2.2 hp/cfm

[twl]

I purposely set my CSA to be give .6 mach where I want the hp to peak.

Here's something interesting, at least me. I tend to concern myself more with velocity at the seat and then taper the port from the set out so the port just is what it is. When I go back an look at my stuff while the peak in the port is .95 mach, the seat is almost exactly .6 mach at the hp peak.

I have a vintage low revving carbureted 500cc application with sidedraft carb pushrod single cylinder hemi. Some street bikes and some road racers are my customers.
I set the CSA at the carb venturi for max atomization over the jet, and then widen the port at the bowl to the large valve(valve larger than the carb).
I always used to think that this was a "backwards" port design, but it actually makes perfect sense for this kind of application, and it delivers very good flow and power like that. The carb size sets the choke point rpm for max hp, and the port is made to flow the necessary cfm to feed the demand at that rpm. This yields good power at the top, and retains excellent lower rpm and midrange rpm torque because nothing is oversized beyond the intended limits which are either mechanical limitations or desire for low rpm tractability on the street.

[CamKing]

Cam King so port flow is totally meaningless at 28" so we can use any flow we want and make any Hp we Want (of course with the right cam)!!!!!

Reading is FUNdamental.

It is meaningless to look at only the peak flow of the head, and try and calculate available HP.

Do you really think a port that maxes out at 400cfm@.900" lift has the same power potential as a head that maxes out at 400cfm@.700" lift ?

Looking at the max flow of a port is nothing but a marketing tool for head porters.

[randy331]

All from the same flow bench and same dyno.

4.030 bore x 3.75 stroke, 383 ci
10.5/1 comp, 230/234/ .590 lift hyd roller
Ported single plane, 4150 carb
565 HP / 274 cfm = 2.06 HP per cfm

4.040 bore x 3.75 stroke. 385 ci
10/1 comp, 256/270 .650/.620 lift mild solid roller
Unported single plane intake ,750 carb
609 HP / 290 cfm = 2.1 HP per cfm

4.060 bore x 3.75 stroke, 388 ci
13.2/1 comp, 259/272 .690/.640 lift solid roller
Ported single plane intake, 950 carb
650 HP / 318 cfm = 2.044 HP per cfm

4.030 bore x 3.75 stroke, 383 ci
11.5/1 comp, 258/270 .705/.640 lift solid roller
Ported single plane intake, 1050 carb
690 HP / 290 cfm = 2.38 HP per cfm

So,.. Maybe the simple HP/CFM math problem in meaningless, BUT, the reasons that one engine takes a cfm and makes 2.38 HP out of it, while a different engine can only make 2.04 out of that same 1 cfm, is anything but meaningless to me. So, I choose to track HP/CFM to look for things the good HP/CFM engines have in common. As Madbill mentioned there are no prizes for most HP per cfm, and gaining %2 power with a %10 flow increase is still a gain in power. However the best running engines, and the best running for the sum of their parts that I have seen, all make a lot of HP/cfm. Comparing the bottom 2 on my list shows %10 more flow making %6 less power. The 690 HP engine made more power from the start of the pull to the end.

So,.. WHY that happens is not meaningless to me.

Different engine on the same dyno but different flow bench.

4.155 bore x 3.940 stroke 427 ci
15.2/1 comp, 273/285 .860 lift solid roller
Ported single plane intake, 1350 carb
825 HP / 345 cfm = 2.39 HP per cfm

This engine surprised us all a little on the power/TQ it made and there it is making a lot of HP/cfm and a nice power curve to go with that 825 HP, and it is a bullet in the car.

Randy

[groberts101]

First off is my straight up disclaimer. Being just a lowly hobbyist and "making a little cash on the side" in my construction off-seasons, I certainly can't speak from all the years of work experience many here may have under their belt. Luckily for me and my occasional customers, I did receive some rather important lessons from an offshore power boat racing teams lead engine builder(numerous world records) that set me straight very early on when I started porting heads over 25 years ago. I have to say it was rather awesome to watch him hand form 15:1 piston blanks to perfectly match each specific chamber and port cylinder heads like a human machine as he talked a mile a minute without missing a beat or making mistakes. Most builders I know can usually only do one of those things rather well at a time.. lol. Anywho, it's probably all textbook to many here but his overview was this.

The rpm range/powerband doesn't completely dictate the optimum port/seat speed. The turn radii, length of throat section, and curtain area does. And as I learned later on with induction systems, so does the turn angle and CSA from manifold port into the heads inlet. In other words and comparatively speaking.. a poorly designed low angle entry port needs an overly large CSA to help slow and trick gaseous mass around corners. Whereas a more steeply inclined, or even taller entry point location designs(longer/straighter throats) can often cope with the higher mach speeds associated with smaller CSA before everything starts going south. If you're really good about an inch or so just before and after the valve?.. the port can often be made smaller than would be typically accepted as optimum sizing.

