Flow Benches and head porting

[zefarelly]

ANother Dark art . . . . but what experiences do people have ? how much have you got from an old nail ? or how little for that matter

regardless of total flow, I just flowed my race head which is nearly three years old now, it worlks well but I wanted to record what it did as a bench mark to start a new one and see if I can better it.

Inlet flowed 138 CFM with no valve in and 135 with the valve in, I believe its possible ( but rare) to gain flow with valves in, and I know If I could get the ports bigger I could flow more, but I think its respectable for the shape of the port and size of the engine and valve size etc ( 39.6 inlets)

the curve and figures almost exactly followed a Pug 106 rally head which has the same size valve

[profjohn]

You might find that when comparing port flow for different valve sizes that it is convenient to use a ‘flow coefficient’. This is usually defined as the ratio of the flow rate of the port and valve and the flow rate of a straight tube of the same diameter as the valve seat.

The flow coefficient is calculated from the ratio of (the volume flow rate measured through the port and valve in m^3 s^-1) divided by (cross sectional area of the port at the valve in m^2 times the square root of (the pressure drop across the port and valve in Pa, or Nm^-2, times the density of the fluid in kg m^-3)). This is a bit hard without an equation editor!

So, a flow coefficient of 1 means that the port flows the same as a straight tube of the same diameter. In practice a value of 0.6 is pretty reasonable. A FF 1800 Zetec has an inlet flow coefficient of 0.62 at 11mm lift, for example.

Good luck.

[grelley]

I would suggest you get a book called Practical Gas Flow published by MRP Speedsport. This has a lot of good information in it. Well worthwhile.

Regds

[greenamex2]

Bigger isn't neccesarily better as it can cause an unacceptably low gas speed.

An example is on my previous engine we opened the exhaust ports a bit and gained a couple of BHP at the top due to the peak revs raising. In the mid range we lost 20-30BHP and the car was slower and more difficult to drive.

Without knowing what 'pressure' you are using it is hard to compare numbers but my 144.9 open or more than 50% valve lift with no inlet system. The exhaust flows 108.6 open or at more than 50% valve lift with no exhaust system. These are corrected figures.

My valves are smaller than yours but I have four of them!

The head was done by an ex McLaren F1 head man and made 20BHP more than the previous effort so I am pretty sure it is somewhere near optimal.

[Dennis.Doyle]

Originally Posted by Denis Bassom

...
Without knowing what 'pressure' you are using it is hard to compare numbers
...

I was wonderig that. Is 10" of water the accepted "standard" for comparison?
...

The head was done by an ex McLaren F1 head man and made 20BHP more than the previous effort so I am pretty sure it is somewhere near optimal.

I was wondering that. Is 10" head of water the accepted standard for comparison?

Are the tests all done with air or with different fluids corrected for Reynolds numbers of whatever.

Hmmm, 20 bhp - could use some of that myself. Can you tell us which engine?

[greenamex2]

Toyota 1.6 16 valve (aka 4a-ge).

It would probably make more but the engine was wet sumped and was losing oil pressure the more it revved.

It's dry sumped now but I haven't got the guts to rev it any higher.

[Notso Swift]

Originally Posted by Denis Bassom

Without knowing what 'pressure' you are using it is hard to compare numbers but my 144.9 open or more than 50% valve lift with no inlet system. The exhaust flows 108.6 open or at more than 50% valve lift with no exhaust system. These are corrected figures.

My valves are smaller than yours but I have four of them!

1 x 39.6mm valves is 1231 sq mm
2 x 28.5* valves is 1276 sq mm
so about 3.5% difference
If we are talking at the same pressure (vacuum)
138 cfm
144.9 cfm
About 5% difference
Considering it is generally agreed that a 4 valve motor is about 10% more effecient than a 2 valve engine, I reckon Zef has done a pretty good job.

To work out if your head is near optimum I would look at how much torque it is producing, if it is up around 90 ft/lbs/litre (120 Nm/Litre) you will be hard pushed to beat that. The best 2v head will be about 10% behind, as said before. Of course WHERE and HOW LONG you hold that torque is just as important for making a car nice to drive, (and for the big BHP number)


* Standard 4AGE, you could be bigger, of course.

[zefarelly]

all good points chaps, my head has been tested on a flow pro quick machine at QEP and its callibrated with a fixed tube each start up, (and computer controlled)

90lb/ft per litre is an interesting stat, early engines where 120 BHP and 120 lb/ft @: 6k and 4k respectively, and this was a fast road engine, I think I'm up to 140 and 130 now which from 1500 is close . . .we'll see, the important thing is it got me into pot winning situations, and its reliable.

but I still want more !!!

[greenamex2]

I am about 15lb/ft off my theorhetical 144lb/ft.

