High Flow, Low Cost, 8V Cylinder Head Project
 

I have been doing research for quite some time now on a high flow, low cost 951 cylinder solution. LR sell a stage III head that flows great but $1,619 with a core was quite a bit more than I wanted to spend. Nize did a custom high flow head a while back but the cost on that was also very high ($2,500). From my days of building Chevy's I just believed there must be a way to get a high flow solution that is much more affordable. So here is where I am at:

Bought a 951 8V head for $50 as all the valves were bent from a broken timing belt.

As I stated in another thread, through A LOT of research I found a company that manufactures SS 951 intake (and exhaust) valves in the OEM size (45mm) but also in +1 (46mm), +2 (47mm), and +3 (48mm) for $26/each! These valves have the OEM 9mm stems, are OEM length, have OEM keeper grooves, and are undercut. As many of you know 2.7L intake valves (47mm) are hard to find and cost about $130/each new. The company that sells the valves is SI Valves:

http://www.sivalves.com/index.html

The valves are not listed on their website but they have them and they have them in stock. I ordered the valves Wednesday and received them today. Here is a few pictures.

Notice the undercut to improve flow. The orginal valve was not undercut.

http://www.refresh951.com/Valve2.jpg

The original 45mm valve head (left) and the VTM 47mm head (right)

http://www.refresh951.com/Valve1.jpg

This solution will be significantly cheaper as only the intake seats will need to be replaced. The OEM guides, springs, and retainers can all be used and again the intake valves are only $26/each. If custom or Chevy valves are used custom guides would also be necessarily but in this case that is not required. I originally tried to purchase custom valves from Ferrea but they do not offer custom valves in 9mm stems as they do not have 9mm blanks. Therefore I would have been forced to go with something like 11/32 stems with custom guides which would be A LOT more expensive.

I have been doing a ton of research on porting as my plan is to do the porting on the intake ports myself. A machinist with 30 years experience also gave me a lot of good advice. I was able to purchase the necessary equipment as shown below at Harbor Freight for about $50. I will post final pictures of the porting by Sunday. The porting will be similar to the pictures shown here:

http://www.aftracingheads.com/Project9.html


http://www.refresh951.com/HeadW.jpg

Lindsey has great info and flow data for different modified heads here:

http://www.lindseyracing.com/LR/Parts/LRHEADS.html

I will be flow testing my design under the same conditions as LR to see how it compares. I should have these flow number in 1-2 weeks and will keep this thread updated.

Keep us posted.

Definitely looking forward to seeing what it does on a bench. I've got a n/a head laying around full of bent valves.. hmm...

Really been liking the new life that's been breathed into the 944 world over the last year or so, lot of great low cost stuff.

Keep up the good work. I love how there are so many people now breaking away from all of this high dollar its a "Porsche" stuff. Very cool.

sweet keep it coming sir!

definitely watching this..have a spare head upstairs...maybe you could offer a service :thumbsup:

i am so watching this thread!

Not really planning any kind of service but I do plan to document everything and will be posting it here. I really would like to contribute something useful back to the community as so many here have helped me.

Thanks. Blown944's Hybrid motor is a pretty impressive demonstration of getting around the "Porsche Tax".

I attribute a lot of the resurgence to Rogue and the M Tune. Affordable True MAF has opened up a lot of possibilities.

Thanks mate! :cheers:

are you going to do a full write up on this, because if it works out good, my dad might have his machine shop back next year and might be able to do this!

Great thread. Will find my self a NA head to hack up and install 48mm valves in :-D

Should have used a 8V NA head.

And you'll need more than 2mm larger valves to make it a good flowing head.

I made with stock valve sizes on a 2.5L engine with a ported 8V NA head 400RWHP at "only" 17psi (with the rest of the top end flowing well).

TonyG

Thanks for the input. That is impressive power for a 2.5L.

I guess good is a relative term. I want to stick with the ceramic liners. The exhaust port flows better than the intake port on the stock 951 head. I think this modified head intake will flow about 30% more at 28" H2O vacuum. I will be flow testing the head to see how it performs. This is my first attempt of probably several. Anyway, the valves are the key to this thread.

Got a few hours in my shop this afternoon and made some progress on the porting.

http://www.refresh951.com/ShopPort.jpg

Here is of picture of the tools I have been using, all from Harbor Freight.

http://www.refresh951.com/ToolsPort.jpg

Intake port before pictures.

http://www.refresh951.com/PortBefore.jpg

Intake port after pictures. (note: all valve guides will be replaced)

http://www.refresh951.com/PortAfter.jpg

are you going to polish those ports to a mirror finish?

The intake ports will be close to mirror (EDIT: smooth) but not mirror when completed.

Bit rougher is advised.

I was always under the assumption that the "rough" texture being better was kind of a myth. Air with less turbulence and the smoothest path, is best, no?

I've heard in a carburated engine, the turbulence helped mix the fuel with the air, since the fuel was mixed with the air so far away from the cylinder. In a fuel injected engine, with the fuel being injected right above the valve, air flow/speed is most important.

EDIT

This is the best I could come up with. An image of an old (No year given, but 3 valves per cylinder haven't been used since the 80's) F1 intake port. Its not "mirror", but I wouldn't call it rough either.

http://image.circletrack.com/f/14889...fold+guide.jpg

Moar edit

NASCAR "upper level" engine. Note the smoothness.

http://image.circletrack.com/f/31128...ad-gaskets.jpg

what I heard is that if you polished to a mirror finish you wont have a good atomization of the air/fuel is it a myth?

That's taken care of by the injector now days. The fuel hits the intake valve (or damn close to it). It doesn't need help of the intake, and intake ports to keep atomized.

I'm not saying that I'm right. I'm posting as an opportunity for someone to change my mind. I think the "rough" texture is a myth carried over from olden carburator days.



Granted, there is the subject of the boundary layer affecting the speed of the intake flow. However, the "fix" for this is having an impedance in the flow path of the air. I'm not sure whether there is a gain, or a drawback to that. "Roughness" isn't enough to fix it though.

Golf balls are dimpled because it disrupts the boundary layer of the air around the ball, making it fly farther. However there is a difference between air flowing AROUND something and THROUGH something.

Maybe someone should dimple their intake port?

under the same context the intake valve chamber could be mirror polished right?

