|
-
POSTING RULES

-
Donate yearly (please).
-
Advertise in here!
-
Today's Posts
|
Insert Pics
|

05-17-2015, 09:45 AM
|
 |
|
|
Join Date: Oct 2005
Location: 08A
Posts: 9,500
|
|
Quote:
Originally Posted by rcpaisley
1. The beauty of electronic fuel injection is that you get the same fuel to all cylinders (balanced delivery) under all operating conditions from starting to full power.
|
Untrue. Flow +-3% (6% variation) is an oft-quoted tolerance for new fuel injector static flow (100% duty cycle), and large injectors running at short pulse widths tend to show significant flow variation even when the static flow rates are near identical.
The system is supplied with 60 lb/hr Siemens injectors. Let's use Mr. Carroll's engine as an example. At WOT, a BSFC of 0.52 would suggest 16.9 GPH total for 195 hp, which is roughly 101 lbs/hr total, or 25 lbs/hr for each individual injector at WOT. 25/60 describes a large injector running at a short pulse width at WOT. The pulse width would be correspondingly shorter at cruise fuel flows.
Quote:
|
This is not possible with any other type of fuel delivery.
|
Clearly precise delivery is possible. We routinely swap restrictors in mechanical FI systems to achieve any desired delivery for each individual cylinder. My own peaks all cylinders within 0.2 GPH, and a lot of that is flow meter indication dither. A first-to-last peak range of 0.2 is a 1.7% variation across all cylinders for the 390 in best power cruise(11.2 GPH), or 2.8% for LOP cruise (8 GPH) with 23 deg fixed mag timing.
Where EFI does better is total flow below about 7 GPH (less with smaller restrictors), the point at which the mechanical system is transitioning between flow division based on nozzle restriction, and flow division based on the metering slots in the flow divider.
Quote:
|
You also have a very easy way to tune the fuel delivery at different operating points - also a huge benefit.
|
Absolutely true in terms of changing overall fuel flow at the default value. It is almost always easier to make a software table change than a hard parts change. Either system can be varied from the default in flight, as both systems have a mixture knob; one is push-pull, the other is twisted. However, with a batch fire EFI, changing flow to an individual cylinder will still require swapping the individual injector, just like swapping a restrictor.
Quote:
|
2. GAMI spreads are a method to attempt to balance mechanical fuel injection systems.
|
A GAMI spread reveals the operating fuel air ratio of each individual cylinder. It merely measures a result, regardless of how the fuel was delivered.
Quote:
|
This technique does not take into account that fact that you may have (probably do have) imbalanced air flow to the cylinders as well.
|
Definitely have, unless somebody went to a lot of trouble to port cylinders and air system for matched flow. Our task is to match fuel delivery to the available air delivery.
Quote:
|
GAMI spread information is not applicable to electronic injection - you already have balanced fuel delivery.
|
Just for fun, let's humor your assertion of perfectly balanced fuel delivery squirted into a sadly unbalanced set of air pumps. The recommended EFii tuning method is to rely on a wide band oxygen sensor installed in a single headpipe, near the exhaust port, so high temperature will delay lead poisoning of the sensor. The single point measurement means that if the operator chooses to play ostrich and ignore multi-point EGT indications, he is setting fuel air ratio based on that one cylinder's air pumping capacity, and you say, the other cylinders are not likely to be the same.
It would be a lot like going back to the single-EGT Alcor panel gauge of the 70's. We assumed the rest of the cylinders were doing the same as the one we were monitoring. Ignorance was bliss.
Quote:
|
The best that can be done is to have balanced fuel delivery and a high energy ignition to ensure complete combustion even if there is an air/fuel ratio variance between cylinders. This is exactly what the EFII system does.
|
That is exactly what high energy ignition does...successfully light mixtures across a wide range of AF ratios. A good ignition will hide poor mixture, except on the torque meter or the EGT display.
Here's the catch. With wide GAMI spreads, an operator can twiddle the EFII mixture knob and park the engine with one cylinder over on the rich side (say 40 ROP, an ugly place), two more cylinders 40 LOP, and maybe a fourth cylinder 75 to 90 LOP. The first cylinder will run hot, the middle pair will be happy, and the last will be a weak sister, not making much power. But hey, it will run smooth.
The above is why George Braly and friends introduced restrictor tuning. I'd love to be a fly on the wall when he hears that your system makes EGT spread irrelevant 
__________________
Dan Horton
RV-8 SS
Barrett IO-390
Last edited by DanH : 05-17-2015 at 09:49 AM.
|

05-17-2015, 10:28 AM
|
|
|
|
Join Date: Jun 2010
Location: Sunman, IN
Posts: 2,189
|
|
Here we go!
Can't we all just get along?
Seems to me, YEARS ago, when people first started running tuned injectors and the GAMI spread first came into being, there were those who said that it was crazy nonsense because that is not the way we have always done it. Then there were the ROP LOP "discussions". Now, a "new" (not really) tech comes along, that has the potential to makes things better. What do you know? It is crazy nonsense again...at least until it becomes the new status quo. Then something better will come along, and BOOM, it will be crazy nonsense again...
Man, I LOVE this EXPERIMENTAL thing! 
__________________
Bob
Aerospace Engineer '88
RV-10
Structure - 90% Done
Cabin Top - Aaarrghhh...
EFII System 32 - Done
297 HP Barrett Hung
ShowPlanes Cowl with Skybolts Fitted - Beautiful
Wiring...
Dues+ Paid 2019,...Thanks DR+
|

