Blow bench results for 2V injected heads

Any time you increase the intake valve size over stock you will adversely impact low/midrange power. If the cams are stock, then you will reduce low midrange AND have no additional top end. It is never a good idea to "improve" head flow unless other changes are made to take advantage of the change - more cam. But, of course, more cam also means less low/mid but increased top end HP. There is no magic...

 

The short turn radius in the intake port is almost non existent on the Ferrari head. I was able to increase the radius by using a larger intake valve. That brought the flow numbers up. The ports on these heads are pretty bad and there isn't a whole lot that can be done to improve the shape. My reworked Corvair heads flow considerably more air! The stock valve shape is not correct either. The exhaust side is fine as is with minor clean up and a good valve job.

You did hurt yourself since the low lift and mid lift numbers are off from stock on the intake side. A proper valve job and valve shape will improve the low lift numbers, assuming the porting hasn't been screwed up.

This is what I see for intake flow with my reworked 308 heads on my flow bench, SuperFlow 1020 ProBench @28"

.050" 33.1 cfm
.100" 65.4
.150" 98
.200" 125
.250" 144
.300" 155.7
.350" 163
.400" 168
.450" 174
.500" 177

I went with a .445" gross lift intake cam on the 3.5 liter engine. The stock 308 cams are only .300" lift as I recall. The average flow numbers are just, if not more important than the peak flow number.

Steve

 

I am not disputing your flow bench results and the fact that the Ferrari heads may not flow as well as the Corvair heads, but as you know, flow on a flow bench and performance on an engine are often quite different. I have seen many cars "slowed down" by "improved breathing" with modded cyl heads but without other appropriate changes. Yeah, they really flowed well on the bench but I'm not aware of anybody driving flow benches around.

As with any other mods, the only thing that matters is the actual performance of the car so a track or dyno is the final arbiter of what is really happening in the engine. That doesn't mean that there couldn't be some improvements in otherwise stock performance with improved cyl heads but in my experience it is very unusual that in an oem engine the heads are, by themselves, the "power restricter."

 

If air flow is increased by hogging out the ports, you are correct, the car will run worse than before. The idea is to increase the air flow while keeping the port cross section as small as possible consistent with the rpm range the engine will be running in. The engine should also be cammed along with some other changes to take advantage of the increase in air flow. I don't think anyone expects great increases in power with just head work alone.

Also, It makes no sense to increase air flow in lifts that are impossible to achieve. Good luck getting .600" valve lift on the 308 head!

You might want to check out the dyno run of Nick's 3.5 liter on Carobu's dyno: http://www.carobu.com/html/308_gtb_3_5l.html This is a street engine with 11:1 compression ratio, not a race engine. Compare the horsepower and torque to the dyno runs of the 308 racing engines on Carobu's site. The improvements came from carefully matching the components used. I have some pretty good software for this plus over thirty five years of experience. Notice that torque was substantially increased throughout the rev range. Torque is what makes the car move when you step on the gas.

I can get the Corvair head to flow over 200 cfm @.500" lift with the same size valve as used in my Ferrari 2V head by the way. That's 200 cfm through the intake that is cast integral with the head. Flow is much better with the manifold removed through individual Weber style runners.

Steve

 

The subject of flowing 2v heads and matching correct cams ect has been beat into the ground here. There are some fantastic engineers here who have reviewed methods of doing so and have made some very interesting theoretical calculations. I believe it was mk_e who calculated that reducing the size of the intake valve would yield superior flow, but I am unable to recall how many millimeters he calculated in his models. Regardless, the posts he made commenting on the subject yielded flow number far greater than that of using a larger valve would yield. Also, the Michelotto team used 208 heads for their 3L competizione engines specifically for that reason of needing a small intake and they made some fantastic and reliable race engines. They might know a thing or two.

