Author Topic: End of injection timing matters  (Read 24191 times)

bradtpt

  • Administrator
  • Member
  • *****
  • Posts: 469
  • Without data you only have an opinion
    • Turning Point Technology
End of injection timing matters
« on: May 04, 2012, 10:23:44 PM »
Here I am waving my "EOT is important" flag again.

I tuned a car today which was really sensitive to EOT (End Of injection Timing).  Engine is a Nissan CA18DET with 730cc injectors running pump fuel and a Garrett GT2860-5 turbocharger at wastegate pressure of 10lbs building to 15lbs by 7500rpm.  EOT can make a really big difference when running big injectors on small engine engines.

Boost on both runs was identical.  The red line is with EOT numbers in the low 300's, the green line is with EOT numbers in the mid to high 400's AND with 2 degrees less timing!!  You can imagine my suprise when I looked up from the run to see this difference as I expected the power to drop a bit as this car is going to get flogged and I was putting a little more knock buffer in there even though it had on the previous run survived a 250rpm/sec ramp with no knock at all.  These runs were both done at the same operating temperature, same ramp rate (500rpm/sec) about 2 mins apart from one another.  All the bad manners and bit of a cough it used to have up over 6000rpm totally gone.  It also totally transformed the transient throttle response.  It is also interesting to note on the 2nd graph that the lambda trace from the green line with EOT set correctly reads richer throughout with no change made to the fuel delivery quantity, only the fuel delivery timing.  This is due to achieving a more thorough burn as the fuel is entering the chamber in suspension and not pooling off the back of the intake valves.

Moral to the story
e1280s owners: Know this and leverage it to your advantage.
Andy:  When possible, PLEASE bring EOT adjustment to the 4x0 range!!
Regards,

Bradley Smith
Proprietor
Turning Point Technology

vagman

  • Member
  • *
  • Posts: 46
Re: End of injection timing matters
« Reply #1 on: May 04, 2012, 11:02:47 PM »
Hi Bradley,
According 1280 manual  300 deg EOT means 300 deg BTDC compression of the corresponding cylinder. Am I correct? So EOT 400 means you're ending injection even 100 deg earlier, actually one + 40 deg crank turns earlier and definitely before the inlet valve opening!? 
Ivo

bradtpt

  • Administrator
  • Member
  • *****
  • Posts: 469
  • Without data you only have an opinion
    • Turning Point Technology
Re: End of injection timing matters
« Reply #2 on: May 05, 2012, 12:01:20 AM »
You are reading the manual correctly yes, however at 400 degrees we are not injecting the fuel that much before the intake valve opens.  In the example I'm about to write, it's only about 3 deg before the valve opens if injected at 400deg.  The reason is because at 360deg BTDC compression stroke you are at 0deg BTDC exhaust stroke and this is actually the very beginning of the intake stroke.

I've just grabbed some cam specs here from another client's engine to use as an example.  This cam has 232deg duration at 0.050" lift.  The cam card reads that at 37.59 deg BTDC the intake cam lobe should have 0.006" lift.  Now that's BTDC exhaust stroke which is actually 360+37.59 degrees BTDC compression stroke = 397.59 deg BTDC compression stroke which is what the EOT numbers refer to.  It has 0.020" lift at 24.45 deg (384.45 deg) and at 0.050" lift we are at 9.28 deg (369.28 deg).  Given there is some distance between the injector nozzle and the intake valve, my EOT numbers in the low 400's now make sense as the engine will move through some degrees of crank rotation between the fuel leaving the injector nozzle and it making it to the intake valve curtain area and ultimately the combustion chamber.

Hope that helps to paint a clearer picture for you, it's easy to get confused when working in the 720 degree cycle sometimes.
Regards,

Bradley Smith
Proprietor
Turning Point Technology

stevieturbo

  • Member
  • *
  • Posts: 133
Re: End of injection timing matters
« Reply #3 on: May 05, 2012, 03:17:39 AM »
Some chat here

http://www.motec.com.au/forum/viewtopic.php?f=39&t=769&p=3583&hilit=end+of+pulse+lambda#p3583

and here

http://www.syvecs.com/forum/viewtopic.php?f=6&t=296

And a very good diagram courtesy of Link

http://www.linkecu.com/support/documentation/technical-drawings/G30%20Injection%20Timing%20Diagram.pdf

Consensus seems to favour end of injection before the valve opens. This time will need to increase with rpm, as the time window gets less.
And most suggest adjust timing until AFR's appear richest, that means the fuel that is injected, is getting burnt more completely. Although I'd imagine you also need to check hydrocarbons as well to get a bigger picture.

Never really tried it myself though.