I think mirror polished is extreme, but I wouldn't worry about purposely leaving it rough. Smooth it out

(Though I am anxious to see what others think about it... with some data to back it up :cheers:)

It depends. Smooth ports are great if the air doesn't need to turn. If it has to turn, it can separate very easily, creating a vena contracta (essentially a choke point from stagnant air) A rough port will disrupt the boundary layer, and prevent the air from separating as easily.

So if the port is tight and has a small radius, you don't want it to be very smooth. If it is darn near upright, and the air has to turn very little, smooth is probably better.

While the 951 has some of the best ports ever because of the shape of the head, it is best to err on the side of keeping it a tad rough. The extra flow from a perfect surface is minimal, but the penalty for separation is HUGE.

Interesting, everything I have read states with forced induction the boundary layer does not play the same role as with normally aspirated. I agree that mirror would be overkill. What would be your reasoning for a rougher surface?

Thanks Patrick. Your pic has what I would consider a smooth surface. Your pic is exactly the surface finish I an shooting for. Where did the 4 valve head come from?

nice thread. definitely love the work...

still not sure if you're really tapping into the most direct route for affordable power.

in my mind the best bang for the $$$....


1. displacement; Gen IV standard rebuilt or factory block w/4.06~8" bore w/ LS-3.662" (or larger) stroke.

2. stock LS3 heads (flow huge for low $$$$ from a discount GM parts store).

3. stock LS3 intake manifold (they flow huge for the money and sell for pennies on ebay).


simple enought to just get an LS3 crate engine... then,


4. a hotter cam and better valvesprings/pushrods, ect...


then:

5. head porting or even

6. moving up to an LS7 block w 4.125" bore and 4.0" stroke.

:banghead::banghead::banghead::banghead::banghead:

But to overcome the boundary layer, you need actual disruption of the air. A golf ball isn't just "rough". It has dimples.

The air just simply flows over a rough surface.

really ?

a friend just built a complete 379 c.i. LS engine for a bit over 5 grand by;

1. rebuilt LS2 block.

2. GM parts store.

3. auto salvage yard.

Why did you friend not buy a vette or a truck in stead ? This thread is about the 944 valves not LS engines.

his work is beautiful. and it sounds terrible to say.... but on the matter of an improved head flow via porting; (short of a full racing program) with this particular engine, you're talking about fractions that can easily be handled purely via the turbine/wastegate/tuning. anything you do to the cylinder head would make only the slightest improvement, when all you have to do is increase boost. if he want's to go past what the turbine can theoretically do on it's own, and he isn't limited to 2.5 litres, then going to V8 is the fastest and least-expensive way to more power.

Pretty cool project good job on the valve source. I agree keep the surface a little rough. The small imperfections will help with the flow. As for the LS motor option, I agree best power to cost around. But I am not going to pollute my engine bay with a non-Porsche motor.

put it this way, id rather make 400 rwhp with 18 psi then 25 psi.
less boost, more power is the way to go.
that's where head work.. getting more air though the head.
this is where the head is a bottle neck by forcing the air to go in, where opening it up is going to allow more air to flow through with the same pressure.

if that's true, then that would be sick.

i was thinking more like 350 rwhp with 21 and a half instead of 22. numbers like that.

You don't "overcome" a boundary layer. It is there any time air flows. The size of it depends on the surface texture, the velocity, and the density of the fluid (note that is relative - air relative to, say, crude petroleum - so there is no notable difference in naturally aspirated or FI applications). And to make air change directions, a thicker boundary layer is better. Disrupt was probably a poor choice, but that's the term that is often used in fluid dynamics when you are talking about making a fluid turn a corner. The typical demonstration is smoke in a chamber with a turn, which shows the vena contracta, comparied with putting a wire on the floor of the champer, which increases the size of the boundary layer in that spot, which allows the air to "stick" to the floor and avoid the vena contracta. There is a WHOLE lot more going on that "air just simply flowing over rough surfaces."

This is the guy who lectures at Ford to the guys who do port work: http://engine.osu.edu/ He is a gret professor, if you want to learn more, go read his stuff.

Interestingly, had some friends working on the new Ford 6.2L several years ago make a bet, one thinking that a dimpled surface would flow well, so they mocked one up and perform CFD. It increased drag drastically, and flowed like crap.

By the way, one needs to be careful with valve sizing. Just throwing bigger valves in something may be an old hot rod trick, but it changes your turning, and slows down the air (at a given RPM). You need to have an intake system designed to allow the engine to breathe at the right RPM for the valve size you stick in there. I'm going off of memory, but when I was messing with a different 2.5L engine (stroked Ford 2.3L), I ended up using 47mm intake valves for peak power at 6500RPM, but had to drastically shorten the intake runners to, going off of memory, 13 or so inches for peak torque near 5250.

Here's some work I did on a 3.0L BMW. Red is before, blue is after. Changed the intake and increased valve size very slightly. Point being it is a system approach. This was with minimal tuning, it now makes 358RWHP. https://www.facebook.com/media/set/?...1250448&type=3

Porsche already did that with the 924 :thumbsup:

+400
boost numbers get higher with the more work that has to be done to get the air in/out of the motor. it is not uncommon on prettymuch any forced-induction car to see boost drop with the same or better power output by opening exhaust up or cleaning intake side.

The boundary layer is always there. However, you can manipulate its size, and effects.

Yes. I realize that you can use the boundary layer to help the flow of air, rather than hinder it. You can slow the air in the inside of the curve in the port, to help the air turn. However, working in broad generalizations (since cylinder head work is one of the best kept secrets in racing, and no one here is Robert Yates) keeping a smooth port is best.

And, for the green portion of what I quotes..... OMG YES!! LOL. Fluid dynamics is unbelievably complicated, that won't be solved in an internet forum. I got interested in it from watching F1, and now its kind of a hobby. Its one of those things that, the more you learn, the more you figure out you don't know. Knowing what little I do, I know that I know very little about it.

The initial conversation was "rough" vs "smooth"....as a whole. I still say, without being Ross Braun or Robert Yates, that smooth will give greater benefit than rough...as a whole. I know, there are areas that SHOULD be rough, but by making the entire port rough, you're eliminating the advantage you would have made by only making the needed sections rough...and hurting air flow/speed as a result. Am I right in that thought?

Every racing engine I have seen had smooth ports.

