Hi everyone,

I'm looking for advice on the best way to upgrade a Polo Mk2 Coupé GT 1.3 8V VW engine for high power and torque in the 4,000–7,000 rpm range. I currently have a Volkswagen MH engine from a Mk2 Jetta with a GK intake and carb and a GK exhaust from a Polo 86C GT. I also have a GK block lying around, but its cylinder head is mechanical, while the MH head is hydraulic.

Which setup would be better for building a race engine? I’m also planning to run ITBs with a standalone ECU or bike carbs. My goal is around 110–130 hp, but torque is equally important.

Are you building your own exhaust? You can't get 4-1 manifolds AFAIK. The other thing that has occured to me is to move the oil filter to a remote location so that the headers can be symmetrical(ish).

I run these valves, they make a big difference, i waist and reprofile them on the lathe at work.

I’m going for an aftermarket exhaust like this one, mainly because they’re not too expensive. I’ll also go with a 4-2-1 manifold for more torque instead of just peak power. https://www.gravityperformance.co.uk/exhaust-manifold-volkswagen-polo-86c-6n-1-0-1-1-1-3-g40-75-99

The 4-2-1 will help but the exhaust ports and valves are the outlet restriction, though adequate for the inlet valve size.

And I think you need more revs than 7000rpm to get over 103hp from 1272cc 2 valve per cylinder.

I may be biased (because I have a GK) but the GK mechy head will rev higher. No need to convert from hydraulic to solid lifters for one thing (though new mechy rocker arms are very hard to find).

Even with its standard 22/26 Pierburg 2e3 carb the High Compression (11:1) and Squish pistons (effectively only 1263cc) the GK engine gives 75hp at 6800rpm at which point the Bosch Dignition's rev limiter cuts the sparks. It's factory torque is 77lbft at 3600rpm.

You can move that torque up the rev range with bigger inlet valves, and ITBs. Engine management will get over the rev limiter issue. With twin Dellorto DHLAs I saw sustained 75 to 77 lbft from 3400 to 4800rpm, which used to be produced in the 2400 to 3900rpm band with a smaller single Weber DMTR carb. All with the same GK camshaft, but 36.2mm inlet valves.

Then pick a new camshaft.

If really racing something over 300 duration

Approx

300 for rally

320 for hillclimb

330+ for track.

Or send the mechy head to KWL Automotive

https://www.kwl-motorsport.de/

(Polo Straßen Tuning tab)

to get an exchange cylinder head with ports enlarged, smoothed, polished, 38mm inlet / 32mm exhaust valves, valve seats maximally enlarged and ground, head planed. For €1350+VAT+any import duty. Perhaps a machine shop could do that work here in UK, or Belgium?, for less with the right parts.

Or for 7000rpm work out how to add 10psi boost to 75hp engine for ~126hp.

Edited 9 September, 20251 yr by dvderlm

On 08/09/2025 at 07:21, yarne_l said:

I’m going for an aftermarket exhaust like this one, mainly because they’re not too expensive. I’ll also go with a 4-2-1 manifold for more torque instead of just peak power. https://www.gravityperformance.co.uk/exhaust-manifold-volkswagen-polo-86c-6n-1-0-1-1-1-3-g40-75-99

You can't argue with the price, but i've always thought there was room for improvement on that particular design. The headers aren't matched very well on 1 and 4, and the 2-1 collector is a bit nasty looking. I'd like to know what company came up with that design and whether it was based on anything mathematical. I used an exhaust manifold calculator to come up with a design for me a while ago; i had to enter cam timing data etc. which made it a bit tricky, i think i used a Schrick. The results were 4-1, one collector, 1-3/8" primaries at a length of 470-520mm. The primaries length can vary according to where you want the power to peak, i went for the area around 5500-6000, so road revs in other words. I'll see if i can dig out the results.

Yes, I noticed that too – it looks like cylinder 1 (the far left runner) actually has a noticeably shorter primary compared to cylinder 2 (the far right one). That mismatch is probably to get around the oil filter, so yeah the filter would need to be relocated if you wanted equal-length primaries. What’s the best way to do that – sandwich plate with a relocation kit?

I think the alternator or G-lader placement also influences number 1 pipe being bent. The block is not central in engine bay as transverse with gearbox so the pipes have to be angled to match the central exhaust tunnel and miss the gear lever rod.

But yes, Relocation plate and a filter header. It's 3/4:UNF spin on filter.

But the oil tubes to remote filter could also foul.

Using the GK cam 47° BBDC exhaust valve opening

with 111 centreline, a guessed advertised duration of 270 at 0.1mm, 29mm exhaust valve diameter, 6000rpm peak power...

http://www.wallaceracing.com/header_length.php

says

Your Primary Tube Length is 34.77 inches : 883.16mm

Your Primary Tube Diameter is 1.14 inches : 28.86mm OD

Your Collector Diameter is 1.95 inches : 49.62mm OD : One Collector

I have the Philip A Smith's Scientific Design of Exhaust and Intake Systems

book, so I'll try to find the formulae.

A Schrick 332° race cam has 86° BBDC closing if you want to experiment.

The Wallace one is the calculator i used. An austin mini fanatic told me that the header diameter i had got was too small, and that it needed to be 1"-3/8. The mini world seemed like quite a reliable source of info to me, different designs will have been tested to death on and off the road. My most valuable discovery was that the one collector specified could potentially be back behind the sump, as sometimes the calculator gives you a very long header length, so long that the collector could potentially be tucked away out of harm's way around the space currently occupied by the CAT. The headers would need to be turned off the ports as soon as possible so as not to waste length. It was just a theory at the time!