« Last Edit: May 05, 2012, 03:21:12 AM by stevieturbo »

vagman

  • Member
  • *
  • Posts: 46
Re: End of injection timing matters
« Reply #4 on: May 05, 2012, 06:36:17 AM »
Thanks Bradley,
My English is definitely not the best I can BUT, if  I understand well enough your explanations you're getting power gains by injecting the same fuel but 100 deg earlier (in the exhaust stroke). Compared to 300 deg EOT(later in the inlet stroke) setting this means (at least to me) that most of the fuel is injected over the closed inlet valve, so actually just a small part from the end of the inj quantity is taking place when valve have some stroke/opening. With other words most of the fuel "boils" on the back of the valve and just small part is entering the chamber/cylinder directly.
So, am I wrong to conclude that your improvements in term of manners and cough are exactly because most of the fuel was injected with valve closed and later pooling off there?
Strange, according provided information EOT shouldn't change so much the things especially at high rpms. This is not the case though with your recent dyno experience.
Ivo

bradtpt

  • Administrator
  • Member
  • *****
  • Posts: 469
  • Without data you only have an opinion
    • Turning Point Technology
Re: End of injection timing matters
« Reply #5 on: May 06, 2012, 03:28:32 PM »
Thanks Bradley,
My English is definitely not the best I can BUT, if  I understand well enough your explanations you're getting power gains by injecting the same fuel but 100 deg earlier (in the exhaust stroke).
In this instance where we have relatively large injectors on a small engine, yes that is the case, we made more power injecting the same quantity of fuel about 100 deg earlier.

Compared to 300 deg EOT(later in the inlet stroke) setting this means (at least to me) that most of the fuel is injected over the closed inlet valve, so actually just a small part from the end of the inj quantity is taking place when valve have some stroke/opening. With other words most of the fuel "boils" on the back of the valve and just small part is entering the chamber/cylinder directly.
So, am I wrong to conclude that your improvements in term of manners and cough are exactly because most of the fuel was injected with valve closed and later pooling off there?
What you have described there could very well be the case.  What I haven't sat down and worked out is the theoretical delay in time or crankshaft degrees it takes for the fuel to reach the inlet valve curtain area from the injector nozzle.  This is just a theory but it could be that having the fuel "waiting" at the back of the closed inlet valve ready to enter the chamber means it is among the first air to be inducted into the combustion chamber and hence is mixed exceedingly well by the whole volume of air entering the chamber for that cycle, compared to entering midstream and not mixing as well as if it had've entered at the very beginning of the induction stroke.

Strange, according provided information EOT shouldn't change so much the things especially at high rpms. This is not the case though with your recent dyno experience.
The reason I believe EOT made such a difference on this setup is because at 6000rpm a 5ms injector pulse will only last 180 crankshaft degrees.  I can't remember the injector on duration at 6000rpm on this engine, but with 730cc injectors I'm confident it would be between 5 and 10ms given it was at about 14psi by that stage.  The reason EOT made such a big difference on this engine is because the injection duration is very short due to the flow capacity of the injectors.  If you can picture an injector half the size delivering say 20ms worth of fuel, then it's injection duration lasts the whole 720deg so changing EOT on a setup like that would yield no change at 6000rpm/14psi.  Because with this example we are seeing the same quantity of fuel delivered in half the time, timing it so that it all arrives and lands in the cylinder well mixed made a world of difference.

This is an interesting subject when you delve right into it.  I have a 4 gas analyser at work which this discussion has made me want to go to the effort of replacing the failed sensor in it so I can do some more testing when I get time.
Regards,

Bradley Smith
Proprietor
Turning Point Technology

rvengineering

  • Member
  • *
  • Posts: 625
    • RV Race & Rally Engineering
Re: End of injection timing matters
« Reply #6 on: May 06, 2012, 06:36:08 PM »
To begin with the 1280ECU dos begin of injection but we done a calculations strategies in de ECU so the ECU is set-up on end of injection over duration.

Using the inlet valve as a heater element to evaporate fuel works well so having a start of injection works best al do option B can be using smaller injectors or more uncommon change fuel rail pressure as we do on DI systems.

We run a lot of E85 projects on the 420 and 440ECU mostly staged injection as well and retarding the ignition on hi RPM sound silly but it works when you have evaporation problems with fuel. Remember un evaporated fuel can course detonation and unburned fuel can give a lean o2 reference so yes measuring HC is important.
Option can be to lower the fuel pressure if big injectors are used so you get a 75% duration over 720 degreed crank or use stage injection.
Ruud Visser Engineering.
Dr Vermetstraat 14
4681 CD Nieuw-Vossemeer.
TEL: +31(0)6 13 90 61 94
E-mail adaptronic@gmail.com
www.adaptronic.nl
www.facebook.com/ruudvisserengineering
www.facebook.com/DieselEngineTestSystems

Andy

  • Administrator
  • Member
  • *****
  • Posts: 4,319
Re: End of injection timing matters
« Reply #7 on: May 07, 2012, 09:07:56 AM »
Wow!!! I didn't expect the difference to be so huge.