I have never seen a 4a-ge on a UK dyno make more than about 135 so I suspect the natural design of the engine just precludes good torque figures in place of stupid revs (the engine in early form has HUGE inlet ports to start with, they are smaller on the later heads and these usually produce higher torque figures for the same engine spec, they just don't rev as high).

[zefarelly]

I've tried a 1200 casting with the same size valves but the port size restricts maximum flow, it matches the bigger head at low lift but gets left behind after 2-3mm lift

[trikesrule]

I'm assuming most you race on the black stuff. The team I'm with race dirt track speedway where torque and a nice wide power band that makes it easy for the driver is more important than hp. Lately we've found by keeping the port velocity up has made the car quicker out of the corners and more importantly our driver is able to get on it earlier and harder than last season. Our lap times have come down to 1 or 2 tenths off the record where as last season when we went looking for hp we were fast (1/2 a sec outside the record) only when the 'track' was flat and fast. Now we can run that pace in most any track condition. One thing that a lot of people have said to me is our car puts a pass on any car easily exiting a turn where as last season we had to round them up at the end of the straight and do a desperate at turn entry. By the way the cam is still the same, the only thing different is the port velocity. I always thought that torque is what actually moves the car and hp was along for the ride. One old chap said to me 'hp is how fast u hit the wall and torque is how far you move it'. Perhaps you guys might sit back and take a breath before you go searching for more hp or more revs. Or maybe thats just me thinking outside the box again.........trikes

[phoenix]

Originally Posted by Notso Swift

1 x 39.6mm valves is 1231 sq mm
2 x 28.5* valves is 1276 sq mm
so about 3.5% difference

Surely, the area that affects the flow into the cylinders is not the surface area of the valve, but the valve "curtain area" - i.e. the circumference of the valve(s) multiplied by the valve lift.

In that case, the circumference of one 39.6 mm valve is 124.4 mm and the combined circumference of two 28.5 mm valves is 179.0 mm. So for a given valve lift the 2 valves have a flow area nearly 44% greater than the single valve.

[trikesrule]

I was with a chap that raced a Mini Sprint (they look like sprintcars but with m/c engines). The engine was a FJ 1100. The reason I mention this is I think perhaps u guys have already found the 'pot of gold under the Rainbow' concerning heads to suit your various engines. Back with the m/s the FJ head had huge intakes and was hurting us badly coming out of corners. We talked to a camshaft man. We closed the intakes a little earlier and left the exhaust open a little longer. Boy what a change. The chap who they call 'Cruise Missile' was the king suddenly found we were lapping him in a Feature race.....trikes

[Dennis.Doyle]

Originally Posted by phoenix

Surely, the area that affects the flow into the cylinders is not the surface area of the valve, but the valve "curtain area" - i.e. the circumference of the valve(s) multiplied by the valve lift.

In that case, the circumference of one 39.6 mm valve is 124.4 mm and the combined circumference of two 28.5 mm valves is 179.0 mm. So for a given valve lift the 2 valves have a flow area nearly 44% greater than the single valve.

Absolutely spot-on!

However Notso is dead right in that, it's not the max power, torque or flow that matters it's the area under the power, torque or flow curve that counts. Also, since power == torque * revs & all heat engines depend upon how much fluid flows into them the different curves are all just different ways of expressing exactly the same thing.

[Dennis.Doyle]

Apologies - getting a bit carried away there. Of course power & torque are nailed together, the relationship of either to flow aint quite so simple.

But, getting back on thread, some of you might be interested in the flow numbers at:

http://users.erols.com/srweiss/tablehdc.htm

They are mostly US models that don't mean much to me, but there are a few Euro & Jap ones too, including the 4 AGE. Sadly, this man's flow bench has been turbocharged - it runs at 28" of water pressure, hence I reckon the flow rates quoted represent a few bar of boost! Still, does show what big ports/valves the Toyota has.

[Notso Swift]

Originally Posted by Dennis.Doyle

Quote:
Originally Posted by phoenix
Surely, the area that affects the flow into the cylinders is not the surface area of the valve, but the valve "curtain area" - i.e. the circumference of the valve(s) multiplied by the valve lift.

In that case, the circumference of one 39.6 mm valve is 124.4 mm and the combined circumference of two 28.5 mm valves is 179.0 mm. So for a given valve lift the 2 valves have a flow area nearly 44% greater than the single valve.


Absolutely spot-on!

However Notso is dead right in that, it's not the max power, torque or flow that matters it's the area under the power, torque or flow curve that counts. Also, since power == torque * revs & all heat engines depend upon how much fluid flows into them the different curves are all just different ways of expressing exactly the same thing.