05-17-2015, 10:57 AM
|
|
|
|
Join Date: Jan 2005
Location: Shorewood, WI (Milwaukee area)
Posts: 1,066
|
|
Mixture
Dan - Thanks for that exposition. My -8 is at 32 hours. Cruise CHT's within 9 or 10 degrees (320-330) on Lycon IO360M1B with 10:1 compression. EGT's vary more, with 1 and 4 being higher (1229,1180,1149,1241). All @ slightly below peak EGT, OAT 55F - the pattern holds @ widely different speeds and mixtures. Seems time to perform GAMI lean test and consider swapping/changing injectors. Would you do anything else first? Please pitch in, everyone. Thanks.
__________________
Bill Dicus
Shorewood (Milwaukee) Wisconsin
RV-8 N9669D Flying 12/4/14!
Flying Pitts S-2A, Piper Lance
|

05-17-2015, 11:19 AM
|
 |
|
|
Join Date: Aug 2006
Location: Taylor Texas
Posts: 812
|
|
Can the injectors be tuned easily/cheaply?
So Robert:
Can the system be tuned to bring it into line with the GAMI test? How might this be accomplished, and what might that cost?
I recognize that the exact same fuel delivery amount to each cylinder is not the optimal process, no matter how much effort is put into the air side of the equation, with round engines possibly being VERY good, and possibly as close to optimal as can be designed.
My involvement with your system in minimal, but it is intriguing!
Carry on!
Mark
|

05-17-2015, 11:43 AM
|
|
|
|
Join Date: Aug 2012
Location: Dallas
Posts: 196
|
|
Quote:
Originally Posted by rv6ejguy
This is a user programmable system so it's not like an OEM engineered and programmed system like Powerlink or IE2 which is fitted and programmed to specific engines. It's also different from aftermarket aviation solutions like the Eagle EMS which is around $11K.
The idea is to give users in Experimental Aviation many of the benefits offered by expensive, factory developed FADECs at a fraction of the price. One of the challenges in programming for a wide variety of experimental engines is the variety of displacements, induction, porting , cam and exhaust system differences. The OEM systems only have to work on one engine configuration and their engineers can spend many hours mapping to that end.
This article may be helpful for some people to better understand the system: http://www.sdsefi.com/aircrafttuninglyc.htm
Information sharing and feedback from users is helping to shape future improvements and developments and hopefully improve initial setups for different engine configurations.
Also, just coming available now is 8 channel data logging to PC capability to further assist in mapping the system.
|
Ross/Robert,
Is the EFII system a collaboration between you two?
It seems that Ross knows a lot about it, is why I ask. Or is the EFII just using the SDS "box"?
If so, is there an advantage of one vs the other?
Pardon the ignorance on my part
Thanks
Chris
__________________
RV-9A finshed! For sale
|

05-17-2015, 11:45 AM
|
|
|
|
Join Date: Jan 2005
Location: Shorewood, WI (Milwaukee area)
Posts: 1,066
|
|
Sorry for "drift" of my post. Re-locate if appropriate... Thanks.
__________________
Bill Dicus
Shorewood (Milwaukee) Wisconsin
RV-8 N9669D Flying 12/4/14!
Flying Pitts S-2A, Piper Lance
|

