What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

uneven intake airflow corrective action options

L

Larry DeCamp

Well Known Member
My question got lost in all the superb discussion about cutting edge FI activity. Dan H made a convincing argument that intake imbalance is a composite of manifold, runners , heads and exhaust characteristics. the vertical sump is noted to probabably be the worst case.

Don R reported some very encouraging numbers (egt balance ) with mechanical FI. One post referenced a scenario where in a significant CHT imbalance was improved buy manipulating the rubber gaskets in the intake system.

My question is, Why not use restrictors in the tubes between the sump and head to "tune" the air flow rate and thus optimize egt spread. For a carb engine, it would be an option although not as convenient as nozzle change for AFP or pushing buttons for EFI.
 
More Data Please

Larry DeCamp said:
My question got lost in all the superb discussion about cutting edge FI activity. Dan H made a convincing argument that intake imbalance is a composite of manifold, runners , heads and exhaust characteristics. the vertical sump is noted to probabably be the worst case.

Don R reported some very encouraging numbers (egt balance ) with mechanical FI. One post referenced a scenario where in a significant CHT imbalance was improved buy manipulating the rubber gaskets in the intake system.

My question is, Why not use restrictors in the tubes between the sump and head to "tune" the air flow rate and thus optimize egt spread. For a carb engine, it would be an option although not as convenient as nozzle change for AFP or pushing buttons for EFI.
Click to expand...

I'm not sure we can draw any solid conclusions on the matter at this time. We have a few diverging data points and cases and some conjecture.

In Clark's case, 2 different sumps tried and swapping the "hot" and "cold" injectors made no difference. Why the huge imbalance here? What are other Bendix and AFP users seeing with the same combination of parts? Is there a big piece of casting flash in one of Clark's ports? I don't know the answer here.

Interested to see FI data from someone with flow matched heads, equal length exhaust tubes and the same sump.

For carbs, I think we almost always see large GAMI spreads without carb heat. Flowing wet mixture through unequally shaped runners with a small plenum almost universally has to be bad. I'd like to see some carb folks post their GAMI spreads here, with and without carb heat.

Carb users often report throttle angle mixes up the EGT spreads too. Is that what most people see?
 
It's likely unpossible

The engine in my plane is the TCM IO-550 out of a Cirrus. VERY good induction system, with factory balanced injectors. The thing is designed to run LOP throughout it's life.

All this magic happens at 2500RPM. It does not happen at 2200RPM, nor does it happen at 2700, tho at 2700 it is close. I think that's kinda strange for an engine that has an effective RPM range of only ~700RPM, don't you think?

Looking at the induction on my wife's hemi - that setup is likely to be what we want if we are serious about even airflow thru some range of RPMs. Given that the oil sump is where it is, getting airflow corrected might not be so easy. Kevin's sump on the R motor might be correct - or close - I have not seen what it takes to get the GAMI spread right.

The restrictors mentioned by the OP might do the trick, but at a reduced power level (think NASCAR restrictor), and probably at a specific RPM.

One question I have not seen addressed by the folks who have tuned for an even GAMI spread: does the tuning work well over a wide RPM band? I recall tuning the 540s for 2300RPM, but I never took a look at higher or lower RPMs..

Carry on!
Mark
 
Since the original thread started some time ago, many people have contacted me to say that their mechanical FI with tuned injector sizes runs a close GAMI spread at one rpm/ MP range but often does not at other rpms/ MP.

What are other people's experience with that? This would be a good place to share data and experiences. If you can, let us know your setup specs- AFP, Bendix, EFI, carb, other, sump type, engine model, exhaust type, flowed matched heads etc.

I'd like to understand the the likely causes and magnitude of imbalances that exist here.
 
If you have a Vetterman exhaust, give Clint a call and talk with him about this topic. They have done lots of dyno testing with different setups.

I recently had a discussion with him when trying to decide if I would go with a cold air induction system and associated trombone exhaust from him versus sticking with the vertical intake and using the less expensive cross-over exhaust. Clint said the cold air induction primary improves performance because of the increase in plenum area; not the temperature of the air. I also had a discussion with Don from AFP about the same decision.

