EFI University
Electronic Fuel Injection Tuning Forum
|
| View previous topic :: View next topic |
| Author |
Message |
Pantera EFI
Joined: 12 Feb 2005
Posts: 1268
Location: So. California
|
 Posted: Wed Mar 01, 2006 8:18 am Post subject: Injector Timing |
 |
|
Which is better,
To specify the " start of injection " then add PW the change the closing point ?
To spectfy the " end of injection " then add PW the change the opening point ?
To calibrate both points ?
What are the benefits of each method ?
What are the EMS systems that can do this ?
Lance |
|
| Back to top |
|
 |
Hic
Joined: 09 Jan 2004
Posts: 152
Location: Zagreb, Croatia
|
| Posted: Wed Mar 01, 2006 12:08 pm Post subject: |
|
|
| my vote is to end of injection angle. Let the ecu calculate start, and always end in angle you choose. |
|
| Back to top |
|
|
MANOS
Joined: 23 Oct 2004
Posts: 27
|
| Posted: Wed Mar 01, 2006 12:17 pm Post subject: |
|
|
maybe selecting end of injection would be better. Because as pw increases the end of injection will "go" past the inlet valve closing if we use non altering start of injection,so remaining fuel will be useful for the next cycle waiting in the back of inlet valve. good for fuel economy and emmisions.i ve read that batch fire injection with the same afrs and advance makes about 3% more power ,can someone explain this?any ideas? 
thanks for attention. |
|
| Back to top |
|
|
d_sheffs
Joined: 04 Aug 2004
Posts: 111
Location: Las Vegas
|
| Posted: Wed Mar 01, 2006 10:02 pm Post subject: |
|
|
obviously being able to control delivery at either end is advantageous. the question is, will the engine you are working on benefit from it? I can not believe that every engine in various applications will reap benefit from this.
as far as what manufactures offer this, I would like to know. |
|
| Back to top |
|
|
rotorhead
Joined: 12 Aug 2004
Posts: 340
Location: Adelaide South Australia
|
| Posted: Sun Mar 05, 2006 12:51 am Post subject: |
|
|
Can anyone quantify the best way to tune injector phasing?
Are we targeting leanest or richest mixture at given pulsewidth vs engine rpm?
Or are we just targeting max torque injection angle?
Is it a compromise between both?
I guess it is a function of what the car is used for.
Can anyone also define how this would be advantageous in a rotory/2 stroke where there are no valves involved (just a rotor/piston closing the port).
Cheers 
_________________
13BT RX3 sedan pure circuit (race)
13BT Courier Ute (work)
MY97 WRX Hatch (wife)
HSV TWR Group A Commodore 304 (show)
And now a Mazdaspeed 6 MPS |
|
| Back to top |
|
|
Pantera EFI
Joined: 12 Feb 2005
Posts: 1268
Location: So. California
|
| Posted: Sun Mar 05, 2006 10:52 am Post subject: PW vs Injection Angle |
|
|
What I have found, when any engine is throttled (low PW) it is better to
" inject " fuel during the highest airflow speed within the port. When RPM increases the correct
" start " of injection point becomes earlier in the cycle. When airflow increases more fuel may be injected into the cylinder
" later " into the cycle.
What I am looking for is highest torque/lowest BSFC at part throttle, with a fixed throttle angle.
I agree that when a EMS may not be capable of this type of calibration,
it may be better to just " batch fire " the injectors.
Rotory/two stroke engines need to inject fuel " each " crankshaft revolution.
Lance |
|
| Back to top |
|
|
yawpower2
Guest
|
| Posted: Wed Mar 22, 2006 9:46 am Post subject: |
|
|
Opinion stated below.
It only seems sensible to me to center the injection event over the point of maximum velocity in the intake runner. That being said, center of injection would make me quite happy. Motec is not interested, or maybe it is just quite hard to accomplish in the software.
