Building a 10,000 rpm NA engine.

[bazza]

Originally Posted by oab_au

The way that the engine is developed depends on, what work it has to do. So the torque has to be developed for the rpm range that the engine will run, in the car, running the type of courses that he wants the car to do.
I know that it will be fitted to Tony’s ‘Targa Tasmanian Rally’ car, to do road rally courses, also to do some track work.

What rpm range that we use is really set by the gearing, No point in developing the torque at 8000 if the gearing limits you to running between 4000 and 6000. If it were for an open wheeler, we’d certainly find the gearing.

I'm trying to convince Tony to put the close ratio DCCD box in the racecar. These are the boxes found in 02-05 STI and will hopefully allow him to use all the gears.

With this in mind I think that if we set the torque to center at 6000?, we can then stretch the torque around between 5500 and 6500. This should be a good comprise for the two types of work. We can do this with adjustable inlet cam gears, and changing the ram tube lengths, to vary the resonate timings. Moving the torque peak, down for rallies, and up for race.

I think swappable ram tubes is a great idea. For high speed stuff like Phillip Island you want peak power and lots of RPM. For medium speed stuff like Winton / Targa tracks you will probably want a fair whack of peak power but also a heap of torque to drop back into to throw the car out of the slow hairpin corners. For either setup you want a flat torque plot which is obviously an increasing power plot to redline.

It will run E85 so we need to get the compression ratio as high as it will allow, and having less overlap will make that a bit easier. I would not like to take too much off the head as it may give sealing problems in the future, piston can do the rest of the filling. The oil system will be ‘World class’ when Tony’s finished, won’t suffer from sump surge, or heat.

We have three different zones that the engine has been used. Matt’s dirt rally car, Bazza’s tar car, and Dan’s road race car. I would like to hear from the drivers on how they found the engine to preform in there type of use.

Dan’s
. ‘Shotgunslade’ Dan’s car was road raced for a couple of years, with a 5 speed box and the first Delta cam, 8mm lift, 247* with 10* on the closing. The old engine lived through 23 track days, each of which had at least an hour of track time, some, as much as 2 hours. I routinely shifted about 7200 rpm.

The second engine that Y Tom did, had unknown timing/duration, just Big. With 11mm lift?
http://i203.photobucket.com/albums/a...nslade/SVX.jpg
Dan said;
The first session was really slow as I signed up for a novice class to get an instructor because I'd been off the track for a while. Also, I wasn't routinely taking it up to 7200 rpm as I did with the old engine. The torque curve drop off is such that I was generally shifting about 6800 rpm to get back to the sweet spot. Prior to this track day, the engine was babied for break-in, never going over 6000 rpm, and usually held to about 4000.

Matt’s
.Matts first had probably standard duration with 10.5mm lift, not too sure what the actually timing was, but it looked like it had the inlet closing at 70* ABDC. Torque dropped off after 6500, probably held down by the inlet tract length.
http://www.subaru-svx.net/forum/atta...9&d=1181993643
What Matt said about the timing.
Factory duration at 0.050" is 190 ish...can't recall exactly, but we did map it.
Mine are 230 degrees at 0.050"
The factory inlet opening point in the manual of 4* is *assumed* to be seat to seat duration, not 0.050". If you do the maths there, it becomes a seat to seat duration of 236*.
Given that, we picked 2.5* inlet opening, but at 0.050"...so to compare properly, it would be (very roughly) about 28* inlet opening and about 70 something degrees closing.
Does that make sense?
I think we aimed for a lobe separation angle of 112 degrees...versus factory of 114.
Of course we have no idea what was actually installed!!!

The later one was a standard engine had headers, and throttle bodies with the tract length set at 30cm. This one revved to 7500.

From all this info, I think if we can get the timing, and lift that we want, and set the inlet length to suit the higher rpms we stand a good chance of ending close to the best set-up, that we can be put to the best use.
What do you all reckon?

Harvey.

A couple of things here. With the cams I think we should start fresh and ignore the current results, it would be nice to include Jack's although it's hard to get those specs. My reasoning is simply because none of the above engines show plots of them pulling anywhere near 8000 rpm (they all fall off around 5-6000 rpm) and Tony is chasing 10,000 rpm as he's mentioned. The Honda S2000 motors run massive lift (12mm+) and up to 300+ duration... are you able to put some different profiles, maybe some proven units from Kelford http://www.camshaftshop.com/products...tid=1170#specs into your computer program and post the results. It would be great to see a few off the shelf proven units. You always email me the setup file with the stock measurements and I'll assist in producing a few different setups.

