Anyone want to have a TB/inj. discussion?

[uglyDougly]

  Admittedly an injector discussion could be percieved as a pi**ing contest, but is anyone game to share a little?

   Doug

[Vosselman]

Hi Doug,
That would be nice. But do you expect another result than you did when testing the Duc injectors. Or is the spread in delivered fuel not what you mean?

[uglyDougly]

  <Or is the spread in delivered fuel not what you mean?>

  It could be that and much, much more.

  Doug

[Scramjet]

I do not know if this is on topic but I do have a question regarding the TB/inj size for a 103" and 107" build.  It may kick off some discussion.

Looking at pulse width on a SERT data recording, what are you looking for to determine if pulse width is near maximum?

I typically have seen numbers in the 6-20 ms range.

ex. 103" with stock 46mm TB 3.9 gm/sec inj pulse width around 6-20 ms under load.
ex. 107" with SE50 TB 4.9 gm/sec inj pulse width also around 6-20 ms under load.

I have heard the number 85% at WOT thrown around.  85% of what.  Is the duty cycle maximum 100 ms?

What should a self tuner look for?

B

Edit: added examples

[uglyDougly]

  First you have to figure out cycle time.
  A cycle is two revolutions of a 4-stroke engine.

  RPM divided by 60 = RPS (revolutions per second)  (6000 RPM / 60 = 100 RPS)

  The time of one revolution = 1 divided by the RPS That will give you the time in seconds. (1 / 100 = 0.010 or 10 thousands of a second or 10 milliseconds)

  10mS X 2 = 20 mS

  85% of 20mS = 17mS

  At 3000 RPM the cycle time is 40mS (check my math) so 85% duty cycle is 34mS.

  The injector is just a solenoid, a doorbell chime mechanism. When you use someone's doorbell you don't notice that the striker hesitates for a couple milliseconds while the field builds up enough to move the core or striker.
  In an injector that millisecond matters. When you're trying to fire the thing once every two revolutions at 12,000 RPM 1 mS is 10% of the time available. at 6000 RPM it is still 5% of the time.
  This is referred to as 'dead time' and it is dependent on battery voltage. There is a table in the calibration that accounts for the variation in dead time thru the battery voltage range. It is part of every firing of both injectors.

  Let's not move on before this is covered to satisfaction?

  Doug

[FLTRI]

Doug,
Maybe it would be a good thing to address how a guy would be able to determine if what he has for injectors are big enough to do the job he's asking them to.
In other words there seems to be a lot of guys who buy bigger injectors because they went from and 88" engine to a 95" or even a 103" engine.
How does a guy know whether he needs bigger injectors or the ones he originally got in the bike are big enough?
Is there a way for him to make a self-determination or does he have to rely on propaganda and/or heresay?

[FLTRI]

Another question:
Since the economy is in a little funk right now, there are still guys who want more power from their bike but now think twice before pulling the wallet out for bigger TBs and want to know if they really need one.

Let's use what I consider the most common example:
Joe owns an '07 bagger. He likes to ride 100-600 miles per day on weekend trips that include back roads and interstates. He doesn't drag race the bike nor is he interested in revving the engine past about 4500-5000 when he's performance riding, mainly because it feels really good to him from 2000-5000 rpms and he doesn't feel there's a need to go any further into the higher rpms.

His bike came stock with a 96" (strokes 88") engine. He figures he can justify going to 103" simply because it only requires a big bore kit to buy and install, rather than a complete tear-down to install a stroker crank as it was with the 88" engines of the past.

He's looking for more bottom end torque and after researching here among other places and people opts for a Woods TW6-6.

He doesn't like the ear-piercing sound of some of the highest peak number producing exhausts, so he opts for a Supertrapp Supermeg because he can adjust the noise level himself before he has it tuned for best performance.

So now he has an engine that is bigger and produces more HP/TQ, so the next thing he naturally thinks about is the size of his intake (TB).

Based on this scenario will he benefit from a larger TB and if so how much and where will he experience the increase?