OK, I'll advance that up the list :)


johnG

  • Member
  • *
  • Posts: 37
    • www.dynotuning.net.au
Re: End of injection timing matters
« Reply #8 on: June 17, 2012, 08:18:30 AM »
Hello All

In responce to bradley , I have the 1280 installed  in a  Porsche 928 GT (5 litre v8 quad valve) ,( must get around to putting this in the intros) and was having probelms with the engine misfiring at the high end rpm , and after much discussion (with Andy) I sat down and worked out the injection cylces (opening and closing od the valves) , in time with the opening and closing of the intake valves , and made the appropiate changes in the end of injection tables and low and behold the misfire disappeared .
WHen you research this you will get a lot of information that will tell you that this does not matter!!  a lot of tuners will tell you this ( one of the black secrets) but it apparently makes a lot of difference , in this case the injection is sequential. Interesting I have not comes across this problem with group fired injection .

THanks
JG
Sincerely
John Gill

www.dynotuning.net.au

bradtpt

  • Administrator
  • Member
  • *****
  • Posts: 469
  • Without data you only have an opinion
    • Turning Point Technology
Re: End of injection timing matters
« Reply #9 on: June 17, 2012, 10:40:22 PM »
Thanks for sharing your experience on this topic John.  Out of interest were your period offset angles incorrect, or simply EOT wasn't where the engine responded best?  Also, what size injectors are you using?

In a group fired engine, issues like this are somewhat masked.  Using your engine as an example: To be clear I'm taking "group fired" to mean each bank of four injectors are fired alternatively every 360 crankshaft degrees, as opposed to all eight injectors fired once every 720 degrees.  In this setup you would likely find that only two out of your eight cylinders are not be burning optimally at certain load/rpm and it will probably for the most part go unnoticed other then the very occasional pop.  Now enter sequential injection where you have each cylinder receiving fuel at the same time in terms of crankshaft degrees for that cylinder's intake stroke and all of a sudden the occasional pop (from the two unoptimally timed cylinders described earlier) has become a terrible, inexplicable misfire as it now affects all eight.  This might initially make one think group fire is a better idea and chances are the people/tuners who are telling you it doesn't matter have probably primarily worked on group fire setups where it "doesn't matter" because you can't change it.  They're also likely the ones who have to live with the occasional pop they can't do anything about.

However, with sequential injection we CAN do something about it by changing the injection angle and when you get it right, you are rewarded with an engine which will have good manners everywhere because whilst a misfire like what was just described is magnified by an incorrectly timed sequential injection setup, so is any gain :)  And that is where your correctly timed sequential setup will run smoother, make more torque and generate less HC's at the tailpipe as you will be burning more of the fuel you inject compared to a group fired setup.
Regards,

Bradley Smith
Proprietor
Turning Point Technology

stevieturbo

  • Member
  • *
  • Posts: 133
Re: End of injection timing matters
« Reply #10 on: June 18, 2012, 02:55:36 AM »
Hello All

In responce to bradley , I have the 1280 installed  in a  Porsche 928 GT (5 litre v8 quad valve) ,( must get around to putting this in the intros) and was having probelms with the engine misfiring at the high end rpm , and after much discussion (with Andy) I sat down and worked out the injection cylces (opening and closing od the valves) , in time with the opening and closing of the intake valves , and made the appropiate changes in the end of injection tables and low and behold the misfire disappeared .
WHen you research this you will get a lot of information that will tell you that this does not matter!!  a lot of tuners will tell you this ( one of the black secrets) but it apparently makes a lot of difference , in this case the injection is sequential. Interesting I have not comes across this problem with group fired injection .

THanks
JG

So what were your EOI angles before and after ?

johnG

  • Member
  • *
  • Posts: 37
    • www.dynotuning.net.au
Re: End of injection timing matters
« Reply #11 on: June 19, 2012, 07:53:09 PM »
Here is my table attachment here

Sincerely
John Gill

www.dynotuning.net.au

DVS JEZ

  • Member
  • *
  • Posts: 7
Re: End of injection timing matters
« Reply #12 on: June 19, 2012, 09:23:39 PM »
where in the software can i find the end of injection table, i must be going blind, cant find it

bradtpt

  • Administrator
  • Member
  • *****
  • Posts: 469
  • Without data you only have an opinion
    • Turning Point Technology
Re: End of injection timing matters
« Reply #13 on: June 19, 2012, 09:44:54 PM »
DVS JEZ:  At present this feature is only available on the e1280s but I believe it's planned to make it's way to the select platform (not sure about e420c) by way of updated firmware.  I don't know how far off it is, but I'm fairly confident it's on the roadmap!  I believe there are a few other things in the development queue first.  Andy might be able to give you an approximate ETA...
Regards,

Bradley Smith
Proprietor
Turning Point Technology

Andy

  • Administrator
  • Member
  • *****
  • Posts: 4,319
Re: End of injection timing matters
« Reply #14 on: June 20, 2012, 07:58:40 AM »
Yes, basically the next things are faster comms, then end of injection timing, then arbitrary map spacing.

I will have to see if we can do it on the e420c; the e420c is already a bit overloaded with calculations which is one of the reasons we went to the Select platform.