An over simplification of my process, because the bottlenecks always vary, but generally I try to first "give" a port the overall shape the casting will physically allow, and then work towards the turn and throats required CSA to get the job done within the valve size/lift(curtain) and intended rpm range. Then come even more compromises as I work the pinch point CSA based off what the turn and throat can handle. Single carb common plenum manifolds also force me to increase optimum inlet size/CSA to help make that turn into the head. On the streetable stuff I usually grind on, anything much beyond that will just raise and narrow the power band. Too often guys trade velocity and efficiency for unneeded.. and totally unusable.. cfm. I see and hear it every year in shops and on the street cars. "Huh?.. is that all they flow?.. you should get bigger heads. Hell.. even the little heads on my awesome 7,000 rpm 327 Chevy flow 325 cfm @.800 lift!". IMHO, many of the designs are oversized for their intended application. Aside from telling me where I'll end up if I keep grinding, who the hell really cares if the port goes well past .800 lift until it stalls when the cam spec's @ .550 thou?

In a nutshell. If the short turn, throat, and curtain area can handle higher main section and manifold runner speeds?.. that's usually how I bag them up. If not?.. most don't.. I just have to crutch it and blow the port out like many others seem to be doing. Of course, stroker motor popularity has changed much of the cylinder head requirements and helped many of these overblown castings to better fit into the mix. Course.. now that everything is moving towards CNC.. many are now buying overblown castings that are blown out even further. The ghost of Stroker McGurk seems to be everywhere now that he's gone digital.

For me, the main point of all this is rambling is that aside from the shortblock and induction.. HP/CFM potentially shows direct correlation to the efficiency of a ports design. And all else being equal.. from there we start cross referencing other components impact and looking at CFM/CSA. And ultimately CFM/curtain area. Discharge coefficient numbers are very very important to me and I purposely keep my ports smaller to give better bang for the CSA used. CFM under the flow curve!

I'm working on my Kasse P38's off and on lately and believe more than ever that cylinder head mfgrs are doing it backwards for the street guys. If it wasn't for capitalism/product placement, they could easily take a larger top-line design and shrink it down with great success on wider powerband setups. Not everyone needs shaft rockers either.

[LCaverly]

Mike If flow is just some arbitrary number and is meaningless then why do you request it on your cam application sheet at 28" even ??????
Randy Thanks for your reply ,you stated far better than I could all the reasons why it is good to have a bench mark to judge things by..It also would have taken me a day and a half to type that with my superior typing skills.
Len C

[LCaverly]

Brian P 1.4 cfm/hp is actual flow through the whole intake tract, but is also useful for choosing a carb. If you were to put a flow hat on the engine while testing you would see this follows very closely with the hp at any rpm unless there is something amiss with the engine.
Len C

[mk e]

If flow is just some arbitrary number and is meaningless then why do you request it on your cam application sheet at 28" even ??????

A peak flow number all by itself is kind of meaningless and partly explains why the hp/cfm range people are reporting for pretty good engines is 2.0 - 2.4 but I keep insisting 2.3 is a good number on my builds.....where I do the heads all pretty much the same just scaled to the size/rpm of the engine they're going on and know the cam duration/lift/timing match the head. hp/cfm numbers assume a lot of things about the system optimization that's been done or maybe more accurately the way the system has been designed. I like to use it to guide my builds and I look at it on others builds but at the end of the day it's the dyno curve or track results that actually matter.

[Brian P]

Brian P 1.4 cfm/hp is actual flow through the whole intake tract, but is also useful for choosing a carb. If you were to put a flow hat on the engine while testing you would see this follows very closely with the hp at any rpm unless there is something amiss with the engine.
Len C

Yes, and that agrees pretty closely with the math that I did earlier.

This "rule of thumb" may work on a carbureted pushrod V8 within a certain size range and having a carb and plenum intake design, but it will probably not work on a DOHC 4-valve motorcycle engine with an individual throttle and intake runner per cylinder. No one rates motorcycle carbs / throttle bodies in "CFM", we just use the venturi diameter (or the equivalent diameter if it is not round).

That the 1.4 cfm/hp "actual" is rather different from the 2.2 hp/cfm "nominal" is indicative that the real world flow conditions are not as they are on a flow bench. This shouldn't be surprising to anyone.