[axtell]

Is scary to think that 3/4 of the pulse width is directed @ a closed valve

[Scramjet]

  First you have to figure out cycle time.
  A cycle is two revolutions of a 4-stroke engine.

  RPM divided by 60 = RPS (revolutions per second)  (6000 RPM / 60 = 100 RPS)

  The time of one revolution = 1 divided by the RPS That will give you the time in seconds. (1 / 100 = 0.010 or 10 thousands of a second or 10 milliseconds)

  10mS X 2 = 20 mS

  85% of 20mS = 17mS

  At 3000 RPM the cycle time is 40mS (check my math) so 85% duty cycle is 34mS.

  The injector is just a solenoid, a doorbell chime mechanism. When you use someone's doorbell you don't notice that the striker hesitates for a couple milliseconds while the field builds up enough to move the core or striker.
  In an injector that millisecond matters. When you're trying to fire the thing once every two revolutions at 12,000 RPM 1 mS is 10% of the time available. at 6000 RPM it is still 5% of the time.
  This is referred to as 'dead time' and it is dependent on battery voltage. There is a table in the calibration that accounts for the variation in dead time thru the battery voltage range. It is part of every firing of both injectors.

  Let's not move on before this is covered to satisfaction?

  Doug

Doug,

Thank you.  That makes sense.

That tells me that at 6,000 RPM (an assumed rev limit) you should look for a injector pulse of no more than 17 ms.  If the injector pulse width is greater than 17 ms a larger injector (or increased fuel pressure) is needed?

At 3,000 RPM a pulse width of 34 ms calls for a larger injector?

Since my VE's are max at 3,000-3,500 RPM, that is where I should look for 85% duty cycle?

And at the rev limit for a reality check?

B

[uglyDougly]

   Good suggestions Bob.

 Injector sizing.

 The amount of fuel required is directly, inextricably tied to the amount of air the engine can process. (This, of course, assumes you actually have control of the fuel-air ratio.)

 Most Otto Cycle engines (spark ignited gasoline fueled) will use around 1/2 lb of fuel per HP per hour. A 40HP load should use 20 lb of fuel in 1 hour.

 The engineering community refers to this as Brake Specific Fuel Consumption. Now you see why injectors are often rated in lb/hr!?!?!?

 The grams-per-second value in the ECM Tuning Constants window of the Master Tune software describes the injector size for the equation that calculates injector pulse width.  When we're done figurin' how much fuel we need we'll convert from lb/hr to gr/sec.  Weight-to-weight/ time-to-time.

  Since the predominant device for measuring the output of our motorcycles is the chassis dynamometer we measure the power at the rear wheel. There are losses incurred through the drive-train.
  My injector spreadsheet assumes 15% loss from crank to rear wheel. This comes from the 101 HP Buell White Lightning which pretty consistantly showed 86 HP on a Dynojet dyno.
  Let's not get side-tracked with an argument about what the losses are, 15%?, 10%. For injector calculation where you don't want to run out of fuel, it works to be conservative or optimistic about the power you will make.

  For a 100HP build I assume 115 HP at the crank.
 A BSFC of .550 works and is more conservative for an air-cooled engine.
 Since we know about injector 'dead-time' it's obvious that at 6000 RPM with 20mS a dead-time of 1mS is 5% of that. There is some time for the injector to open and close also. That time remains constant whether you leave it open for 16mS or 3.5 mS as does the 'dead-time'.

 'They say' a duty cycle of 80% (some of 'them' say 85%) keeps you out of trouble with 'dead-time' and open-close time.
 At low RPMs 80% works fine, but it gets you deep in trouble at high RPMs.

 The problem, or so I've read is literally valve float. When the disc or pintle is between open and close and switched it will float and damage itself. Just like the poppet valves in our engines?

 115 HP X 1.15 + 115 HP @ the crank

 BSFC of .500 X 115HP = 57.5 lb/hr

 57.5 lb/hr divided by 80% Duty Cycle = 71.875 lb/hr

 71.875 lb/hr divided by two injectors = 35.9375 lb/hr injectors.

