.ru/AJ6.html

As for Triumph trick exhausts, put it like this,-
The whole theory of secondary length and convergence on a 6/8/12 cylinder is a very tricky one.
The hotter the cam the more critical it gets.
On Lambos this reaches burglar alarm set-off levels (great laughs!).

Because of small valve areas the Triumphs HAVE to use hot cams, or they don't make any torque.
On a 6-3-1 exhaust the manifold works by complimenting this hot cam need, and it's best presented with a cam with a lot of overlap, so you get a BIG increase in torque, just like on small valve 4 cylinder engines.

(By contrast, On any F1 V8 today it's sheer physics of 4 cylinders, huge valve area, and supersonic gas, so there we are entering the area of shock extraction+advanced materials/rocket propulsion resonance. No comparison.)

However on road motors;-
What we are trying to do after all, is convert velocity to pressure and as you well know this is related to temperature ,and surface area.

Clearly if you constrict the pipes down into one small outlet you are doing the exact reverse of this cardinal rule.
Even on the Murcialago, Audi in their "manifold" wisdom screw up on that, never mind those guys from Maranello (!!!) and who's the great brains behind the latest Jaguar V8 creations....??
I sometimes wonder.....(!!!)just where do R+D depts spend their money?

If you do things the right way, you get a quiet exhaust and even more torque across the entire power band.
This translates into higher real engine compression ratio, better combustion, and hey presto lower fuel consumption.

Do it the wrong way, and you run into back pressure early, have a noisy exhaust, and get a series of unidentifiable but detectable torque flats all the way up.

GT

Yep, the system uses Megasquirt ECU. Actually pretty easy to map as you have feedback from O2 sensor to help, if you use wideband (I don't so far), even better.

I have been slowed a bit by mapping on my own using data logging and some clever software written by another user (the joys of open format!). Couple of hours on a rolling road would have sorted it.

Not sure whether your comment about running all in to one is directed at the inlet or exhaust manifold. I commented more about the inlet 'design' in my post on the PI plenum thread.

Horses for courses really, this being a bog standard, much thrashed Vitesse lump other than the external plumbing. Incidently, inspiration for this exercise came partly from the results mentioned in your tuning manual for a standard Vitesse engine with your 6-3-1 manifold and PI. I didn't expect to match your quoted 140 bhp (and I'm sure I haven't) but I figured I could make some gains and I reckon I have. I'll put on a rolling road later this year and we'll see...... Can map the top end properly too while we're at it, not really possible on the road!

I'm now a true believer in engine management (Megasquirt ECU does sparks too) especially Megasquirt which is awesome value.

Nick

Nick Jones
Somerset UK

www.tengaston.plus.com

I'm getting one for evaluation next month for a compressor engined project.
I've got 3 different systems to evaluate, 2 from UK, 1 from OZ.

My comments were ambiguous, in fact they were directed at the exhaust system not the inlet

GT

2 1/4" inch pipe, as above, along transmission tunnel. Three pipes would be a very tight fit, and a silencer, even a 'cherry bomb' would be even more so, though the original setup was for a small expansion chamber in there. Y-splitter under diff, to two large straight through silencers under boot floor. Car gives 98dB at 4.5K, from its 2.5 straight six, if that's any measure of the system.

Thanks for your informative input!

JOhn

Being as the exhaust port is one of the worst features of those Triumph engines, and increase in efficiency is bound to be very noticeable, and that's why the secondary length is very important.

The 3 secondaries HAVE to be extended back as far as roughly the end of the gearbox area. At this point they should join up 3-1 into a pipe of 55-65mm size (preferably a large sports type silencer, if not possible then splitting it all 2 ways under the floor is quite acceptable.

There should then be a section of pipes (can be 2, and then injecting into twin silencers at the rear of the car, with a minimum pipe diameter of 2"

This needs a drawing, and as I don't have a car here, or accurate dimensions, this is not the best place to discuss it.

GT

I'm not familiar with your work, or products. I have a TR4 that I'm building right now for vintage racing and would really like to learn more about header design for it.

Do you sell or design headers for a fee? If so, and I could take care of the fabrication.

I'm sure the thinking for a 4cyl. vs. a 6 is most likely a little different. Do you have any thoughts on this?

Thanks!