Tu Cylinder Head Flow Results

[christopher 5 1 Cars]

For those who are running modified TU heads I would be interested to see flow results for (before and after modifications) for any of these TU heads at various lifts.

 

I'm interested so see what kind of % increases can be achieved for various TU heads.

 

Or even how they compare if someone has tested a few on the same flowbench.

 

Anyone have any?

[christopher 5 1 Cars]

Come on chaps SOMEBODY must have some results

[bren_1.3 1]

bloody hell chris, thread from the dead here.

 

maybe pumaracing has something for 'that' citreon ax he built?

[TKH 0]

It is virtually impossible to subjectively use flow bench data. It does not take in to account your inlet tracked, cylinder or piston shape or how the exhaust flow effects the inlet flow. You will even get totally different results depending on how the rig is setup. Especially on the TU 8V head. For instance you can increase the inlet port size and see absolutely no gain in power. It can actually loose you power. If your looking for a very good head though I would recommend calling GMC. My 1360cc has been numbed down due to being limited to a 5 speed box but is still running ~130bhp at the wheels.

[RossD 44]

For instance you can increase the inlet port size and see absolutely no gain in power.

 

Increasing the inlet port size does not automatically mean more flow, you can flow a head and get much better results without even touching the port size.

Its all about velocity too, increasing the port size slows the velocity....

 

A before and after graph would be good to see for the TU 8v head, there are plenty floating around for the XU heads but I have yet to see one for a TU.

It doesnt really matter how different they are, it would just be good to see some rough numbers, purely out of interest to be honest!

 

Sam (If you read this!!) Did you get any before/after results from your QEP head that was on your old 205?

Edited March 16, 2007 by RossD

[PumaRacing 2]

bloody hell chris, thread from the dead here.

 

maybe pumaracing has something for 'that' citreon ax he built?

 

Nope. I very rarely use the flowbench these days. Don't really need to anymore. It teaches you the shapes you need in the first few years and then you can achieve them without it.

 

I did however have a look at Matt @ QEP's bench for the first time last month and checked it out with my flow calibration plates. I pretty much knew there was a 5% to 10% error anyway from figures posted previously (Mi16 in particular) but decided not to say anything until I'd checked it myself. It turned out to be about 7.5% high on average but with a varying error at different flow levels. A look inside revealed a design fault in the way the pressure readings were taken. I told him how to fix that and he's just got back to me after doing so and the flow readings have dropped by.......... 7.5%. Well fancy that. So it should now be accurate and comparable to mine. I'm sending him a calibration plate to get it spot on.

 

That's one of the problems with flow benches just as with rolling roads. Half of them don't read true. I've checked out numerous flow benches over the years and found them to be anywhere from accurate to 15% high. So I tend to ignore anyone else's flow figures in the same way as I ignore their power ones unless I know the dyno in question.

[RossD 44]

Nope. I very rarely use the flowbench these days. Don't really need to anymore. It teaches you the shapes you need in the first few years and then you can achieve them without it.

 

I did however have a look at Matt @ QEP's bench for the first time last month and checked it out with my flow calibration plates. I pretty much knew there was a 5% to 10% error anyway from figures posted previously (Mi16 in particular) but decided not to say anything until I'd checked it myself. It turned out to be about 7.5% high on average but with a varying error at different flow levels. A look inside revealed a design fault in the way the pressure readings were taken. I told him how to fix that and he's just got back to me after doing so and the flow readings have dropped by.......... 7.5%. Well fancy that. So it should now be accurate and comparable to mine. I'm sending him a calibration plate to get it spot on.

 

That's one of the problems with flow benches just as with rolling roads. Half of them don't read true. I've checked out numerous flow benches over the years and found them to be anywhere from accurate to 15% high. So I tend to ignore anyone else's flow figures in the same way as I ignore their power ones unless I know the dyno in question.

 

Ok then, how about this.....

 

How does the TU head flow, comapared to say the 8v XU heads? No numbers needed here really. I know the XU5 head has larger valves than the TU3/TU5 head...

