Porting School #5 Identifying Primary Restrictions

[rookie]

Originally Posted by MaxRaceSoftware

its also the same "Basic" reasons or part of the reasons why Engines
have certain RPM ranges. Example-> 500 CID ProStock Engine,
they can make around 820+ Peak TQ at 7800-8000 RPMs,

So giving this example, on a 500 CID ProStock Engine if you reduce only low lift flow by 20% up to say .250 lift but keep all other things the same, how much horse power will you gain?

[MaxRaceSoftware]

Originally Posted by rookie

So giving this example, on a 500 CID ProStock Engine if you reduce only low lift flow by 20% up to say .250 lift but keep all other things the same, how much horse power will you gain?

You might be misunderstanding my Post ??
what i meant to point out was that increasing Low-Lift Flow
does not always equate to low-RPM Torque or HP gains !

Just reducing Low-lift Flow to gain HP ???...i don't know how much
HP you might gain, every Combination can react to that differently .

The 3 main things that could go wrong
with major increases in Low-Lift Flow are these
1- Too much Overscavenging during Overlap Period
2- Too much reverse Flow after BDC at lower RPMs
3- Loss of enough Cylinder depression (Energy potential) to create enough
System Velocities

The NHRA SS type Engines/Heads are relatively lower Compression Ratio
Engines and the RPM Ranges are also relatively lower,
so those Engines like the Intake Valve to close in the low 60's ABDC @.050"
to gain Torque from effective CR.
The Trend i see on those Engines, if there's a CSA Choke problem somewhere
they will like Heads with better Low -to- Mid Lift Flow and not as much High Lift
Flow, and things like trying Rocker Ratios changes don't show you much if any
HP differences as well as trying higher lift Cams,
and the Trend is that the amount of AirFlow the Engine is accessing
as calculated in PipeMax is lower than what the Heads Flow on a FlowBench.


If those Heads have no CSA major problem areas,
then they Trend towards liking Mid -to- high Lift Flow
and respond more to higher lift cams and rocker ratios,
more sensitive to Dyno changes.
Theses Engines will access just about all the Heads Flow on a FlowBench,
and PipeMax will then correlate with the FlowBench and the Dyno Numbers.
Example=> #041x Heads 1.940/1.500, Flow test around 250 to 270 CFM @28"
on a FlowBench ,
a SS/IA 362.7 CID @ 7200 RPM
will in PipeMax require between 250 to 270 CFM @ 28" to make
570 to 590 Peak HP @7200 RPM @600RPM/SEC Dyno Test Rate

but if there's some kind of CSA Choke problem,
the Heads might still Flow Test at between 250 to 270 CFM on a FlowBench,
but won't calculate out in PipeMax that its accessing 250-270 CFM,
but instead lower CFM Flow because its now only making in the 530 to 550 HP
Range because of CSA Choke problem.

The SS Engines can tolerate or use relatively large Overlap Periods
and tight Lobe Centers even at lower RPMs
because of smaller valve sizes and lesser Low_lift Flow

The ProStockers close the Intake Valve in the high 70's ABDC.
The ProStock 500cid Engines turn very hi RPMs and have about the same
Overlap Periods or less, than lower RPM SS type Engines.
They have huge Intake Valve Diameters and Exh sizes
and the "potential" of tremendous Low-Lift Flow,
but instead use around 55deg seat angles that kill Low-Lift Flow,
increase the Throat diameter and still maintain proper bottom angle widths
and transitions, all this results in better mid to high lift Flow.

Suppose you have the fastest ProStock Engine Combination on the Planet
and the Cyl-Head's Flow CFM and CSA is optimized to the point its on the
ragged-edge of Choke. .... your keeping Port FPS as high as possible
to create enough Ram-effect at the Intake Valve closing point
to get in the 120+ VE range.

Now you go change that Combination by increasing Low-Lift Flow,
since the Head's Port velocity is already as high as it can be without
creating too much of a HP Choke, the increase in Low-Lift now
causes a Choke problem by wasting a little more Mixture thru
overscavenging during Overlap and you've started the Mixture moving sooner
changing the end timing at the Intake Valve closing point.

