blau928

Brendan,

Depending on your ECU, I think 2 at the flange with the cone angle calculated to be at the back of the valve should be fine. The Siemens Deka injectors at 80#/hr should be good for the combined 160#/hr you wished to get..

On the other hand, I hope your fuel lines have been upgraded to garden hoses..!

BC

The flange is a done deal. It has one hole. So if I will be using the dual injector idea, the only other place will be from the intake manifold, in a less-controlled injector set simply switched on by the ECU (Probably a Vipec V88) I was concerned over that as it would reduce the control over the injectors - as I would want all 8 injector drivers sequential at the port. I only get 8 on this computer, so to trigger more injectors would be like an on-off deal. Not pretty.

Richard - The lines are all -10. The Fuel rail is -10 as well. I have been warned its simply too big. It may be, because of the volume the pump will need to fill to still create good pressure (though - the other good thing about BIG injectors is that you don't need insane pressures)

I guess we'll see. Its far enough along that I will just do 120s or 150s and then if it won't idle, and has transience problems, I will have to rethink.

dprantl

Why not run batch-fire then? You won't need nearly as large injectors.

Dan
'91 928GT S/C 475hp/460lb.ft

BC

I will always need injectors larger than most would assume to need because of the ethanol. And SINCE they need to be as large as they do for the entire rev range power band, I am sort of cornered into the sequential use anyway. Kind of like a catch-22. Large injectors because of the fuel and power level, which makes sequential usage make sense - but the sequential injection only is useful with injectors sized larger than most OEMs run.

They run sequential not for power, but for emissions and idle quality.

blau928

Hi Brendan,

I am making -10 rails as well..

I was originally going to use an HKS AIC on my dual injector setup. It can control pulsewidth as a function of RPM, and Load (boost pressure). I will not be using it any longer, as I now have a Motec. If you mount your secondaries further upstream as your hole is already set, and use the AIC, the secondaries willnot be sequential, but will still be correlated to load independent of the main ECU....

I would be happy to sell you my AIC if you're interested, as I'm not going to use it. Eventually, all the extra parts will go on fleabay, so no biggie..

Look forward to seeing your setup when ready..!

Cheers

BC

Thanks Richard. I am a ways away from that right now. Right now I am in Cam Timing hell.

blau928

Keep the faith in the valley of doubt..! I can relate, as I have een there. However, I am now closer to the otherside known as the promised land with my setup... A bit more engineering on some minor touches, then to the 5 axis cutters for feedback...

Cheers,

slate blue

Just going back to the port and the dead spots etc, where Tuomo has labeled 1 and 2 and at the bottom of the numbers there is a transition. The area around the guides should be blended so that there is no step but just a straight line until the radius for the port starts.

In respect to the guides, two things could be tried, the easiest would be to turn down the guide area that is exposed in the chamber to a V. Obviously leaving enough material to support the valve. The second design may or may not be better, Nascar guys often teardrop the guides. So you would make large from the port looking in and narrow as the air goes down to the valve. This should improve flow and the reason that happens is because that area will not be so dead.

Just try a see what is happening there, the air is approaching at say 500 FPS and it has to change direction and run into a guide, that is not going to be pretty. Also if the floor is built up like Tuomo said that will help the low lift numbers, the problem there is doing it. I am going to have a look at buying the smallest tig torch I can get but anybody that has welded aluminum will know that the arc just follows the molten pool and with restricted movement in there it will be tricky. Epoxy is then the only answer, I will be coating mine if I use it.

I had no idea that the injectors leaked and dripped. Maybe that is an old study and it doesn't happen these days? With fuel injection and what different engines want, it is true that race engines have different injector placements. The faster the engine spins the further the injector gets from the valves unless it is a DI engine.

So with a F1 engine the injectors are outside the trumpets only, the motobike injectors do have power and economy advantage of around 2% over standard injectors and I would think that emissions would improve too.

The reason the motorbike injectors have such small holes is that they or the engineers could not justify the expense of a high pressure pump and injectors. In F1 until it was capped it was rumoured that some engine makers were running up to 2000 psi, it is now capped at 1450 psi which is a bit more than our 70 psi or so. So clearly this is an important issue.

In a high performance motor bike the runners are very straight, just like an F1 engine, the F1 being the ultimate.



This is Yamaha solution, it doesn't use secondary throttles but has a basic adjustment to trumpet length. It also uses only one injector, these are the injectors I would use. Apparently it is quite hard to tune in Hi-Lo injectors, they also catch on fire quite a lot too so they don't really appeal.

With E85 I know the V8 supercar guys were worried when they changed over but apparently for nothing. They were concerned about atomization and seat wear due to fuel pooling and lubrication differences, I would have to re-read the article but it didn't cause the trouble they thought they were in for.