Just really thinking whats the better head to get power from, if I was to compare a XU5 to a TU5 :-) (TU5 head is identical (same part number) to the TU3 head Pumaracing did for Garrys AX)

Edited March 16, 2007 by RossD

[christopher 5 1 Cars]

Yeaaahh A start.

 

Puma...

So its hard to compare flow figures from different flow benches since they may have different levels of calibration and use different pressures. Well what if each figure from a flowbench is compared to the stock OE flow on that bench. Then at least it will give a % increase. Isn't that something? Or still impossible to compare?

 

Best case senario is of course (as mentioned in orginal post) if anyone have measured flow of different TU heads on the SAME flowbench. Then it would give an idea how the different TU heads flows compare at various lifts and perhaps what kind of % gain is achievable.

 

Seen lots of flow graphs of XU engine but none from TU engines.

Edited March 16, 2007 by christopher

[Sandy 191]

This is starting to allude to the futility (IMO) of trying to quantify flow as a way of judging heads. The TU (106 Rallye/XSI etc) head chamber is a much better design than the 205 XU and is capable of safely running much higher CR without detonation, the flame spread is clearly better too with best power coming at much lower ignition advance values. The Valve sizes on the 106 Rallye/XSi heads are good as they are as well, especially given the port shape, which clearly makes good use of the valve perimeter. This allows fairly mild cam profiles to work very well.

The standard head is very effective IME and I doubt many people working the head are getting any real gains from it, despite the claims!

Edited March 16, 2007 by sandy309

[RossD 44]

This is starting to allude to the futility (IMO) of trying to quantify flow as a way of judging heads. The TU (106 Rallye/XSI etc) head chamber is a much better design than the 205 XU and is capable of safely running much higher CR without detonation, the flame spread is clearly better too with best power coming at much lower ignition advance values. The Valve sizes on the 106 Rallye/XSi heads are good as they are as well, especially given the port shape, which clearly makes good use of the valve perimeter. This allows fairly mild cam profiles to work very well.

The standard head is very effective IME and I doubt many people working the head are getting any real gains from it, despite the claims!

 

Perfect, just the info I was after really

[PumaRacing 2]

Yeaaahh A start.

 

Puma...

So its hard to compare flow figures from different flow benches since they may have different levels of calibration and use different pressures. Well what if each figure from a flowbench is compared to the stock OE flow on that bench. Then at least it will give a % increase. Isn't that something? Or still impossible to compare?

 

It'll be a reasonable guide to the % improvement but as I said if the bench has a non-linear error i.e. the percentage inaccuracy changes at different flow levels then this will skew the gains, especially on heads where a very large increase in flow is possible.

[PumaRacing 2]

Ok then, how about this.....

 

How does the TU head flow, comapared to say the 8v XU heads? No numbers needed here really. I know the XU5 head has larger valves than the TU3/TU5 head...

Just really thinking whats the better head to get power from, if I was to compare a XU5 to a TU5 :-) (TU5 head is identical (same part number) to the TU3 head Pumaracing did for Garrys AX)

 

It's no contest really. A big valve XU has 20% more valve area than a std valve TU and AFAIK no one does big valves for the TU. There are obviously completely different valve train systems and minor differences in other things but nothing can overcome that advantage in valve size.

 

The XU head needs a lot more work to flow well than the TU but as long as you get that right the XU engine will always produce more power than a similar TU one.

[TKH 0]

I get my valves from GMC. Some one I know is using a combination of modified Vauxhall valves and valves from another PSA engine in his 205 XS engine. Both Longman and Lowflow I believe sell bigger valves as well. I think cc for cc the TU is a better engine. The newer TU engines are not as strong as the old ones though.

[PumaRacing 2]

This is starting to allude to the futility (IMO) of trying to quantify flow as a way of judging heads.

 

It's a vitally important part of the evaluation of any head but it helps if you 1) do actually have a flowbench and 2) know how to interpret the flow figures over the full lift curve. With the correct algorithms and within certain limits you can predict the effect of flow on power to very high accuracy. When I was building a lot of CVH engines for the XR2 Challenge and reliant on the base flow of the casting as you weren't allowed to modify the heads I developed computer programs to analyse flow which predicted power to 1 bhp over a range of about 80 bhp at the wheels to 95 bhp at the wheels. I could select castings based on how they flowed and tell the customer exactly what power he would get from the engine within that range and if there were better castings still to be found. Something no other engine builder could do apparently which is why mine took 1st, 2nd and 3rd in championship every year I built for it.