In the end, I opted to save a few hundred bucks, sticking with the vertical sump but using an elbow and spacer from Don to point the FM-150 forward (similar to the IO-550 setup) and using the Sam James intake, which provides more pressure to the induction system versus the snorkel.
 
Ross, a detail please. The new software allows a discrete injector trim setting for each point on a MP/RPM map, or a single +/- trim value for each injector applied across the entire range?

For example, an individual injector can be set with a +2 at 2400 RPM, and -1 at 2000 RPM? Or would both RPMs share the same trim value?
 
rv6ejguy said:
Since the original thread started some time ago, many people have contacted me to say that their mechanical FI with tuned injector sizes runs a close GAMI spread at one rpm/ MP range but often does not at other rpms/ MP.

What are other people's experience with that? This would be a good place to share data and experiences. If you can, let us know your setup specs- AFP, Bendix, EFI, carb, other, sump type, engine model, exhaust type, flowed matched heads etc.

I'd like to understand the the likely causes and magnitude of imbalances that exist here.
Click to expand...


One of the tests that Don had me performed is to measure at multiple altitudes and settings. For example, 23mp/2300rpm/3,000 agl versus WOT(21/22mp)/2300rpm/ 8,000 agl. The results were similar. I've got all the data at home if anyone really wants to see it.
 
DanH said:
Ross, a detail please. The new software allows a discrete injector trim setting for each point on a MP/RPM map, or a single +/- trim value for each injector applied across the entire range?

For example, an individual injector can be set with a +2 at 2400 RPM, and -1 at 2000 RPM? Or would both RPMs share the same trim value?
Click to expand...

The same trim percentage is applied at all rpm/MAP points. It just adds or subtracts a percentage to the base pulse width currently being outputted from the ECU.

The software was designed to allow people to optimize AFRs in all cylinders in cruise during LOP operation primarily. In Clark's case, he found ROP operation also required similar trim amounts but I am not sure if they varied rpm during those tests.

Will have some additional data next week I would guess.
 
Larry DeCamp said:
My question is, Why not use restrictors in the tubes between the sump and head to "tune" the air flow rate and thus optimize egt spread. For a carb engine, it would be an option although not as convenient as nozzle change for AFP or pushing buttons for EFI.
Click to expand...

I think, for a carb engine, your proposal would "tune" the total volume but would not alter the fuel-air ratio.
 
Bob

Wouldnt EGT be a relative indicator of AF ratio. Admittedly, this does nothing to assess the power created by different cylinders, but air/fuel is proportional to EGT, Yes? No?
 
Larry DeCamp said:
Wouldnt EGT be a relative indicator of AF ratio. Admittedly, this does nothing to assess the power created by different cylinders, but air/fuel is proportional to EGT, Yes? No?
Click to expand...

If we know peak EGT, then we can roughly correlate with AFRs however as I said previously in the other thread, other things affect EGT and some to a large degree like ignition timing and since EIs are likely changing timing with both MAP and RPM, this makes EGT vs. AFR potentially less accurate. With fixed timing, the other variables only have small effect within a short time span so could generally be ignored.

The only place we know for sure EGT is accurate is peak/ stoich, assuming rpm and MAP don't change while you are doing the lean pull.
 
rv6ejguy said:
The same trim percentage is applied at all rpm/MAP points. It just adds or subtracts a percentage to the base pulse width currently being outputted from the ECU.
Click to expand...

Thanks, got it.

So, the pulse width trim does the same thing as varying the restrictor diameter in a constant flow system...an across-the-range addition or subtraction to the delivery of that injector. There should not be any difference in EFI and CFI response to changes in engine power setting. If some internal airflow issue changes the percentage of total air delivery to a particular cylinder with changes in throttle, it should it shouldn't matter if the fuel delivery is via EFI or CFI...the EGT should respond the same.