I'm not saying that the "rule of thumb" is useless if pushrod V8 / plenum / common carb arrangement is what you do, and I am certainly not saying that flow bench numbers are irrelevant, just beware that if you use this "rule of thumb" outside the range of situations that it was meant for, it may not be useful ...

[CamKing]

Mike If flow is just some arbitrary number and is meaningless then why do you request it on your cam application sheet at 28" even ??????

You seem to have a hard time understanding the written word.
Flow is not meaningless. The CFM at which a port maxes out as is meaningless, in regards to how much power that port can produce.
The flow curve of the port is anything but meaningless.

[CamKing]

BUT, the reasons that one engine takes a cfm and makes 2.38 HP out of it, while a different engine can only make 2.04 out of that same 1 cfm, is anything but meaningless to me.

So, in your mind, the CFM the engine is actually using, is directly related to the flowbench CFM measured at one static lift point ?

Please tell me how peak flow CFM is any more relative then the flowbench CFM at .300" or .400" valve lift ?

[LCaverly]

Mike Just giving you a hard time. Butttttt if an engine is designed for optimum efficiency at peak flow (at any lift) I bet you will find that it works out to about 2.2 CFM at 28" depression of maximum head flow. As i said before many engines will fall below that number,some engines will be very close and a few will exceed that number. Its all about efficient use of the flow available.
Len C

[mk e]

BUT, the reasons that one engine takes a cfm and makes 2.38 HP out of it, while a different engine can only make 2.04 out of that same 1 cfm, is anything but meaningless to me.

So, in your mind, the CFM the engine is actually using, is directly related to the flowbench CFM measured at one static lift point ?

Please tell me how peak flow CFM is any more relative then the flowbench CFM at .300" or .400" valve lift ?

All the lift's matter but the assumptions are that you chose a cam with adequate lift to work properly with the head and that the head is well optimized. If that's true the engine will choke when the mach number gets much over .6 mach and you'll make what you make.

To your point and the question of the OP the way you go from 2.2 to 2.4 is by taking advantage of flow at ALL lifts and sizing the head to be the restriction. You get lower numbers with great heads and cams by using higher flow heads to shift the peak VE point closer to the peak hp rpm yielding a maybe higher hp but probably peakier engine which might be exactly what you want.

It's just a guideline.

As an aside, the frankenferrari I'm working on is a bit of a departure for me with the head having more flow than I would normally use for the given displacement and rpm. What I'm trying to do with this one is see if with proper tuning I can convince the engine to use more everywhere. 20 years ago the first sim program I messed with had a bunch of assumptions baked in and basically sized stuff based on redline. Playing with it i tfound that if I lied about the redline and said my 8000 engine was a 10k engine then the sim predicted a lot more power across the board and I could just spin to 8k like I planned and be better off. Obviously there are problems with that thinking and that sim......but all these years later I find myself messing with the same idea. My quick math says this combo should peak at 11200ish and in dynomation I can wring 1006 @11400 out of it with the cams and head I have. But I have a 9500 redline and have been able to coax 942-970 @ 9500 out of dynomation depending on what the I tolerate in the 2500-4000 range where I'll actually drive it so I'm going with the lower 940ish setup. I don't think it''s going to deliver what dynomation is predicting because it doesn't match my normal math and it's a new version of the sim error I saw all those years ago.......but we'll see.

[randy331]

I gotta say from one construction guy/hobbyist head porter to another construction guy/hobbyist head porter, great, great post!!!!

First off is my straight up disclaimer. Being just a lowly hobbyist and "making a little cash on the side" in my construction off-seasons, I certainly can't speak from all the years of work experience many here may have under their belt. Luckily for me and my occasional customers, I did receive some rather important lessons from an offshore power boat racing teams lead engine builder(numerous world records) that set me straight very early on when I started porting heads over 25 years ago. I have to say it was rather awesome to watch him hand form 15:1 piston blanks to perfectly match each specific chamber and port cylinder heads like a human machine as he talked a mile a minute without missing a beat or making mistakes. Most builders I know can usually only do one of those things rather well at a time.. lol. Anywho, it's probably all textbook to many here but his overview was this.

The rpm range/powerband doesn't completely dictate the optimum port/seat speed. The turn radii, length of throat section, and curtain area does. And as I learned later on with induction systems, so does the turn angle and CSA from manifold port into the heads inlet. In other words and comparatively speaking.. a poorly designed low angle entry port needs an overly large CSA to help slow and trick gaseous mass around corners. Whereas a more steeply inclined, or even taller entry point location designs(longer/straighter throats) can often cope with the higher mach speeds associated with smaller CSA before everything starts going south. If you're really good about an inch or so just before and after the valve?.. the port can often be made smaller than would be typically accepted as optimum sizing.