I think we are way beyond crap that should be posted on a message board, and I think we are on the same page rather than different ones, semantics notwithstanding. Yes, you are right, you want the lowest loss total you can have, and that means as few rough sections as possible. For similar reasons, you actually want to shape the port differently in different sections, too.

Anyway, the conversations I'd have in a professoinal environment versus a forum are differrent. Like you said, it is extremely complicated, and people generally take a snippet and run wiht it.

As far as the race engine and its ports, that goes back to what I said in my first post - it depends. A purpose built race engine like an F1 one (have you ever gotten your hands on an F1 head!? Great googly-moogly. Was lucky to get my hands on one of the Jag ones when I was at Ford) will have very different ports than a production based race head. The more the port lays down, the rougher you need the ports.

Good for you for picking up this stuff as a hobby! I had no need for most of the stuff I learned when I was at Ford, but I was a bit like you, and wanted to, so it was initially largely self taught, albeit with some great mentors at my disposal (then I started using it when folks realized I knew what I was doing). There are far too many people, at every level, that just want a cook book, and don't want to learn the details behind why.

HA! The Wynn has a Ferrari dealer in it. They sell old, used F1 parts. They have a v10 Ferrari F1 engine for sale ($20k IIRC). I looked at it for an hour, with my small maglight, and strange looks from the people working there :thumbup:. Even that "old" technology is unbelievable.

will it fit in an NA?

I appreciate your feedback Rob. Been following closely your axle and suspension threads and always carefully consider your input. But don't worry too much for me, I do all this for FUN. For me FUN is grinding away in my shop or reading article after article about head porting. I find it to be a BLAST. Also, this head is not going on my current motor. I have a solution for the loss of torque. I in no way plan on giving up my low end performance. :D

They used to cast the F1 heads in one of Ford's labs in Windsor, CA. Can't recall the name of it, Casting Research Center, maybe? Anyway, I was doing cylinder head work back then, and had to go there for some reason, and got to, I think "grope" is probably a good word!, the head.

You could fit 2 of them in an NA. They are so small, for the power they put out.

This head in this picture was done for Alain by Parker in Canada. Anyone know anything about it. Looks like something is going on around the intake valves in the chamber but hard to tell.

http://www.refresh951.com/Parker.jpg

By the way, I should also advise you that you need to make sure those are 1 piece valves. The majority of cheap valves are 2 piece, where the head is friction welded to the stem. Not as big of a deal in an NA application, but there's a lot more heat in a boosted application.

It looks like they're seated in low

Oh just another project lying around.

It would appear that the intakes in this head are too big. The valves are almost touching the side of the combustion chambers.

Yep, that's what it looked like to me too.

Yes, it appears like a lot of shrouding, but word is this was an awesome flowing head.

https://rennlist.com/forums/944-turb...-7l-16psi.html
Still don't believe 450whp a@ 16psi but by all accounts it's a fast car.

The VTM intake valves are 1 piece, made from 21-4N Stainless Steel.

:thumbup: Good stuff!

+1, good reads. I did a Helmholtz/perforated absorber project in college. Always fun to mix my job in acoustics with my hobby of cars.

Naturally aspirated, stock bore and stroke 3.0L from a 99 328. This car:

http://www.racersedge-inc.com/racers....nsf/REBMW.jpg

How much material did you remove?

Finished cleaning up the exhaust ports. Three of the liners were off about 0.030"-.040" as show in the "before" picture below. Also cleaned things up a bit at the header interface.

http://www.refresh951.com/ExhaustPorts.jpg

The intake: I removed casting lines, defects, machining marks and smoothed the surface, generally without changing the geometry of the ports. I also port matched the intake manifold. The new seats go in tomorrow and the plan is to look at the intake ports again after for final blending.

The exhaust: I cleaned up misaligned liners as shown above and cleaned up a bit of casting defects around the inside of the header interface.

The plan is the remove about 1mm from the combustion chamber wall to un-shroud the intake using a mill.**

**NOTE: I am able to un-shroud the intake in this way as my cylinder bore is just over 102 mm.

Deleted: Reported spammer

looks like you're getting there

Yep, and with only one arm too. LOL

Thanks Sid.

Okay, I looked it up. It is an M54B30. Here is a thread Karl made when we built it: http://forums.bimmerforums.com/forum....php?t=1422245

FIFY.

Had some time to kill so I worked out a cheap way to hone the stock intake manifold. I used a $2 flexible bolt retriever (hacked up) from Harbour Frieght and a roll of 80 grit emory cloth to make a flexible "flap wheel" as shown below. It is not quick but pretty happy with the results. I have no problem running through the runners around bends. It is doing an effective job at cleaning up the surfaces.


http://www.refresh951.com/Intake.jpg


http://www.refresh951.com/FlapWheel.jpg

you sir are awesome. When I grow up I hope to be like you :D

Nice work Shawn. Interested in how this project turns out. Do you have a way to test before and after the port job?

My High Flow Head Project is complete. We used a mill to do a bit of un-shrouding of the intakes as shown below.


http://www.refresh951.com/HeadComplete2.jpg


The guy who is doing the flow testing for me on a SF600 is on vacation this week so the flow testing will be done next week. The plan is to do testing as follows:

25" (water column) at 0.472 Intake lift and 0.437 Exhaust lift to compare with some data posted by m42racer on a stock head. The testing was performed by Performance Developments.

28" (water column) at 0.480 Intake lift and 0.480 Exhaust lift to compare with some data posted by LR website for different performance heads.

I will post the flow results in this thread. I do not have all the machining costs as of yet. When I do, I will post the final cost breakdown.

Great work and thanks for the updates. Be great to test the stock intake on this head too?

Shawn try to do some lower lift numbers as well. Ie .050,.100,.250

It would be nice to see how they look with the larger valves. If you need a stock head to compare with on the same bench I can send you one.

Yes, I was planning every 0.050" just as m42racer had done. Here is his numbers on a freshened up stock head:

23.1 29.8 19.9 24.7
47.0 56.0 44.6 55.1
77.7 87.8 70.6 86.2
105.8 116.3 93.8 109.0
133.7 143.6 105.8 116.8
147.6 161.3 121.7 134.6
162.3 174.2 132.9 144.8
171.0 187.2 143.8 157.2
178.7 196.8 149.4 163.6
186.7 202.9 151.2 166.8

This was done on a SF600. I was thinking if I had it done on the same model Super Flow bench, I could use his data for comparison?