05-17-2015, 11:51 AM
|
 |
|
|
Join Date: Mar 2005
Location: Calgary, Canada
Posts: 5,766
|
|
Food for Thought
I mentioned in a previous post about the GAMI spread being essentially the same as with mechanical injection using standard nozzles. That's a reality.
The real question comes down to how the engine operates when one cylinder is running 16 to 1 AFR and another at 16.5 for example. If the ignition still lights off the mixture, it works fine though there will be a variation in torque between cylinders but this is already happening due to the different air mass flow between cylinders and in fact if you look at individual combustion pressure traces on a scope, they also vary noticeably between each firing event.
While a small EGT spread seems or is ideal in our minds, you can't change the fact that you are attempting to balance the point of peak EGT between all cylinders due to an imbalance in individual cylinder air mass flow in the first place. Changing fuel flow to each cylinder does not change the fundamental issue here which is less than optimal manifold/ head design, it primarily makes you feel better.
The cylinder with the least mass flow is likely to make the least mean torque of the 4 or 6 no matter what you do with fuel flows. Richer mixtures (relatively) when running LOP will help make a bit more torque in the lower mass flow cylinder which would help bring the mean torque variations closer together. In this case, we are talking about engine smoothness. With a stackup of a 2% lower flow injector married to a 3% higher flowing head/ runner, we get a 5% variation in AFR for that cylinder compared to the others (just an example). This jug will be making a bit more torque than the others at any given AFR. The leaner AFR will help reduce this in line with the others in fact, leading to better engine smoothness. So you see, in this case, it's not all bad.
If we had that 3% better flowing cylinder combines with a 2% better flowing injector. AFRs might be closer to the others but the mean torque variation is actually higher than the other cylinders.
I submit the goal should be to get all cylinders producing as close to same mean torque as each other. The best way to do this is equal airflow, and equal fuel flow to each cylinder, equal CR, spark happening at the same place precisely (crank triggered ignition) and so on.
I think people have become fixated on this small GAMI spread without considering the true operational ramifications. Changing injector flow rates is simply a patch for another problem. Perhaps someone should offer better manifold designs which, when combined with flow bench matched heads, would yield better overall results in mean torque between cylinders and peaking closer to each other.
That being said, we usually work with what we have unless we have unlimited budgets. Dan is right. If we have a wide spread caused by a large mismatch in AFRs between cylinders, no matter what induction and fuel metering system we use, it's a bad idea to have some cylinders running ROP and some LOP. However if they are all LOP and the high energy ignition lights off the mixture, how much does this matter? Obviously the degree matters if extreme. The true test would be to measure TAS vs. FF and engine vibration signatures.
Using the WB to tune this system works best. Averaging the AFRs doesn't matter since we have no control of individual injector flow anyway.
Regarding the dynamic injector flow vs. static flow at low pulse widths, I'll make a couple observations: First, the average duty cycle in aircraft during cruise (our main concern) is around 20-25%. This is higher than most cars cruising on the highway. Second, I have not seen big differences in these 2 figures on our flow bench nor on the test stand, running the engine at idle where pulse widths may be down around 2-3ms. The idle quality and EGT spread (engines with better intake designs and flow matches heads) is excellent but it wouldn't be if there were important variations in fuel delivered at each injection event. EFI meters fuel and atomizes it better than any carb of mechanical setup in my experience and many people who've fitted EFI to their Lycoming on Continental mention how noticeably smoother it is. Not scientific of course but it's a common comment.
I'd go out on a limb here and say that the vast majority of EFII users here have individual EGT and CHT probes attached to a modern digital engine monitor so they can see what's happening in this regard.
Hopefully this has made people think a bit more of why they are doing something and other possible results of doing that.
Last edited by rv6ejguy : 05-25-2015 at 09:50 AM.
|

05-17-2015, 12:03 PM
|
 |
|
|
Join Date: Jan 2008
Location: Upland, CA
Posts: 286
|
|
EFII SDS
Hi Chris,
We use the excellent SDS ECU in our system. We use the best available components in each portion of our system. This includes:
Walbro fuel pumps
Siemens fuel injectors
Bosch temp sensors
Tefzel wire harness
There are many other examples.
Bottom line, the EFII system uses the best proven components available throughout the system, including the SDS ECU. There are many portions of the system that we manufacture in house as well, including other electronic and mechanical components of our system. Our goal is to provide the best possible engine management system that we can.
Robert
|

05-17-2015, 12:10 PM
|
 |
|
|
Join Date: Jan 2008
Location: Upland, CA
Posts: 286
|
|
GAMI
Hi Mark,
As previously noted, the GAMI test is really irrelevant to electronic injection, you already have the fuel balance goal achieved. This allows you to focus on the next step of your engine operation, whatever that may be. For your race plane, that may be looking at getting the air delivery to each cylinder better balanced. There may or may not be much more to gain in air and fuel delivery optimization. If you learn something, let us know!
Robert
|

05-17-2015, 12:12 PM
|
 |
|
|
Join Date: Mar 2005
Location: Calgary, Canada
Posts: 5,766
|
|
Further to my rather long post, we should consider that flame speed varies with AFR considerably so ignition advance is important to achieve PCP at the optimal time. This also plays into EGT variations between cylinders where AFRs are not close.
Best economy (fuel flow vs. torque) on SI engines is obtained around 17 or 18 to 1 AFR. In the past, many magneto equipped engines simply could not reliably light off these lean mixtures and with fixed timing, even if you did, fuel economy was not optimized because the burn rate was too slow. Rough running and reduced speed was not an uncommon result under these conditions.
Integrated EFI/EI gives a lot more control of these parameters.
Our ideal engine would have equal mass flow rates of both fuel and air. Clearly to achieve that, we need injectors that flow exactly the same amount as the others in the system and we need heads and manifold designs which also do that. Given the use of OTS parts and no more work, this is not a reality. Given lots of work and specialized equipment like injector and air flow benches it can be done.
After fully optimizing the fuel and spark tables, it would be interesting to correlate and compare data from a couple different flying examples using comparable engines and airframes to get TAS vs. FF numbers and compare them to legacy mechanical FI and carb systems running LOP. The nice thing about the EFII is you can experiment to your hearts content without leaving the cockpit or getting your hands dirty.
Who's up for that?
Last edited by rv6ejguy : 05-17-2015 at 02:23 PM.
|
| Thread Tools |
Search this Thread |
|
|
|
| Display Modes |
Linear Mode
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
All times are GMT -6. The time now is 08:06 AM.
|