Bernoulli - Intake vs Exhaust Flow

[Rick360]

I have read many posts explaining that a port has a limit of 146cfm/in² at 28"wc and use Bernoulli equations as the explanations for this. I was all happy with this and didn't question it until someone pointed out that exhaust ports flow well in excess of that at 28"wc. I hadn't really thought of it that way before, but after a few calculations I see they are much higher and I've seen ~180cfm/in² on an exhaust and others said they had seen higher. I think I understand why an exhaust port would flow more than an intake port, but how can it surpass the Bernoulli limit of 146cfm/in² at 28"wc?

Does anybody have a logical explanation for this? One you could explain simple enough for me to understand?

Thanks,
Rick

[FlowSpecialist]

Certainly. Any figures you saw in excess of 146 CFM per sq inch, or even close to that, were wrong. I trust that's simple enough.

Dave

[Stan Weiss]

146cfm per square inch of area (@28" H2O) and the Valve / Curtain Area are used to calculate Discharge Coefficient for the port, but have never seen the 146cfm per square inch of area with valve area.
Here is my table of Cylinder Heads Flow Data at 28" H2O. Stan Weiss' - Cylinder Head Flow Data at 28 Inches of Water
It contains over a 1000 heads. I calculate flow per sq/in of valve area at all the valve lifts when I have a valve size listed. On the exhaust I see a number of exhuasts in the 14x area,
1) The HRD Ported GM#615x 15 Deg Rolled to 13 Deg shows 147 cfm per sq. in of valve area.
2) The Chevy / GM SB2.2 Hendricks - Darin Morgan shows 121 Intake and 143 exhaust cfm per sq. in of valve area .
3) Chevy / GM Splayed Valve 12 Deg MBE Port shows 148 cfm per sq. in of valve area.
4) Dart Little Chief 11 Deg 275 Alum, Dart Little Chief 11 Deg 315 Alum both show 148 cfm per sq. in of valve area.
Doesn't at around .25 D/L the Valve / Curtain Area becomes greater than the throat area? Look at the Discharge Coefficient of the Exhaust for these heads.

[Rick360]

The 146cfm/in² limit would be based off of the actual mcsa of the port, which should be the throat area, not the valve area. Don't forget to subtract stem area from the throat area.

With a 1.600" ex valve and a 92% throat (very large throat) and a 11/32" stem, the max flow possible (with or without a pipe) would be 234cfm using the 146cfm/in² rule.

1.6 valve x 92%= 1.472" diameter throat for area of 1.701788in² and .343" stem = 0.092401in² for a net effective area of 1.609387in² x 146 = 234.907 cfm.

There are many (nearly everybody) head porters who claim bigger numbers than this. 270-280 cfm is not uncommon for 1.6" diameter ex valve flows on 15º/18º heads. This would be ~ 175cfm/in² in this example. Do all SF benches measure this wrong? What am I missing?

Rick

[Rick360]

Wow, nobody has an explanation for this short of "the whole world is wrong"?

Rick

[FlowSpecialist]

I've commented on exhaust flow already in other threads, that I have doubts whether all (any?) flowbenches measure it correctly and that any proper calibration checks should do exhaust as well as inlet flow. When I checked a Flow Quik system last year, after we got it calibrated on the inlet side it was reading about 20% high with the motors reversed on the same orifice plates.

You know already that 146 cfm per sq inch is a limit imposed by physics so what else we can say to you I'm not sure. You must ask the people who claim higher figures to explain themselves.

Dave

[thadeausmaximus]

I may be a little dense but what is so special about the speed of 238 mph(146 cfm per sq inch) in the pressure range of 14.7psia(1 std atm) and 13.7 psia(1 atm - 28"h2o). Is 238 mph the max speed air moves from 14.7psia to 13.7 psia? and where can I learn about why?

[Dusty]

isn't that 146cfm/in^2 based off an orifice not a venturi?

