Race motors, Steve, Henry and the lot

[Porschekid962]

So I have been thinking about various engine configurations as of late. All air cooled, all rather towards the ragged edge but with the goal to keep things streetable and long lived.

I asked not too long ago about people running engines with other than stock rod to bore ratio. No answers. How about thinner valve stems? What about custom valve rockers?

I am just wondering to what lengths people have gone, what suppliers and machiners they have used to get there and what the results have been.

Envision a 3.5 litre short stroke air cooled two valve motor, twin plugged, high compression, aftermarket parts, ITB's, bla bla bla. Doesnt sound too out of range.

Ok so how about this:

3.5 litre motor, 3.0 78sc case, 3.2 crank, lightened and knife edged, 996gt3r oil pump, lightweight flywheel and clutchpack. 3.2 litre heads ported and flowed, larger valves, larger ports, smaller valve stems, racing valve guides, racing springs, Ti retainers, 3D valve grind, racing cams. Thats just the heads.

Smaller rod journals on the crank, CP/Pankl rods with Pankl pistons in LN bores, possibly DLC coating on valve stems, smaller valve stems, custom rocker arms, smaller piston pins, smaller main bearings, dlc coated main bearings, one off intake system with itb's and resonance chamber, equal length stainless header, all the bells and whistles of electronics, coil on plug, darn close to direct injection fuel injection, closed loop engine control, all the fun stuff.

So has anyone done this on a street motor? A non race, non turbo car that runs on the street? I am hoping for a motor that runs strong for a long time and makes plenty of power. The great thing about it is if and when I get tired of NA power the option of turbo's is always there.

So what do you all think? What have you done? Whose parts do you use? What are the secrets that no one is talking about?

[Henry Schmidt]

Custom rockers are the only thing you have listed that is out of the ordinary. BurnBros is looking at a roller rocker design but I don't know of anyone making them yet.
7mm valve stems are surfacing more and more as of late.
We've built a couple of sets these heads and they show promise.
The engine you describe sound very cool all be it very expensive.
Keep us posted on your progress.

[Steve@Rennsport]

Another engine builder I know did extensive engineering work on the possibility of roller rockers and came to the conclusion that due to the position of the cams and rocker shaft bores, there was insufficient room and these were not feasible.

We've done quite a number of things you mentioned and the end result is the combination of head flow, camshaft profiles, intake system and exhaust configuration. In these instances, these were all race engines that ran very well,............

As Henry said, such engine projects are decidedly not cheap to do and require a realistic budget to make big power and remain somewhat streetable.

[cnavarro]

I actually have a lot of customers who go "all out" on their type 4 builds, twin plug, fi, coil packs, etc... they spend 20k+, some spending over 30k on their engines with all kinds of specialty coatings, like the DLC, on every part that moves, TI rods, billet cranks, obviously our Nickies cylinders,...




One such engine is a 2.4L done by Aircooled Technology last year(from the above pictures came from for this engine), that make 257hp n/a @ 8000 rpm, for an aluminum bodied coach built spyder replica which was run at LeMans this year.

http://www.shoptalkforums.com/viewtopic.php?t=100644




Best of all, head temps were under 300F with a full load, so by keeping it cool, it should last a long time.

There is no reason why this couldn't or hasn't been done on a 911, except that these projects are time hogs and cost significantly more to eek out every last HP at that bleeding edge.

[cnavarro]

As far as 7mm valves, we've been using them for years on the type 4 stuff and even have them for our 356 heads. The weight savings are great, but the ports don't flow any better than with an 8mm valve. We've done all the research in conjunction with a few of the shops I work with.

As far as rockers, they aren't hard to make, it's just how many compromises must be made and are tolerable in a final design to make them fit. For example, we've been approached to do them for the type 4, and they only way I'll make them is to have them to have oilers like the factory forged rockers, which are pressurized through the pushrod through the rocker to the valve stem, retainer, and spring, providing cooling and lubrication. To do so would make the rockers way too expensive to make and thus, to sell.