My approach prior to Motec's version 3 software was to Tune the fuel curve, and then use logged data to determine pulsewidth at all points on the map. This data would then get dumped into a spreadsheet to determine the end of injection point that would result in centering the injection event over the effective intake event. (Defined by me as intake "valve open" minus overlap.) This will delay the injection event from the point of maximum intake velocity slightly, but it seems to me to be the best (Easiest) compromise.
Once the injection timing was tweaked, the fuel curve normally required a bit of touch up.
Thanks to version 3, we can now specify injection timing in a table where the x axis is rpm, and the y axis is pulsewidth. A table can be generated in Excel using engine cam/port specs, and the ECU will now automatically center the injection event over the intake event.
In theory at least. Have done all the math, but probably will not get to try it in a real engine for a week or so.
In reality, this is much more complex than it seems. I have to wonder if it is a good idea at all to have an injection event which is much shorter than the intake event. It seems likely that a 2msec injection event in the middle of a 50 msec intake event would lead to a highly stratified charge. What does that do to ignitibility? Or flame travel? For now, we just dump it in, and hope that turbulence in the runner and chamber do the work of homogenizing the mixture.
Maybe a variable flow rate injector would be best so that air and fuel enter the chamber in a constant ratio, regardless of intake velocity.
Gee...where can I buy one of those???
The more questions I ask, the more confused I become.
Just something to chew on. Would love to hear comments from anyone with a brain, and a curious interest in injection timing.
PY |
|
| Back to top |
|
|
APEX Speed Technology
Joined: 13 Nov 2004
Posts: 726
Location: Venice, CA
|
| Posted: Wed Mar 22, 2006 11:27 am Post subject: center of injection |
|
|
| Quote: | | It only seems sensible to me to center the injection event over the point of maximum velocity in the intake runner. |
Sounds reasonable at first, but it turns out that is not usually the case. One time long ago, I undertook the same kind of approach you have outlined, using cam profiles to determine maximum intake air velocity. On the dyno, it did not work, and max power was actually realized through emperical testing. Recently, friends at Sunbelt racing engines ran into the same thing with a Miata.
Eariler this year, I was working with some engineers from the Jaguar Formula 1 program, and we discussed this. Their experience showed the following (from my feeble, overtaxed memory...)
1. Often, your injectors are open longer than intake valve duration, so by centering injection with valve timing, you are pooling fuel on the back of the valve before it opens then dealing with reversion afterwards. This results in inconsistant atomization.
2. Becuase air has mass and the air/fuel mixure is far from homogeneous, the maximum intake air velocity does not always correspond with valve position.
3. Under some conditions, it has been found that to inject completely before the valve opens actually increases power, as the hot valve helps atomize fuel more completely.
Basically, the premise that the best time to inject is during maximum intake air velocity derived from cam profile is an oversimplification. My experience and theirs both agreed on this - the best way to set it is emperically on the dyno.
_________________
Neel Vasavada
Apex Speed Technology
2947 S. Sepulveda Blvd
Los Angeles, CA 90064
310-314-2005
www.apexspeedtech.com |
|
| Back to top |
|
|
yawpower
Joined: 22 Mar 2006
Posts: 17
Location: Phoenix, AZ
|
| Posted: Wed Mar 22, 2006 2:29 pm Post subject: |
|
|
That's good stuff Neel. Hopefully others will chime in on the subject.
I think it's fair to say that most of us don't spend enough time in this area, because it is so time consuming to optimize injection timing at every point on the map.
I have put a lot of thought into this lately, and have been looking for a way to at least get in the ball park mathmatically.
In the past, I have used this approach, and found that at least at idle, and part throttle, the end result was within about 10 degrees of experimental optimum. This only relates to the Mazda rotary engine. I have no experience attempting to optimize injection timing on a piston engine.
| Quote: | | 1. Often, your injectors are open longer than intake valve duration, so by centering injection with valve timing, you are pooling fuel on the back of the valve before it opens then dealing with reversion afterwards. This results in inconsistant atomization. |
I have not put much effort into full throttle high rpm timing for that reason. It just didn't seem to make much difference on the dyno. I am working on a project right now where the injectors are sized so that the injector pulsewidth easily fits into the intake window even at peak power.