Thinking out loud, but there is one wild card only indirectly related to power setting. CFI nozzles have air bleeds, and those bleeds are subject to changes in dynamic and total pressure in the cowl volume above the engine, as those pressures vary with airspeed. Heck, distribution of cooling plenum pressure near individual cylinders probably varies with airspeed and AOA. The EFI system knows nothing of cooling air pressures.
 
Last edited:
DanH said:
Thanks, got it.

So, the pulse width trim does the same thing as varying the restrictor diameter in a constant flow system...an across-the-range addition or subtraction to the delivery of that injector. There should not be any difference in EFI and CFI response to changes in engine power setting. If some internal airflow issue changes the percentage of total air delivery to a particular cylinder with changes in throttle, it should it shouldn't matter if the fuel delivery is via EFI or CFI...the EGT should respond the same.

Thinking out loud, but there is one wild card only indirectly related to power setting. CFI nozzles have air bleeds, and those bleeds are subject to changes in dynamic and total pressure in the cowl volume above the engine, as those pressures vary with airspeed. Heck, distribution of cooling plenum pressure near individual cylinders probably varies with airspeed and AOA. The EFI system knows nothing of cooling air pressures.
Click to expand...

I don't understand all the nuances of Bendix/ AFP injection but I've had a dozen + people tell me their systems do not have equal distribution across all MAP/ rpm ranges, including Dave Anders (one reason he wants to try EFI). I'd like to see more feedback on this. One other poster with a Conti already said as much in this thread. I think we'd be assuming that the mechanical metering system is perfect- a bit of a stretch there perhaps.
 
rv6ejguy said:
I don't understand all the nuances of Bendix/ AFP injection but I've had a dozen + people tell me their systems do not have equal distribution across all MAP/ rpm ranges, including Dave Anders (one reason he wants to try EFI).
Click to expand...

So why would a trimmed EFI be any different?

We're drifting the thread pretty badly here. The OP's question had to do with restricting one or more flow passages to balance the flows. My guess is that it could work, at one particular mass flow. Nobody does it because it goes against a desired result. Much better to figure out which ones are restricted and decrease their restriction to match the best...tuning 101.
 
Thanks Dan

That answers my question. If you don't care about horsepower ( light RV-B) , the cab could be optimized at cruse power for optimum Fuel utilization and possibly smooth operation
 
DanH said:
So why would a trimmed EFI be any different?
Click to expand...

I guess at least with this EFI trim, you can re-trim it for a different flight condition "on the fly", as it were. That doesn't seem feasible with mechanical injection.
 
Larry DeCamp said:
That answers my question. If you don't care about horsepower ( light RV-B) , the cab could be optimized at cruse power for optimum Fuel utilization and possibly smooth operation
Click to expand...

Whoa bro. Wet flow and dry flow is not the same thing. Restrictors might equalize dry flow, which could result in equal FA ratio if equal fuel is supplied at each intake port, i.e. with fuel injection. Wet flow equality (with fuel supplied at a single point, like a carb or the single point Precision injection) would only match the dry flow if all the fuel was fully vaporized...meaning it ain't really wet anymore. The mass of liquid droplets are a problem; Newton's 1st, etc.

rmartingt said:
I guess at least with this EFI trim, you can re-trim it for a different flight condition "on the fly", as it were. That doesn't seem feasible with mechanical injection.
Click to expand...

Yep. A few keystrokes to trim an individual injector is pretty cool.
 
rmartingt said:
I guess at least with this EFI trim, you can re-trim it for a different flight condition "on the fly", as it were. That doesn't seem feasible with mechanical injection.
Click to expand...

Yup, that's the big difference. We can make the GAMI spread zero from the cockpit in 45 seconds at ANY power setting.
 
Exhaust Primary Tube Length Effects

I'm reviving this old thread again to ask for more input on what people are seeing for GAMI spreads as related to exhaust primary tube length differences. Some testing is indicating that this may be a significant factor in "airflow imbalance" between cylinders, where we already have good intake designs with equal induction tube lengths such as the Sky Dynamics setups and flow matched heads.