It should be relatively close. However, the humidity, and overall density of the air will change slightly. Having back to back comparisons is best IMO.

Shawn, sent you an email.

Go over on the 928 board and search: 2 V headflow figures - project update

^^Interesting read. Just getting into it, but a quick question. Am I to understand that he got 298cfm on the Intake of a 2 valve head? Is this to be directly compared to our 2V heads? How has he achieved this figure which is waay higher than our best? (Approx. 250cfm) Just doesn't seem feasible??

I shall continue reading the thread...

928 Euro S heads are probably pretty comparable to a 944 head. The fellow doing the work is a real phanatic and creditible. He had posted some pictures but it was so long ago they no longer are there. I imagine he would be more than happy to forward them as he always shared his ideas.

Slate Blue has some interesting stuff posted. Thanks for the info.

Interesting post on Pelican that he referenced:

http://forums.pelicanparts.com/911-e...-dark-art.html

I am still at a loss that this supposed 300+cfm is readily available on an 8v head?

Those numbers were at very high lift though so it seems to be an apple and oranges comparison (luckily :)).
I'm sure that it will flow more or less similar to the best heads seen around here when comparing the same lift.

Yes I saw that Gustaf but even comparing it at say .500" to the head on my 3L which was 240cfm, his is 272cfm. Is this still Apple to Oranges? Or does the different cams make comparing at same lift unreasonable? What makes it further interesting is that from what I can see on post #32 he states the standard intake at .500" flows 212cfm when our standard 8v head is cited to flow ~180cfm...so I'm sure I'm not reading this correctly. https://rennlist.com/forums/928-foru...-update-3.html
Does the 8v head on the 928 flow differently due to the V8 configuration? Also wondering what the our standard head is meant to flow on exhaust, this being the lesser number?

Hmm yes that does sound weird?

Something is either staring me in the face or it just doesn't add up? Can flow benches be subjective like dyno's? I wouldn't have thought so as there is no actual load apart from air or water being sent through at a predetermined rate. 28 somethings...

Contact him he really enjoys tecnical discussions. In the 928 world there was an enthusiasst named Bob Devore. He built a 2 valve 928 that supposedly had over 700 hp NA. He became apart of a company called Devek that pioneered large bore stroker 928's decades ago. His big motor ran 944 heads.

Yes I've heard and read of Devek. 700hp n/a is crazy stuff. What size do they stroke these motors out to?

Also, I wasn't banging on about the cfm as if it was the be all and end all. As we know velocity is just as, or more important than volume...but numbers is numbers...I still wonder what they're doing that is so different to the 944 world?? I'd like to engage Mr S. Blue in a conversation but he'd run rings around my tech knowledge. Be more than happy to have one of our more credentialed colleagues have a chat with him and watch on. Anyone care to open up the forum?

I'd like to see a flow chart for Henks 400+ hp 3.2l engine :)

Indeed!!
I sent a pm to Mr S.Blue to see if he'd engage us in a discussion. Not sure he is a super regular poster so we shall see what happens. Not that I'll understand everything but I like to learn and engender these sort of threads. Hopefully Shawn won't mind some outside interaction. :)

Ok I was asked to clarify some details of the development that was done years ago on the 928 2 valve heads. First the 944 and 928 heads are similar but not the same due to their valve spacing and combustion chamber.

The 928 head traditionally was fitted to 95 or 97 mm bore whereas the 944 head gets fitted to 100 or 104 mm bore. So the factory designers changed the valve spacing, there is extra width 3.5 mm of valve spacing. I haven't looked at this for quite some time so my memory is a bit hazy on how they moved them.

I know they moved the valves 1.5 mm and 2 mm and if I was to have a guess the exhaust was moved over 2 mm closer to the cylinder wall and the intake moved over 1.5 mm closer to the wall. So more shrouding effectively, however sometimes the way the wall can interact with the flow to assist, i.e forms part of a diffuser, not saying that is happening here though, also it will probably change with different lifts.

The flow of the standard port was done on a decarbonized port and valve with the standard 97 mm bore and I am aware others on the 928 forum have achieved similar numbers. Again from memory we then tidied up the port, different cuts on the seat etc and got around 230 cfm.

The fellow at the race shop then agreed that a larger valve may well help with the ability to shape the port as much as just making the "hole" bigger. He didn't make the port bigger at all but raised the roof quite a lot and was able to develop a better short side which we checked with a silicon mold. The valve used was the 944 2.7L 48 mm valve.

That yielded a result at peak lift of 246 cfm. I believe we used the lift figure of 0472". I was really on a learning curve at this early stage. I believe heads like these would really preform very well and judging by what the cleaned up heads did (which flowed around 230 cfm) with standard valves on a standard bottom end. The engine had excellent pulling power and drivability. It did have other mods such as an elaborate exhaust with variable flow. It also had a cleaned up plenum and intake runners.

I intended to build a separate much bigger engine and had a stroker bottom end waiting to go. The guys that were advising me were always stressing the right port sizing and saying go smaller rather than larger given the intended use.

We decided to go a larger valve, larger on both sides in fact, 2.10" and 1.65" the bore was 103 mm. I used the standard numbers as a reference. A key number was the flow per sq inch of valve area. 87 cfm was the number at 0.500" and that was my immediate goal to achieve that number with the bigger valve.

It took a lot of experimentation to get their. The best figure was 304 cfm at 0.500". However the flow tapered off above this lift, the planned lift was 0.650". That port achieved around 315 cfm at peak lift whereas another port did 292 and 325 cfm respectively. Shows how you can tailor you port to your cam.

The CSA was really too small to take advantage of all that flow, average CSA was 2.15 sq". The port couldn't be increased in size any further due to the casting. The 944 standard casting would probably be better. I did consider using the bigger heads but that never happened.

The only way I was able to achieve such numbers was to fill the floor and raise the roof. You also need a development head which you can section and port molding material to see the shape of your ports. The early ports were not filled.

The idea I had but never tested was with the 2.7L head, these heads like all of the similar series of heads starts to shroud the valve at the higher the lift. A NASCAR head is completely different, the more lift equals less shrouding. This is achieved by canting the valve and the twisted placement of the valves.

