8v cylinder head porting with flow test results

As the title, this thread is dedicated to porting the 8v cylinder head, be it for road, race or hillclimb/sprint use.
It’s a little different to the last thread I started that covered the “how to” aspect. Here, I plan to show what changes give what results, and what can be achieved from minor mods during basic overhauling to full blown race specs.
There will be comparisons with other tuners work to show that we’re in the basic ball park, but we’ll be able to see what does what in the form of basic seat & valve shapes, to port sizes and the limitations in that area and what effect big valves really do have, including the fabled 35mm exhaust mod and what difference it really makes.
I will also try to add as time goes results with manifolds, both inlet and exhaust, standard, ported, DCOE etc so we can see what happens in that area too.
Hope you enjoy the ride...

First off we have to start with a baseline flow test. The donor head I used for this test came off an ’84 MK2 Gti, but subject to basic casting differences and from previous tests I’ve seen its representative of the type. The basic inlet port with the standard 40mm inlet valve flows like this...


.050 15.3 CFM
.100 27.2
.150 39.7
.200 52.4
.250 63.5
.300 68.1
.350 70.3
.400 71.2
.450 71.5
.500 70.5

I’ve seen figures in the region of 75-76cfm quoted for these heads on other flow benches using the same software to calculate the results, and with a very similar flow curve, so I’m happy with the above figures as a baseline as I doubt two flow benches will show the same results, a bit like rolling roads, due to calibration differences etc, and also differences in head castings, but at least my bench doesn’t show wildly inflated numbers so I’m happy we have a starting point with figures that wont mislead anyone.
I will be adding to this over the next few days so bare with me while I adjust the txt and juggle pictures to make some sense of it for you out there in reader land.

The first basic stage of modifying comes with tidying the throat area around the seat insert where it joins the head, and gently blending the short side turn, which gave the following results...



.050 14.2
.100 27.5
.150 39.5
.200 52.0
.250 63.6
.300 71.9
.350 72.4
.400 72.2
.450 73.2
.500 73.3

As can be seen, we have some positive results from .300 up, with some very minor losses & gains lower down.
The next step comes with tapering the valve guide, does it give the results you may think....

 

Legendary and +1 for sticky/engine FAQ IMO.

 

Excellent thread! - more please

 

Oh there's plenty more to come!... and thanks.

Rite, tapering off the valve guide, getting rid of that square shouldered edge must allow more airflow right?.....

Well, not quite like you'd think, at least not at this stage.

.050 13.5
.100 26.1
.150 38.7
.200 51.1
.250 62.5
.300 68.5
.350 71.8
.400 72.6
.450 73.5
.500 73.5

As can be seen, some "very" small gains were picked up at the very top of the lift range, but some bigger losses have appeared lower down. Why?... well my thinking is we have indeed allowed more air to pass around the guide tip, but in doing so the extra air has clattered into other restrictions and caused more turbulance, and thus less total flow.

We redeem ourselves a bit with the next mod though... back cutting of the valve.

A 30 degree back cut was made on the standard valve and a blended into the seat with some significant improvements in the lower part of the flow curve...



.050 14.5
.100 30.2
.150 45.0
.200 57.1
.250 63.4
.300 67.6
.350 73.0
.400 74.1
.450 74.1
.500 74.3

We have some good gains from .100 thou lift, .300 thou is still an issue with another loss but flow is then pulled back from there up. Maybe if the guide tip had been left alone at this stage we may have seen improvements all through, or maybe less of a loss around .300?... I'll check that one out at some point, but for now, your seeing the results in the order the mods and testing were carried out, no cheating in the JMR camp!

The next round of mods are carried out around the lower seat area, again with some surprises...

 

Mr. HC , how do you back cut the valve?

 

Great topic

 

top thread fella.. might ask santa for a die grinder this year. is there much difference in porting early mech heads and later hydro heads?

 

There are a number of ways, depends on the equipment to hand. The home way is to spin them up mounted in a drill clamped in a vice, then use a ginder to back off the edge. I mounted the test valves in a pillar drill and backed them off with a belt sander, but again a grinder or die grinder will do the job. Or if none of the above take them to a machine shop.

 

None that I'm currently aware of, but I'll be playing with a hydro head at some point when I do some more testing.

 

excellent thread

 

The valve guide bosses are smaller in mechanical heads, than in hydro heads. Other than that, there shouldn't be much.

Mr H - great thread. Keep up the good work.

The late Bill Blydenstein once told me (actually it was more than once) that 80% (or 90%?) of the power gains are had within 1/8" of the valve seat. I'd be really interested to see the results from doing that area first, then finishing off with the parts further away from the seat. It makes logical sense to me, because the narrowest part of the port tract is the valve opening, so by working there, you're freeing up the bottleneck.

 

Thanks chaps.

I'll try later today to write up the rest of the inlet testing, followed by road vs race head shoot-out... there "are" some surprises...

 

I love these threads...

On back cutting VAG valves , are they too hard for a lathe?