Ah, but Mini A-series engine has Siamese ports so need a larger pipe for 2&3 to share in a long centre branch 3 to 1 design. 1&4 made same diameter so gas travels at same speed. Limit may be what tubes one can get mandrel bent.

Page 142 of the Smith book has a rough formula using a constant based on peak exhaust pressure as valve opens and speed of sound.

A x S x D^2

__________ = l

1400 x d^2

with inputs in inches and answer in feet length of primary header to a collector from exhaust valve.

A exhaust duration

S stroke

D bore diameter

d exhaust valve diameter

On pipe diameter Smith suggests about 1/8" larger than exhaust port throat dimensions.

I think this is a length for third harmonic at up to 6000rpm.

This is the kind of build I want (https://www.youtube.com/watch?v=FkhNcL4gZ6g).

So 77mm overbore and 1w crank to get 79mm stroke for 1471cc ?

On twin Weber 45 DCOEs to big valve head.

Yes that will hit 130ps at 6500rpm with a 290+ cam.

Something like 34 chokes, with 5 aux venturi and 145/F16/180 main jets.

A bit bigger than Fiat Ritmo/Strada 105TC.

On 12/09/2025 at 08:37, dvderlm said:

Ah, but Mini A-series engine has Siamese ports so need a larger pipe for 2&3 to share in a long centre branch 3 to 1 design. 1&4 made same diameter so gas travels at same speed. Limit may be what tubes one can get mandrel bent.

Page 142 of the Smith book has a rough formula using a constant based on peak exhaust pressure as valve opens and speed of sound.

A x S x D^2

__________ = l

1400 x d^2

with inputs in inches and answer in feet length of primary header to a collector from exhaust valve.

A exhaust duration

S stroke

D bore diameter

d exhaust valve diameter

On pipe diameter Smith suggests about 1/8" larger than exhaust port throat dimensions.

I think this is a length for third harmonic at up to 6000rpm.

What book is that?

What value do you get for header length on a 3F with standard sized valves and a Schrick cam?

Scientific Design of Exhaust and Intake Systems third edition

Philip H. Smith FMech

John C Morrison MIMechE

Bentley Publishing

ISBN 978-0-8376-0309-4

valve sizes here.

https://www.clubpolo.co.uk/topic/347124-mk1-3-valve-specs/

29mm exhaust 3F is 1.142"

29.1mm exhaust GK is 1.146"

75mm bore is 2.95"

72mm stroke is 2.83"

For a 3F

With a Schrick 280

3 feet 9 1/2 inches

With a Schrick 332

4 feet 6 inches

from the valve to a 4 way collector for independent headers.

I think it's around 150mm from valve to head port (average of inner and outer radius) on GK but I can't find my diagram with the measurements.

Subtract that and flange thickness to get pipe length needed.

2 hours ago, dvderlm said:

For a 3F

With a Schrick 280

3 feet 9 1/2 inches

With a Schrick 332

4 feet 6 inches

from the valve to a 4 way collector for independent headers.

I think it's around 150mm from valve to head port (average of inner and outer radius) on GK but I can't find my diagram with the measurements.

Subtract that and flange thickness to get pipe length needed.

Thanks. What are you considering as the max rpm? This seemed to alter header length quite a lot in a calculator.

On 12/09/2025 at 08:37, dvderlm said:

Page 142 of the Smith book has a rough formula ....

I think this is a length for third harmonic at up to 6000rpm.

The preferred powerband is between 4-5k and 8k rpm

https://www.lms-eng.com/header-calculator

with same figures 280 cam and 300 feet per second at 1200°F exhaust gas at 5psi with optimistic 90% VE

at 6000rpm says 46.2 inches for 3rd harmonic which is 3 feet 10 inches.

And 22.5mm ID pipes.

At 7000 rpm it says 39.6" = 3 feet 4 inches using 25mm ID pipe.

4th harmonic shorter by 10 inches but more interference...

I wonder if you can trade a wider pipe where air moves a bit slower for shorter length?

Really top end motorbike systems (bike engines will rev to 11,500rpm as short stroke) like Akrapovic look to have divergent wider chambers after a short narrower section.

I helped an Austin Healey hillclimber at Shelsley Walsh Retrorides event with the low speed ignition timing. Twin SUs or Strombergs feeding a short (under 12inch) trumpet exhaust vertical out of the bonnet straight off the manifold that I think is cast iron 4-2-1.

It would spit flames on the rev limiter or when pumping the gas pedal when stationary. Illegal on a road car (not sure ... first use pre August 1986 emissions rules are more tolerant and noise test is quite subjective : noisier than expected for vehicle of its age at 2500rpm)

Scrutineer in track race would fail it for noise.

While moving I think there's a low pressure area above bonnet. Quite sluggish below 3000rpm, but then it screamed.

Edited 19 September, 2025Sep 19 by dvderlm

3-1 on 1500cc cast iron (siamesed)

scparts.co.uk

Exhaust manifold for MG | 318609 | SC Parts Group

Exhaust manifold (used item) shop at SCparts.co.uk kaufen · best Quality · top Consultation

The flames

Austin Healey at Shelsley Richard Woolmer flames

Edited 19 September, 2025Sep 19 by dvderlm

Measured a friend's 2006 Honda Fireblade CBR1000RR exhaust. It's 16 valve with peak power of 170bhp at 11,250rpm and max torque of 85ftlbs at 10,000rpm.

It's 4-2-1 system. 38mm OD primary headers 4-2 about 63cm from cylinder head exhaust ports to wider secondaries (could not get calipers in but guess 50mm OD), then about 30cm length to where these join 2-1.

So 3 feet 8inches (+ exhaust valves to cylinder head port) to where all pipes join.