In my personal opinion,
looking at Chrysler's "falsely advertized Hemi" ProStock Chamber/Valve layout
design -VS- Chevy Chamber/Valve layout design,
its the Chrysler design that has the potential to waste more Mixture out the exhaust
during Overlap because the Exhaust Valve is located off center across from
the Intake Valve creating a straighter shot out the exhaust than the Chevy layout.
So its my opinion that the Chrysler layout would not run as fast as Chevy
under all Weather Conditions all year long.

With the Chrysler layout you would tend to need a larger CSA because of potential
overscavenging, that would make it more sensitive to Tune and Weather Conditions
and probably need to work the Chrysler layout at higher RPM
most of the time than the Chevy.

[FlowSpecialist]

I've been digging through my dyno data to see if there's anything in there of relevance to this topic. I came up with two engines tested on the same rolling road and in a similar state of tune except one is a 4v and the other a 2v.

The first is a 4v Ford Zetec 1800cc with 33mm inlet valves, throttle bodies, fast road cams and standard unported head. It puts out 192 bhp or 106.7 bhp per litre and 77 ft lbs per litre.

The second is a 2v 1360cc Peugeot TU engine with 39.4mm inlet valve, throttle bodies, rally cam and fully modified head. It puts out 139 bhp or 102.2 bhp per litre and 74 ft lbs per litre.

Both engines also have very similar valve area per cc. The Zetec at 3.8 sq mm per cc and the TU at 3.6. By way of comparison a 350 Chevy with 2.02" inlet valves would only have 2.9 sq mm per cc.

Both engines produce peak power at 8000 rpm and peak torque at 6000 through 7000 rpm. In all respects they seem to be nearly identical.

Engine Comparison.....4v.....2v
BHP per litre..............107...102
Torque per litre...........77....74
Valve area per cc........3.8....3.6
Peak power rpm........8000..8000
Peak torque rpm........6/7k....6/7k

The only significant difference is going to be low lift flow because the 4v engine obviously has much more valve circumference for a given valve area than the 2v. In fact the Zetec has 0.46mm of valve circumference per cc and the TU only has 0.36mm, a difference of 28%.

So if low lift flow is a bad thing then the 4v engine in a similar state of tune will be much less tractable at low rpm? Let's see if it is. What I did was calculate the percentage of peak torque that each engine had at every rpm. The table below shows the results.

RPM.....4v.......2v
2000....65%....54%
3000....75%....69%
4000....83%....88%
5000....92%....96%
6000....99%....100%
7000...100%...100%
8000....91%.....91%

At low rpm the 4v is much more tractable. It has relatively 20% more torque at 2000 rpm and nearly 10% more at 3000 rpm. It should by rights also be better in the midrange but I do know that the 2v had a lot of development time spent on it in terms of induction and exhaust length tests and the 4v was just built and tested as it came out of the box. I suspect it needs a lot more rampipe length or a different exhaust system to give of its best. It should be well over 80 ft lbs per litre if everything was right. 77 ft lbs per litre is desperately low for a tuned 4v engine.

Still, even with its lack of development work the 4v engine with much more low and mid lift flow and about 10 to 15 degrees less camshaft duration to compensate is producing similar top end power per litre with much more low rpm power.

With a ported head and a bit more cam duration the 4v would crucify the 2v in every respect. I'm sure there's at least 20 to 30 more bhp tucked away there because the Zetec head doesn't flow that well in standard trim.

Anyway, comments invited.

Dave

[rookie]

MaxRaceSoftware,

Sorry to make you write such a long post.

My question was kind of a loaded one.

We both know that reducing air flow on a well designed combo would also reduce power, if not Pro stock and Cup teams would quit porting heads.

I find it interesting that almost every time you see the quote “Low lift flow hurts horse power” it is made by someone working with, or a follower of someone that works with heads that would consume a traditional small block Chevy head in terms of CFM.

Then the reason they give points to a cam with too much duration, overlap, wrong install centerline or a combination of the three.

