Superflow Conundrom

Author
Discussion

PeterBurgess

Original Poster:

775 posts

145 months

Wednesday 1st April 2015
quotequote all
Here is an odd one, V8 heads flow tested on a Superflow flowbench. Flow figure according to Superflow operator could only give a max bhp of 330.8. The heads were fitted to a block and run up on the same company's Superflow engine dyno and showed 360 bhp, any thoughts why the difference when both pieces of equipment are Superflow?






Peter

anonymous-user

53 months

Wednesday 1st April 2015
quotequote all
Static airflow != power



(No way would any engineer worth there salt make any statements regarding static airflow and the likely power potential, unless they were backed up by a LOT of real data to support than assumption)

Kokkolanpoika

161 posts

150 months

Wednesday 1st April 2015
quotequote all
Flow bench measure only direct airflow, not moving inside/backwards flow like engine will make..

Also port CA will have effect to this.. Big port may flow less and gas speed is poor, but gas has got more time to toutch port wall´s and not want to move straigt when port is making to turn valve.. = More power.. You can do some small modifications on you´r valve seat´s and you wont see them on flow bench, but in real life it will make more power.. I have seen one Opel engine witch has got Hand finished valve seat´s angle, in the flow bench you don´t see any difference, but in real life it will make +11hp more.

Pumaracing

2,089 posts

206 months

Thursday 2nd April 2015
quotequote all
PeterBurgess said:
Here is an odd one, V8 heads flow tested on a Superflow flowbench. Flow figure according to Superflow operator could only give a max bhp of 330.8. The heads were fitted to a block and run up on the same company's Superflow engine dyno and showed 360 bhp, any thoughts why the difference when both pieces of equipment are Superflow?
You mean other than the flowbench operator not actually being capable of following the Superflow manual properly? The actual Superflow equation for potential power per cylinder out of the old SF110 manual is peak CFM (@10" water pressure drop) x 0.43

At 25" pressure drop this would change to CFM x 0.272. (0.43 / 1.581)

At the 550 thou flow figure marked with a cross of 165.4 CFM this would equate to 165.4 x 0.272 x 8 = 360 bhp. Exactly what the engine showed.

What the operator obviously did instead to get 330.8 bhp is just multiply 165.4 by 2 which must be something of his own devising and clearly pointless. In actuality I suspect he used 0.25 as a multiplier instead of 0.272 so it was really 165.4 x 0.25 x 8 hence about 10% too low. Hardly a conundrum of Holmesian proportions.

PeterBurgess

Original Poster:

775 posts

145 months

Thursday 2nd April 2015
quotequote all
Interestingly the dyno wasn't taken up to peak power revs for some reason, as can be seen the power is still climbing, the builders reckoned around 7000 rpm peak power and 370+ bhp, designed to be useable to 8000rpm. More power to come with some valve redesign.
The pic attached seems to show 30 cu in engine with the measured flow should give around 344 bhp at 25" flow and 7000 rpm peak.


PeterBurgess

Original Poster:

775 posts

145 months

Saturday 4th April 2015
quotequote all
Thinking further about this, the Superflow chart is for full intake/inlet system flow, the flow figures I showed were for port flow alone so one could pretty much assume the flow would be a little less with the full inlet system attached. So it looks like the flow to bhp conversion is even further off in this case. It also makes one think, if someone used the system or system maths to dictate maxima they may well fall short of the bhp that can be achieved?

Peter

anonymous-user

53 months

Saturday 4th April 2015
quotequote all
The problem with getting from "airflow" to "power" is that it involves combustion. Even today with 3d simulation on high speed computers, the subtleties of fully simulating the combustion process are myriad. We have really very accurate Fluid dynamics models that can generally predict cylinder filling to within a few percent, but still struggle to get to a similar accuracy for power prediction. This is because the combustion process is highly chaotic, with tiny changes in initial state leading to large changes in heat release profiles. Particularly in the case of detonation limited ignition timings, you need a very fine simulation grid to accurately determine critical pressures/temperatures in the end gas regions.

Outside of this, the best you can do is to look at similar engines, where you know the intake mass flow and the power and form a "rule of thumb" value for specific pair consumption. However, this will always be highly inaccurate, and changes to engine specification and calibration parameters can easily give +- 10% change in BMEP.......