This is a bit easier on a rotary because our intake cycle is 270 degrees long, (TDC intake, to BDC intake) so for a given rpm, we have 50% more time to inject fuel.
It will be interesting to see how injection timing affects the power in that case.
| Quote: | | 2. Becuase air has mass and the air/fuel mixure is far from homogeneous, the maximum intake air velocity does not always correspond with valve position. |
While I can't prove it, and I don't expect you to believe it, my gut feeling is that this will have little effect on optimum injection timing. At least in my application. For the most part, I consider air to be non-compressible, as long as the velocity is sub-sonic.
I never thought about the relationship in a piston engine, but in a rotary, the point of maximum rate of change in chamber volume is also the time when the port is completely open.
| Quote: | | 3. Under some conditions, it has been found that to inject completely before the valve opens actually increases power, as the hot valve helps atomize fuel more completely. |
I think this is quite important, and the situation is very different on a rotary. (I should have noted in the previous post that I am working on Rotary engines.)
I can see the value of letting the fuel puddle up on the intake valve. The heat is probably helpful, and it may also help to have the fuel "shear" past the valve/seat in the early stage of valve opening.
In a rotary, getting the fuel to the port/chamber interface any time the port is not completely open results in fuel making its way to the side of the rotor, and from there, it is an easy shot past the oil seals, and into the sump. In practice, this may be a non issue, but in theory, it certainly seems like a bad thing.
There is one other issue in a rotary that makes early fuel delivey problematic. The air/fuel mixture must travel upwards to make its way into the chamber, rather than simply flowing down past an open valve.
It seems to me that letting fuel puddle at the bottom of the port is a bad thing. It is not at all unusual to remove the intake manifold from a side port rotary, and find a puddle of fuel sitting in the bottom of the runner.
Mazda even went to the trouble of placing air spray jets at the bottom of the primary runners to help the situation in the latest Renesis engine.
I'm realizing now that injection timing in a rotary may be a very different issue as compared to a piston engine.
Still, I'm interersted to hear what others have to say.
Thanks for your input.
PY |
|
| Back to top |
|
|
APEX Speed Technology
Joined: 13 Nov 2004
Posts: 726
Location: Venice, CA
|
| Posted: Wed Mar 22, 2006 2:43 pm Post subject: rotary |
|
|
| Quote: |
I think this is quite important, and the situation is very different on a rotary. (I should have noted in the previous post that I am working on Rotary engines.)
|
And of course, I completely spaced on the fact you were talking about a rotary...
| Quote: |
While I can't prove it, and I don't expect you to believe it, my gut feeling is that this will have little effect on optimum injection timing. At least in my application. For the most part, I consider air to be non-compressible, as long as the velocity is sub-sonic. |
The problem is not about the compressibility of the air (though it is far from being non-compressible) or the velocity. The problem I've seen is that your mixture is far from homogenous. To assume that the transport propoerties of the air are the same as those of the fuel could be invalid.
_________________
Neel Vasavada
Apex Speed Technology
2947 S. Sepulveda Blvd
Los Angeles, CA 90064
310-314-2005
www.apexspeedtech.com |
|
| Back to top |
|
|
furpo
Joined: 06 Aug 2004
Posts: 1082
Location: Mt Maunganui, New Zealand
|
| Posted: Wed Mar 22, 2006 3:35 pm Post subject: |
|
|
well i am jumping in on a topic which is well over my head so i will avoid the issue and bring up another factor.
before i started uni i got quite serious with the development of my home made flow bench because i beleived there was a better way to optimise a cylinder head than making static measurements. i ended up basically combining my flow bench measuring equipment onto my dyno. i could then measure mass air flow per cylinder in relation to crank angle, valve lift, engine torque, engine rpm (and what ever else i felt like logging). i have sinced saved this project for a rainy day but not before i found a very interesting phenomenon.
