Adjusting Pilot Timing when Base Timing is changed?

[koolky]

Adjusting Pilot Timing when Base Timing is changed?
I would like to know how important it is to scale back pilot timing as you increase base timing. One of the first thing I experimented with on my tunes was to use a timing calculator to set my timing tables with a minimum of 50% injection event BTDC while retaining any stock value that was above 50% BTDC. After reading about piston crack and large PW’s causing fuel to be injected above the piston bowl I realized that when base timing is increased so is pilot injection timing. So here’s where my thought go crazy. I was think that I could recalculate the Pilot 1 Timing Base (B1201) and reduce the amount of microseconds before the main injection event so that even with a modified Main Timing (B0909) I could time the Pilot injection event so that it would take place at the same time as a stock tune. Would this work? Would PW of Pilot overlap the Main injections?

I did a little bit of work calculating this but I am not sure if this is worth pursuing.
So far I made an excel spread sheet that converts main timing base from degrees to microseconds. Here's what I have for notes:

Pilot Injection Timing/PW Calculations
(Alt 0181 µ)
Convert rpm to deg/µs (*/µs)
((rpm/60seconds)/1000000microseconds)x360*= degrees per µs. Simplify equation: (3608/100000µs=0.00036) 60/0.00036=166666.666
Therefore rpm divided by 166666.666 = */µs (degrees per microsecond)

Example:
-0.5* of timing from the table divided by */µs at say 800rpm (800rpm = 0.0048*/µs)
(do not divide by -0.5. divide by 0.5 and then convert it to negative).
0.5 / 0.0048= -104.166µs ATDC

Now look at Pilot 1 timing base. At the same rpm/mm3 it calls for pilot injection to start 900µs BMI before main injection.
So, if main timing it happening at -104.166µs BTDC (in this case ATDC) add 900µs to that to find out when the pilot injection event starts. In this case 795.834µs BTDC.
Take the above µs BTDC and multiply that by */µs for the given rpm and you will get the pilot injection timing in degrees
795.834 x 0.0048= 3.82* BTDC

So what im thinking is that I can calculate the stock timing in *BTDC of the pilot injection event and keep it constant while adjusting base timing. Im sure this could affect some other things so I would appreciate any input.

[Wheelz]

Do some google searches for effects of pilot timing on combustion. There are some very interesting technical articles out there. They may take a couple times to read before you understand it. At least it took me a few to fully grasp the concepts.

But in short, pilot timing and amount is just as crucial as main timing. Also, from the factory it most likely isn't optimal. It will be a good average to achieve their emissions goals. But pilot injection (timing and amount) effects the ignition delay and burn rate, which in turn effects rattle, mpgs, and emissions output. It is well worth looking into and should be changed along with main timing. Also, pilot injection events should always be referenced Before Main Injection Pulse. It's timing should remain consistent relative to the start of the injection event, not TDC.

Also, the pilot injection is such a small amount you don't have to worry about it damaging a piston or gasket.

[robof16]

The articles I’ve found seem to be most concerned with emissions (NOx IIRC) and that can be at odds with efficiency/performance. This serves to make it more difficult to discern from the reading what adjustments to make.
I’ve seen what I believe to be a noticeable improvement at low RPM, low load conditions (light acceleration and cruise) with pilot timing set to 15-20 CAD before the main injection pulse. This is in addition to my already advanced-over-stock timing BTW. Supposedly a short-duration pilot delivered with higher pressure and a longer dwell before the main pulse will yield better fuel economy.

[Wheelz]

That seems on par with what I remember, but the article I read only had data at a given rpm, which I would think the ignition delay would be about the same, which would require advancing pilot timing for increasing rpm.

But 15-20 CAD prior is about right for cruise rpms. I know it's weird to think of timing in terms of ms advance, but with pilot timing it will make the math easier and the map less complex.

[koolky]

Thanks for the posts guys. The thing that gets me about the stock pilot injection timing B1201 is that it is not consistent in the way that makes sense to me. Example: At 300rpm and 1000µs BMI (Before Main Injection) the result is PI taking place 1.8* BMI. At 1400rpm and the same timing (1000µs) the result is PI taking place 8.4* BMI. At 2000rpm it is 12* BMI. That is if my calculator is set up properly. Can anyone confirm these numbers?