The difference is and forgetting about the cant, with a 944 or 928 head the valve starts in the centre of the bore and then migrates away from the centerline and gets closer to the bore as it decends on it's inclined angle which is 20 degrees.

Whereas a NASCAR head valve starts off the centre of the bore, (the reason for the twisted design along with a better steeper short side) and as it decends it comes to the centreline of the bore for maximum unshrouding. They then cant the valve for further bore clearance. These heads can achieve around 100 cfm at 0.500" per sq" of valve area and at 0.800" around 114 cfm.

So my idea was to move the intake valve over and away from the bore and make the bore bigger, much bigger like others have done in the 109 to 110 mm range. With a 2.20" valve and the flat floor the 2.7L port has and around 5 mm to 7 mm wall clearance, I would estimate flow up to 330 cfm at 0.500" and around 350 cfm at a peak lift of 0.650". In a big capacity high speed engine that would have made a lot of power.

Very interesting info indeed. Thanks for your input. There has been quite a discussion in this thread about intake port surface finish. From the testing and research you have done what would your thoughts be on this subject? TIA

Final cost fot the project which included used head (bent valves), new guides, new intake valves (47mm), new exhaust valves, new seals, competition valve job, deck (0.010") was $693 complete. This does not include flow testing costs.

This kind of discussion is what this thread is all about! Thanks Patrick.:thumbup:

With regards to port finish, rough on the short turn radius, I used 40 grit paper on the floor and turn in the direction of the airflow to try and create rib lets to keep the air attached over the turn. The rest of the port is fairly smooth but is much straighter than the short turn so you don't want such turbulence there.

I'll have to sign off here, I not well and that is the reason the project never proceeded and the reason I don't post much anymore, not looking for sympathy but I don't want you guys to think I am rude by not responding any further. Thanks

No worries at all. :thumbup: Thanks for taking the time to give us some feedback. Hope you feel better soon.

Yes, thanks very much for your posts. Really interesting and I'm still incredulous over the results compared to what we are used to in our world. Be interested to hear what you think about the differences the head plays between n/a and forced induction. Anyway, hope you recover asap and can get back into your projects.

Wonder what the limits are on the 16v heads?

Would be interested to know how much more the 2.7 head flows than the std head. Intake ports are larger, so surely it should flow slighly better than a std head, assuming fitting the same 48mm intake valves in both?

Honestly, it seems a cheaper overall way to go is a 944S2 engine with custom turbo pistons.

Did you flow test this yet?
Were you planning on dynoing this head before you swapped bottom ends?

Hi i've tried to contact VTM but they email seems to be broken... is there another way to contact them ?

Not yet. Been crazy busy with my build and work.

Call them -- 855-361-4000 Toll Free

Shawn. I know you have been very busy but did you ever flow this head before installing on your stroker?

Where did you get the new valve seats?

Unfortunately no :(. I have a long sad story about this but oh well....

Valve seats were provided by the shop that did the work, Goza Auto Machine in Acworth GA.

What's up ?

Update - looks like VTM no longer sells the valves but they are available from their supplier SI Valves:

http://sivalves.com/

Intake: stock, +2mm, +3mm, +4mm, and +5mm
Exhaust: stock, +1mm (non-sodium filled)

I bought some SI valves, they were very heavy and they were not true. The could have been fixed but it just isn't worth it. I don't know where the machining is done but....... anyway in the bin now.

Not good Slate. Called VTM and phone was redirected to SI. The above was story communicated to me. I bought valves from VTM a little over a year ago and was happy with them. How much heavier were they? Did you discuss the "not true" issue with them?

What's the point with discussing it with them, how long is that going to take to resolve. I am not in the US and sending them back and forth with all the time delays and costs, not worth it. They weighed around 132 grams, I think they were 52 mm and I had the identical size custom made and they came in at 99 grams.

Planning a High Flow Head 2.0, aka Ultra High Flow 8V Head :) Hoping to have a really great flowing 8V head (yet low cost) ready for the Spring. Who is using 50mm intake valves?

My thinking is that this would apply equally to the intake runners? Would this not be a downside to a extrude honed intake as the short side radii are smoothed?

I am hopeful that you are able to dyno test the extrude honed intake to let me see how much money I wasted after reading Harry's post. ;)

However, all the new composite intakes are very smooth bore on the inside and have many curves, so I am still hopeful as to positive results.

It isn't the number, it is the radius of curvature. And also you want to avoid S-curves, unless you have a straight about 4*ID long. I think 3 and especially 4 may be pretty compromised on our cars. If you have to make curves, making them in the same direction is always the way to go.

Harry:

So what is your prediction on the extrude honed intake, good, bad, no difference at all???

Eric, I think you'll be fine. Had to go back and re-read for the context. Extrude honing from my experience (which is minimal with it) is great for cleaning up cashing flash and stuff. It generally doesn't give a polished like finish. I think to get a really, really polished surface, you'd have to run multiple passes, which would remove more material at turns, and make the resutls bad. But overall, I think you'll probably see a small improvement. My "too smooth" caution was really that some folks will port and polish the heck out of intake ports, and if the short side is too tight, air will separate.

Whew!!! Thanks Harry, guess we will find out soon, Shawn has it for testing. I put the most likely too smooth LR intake on the FL car and took this one off. I just liked the way it looked, really no other reason. If the EH intake really has any benefits, I might through it on the CA car later.

What you'll find, with enough flow bench testing, is that it's all about the shape of the port rather than the surface texture. Work on the shape!

Keep in mind that head porting is really 2nd semester stuff because it's not intuitive. There are people that do nothing but cylinder heads and some of them can really boost the power. Sometimes you're way ahead if you farm some of the work out. Cylinder heads, in my opinion, fall into this category because it's so easy to go backwards.

Michael - I am sure you are right. In the end I am pretty sure I will be having you do an 8V head for me. As always, really appreciate you input.

I think I am going to spend a bit of time on this and see what I can do. No matter the outcome, I am sure I will learn a lot and to me that is the fun of it. :)

Be interested in you take on the differences between heads/cams and perhaps inlet manifolds for turbo cars vs n/a Michael?

Shawn, if you can section some heads, you'd learn a whole lot. Was able to do that with some old Ford 2.3L heads several years ago. Also, to Michael's point, he is absolutely right. There are some exceptions, basically older cast iron heads such as older American V8's. But for anything resembling a modern head (and that includes ours), tread very carefully!