Greg

BC

There is a racing article on E85? Can you point me to it please?

slate blue

Brendan, here a link, issue 36 RET
http://www.epi-eng.com/epi_general_i...ack_issues.htm
Without searching for it I would have to re-read it to quantify if it is of any value to you but these engines run E85.
Greg

BC

I have been searching around for a non-biased, non-luddite based article on using E85 for POWER. Thanks

slate blue

Guys, I thought I would share a response to a PM I sent to a very knowledgeable fellow, he was building 700 hp engines when they were not common, a real engineering type, somebody I learnt a lot off quite quickly. I deleted his email address for privacy reasons.

"Hi Greg,

Sorry for the long delay. Started a new job recently, been quite busy, missed your PM.

I discussed a Windsor 366ci engine a friend of mine built recently for a GT40 circuit racer with you that Saturday morning at Racetech a few months ago if you remember. It had Air Flow Research 205cc SBF heads, you can check flow figures on their website. I flowed them at Racetech and found their numbers to be spot on up to 0.600" lift but backed up with major turbulence at 0.700". Just over 300cfm at 0.600" lift. However, maniflod wasn't made properly which dropped quite a bit of flow, about 10-15 cfm from memory. I measured Min CSA, was 2.23 sq" at pushrod pinch. IR induction, 14" total length from bellmouth to valve seat, 4.25 sq" entry where radius of bellmouth meets straight section of ram tube. Camshaft was 258/268 @ 0.050", about 0.670" lift. This engine made 670HP at 7000rpm and 560ft/lbs on VP MS109 fuel. This was dynoed at Kim Bakers.

I really can't see a problem with your engine making more than 700HP if camshaft and induction & exhaust diameters & lengths are chosen correctly, you certainly have the flow to support it.

I can't remember if I showed you the plaster casting I made back then. You may have seen it sitting under the flowbench at Racetech. I've been flow testing various mods. It's a 408ci Cleveland. We dynoed it 2 weeks ago. 709HP at 7200rpm. It only had about 330cfm at 0.700" lift through the manifold. 2.65 sq" min CSA in inlet, 266/270 @ 0.050" camshaft. Single plane manifold. Dynoed at Westend which gives same figures as Ned's dyno (One of Kim Bakers)

Another 427 SBC which was dynoed recently at Westend. I measured the head, 313cfm @ 0.700", very small min CSA for that capacity at 2.3 sq". Single plane manifold. Made 717HP. These were 18 degree Chev heads which are fairly high port and so can handle higher velocities before choking but had quite a big cam which also compensates somewhat for small CSA.

I use Engine Analyzer Pro, Dynomation, Pipemax and Engine Pro when designing engines. They help but I find that past exeprience (mine and that of others) is much more accurate. That is why I'm always down at Racetech flowing heads, making rubber moulds, getting camshaft, exhaust, etc specs and dyno results from every engine builder I know. The programs can lead you down the wrong path if you're not carefull.

If I can be of any assistance, let me know ( I'll check my PM's more frequently)

Can you send me all of the engine specs you have? I've dynoed enough engines over the years that I have designed with the software's assistance to have a good idea if they're giving me an accurate prediction or not. Can you send bore, stroke, lengths, diameters, flows, etc

Rgds
Alex"

Kim Baker is quite a big time engine builder here, equivalent Nascar stuff, no nonsense guy, so to speak. His dyno is tough, no fairy tale hp there. I sent Alex my 4V numbers and 2 V numbers, my 2V numbers are 322 cfm at max cam lift, the 4V cam numbers are around 370 to 375 depending on where it comes out. I do think with more R&D we can get 380 to 390 cfm though, just takes more money. Err out of my pocket!

You can probably see why I am busting at the seams to build these engines, oh well have to wait. My 2V average port size is 2.20 sq" and the 4V is 2.46 sq"

Greg

slate blue

If I get time I will post some details from an SAE paper I have that details results from tests regarding engine speed and the different results and effects they got from pulse tuning at various engines speeds. It appears if I am reading it correctly at low engine speeds you cannot achieve high volumetric efficiency or may be that should be rephrased as the highest volumetric efficiencies are achieved at very high engine speeds.

That would probably answer the question a posed to the board about what dictated volumetric efficiency. It looks like revs play an essential part. Cosworth achieved 145%, the Moto GP engines don't run the air any faster than other top engines, the pulses are stronger the faster you go, so a 8,500 rpm engine will always beat a 6,000 rpm all things being equal just by this fact it seems.

This of course has a dramatic effect on BMEP, as I have mentioned before the BTTC engines were achieving 165 hp per litre, from memory close to 17 bar, an astounding number. Camshafting does appear to play a key role.

Greg

slate blue

A further bit of empirical evidence about what can be achieved with our engines is supplied here courtesy of engine builder Michael Mount and owner Henk, this 3.2 liter build is very similar in power output to the Nissan Maxima engine.


https://rennlist.com/forums/924-931-...for-400hp.html

He achieved around 120 HP per liter, he also took on a number of challenges that have faced builders of our family of water cooled engines. He hit 8500 rpms without any trouble, he used lightweight components to ensure it all stayed together and had a advanced dry sump to ensure proper oiling during this hi rpm duty along with lowering the oil pressure to save on losses.