PeterBurgess

Original Poster:

775 posts

145 months

Sunday 5th April 2015
quotequote all
Interesting what you say about +/- on BMEP, personally I have never understood calculated/guesstimated power limitations, just constraints to push forward, lower bhp advances as we get down more to the nitty gritty smile

I attach flow figures from a head tested at same time as the other one. This was predicted to make 343.8 bhp, in practice the dyno operator has seen max of 340 bhp from this company's heads. Maybe a lot in what you say about +/- effects from combustion, especially it would seem with the humble old Rover V8 engine. Main differences I could see with the two heads was little or no top cut on the valve seat on the '343.8' heads, all the protruberances surrounding the spark plug had been removed on the '343.8' heads and I had the 'feeling' the valve throats were a little wide/big compared with the expected flow limitations (vol of flow and flow effects) I envisaged with the rectangular ports.



Peter

Pumaracing

2,089 posts

206 months

Sunday 5th April 2015
quotequote all
That second head has an identical flow curve to a John Eales (JE Motors) 43mm head I tested nearly 25 years ago and which I actually found nigh on impossible to beat which caused me some chagrin.

Lift Flow
50...25.0
100..49.1
150..77.7
200..104.5
250..127.8
300..144.8
350..152.8
400..164.1
450..169.9
500..172.4
550..170.2


Pumaracing

2,089 posts

206 months

Monday 6th April 2015
quotequote all
Max_Torque said:
Static airflow != power



(No way would any engineer worth there salt make any statements regarding static airflow and the likely power potential, unless they were backed up by a LOT of real data to support than assumption)
Hmmm. Obviously completely different engine designs, chamber burn efficiencies and CR etc will affect the amount of power possible from a given amount of port flow but within reasonable limits I've always found it to give a pretty good approximation of bhp potential. For a given engine if all that changes is the port flow then it can be very accurate. When I was building a lot of Ford CVH race engines where porting was not allowed but different castings flowed different amounts of air I could predict bhp from the flow curve to about 1 bhp over a possible spread of almost 15% bhp from the best heads to the worst.

However this involved much more sophisticated algorithms than just taking the peak flow number. Low and mid lift flow is also very important.

tonyfoale

1 posts

269 months

Friday 29th May 2015
quotequote all
It is unreasonable to expect that one can predict engine power in general by only looking at max. valve lift static flow. If that were the case then who would bother with cam timing/duration, compression ratio, combustion chamber shape and the miriad of other things that affect power. I doubt that many here would expect flow and bhp to be proportionally related. That is, who here would expect a 20% power increase to come from a 20% increase in static full lift flow. If I got a 10% bhp increase I'd be over the moon.
However, it is reasonable to say that UNLESS we have at least x CFM static flow @ n" WG we cannot get y BHP. You can't produce power unless you get air in. It is what we do with air once inside and how it travels in the prot that determines just how much power we get.

stevesingo

4,848 posts

221 months

Friday 29th May 2015
quotequote all
13years 7months a member of Pistonheads and this is your first post.

Gentlemen, we have a World Champion Lurker! Congratulations!

BTW, I agree with you points.

PeterBurgess

Original Poster:

775 posts

145 months

Monday 11th April 2016
quotequote all
An update. Last year the car struggled to fuel correctly. This year the heads now have my own shape inlet valves, larger ports that show a theoretical max of 420 bhp and a set of shiny new Webers. On HT racing rolling road flywheel power last year was 380 (close to the superflow dyno figure?) and this year with the larger ports and nice shaped valves we now have 408 flywheel power and 326 lbs/ft torque. First race was at Rockingham last Sunday, pole position, race win, fastest lap and lap record. Close race though only half a second 1st to 2nd after 10 laps! Looks like the Superflow flow bench prediction was miles off smile

Peter

hoffman900

23 posts

97 months

Tuesday 12th April 2016
quotequote all
Peter,

Where did the numbers peak vs before? Same dyno? Are your flywheel numbers back calculated or were they measured on an engine dyno?

I fail to see how ports will cause the engine to fuel properly - unless the port to port variation was terrible.

Fueling is typically an issue of the exhaust system and camshaft - which leads me to believe your low lift figures are different.

Also, why are you using figures from the head alone? In an ideal world, with the intake and carburetor attached, you'd hope for under 10% flow losses (will increase with lift), but this isn't always the case. Any engine modeling program, camshaft designer, etc, will want flow figures with the everything attached. It also can make a difference on the port behaving or misbehaving..