Carobu has built some interesting engines and their dyno numbers are consistent with what others have reviewed in the past, but I have yet to see actual dyno figures for engines with smaller intake valves being used. The result is superior but I cannot be sure to what extent. The highest output a somewhat streetable 2v 308 with 308 heads naturally aspirated engine yields around 270-280bhp and a full race configuration being 310+. Michelotto claimed a bit more from their race engines and were running Bosch mechanical injection on at least one of their engines as well as I am sure their own camshaft cocktail. Mk_e seems very confident that with his calculations, a 300+bhp street engine is very possible.

Maserati was in the same 3L head flow situation as well. It had a decent stock engine but the heads did not flow enough for an agreeable racing configuration. The Ligier team used the Maserati Merak V6 engine in the JS2 to run at Le Mans ('74 '75) and instead of changing the 2v head to flow differently, Maserati fashioned a 4v head for their use and pretty much blew all other 3L engines out of the water with it yielding over 400bhp at over 10,000rpm. The engine ran far too hot but still won its class and finished 17th overall in I believe '74. Too bad this engine has not received more press as it truly is a remarkable feat.

Regarding Corvair head flow you seem to be so interested in. You can make it flow three as must as a 2v 308 head but it still will not make the same naturally aspirated power which tells you that not all power is just displacement and flow. I have yet to see a really good naturally aspirated American (and many Japanese) engine that can yield the same bhp per liter that these old Italian and German engines make. Mike996 is dead right that performance on a flow bench and real life performance scenarios are totally different.

Nick's engine is modified on several ways. His increase in torque comes from the fact that it is a 3.5L (not quite 11:1 compression btw) and he has employed ITB injection and a plethora of other arrangements to make a fine engine. My hat is off to him, but if I wanted to go that fast I would have bought a 355 or 360 which probably would have been cheaper in the long run.


J

 

Exactly right, cheaper, but in the visceral impact department, well let's just say it wouldn't be a 308.

Ciao,
George

 

My numbers were not really for a super hot build, but what would work best with the early stock cams. With those cams a 36mm intake would be about right but I've never tried to flow the heads with those valves....Steve may very well be right that ferrari buggered the ports and they just won't work with a small valve .....without some welding or epoxy anyway

The QV heads have a similar problem with the short side radius, touch it with the grind and you kill the flow....the whole bottom of the port really, but particularly in the last inch nearest the seat is a no touch zone.

 

Apparently you don't seem to understand that I am the one who did all the work developing Nick's 3.5 engine. If I tell you the compression ratio is 11:1, it is 11:1. Tate has it wrong on his site.

In any event, if you use technology from twenty to thirty years ago, you will end up with the power levels of twenty to thirty years ago. There has been tremendous progress made in camshaft designs, porting techniques, piston and ring design and on and on, that the folks who modified these engines years ago had no clue about. I remember this quote from years ago: "There is nothing in this world that is made so well that it can not be made better". What I did was use modern concepts and a clean sheet to come up with the 3.5 liter street engine which made 340 hp at the flywheel on Carobu's dyno. There is more left in that engine. It was dynoed with a stock exhaust system including muffler. If I made a nice stepped tube equal length header with Burns merge collectors, there is another twenty horsepower to be had. One exhaust lash cap was out of place during those dyno runs by the way. The shop that assembled the head screwed up. I repaired the head for Nick and hopefully he will get the engine back together shortly.

Yes, the engine does have individual throttle bodies. Mark Lewis' four liter will have a plenum style manifold by the way. A properly designed plenum manifold will make more torque and more horsepower. The individual throttle bodies look nice and are easier to adjust runner lengths. I put a 3/4" spacer between the throttle bodies and manifold. That's the reason for the TWM assembly on the 3.5.