 The H-D Weber injectors from '02 -'06 flow 34 lb/hr at 3 Bar.

 The flow thru an orifice changes as the square root of the pressure differential.

 sqrt of 3 Bar(rating pressure) = 1.732,   sqrt of 4 Bar (H-D regulator) = 2

 2 divided by 1.732 = 1.155

 34 lb/hr X 1.155 = 39.26 lb/hr

 The Weber IWP 043 and IWP 162 flow enough.

  Let's see how the later injectors will work.

 The H-D PN 27625-06 injectors flow 28.5 lb/hr.

 28.5 lb/hr @3 Bar X 1.15 (flow increase with 4 Bar) = 32.78 lb/hr
 
 Go back to our requirement, 35.9375 lb/hr injectors

 We're 9% short, but it will work, maybe.

  I would guess you would be able to get the AFR you need at WOT until the PE-afr table kicks in, then you will not be at the target AFRs of that. But, then again, if you've just built your engine why would you want to kill all the power with too much fuel?
 That's my opinion showing!!

 Get it?  we'll cover this 'til it sinks in.

 Doug

[uglyDougly]

  Thanks for chiming in Ron.

  Way back in the VW Bosch CIS (that's the Constant Injection System) days, Bosch said they had tested the performance of Otto Cycle engines and there was no (maybe they said 'little' I don't remember) difference between short injector duration and constant injector flow.
  Those were pre-emissions days, so it probably does have an affect on emission, but not so much on power.
  Chris Schroeder of DTT, says you shouldn't have greater than 50% duty on 'Y' manifold engines. That would be from the fuel-sharing/robbing standoint. Any of us who have done custom calibrations, know that is not the practical limit. But, it probably doesn't make it easier to tune them well.

  But, if shorter duty cycle were critical for emissions why did the MoCo put in smaller injectors from '06 onwards.
  The emissions thing is way over my head.

  I have read (don't remember where) that spraying some fuel on a closed, hot valve works to help vapourize the fuel and Reggie has put a digital scope on one and says that the injector spray starts earlier with an increase in engine size in the ECM Constants section.
  That implies something, although I don't know what.

  Doug

[axtell]

My honest opinion is that there is probably very little over your head Ron

[Scramjet]

Am I running out of injector with my SE50 TB and 4.9 gm/sec injector on a 107" build?

A data run is attached.  At only 5108 RPM the pulse width is at 19.08 ms and the MAP reads 98.2.



[attachment removed after 60 days by system]

[-SeabrookTrickBagger]

Doug,

Hopefully, my question below does not divert your intended path in regard this discussion.

I find this part of your explanation of interest to me since it addresses reality of the HD FI engine:  The flow thru an orifice changes as the square root of the pressure differential.

  sqrt of 3 Bar(rating pressure) = 1.732,   sqrt of 4 Bar (H-D regulator) = 2

Would this be a fairly accurate calculation:  an injector flowed/rated at 6.1 gps at HD pressure would equate to about 5.7 at 3 bar?  I am not a mathematician and I only had one college level math course in quantitative math skills which was not enough.....  .

FWIW, the DTT TCFI does not ask for the pressure used to achieve a flow rate at the time you enter the injector flow rate (in grams per second) in parameters.  Since the "industry standard" appears to be 3 bar, then the 3 bar flow rate is used in this parameter for injector flow rate without any real thought.  When Schroeder was asked about this, he said it did not make any difference:  use 3 bar or HD pressure injector flow rates.  He probably meant it was "mot snix" as far as a street engine is concerned.

Of course, Schoeder is no dummy.  His printed recommended injector sizes for common plenum  Y manifolds is considered WAY too high in actual practice for engines which operate at rpms other than WOT, though.

[uglyDougly]

  Scramjet asked; 'Since my VE's are max at 3,000-3,500 RPM, that is where I should look for 85% duty cycle?'

   Your pulse width will be greatest at your torque peak, which is your Volumetric Efficiency peak.

  Hint: the fuel table is an analogue of the torque curve.

