MSP volumetric efficiency

[TurfBurn]

We didn't as the flow bench arrived when my head was already done.. I wasn't worried about the overall improvement as simply put it was getting ported to the math and that's the best you can do... (I can dig up the equations if you want them)

Flow Bench's are actually quite terrible measures... in fact Ford use to do that with all their high performance heads they sold for big blocks.. they would keep the highest flowing heads for themselves. Well they found out later that they weren't the most POWERFUL heads... they found they got better dyno results from some of the heads the flowed considerably less air. So they started test stand dynoing all the high performance heads and that way they could keep the best ones for their racing team... so the flow bench lied.

Also then you have to ask if you want a valve installed or a valve out flow reading.. and what about during the different stages of lift? All those things will affect the overall flow of the head and how much throughput it has... so in other words... Flow Bench = useless numbers

Kind of similar to peak horsepower and big turbos etc... you can make a high number... but a lower peak vehicle can still outrun that big one because of more overall area under the curve... again that isn't something you can measure on a flow bench with out a ridiculous amount of effort and logging.

 

[TurfBurn]

Yep, that's usually what most non-performance oriented heads are designed to do. Unfortunately, Mazda didn't create a new head optimized for a turbocharger. They just slapped the ol' Protege head on and called it a day.

Cost benefit... for what they were creating for us there was no point in spending the 100's of thousands of dollars in creating a new head casting design for what their goals were with the car. Since there are people flowing hundreds of horse through un-modified head I wouldn't worry about that too much.

 

[JDM Sam]

Read and learn
http://e30m3performance.com/tech_articles/engine-tech/flow-1/index.htm

 

[Spooled]

We didn't as the flow bench arrived when my head was already done.. I wasn't worried about the overall improvement as simply put it was getting ported to the math and that's the best you can do... (I can dig up the equations if you want them)

Flow Bench's are actually quite terrible measures... in fact Ford use to do that with all their high performance heads they sold for big blocks.. they would keep the highest flowing heads for themselves. Well they found out later that they weren't the most POWERFUL heads... they found they got better dyno results from some of the heads the flowed considerably less air. So they started test stand dynoing all the high performance heads and that way they could keep the best ones for their racing team... so the flow bench lied.

Also then you have to ask if you want a valve installed or a valve out flow reading.. and what about during the different stages of lift? All those things will affect the overall flow of the head and how much throughput it has... so in other words... Flow Bench = useless numbers

Kind of similar to peak horsepower and big turbos etc... you can make a high number... but a lower peak vehicle can still outrun that big one because of more overall area under the curve... again that isn't something you can measure on a flow bench with out a ridiculous amount of effort and logging.

Good points, thanks for the info.

I'm curious about exacly what you did to your head. 3 angle, 5 angle, etc?

 

[Spooled]

I know a lot of you might not like it, but this is a great book. I've read it a couple times all the way though. There is a ton of great info everything from cams to valves to turbos. It may be Honda biased, but the info is very generic.

https://www.amazon.com/dp/ (commissions earned)

 

[TurfBurn]

Good points, thanks for the info.

I'm curious about exacly what you did to your head. 3 angle, 5 angle, etc?

3 angle on the intake due to the cutter size... couldn't get more in there.. but the exhaust is a full radius. My exhaust valves are backcut, necked, and polished Ferrea's.

 

[TurfBurn]

Read and learn
http://e30m3performance.com/tech_articles/engine-tech/flow-1/index.htm

Great article!!! Nice read and good to see another confirmation of exactly what I was saying . They put a lot of work together to make the flow bench numbers useful in that. But I can guarantee you that 90% of the shops out there don't go anywhere NEAR that level of tech and thus you'd have nothing of use from that set of flow numbers other than "bragging rights." Great post JDM!

 

[TurfBurn]

I know a lot of you might not like it, but this is a great book. I've read it a couple times all the way though. There is a ton of great info everything from cams to valves to turbos. It may be Honda biased, but the info is very generic.

https://www.amazon.com/dp/ (commissions earned)

One of the best books out there that I highly recommend is by A. Graham Bell and it is called Performance Four Stroke Tuning. Fantastic book.

 

[Spooled]

This is a great point. I love how they show the importance of "time at lift" and how it affects where you want to concentrate on (low lift vs. high lift). If you have a cam with a long duration and a flat high section, you want to focus on getting good flow when the valve is wide open.

Edit: http://e30m3performance.com/tech_articles/engine-tech/flow-1/chart-4.htm

That page.

 

[TurfBurn]

One of the other things they briefly mention in the beginning of the article is that keeping your port speeds up is often even better than higher flow... although you ideally want both.. but the improved mixture and better inertia makes a significant difference in performance.

 

[Chojin]

Ahh, yes, I almost forgot; I also made these charts to indicate flow rate requirements at various levels of boost... It was a while ago, so I nearly lost them.