 

Those same programs applied to other engines also help predict power very accurately although of course each engine needs to have factors added in for its capacity, cam profile, burn effectiveness etc. Nevertheless, flow is always the base on which power is developed and anyone who can measure flow and improve on it has a huge advantage over anyone who can't.

 

The TU (106 Rallye/XSI etc) head chamber is a much better design than the 205 XU

 

They aren't that different really. In fact the XU is more compact.

 

and is capable of safely running much higher CR without detonation,

 

Again, allowing for the inherent TU advantage of smaller bore size, considerably so in the case of the 1300/1360, I don't see much difference between them.

 

the flame spread is clearly better too with best power coming at much lower ignition advance values.

 

Data to support that?

 

The Valve sizes on the 106 Rallye/XSi heads are good as they are as well, especially given the port shape, which clearly makes good use of the valve perimeter. This allows fairly mild cam profiles to work very well.

 

What allows mild cam profiles to work well is the ratio of flow to cylinder capacity. That's mainly another inherent advantage of smaller engines benefiting from the rule that volume is proportional to the cube of the linear dimension and area, valve area, only to the square of it. That means a 1360 TU has a completely different cam requirement to a 1900 XU. However in the case of two 1600 engines, one a TU and one a BV XU, the XU would want the lower cam duration, not the TU. The other big factor affecting cam design requirement is that the TU has 30 degree inlet valve seats and the XU has 45 degree ones. What exactly this affects I'll leave you to ponder though.

 

The standard head is very effective IME and I doubt many people working the head are getting any real gains from it, despite the claims!

 

I haven't seen anyone making any specific claims for gains from their TU heads. I certainly haven't. There will be a flow development exercise at some point when time allows but at the moment Garry's engine is beating what it needs to so there's no pressure to make it better still. If someone wants to move the goalposts of course we'll do what we need to.

 

I think before you want to be saying too much about flow you need to have a flowbench and spend a lot of time learning what it can teach you. That's a long process though and the flowbench doesn't actually give you any answers, it just tells you what you've achieved so far and nothing about how much further you might be able to go. It's basically just a dumb measuring device. The intelligence and the knowledge to interpret what it says has to come from other sources. I set my bench up in 1991 and for ten years it was an invaluable tool in helping me build better engines. Eventually you move on to other factors though. Gas speed in certain areas of the port is key as are valve seat profiles.

[RossD 44]

Now this is more like it :-)

 

Thanks for the info Sandy/Puma. Its always good to get some proper experienced input in on a thread like this, its what makes this forum stand out over the others out there

[Sandy 191]

I think cc for cc the TU is a better engine.

I would concur with that, both the TU 8v and 16v heads seem to out perform their valve sizes.

[Guest JasonS]

I think before you want to be saying too much about flow you need to have a flowbench and spend a lot of time learning what it can teach you. That's a long process though and the flowbench doesn't actually give you any answers, it just tells you what you've achieved so far and nothing about how much further you might be able to go. It's basically just a dumb measuring device. The intelligence and the knowledge to interpret what it says has to come from other sources. I set my bench up in 1991 and for ten years it was an invaluable tool in helping me build better engines. Eventually you move on to other factors though. Gas speed in certain areas of the port is key as are valve seat profiles.

 

I know nothing can truely subsitute experimental data which can be obtained on the flow bench, but in finding a guideline to your limitations of modifying heads for ideal flow and gas velocity, without going too far and ruining a head, do you consider using CFD to model it? Granted that you probably have enough experience now with common heads but if it was on something you werent used to porting it could work well.

[PumaRacing 2]

I know nothing can truely subsitute experimental data which can be obtained on the flow bench, but in finding a guideline to your limitations of modifying heads for ideal flow and gas velocity, without going too far and ruining a head, do you consider using CFD to model it? Granted that you probably have enough experience now with common heads but if it was on something you werent used to porting it could work well.