I understand that some combos like SS where head, valve or lift rules dictate that the cam has to compensate for head flow would be an exception.

For me to go into anymore detail would be to make claims without #’s to back them with, and this seams to be the case with this discussion each time it comes up so I will leave it at that, I’m sure DV will cover all this in his porting school.

I think the problem is the quote it’s self, not the low lift flow.

[FlowSpecialist]

In the previous two posts I've shown how a 4v engine with the same total valve area has 1.41 times as much valve circumference and therefore low lift flow as a 2v one of the same size. I've also hopefully demonstrated that such an engine achieves a given power output with a wider power band and more tractability than its 2v counterpart and that this is due to having more flow per cc and especially more low lift flow per cc. This means the engine can use less cam duration for a given specific output and that benefits low rpm torque.

However, I hear you say, there are other factors at work in 4v engines such as flow pattern, chamber shape, tumble instead of the swirl we see in 2v designs and how can I prove that it's not these factors that are responsible for the change in the shape of the power curve. To do that I'd have to design a 2v engine with lots more low lift flow, say 1.41 times as much of it or even more, and see how it runs. That's surely impossible to do though because you can't just magically make a valve flow lots more at low lift.

Well I'm about to show you how you can indeed design a 2v engine with massively more low lift flow, in fact almost unlimited amounts of it. I'm then going to wave a magic wand and turn that 2v engine back into a 4v one with identical operating characteristics, ultimately proving that good low lift flow is not just the preserve of 4v engines and that when you have it in a 2v one you also gain both extra power and extra tractability at the same time. That in fact good low lift flow doesn't hurt low rpm power, it helps it.

I'm going to start with something akin to a big block Chevy engine. 1 litre cylinders and a single 60mm inlet valve. Simple maths tells us that this has 2827 sq mm of valve area and 188 mm of valve circumference. Relative to cylinder volume it has 2.83 sq mm of valve area per cc and 0.19 mm of valve circumference per cc. That's not a lot of valve circumference to get low lift flow through when we're trying to pack the cylinders with air at the end of the inlet stroke. Only about 8 thou of circumference per cc, not even as big as a piston ring gap. So how do I magically increase this valve circumference per cc to get more low lift flow and fill the cylinders more easily?

What I'm going to do is take the blueprint drawings for this engine and make a scale model of it with every linear dimension reduced by a factor of two. The bore halves, the stroke halves, the valve becomes a 30mm one. Cylinder volume reduces by a factor of 8 to 125 cc. Valve area reduces by a factor of four to 707cc but valve circumference only halves to 94mm. Run the calculations again and we now have 5.65 sq mm of area per cc - twice as much as before and 0.75 mm of circumference per cc - 4 times as much as before. A much bigger change than going from a 2v engine to a 4v one with the same valve area.

Now it's true that if we also scale the cam exactly then the valve lift at any point in the engine cycle also halves so the low lift flow per cc comes down to twice as much as the Chevy cylinder - the same increase as in the valve area per cc. However it's much easier to lift small valves to a high percentage of their diameter than big ones as the forces involve reduce.

What in fact we now have is one cylinder of a Honda CB500 2v motorcycle engine, or with a small margin of error my own CB550 K3. We know what sort of power curve this produces because Honda built it. Nearly 100 bhp per litre in standard form, a power curve that starts from tickover and revs to 8000 rpm. Similar specific output to a full race big block Chevy but with much more tractability and a massively wider power band. Why, because it has much more low lift flow and can get away with a much shorter cam duration. If we tune this engine with ported head, race cam, better induction and exhaust systems the power potential is nearly 100 bhp or 200 bhp per litre. A big block Chevy can't get anywhere near this.

Scale the engine down by another factor of 2 and we arrive at a 15.6cc cylinder with a 15mm valve. The valve area per cc has doubled again to 11.3 sq mm and the circumference per cc has quadrupled again to 3mm. That's now a seriously large valve gap to fill just one cc of volume through. The power potential is now up to 400 bhp per litre and peak power rpm will be somewhere around 25000. This is similar to a model aircraft engine and we can continue this process of scaling the engine down almost ad infinitum with power potential and valve area per cc rising each time.

