The Flowbench thread.

So where does this all lead to ? The full inlet tract test.








This was tested with the valve lift at 0.310"

Interestingly at a 10" there is not sufficient depression to fully lift the CD125 Piston





This results in an artificially low flow figure - could have fixed the piston fully open but decided to just wind it up.





It was not until we reached 20" that the piston was fully lifted








So there we are - all together this stock assembly flows 67CFM @28" - probably a bit higher than it should be due to the 74mm bore unshrouding things so this would be representative of this assembly on a 998 engine.


If we take the rough guide of 0.2 - 0.225 BHP per CFM@28" we get 13.4 - 15 BHP per cylinder - 52.6 - 60 BHP

All done extensively earlier in this thread.

I'd guess that when you flow a component in isolation you get a relatively uniform distribution of flow from the open end which would help use more of the diameter. When you put another component in front it'll bias the flow in certain directions or introduce turbulence that'll influence how well the component behind it flows as the air's no longer entering the original component in an ideal way.

That'd be my guess anyway!

If you tested a section of pipe 100mm long and 50mm diameter you would get 'x' flow if you tested 5 they would all be 'x' flow - when you assemble them 500mm long you would get far less flow due to the combined internal frictional losses. For flow in any pipe you will get a loss per M value.

With any system like an intake port the length of the port all added up reduces flow - changes in velocity reduces flow - changes in CSA - (much the same as changes in velocity) reduces flow and imposed directional changes - reduces flow. Mismatches in the port sections etc. etc. all reduces flow.

Add it all up and the air has to work harder to get to where its needed - work takes energy so a lot of that atmospheric push that you started with at the inlet to the filter has been well used up by the time it gets bounced off the cylinder wall.

Make a port section larger from smaller - the air slows down and uses up energy slowing down - make it smaller again and it has to speed up again using up energy - consider all of the sectional changes from the filter to the valve - Half way down that carb the port is square ! And that's within 38mm of the start of the port.


In this test

Total flow has been shown to be worse than the worst performing component - Improvements in individual components will show an overall gain but in this case it looks like the manifold is where the main restriction is - Well, we actually know this to be the case by physical measurement. Trouble is that this manifold is not mine so I am not about to go opening up the ports to prove what is already known.

What I will do next is fit the banana manifold and see what happens.

So you are feeling flush and are fed up to the back teeth with Strombergs leaking fuel and going out of sync. You spend mucho hard earned on a new manifold and refurbed an old 28/36 DCD. ( DGV used so use your imagination)




And we get




But in reality you would not have because this manifold has had the ports opened up a little and using the gasket as a gauge the ports provide a pretty good match - so Even if it looks like we have pulled 6CFM out of the hat. I would estimate that if you had taken the same care to match the stock manifold to the head that improvement difference would be down to more like 2CFM.


Port length with the Banana - 410mm

Still if we use the formula - we are at 73CFM And should be seeing an extra 3-5BHP

As a quick explanation of how the assembly affects total flow.


When we test each part individually we have 14.5 PSI available to us at the front end of each component. When we start adding them all together we have a drop in pressure across them individually no matter how well they flow. None will be increasing the pressure available for the next - they will all be reducing it.





The only way that will ever get the flows at the valve seen when just testing a head is if you can design an induction that has virtually no resistance to flow - something like this taper set up will get close.





We will see how close we can get with our Imp heads when we put the individual runner DCOE/DHLA on it.

Fully modified sport manifold for CD150's





Which flowed





The attentive may notice that the depression had been turned up quite a way. Reason for this was to see what it was going to take to get the piston to fully lift.


Head cannot flow enough air for a CD150 - The bench confirming what's written about the CD150's being to big until the engine has been significantly modified.


As we did not get the full lift with the CD125 at 10" - the CD150 was only just lifting with a greater flow rate.








Same as with the DHLA/DCOE - there was a step down to the head due to the manifold being for a ported head




For a final test I turned the bench up to about as far I am happy with to see if I could get flow for full piston lift.

Even at 34" and pulling 86CFM through the carb it was still not opening fully








The expectation being that you will not start to approach the capacity of a CD150 until the head can flow enough to support 75-80 BHP

Next.

Testing with the full race manifold will take a bit more effort as it clashes with the fixture due to having an exhaust - even just stubs - fitted to it. I will need to grind some steel off it to get sufficient clearance.

Cheers Phil,

It's not a thread that makes easy reading but my own preference is toward the more technical threads a few contributors post from time to time so hopefully there are few on here that also find it interesting. I'm very fortunate to have found myself with access to a lot of old Ginetta G15 (Imp) race parts and its about time I used this flowbench for something.

Figures for the full race prepared big valve Chesman head ACC212 @28" corrected - average actual depression on the bench 16"

050 - 25CFM
100 - 46CFM
150 - 66CFM
200 - 85CFM
250 - 95CFM
300 - 105CFM
350 - 112CFM
400 - 116CFM
450 - 119CFM

Pulled the bench up to 28" at 0.360" and it was flowing 114CFM

Tomorrow I will look at dropping the comp manifold on to see what it does to the numbers.

Port diameter on this thing is huge - there is a little pinch point in an Imp port about 3/4" into the port on the Race head this pinch diameter is 27.5mm - On the Sport std head its 22mm

In most of the Imp tuning manuals the advice is for a full race head you take the port out to 25.75 to 26mm - this is 27.5 at its narrowest point. Exhaust port diameter should be taken out to 23.75mm - This head it's 25mm. Now bigger is not always better with ports so we will need to delve a little deeper at some point.