The primary tube length will theoretically alter scavenging of the cylinder and intake pumping at different rpms.

Looking for data on what intake you have, flow matched heads or not, injection system used, GAMI spread and primary exhaust tube length from head to collector or end of the pipe.
 
rv6ejguy said:
Looking for data on what intake you have, flow matched heads or not, injection system used, GAMI spread and primary exhaust tube length from head to collector or end of the pipe.
Click to expand...

Stock Van's IO-540 with Precision injection and stock exhaust system. Out of the box the GAMI spread was over one GPH and very rough on my LOP tests. After replacing 4 of the 6 stock 0.028" injectors nozzles of various sizes from Air Flow Performance, GAMI spread is 0.0 to 0.2 GPH. The SkyView data recording provides a superb tool to balance injectors.

I run LOP 90% of the time.

Carl
 
Carl Froehlich said:
Stock Van's IO-540 with Precision injection and stock exhaust system. Out of the box the GAMI spread was over one GPH and very rough on my LOP tests. After replacing 4 of the 6 stock 0.028" injectors nozzles of various sizes from Air Flow Performance, GAMI spread is 0.0 to 0.2 GPH. The SkyView data recording provides a superb tool to balance injectors.

I run LOP 90% of the time.

Carl
Click to expand...

Do you know the nozzle size variations used?

Is your GAMI spread equal at 2000 rpm and say 2500 rpm?

Could you measure your primary exhaust tube lengths next time you have the cowling off?

Thanks for the data point Carl.
 
The original poste was about trying to even out the A/F ratio and charge in a carbureted engine. I do not think adding restrictors to the intake pipes would work across all throttle settings, and it would reduce total airflow and therefore power.
I have noticed on my O-360-A1A that cylinders 1&2 are closely matched, and 3&4 are closely matched. I have a theory that needs testing, but I lack the fabrication skills. I believe capping the 1&2 intake pipes and then fabricating new pipes the feed 1&3 and 2&4 together with even geometry would balance a carbureted engine as good as a FI engine - if not better.
 
rv6ejguy said:
Do you know the nozzle size variations used?

.
Click to expand...

Ross,

The Dynon SkyView data log function provided unbiased information on the fuel flow when each cylinder peaked. I also looked over a series of data runs (as in the each cross country) to make sure the information was repeatable. The standard injector nozzle for this engine is 0.028". This is my final nozzle set up for each cylinder:

#1: .028
#2: .029
#3: .027
#4: .027
#5: .029
#6: .0275

Note this is for my specific engine - your results will most likely be different.

As you can see, I worked with the 0.028" stock nozzles and made adjustment either bigger or smaller. Don at Airflow Performance suggests that, depending on your flying, you start with perhaps 0.026" and adjust from there.

Don provided valuable information on this effort as well as the nozzles themselves. His nozzles fit both Bendix or Precision fuel injection system.

Carl
 
Carl Froehlich said:
Ross,

The Dynon SkyView data log function provided unbiased information on the fuel flow when each cylinder peaked. I also looked over a series of data runs (as in the each cross country) to make sure the information was repeatable. The standard injector nozzle for this engine is 0.028". This is my final nozzle set up for each cylinder:

#1: .028
#2: .029
#3: .027
#4: .027
#5: .029
#6: .0275

Note this is for my specific engine - your results will most likely be different.

As you can see, I worked with the 0.028" stock nozzles and made adjustment either bigger or smaller. Don at Airflow Performance suggests that, depending on your flying, you start with perhaps 0.026" and adjust from there.

Don provided valuable information on this effort as well as the nozzles themselves. His nozzles fit both Bendix or Precision fuel injection system.

Carl
Click to expand...

The range of nozzles represents about a 15% difference in area and from my experience with Hilborn systems, something like a possible 12% flow difference between smallest to largest. This backs up what some others have found with EFI and certain manifolds.

Do you see the same GAMI spread at 2000 and 2500 rpm?