 

I imagine that's a powerful tool if you have the budget and the right program but it isn't something I need to be able to beat the sort of engines I come up against. I'm sure in F1, Nascar and the like it has its place but most club racers over here don't have engine builders who even know how to use flow benches, or have them, so the level of competition isn't really that high. Even if you have a flow bench you need very advanced maths to help you interpret the data and again most race engine builders are really fitters who came from car servicing backgrounds and fancied the extra money in race work. They can't program algorithms to evaluate flow curves. Their flow bench, if they have one, might tell them they've gained 20% at some lift or other but they still don't know what the targets are and how close they have got to them.

 

Ultimately it's your IQ and your maths and physics knowledge that determines how good your engines are, not the toys you have to measure things you barely understand.

[Guest JasonS]

Yes it is a powerful tool for fluid dynamics. Some experimental work would have to go into modelling the head casting, ports and chamber and possibly head operating temps if the effect on inlet temps was also looked at, to get the correct boundary conditions for analysis. And yes i also believe that it is the theoretical background and understanding that would make a successful engine builder. To be honest, from hearing that, there must be alot of engine builders out there that really dont know what they are doing. If Inlet velocity, Boundary layer thickness, shear stress at the port wall, skin friction co-efficients etc. data was asked about for their porting i'm sure they wouldnt know. All of which are essential to fluid flow.

[PumaRacing 2]

Yes it is a powerful tool for fluid dynamics. Some experimental work would have to go into modelling the head casting, ports and chamber and possibly head operating temps if the effect on inlet temps was also looked at, to get the correct boundary conditions for analysis. And yes i also believe that it is the theoretical background and understanding that would make a successful engine builder. To be honest, from hearing that, there must be alot of engine builders out there that really dont know what they are doing.

 

You can't imagine the half of it. If you'd seen what I have in the 20 years I've been doing this. I've spoken to long time engine dyno and rolling road operators who still had no idea power and torque were mathematically linked by a simple formula. They'd never twigged the curves always cross at 5252 rpm for a reason. One I knew printed the flywheel power but the torque at the rollers and never realised the two were not properly related. However not a single customer ever realised there was a problem either. Very few people actually understand the printouts they get given other than the headline bhp at peak power number.

 

I remember a chat with a race engine builder over a beer one day and the subject, for god knows what reason, got onto the vacuum of space. "Can't be a vacuum" he opined, "stands to sense." "Why's that?" I asked. "Cos if it was really a vacuum what would rockets have to push against?"

 

It's hard to even know where to go when someone's knowledge of basic physics is so flawed they don't understand Newton's laws of motion. I tried briefly to explain but gave up in the end.

 

Once in a while you met someone who has enough maths and physics knowledge to have a half decent conversation with but it's rare. Most people build their engines by trial and error until they find what works. They then call themselves engine builders. Computing the requirements from basic principles isn't something our school system manages to teach. It doesn't even cover basic spelling if newsgroups and forums are anything to go by.

[Sandy 191]

Nothing wrong with experimentation, it accounts for all variables (if the ambients are allowed for). Modelling can be interesting, but it can also waste alot of the time it's claimed to save.

Some of the best engines i've seen recently have heads that haven't been near a flow bench.

 

Please continue telling us how clever you are Dave.... I never get bored of it.

[Guest JasonS]

Nothing wrong with experimentation, it accounts for all variables (if the ambients are allowed for). Modelling can be interesting, but it can also waste alot of the time it's claimed to save.

Some of the best engines i've seen recently have heads that haven't been near a flow bench.

 

Definitely, i'd go as far to say as experimentation is always better than numerical modelling. It should only be used as a guide and in this case a tool in combination with experiments. I feel in this case if you are working on for a example a head that you havent had experience on enough to adequately say you know where to take off material and leave material for good flow results it would come in very handy. You could model the ports numerically, and try removing material in places to see if it improves flow. If it doesnt improve flow or reduces flow, this area can be disregarded, or it would give a base for the amount of material needed to be removed in certain areas. I would never take it as being exact but it just gives a very good guide to start experimental work with and in particular it would reduce the amount of heads that you would go through if you just tried experimentation alone.

