HOT AIR: More About Six Cylinder Branch Manifolds

I would like to dwell for a bit on another 6 cylinder animal, the VW VR6

When these cars were first introduced into the S.A. market everyone was clambering for a set of branches to enhance power. The power delivery in stock form was smooth and torquey, but a little disappointing. An imported branch was available (at considerable cost) but power gains were negligible. My good buddy Joos Claasens had a VR6 Golf ‘lying around’ in his garage covered in white sheets waiting for the next Concours d’elegance competition…. He was assured his baby would be needed for not longer than two weeks to design a proper performance exhaust system…

Little did he know…

A good VR6 produces 76 KW off the wheels at reef altitude.

After two weeks of slogging on the dyno, a negligeable 1.5 kw was the best result obtained in power gains!

After four weeks 1.5 kw was still the best result.

After eight weeks 1.5 kw was still the best result!

By now, existing formulae were not even good enough for a baseline result. This was good theory gone awry!

Joos “borrowed” back his VR6 from time to time for Concours competitions, each time returning the car for further development, and each time trusting his mate, that the worst that could happen to his Concours winner was the collection of a thick layer of dust on the immaculatey polished duco. His car was the subject of much deliberation in my social circle for months.

After ten weeks it was decided that we should go back to basics:

A resonance is set up in a pipe when the firing frequency coincides with the natural frequency of that pipe. The natural frequency of a pipe is set by its length and diameter. The shorter the pipe the higher the frequency. When two or more headers are joined, the firing order becomes significant. So, all that needs to be done is to design the header pipes to resonate at a reasonable RPM so that the pressure waves reflected from the pipe-ends create a strong negative pressure wave at the exhaust valve at the correct time during the valve overlap period. The VR6 has a common cylinder head for both “banks” of cylinders, making the exhaust ports on the front (inlet) side of the cylinder head considerably longer than those on the rear or exhaust side of the cylinder head. This causes a mis-match in the interaction between the resonance of the inlet and exhaust tracts.

It was decided that header pipes of differing diameters would compensate. But first it was necessary to increase the energy of the reflected pressure wave. To do this, the diameter of the free-flow exhaust system which we had been using all along was reduced from 63,5mm to 57,15mm.

This alone was worth ad additional 1.5 kw through the entire RPM range: definitely a sign that we were heading in the right direction. Next the diameter of the secondaries were reduced from 50,8 to 44.45mm. This mod increased power by another 4 kw.

By the time development was concluded (which was approximately 4 months later!) with fine tuning, peak power was increased from 76,6 kw @ 5500 RPM to 88 kw @ 5000 RPM with amazing power gains through the whole rev range:

  1ST  RUN (KW)         LAST RUN (KW)
2000                23.0                     28.1
3000                37.6                     45.0
4000                57.0                     71.0
4500                67.9                     80.0
5000                73.7                     88.0
5500                76.6                     87.3
6000                74.2                     85.5
6500                66.2                     77.4

When Joos finally sold the car it had 5500km on the clock, most of which were done on the dyno!

That’s all for now!

Abel dos Santos