Equal length headers for a 911, properly tuned would be awesome. Even for the 356 they would be great. We've had them for the type 4 with a true 4-2-1 equal length header since 2000. Pricy, but on an 1.8L F-production race engine they are worth 20-25hp.

I would really like to see direct injection brought to an aircooled engine. I guess eventually Porsche will do it on their newer water cooled flat-6's and maybe we can pawn some of the technology from there.

[cstreit]

How much less do the 7's weigh compared to the 8's?

[cnavarro]

The only numbers I have in front of me were for the last set of 356 heads we had prepped and the 40 mm intake valves (7mm) were 24 grams lighter than the stock ones and the 35 mm exhaust valves (8 mm) were 14 grams lighter than stock.

The stock valves had 10mm stems. Here is the flow data for the exhaust port with 8mm (No. 2) vs. 10mm (No. 1) stems:

CCFM DEPRESSION
Lift No. 1 No. 2 No. 1 No. 2
-------- ------- ------- ------- -------
0.0500 22.6 24.9 28.00 28.00
0.1000 44.3 46.2 28.00 28.00
0.1500 63.3 62.9 28.00 28.00
0.2000 81.7 77.6 28.00 28.00
0.2500 95.8 92.0 28.00 28.00
0.3000 109.3 106.2 28.00 28.00
0.3500 120.8 119.2 28.00 28.00
0.4000 130.0 128.0 28.00 28.00
0.4500 136.6 134.3 28.00 28.00
0.5000 140.7 140.1 28.00 28.00
0.5500 144.1 146.8 28.00 28.00
0.6000 146.2 152.7 28.00 28.00

[cnavarro]

On the flow bench, for the numbers above...

[304065]

Wow, so there was only a difference at the higher lift?

I love Excel.

[Henry Schmidt]

911 valves up through 964 are 9 mm.
The weight reduction from 9 mm to 7 mm is about 15% depending on the valve head diameter.

Real savings come from the use of titanium.

A TRW 49 mm 964 intake valve weights 118 grams
A Del West 49 mm titanium valve weighs 79 grams (30% less) with the exact same configuration. Three groove retainer and all. Add to that a ty retainer (steel 19 grams & ty is about 12 grams) and the weight reduction is substantial.

118+19+ 137 stock

79+12=91 Ty about 33% reduction

Less valve package weight means less reciprocating weight and lighter spring rate is necessary. Lighter spring means less friction, heat and wear. More horse power.

Valve on the left is standard and right is Del West ty

[kenikh]

Henry, what is the service life of a Ti valve (properly seated, of course).

[cnavarro]

Normally, Ti valves really have been limited in their life spans, measured in hours in a race engine, but with the right coating (Anatech Ltd's Casidium coating comes to mind), a normal lifespan can be achieved (with the right seats and minimum of valve spring tension). The Corvette Z06 comes to mind, as it has Ti intake valves, right from the factory.

[kenikh]

Originally posted by cnavarro
Normally, Ti valves really have been limited in their life spans, measured in hours in a race engine, but with the right coating (Anatech Ltd's Casidium coating comes to mind), a normal lifespan can be achieved (with the right seats and minimum of valve spring tension). The Corvette Z06 comes to mind, as it has Ti intake valves, right from the factory.

Pant, pant...droool...MUST KNOW MORE.

I will ahve to do some more reserach on the Vette drivetrain; I had no idea that the pushrod motor was running Ti valves.

[Henry Schmidt]

30-60 hrs = safe @ 8000 RPM

60-100 hrs = reasonable @ 7600

100-150 hrs @ any RPM = what the hell are you thinking?

Alloy 25 intake seats
Alloy 3 exhaust seats.

It's hard to find a machine shop willing to work with beryllium.

We're thinking about Amco 45. Just thinking.

[kenikh]

Originally posted by Henry Schmidt
30-60 hrs =safe @ 8000 RPM

60-100 hrs = reasonable @ 7600

100-150 hrs @ any RPM = what the hell are you thinking?