Stock 2.7l head
 

Hey...since your original intention was a 2.7l head, do you know if the company mentioned (SI) makes valves that will work in stock 944 2.7l heads? I need 3 exhaust valves (40mm, 944 105 418 00) and 1 intake valve (48mm, 944 105 409 03). Looking for options here so I don't have to spend over $700 on 4 valves...:icon501:

Thanks,
Bob

Yep, call them.

Shawn, bumping this thread for info on 50mm intake valves.

Has anyone run any valves this large? I am looking into doing 50mm valves on my 2.7 head, which with the increased bore size, leaves some room to play.

https://rennlist.com/forums/928-foru...ld-test-2.html

52 mm intake valve in a 104 mm bore.

I think this confirms what Jon Milledge has said, no reason to really go more than 48mm on a 104mm bore. I also saw this on my flow bench. I think especially when you can get 48mm valves for $29/each. Just my opinion. The reward past 48mm is minimal at best. The cost goes up dramatically and the valves just get heavier. If you want much more flow than 250cfm then the 16V head starts to really make sense. In my mind the beauty of the 8V head is fairly good performance at a bargin price. If cost is not really a consideration then the 16V head is the way to go.

Well the 52 mm is flowing around 275 cfm, which is quite a lot more, the 52 mm valves I used weigh less than the 45 mm factory intakes and a lot less than the factory 48 mm intakes. The other head I have, if you read the whole thread used 48 mm in a 100 mm bore as a comparison flowed just under 250 cfm but the head was ported before and the current porter did complain about some of the workmanship.

I don't disagree that the 4 valve is much better but more complicated and the velocity is better in the 2 valve head which is the main reason I believe my 2 valve engines will make excellent power.

Clearly you are getting outstanding results. You have done a lot a great work! Just making a point about cost vs reward for the 951 turbo motor. We have now built a couple 951 heads that flow near 250cfm for $800. Cost is one of the main points of this thread.

Not sure when Jon would have said that, must've been a while ago.

48 vs 45 factory weight. Not that 'a lot'.


https://cimg0.ibsrv.net/gimg/rennlis...ae3b978d3f.jpg

I believe it was around 2007. Some of us shared some emails with Corleone and he told us that this is what JM told him. He said they discussed it extensively. He told us JM did a lot of testing looking at 48mm, 49mm, and 50mm valves and that 48mm was always better.

To reach the same inlet valve area of a stock 968 head (using 39mm intake valves) we would need a 55mm single intake valve (!). This plus JME's experience about the 48mm being the largest size that may be used suggests to me it's probably going to be even more cost-effective to build right away a 16V engine than spending a fortune on major port work, aftermarket valves & springs and a suited camshaft. Both set ups will require an aftermarket intake to make the most of the top end, too.

I am living this nightmare right now! The giant valves in my head were simply too big even with my giant bore. We all thought with the 106.5 mm bore it would work, but it didn't. At least not where I wanted it. We are starting over now and going all out. Nothing but the best going into this head now. Hopefully we can make it special and not just expensive! Just the new Titanium valves and BeCu seats cost more than most complete worked heads.

Well I presume that you are talking about a giant 2 valve engine. I have two 2.7 litre heads for the 928. These are in the process of being modified and fitted with 2.20" intake valves. The secret I believe is to move the valve's position in the bore. We made alloy plugs and welded them into the head. Then we will use the cnc to shuffle the valves across the bore to create enough room to unshroud the valves. The bore is 4.285" as such we will be looking for flow around 350 cfm. The 944 for some reason has different valve placement to the 928. The intake valve is 1.5 mm closer to the cylinder wall.

The quoted 350 cfm is at or around 0.650" lift. I have achieved 305 cfm at 0.500" lift with 2.10" valve in a 4.060" bore and got to peak flow at 315 cfm for that port. However it does depend on the port work as I did another port that flowed 292cfm at 0.500" and flowed 325 cfm at 0.650". You only have to look at Harley heads and their flow numbers to realise this is doable but you need all the key ingredients.

O.k then my 52 mm valves are a lot lighter than both at 99 grams.. :cheers:

Yes referring to a very large valve 951 head. I am not familiar with the differences in the 928 heads.

Attachment 921414

This app crashes every time I try to add a picture. This is like the 10th time I have typed this. I hate this App. The powers that be who refuse to fix this issue should be fired!

55.88mm intake valves! You moved them quite a bit! I thought mine were big, congrats!

The bore size you are referring too is very similar to mine but you are using a 53.34mm intake? How did you achieve such giant lift without coil bind? It took tons of research for Shawn to locate good valves to allow for .550 and still keep correct seat pressure. How far did you move inward??? Which exhaust valve did you chose and what kind of numbers there? Getting it in without getting out doesn't help much.

Probably Bee-Hive springs. The key for me was finding springs that allowed 0.550" lift while using the stock spring cups and retainers and adding a bit of seat pressure for $80 :) With enough money you can do a lot of amazing things.

Easy question first, the exhaust flows very well, you will need to look up some of my flow tests, I don't have them on my iPad. There can be a big difference in flow figures on the exhaust side, I normally get them to use a flow pipe which is bent 90 degrees

Well on the coil bind issue, the first set of springs I bought were for my 0.550" lift cams and they were from the Ford modular engine using beehive springs from Comp Cams. To go bigger and raise the roof enough I have welded the spring seat recesses up and will either add a sandwich plate between the head and the cambox. It would be around 10 mm thick. 5 mm for raising the roof the other 5 mm for extra travel on the springs.

I may just get custom billet cam boxes made with bigger bearing journals or modify the design somewhat to include bearing caps so that I don't need such large bearing diameters and add extra height into this cam box design. The belt length is a non issue as I am not using the factory oil pump and this allows me to gain some extra length without penalty. I realise this aspect of bespoke design is not for everybody....

The other thing I am looking at us using GT3 radiused lifters which are hydraulic and only weigh 30 grams. This way I don't need to look at constantly adjusting lash.

Before this work can be done I need to get the two engines back together that are the more simple builds and be able to put that aside as I have parts everywhere.

Very interesting stuff! Thanks for the information!

What I don't get with these head projects is why going to all the trouble with the head if these engines very rarely break the 600 whp barrier and 951 8V head with few modifications that OP did, is at 550whp at the moment.