   But, as you start calculating cycle time you realize that time is the problem. At your torque peak you have lots of time realtive to your power peak.

  Guess what? The injector flow capability relates directly to max air-flow, max-power where time is getting short. That is where you should expect to run out of injector capability.
  If you're running out at the torque peak, and not at the power peak your torque is dropping off too fast by conventional IC engine standards, but it doesn't matter, you need larger injectors.

  If your torque drops off typically, it's all over already. Right now you're at 81% @ 5108 but it shouldn't get better.

   STB, my lb/hr numbers come from an injector flow bench. Here we go with the 'flow-bench', dyno, flow-bench, discussion.
   The short story is, if the numbers are repeatable you can work with them. In this case they fall close to the calculated gps and cc/min values given by the mfg.
  With injectors the point is to size them large enough to not run out of fuel. We could probably run up to 90% on a 6000 RPM engine, but if our values vary a little, why take a chance?

   About injector values in MT; The calculated Gr/Sec (gps) for the Weber IWP069 (yellow band)  is around 6.1. In fact they work better in the MT-RT-DL with a value more like 5.6-5.8.
   Why? We're using that number to manipulate the software to give us enough VE overhead to tune. And that is about H-D/Delphi assuming that these engines will never process as much air as you guys are getting them to.

   I can say 'you guys' 'cause I don't build H-D engines. If I need that done I'll just see Bob Wright.

  Quote from dirty Harry; 'a man's gotta know his limitations'

  Doug

[axtell]

Doug...Could you call me @ 1-800-704-3201 Ron

[Scramjet]

  Scramjet asked; 'Since my VE's are max at 3,000-3,500 RPM, that is where I should look for 85% duty cycle?'

   Your pulse width will be greatest at your torque peak, which is your Volumetric Efficiency peak.

  Hint: the fuel table is an analogue of the torque curve.

   But, as you start calculating cycle time you realize that time is the problem. At your torque peak you have lots of time realtive to your power peak.

  Guess what? The injector flow capability relates directly to max air-flow, max-power where time is getting short. That is where you should expect to run out of injector capability.
  If you're running out at the torque peak, and not at the power peak your torque is dropping off too fast by conventional IC engine standards, but it doesn't matter, you need larger injectors.

  If your torque drops off typically, it's all over already. Right now you're at 81% @ 5108 but it shouldn't get better.

  Doug

WOW, when I do the actual calculation using 6200 rev limit, VE relatively contant and AFR demand constant I get 100% duty cycle of 19.35 ms.

85% of 19.35 ms = 16.45 ms ------ I will be over well over 85% duty cycle and will be close to 100% duty cycle even before Power Enrichment starts droping the AFR to 12.5 over time at WOT.

Is this correct?

B

[ederdelyi]

These aren't bad for ball park figuring ... remember that you need to use crank power, not wheel power.

http://www.injector.com/injectorselection.php

http://www.rceng.com/technical.aspx



If you look around there are some on the net that use other methods to arrive at the figures, but they should all arrive at pretty much the same results. I have a fuel injector calculator tool from Performance Trends that is pretty nice, works well with my engine simulation software and gives the option several different methods of calculation. Just makes it easier, so you don't have to remember the formulas. I found HD's BSFC to run in the .55 to .6 range, getting them lower is a challenge, but it can be done.

[nc-renegade]

 
  I have read (don't remember where) that spraying some fuel on a closed, hot valve works to help vapourize the fuel

I think that has merit.

[uglyDougly]

   Rob's Dyno has my spreadsheet on his site;  http://www.robsdyno.com/injectors.htm
  Again, I did mine for some arbitrary loss factor for motorcycles.

  Scramjet, it depends on how fast your torque drops away after the torque peak.
 The pulse-width will be roughly 20% shorter if your torque drops 20% from 4000 to 6000, but the time decreases 33.3% from 4000 to 6000. Looks like a loosing proposition to me.