LOL

[Porschekid962]

Didnt know you dealt del west parts Henry, good to know for future reference. I just got the latest issue's of Racecar Eng. and Race Tech and each feature an in depth article of last years Cosworth CA 2.4 V8 F1 engine as used in the Williams. There are tons of detail pics and cutaway shots of the guts of the motor. Awesome to see such recent pics of an F1 motor in public. It is downright dirty engine porn, XXX hustler style.

Enough digression. After looking through all the pics and reading some books and comparing my budget (current and future) and rebuild ideas with longevity goals it seems raw Ti valves of any alloy are out of the picture. I am sure DLC coating the tips and stem would help but $$$$. I am now thinking properly sized hollow stem stainless or sodium filled valves are a much better bet for a hot street rod. I think the big place for gains in terms of weight reduction are in the rods/pistons and a race cut crank.

There really is only so much you can do with a Porsche head. I figure 6 or 7mm hollow stem stainless valves, sodium filled exhaust, Ti retainers and keepers, valves and seats cut to create more "tumble" than "dump" and carefully sizing the valves and ports to the engine size and power characteristics will pay off more than superlight short lived Ti valves.

As great as DLC coating can be I dont intend to run this motor to the stratosphere so I think Ti would gain a tiny bit in the long run it would be moot because of its serviceable lifespan.

What I am really interested in is designing a piston/rod combo with minimal spacing between the rings, thin rings, minimal distance from the pin bore to the top of the stack, a 20mm pin, x forge design, longer rod, smaller rod journals, H or I beam in steel (Ti once again losing out because of cost/lifespan factor) and hopefully rather light. I do have a rather large piston, 100+mm bore albeit for a 4valve motor that weighs less than 350grams, it does only have two rings but under 400 sounds doable for a 3ring air cooled 100mm piston doesnt it?

I dont have a scale accurate enough to weigh the rod that goes along with it but its certainly light.

As a point of reference, what does a stock size Del West Ti Porsche valve go for?

[kenikh]

I am vascillating Titanium performance. If all of these factory cars, with huge warranty liability, are using Ti rods, springs and valves, what rpecludes a 100K mile Porsche 911 motor from using it too?

But, yes, a lot of buck, with questionable bang for Ti parts...

[Porschekid962]

Well as far as I know the only production porsche using Ti rods is the GT3. There is a rebuild schedule that goes along with that motor and if I remember correctly it isnt 100k miles. The fatigue life of Ti is much shorter than steel or aluminum. Porsche race motors which utilize Ti rods and valves have a rebuild schedule of 40 hours, after 120 hours the parts are considered "houred out" regardless of spec. Yes we are talking about motors that live on the ragged edge and 40 hours is rather impressive. For myself I dont want to put an hour date on a rebuild for a street motor. I dont see a problem rebuilding a really hot street motor every 25k or 50k miles but I dont want to be replacing rods and valves at that point. If it were financially feasible to use DLC coated Ti valves that could last that long and give me the performance I want I would do it.

As a reference point, is there anyone running Ti valves in a street motor, or has anyone torn down a street motor with Ti valves and noted the wear on them? How about running pistons with two rings instead of three? Who makes custom bearings for odd sized journals with specialized coatings and does anyone do this on a street motor?

This is half wet dream, half school project and most likely going to happen in the next two years but I really want to pack modern technology into an old engine and see how it works. So far I have come up with a few ideas.....

