Airflow and rpm vs power

[timdog]

A lot of the development i've done in the last year has been suggesting that the very high revving engines i work on require substantial port velocity to work efficiently. Increasing cfm can actually cost serious top end power (past 14k revs) without affecting power much below that. Port velocities are well over 350fps at 28" before the choke point.
A guy i know in the piston business reckons the current F1 engines use relatively small and high speed ports. He does supply a team with pistons but i have no idea if he giving me BS on the heads.

Another thing is the wave tuning on these engines is super critical.

As i don't work on slower revving engines i'm wondering if this is the case with these as well.

Is anyone else working with these rpm and have any useful data that they want to share? I have read threads on here suggesting max port velocity from maxracesoftware which were very informative but obviously these engines were much bigger cc.

[maxracesoftware]

Increasing cfm can actually cost serious top end power (past 14k revs) without affecting power much below that. Port velocities are well over 350fps at 28" before the choke point.

what you are describing ...i've seen this also every once in awhile

by the time i get to work with Heads , its like 3rd party.

1st the Guy calls up someone like Cam Motion
or anyother Cam Company, and tries to solve that
problem with running a bunch of different Cam Lobes
on the Dyno/Track..but can't fix it with Cams !

usually Terry at CamMotion sends them to me,
and so far its always a speed FPS problem in the Heads.

everyonce in awhile you pickup the Flow maybe 20-30 CFM @ 28"
but at the Track/Dyno it runs the same or "slower"

you might see 10-15-20 "Peak HP" gains after fixing problem
but you might also see 40-50 more HP "after" the Peak HP RPM point too
(thats mostly on engines turning 8000+ RPMS,
where pumping loses rapidly increase )

sometimes much more gains like that SS350 example on another Post
and that was at 7500-7600 rpm


from other thread=>
FPS = (CFM * 2.4 ) / CSA

if Head Flows 127 CFM at .200" Lift = 156.6 fps @ 1.948 csa

if head flows 260 CFM at .700" Lift = 320.3 fps @ 1.948" CSA

looking at it another way like you are describing =>
suppose your Heads are right at 350 fps and that FPS
happens to be on verge of Choking
so you then do more careful shaping Porting or change to a better
valve job or Chamber Porting...so now you've not really increased
any CSA Area inside the Port..but now the Speed FPS is too fast
..now the Engines choked.

old heads
FPS = (CFM * 2.4 ) / CSA
350 FPS = ( 284 cfm * 2.4 ) / 1.948

after new Work
369.6 FPS = (300 cfm * 2.4) / 1.948

now the Engine makes no difference in HP
or actually looses HP , especially after RPM point of Peak HP

thats the "Basic Concept"
not actual FPS each and every Engine will choke at

the Formual 1 Heads have much straighter ports that
can handle much greater FPS without Choke setting in

A guy i know in the piston business reckons the current F1 engines use relatively small and high speed ports. He does supply a team with pistons but i have no idea if he giving me BS on the heads.

yeah but , whats "FPS Speed" does he call "High Speed Ports" ???

Another thing is the wave tuning on these engines is super critical.

yes because of "Stroke Time" of each stroke
and dealing with Friction/Pumping Losses
..any little loss or variation is "amplified" by RPM

HP = Torque_Loss * RPM * .0001904

Is anyone else working with these rpm and have any useful data that they want to share?

PipeMax calculates what CFM @ 28" the Piston is moving
so you know what the Engine really needs at various RPMs
and your goal is to achieve supplying that at a "reasonable velocity FPS"

thats only 1 basic variable, then you have to take care of all the
other Engine variables..but at least you know CFM required.


note> because of Lag-Times and differences in various FlowBenches,
the Piston CFM Demand maybe 20 to 40 CFM greater than PipeMax
calculates at a certain RPM

[timdog]

Numbers from memory here: An example head is flowing about 160cfm inlet and 120cfm exh both at .4" lift peak. PFA is saying pseudo port velocity to be around 330fps on inlet but i remember it reading much higher than this at the ssr. The higher speed port is about 85% the volume of the slower one. Engine is 36cu in, with bore and stroke of 2.64 & 1.67
If i increase the cfm then power drops. FPS is also a traditionally more sensible figure as well with the larger cfm.
When i dyno these engines, they generally gain a lot of top-end but don't show much midrange. However, on track the guy comes back in claiming huge throttle response and stacks of midrange!