[Wheelz]

Thanks for the posts guys. The thing that gets me about the stock pilot injection timing B1201 is that it is not consistent in the way that makes sense to me. Example: At 300rpm and 1000µs BMI (Before Main Injection) the result is PI taking place 1.8* BMI. At 1400rpm and the same timing (1000µs) the result is PI taking place 8.4* BMI. At 2000rpm it is 12* BMI. That is if my calculator is set up properly. Can anyone confirm these numbers?

Why does that not make sense to you? The math makes a logical progression that for the same ms advance, with increasing RPM, PI CAD timing must also advance. The theory is your pilot event will have a consistent ms delay before ignition, which is what you are trying to control so that the combustion chamber is primed to the same point, at any given rpm

[koolky]

After coming back to it this morning with a clear mind I see what you’re saying and don’t really know what I was getting at. Maybe it was that B1201 seems stair step when I thought it would make more sense if it looked smoother like a main timing table. When I look at B1201 now that it’s been converted to degrees I can see that it is doing exactly what I expected it to. Low numbers at low rpm/mm3 increasing at a fairly even rate as rpm/mm3 increase.

[Wheelz]

I do agree, the stair stepped nature of the factory PI timing makes zero sense to me as well.

[koolky]

That seems on par with what I remember, but the article I read only had data at a given rpm, which I would think the ignition delay would be about the same, which would require advancing pilot timing for increasing rpm.

But 15-20 CAD prior is about right for cruise rpms. I know it's weird to think of timing in terms of ms advance, but with pilot timing it will make the math easier and the map less complex.

Do you ever play with PI timing? Stock seems to be 8-14 CAD BTDC in the cruising range. I am experimenting with upping PI timing to 1100 µs for 0-30 mm3 up to 1900 RPM. I know it’s not a lot but I can’t find much information about LBZ PI timing. I might try 15-20 like you mentioned. I’m assuming that there is good potential for improved efficiency by modifying PI timing since stock is likely kept retarded for emission reasons.

[Wheelz]

I've changed it up quite a bit. Ill post up some of the tunes I've been working on when I get home tonight, and try to find the tech paper I had been working from. I'm working late today so it'll be around midnight before I get a chance.

I've found that keeping a short duration between PI and MI makes the engine run quieter. My dads truck, which is the LBZ I've been tuning, will get around 18-19 mpg highway running 33" BFG ATs and a bone stock motor/exhaust.

Also, what I've been discovering, is that for a given engine load, the ms duration between your pulse events should remain close to the same throughout the RPM range of the motor. CAD advance really isn't a good working unit because ms/CAD is constantly changing with RPM. I used to try to think of PI timing in terms of CAD and it's just not necessary. I haven't figured out the best way to extrapolate the PI timing for increased engine loads. I need to do some more experimenting to understand how they are related to each other so the MI burn is as consistent as possible throughout the load range.

[Wheelz]

Here is what I've been trying. #10 was when I was thinking in CAD units. #12 is using the logic I mentioned in the above post. Also check out this paper. It took me a couple times to read it before I started to grasp the concepts.
http://archcomb.itc.pw.edu.pl/downloads/46_2010.pdf

16034
16035

[CwF]

Interesting and good work. I'd really like to see a detailed log of varied conditions. I haven't touched pilot so far mainly since I haven't decided on a suitable feedback to help make decisions. I think you're right to think it is maybe not optimized for efficiency but my take it a little different than yours maybe. I don't think it is nearly as important as main timing. From my readings the factory got it right enough.

First, those two tunes register the same in a compare?

I see you upped the quantity in the meat of the operating area. What have you seen to support the change? You do know that the pilot quantity is taken from total calc, so the main at a given calc mm3 will be smaller? So for example, at 40mm3 calc, under a steady cruise a 1.5 pilot gives a 38.5 main, and your 3.0 gives a 37.0 main. A log may see a difference?

Also I see you have greatly lengthened the gap, or PILOT1MT. Again, any log support? I understand the setup the pilot performs is time dependent, and have seen dat to suggest it is long enough, or not, and a few µ or ° (alt 248) extra has little if any effect that main timing adjustment couldn't also achieve. Actually, to far ahead and it is now extra resistance, though miniscule.