The final part of the story is to go back to our 125cc 2v engine and turn it into a 4v one with the same flow geometry. We do that very easily by doubling the cylinder size to 250cc and adding two more valves, one inlet one exhaust, the same size as the 125cc engine's. The critical measures of valve area per cc and circumference per cc stay the same. We now have one cylinder of a modern 1 litre 4v road bike. 120 bhp or more out of the box and 200 bhp in race tune. The same brake specifics as the 125 cc 2v cylinder.

What this shows is that 4v engines don't scale with similar sized 2v ones, they actually scale with 2v engines of half the capacity. It's also why race engines ideally use lots of small cylinders instead of a few big ones for a given engine capacity. It increases the valve area per cc by a big margin but it increases the valve circumference per cc and hence low lift flow by even more.

In no case does this extra low lift flow hurt anything. It vastly improves power potential and tractability at a given power output. It just means you have to close the inlet valve earlier because the cylinders have filled sooner.

Dave

[MaxRaceSoftware]

Originally Posted by FlowSpecialist

I've been digging through my dyno data to see if there's anything in there of relevance to this topic. I came up with two engines tested on the same rolling road and in a similar state of tune except one is a 4v and the other a 2v.

The first is a 4v Ford Zetec 1800cc with 33mm inlet valves, throttle bodies, fast road cams and standard unported head. It puts out 192 bhp or 106.7 bhp per litre and 77 ft lbs per litre.

The second is a 2v 1360cc Peugeot TU engine with 39.4mm inlet valve, throttle bodies, rally cam and fully modified head. It puts out 139 bhp or 102.2 bhp per litre and 74 ft lbs per litre.

Both engines also have very similar valve area per cc. The Zetec at 3.8 sq mm per cc and the TU at 3.6. By way of comparison a 350 Chevy with 2.02" inlet valves would only have 2.9 sq mm per cc.

Both engines produce peak power at 8000 rpm and peak torque at 6000 through 7000 rpm. In all respects they seem to be nearly identical.

Engine Comparison.....4v.....2v
BHP per litre..............107...102
Torque per litre...........77....74
Valve area per cc........3.8....3.6
Peak power rpm........8000..8000
Peak torque rpm........6/7k....6/7k

The only significant difference is going to be low lift flow because the 4v engine obviously has much more valve circumference for a given valve area than the 2v. In fact the Zetec has 0.46mm of valve circumference per cc and the TU only has 0.36mm, a difference of 28%.

So if low lift flow is a bad thing then the 4v engine in a similar state of tune will be much less tractable at low rpm? Let's see if it is. What I did was calculate the percentage of peak torque that each engine had at every rpm. The table below shows the results.

RPM.....4v.......2v
2000....65%....54%
3000....75%....69%
4000....83%....88%
5000....92%....96%
6000....99%....100%
7000...100%...100%
8000....91%.....91%

At low rpm the 4v is much more tractable. It has relatively 20% more torque at 2000 rpm and nearly 10% more at 3000 rpm. It should by rights also be better in the midrange but I do know that the 2v had a lot of development time spent on it in terms of induction and exhaust length tests and the 4v was just built and tested as it came out of the box. I suspect it needs a lot more rampipe length or a different exhaust system to give of its best. It should be well over 80 ft lbs per litre if everything was right. 77 ft lbs per litre is desperately low for a tuned 4v engine.

Still, even with its lack of development work the 4v engine with much more low and mid lift flow and about 10 to 15 degrees less camshaft duration to compensate is producing similar top end power per litre with much more low rpm power.

With a ported head and a bit more cam duration the 4v would crucify the 2v in every respect. I'm sure there's at least 20 to 30 more bhp tucked away there because the Zetec head doesn't flow that well in standard trim.

Anyway, comments invited.

Dave

In no case does this extra low lift flow hurt anything. It vastly improves power potential and tractability at a given power output. It just means you have to close the inlet valve earlier because the cylinders have filled sooner.