Also we had some stock cams profiled at Tighe last week - they're different to the above published "stock" specs. I'm a little worried if we rely on these figures we will end up down the wrong path.

[oab_au]

Congratulations Bill.


Now you will have the time, to put it on the dyno.

Harvey.

[SVXRide]

Originally Posted by oab_au

Congratulations Bill.


Now you will have the time, to put it on the dyno.

Harvey.

Harvey,

Thanks! YT made getting the car to a dyno soon a condition of his installing the Hydra

Just need to modify the fg hood to clear the new intake.

Bill

[oab_au]

Originally Posted by bazza

I'm trying to convince Tony to put the close ratio DCCD box in the racecar. These are the boxes found in 02-05 STI and will hopefully allow him to use all the gears.

If we can get the torque curve wide enough, it won't need close ratios. As they are now, most 6 speeds go 1,3,5, or .2,4,6.

Originally Posted by bazza

I think swappable ram tubes is a great idea. For high speed stuff like Phillip Island you want peak power and lots of RPM. For medium speed stuff like Winton / Targa tracks you will probably want a fair whack of peak power but also a heap of torque to drop back into to throw the car out of the slow hairpin corners. For either setup you want a flat torque plot which is obviously an increasing power plot to redline.

The ram tube length lets us, not only to move the resonate peak to another rpm range. it also gives us the choice to have them, all the same length to max the Hp, or three different lengths to spread the power over a wider rpm range.

Originally Posted by bazza

A couple of things here. With the cams I think we should start fresh and ignore the current results, it would be nice to include Jack's although it's hard to get those specs. My reasoning is simply because none of the above engines show plots of them pulling anywhere near 8000 rpm (they all fall off around 5-6000 rpm) and Tony is chasing 10,000 rpm as he's mentioned. The Honda S2000 motors run massive lift (12mm+) and up to 300+ duration... are you able to put some different profiles, maybe some proven units from Kelford http://www.camshaftshop.com/products...tid=1170#specs into your computer program and post the results. It would be great to see a few off the shelf proven units. You always email me the setup file with the stock measurements and I'll assist in producing a few different setups.

You think we should "ignore the current results" and have another guess?
All these results give us information that tell us a lot. Dan's first had low torque but higher rpms, the second had high torque but low rpms. which was the most drivable?
Matt's first had every thing but rpms, the second had nothing but rpms. Matt was the only one to actually check the timing that the cams had. We know that it had inlet timing of 20*/70*. The 20* opening was not needed, but the 70* closing was right on the money. If the opening was knocked back to say 10* or less, and the throttle bodies were used to raise the resonate peak, I am sure it would give a good wide torque curve at a higher rpm, to produce the HP.

Originally Posted by bazza

Also we had some stock cams profiled at Tighe last week - they're different to the above published "stock" specs. I'm a little worried if we rely on these figures we will end up down the wrong path.

Strangely enough I have been trying to get some one, to put a degree wheel on and measure the standard timing, to be sure what it is.

Harvey.

[dynomatt]

Bazza,

When Tony gets his 10,000rpm engine going, the 6 speed DCCD box will be useless. He'll want much closer ratios...heaps closer. Theoretical top speed in 6th gear would be 360kmh. In reality, he'll max out in 4th at best. Carrying an extra 2 ratios and 60kg of weight for nothing.

i'm thinking he'll be wanting something like a 2.5 or 2.2:1 first gear and maybe stop at an underdriven 5th or 6th ( like 1.2) depending on what final drive he runs.

I put together some ratios for my 5 speed that better matched my proposed (at the time) 8000rpm limit.

They were:

2.67 1st
1.94 2nd
1.43 3rd
1.17 4th
1.00 5th

Even with that, at 8000 rpm in 5th the top speed was 230kmh...never going to happen in a rally.

[bazza]

Originally Posted by oab_au

If we can get the torque curve wide enough, it won't need close ratios. As they are now, most 6 speeds go 1,3,5, or .2,4,6.