What would be the reason for switching to 16v head on 3.0 build?

With a 16V head I could do the same power with less boost. The turbo is a great heat producer. Less boost means less heat. There are very valid reasons to go 16V. The 16V head is more knock resistant as well it is stiffer and can handle more pressure. In my mind the 8V makes a ton of sense for a street car/DE car where cost is a significant factor.

Are you sure it is stiffer? 16v is more complex structure and the more the head is complex the more it is prone to fractures.

Also 535whp is monumental for these engines, I don't know if I ever heard anybody succeed to produce reliable more regardless of head type.

Then there are non turbo exhaust valves that are a risk for catastrophic failure on high power turbo car + lack of 16V turbo pistons.

Considering all together I can not find a reason to switch to 16V. I can get 16V head for peanuts but I can not think of single real benefit that would be worth of extra complication.

maybe going one size-up on the compressor wheel will get you closer to 600whp :)

If you have the skill and intelligence (I have neither) to essentially hot rod your car on a budget like Shawn has then this is all the answer you need. The question 'why' will be answered by 'because its there'. The way that Shawn and Sid (and others) have built their cars up on a relative budget is to be applauded.

I think you're probably right in that there is little need to switch to a 16v...but I would preface that by saying 'for a roadcar'. For a track car, the 16v is definitely a better option for the facets mentioned. Some people have done this while retaining much of the stock running gear and had good results. Some have made more changes for hard core race cars. I think for 95% of us the 8v is totally fine for the street.

Voith, can you mention at least one high HP 16V engine on here that has suffered catastrophic exhaust valve failure?

What extra complication are you talking about?
Don't you think that headaches related with head work, valves, cam, etc and the cost related is not an extra complication over a stock 16V head that will breathe better right away?

If you are able to properly read through the archives on here it should come to you that there is zero reason to prefer an 8V over a 16V when building a 3L turbo engine. You may as well purchase a Golf GTD.

I did 514 rwhp at 1.1 bar of boost on pump gas using stock 968 head (stiffer springs) and stock 968 cams with Variocam. I guess that pretty much sums up the potential :thumbup:

For the block you either go Alusil or sleeve it. And there's the same options for 16v as for 8v excluding the odd Mahle 3.0l 8v batch that can be found from time to time. So it's not really a matter of less piston choices either. Sure, you need to adapt the headers and intake manifold. But considering you can leave the head and cams stock it can be done on a reasonable budget similar to a reasonable 8v route.

The low cost, high flow, 8v head in this thread is might impressive in terms of ROI. But I will never understand spending a huge amount of money on a 8v head.

8V fits together exactly like 951, there is no need to alter anything else. 16V not so much.

Ive been all smart with BMW 24V engine and its exhaust valves and I burned one valve in a way that its pieces fell in the combustion chamber, stuck between piston and cylinder and made catastrophe in my wallet. It is why all turbo cars have sodium or other means of cooling exhaust valves & seats, it is not there for nothing.

Also 16V turbo crowd is too small to draw any significant statistic from them. How many of 10 16V engines from the 3.0 list is still running?

By all means if money, expertise and learning trough failure is not a big problem, 16V is great, but since in most cases reliable 500~550whp is absolute ceiling for these engines, I do not see any benefit in all extra work and price involved.

It is also very strange JM built engines and US 944 GTR used 8V engines and not GTDs..

https://cimg6.ibsrv.net/gimg/rennlis...09741ba993.jpg



https://cimg7.ibsrv.net/gimg/rennlis...f48835416e.jpg

Been a good read so far. I have a lot more learning to do.

Here's another topic to bring up. The topic of Stainless steel vs sodium filled valves. I have done some reading on Corvette/other forums and the topic seems to be split.

Sodium filled valves should be able to transfer more heat, but some seem concerned about the heat being transferred to the guides. Stainless steel seems to be a compromise of the pros and cons of the sodium filled.

My plan is to stick with the sodium filled 951 exhaust valves and go with the stainless Intakes sourced from SI, unsure the size just yet.

School me.

Point of sodium valve is to get as much heat to guides as possible. That way seat area stays ~150°C cooler and that is a lot.

https://cimg1.ibsrv.net/gimg/rennlis...af257a6216.jpg

I understand the theory. Heat at the guides, should transfer the heat to the water/oil passages, thus being cooled. As far as I understand it. I should say, the amount of heat transferring to the guides that may cause increased wear.

Doesn't seem like an issue as plenty of stock 951's are running around without guide problems. And sodium filled valves are usually the choice of high performance engines. The cheap price of the SI valves does make the SS valves something to consider. Wondering what cons the SS valves would have to performance.

Not really... if used with a 2.7 head the 951 intake must be port-matched, and if using a 944/951 heads needs port-matching the water passage. And that's not even saying the 951 intake is in itself a poor design.

Which of the running 8V engines in the list do you base yourself on to come to the conclusion the proportion of 8V engines still running is higher than the 16V's?
In fact the folks who have built 16V engines have properly put a lot more attention to detail than some of the 8V folks who may have foolishly believed that building an 8V was going to be a lot easier/cheaper and have gone through issues along the way. At least with engines with a power output and durability worthy of a 3L capacity.

Haha, the 944 GTR head was designed in the US on blank castings coming from Germany and has nothing to do with the stock design. It's probably the only 8V head that might flow about as well as a 16V head but no stock 951 intake, headers, fuel rail, etc will fit.
JME and others on here have built 16V engines as well, FYI.

The stock 16V outlet valves may not be sodium-filled, but to destroy them you will need to run your engine at WOT for several hours.

I am not saying it is useless to source sodium-filled exhaust valves, I am saying that staying 8V instead of going 16V for a 3L build because of the intake manifold and outlet valves is wrong.

There are a lot more successful 944 8V race cars with recorded long life, than 16V.

Why mould and produce a new iteration of 8V head and spend hundred thousands of deutsche marks or $ if 16V was developed and was available at the time?

What is the reasoning for 8V with unlimited budget?

It's the proportion that matters, not the outright number.

The 944 GTR engines, which were an initiative from the US at a time when Porsche was not interested anymore in a full racing program with these cars, were running before the factory production 16V head was designed.