peak intake velosity is affected by ram charge and is therefore and function of engine speed and intake volume and length. i noticed at some points (lower) in the engine speed that peak velocity would lag by as much as 30 crank deg. i also noticed that at standing wave (intake port) engine speed that the peak velocity point would change dramatically compared to the peak velocity point for engine speeds quicker or slower. such veriations much have an effect on the injection timing.
the only engine i tested was my 1293cc BMC A-Sereies as this is what i was racing at the time. i never managed to test a non siamiesed intake port engine.
i hope that does not through a cat in with the birds.
roger
_________________
don't mind me, i always need help |
|
| Back to top |
|
|
yawpower
Joined: 22 Mar 2006
Posts: 17
Location: Phoenix, AZ
|
| Posted: Wed Mar 22, 2006 3:43 pm Post subject: |
|
|
| Quote: | | The problem is not about the compressibility of the air (though it is far from being non-compressible) or the velocity. The problem I've seen is that your mixture is far from homogenous. To assume that the transport propoerties of the air are the same as those of the fuel could be invalid. |
I see. I didn't get your point. My reference to "incompressible" Just meant that the delay from point of maximum pressure drop in the cylinder, and maximum pressure drop in the manifold was insignificant because we are in the range of "incompressible" flow.
Transport properties....There's a big issue. I guess the question is, if we were trying to center injection over intake, how much would we have to lead the injection to account for that? If we lead too much, the fuel has a prime opportunity to puddle, or at least coat the runner walls, and not get picked up properly by the incoming air. If we don't lead enough, the fuel may end up puddled on top of the valve (Or side of rotor) anyway.
It may be as you suggest, that the whole situation is so complex that trial and error is all there is. Still, when you start with a fresh project, you can do a few calcs, and come up with a decent starting point for a fuel map.
It would certainly be nice to do the same with injector timing.
On the little round motors, I can calculate something very close to optimum, as long as the injection event is a fair amount shorter than the intake event.
I'm curious about the engines you work on. Do you find that there is a typical range that is close to optimum, at least in the idle/part throttle region where the pulsewidth is small compared to the intake event?
Or...is optimum timing all over the place from one engine to the next, even at low duty cycles?
If you find that there is a relatively small range that "optimum" fits into at low duty cycles, is it possible that there is also a valid "range" if the injectors are big enough that the injection event is always considerably smaller than the intake event, even at high rpm/load?
(How's that for a run on sentence?)
Whatever the case, I know that my understanding is limited, and I assume that is true for many others. This seems to me to be the "mystery" area of injection tuning.
If I can bounce this off enough people, sooner or later, the light bulb is going to go off over my head.
Has anyone out there tried using oversized injectors and truly optimizing the injection timing? Better yet, done that, and then compared it to a setup using smaller injectors that were pushing the limit on duty cycle?
Maybe the injection timing issue is very different in these two situations.
Neel, have your own experiences shown any trends?
PY |
|
| Back to top |
|
|
EFIOZ
Joined: 03 Aug 2004
Posts: 407
Location: Melbourne, Australia
|
| Posted: Wed Mar 22, 2006 6:11 pm Post subject: |
|
|
I'd assume it would be easier to empirically calculate this on a rotary (or even a two stroke) as a port actuated intake would be a lot easier to predict than a camshaft/valve mechanism. I say that mostly as the valve doesn't quite follow the camshaft profile in the same manner throughout the rpm range. Could this be why the only way to get it is empirically?
Does the injection timing become more vital as you go to oversized injectors? I've done a couple of engines lately that ran quite high fuel pressures with fairly small injectors. The theory is the smaller injector is more accurate and the higher pressure gives better atomisation. I haven't done anything back to back yet but I think I'll get the opportunity in the next month or so. BTW, I pinched that theory from a couple of discussions I had with Chip from Elanpower about the Ferrari V12's they did for the ALMS.