On a lower revving engine of say 11-12k this high speed issue is different. I can chuck more flow and less velocity for the gains i require but the really high rpm just seems to want more velocity. Almost as if the air needs huge inertia for the breathing to keep up with events.

Guess i just need to keep collecting data. Would be nice to work it out because revs climb more every year. This season they are running 17k still with conventional springs!

On the F1 stuff, i have no idea on fps and doubt he does. Reckons he'd seen some heads. I haven't so can't comment. Would be nice to have a look at trends though.

Thanks for all your input so far, most appreciated.

[ron hamp cycle]

i just moved the power up a full 1000 rpm on a small engine from 9000 to 10000 rpm and gained 3 hp the head flowed the same as previous heads but i reshaped the floor @ the apex and reduced the measured velocitys a bit the engine also required less ignition advance @ the uper end of the igniton curve and a considerably smaller main jet there had to be some choking before hand .maybe with the increased flow on the small 4 cylinder engines there needs to be a slight increase in port cross section but thats going to move the entire curve up again .recently i had found on the old xr 750 hd engines that the increases in flow had made know hp gains and the measured velocitys @ the port were nearing 400 fps @28" i did not want to loose the aceleration off the corners bys enlarging the ports more so i reduced the lift of the cam to where the measured velocitys were more inline the engine actually made more power and acelerated to its maximun rpm sooner .the bike also finshed 3rd on the Ducoin mile this year which ended up beeing the fastest race of the year and did it with .050" lift on the cams using a shorter rocker ratio.

[maxracesoftware]

and did it with .050" lift on the cams using a shorter rocker ratio

from other thread=>
FPS = (CFM * 2.4 ) / CSA

if Head Flows 127 CFM at .200" Lift = 156.6 fps @ 1.948 csa

if head flows 260 CFM at .700" Lift = 320.3 fps @ 1.948" CSA

yeah, that just about covers all 3 aspects of the same Concept

that is, => likwise if you increase max lift with an already
too-fast Choke problem...it hurts top-end HP
even Peak TQ

the "best fix" in the long run, ...the fix i feel makes the best overall HP/TQ Curve...is to actual fix the Speed FPS problem....
then that will let you go back and try the Lift..might help afterwards.

[prochamp]

can the min csa be at the short turn opening to the bowl area or is it either the pushrod pinch area or valve seat choke area? also what mach index #s are you getting at 350 fps? thanks rob

[maxracesoftware]

can the min csa be at the short turn opening to the bowl area or is it either the pushrod pinch area or valve seat choke area? also what mach index #s are you getting at 350 fps? thanks rob

from my memory reading Jenkins's Book about 20 years ago,
i think Bill Jenkins and other years before him were saying
you should "funnel" the Mixture, that is use Taper from the
Plenum's Entry ...then "neck it down" inches before
the Short Turn Apex and Bowl Area..and let it expand
to the Short Turn Apex + Bowl
...in that way, you slow the velocity and turn the velocity into pressure before the Short Turn Apex and Bowl ,so it can turn , so it increases pressure differential between Bowl and Cylinder .

then others say just use constant Taper all the way to the Seat

from working with old 23 deg SBC heads with pushrod choke
..if its too much choked at the pushrod area, thats about
4.000" inches before the Curtain Area....the Pushrod Choke problem
seems to loose more HP than the Short Turn FPS speed problem...
so from just those experiments,..i'd say it appears if the choke location is
located too far from the Curtain Area..that it must also
be influencing Wave Tuning .

in other words, you Tune your Intake Length from
Entry -to Seat Angle or Curtain Area....but if you put too much Choke
midway between them...i think it disrupts or change wave tuning
enough to hurt Torque, by changing its pressure wave "timing"

you can successfully run a lot more Short Turn Apex FPS
than you can get away with at the Pushrod area

i never checked out the "reverse", that is,
to put the actual Choked min. CSA at the Plenum's Entry,
and expand the Total Induction System Length from there
to the seat....i think it would be obvious the Heads would be
CFM Flow Choked + weird wave tuning problems coupled
with strong reversion pulses in Plenum ???


the Mach # at 350 FPS = approx 350 divided by 1116
thats .3136 Mach ..the 1116 FPS Speed Of Sound depends
on local mixture temperature and density

the FlowBench is steady-state constant depression/velocity
and 350 fps = 28" Test Press. approx.
but the correlation to Live Engine is that
350 fps at 28" on FlowBench maybe the same as between 614-700 fps
.627 Mach = 700/1116
.55 Mach = 613.8 fps

most People think that just because you FlowTest at 28" inches
that everywhere inside that Port air will be flowing at 350 fps...
thats not going to happen unless you have a perfect, constant area,
perfectly straight piece of Pipe.


what you'll actually discover with Pitot Probe is that
Air Speed FPS inside Port is very different at different locations
..one spot might be at 350 FPS, but another might be at 400+ FPS,
another might be at 280 fps

i have seen a few Heads go to what calculates a little over .68 Mach
and make their Peak HP there

[timdog]

Interesting comments of the port choke position/place of minimum csa.
Here is a port reduction that gave mixed results.

