Bernoulli - Intake vs Exhaust Flow

[Rick360]

I have seen in many places the calculated (using Bernoulli theory) limit of airflow of 146cfm/in² with a D/P of 28"wc. This holds very true when flow testing intake ports on cylinder heads. The very best intake ports can flow close to 140cfm/in² at the throat area (mcsa) of the port.

Why is it that exhaust ports will flow (on a flowbench) much more than the 146cfm/in²? I've seen them around 175 - 180 cfm/in², 270cfm with throat area of 1.50in². I've heard some of the best will flow 300cfm with the same area (200cfm/in²). These tests were done on SF600 flowbenches and agree with many head porting shops as well so these are not off-the-wall numbers, but very typical of racing exhaust ports.

What is going on here for the exhaust to flow so much more cfm/in², and even more than Bernoulli says?

Rick

 

[CCycle]

The maximum gas velocity is for a given pressure drop is given by Toricelli's formula which is a subset of the Bernoulli equation.
The difference you are seeing is probably because when the engine breathes combustion air in, the air is starting with zero kinetic energy. All air movement has to be generated by the energy available in the pressure drop. Exhaust on the other hand exits the engine with considerable kinetic energy due to the cylinder blow down when the exhaust valve opens.

 

[patprimmer]

I think Rick is talking about flow bench measurements, not real running engine conditions.
The flow bench data he quotes is at 8" water pressure differential for both inlet and exhaust, hence the confusion.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Bribyk]

Could it be because the exhaust ports are larger? I know the flowrate is corrected for area but, from my understanding of fluids, two 1" pipes will flow less than one 1.414" pipe for the same DP even though they have the same equivalent flow area (the single pipe has less wetted area and, therefor, less losses).

 

[patprimmer]

Exhaust ports are normally smaller, but taper out.

Inlet ports are normally bigger but taper in

I have a typo in my previous post. It should say 28" not 8"

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Bribyk]

You're right, I had it backwards.

 

[Rick360]

Yes, patprimmer is correct about my question being about measured flow on a flowbench @28"wc d/p, not on the running engine.

Since the original post I tried flowing an exhaust port in the reverse direction...from atmosphere back into chamber/cylinder on flowbench, just to see how much it could flow that direction since the exhaust port on a race head has very little bend. It flowed 141.3 cfm/in² vs 175cfm/in² in exhaust direction.

Thought of all the places to ask, this is where someone would know the answer.

Even if you don't know the answer for sure how about some educated guesses?? Anybody?

Rick

 

[patprimmer]

My wild guess is that it's related to the taper direction since as you reverse flow direction you change the effect.

What happens to the inlet if you change direction of flow.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[dgallup]

I have never done benchflow testing so this may be a stupid question. I believe for inlet testing you supply the inlet with atmospheric pressure and pull 28" of vacuum on the head. For exhaust testing do you supply the head with atmospheric pressure and pull 28" of vacuum on the exhaust? If so I would expect similar airflow limits. However, I would think it would be much more logical to supply the head with a positive 28" of water column and have the port at atmospheric pressure. In this case the supply density is much greater and I would expect (and Bernoulli would predict) a much greater airflow limit corrected back to SCFM.

 

[Rick360]

dgallup,
Yes, air is supplied to the cylinder/chamber under +28"wc pressure above atmosphere.

Could you expand on the "why" and if Bernoulli predicts a limit much higher, how high?

Thanks,
Rick

 

[dicer]

When testing in flow what was used at the entry? If its just a bare port with nothing other than a flat surface for entry then of course it will be less.

 

[dgallup]

In the case of the intake test the supply pressure is 1 Atm - 28" WC = 378.78" WC. In the case of the exhaust test the supply pressure is 1 Atm + 28" = 434.78" WC. The ratio is 1.1478 so I would expect exhaust ports of similar flow capabilities to flow about 15% higher in SCFM. Your 140 cfm/in2 intake would flow 160.7 cfm/in2 under the exhaust test conditions. Not quite as high as the 175-180 you claim for some exhaust ports.

 

[Rick360]

Well I think I have the answer for those who are interested. I'll see if I can explain it like it was explained to me.

The flow can exceed the 146cfm/in² in the exhaust direction because of pressure recovery due to the exhaust port and attached pipe acting as a conical diffuser downstream of the throat venturi. This allows the static pressure at the throat (mcsa)to drop well below atmosphere (making Bernoulli happy) since the overall test pressure is adjusted to 28"wc.

When flowing in the intake direction there is almost no chance for pressure recovery due to the sudden csa change entering the cylinder adapter, therefore the static pressure at the throat will be nearly the same as the flow bench test pressure which is measured inside the flow test bench.

Thanks to Patrick Hale(DragRacingPro.com) for solving and explaining this to me.

Rick

 

[willeng]

Rick360

Just a little more on your Reverse Flow EXH test, the greater the percentage difference in the two readings the better as an anti reversion measure.
Anything you can do to stop "Reverse" flow in the Exhaust & also the Inlet port will help with performance.
Always have a look at your Reverse flow & do what you can to slow it down. View it as a one way only open centre valve where turbulence, flow direction etc etc stops the flow one way instead of a ball & spring doing it in a proper one way valve.

It's hard enough as is getting the cylinders full efficiently without anything wanting to flow in the wrong direction against it.

The results are worth every minute working on it.

Pat,
Your correct with the test proceedure comment on flowbench figures compared to real life running conditions.

 

[jbond]

In response to Willeng's above post, and one way ports, yes I learnt that on motoman's site - has anyone read that? I am rebuilding an engine at the moment, but didn't have the guts to trust metal putty on the intake ports of my daily driver 500+hp baby

 

[howboycat]

It's going to get out.

Not much of a concern to me.

A turbo is the worst exhaust restriction I can think of. In fact by benifiting only the intake side they make tremendous power.
Pretty much a solid case for spending 95% of your time trying to get more in.

Pressure is always tying to equalize.

In one case atmosoheric pressure is trying to fill say an 800 cc low pressure area.

On the other hand pressure greater than atmospheric is trying to equalize with the entire earths atmosphere`