_________________
Competition Systems
Melbourne, Australia
www.compsystems.com.au |
|
| Back to top |
|
|
APEX Speed Technology
Joined: 13 Nov 2004
Posts: 726
Location: Venice, CA
|
| Posted: Wed Mar 22, 2006 8:45 pm Post subject: chip and rotaries |
|
|
First off, I think Rotaries are a great testbed for looking at this issue due to the smaller number of variables. I think this would be great to investigate.
My experience shows me one things for sure - idle and low speed timing can be quite different from mid-high speed/load. Which is to be expected.
Pat, you know Chip? Chip defected to Porsche for 6 months or so before going back to Elan before Sebring. Elan has some really, really nice dynos.
_________________
Neel Vasavada
Apex Speed Technology
2947 S. Sepulveda Blvd
Los Angeles, CA 90064
310-314-2005
www.apexspeedtech.com |
|
| Back to top |
|
|
Pantera EFI
Joined: 12 Feb 2005
Posts: 1268
Location: So. California
|
| Posted: Wed Mar 22, 2006 9:20 pm Post subject: ECU-882 " injector timing " |
|
|
What the ECU-882 does is the specify the " start " of injection in degrees.
The accuracy is 1/4 of a degree when a 60 tooth target is provided.
Then as the RPM increases that point is advanced, similar to the " spark instant ".
I know this may not be " perfect " though it meets most of the above requirements stated in the previous remarks.
Lance |
|
| Back to top |
|
|
EFIOZ
Joined: 03 Aug 2004
Posts: 407
Location: Melbourne, Australia
|
| Posted: Wed Mar 22, 2006 9:53 pm Post subject: |
|
|
I only know him from numerous phone conversations. He's quite a clever guy. What's he doing at Elan now?
_________________
Competition Systems
Melbourne, Australia
www.compsystems.com.au |
|
| Back to top |
|
|
rotorhead
Joined: 12 Aug 2004
Posts: 340
Location: Adelaide South Australia
|
| Posted: Thu Mar 23, 2006 2:54 pm Post subject: |
|
|
Hey Yaw,
Are you working on rotors with OEM intake manifolds?
Obviously in this case the primary injectors are right in the inlet port so there is less delay to get into the port compared to the secondary injectors - further away. The secondary runner is also a fair bit larger than primary so air speed may be less depending on throttle angle.
This is difficult to incorporate into injector timing mapping as at any rpm you could be running on the 2 primarys or all 4, which requires a bit of a different strategy. It depend if you are just target peak BSFC at WOT or other. I am interested to hear your strategy. I dont spend much time on this with rotors usually due to the long duration and the position of OEM injectors (and because they run so damn good anyways!!)
I have a 13BT with quad 35mm throttle bodies in my RX3 which I have an E6K on. I have to run batch fire due to all 4 ports working simultaneously (engine is siamese ported). I will soon be finalising the tune (upgraded from F9A) and will spend a bit of time on the injector timing to see what happens in this application. I will keep you all posted.
Any predictions what I will find?
Cheers
_________________
13BT RX3 sedan pure circuit (race)
13BT Courier Ute (work)
MY97 WRX Hatch (wife)
HSV TWR Group A Commodore 304 (show)
And now a Mazdaspeed 6 MPS |
|
| Back to top |
|
|
yawpower
Joined: 22 Mar 2006
Posts: 17
Location: Phoenix, AZ
|
| Posted: Mon Mar 27, 2006 9:04 am Post subject: |
|
|
Been out of town for several days. Glad there was some activity while I was gone. This forum seems like a gold mine of information. I can't believe I just found it last week.
To Furpo, all I can say is WOW! That's good stuff, but unfortunately, you have twisted my brain into a knot. This was a lot easier for me to grasp before your dose of reality. It reinforces Neel's assertion that emperical testing is the only way to get it right.