You need to have close ratios for tight track and targa to help throw you out of the hairpins and very slow turns. Long ratios tend to only work at the drags or very high speed tracks like Phillip Island and would compromise the slow speed requirements of a Targa type course. Even with a wide torque curve the car will be quicker with close ratios rather than long ratios around a tight circuit. Close ratios give the driver more options coming into a corner as to what gear and RPM. A long ratio box leaves you with one available gear to use.

As for skipping gears whilst racing with a 6MT... that's crazy talk and you would go slower. Maybe take your car down to the drags and you'll see for yourself or go get a PS3 with GT5.

You think we should "ignore the current results" and have another guess?
All these results give us information that tell us a lot. Dan's first had low torque but higher rpms, the second had high torque but low rpms. which was the most drivable?
Matt's first had every thing but rpms, the second had nothing but rpms. Matt was the only one to actually check the timing that the cams had. We know that it had inlet timing of 20*/70*. The 20* opening was not needed, but the 70* closing was right on the money. If the opening was knocked back to say 10* or less, and the throttle bodies were used to raise the resonate peak, I am sure it would give a good wide torque curve at a higher rpm, to produce the HP.

Harvey.

These above results were on dynoes on different sides of the world - most know that you CANNOT use these results in any proper comparison and if you do there are major risks. For example, I've done a dyno run on one local dyno then driven to another dyno shop the next day and done another run and the results were completely different. This is something I've seen numerous times over the years when tuning. The only way to do it properly is to do runs back to back on the same dyno changing the setup as you go - then you start understanding what works and what doesn't. However even after that dynoes are only a tuning tool and you still need to hit the drags to get split point mph's and see if the setup is really working.

Now with Matt's he did a run with the stock manifolds / fancy cams then another run with the fancy manifolds / stock cams. If I understand correctly he did not get a chance to do a dyno run with the fancy cams / fancy manifolds and he did not get a chance to tune ecu tune to suit the nice setup. So again it's very risky to try and determine valve openings of all things when the car never showed it's full potential - I think with the right tune and everything sorted it would've taken off.

Dan's made power to 5800 rpm... still a long way off 7000 rpm and it's very hard to form proper theories based on the driver report.

I posted this in the other thread - very good info. I'd be far more inclined to use a camshaft recomendation from these guys rather than blindly guess what "might" work from the above results. I'm the type of person who needs to see proven results rather than guess.
http://www.tighecams.com.au/helpful-hints.htm
http://kelfordcams.com/applications.aspx

Originally Posted by dynomatt

Bazza,

When Tony gets his 10,000rpm engine going, the 6 speed DCCD box will be useless. He'll want much closer ratios...heaps closer. Theoretical top speed in 6th gear would be 360kmh. In reality, he'll max out in 4th at best. Carrying an extra 2 ratios and 60kg of weight for nothing.

i'm thinking he'll be wanting something like a 2.5 or 2.2:1 first gear and maybe stop at an underdriven 5th or 6th ( like 1.2) depending on what final drive he runs.

I put together some ratios for my 5 speed that better matched my proposed (at the time) 8000rpm limit.

They were:

2.67 1st
1.94 2nd
1.43 3rd
1.17 4th
1.00 5th

Even with that, at 8000 rpm in 5th the top speed was 230kmh...never going to happen in a rally.

Matt, there are 3 ratios availabe in the 6MT DCCD box (that I know of, maybe more).

3.636 2.375 1.761 1.346 1.062 0.842 (MY02-05 STI DCCD) - My current ratios.
3.636 2.375 1.761 1.346 0.971 0.756 (MY04 STI USA DCCD - maybe same as MY06 STI AUS, can't recall.
3.636 2.235 1.521 1.137 0.891 0.707 (MY08 STi DCCD onwards) - My old ratios.

Having driven all 3 whilst racing (and most of the available 5MT stock and PPG ratios) this is exactly why I swapped out the long ratio box and replaced it with the close ratio box.

[dynomatt]

Yep, the closest ratio is in the original 6 speeds...and for a 10,000rpm engine, even if it's only going to 8000 or 8500...none of them would be any good. The rev drop between changes, and the possible top speeds render them inappropriate for this engine.

Targa is somewhat different I agree...but he'll struggle to get out of 4th...why would you bother?

M

[dynomatt]

I should add, your needs, with a lower revving turbo engine are different. You have torque...us NA folk don't!