[PumaRacing 2]

Nothing wrong with experimentation, it accounts for all variables (if the ambients are allowed for). Modelling can be interesting, but it can also waste alot of the time it's claimed to save.

Some of the best engines i've seen recently have heads that haven't been near a flow bench.

 

... and all of the worst engines I've seen had heads that hadn't been near a flow bench and all of the exceptional ones had heads that had. I also know people with flow benches whose porting and seat cutting is diabolical and ones without who do very nice work. Experience, common sense or being taught by someone else who knows can enable you to do very nice head work, however it'll always be improved on by someone with a flow bench who knows what they're doing. You might as well try to claim you can judge distances by eye so you don't need a micrometer to measure crank journals with. If you think you can shape ports by eye then you'd be amazed what a flowbench would tell you in just the first few weeks of having one.

 

Please continue telling us how clever you are Dave.... I never get bored of it.

 

Similarly it's interesting seeing how stubborn you are, how you think anyone who doesn't share your views is 'getting at you' and how instead of trying to learn from other people's knowledge you keep your head resolutely buried in the sand. There's one thing that stays true throughout all our lives. No matter how much you know or think you know there's always someone out there who knows more. When you stop being willing to learn from them is when you stagnate. There are things I do with engines, and have done for many years, because I think it's the best way. However if someone were to prove to me a better way I'd change in the blink of an eye.

 

That actually happened a couple of years after I built my flowbench. Someone very experienced showed me a better way to cut seats. It added a massive 10% to the low lift flow I was getting and not much less to the high lift flow. You see, a flowbench can only tell you how good something is if you think to try it out. It doesn't tell you what to try and every port is an exercise in the infinite so there will always be things you can't try or would never think to try.

 

What improves us all is competition. Someone else beating you is the best thing in the world to upset your complacency. Unfortunately schools can't be doing with that sort of thing anymore because it upsets the thickos so everything gets dragged down to the lowest common denominator. That's why kids still come out of it after ten years unable to spell or do simple maths.

[TB_205GTI 1]

Dave a few questions from me

 

1) How come you never have seen a flowbench read LESS that the "actual" flow? You say they always are spot on or read high, that doesn't make sense to me. There will always be variation on the benches, and some WILL read low and some WILL read high. But in the end it doesn't matter (IMHO) if it reads too high or too low, you always need the differences between the starting point and the progress. I really do not care if it shows 200CFM from start and it really only flows 180CFM - what I need to know is if the changes I make makes a difference, that I can see from before and after flows.

 

2) Why are engine builders that get their hands dirty bad engine builders in your eyes? I have always done my own engines, and so on the price/performance ratio has been pretty good. Sometimes you need to try different approaches to get experience.. I do not suppose your first head performed super duper - you must also have spent some time with trial and error in the beginning?

 

3) I still do not understand why you not are willing to share flow data with others? Guy Croft does it in his forum (along with very good high-res pictures), I really can't see why you do not want to share your knowledge with others? What do you have to loose if all other engine builders do not understand the maths behind your work?

[Sandy 191]

Similarly it's interesting seeing how stubborn you are, how you think anyone who doesn't share your views is 'getting at you' and how instead of trying to learn from other people's knowledge you keep your head resolutely buried in the sand. There's one thing that stays true throughout all our lives. No matter how much you know or think you know there's always someone out there who knows more. When you stop being willing to learn from them is when you stagnate. There are things I do with engines, and have done for many years, because I think it's the best way. However if someone were to prove to me a better way I'd change in the blink of an eye.

I'm stubborn about cutting through the numbers game sales pitch. The more informed people can see through it, but the less informed are vulnerable to being unduly impressed by the appliance of science. You are apparently suggesting that I am of a closed mind? You often seem to resort to the suggestion that i'm paranoid or have fixed ideas; I can't help thinking that's a reflection of your own insecurities? Just as likely as being mine!

I admit I do find your self promotion quite nauseating and feel compelled to resist it. I've just as much right to do that as you have. Maybe that's grounds to sue me?

 

The results are all that matters and how you get there is up to you.