An over simplification of my process, because the bottlenecks always vary, but generally I try to first "give" a port the overall shape the casting will physically allow, and then work towards the turn and throats required CSA to get the job done within the valve size/lift(curtain) and intended rpm range. Then come even more compromises as I work the pinch point CSA based off what the turn and throat can handle. Single carb common plenum manifolds also force me to increase optimum inlet size/CSA to help make that turn into the head. On the streetable stuff I usually grind on, anything much beyond that will just raise and narrow the power band. Too often guys trade velocity and efficiency for unneeded.. and totally unusable.. cfm. I see and hear it every year in shops and on the street cars. "Huh?.. is that all they flow?.. you should get bigger heads. Hell.. even the little heads on my awesome 7,000 rpm 327 Chevy flow 325 cfm @.800 lift!". IMHO, many of the designs are oversized for their intended application. Aside from telling me where I'll end up if I keep grinding, who the hell really cares if the port goes well past .800 lift until it stalls when the cam spec's @ .550 thou?

In a nutshell. If the short turn, throat, and curtain area can handle higher main section and manifold runner speeds?.. that's usually how I bag them up. If not?.. most don't.. I just have to crutch it and blow the port out like many others seem to be doing. Of course, stroker motor popularity has changed much of the cylinder head requirements and helped many of these overblown castings to better fit into the mix. Course.. now that everything is moving towards CNC.. many are now buying overblown castings that are blown out even further. The ghost of Stroker McGurk seems to be everywhere now that he's gone digital.

For me, the main point of all this is rambling is that aside from the shortblock and induction.. HP/CFM potentially shows direct correlation to the efficiency of a ports design. And all else being equal.. from there we start cross referencing other components impact and looking at CFM/CSA. And ultimately CFM/curtain area. Discharge coefficient numbers are very very important to me and I purposely keep my ports smaller to give better bang for the CSA used. CFM under the flow curve!

I'm working on my Kasse P38's off and on lately and believe more than ever that cylinder head mfgrs are doing it backwards for the street guys. If it wasn't for capitalism/product placement, they could easily take a larger top-line design and shrink it down with great success on wider powerband setups. Not everyone needs shaft rockers either.

Randy

[utmost]

Cam King so port flow is totally meaningless at 28" so we can use any flow we want and make any Hp we Want (of course with the right cam)!!!!!

Reading is FUNdamental.

It is meaningless to look at only the peak flow of the head, and try and calculate available HP.

Do you really think a port that maxes out at 400cfm@.900" lift has the same power potential as a head that maxes out at 400cfm@.700" lift ?

Looking at the max flow of a port is nothing but a marketing tool for head porters.

I would hope most people would know that.

[randy331]

BUT, the reasons that one engine takes a cfm and makes 2.38 HP out of it, while a different engine can only make 2.04 out of that same 1 cfm, is anything but meaningless to me.

So, in your mind, the CFM the engine is actually using, is directly related to the flowbench CFM measured at one static lift point ?

Please tell me how peak flow CFM is any more relative then the flowbench CFM at .300" or .400" at least on an engine with .700 lift valve lift ?

To a large extent yes, peak flow will dictate power way more than flow at .300-.400. On the heads I've seen/done, the flow curve doesn't just spike at peak lift. If the flow at peak lift is good, (say .700") it won't be a turd at .500-600 either.
On an engine with .700 lift and some duration, the piston cfm demand at .300-.400 lift is zero, so chasing flow at those lifts would assume that something besides the piston is demanding cfm. BTDC one could assume the ex. system is creating a cfm demand, but how full will you get the cylinder BTDC?
So what is creating a demand for cfm ABDC? The inertia of the intake tract should be, so I concentrate my efforts on seeing to it that happens first and let the flow at .300 be whatever it ends up being. Since within a given cylinder head type (say 23* sbc) there is little you can do to increase .300" lift flow other than go to a larger valve, shallower seats, or scallop out the chamber all around the valve. None of these are good for Higher lift flows or D/C. Going to shallow seats, big valve and throat chasing flow at a lift when the piston is going the wrong way don't make sense to me.

If you look at HP/ average cfm between the 650HP/319cfm engine and the 690/290cfm engine it shows even a larger difference.
650/257 average cfm = 2.53 HP per average cfm
690/232 average cfm = 2.97 HP per average cfm
So is all that extra flow below peak lift helping, or hurting power?

Like Goberts said, HP/cfm is good indicator of proper port sizing and shape for the application.

I know you won't agree with me, and that's just fine with me, and no offence take.

Randy