M

[oab_au]

Originally Posted by dynomatt

Yep, the closest ratio is in the original 6 speeds...and for a 10,000rpm engine, even if it's only going to 8000 or 8500...none of them would be any good. The rev drop between changes, and the possible top speeds render them inappropriate for this engine.

Targa is somewhat different I agree...but he'll struggle to get out of 4th...why would you bother?

M

Matt if this engine hits 10,000 rpm, it will be a very odd time.
We can get it to run to 10k, but it won't operate up there, for long. 10k is 25mt/sec, and we like to have 20mt/sec as the upper limit for constant operation, thats about 8k. The loads on the conrod, and piston are huge at that piston speed, for any sustained running.

The cylinder head is the limiting factor, the ports are too flat, to the chamber, and the valves need to be around 39/40mm. If you could get 37mm it there, the valves would be almost touching. Thats why you don't see 75mm stroke engines at that rpm. A shorter stroke and bigger bore gives bigger valves, so 41mm is what is used.

So as I said before, the gearing and the course will set the rpm that it operates at. What ever that is, but I guess that it will be between 4 and 8k

Harvey.

[bazza]

Originally Posted by dynomatt

I should add, your needs, with a lower revving turbo engine are different. You have torque...us NA folk don't!

M

Hahaha yes I have torque but it's not really useful. The power delivery isn't ideal at the power figure I wish, it just falls flat on it's face and doesn't rev and use the gearing effectively. The new cams should solve this and allow it to rev to 7K - I'll let you know.

[oab_au]

Originally Posted by bazza

These above results were on dynoes on different sides of the world - most know that you CANNOT use these results in any proper comparison and if you do there are major risks. For example, I've done a dyno run on one local dyno then driven to another dyno shop the next day and done another run and the results were completely different. This is something I've seen numerous times over the years when tuning. The only way to do it properly is to do runs back to back on the same dyno changing the setup as you go - then you start understanding what works and what doesn't. However even after that dynoes are only a tuning tool and you still need to hit the drags to get split point mph's and see if the setup is really working.

Well I think it is what results you are looking at. You are just looking at the numbers, you seem to rely on them, but they are unreliable.
I look at the torque line that runs across the graph, regardless of the numbers. That line is comparably between all the engines.
That torque curve, is the same as the graph, that shows the engines breathing, it shows how much air is filing the cylinder. A dip in the wave means that the breathing has reduced at that rpm. I look at the rpm to see why it has reduced, run back over it to see why, and try to improve it. It shows where the breathing drops off, and why it does. It gives the clues to what has to be done to get that line as high, and as long as we can, at the rpm that the engine has to run.

I posted this in the other thread - very good info. I'd be far more inclined to use a camshaft recomendation from these guys rather than blindly guess what "might" work from the above results. I'm the type of person who needs to see proven results rather than guess.
http://www.tighecams.com.au/helpful-hints.htm
http://kelfordcams.com/applications.aspx

"I'm the type of person who needs to see proven results"
Do you believe that they have tested these cam profiles on a EG33 with the independent inlet tracts, that we will use, to produce the torque that we need, at the rpm that we will use. I think not.

We will determine, what valve timings and lift are needed to do the work that we need. We will then know what the duration will be, then we look for a cam the is close to that, if we can't find one we have it ground.

Harvey.

[bazza]

I think we'll just have to agree to disagree. Good luck.

[SVXRide]

Bazza isn't related to Trevor, is he?

Bill

[dynomatt]

Fair suck of the sauce bottle Bill.

I've met Barry in person...and he's keen to explore the potential of an EG33. I think we should welcome new folk with open arms.

And for him to question Harvey isn't a sign of disrespect or "Trevorness" it's merely wanting to learn more and test ideas. After all, didn't some folk question the old guard who said the earth was flat?

M

[SVXRide]

Originally Posted by dynomatt

Fair suck of the sauce bottle Bill.

I've met Barry in person...and he's keen to explore the potential of an EG33. I think we should welcome new folk with open arms.

And for him to question Harvey isn't a sign of disrespect or "Trevorness" it's merely wanting to learn more and test ideas. After all, didn't some folk question the old guard who said the earth was flat?

M

Matt,

Fair enough

I was just starting to get a little concerned that the tone of the conversation was going in a bad direction -- i.e. away from reaching the objectives stated way back in post #1

Civil discourse is always a good thing in my book! Let's get back to it!

Bill