Length is 3.75" and it would be very similar to data given earlier, although i seem to remember this had maybe slightly higher velocity again. Obviously the upper mold is the standard port.
The small port netted a 5% power gain at peak with a steady increase all the way from the start. The one thing it did struggle with was fuelling lower down up to nearly half way up rev range. Very bad pulsing i think. Was all over the place. Cleaned up at the top. It also straightened out the trough half way up the power band that normally you can't get rid of.
I've now moved this shape on quite a bit and its straighter on the bottom and top with a less agressive entry, but still a slight taper all the way. The point of minimum csa is closer to the valve stem too and slightly larger than shown here. This latest one does not suffer with any undue wave fuelling problems and has netted an 8% power increase over stock.
I definately agree that the port needs to open out again in preperation for the back of the valve and bowl area. Failure to do this robs power everywhere but more so up top.

[prochamp]

thanks larry,that makes total sense.do you think those high mach # heads were successful due to the total engine combo,especially cam profile? do you usually try to stop at .55 mach or what do you shoot for? how much effect does the valve lift & closing time have on min csa air speed? it seems to me that choosing a cam without knowing what & where your min csa is is somewhat of a crapshoot.is there a formula for figuring this out? thru the years on most of the engines i've built the majority of the quicker ones i was conservative on the camshafts,now i realise why they ran the way they did (min csa).it's kind of enlightening to now understand that.it brings to mind some of the bench racing stories of guys telling me about their 9000 rpm 283's.if they were doing that they had to bo sonic choked a few thousand r's before that.

[SStrokerAce]

Interesting topic here.

[steelcomp]

Interesting comments of the port choke position/place of minimum csa.
Here is a port reduction that gave mixed results.

Length is 3.75" and it would be very similar to data given earlier, although i seem to remember this had maybe slightly higher velocity again. Obviously the upper mold is the standard port.
The small port netted a 5% power gain at peak with a steady increase all the way from the start. The one thing it did struggle with was fuelling lower down up to nearly half way up rev range. Very bad pulsing i think. Was all over the place. Cleaned up at the top. It also straightened out the trough half way up the power band that normally you can't get rid of.
I've now moved this shape on quite a bit and its straighter on the bottom and top with a less agressive entry, but still a slight taper all the way. The point of minimum csa is closer to the valve stem too and slightly larger than shown here. This latest one does not suffer with any undue wave fuelling problems and has netted an 8% power increase over stock.
I definately agree that the port needs to open out again in preperation for the back of the valve and bowl area. Failure to do this robs power everywhere but more so up top.

What would an intake runner look like for this type of port? Is this specifically for a short IR, or will it work with conventional cast intake/plenum designs?

[maxracesoftware]

larry,that makes total sense.do you think those high mach # heads were successful due to the total engine combo,especially cam profile? do you usually try to stop at .55 mach or what do you shoot for? how much effect does the valve lift & closing time have on min csa air speed? it seems to me that choosing a cam without knowing what & where your min csa is is somewhat of a crapshoot.is there a formula for figuring this out?----ProChamp

Here are Specs from a well-tested small block Chevy SuperStock
engine Combination that i have a ton of FlowBench/Dyno/Dragstrip data
on.