That could be right if maximum flow occured only at maximum lift, but it dosen't. Generally the surface area is proprtional to area of the valve, and it translates pretty well to the capacity for flow. If the port is right I think (from memory here so don't cricify me if i have it wrong) maximum flow occurs from either 1/3 or 1/4 of the valve diameter. Generally, the steeper the ports (closer to the vertical of the valve) the earlier it can happen

Dennis B
135 is that at the wheels? Rolling roads are so crap with so many variables the best you can do is use it to compare one run to another on the same unit.
One thing you can do with the 4AGE is give it mega revs, but as always, mega revs are mega money

[greenamex2]

That 135 was on a rolling road whilst my 127 is on two separate engine dyno's. The rolling road is a very reputable one but you do have a point.

As you said, whilst my flow figures aren't 'massive' it achieves them from 50% of the maximum valve lift or less.

And my inlet valves are quite a bit bigger than standard (the exhaust valves are only a bit bigger).

Mega revs aren't that expensive if you buy the right bits the first time. It's getting it wrong and taking four blown up engines before going to a better engine builder that's expensive. My 9500RPM engine was only about £1500 more expensive than my 8500RPM one but makes more than 50BHP more, 1-2 seconds a lap and the difference between winning and not winning.

[zefarelly]

my head flows right up to 13mm lift although I'd never open that much, 11.3 with the new cam, 10.4 with the old, the old written guide was 1/4 of the valve diameter was the best/max lift, but I think that was thought up before people designed better heads and 4 valve affairs, 1/3 makes sense as 13 x 3 is pretty close to my valve size

[Notso Swift]

Fair point, I often see stuff that is based on US type 2 valve V8's. This is probably because I know people who have spent more time, both doing it and just as importantly telling everybody about it!
As an example with the local big Fords a lot still use Windsors (some Cleavelands) which is proven and all (the VE Supercars use them, so there is also a trickle down effect) While the current Road cars have a 5.4 litre 4 valve V8 based on the Ford Modular unit. 350 or 390 hp straight out of the box, thanks very much, yet no one seriously mod's them because they would have to do some real work on development as opposed to bolting together a bunch of off the shelf parts. This is despite a reletively low stressed 110 or 120 hp/litre being there for the taking, as with most 4 valve motors. While the best 2 valves can do that it costs a bomb. GM's (350 chev or LS1) are different because they are all 2 vlave anyway.

Dennis what you need is an old Formula Atlantic motor!

[greenamex2]

Originally Posted by Notso Swift

Dennis what you need is an old Formula Atlantic motor!

My engine IS an old Formula Atlantic motor (just wet rather than dry sumped until this year)!

Which is how I know that most of the figures quoted for these engines are 'optimistic'. For example my old engine bloke used to build them, dyno them at around 220BHP and then ship them to the States. Obviously the air in the States has far more oxygen in it because with would suddenly be making 240BHP.

[zefarelly]

most engine output figures are 'optimistic' or at best ronded up to the nearest tasty sounding Number

example . . several people have said 105-108 BHP for FF engines . . . . several engine builders I've spoken to recently have said they've tested other peoples FF engine and a more realistic figure is 95-100 on a good day.

[GORDON STREETER]

The old addage for a 2 valve engine was 100bhp per litre that was first obtained on a Ford engine by Cosworth many moons ago . With more modern
parts and acurate machining you can get 120bhp plus on carbs, but at a price and a compromise on torque.
(Note I am only talking about historic classic engines )

[greenamex2]

Can't a Kent 1600 make 190BHP+ (115BHP/litre) if you spend enough on it/rev it high enough/sacrifice loads of torque?

Or is that just paddock gossip?

[zefarelly]

a lot depends on regs really . . . . essentiall a kent engine can be as low as 40 BHP per litre in std form, right up to hundreds with 4 vlve heads etc

I think 100 per litre is a good mark fior historic spec engines I'm reasonably happy with a bit less and plenty of torque . . .although I'm just starting a new engine which may be a bit nuttier !

[Dennis.Doyle]

Originally Posted by zefarelly

...
I think 100 per litre is a good mark fior historic spec engines I'm reasonably happy with a bit less and plenty of torque . . .although I'm just starting a new engine which may be a bit nuttier !

I'm up the conservative end with Zef on this one for 2 reasons. First, there is a graph of a Wilcox 1800 crossflow fitted with an A9 cam in the Piper catalog. They don't normally understate things, that shows 174bhp & such an engine is likely to cost a 5 figure sum new. Second, take a peek at the figures here (pre-xflows, twincams & BDs):

http://www.race-cars.com/utility/coswrthr.htm

Sure it's as inaccurate as any other source, but I feel it hasn't suffered from inflation as much as some and there not a right lot over 100bhp/litre there.

Obviously it's much easier to get the bhp/litre on the smaller xflows as you can get the same huge valves in the small ones as you can in the big ones and you are asking the gas to do a lot less with lower piston speeds for equivalent revs & less volume to fill.

I too have seen the advert for the "200bhp xflow" but suspect they may be talking old-fashioned gross SAE