Board logo

Manifold design
Mark Allanson - 13/2/08 at 01:08 PM

I am thinking of making my own manifold, doing a 4:2:1 design. The general concensus on here is that the primaries must be the same length, but I found this BMW manifold on ebay and its design doesn't even try to keep the equal.

What happens if they are not the same length? Does someone come and tell you off? Rescued attachment a61b_1.JPG
Rescued attachment a61b_1.JPG

Mr Whippy - 13/2/08 at 01:18 PM

It’s the gas pulses flowing down the pipe, ideally one after the other. If the pipes are of different lengths then at some point in the rev range two or more pulses will arrive (where the pipes merge into one) at the same time which in theory would slow the flow down. I don't know how much it really matters, personally I’m not to keen on them as these tubular exhausts make a right racket unlike the cast manifolds.

You have off course to run the engine with no manifold on (after 11:30 pm) for pure fun. You can usually see right into the engine through the valve gap with all the flames inside, I love that. Next thing you'll find yourself doing is picking up a stick to poke in there yeah lets see how hot it is, oh it's the small things in life...

[Edited on 13/2/08 by Mr Whippy]

muzchap - 13/2/08 at 01:20 PM

I'm that the Impreza RB5 with the Prodrive Plus Pack exhaust had 2 primaries longer then the others which gave it the incredible growl and bark - it truly sounded like bits of the engine were being mangled and thrown out the exhaust - it scared the shite out of people standing around during the day!

I'd hazard a guess that equal primaries are a safe bet, but professional tuners will know how to adjust the primaries to give different engine performance/noise characteristics.

HTH

M

dhutch - 13/2/08 at 01:27 PM

I can understand the reasoning, as you say, all the cylinders fire at diffrent, equally spaced, times around the cycal. And that is how you want the exhast gasses to join, idealy. In order to keep the backpresure constant, and as low as possable. etc etc.
- That said, if your plumbing it up in 2inch ruddy drainpipe anyway, does it matter. Gasses are compressable anyway, and it will proberbly sort itsself out.
- I dont know. Does anyone have any proof, dyno printouts with two diffrent manifolds? What about the exhast note?

How do you mesure the lenght anyway? String lenght along the center line of each pipe?


Daniel

iank - 13/2/08 at 01:27 PM

quote:
Originally posted by Mark Allanson

What happens if they are not the same length? Does someone come and tell you off?


Absolutely, that and you'll restrict the power and torque you can produce.

There's a lot of variables involved but as I understand it:
The cylinders of the engine spit out a plug of hot gas into their header, this travels down until it hits the collector and wanders off down the exhaust. So you have a series of these plugs coming down the headers slightly timed apart.
If they collide in the collector (which they can if headers are different lengths) it causes back pressure at certain RPMs which is bad, but better since they create suction behind them if the headers are the same length at some RPM they actually suck the next plug faster in the other header which helps to empty the cylinder quicker giving the possibility of more power with a cam that can take advantage.
This is also why you tie 1-3 and 2-4 together as the pairs of a 4:2:1 system rather than 1-2 and 3-4 (or 1-4 and 2-3)

The above is horribly simplified no doubt, but reality probably involves lots of maths and computer simulation at which point just glue some pipes together and see what happens (which is the way a lot of the manifold companies no doubt work, those that don't just copy other peoples designs)

jambojeef - 13/2/08 at 02:01 PM

I wouldnt worry too much about the BMW manifold you have found.

If its at all a recent one it will be from the v8 engines and so you have the cylinder firing order split across the two banks making it difficult to adapt to your 4 cylinder.

Have you got a copy of 4-stroke performance tuning by Bell?

All good stuff and explained very clearly in there.

Geoff

r1_pete - 13/2/08 at 02:32 PM

quote:
Originally posted by iank

This is also why you tie 1-3 and 2-4 together as the pairs of a 4:2:1 system rather than 1-2 and 3-4 (or 1-4 and 2-3)


This does of course depend on the firing order.

twybrow - 13/2/08 at 02:41 PM

That book is a great read (although quite hardcored in places). Pick it up and answer all the questions you could ever dfream of....

DarrenW - 13/2/08 at 03:39 PM

Doesnt vizard quote some good lengths to aim for on a pinto manifold as well? I seem to recall my manifold, supplied by Mac#1, is based on the vizard sizes as far as is possible in a 7.

grub - 13/2/08 at 04:57 PM

well thats me f---ed then

Phil.J - 13/2/08 at 05:01 PM

Exhaust pulses travel down different diameter pipes at different speeds, and Ducati exploit this fact to get around having to package primaries of the same length when cylinders exhaust from different places on the engine.
I suspect (but don't know for sure) that primaries of the same volume will do the same job as primaries of the same length.

sebastiaan - 13/2/08 at 06:51 PM

Nope, the speed of sound in the exhaust gas is the determining factor. Same temperature in the headers = same speed of sound = same length needed. Diameter only influences gas velocity, not the velocity of the pulses (as these travel at the speed of sound...) Get the primarier big enough in diameter and there won't be any pulse tuning effect left... Read up on standing or stationary waves and you'll soon get it.

madteg - 13/2/08 at 07:40 PM

Fit a turbo then it dont matter

muzchap - 13/2/08 at 07:54 PM

quote:

Fit a turbo then it dont matter


LOL That's a VERY good point - although I was too chicken and opted for an S2000

DanP - 14/2/08 at 01:13 AM

There is a pretty useful calculator here: http://www.headerdesign.com

you will need to register to access the calculator, for my V8 the lengths I needed for a peak RPM of ~5500 came out at 1metre (which I was a bit suprised at), you can play with the config and see how the lengths change depending on the RPM you want peak HP at.

BTW have a look at the manifold on a BMW M3 if you wanna see if they worry about equal lengths.

Dan

[Edited on 14/2/08 by DanP]