		Boost (PSI)
 
		0	5	10	15
 
	7	189.84	254.39	318.93	383.48
R	6	162.72	218.04	273.37	328.69
P	5	135.60	181.70	227.81	273.91
M	4	108.48	145.36	182.25	219.13
*	3	81.36	109.02	136.68	164.35
1	2	54.24	72.68	91.12	109.56
k	1	28.13	37.69	47.26	56.82
 
		1	1.34	1.68	2.02
 
		Pressure Multiplier

FYI -

Based on that chart, a VE of ~76.6 % is being used.

-Shawn

 

[flat_black]

Forgot to mention that... Though I seem to remember using 80%... It seems about accurate for the stock engine, however.

 

[TurfBurn]

VE changes significantly based on rpm's though... and the turbo also affects the VE.. so it's a hard estimate... it could be anywhere from 60% to over 115%

 

[Chojin]

VE changes significantly based on rpm's though... and the turbo also affects the VE.. so it's a hard estimate... it could be anywhere from 60% to over 115%

O.K. -

To avoid confusion, here's the formula I use to calc. engine airflow requirements:

cfm = [(Engine Displacement)*(VE)*(RPM)*(Pressure Ratio)] / 5660

Taking the numbers from the chart above and solving for VE will give ~76.6 %.

Although this formula over-simplifies things a bit (by not varying VE by RPM), it does give a pretty close approximation.

ex: [(2)*(76.6)*(6000)*(2.02)] / 5660 = ~328 cfm
ex: [(2)*(76.6)*(6000)*(1.68)] / 5660 = ~273 cfm
ex: [(2)*(76.6)*(6000)*(1.34)] / 5660 = ~218 cfm
ex: [(2)*(76.6)*(6000)*(1.00)] / 5660 = ~162 cfm

In order to use this information on a compressor map, you must convert cfm to lb/min:

lb/min = cfm / 14.27

ex: 328 cfm / 14.27 = 22.99 lb/min
ex: 273 cfm / 14.27 = 19.12 lb/min
ex: 218 cfm / 14.27 = 15.25 lb/min
ex: 162 cfm / 14.27 = 11.38 lb/min

Hope this helps.

-Shawn

 

[Rainman]

You guys f___ing rock! This is some educational s***! Subscribing.

R

 

[TurfBurn]

Shawn, I was pointing out that typically any given vehicle will vary in VE over different RPM's. Typically more efficient toward lower RPM's. So in other words, our cars may run around 95% efficient at 3,000 rpm's... but drop off to only around 75% at 6,00 rpm's. Also, modern vehicles, especially imports, tend to run more in the 80%+ range for efficiency, often well over 90%. Add PandP and the like to it and it is possible through inertia charging and the like to actually get over 100% efficiency.

EDIT: So if anything, your graph is probably conservative for the actual throughput of the motor. AND I see you did state right in your last post that you realized VE varied... my bad!

 

[twilightprotege]

FYI, my ported head reads like this for VE :

rpm---VE(old intake cam)--------VE(new intake cam)
2000--81.8%--------------------85.6%
2500--79.1%--------------------85.3%
3000--76.3%--------------------83.7%
3500--79.9%--------------------87.0%
4000--85.9%--------------------89.9%
4500--89.9%--------------------91.8%
5000--93.2%--------------------93.2%
5500--95.1%--------------------93.6%
6000--95.6%--------------------92.6%
6500--95.4%--------------------91.2%
7000--96.4%--------------------89.8%

and then it drops off from there.

 

[TurfBurn]

Wow, I'm surprised by the drops in efficiency on the high end... the cam was definitely catered to your low/midrange. But good to see those values are in the ballpark that I expected... the head was PandP'ed, but any valve work when those numbers were taken?

 

[twilightprotege]

i have a sheet somewhere that says what exactly they did, but it was a pretty basic job. port and smooth, 3 angle valve job, couple of other things i cant remember off the top of my head.

those numbers too were taken from the dyno2003 desktop dyno program. no idea how accurate they are. also those numbers are with my cam gears at 0 0 (where they are now) and also i have lash on the intake at the moment of 0.019" instead of acceptable range of 0.009-0.012". it's no drama coz the ramp rate is very slow at first, but i'd like to get the lash down to 0.012" which will add a bit back to the higher end.

 

[TurfBurn]

I've got my lash running at 10-11 thousandths right now.. with 1 tappet at 9, and two at the high side of 11, otherwise they are almost all perfectly on nominal. I had my exhaust side ports opened up pretty massively, and the intake side was largely just cleaned up and opened a little bit and knife edged a bit to help in flow division. But I don't think being slightly over 90% is at all questionable, so it seems the dyno thing did a decent job.. how is it comparing to your chassis dyno numbers for outputs?