SS Chevy 362 cid 4.065 x 3.493 .013 Legal Stroker C-12 VP Race Gas
#041x 1.940/1.500 Cylinder Heads 165.0 Intake Port Volume
61.2 min Chamber Volume CC Q-Jet Carb Victor E Intake w/adapter
Cam Motion Solid-Roller Cam 1.635 Springs Flat Top BME pistons
3-Step Headers 1st= 1.750 x 14-15 2nd= 1.875 x 7.250 3rd= 2.000 x 8-9
3.500 Collectors x 16.250 long 34 deg to 36 deg BTDC .030 Lash
1.650/1.65 Rockers

made best of 585 Peak HP and 470 Peak TQ (600 RPM/SEC)
just about anytime of Year made 565 Peak HP to 575 HP with 465 to 470 TQ

at 3330 Lbs. 2-Speed PowerGlide
5800-6000 Stall, 5.86 gears , 14x32 Tires, 1968 Camaro

set NHRA SS/IA Record 10.132
incremental ET-Times
(when Engine made 565 HP 465 Peak TQ at that time)
60 FT= 1.330
330 FT= 4.058
660 FT= 6.377
1000 FT= 8.402
1320 FT =10.132

Dyno tested on SF-901 Dyno at 600 RPM/SEC
RPM---TQ----HP----Fuel Lbs
4500-449.5-385.1--166.1
4700-444.2-397.5--164.5
4900-455.2-424.7--177.0
5100-456.2-443.0--158.3
5300-464.1-468.3--169.6
5500-471.1-493.3--192.9
5700-470.2-510.3--199.2
5900-468.2-526.0--199.3
6100-465.0-540.1--204.5
6300-459.8-551.5--209.7
6500-456.6-565.1--216.3
6700-442.6-564.6--223.1
6900-432.8-568.6--217.2
7100-426.6-576.7--215.3
7300-418.3-581.4--224.5
7500-401.8-573.8--238.2
7600-394.1-570.3--231.0

to Calculate what the "baseline Mach Index " was,
use the following Formula=>

rearranging equations around

FPS = ( Bore * Bore * Stroke * RPM * .00353 ) / CA

CA = ( Bore * Bore * Stroke * RPM * .00353 ) / FPS

RPM = ( FPS * CA ) / ( Bore * Bore * Stroke * .00353 )

where;
RPM = point of desired Peak HP
FPS = Feet per Second
CA = Cross-Sectional Area in Square Inches (smallest measured)


i didn't Post the 7200 RPM point Dyno Data, but looking at 7300 RPM=>
FPS = ( Bore * Bore * Stroke * RPM * .00353 ) / CA
763.5 fps = (4.065*4.065*3.493*7300*.00353)/1.948

.6842 Mach Index = 763.5 fps/1116

.6842 Mach = approx. a chance at making
a "theoretical Trapped-Ve%" of = 132.7 % Volumetric Efficiency %

as well as this Combination might have Run,
it still was "Trapping" at best of between 104-105 % Ve

so that tells you that there was at least an approx 28.2 % PerCent Loss
and that Pumping Losses where substantial

to get this Engine to run 7200-7300 , need to find a way to get Heads
to handle very hi velocities and try to "feed" the engine CFM wise.

some engines can't make their RPM point of Peak HP occurrrence
above .55 Mach

yet you have to work the Combo like that or much better,
to Run fast in SS/IA .

if you tried for .55 Mach, that would mean RPM point of Peak HP=5868
(about the RPM it now makes its Peak TQ point)
the RaceCar would be very slow down the DragStrip,
no matter what Cam/Rear+Trans Gear Ratio/Stall RPM

-------------------------------

it's kind of enlightening to now understand that.it brings to mind some of the bench racing stories of guys telling me about their 9000 rpm 283's.if they were doing that they had to bo sonic choked a few thousand r's before that.

yes, but use the following Equation to Calculate what the
"real FPS" were for that 283CID turning 9000+ RPMs ??

FPS = ( Bore * Bore * Stroke * RPM * .00353 ) / CA

[maxracesoftware]

Interesting topic here.

more Math/Equations =>

Velocity_FPS = ( Test_Press ^ .5 ) * 66.2

Port_Velocity = ( CFM * 2.4 ) / Cross-Sectional_Area

FPS = ( CFM * 2.4 ) / CA

FPS = ( CFM / CA ) * 2.4

CFM = FPS * CA * .41666667

CFM = ( FPS * CA ) / 2.4

CA = ( CFM / FPS ) * 2.4

CA = ( CFM * 2.4 ) / FPS

using the same SS/IA Engine Example #041x Heads 1.940/1.500 Valves
165.0 Intake Port Volume CC's

Formula=> Average_CSA = Port_Volume_CC / (Port_CenterLine_Length * 16.387)