[Rick360]

Originally Posted by Dusty

isn't that 146cfm/in^2 based off an orifice not a venturi?

No, a square edged or sharp edged orifice normally flow ~ 60% of that 146cfm/in² (cd of .60).

Originally Posted by thadeausmaximus

I may be a little dense but what is so special about the speed of 238 mph(146 cfm per sq inch) in the pressure range of 14.7psia(1 std atm) and 13.7 psia(1 atm - 28"h2o). Is 238 mph the max speed air moves from 14.7psia to 13.7 psia? and where can I learn about why?

Do a Google search for Bernoulli. The Bernoulli equation is what everyone cites when they mention the 146cfm/in² limit. This is the theoretical limit with a D/P of 28"wc. I haven't actually done the math, but have heard this from many reliable sources.

Originally Posted by FlowSpecialist

I've commented on exhaust flow already in other threads, that I have doubts whether all (any?) flowbenches measure it correctly and that any proper calibration checks should do exhaust as well as inlet flow. When I checked a Flow Quik system last year, after we got it calibrated on the inlet side it was reading about 20% high with the motors reversed on the same orifice plates.

You know already that 146 cfm per sq inch is a limit imposed by physics so what else we can say to you I'm not sure. You must ask the people who claim higher figures to explain themselves.

Dave]

Dave,
Is it your opinion that the way most flowbenches (like the large Blue ones) are setup to measure the depression or pressure (inside the upper plenum just below the head adapter), that it may appear to flow more cfm/in² than 146 because of pressure recovery in either direction (intake or exhaust)?

Thanks,
I shouldn't worry about it and just go on flowing heads with numbers like everybody else and go about improving them rather than trying to understand why something like this doesn't make sense.

Rick

[FlowSpecialist]

Originally Posted by Rick360

Dave,
Is it your opinion that the way most flowbenches (like the large Blue ones) are setup to measure the depression or pressure (inside the upper plenum just below the head adapter), that it may appear to flow more cfm/in² than 146 because of pressure recovery in either direction (intake or exhaust)?
Rick

I covered this in some detail in the thread on High Velocity Porting a few weeks ago.

There will always be some element of pressure recovery taking place but this will be a small amount for intake direction flow over the short length of a bore adaptor. You need several pipe diameters for pressure recovery to fully establish. I did do some tests many years ago with my manometer tapping at the bottom of the 18" long flowbench downpipe instead of at the very top right under the test head where it should be.

With the tapping in the right place at the top my bench reads spot on to the calculated flow of the orifice plates. At 25" of water the CFM results of the two tapping locations on my orifice test plates were as follows.

Plate..Top....Bottom....% Error
#1.....34.2....35.4.........3.5
#2.....78.6....83.7.........6.5
#3....123.9..136.0.........9.8
#4....165.0..184.5........11.8

So as the flow rate goes up so does the amount of pressure recovery that takes place. I have no idea how high the error would be at 300 CFM or more on something like a Chevy head. Anyway as I say over the length of a 4" or so bore adapator it isn't going to be a big percentage. Of course if you use longer bore adaptors than anyone else.....

With exhaust direction flow, provided the exhaust port is dumping straight to atmosphere there can't be any pressure recovery because there's nowhere for it to take place or be measured. However people have now started adding lengths of exhaust pipe after the port because "it makes the port flow better". Complete nonsense of course. You can't make a thing flow more by sticking something else behind it. That second 'thing' can either be a zero flow restriction if it's big enough or an actual restriction if it isn't but it can't be a negative restriction. It should however now be apparent what is actually happening and why none of these results have any meaning whatsoever.

Unfortunately airflow is complex and just going out and buying a flowbench doesn't teach you much if anything about it. It's therefore not really surprising if most of the people who use flowbenches don't actually have a clue about what they're really 'measuring'.

It is however yet one more reason why the best thing you can do with other people's flow figures is ignore them.

Dave