 Doug

[-SeabrookTrickBagger]

 
   STB, my lb/hr numbers come from an injector flow bench. Here we go with the 'flow-bench', dyno, flow-bench, discussion.
   The short story is, if the numbers are repeatable you can work with them. In this case they fall close to the calculated gps and cc/min values given by the mfg.
  With injectors the point is to size them large enough to not run out of fuel. We could probably run up to 90% on a 6000 RPM engine, but if our values vary a little, why take a chance?

  Doug

Thanks.  The correct factors were in your earlier post.  6.1 gps at 4 bar equals 5.3 gps at 3 bar, not 5.7 gps as I "guessed." 

[whittlebeast]

If you have Microsoft Excel this sheet will cover everything that you will ever need in calcing injector sizes.  It was developed when I was doing a 7000 RPM, 140 HP two stroke EFI motor.  In that case, I needed 800 cc injectors.

http://www.ncs-stl.com/fuel/ReqInjectors03.xls

Yellow cells are input cells and Red are critical output cells.  If you use the program MegaLogViewer to look at the data log you can turn on Duty Cycle 1 and Duty Cycle 2 and have the software calc it for you automaticly.  The first time you activate the function you will have to answer t questions.  I think the answers are .5 and 1 for a Harley.

Hope this helps and have fun tuning.

AW

[Scramjet]

AW,

Thanks for the info.

Some of this does not add up.  If you look at the numbers for the commonly used 4.9 gm/sec H-D injectors they can provide about 38.9 lbs/hr max.  That means that these injectors will not support most 107" (and above) builds.  They would run out of injector limits around  115 rear wheel HP (actual not corrected) or 130 crank HP (actual not corrected).

The contradiction is that many builds producing that kind of power are running those injectors.

B

[whittlebeast]

If you push the duty cycle to 90% and you use the opening time of .7 ms that the code appears to be using and things are fairly close.  If you do an o2 enabled SERT data log of one of these bikes at WOT and close to max RPM and email me the DM2 or DM3 file, I can post the duty cycle graphs along with lots of the other stuff we can see.

AW

[uglyDougly]

<4.9 gm/sec H-D injectors they can provide about 38.9 lbs/hr max. >

 I haven't memorized the SERT, MT injector values, but these look like the Weber Pico IWP043 and IWP162 (brown band and white band, pre '06)

 Bob Wright put together his own 95CI engine with K billet heads (seconds) an experimental TC3G cam, and a prototype 54mm TB.
 I tuned it and it made 117HP and the injectors where at 86% duty cycle.

 Doesn't look like there's any disagreement there?

 That would give you a couple percent overhead and if you revert the AFR table to values much richer than 13.2 it might not get you quite the fuel you expect.
 Even in that case it will make all of the power, just go a little lean , just a little, and they don't mind that on the top. (It speeds up the burn rate. Hint; RPM/time)

  When the numbers work out, there have to be other reasons or more to the story. That's where you have to be an expert interviewer to drag all of the information out into the light.

  If the pulse-width is known and it equals the time for a complete cycle, you're done! There is no more, and as the engine increases RPM there is less time, so less fuel.

  Detecting when the injectors are maxxed can be very difficult, until you look at the data.

 That's the reason I bothered to learn this stuff. It saves a lot of time.

 At the V-Twin Expo we (Ed Dahir and I) talked with a guy who's doing his own tuning. He's a Mechanical Engineer so he 'gets it', understands.
  He's using a Twinscan II+ and two O2 sensors to collect data. He was surprised to find out that there was a PE_AFR table that messes with the tuning AFRs. Unfortunately, he had already purchased the Super Pro Tuner which doesn't support access to the PE_AFR table. He understood the effect of that table and was disappointed that nobody else had pointed that out to him.
 He has a 117 with a stock TB and stock injectors.
 Since then Ed told him he needed larger injectors (stock '06 and later Delphi injectors are 24 lb/hr) and I told him he will need larger injectors, plus I informed him that the injectors H-D sell for the 50mm TB are only 28 lb/hr, which he had already ordered from a dealer.
 I offered to confirm the flow rate of whatever PN he ordered but he didn't get back to me.

 Test Time!!

 How much power will the 28 lb/hr injectors support?

 Doug