924 GTP = 1980 = 16V
944 GTR = 1987 = 2.5 litre, SOHC, turbocharged engine. It produces 650 h.p. at 1.8 bar. The aluminum liner-less block was specifically cast for the GTR. It has a 100 mm bore x 78.9 mm stroke.

http://motorsportsmarket.com/pages_n...il.asp?car=131

I don't think this is the place to look at history, so let's not digress any further.

If we get back on topic, Shawn has done a great work developing the standard 8V head, which is even more remarkable considering his focus on a limited budget.
However, would we get more power and reliability with less resource by using a non-worked 16V head? The answer is yes.

Historically 16v race breed head was available 5+ years before big $$ GTRs.

There is no proof 16v head is more reliable. It is an assumption.

The 16V Le Mans factory head was very fragile, they ran only 375hp with reduced boost because of the issues during the various testing sessions. It makes sense that the factory was not interested to supply private teams with non-proven parts, especially when PAG itself had no official racing program. In this context it makes also sense that PAG supplied Holbert and Andial with blank castings to let them design the GTR head by themselves.

"There is no proof 16v head is more reliable. It is an assumption"

How long is a piece of string, etc.

Powerhaus/ David Raines build reliable 16v 3.0 turbo engines 20 years ago.....and not just one.

Porsche spent months on engine dyno torturing these engines to all extents possible and decided 16V head is not good enough for turbo application.

I don't know why, but they certainly do.

im sure they spent a ton of time working out possible permutations of parts and stuff but the 8v turbo was likely a result of budget considerations (reuse a bunch of parts from the NA engines, easy enough to tweak the head casting, chamber design figured out already) along with practicality for a fast street car (the feeling of thrust is what fast car buyers want, not really every last ounce of power at the top end!)

sohc torque tendencies + long runners + small turbine wheel = mega midrange which 80s 944 buyers would have wanted. think about even today there are people buying 944s who are afraid to rev them over 5000 because they think they are delicate.

If the factory was really doing all this back to back R&D to compare the various virtues of the two head configs one can only theorise as to what their findings were?

Also, there are a lot of changes available to us now that weren't back them. Different turbos, ECUs and possibly valves. Lighter pistons and rods too. Great that we have these options available for us to go either way.

This video is interesting from a turbo engine development perspective. Not everything that gets slapped together will work. It will fail more likely than not.

Now that this thread obviously have derailed I must add my 2 cents..

There is absolutely nothing in factory development for a street car that is aimed at maximum power. Turbo was primarily added to engines to give more power without maximizing the rest of the engine.

This "16v head was not proven to be reliable for a turbo" is a bunch of BS and there is nothing that supports this. The same goes for the ceramic insert in the 951 exhaust ports that also have turned out to be done mostly for emissions to get the catalytic converter up to temp more quickly...

Comparing the heads the 16v IS indeed much better to use for more reasons than power as the 8v head has a steam pocket that can and will create issues on a high hp engine. The 16v head does not have this problem so yes, it is a better head to use. Sodium valves are nice but usually heavy. My new 16v head gets the Ferrea SuperAlloy exhaust valves which can be said to be an improved inconel valve and removes the need for sodium filled vales.

Besides the fact that the head flows much better and produces the same power at lower boost levels the pent roof design with location of spark plug is much more efficient. And you can run a little higher static compression. So if you're interested in the full aspect of things you will get out more power compared to an 8v for the same energy level you put in (fuel).

Trying to analyze what engine parts are the "best" to use by looking at different racing programs some 30 years ago is simply not the right thing to do. It doesn't say anything about anything. The same reason the 968 Turbo RS have a special 8v head with tiny ports (and only produces 305 hp). It doesn't say ANYTHING about what is best. It is always a combination of marketing aspects along with rules.

If anyone wants to discuss this further I really suggest a new thread to be started.

Maybe Shawn should separate this part to new thread.

There are so many variables that without extensive testing it is all more or less guess-work. (in my opinion)

Part of full aspect of things is also the fact that 16V combustion chamber has more extremely hot surface than 8V head since the valve head combined area is larger and stock ex. valves are hotter than sodium valves. So is this affecting knock threshold and combustion temps or not?

There is an unbelievable amount of quality disinformation about these engines out there. Proportional to that, percentage of self destructing rebuilt/reworked engines is very high. Even from seemingly very reputable rebuild shops.

I was hoping the next poster in this topic would respect Shawns thread and start a new one. I started one here:
https://rennlist.com/forums/944-turb...rbo-myths.html

Dumb question for people running larger intake valves (47 / 48 / etc mm) in a 951 motor with stock pistons. Do you need to run a thicker HG to account for the larger valve face? I wouldn't think you would but want to double check.

Also - would 47mm or 48mm be a better choice for a street car? I'm shooting for 300-350 whp. I know 45mm valves will get me there, but I want to make that power at a lower boost level.

Edit: even dumber question. I just want to confirm that this is a compatible mod for a 2.5 (stock bore diameter). I was reading in another thread that this is not the case, due to the valve radius exceeding the 100mm bore.

The 48mm valve fits in a 100mm bore just fine, but you need to open up the port and valve seat to get it to fit.

Without appropriate backup work a larger valve can actually COST flow!

Thanks for the info, the price is right for the larger valves and just in time for my head to be rebuilt at a reputable local place. Would "opening up" the port as it reaches the valve get me there? I don't really have the experience to do a full blown port job on my good 951 head (or the wallet to pay someone).

This thread was my first attempt, much better info in the second updated thread:

https://rennlist.com/forums/944-turb...t-8v-head.html

Its all a bit more complicated than most people would think. The 944 head is at a disadvantage compared to the 928 head, normally this is not the case. The 944 intake valve is positioned closer to the bore than in the 928 head. This shrouds the valve and this is quite an issue. We were getting big gains when we experimented by moving the valve away from the bore.

In my research into a truly big valve head with a big bore (109 mm) it indicated flow of around 350 cfm at high lifts. N.B I have only tested a 53.34 mm valve in a 103 mm bore and the best result is a touch over 300 cfm @ 0.500". That head had stock 928 valve position.

The big valve head will be a 55.88 mm in the 2.7 litre head and running radius hydraulic tappets with around 16 mm lift. When I extrapolate the results from previous testing adding in the movement of the valves away from the bore albeit the exhaust gets a touch closer to the bore therefor 350 cfm seems is quite reasonable.

Back to the 48 mm valve, here's our development;

https://rennlist.com/forums/928-foru...fold-test.html