Here is a chart of a personal project of mine:


This isn't what I would call a "max effort" either. RPM goal is 7200rpm, so the MCSA 1.76in^2 (or 38mm in diameter). Wanting a bigger peak, we could open it up to 1.95in^2 (40mm diameter) which would push those numbers even higher. Average velocity through the port at 28" H20 is right around 300fps at 7200rpm.


As for the flowbench, Darin Morgan has an updated site and a nice 'post':
http://darinmorgan.net/frank-talk-about-flow-numbe...

I don't think Darin really needs an introduction, but for that don't know, look here: http://darinmorgan.net/about/

Edited by hoffman900 on Tuesday 12th April 00:50


Edited by hoffman900 on Tuesday 12th April 01:30

eliot

11,361 posts

253 months

Tuesday 12th April 2016
quotequote all
Pumaracing said:
You mean other than the flowbench operator not actually being capable of following the Superflow manual properly? The actual Superflow equation for potential power per cylinder out of the old SF110 manual is peak CFM (@10" water pressure drop) x 0.43

At 25" pressure drop this would change to CFM x 0.272. (0.43 / 1.581)

At the 550 thou flow figure marked with a cross of 165.4 CFM this would equate to 165.4 x 0.272 x 8 = 360 bhp. Exactly what the engine showed.
.
Wow that brings back memories - my dad bought a sf110 back in the 80's and he was fed up hand cranking the calculator- so he paid me pocket money to write a program in BASIC that would do all the corrected valve lifts and calculate the results including bhp - which came close to actual power on a dyno.




PeterBurgess

Original Poster:

775 posts

145 months

Tuesday 12th April 2016
quotequote all
Hi Eliot, I can remember spending hours and hours with a Vic 16 then 20? till the answer came out the same as 10 mins for each lift on a calculator smile

Hoffman900, not my power figures. Last year on HT Racing rolling road identical engine 380 calculated flywheel, they also have Superflow dyno ( not the one from first post which was a different company)and numbers are very similar I am told. Fuelling nothing to do with port sizess and I never said they were. Fuelling problem caused by carburettors and cured with new carbs. Ports were not large enough to go past 400 bhp as they originally were. We worked hard to get em as large as we dare on the inlet side.This year figures now 408 flyhweel calculated. Maybe others Sateside make our stuff look crap but we are pretty pleased with the progress we have made in the last two years starting pretty much from scratch with all aspects of the engine from crank design upwards. We last worked hard on developing the Buick V8 in the early 90s when we dominated the UK TR racing with a TR8. You have to remember the numbers are pretty irrelevant, the race wins, lap records and championships speak for themselves in terms of holistic team effort when compared with the teams who do not win. This must be the same Stateside, it doesn't matter how much folk blow their own trumpet it is for the race track to deliver the obituaries of the front runner also rans. Reading the link you added, it has taken me 32 years, first Championship 1987 to realise I don't know anywhere near as much as I thought about engine tuning and the joys of discovering a little more power here and there do not dull after all this time, I find as I get older and pay off debts I can buy better equipment for machining and tuning and power measurement which also helps push the power levels upwards. I will state again my deepest held 'truism' that it is from the mods that do not work as expected one learns the most about power production and reliability.

What I have found is Stateside B series engines would be very good on a long track somewhere where the high rpm bhp works well BUT on the UK tracks the higher revving, to achieve the power and consequently less grunt Stateside engines get burnt by the lower rpm peak grunty UK spec motors.

Peter

Stan Weiss

260 posts

147 months

Tuesday 12th April 2016
quotequote all
Peter,
I have sent you an email.

Stan

hoffman900

23 posts

97 months

Wednesday 13th April 2016
quotequote all
Hi Peter,

Thanks for the clarification. My questions weren't meant to criticize your work, just get a better understanding of what happened.

Also, the BMC B engine isn't nearly as developed to the same standards at the venerable A-Series.

For example, in SCCA FP specs for the best examples:

1.5" SUs
Stock valve sizes
Stock rod length
No material adding to the port
.040" overbore
Stock stroke

"Typical VE% for FP 1275 is around 102-103%. At peak hp, around 7600 rpm. 5900 rpm, about 106%"

These would be good engines on any sized circuit - chassis set-up and transmission gearing considering.