Regarding air flow. You can say whatever you want about racing flow benches and other BS wives tales. If the heads are ported correctly the engine will be capable of making more power. I'm in this business and have been for over thirty five years. I have a lot of race winning and pole setting engines under my belt. A good rule of thumb, in a V8 engine one cfm of air flow @28" is equivalent to two horsepower. A highly tuned, high compression racing engine can make more but for an engine built for street driving, that's what you are going to see assuming modern components are used for the build. So an engine with 175 cfm air flow is going to give you 350 hp plus or minus.

Do you actually believe that the .5 liter increase in displacement is making the horsepower? If you do, you know very little about engines. A three liter using the same heads would make the same power at a higher rpm. Torque would be less down low. That .5 liter is there to increase torque down low, not peak horsepower.

I design many different engines. I am not a Corvair shop. I used the Corvair as an example because the inlet valve size is identical to what I used on the Ferrari. In fact the Corvair head is lacking as well. Air flow should be well over 200 cfm and the best of the heads I have done for my friend barely got up to 210 cfm. Now remember that the Corvair is a six cylinder engine. The Ferrari has two more cylinders drawing air. That's one reason why the Ferrari makes more power, more total air flow with 8 x 175 vs 6 x 200 or 1,400 cfm vs. 1,200 cfm. The other main reason is the Corvair is air cooled and therefore the inlet charge is much hotter than on a water cooled engine. There is a drag race 3.1 liter Corvair making 375 hp using two two barrel Holley carbs for your information. It is raced by a friend of my Corvair buddy.

I fully realize that many 308 owners do not want to hot rod their engines for whatever reason. Nick knows this too. But there are those few folks who love the body style and want to enhance their driving with modern power. Mark Eberhardt, Sean Dynes, Mark Lewis (4 liter which should be in progress again shortly) are but a few of the guys I am trying to help here. That's the reason why I even bother to post on this site. Otherwise I have better things to do with my time.

Steve

 

I’ll echo the thought that most of the effects of porting “knowledge” out in the general circulation are pure BS and add an engine with properly ported heads will outperform an engine with stock heads at every rpm and under all conditions…..the key word being properly ported as most porting I’ve seen done is anything but correct IMO

Interesting number and very close the “rule of thumb” number I’ve got for 10” flow.

I’ve been using 2.4cfm (at 10”)/hp/cyl. 175cfm at 28” is about 101.5cfm at 10”. That gives 101.5/2.4*8=338 hp vs 350 your way.

Just because I like stuff like this I plotted all the numbers converted to 10” on the same graph as my TR head at 10”. I thought I remembered you had a 44mm intake so I also plotted your head with the flow multiplied by 1.5 which should convert your numbers to about the same valve perimeter area(curtain)I have…and BAM the low flow number are exactly the same (these are the dotted orange and solid green lines). By .150 lift your head is already pulling away from mine, I think that is the 2 valves of the 4V head interfering with each other flow wise and effectively reducing the curtain area. Whatever the reason the shape of your graph is better than mine up to about .250 lift when an inefficiency starts to show on your port and the graph starts to nose over letting me catch you by .400 lift.

Another interesting graph (to me and probably only to me ) is the velocity of the air entering the cylinder at the various valve lifts. For this graph I assumed Steve had about a 44mm intake valve and about a 39mm ID seat, Aaron about 443mm/38mm and stock 42mm/37mm. I have 34mm intakes with 29mm ID seats. I guess a number up around 195 ft/sec is about perfect and higher velocities up at the higher valve lifts are much more important for cylinder filling
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Here’s another graph I found interesting. It’s flow and flow area on the same graph. The flow area is the small of the valve curtain area (valve perimeter X valve lift) or the valve seat ID. It show that as lift goes up and flow goes up the valve is actually more and more in the way until the seat ID becomes the limiting factor and then the valve begins to move back out of the way of the flow and begins to have basically no effect at all at very high lifts. This is why the velocity graph looks the way it does.
Image Unavailable, Please Login

 

highly informative. so should we do our best to locate a set of 208 heads? would they be more optimal/easier to set up as opposed to making a 308 intake smaller?