Lastly, I think the stair step does make sense. As pretty as your curve is on b1201 is, I think smoothing is irrelevant to the computer. The variation in the factory table is really for recognizing the different states those ranges are effective, transitional you could say. As mm3 changes or rpm, there is no technical reason I know of for any difference in delay. The differences are for different uses of various ranges, accelerating through, decelerating through, steady state, etc. If that makes sense. Remember, the studies fix factors we don't have to fix. If they show a difference it doesn't necessarily translate when the option to change the main is available.

I congratulate you for not turning off the pilot as rpm goes up. Again, no technical reason to do so unless we exceed the capacity of the rail. No reason to do so otherwise, plenty of higher speed (smaller, shorter stroke) diesels maintain pilot up to 5500 rpms. It is always a benefit. The tune is not set to rev past stock however. You do tighten the time as rpm goes past the stock range, why? Actually you have it curved with a peak. I don't get it.

In my logs I see times that don't follow the tune that suggest there are a few tables or parameters we don't have access to with efilive. This is very apparent with post injection. Who know whats up with those P2 and P3 variables. I've never seen them. I did find the extents of Post2 a year or so ago in maybe 10 various tunes, then turned it off. Now, with a few months of 170 mile mountain runs a bell went off and I used that knowledge in a new way, turned it back on with and have great results. With some negative numbers I solved my earlier issue and find it very beneficial. At some point I'd like to take a look with WinOLS or something and see what I can find, if I can find the time for such brain numbing.

Keep up the good work.

[CwF]

Oh, the last 158.4 mile run used 7.30 gallons! With 4 burst past 115mm3 up long hills! One to 4003 RPM!

[CwF]

I'm back at looking at this and it occurs to me now I was on a side thought. The title is asking 'should' you adjust the pilot 'because' you adjust main. I'm thinking a definitive NO. The pilot will follow the main.

Back to what page I was on is more like 'Is there any advantage in adjusting pilot timing?' regardless of main timing. And to rephrase what I was trying to say, I'm sure there is in some cases and I can't figure out how to tell in a meaningful way. Large adjustment may yield some feel, some mileage change, or sound, but smaller changes seem within some magic margin of error. I looked at some other papers and think there are three ranges: Not enough time and you can hear it, a few thousand micro second range where it's long enough, and over that with no indication what so ever. I think optimal is at the border of the first two in order to minimize excessive (though tiny) pressure BTDC without any advantage that can not be had by simply adjusting the main, or pressure, or another factor. So I'm leaving mine alone.

On these tunes you have the pilot out to 47°+ before the main at peak. I say too far. Way to far. Damaging no, beneficial is the debate. Note that in my logs on my tune the pilot seems to close in at higher RPM despite a timing steady table. Closing up to half or a third the delay my stock table says to do beyond 3300 rpm. And again, no technical reason to turn it off and I do tune to stay within rail pressure capabilities - there enough there for me - so far. There is another hidden parameter or table causing this....?

[Wheelz]

While I do agree with most everything you said, referencing pilot timing in terms of degrees advance is useless. It should be discussed in terms of ms advance as you are attempting to control the ignition delay. I'm not denying my pilot could be advanced too far to be useful. I'm experimenting with this quite a bit. Just note that when you do your calculations don't worry what the degree advance is because the only purpose of the pilot injection event is to control the ignition day of the main event to create a smooth burn. And in doing that you want to keep a smooth ms distance between the events.

[CwF]

Agree, that is why the tables are in μs. I just calculated it out and thought holy crap on a cracker... I'm interested to hear the results of what the extra 2000μs or so is giving.

[Wheelz]

Just got thru comparing it with a tune that had a max PI timing advance of 1700 ms. Ran with a lot less rattle at higher rpms. Ill post up that tune in a few days.

[koolky]

Any theory on why rattle was reduced with the PI timing advanced so much at higher rpm? Do you think that with the advanced PI timing that it is having an effect on the fuel pressure drop/recovery time between PI and MI, allowing the MI to atomize better?

I didn’t realize you guys had commented here since I last looked at this thread. I am still working on my tunes and have sort of come back to optimizing PI timing and now starting to look at mm3.

Wheelz, I noticed in posted tune #10 that you dropped the PI timing down below stock values in the low rpm/mm3 range. Did you notice any changes at idle with this? A few months ago I was playing with advancing the timing in that range with the idea that the MI event would have more time to recover fuel pressure possibly leading to better atomization without upping rail pressure. I also lowered the PI mm3 at idle and noticed only the slightest increase in rattle but figure that would also mean that the PI event is shorter and leave me with more time between PI/MI.