From Neel | Quote: | | First off, I think Rotaries are a great testbed for looking at this issue due to the smaller number of variables. I think this would be great to investigate. |
I never thought of it that way, but I can see your point. I'll be playing around with this in the next few months. I think it's wrong to assume that I can mathmatically determine optimum, but if I could at least get close, I would be quite happy. If nothing else, It would at least simplify the final "optimizing" which would be done empirically. It's interesting to me that (At least in my experience) a change of 10 degrees or so is required to see any difference.
From Rotorhead | Quote: | Hey Yaw,
Are you working on rotors with OEM intake manifolds? |
Sometimes. Recently, everything has been for a class that requires stock manifolds. SCCA EP, Grand Am cup, etc.
| Quote: | I have a 13BT with quad 35mm throttle bodies in my RX3 which I have an E6K on. I have to run batch fire due to all 4 ports working simultaneously (engine is siamese ported). I will soon be finalising the tune (upgraded from F9A) and will spend a bit of time on the injector timing to see what happens in this application. I will keep you all posted.
Any predictions what I will find? |
I wish I had a clue. I have done very little with turbos. I have wondered recently if the situation may be different in a turbo motor. With the intake being blazing hot, maybe it is better to have a high duty cycle, and benefit from the intake charge heating the fuel for better atomization. Or maybe not, I really don't know. I guess that's why I started this thread. Too many questions, and not enough answers.
| Quote: | | This is difficult to incorporate into injector timing mapping as at any rpm you could be running on the 2 primarys or all 4, which requires a bit of a different strategy. |
This may be simplified a great deal with V3. Let's just assume (Right or wrong, for the sake of discussion) that we will get best results by centering the injection event over the intake event. accomplishing this is easy with V3. We just do a quick bit of number crunching, and base injection timing on FEPW on the y axis, and rpm on the x axis. The ECU will then always center the event, regardless of secondary balance.
I should mention that this works great on my laptop, but I haven't loaded the file, and run the engine yet. Simon at Motec claims that this will work just fine. I'm a little worried that given the opportunity to use any channel on the axis might make the ECU flip out with certain combinations.
I will be trying this out on a daily driver as soon as I get a weekend at home.
For now, I have made a few basic assumptions, and I will be working to prove or disprove those assumptions.
1. Injecting fuel during overlap is bad. At some points in the operating range, the engine will likely be overscavenging, and the result will be fuel in the exhaust. In the case of a rotary, there is a big exhaust belch up into the intake when the apex seal first clears the exhaust port. This would result in fuel being blasted up into the intake manifold, with the "standoff" that we typically see on carbureted engines. This can't be good.
2. Centering the injection event over the point of maximum runner velocity is best, as it gives us the best shot at getting the fuel cleanly into the chamber.
As I said before, these are assumptions.
So far, centering the injection event over the "theoretical" point of maximum runner velocity has served me well at idle/light throttle, but it all gets fuzzy once the duty cycle gets big.
Any input is appreciated.
PY |
|
| Back to top |
|
|
furpo
Joined: 06 Aug 2004
Posts: 1082
Location: Mt Maunganui, New Zealand
|
| Posted: Mon Mar 27, 2006 8:15 pm Post subject: |
|
|
| yawpower wrote: | | To Furpo, all I can say is WOW! That's good stuff, but unfortunately, you have twisted my brain into a knot. This was a lot easier for me to grasp before your dose of reality. It reinforces Neel's assertion that emperical testing is the only way to get it right. |
knots are fun
i think that your theory rests on your 2nd assumption. this is very easy to prove or disprove in an emperical manner. at uni we use little probes (similar to a hotwire maf) that will measure air speed at verious points along the wire. all you need is to place these in your port and emperically tune your injetor timing on the dyno taking note of peak port velocity vs crank angle.
if you want i can find the brand name for the probes
roger
_________________
don't mind me, i always need help |
|
| Back to top |
|
|
yawpower
Joined: 22 Mar 2006
Posts: 17
Location: Phoenix, AZ
|
| Posted: Mon Mar 27, 2006 8:34 pm Post subject: |
|
|
That would be great. That would also "untie" the knots. Thanks you for the help.
PY |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
©2007 EFI University |
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