1.948 CSA area = 165.0 CC / ( 5.168" Port Length * 16.387 )

using PipeMax=>

Crank-Head-Piston-Piston-Piston-Volume-CrankRod-to--Rod-to-
Angle-CFM @-CFM @-Depth-Speed-CC per-Rotate-Bore-Crank
Degree-28 in.-28 in.-inches-Ft/Min-degree-TimeSec-Angle-Angle
49.0-219.4-277.8-0.7552-5968.0-160.60-.00113-13.37-117.63
49.5-220.7-279.4-0.7691-6002.9-163.56-.00115-13.47-117.03
50.0-222.0-281.0-0.7831-6037.1-166.54-.00116-13.58-116.43

--- Cross-Sectional Areas at various Intake Port Velocities (@ 28 in.) ---
137 FPS at Intake Valve Curtain Area= 4.425 sq.in. at .726 Lift
205 FPS at Intake Valve OD Area and at Convergence Lift = .485
253 FPS 90% PerCent Rule Seat-Throat Velocity CSA= 2.394 sq.in. at 7200 RPM
350 FPS Velocity CSA= 1.729 sq.in. at 7200 RPM Port Sonic-Choke with HP Loss
330 FPS Velocity CSA= 1.835 sq.in. at 7200 RPM Smallest Useable Port CSA
311 FPS Velocity CSA= 1.948 sq.in. at 7200 RPM Recommended Smallest Port CSA
300 FPS Velocity CSA= 2.019 sq.in. at 7200 RPM Recommended Port CSA
285 FPS Velocity CSA= 2.125 sq.in. at 7200 RPM Recommended Short-Turn CSA
250 FPS Velocity CSA= 2.423 sq.in. at 7200 RPM Recommended Largest Port CSA
225 FPS Velocity CSA= 2.692 sq.in. at 7200 RPM Largest Intake Port Entry CSA
200 FPS Velocity CSA= 3.029 sq.in. at 7200 RPM Torque Loss + Reversion

311 FPS Velocity CSA= 1.948 sq.in. at 7200 RPM Recommended Smallest Port CSA


between 49.0 and 50.0 degrees after TDC (ATDC)
the Engine will completely "suck" the Intake Port dry
that is, use up approx 165.0 CC's of its Intake Port Volume
if no other Mixture flows in to replinish .
(Pumping HP/TQ Losses will skyrocket with too much Flow restriction)


another thing interesting is PipeMax calculates that
a 362.6 CID Engine @ 7200 RPMs needs
a Total Induction Length = Best Length= 13.504
too long to fit under Hood of 1968 Camaro

so we go to 2nd Best Tuned Length= 2nd= 10.629 (PipeMax)

so 10.629" inches MINUS 5.168 #041x Int Port CL = 5.461 Manifold Length

so you use an Edelbrock Victor E Intake to "Tune-in"
on the 2nd Wave for 5200-7200 RPM Power TQ/HP range

the Victor E about averages out in that 5.461 required
Intake Manifold Runner Length
it has Flat Plenum Floor to handle uneven velocitities coming out of
unequal Throttle Bore QuadraJet Carb
plus has "low underhood" compactness to fit Rules where you can't
run HoodScoop..where you have to becareful how close you put
the top of Carb to bottom of Hood..you don't want to
Choke-Off AirFlow at the Carb

you can get better Fuel Distribution + even Cylinder EGT Temps
with Flat Plenum Floor -VS- slopeing Plenum Floor
with a type of Carb like QJet or ThermoQuad

[maxracesoftware]

This latest one does not suffer with any undue wave fuelling problems and has netted an 8% power increase over stock.
I definately agree that the port needs to open out again in preperation for the back of the valve and bowl area. Failure to do this robs power everywhere but more so up top.---TimDog

you have to be careful "using" or accounting for too much TAPER

that is,..if the TAPER is too great... you are seeing more effect
of "Shorter" Runner..you are effectively shortening up the Runner
..and it may not be the TAPER it likes , but the Overall Length
to some degree ??

[timdog]

What would an intake runner look like for this type of port? Is this specifically for a short IR, or will it work with conventional cast intake/plenum designs?

The runners are very short at about 4 inches or less total length and mounted direct to head. You'd never get this kind of engine to pull its rpm range with a long runner or conventional plenum.

you have to be careful "using" or accounting for too much TAPER

that is,..if the TAPER is too great... you are seeing more effect
of "Shorter" Runner..you are effectively shortening up the Runner
..and it may not be the TAPER it likes , but the Overall Length
to some degree ??

You could be right on that as the overall volume is obviously reduced a lot as a percentage with only a 4" runner on top of it.
It does seem quite critical on length to the bellmouths though and whether port is filled or not seems to have the same effect on peak power with each length bellmouth.