 

Fantastic of you to take the time out to do all this
Ill be chopping up one of these heads(dead) I have here soon also, may be of interest too.

Keep up the good work

Brian

 

Brian , is it a hydro or mechanical? I've probably a spare hydro head here .

 

Thanks guy's.... I maybe able to save you cutting a head up Brian, I've found the limits of how big the ports and throats will go before you find water or air!... I'll expalin later.

Anyway, back to the plot.

After the back cutting of the valve we're into test number 5 (of 11)... a 3-angle seat.

Now these are touted as "the thing" to do as far as valve seats are concerned, but in this case, in this particular order it's not a night & day difference...

.050 14.3
.100 29.9
.150 45.6
.200 54.9
.250 62.7
.300 69.2
.350 74.1
.400 74.9
.450 75.2
.500 73.7

Not the biggie we may think at the moment, but a step in the right direction. With some variable changes lower down, there are some positive gains from .300 up.

The next change comes with opening out the throat a little more, 2mm more to be exact, and the results panned out like this...

.050 15.0
.100 29.8
.150 45.6
.200 53.4
.250 62.5
.300 69.7
.350 73.0
.400 74.9
.450 74.6
.500 76.1

With a small gain just off the seat, we have mostly losses apart from at .300 and right at the very peak at .500, where most cams (certainly the road stuff) dont even lift to... a mod too far maybe?

Well, at least the next mod is positive, with gains at all but 2 points in the lift curve... this is what happens when the guide boss is thinned down to all but the same size as the guide...

.050 14.9
.100 30.8
.150 45.6
.200 57.2
.250 63.3
.300 69.1
.350 74.7
.400 76.4
.450 76.9
.500 76.7

Well, we're getting there, slowly, with a good couple of CFM gained from .350 up and some smaller improvements lower down, it's starting to head (no pun intended!) towards the home straight, just another 10 (ish) CFM to go!

 

silly question what do all of the numbers mean?

 

The volume of airflow , at a given pressure , that is able to flow into the engine through the valve port . In general , the higher the numbers , the more bhp, due tothe head breathing freer . Not always the case though , sometimes the numbers can be too high , leading to a drop of torque and/or the air speed slowing down

Mr HC is just demonstrating how different head porting mods can affect the flow .

 

Back to the plot....

We left the last stage of mods at the narrowing of the guide boss which gave some good improvements from the midrange flow up. At some point I'll grind out the guide boss totally and see if that helps at all, but that's for the next round of mods when I play with valve sizes.

The next step involved raising the roof of the port. I did this in two stages as a bias towards the cylinder wall side of the guide is alleged to improve flow dramaticaly. Well, I would'nt quite say that, but an improvement was had...

.050 15.7
.100 31.5
.150 46.0
.200 57.3
.250 63.4
.300 70.4
.350 75.1
.400 76.4
.450 77.9
.500 77.0

Not a dramatic improvement, but our strongest flow curve at that point.

If raising the port roof was good on one side, maybe both would be better? Only one way to find out!...

Grinding up the roof of the opposite side of the guide boss produced these results...

.050 15.5
.100 31.4
.150 46.0
.200 57.0
.250 63.0
.300 70.0
.350 74.9
.400 77.4
.450 77.1
.500 77.2

Well, apart from a 1 CFM gain at .400, the rest was nothing to get bowled over about.

At this stage I had a very tidy throat with all the dimensions correct for the valve size used, but something was still missing...

Even though we are told that the main straight part of the port should flow well enough, the dramatic improvements wer'nt really coming with the throat and seat work, though we did have a peak flow improvement of 6CFM (thats around 8% up from our starting point), with some good increases at the lower end of the flow curve.

I've been led to believe that around 90 CFM is possible with some basic porting on these heads whilst still using the standard valve. Had I been doing something wrong or was there something missing?

Well, the latter seems to be the case. The only thing I had'nt done was open up the main port runner. A basic rule of thumb is that the port should be around .08 to .82 of the valve size. I ground the port out to slightly under that for a couple of reasons, one because I've broken through on the race heads before (I'll go into that later), and two, as this is a road type port I wanted to keep the air speed up.

So, would opening out the main port get this head to my goal of 90CFM?...

.050 15.5
.100 30.7
.150 43.8
.200 55.1
.250 65.7
.300 73.6
.350 83.4
.400 85.4
.450 87.9
.500 86.0

I had the expected losses up to around .200 thou, there after some small improvements at .250-.300, then take off!... a gain of 8.5 CFM @ .350, and almost 11CFM up @ .450... giving us a total improvement of 16.5CFM at peak flow, or 23%.

With high percentage improvements right through the lift range, was there anymore to come?

 

Excellent step by step thread Jason with some excellent results

What kind of power gain over stock would you anticipate from a head with 90cfm flow?

I wait with interest to see what the big valves do for flow. I wonder now if the standard p&p job I had done on my 8v was as thorough? I never really felt any gain until I went bvh (albeit on a 2L block)