He turned down the crank to use the Honda journals which is something I wanted to do around 5 years ago, you need a reasonable cost machinist to make this work which shouldn't be too hard stateside.

The other thing I noticed was the stroke and rod ratio was the same as my proposed and had my crank made to those specs. Michael also used a smaller piston pin that wasn't shown in the thread but I suspect may have been a DLC coated item

The piston he used is very similar to the pistons I have, so that was also nice to see and interesting, very thin rings, he didn't disclose the ring material or coatings but this is an area of a lot of development. Along with the piston design which a study was released at a recent symposium where it showed a considerable HP increase over other designs due to lower friction and also it has lower stress levels which mean longer piston life.

I suspect it is a very light piston much like the pistons in Mike Simard's engine, similar in many ways I suspect, like comp height, bore diameter and weight. My guess is 500 grams all up, considerably different to the standard 928 piston at 760 grams. It will be interesting to see how ring life goes.

In regards to that subject the 928/944 family have quite a lot of tension on the bore and the rings are quite tall. So if the tension is reduced in proportion to the height, there is the same tension per sq mm of the ring bore interface. There is just less of it and and such the biggest area of friction in the bottom end is reduced.

This is normally done like Michael's engine in association with the improved piston design, it was not disclosed if had had made the next step in reducing losses via the use of piston guided rods, however he may have achieved some gains as those rods look quite wide at the small end in relation to the piston small end.

Michael's engine uses 41 mm intakes, mine in testing used a 41.91 mm valve that was not optimized, it was tulip design versus the nail head more common in high performance engines. With a larger bore a bigger valve would be achievable. The GT3 race engines are using 42 mm valves with a 103 bore so an increase to 105 mm in bore size would make a 43 mm valve reasonable.

That should see the flow increase to 400 cfm when you factor in a better valve design, less shrouding, a bigger valve. The other aspect I think should be addressed is the port on the head, it is very upright. It is that way because of the manifold design of the standard engine.

That is far from optimum, I believe it would be best to mill the ports so that they are at 45 degrees and facing inwards so that the intake can be straighter. If you look at the new Ferrari V12s they are angled differently. Phil Threshie tried to straighten the intake tract but by leaving the ports the way they were from the factory it was always going to be compromised.

The longer tract should allow for better optimization and higher volumetric efficiency like the 944 engine, this optimization doesn't really come into it's own until the revs get above 7,000 rpms. Even michaels engine has the compromise of a 90 degree bend in it.

The aspect where it is possible to improve is the throttle bodies, I would love to use these, this is a design used in F1 about a decade ago, it is still extremely efficient and has numerous advantages, not that many people know about this company. They are ex Lotus engineers that started this company.

http://www.atpower.co.uk/shaftless-design/4542715141

The other area is their throttle control,

http://www.atpower.co.uk/shaftless-d...-co/4542715460


Here's an example of their design along with an explanation on the efficiency.

http://www.atpower.co.uk/shaftless-d...ple/4542716158

These throttles along with a machined suitable port will give the maximum achievable airspeed and volume, as such volumetric efficiency based on other factors being tuned in correctly, i.e headers, cams etc. The best driveability also.

The other bit of 928 technology I have been pondering while I am laid up is the bottom end, I have been wondering about a stroke increase to 100 mm or 3.94" for metrically challenged. It would use a 6" rod with a 1.155" comp height piston. The issue I see and I would like to hear feedback is the fact that the piston skirt will come out of the bottom of the bore by 3 to 4 mm. The bottom of the pin is still within the bore at a depth of 140.3 mm. Now the strokers that used to use the 968 piston or 968 style piston, they have a longer skirt, my piston is 45 mm total in height, so at 100 mm stroke it is 145 mm down the bore. The 968 styled stroker has its piston and pin sticking out of the bore considerably with the skirt as shown by Errka in his video and the bottom of the pin is 3.5 mm out of the hole/bore. I would like to know how tall is a 968 piston? How tall is the skirt? That is from the oil ring to the bottom of the skirt. Normally you measure max piston diameter around the bottom of the pin or 0.500" up from the bottom of the skirt. So I am thinking that my pistons sticking 3 mm out is probably O.K but I still don"t like it but may have to lump it.

slate blue

I just measured my piston skirt and it is 26 mm of useable area, not counting the chamfer near the oil control ring. So there is 22 mm of useable skirt area when the piston is at the bottom of it's stroke. There is no side loadings from the rod but there may be some change of direction forces at play. The rod to stroke ratio is slightly worse or lower than a 968 stroker engine despite the longer stroke, mine uses a 6" rod versus a 5.85" rod.

From my measurements of 968 pistons from pics on the Rennlist the usable skirt height may be the same as a 968 3.75" stroker. My guesstimate from the pic is that if the pin is 24 mm in diameter and the pin seems to go all the way to the oil rail chamfer. In which case the useable skirt area of the 968 style piston is about 20 mm or less than what the 100 mm stroker would have. So on that basis it should work.