Still interested in your primary exhaust tube lengths to see if there is correlation with GAMI spread and RPM on those cylinders.
 
rv6ejguy said:
The range of nozzles represents about a 15% difference in area and from my experience with Hilborn systems, something like a possible 12% flow difference between smallest to largest. This backs up what some others have found with EFI and certain manifolds.

Do you see the same GAMI spread at 2000 and 2500 rpm?

Still interested in your primary exhaust tube lengths to see if there is correlation with GAMI spread and RPM on those cylinders.
Click to expand...

Typical cruise RPM is 2300 - 2400. I don't have 2000 RPM data but at 2500 RPM saw no difference in cylinder balance as compared to 2300-2400 RPM. You decide what RPM range you want for perfect balance.

Primary exhaust tube length - this is a stock system from Van's so perhaps you can get that data from them. While I'm sure specific exhaust systems can effect cylinder balance, I offer that as you look at the engine as a system the prime variable under your control are the nozzles selections. Exhaust system fit up tends to drive tube length compromises. I can tell you that the stock exhaust from Van's has worked well for me and I recommend it.

Carl
 
I've got my GAMI spread pretty close. Previously, I was able to get the spread to near zero at a fuel flow of 8 gph (fixed pitch). But at higher power settings, one cylinder was too lean. After reading some info from DanH on the fuel injection spider, I surmised that the spider was contributing to fuel flow variance at low power settings and my sizing of injectors was compensating for spider related fuel flow variance. I now have the injectors sized for higher fuel flow where the spider influence on flow rate is no longer a factor.
So, at least under some ciecumstances, GAMI spread is influenced by % power.
 
Thanks Carl.

If anyone has a stock Van's 540 exhaust exposed for measurement, it would be great if they could take out a measuring tape and see what each of the 6 primaries are in length.
 
rv6ejguy said:
Carb users often report throttle angle mixes up the EGT spreads too. Is that what most people see?
Click to expand...
I had good success drilling some aeration holes in the main jet for my carburetor - thus turning it into a "peppermill" jet. At cruise, I adjust the throttle ever so slightly to even out the EGTs. Works like a charm.
 
rv6ejguy said:
Do you know the nozzle size variations used?

Is your GAMI spread equal at 2000 rpm and say 2500 rpm?

Could you measure your primary exhaust tube lengths next time you have the cowling off?

Thanks for the data point Carl.
Click to expand...

Ross, I do not have information at various RPM but could get that at a later date.

Current configuration is

Cyl# , Primary length, Injector

  1. 23.5" 0.0275
  2. 19.25" 0.0280
  3. 17.75" 0.0275
  4. 18.25" 0.0275
  5. 23" 0.0270
  6. 31" 0.0280

Runs less than 0.4GPH spread LOP ~2400RPM


34962239736_0b635eaf52_c.jpg

34962240636_740ee05e32_c.jpg
 
Thanks Weasel, this was what I was looking for. Massive difference in lengths but no direct correlation with nozzle size.

Is this the standard sump?

Flow matched heads?
 
I have been curious about finding the optimum settings for multiple RPM's myself. I have one AFR gauge on CYL4. I'm not too keen on finding peak at 30/2700rpm to find out what the other cylinders are doing. I guess for the cost of 3 more wideband sensors and some switching I could have a better answer.

or possibly 2 of these:

http://www.innovatemotorsports.com/products/dlg1.php


Caleb
 
Last edited:
rv6ejguy said:
Thanks Weasel, this was what I was looking for. Massive difference in lengths but no direct correlation with nozzle size.

Is this the standard sump?

Flow matched heads?
Click to expand...

Standard vertical sump.

Standard ECI heads.

Another data point is that I ran this airframe for over 1000hrs with a Lycoming IO-540C4B5 with ECI cyl's which were flow matched and this exhaust system.

~0.4 or less GPH peak spread and the following injectors

  1. 0.0280
  2. 0.0290
  3. 0.0280
  4. 0.0280
  5. 0.0285
  6. 0.0285
 
You must log in or register to reply here.
Back
Top