Bespoke intake system designed by me.
Heavily researched injector location, possibly two.
Coil on plug or coil in spark boot coils.
Twin plugged with much smaller racing spark plugs.
Custom pistons with custom rods on different sized journals.
Lightened and dynamically balanced crank (3.2 i.e. 74.4mm) to match rod and pistons.
Superlight clutch and flywheel assembly.
Equal length stainless or inconel enriched stainless headers, "muffler" system attached via v bands to switch from muffled to open.
Extensively researched port and valve sizes to match powerband wishes and performance goals.
Smaller stem hollow valves.
Finned manganese bronze or beryllium copper valve guides to suit valve material.
Ti or hollow stainless valves, exhaust sodium filled, possibly dlc coated
DLC coated main and rod bearings
Thermal barrier coated piston tops, DLC or other dry film lubricated skirts.
GT3rsr oil pump
Staight cut intermediate gears
Lightweight timing gears and chain roller gears
Cams with dry film lubricant coating
Either highly modified rocker arms, custom made or DLC coated surfaces.
Ti retainers and keepers, springs matched to needs.
Custom lightweight crank pulley for efi.
Case fully polished on inside for oil shedding.
Possibly use of pump to draw air pressure out of crankcase to lower operating pressures and reduce windage losses.
Dual fuel pump system and high pressure motorsport injectors, I would love to get some 100-300bar bosch pieces but I think the TAG/McLaren 10 bar pieces will be a bit more cost effective.
Going the keep the Autronic SM4 brain for now, run closed loop with map sensing, knock sensors, cam and crank sensors, 02 sensors, tps, intake temp, oil temp, oil pressure, fuel pressure amongst other telemetry from the chassis and feed it into an AIM dash, the BOSCH one is just WAY to expensive.


So far I think the most expensive parts will be the rods, pistons and coatings. Valves and such are so prolific these days that custom sizes are not terribly expensive. The intake I am designing and fabricating myself so its just raw material cost at this point. The exhaust will be another matter.

I just wanted to know with the years of porsche building and knowledge on this sight if anyone had run custom sized journals, bearings, valve stems, rods and such. Where you got them, how did it pay off, did they wear well and so on.

Yeah im nuts but what the heck. I have to the chance to make this happen I really believe. Then a couple years after that motor is done I can always slap two turbos on it.......

[jluetjen]

Originally posted by john_cramer
Wow, so there was only a difference at the higher lift?

I think that this makes sense John. At lower lifts, the curtain area between the valve and the seat is the limiting factor. The reality of math is that the curtain area of the valve is about equal to the cross-sectional area of the port at about .5 inches of lift. Above that point, the valve will flow more then the port -- at which point increasing the cross sectional area of the port by taking material out of the valve stem makes a difference. To me it would only seem to really make sense in a race-engine that is port limited (either by rules or head design), and are allowed extreme valve lifts. MGs and Spitfires racing in SCCA production classes come to mind as examples where this might be the case. (911s and 914s on the other hand are venturi restricted in the production classes.)

Just an observation...

[304065]

Hmm, that's really interesting. Let's take my small-valve, small port engine: 39mm intake valve, 32mm port. Area of a circle is pi r squared, so the area of the 32mm port is about 804.25 square mm.

Valve curtain area is basically the circumference of the valve head multipled by the lift, is that right? So circumference of the circle is pi times diameter, or 122mm, times the lift, let's call it 11.557 (OK, .455" for the grumpy decimal-inch crowd) for a Solex cam, or roughly 1,409 square mm.

That's not a perfect formula for valve curtain area, certainly. I did some searching and used one that took into account the seat width (guessed at 5mm, it's in the white spec book) and the seat angle (45 degrees, the middle of the three-angle-grind) and got 1159 square mm, probably more accurate.

So backsolving the calc, the curtain area is equal to the cross-sectional area of the port at about 8.0mm of lift (OK, .315"). Which would imply, if the theory is correct, that past 8mm of valve lift, the port is the limiting factor, not the valve.

Now, we know that that's not the entire picture, because the Super Cup lobe has a .496" lift, has to be a reason for that. I remember you posted the EMPIRICAL head flow data in this thread:

Valve flow potential vs. port flow potential

Which sort of shows a brick wall with increasing lift that's evidently limited by the port.

My own expectation is that the steeper ramp of the modern lobe design will get the valve out of the way, allowing the port to become the limiting factor SOONER, and that this will be responsible for the performance increase. Only the dyno will tell.

I've had a lot of intelligent, experienced people question my decision to stick with the 39/35 valves and 32mm ports with an improved Solex design. Half the reason I'm sticking with it is, I want to see what happens! I can always stab the original Solex cams back in. . .