 

I've got some experience with these. What works best is welding up the port, and making a D shape of the intake. That raises the bottom of the port, thus making it easier for the intake to make the turn around the valve, giving you a greater flow with much higher velocity. I've used Kinetic Analysis to port my racing ducatis which have very similar ports to the 308 Ferraris. What we found is that the flow was only using 1/2 of the valve, because of the low port floor. Raising the port 1/2" increased the flow by over 25%. Putting a bigger valve onto a standard seat, lowers the valve in the head, making the low port floor problem even worse.

Art

 

I don’t know.

There is no question in my mind that for a 300 hp engine you need 90 cfm (at 10”) which is about what a 308 head has but the valve seat ID should be no bigger than 31mm which means about a 36mm valve. For 350hp you want about a 34mm seat ID and a 40mm valve.

That’s the simple math, but if Steve says the 2V 308 head has serious problems I have to believe he’s right. I’m sure whatever the problem is it can be fixed, but whether there is an advantage to starting with the 208 castings I don’t know….maybe.

 

I suspect that welding just up near the seat is all that is required, maybe 1 – 1.5 inch up the port. A few hours on the flow bench with a scrap head, some clay and a grinder will get the answer pretty quickly.

 

I suppose the best way to experiment would undoubtedly be with a scrap 308 head with clay and a grinder. That sounds much easier than taking a 208 head and fiddling aluminum. On the other hand, using a 208 head once the solution to this flow issue is located sounds easier than building up a 308 head with material to accomodate a smaller seat.

 

Yes, if the 208 head has material in the correct places and I'm not sure it does.

 

This is all incredibly interesting, and I'd love to see some more visual examples rather than descriptions of port-shapes and sizes.

But we are still talking about the 2V FI heads, right? Correct me if I'm wrong, but isn't the biggest restriction in the chain the metering flap? How much can that flow?

The throttle body?

 

2Vi heads are exactly the same as the 2V carb heads port wise as far as I know.

On the 2Vi the FI system limits cam options and is a BIG flow restriction., next is the Cams and then the heads but better head will help even with the stock FI and cams.

On a carb engine the issue is heads, then carbs/manifolds then cams.

 

Take a look at this article: http://books.google.com/books?id=GnoUz_DStrYC&pg=PA151&lpg=PA151&dq=harley+davidson+xr+model+d+port+heads&source=bl&ots=ht7flnpc1F&sig=bNxNCvxEIgWhc86jpH6EALoVq5E&hl=en&ei=3JYdSrWuOJSQswOZ44yFCQ&sa=X&oi=book_result&ct=result&resnum=1

The guy who did that is the same guy that did my Ducatis. Worked for both bikes because of the design parameters. We found it was better to weld the ends up, and if you went through the castings, but more weld back, but then you knew the limits. My bevel drive ducati turn the fastest time through the traps at Daytona, at over 155 mph, which at it's time was well beyond what the other bevel drives were doing, and it was directly related to the d shaped ports.

Art

 

I raced an XR for years so I'm quite familar with the heads. My round port heads (done by Jerry branch) out flowed any of the oval port heads I ever came across, including the set I owned. The port shape is pretty much irrelevent as long as it's basically straight and the right cross-sectional area. What really matters is the last inch of so.

 

m ke last time we discussed this wasn't there a piece of a scrap head on it's way to you ?? Once we have achived this ! bending your arm to put it on the flow bench will be quite easy !!! You just love number crunching . Thanks for all your input .

 

The head piece is Wil's and I'd bet it's still at Vic's shop so , we'd need to ask Wil if we can borrow and we'd need to get Vic to ship it which Wil has been waiting for for 2 years now....but maybe.

 

The flow that matters the most is the top 1/4 to 1/3 of the valve lift so you probably haven't done a bad thing....it would be a much better thing with a pair of early carb intake cams added....and better yet if you add EF to the "i" part of 2Vi