FlowSpecialist,

almost 95+ % PerCent of Heads i Port are the 2-Valve varities
of mainly Chevrolet - Chrysler - Ford

the little amount of 4 and 5 valve Heads i've Ported are similar to your Post

i find the same results as you Posted,
i totally agree,
i've Posted a somewhat similar Post about 1 1/2 years ago on SpeedTalk Forum
on 4-Valves and Low-lift Flow and Cams.



here's something interesting ->

i've Ported a few very Hi-RPM max effort 4-Valve Heads,
and one Honda 4-Valve Head for NHRA F/D Dragster Class
which has a bunch more Valve Lift and Duration than the other
type 4 and 5 valve Heads i've Ported in the past.

This particular Honda head was already Ported and previously run

After re-Porting + Valve Job + lite Chamber Porting ....
the FlowBench CFM results were

on Intake side, .200, .300, .400 Lift Flow were same as before
but .450",.500",.550", and .600" were greatly improved

on the Exhaust side, .200, .300, were very much Lower(worse than previous)
.400" was dead even,
and at .500 and .600 Flow picked up a bunch

the DragStrip results were set the NHRA F/D Dragster Record
at 7.996 ET and then 8.120 ET
the 7.996 ET was the first time anyone's ran a 7 second ET in F/D Class

So it went from 8.30's to 7.996 to 8.12
due to mid -to- high lift Flow Number gains
seems as if the Low-Lift Flow Numbers were of no consequence or importance
that much on this Combination
( RPM Range was 10,000 to 11,600 sometimes 12,000)

Even with these Test results,
i still agree with your Posts as i've noticed the same things,
but as we/ve run more Duration + Valve Lift + RPMs
the Combinations seem to Trend towards geting as much
Mid to High Lift Flow as you can even at the expensive of Low-Lift
similar to current 2-Valve Heads in Comp or ProStock
or hi-end ET Bracket and various hi-end NOS Classes.

[MaxRaceSoftware]

[DavidVizard-GFN]

Originally Posted by FlowSpecialist

-------------
In no case does this extra low lift flow hurt anything. It vastly improves power potential and tractability at a given power output. It just means you have to close the inlet valve earlier because the cylinders have filled sooner.

Dave

Well Mr Flow Specialist - talk about driving in through the back door of the garage here. I have never heard an argument for low lift flow the likes of this! The logic is really hard to argue with and as far as several reads of your post goes all seems to make good sense.

I fear however that you are going to get arguments along the lines of: 'the dyno tells a different story' but you are not going to get it from me - my dyno testing looks to support what you are saying.

However I find myself once again in a position were an outstanding post calls for an article from me to make a possible point that can influence subsequent readers posts. It's 8:18 right now so lets see what I can do in the next couple of hours - it's now just past 10.30 and you can check out the story at:- Dyno's Don't Lie - but sometimes it looks like they do -

DV

[rookie]

One thing I have not seen addressed in this low vs. high lift flow issue is rocker ratio, me being mathematically challenged, maybe some of you math magicians can put actual #’s to this.

Because the discussion is more towards race engines, I will use these Lunati cam specs for my question.
Intake only,
Total duration 309
.050 – 276
.0200 – 194
Lobe lift .446, with a 1.5 rocker 669, 1.65 RR it goes to a .735

As Larry and others have shown on other posts they use 1.8 RR, and some engine builder’s use as high as 2.0 ratios with even more aggressive cams.

Say this cam is installed at a 103 cl and opens at 51.5 before tdc.

How high will the valve be lifted at 55 degrees after tdc with a 1.8 RR vs. 1.5 RR.


So even though the head may not flow good at low lift, the engine sees higher flow #’s sooner in relation to degrees of crank revolution with higher ratio rockers and more aggressive cams.

I think that Nascar and Pro Stocks use of 55degree valve angles to control valve bounce has led people to look at it in terms of flow and believe it is being done to reduce low lift flow #’s and has caused people to over look other possibilities.

[rookie]

No one?

Not even a guess?