Sequential injection with only crank position sensor

[NewZed]

Here's how the factory defines it.  Literally, the word only means that injection happens in order, or sequence, of cylinder actions.  Doesn't say anything at all about when injection happens, within a complete cycle.  But each cylinder sees the same type of charge; well-blended, inhomogeneous, straight in to the open intake port, or whatever.  Unlike batch where the cylinders see a different type of charge since there are six possibilities, each cylinder seeing two of the possibilities.

[TimZ]

Here's how the factory defines it.  Literally, the word only means that injection happens in order, or sequence, of cylinder actions.  Doesn't say anything at all about when injection happens, within a complete cycle.  But each cylinder sees the same type of charge; well-blended, inhomogeneous, straight in to the open intake port, or whatever.  Unlike batch where the cylinders see a different type of charge since there are six possibilities, each cylinder seeing two of the possibilities.

Thanks, NewZed - this is how I understood Sequential Injection to work, and what I was basing my posts on.

[madkaw]

http://www.msextra.com/doc/pdf/html/Megasquirt3_TunerStudio_MS_Lite_Reference-1.4.pdf/Megasquirt3_TunerStudio_MS_Lite_Reference-1.4.html

[NewZed]

Links without words attached aren't really worth much.  No offense.  Here's one that offers about as much, maybe more actually than the MSextra link - https://www.google.com/webhp?tab=ww&ei=dq5PVNe5B8j8igLU9ICABg&ved=0CAMQ1S4#q=sequential

 

About the same as posting the word "search".

 

 

Edit - point being, what's in that msextra link that is relevant?

Edited February 19, 2017 by NewZed

[madkaw]

Sorry, I thought the link was more direct thus not needing any explanation

[madkaw]

I was trying to link the MS3 tuner studio explanation of sequence settings

[NewZed]

Oops.  That's the Mega documentation for you.  These discussions all need good definitions of terms as they proceed otherwise people think that they're talking about the same things when they're not.  There was a conversation about sequential ignition a while back that made little sense to me, since a spark out of sequence really isn't worth much.  But the word was used to distinguish from wasted-spark.  Which is still sequential where it matters.  

[seattlejester]

Page 81 to start for those interested. It seems it does have inputs for cam angle so it would indeed be trying to inject during timed valve events. 

 

Very much so a case of definitions.

[TimZ]

Page 81 to start for those interested. It seems it does have inputs for cam angle so it would indeed be trying to inject during timed valve events. 

 

Very much so a case of definitions.

Again - the key words here are "trying to".  

Once the required pulsewidth is significantly longer than the vale open time, then injecting on the "timed valve event" doesn't really do much.  Typically the valve open period is 35 to 40% of the available 720 degree cycle, so it doesn't take that much to exceed it.  This is why I mentioned that the timing is really only advantageous for low load conditions where the injector pulsewidth can fit inside the valve open period.  Even given that, I think you'd be hard pressed to be able to tell the difference if the injectors were timed 360 degrees off.

[madkaw]

I think I have been talked out of a cam sensor for my motor:) One less sensor to worry about.

I think I read early in my FI adventure that performance gains are minimal with full sequential . I'm wired for it if I want to though.

[TimZ]

I think I have been talked out of a cam sensor for my motor:) One less sensor to worry about.

I think I read early in my FI adventure that performance gains are minimal with full sequential . I'm wired for it if I want to though.

Yeah, you won't really see any difference in max power.  If your injectors are big enough that your idle PWs end up in the sub-2mS range, doubling the PW by going full sequential will make a huge difference in fuel control at idle/low load.  

If you don't mind having the timing off by 360 degrees about half the time you could try just programming it for full sequential and simply not hooking up the cam sensor.  That will work with a TEC3, but I'm not sure how the MS reacts.

[seattlejester]

Oh I agree the benefit is minimal, and once you get past a certain RPM I imagine we become sloppy. Not much for power if I remember, but the benefit was fuel economy for timed sequential injection at very low load and engine speed (highway cruise at low engine speeds).

 

I was trying to work this out with math. I think I got it, took me a couple hours and some reference finding.

 

So the formula in long form is 

 

1 / (Engine speed (R/M) * (1Min/60s) * (1000ms/s) * (2 revolutions / 4 stroke cycle) * (Angle valve is open/720 degrees) * (80/100 duty cycle limit) = ms injectors can fire

 

(Volume of engine / number of cylinders) / (22.4L/Mol) * (28.966g/mol of air) / (14.7 air / fuel) * (cc of fuel / gram) = cc of fuel required

 

cc of fuel / ms injectors can fire *  (1000ms/s) * (60s/min) = injector size needed in cc/min

 

 

At 6000 Rotations (of the crank) per Minute you get 100 Rotations per second. Or 10ms a rev. for 4 stroke cycle that gives you 20ms a 2 rev cycle.

 

Using Timz's number of 40% (closer to 33% on stock cams 240*) valve open time you get 8ms of injection when the valve is open at 6000rpm.

 

With that you get 16ms at 3000rpm, 32ms at 1500rpm and so on.

 

Given you don't want to run a given injector more then 80% duty cycle you get.

 

5.5ms at 7000rpm

6.4ms at 6000rpm

12.8ms at 3000rpm

25.6ms at 1500rpm

51.2ms at 750rpm

 

Using a 3L motor one cylinder in question pulls in and uses 0.5L per combustion cycle in one cylinder

 

Using avogadro's law we get Volume of gas/(22.4L/Mole of gas) * (grams/mole of gas) = grams of gas

 

Air has a g/mol of 28.966g/mol

 

0.5L/22.4l/mol * 28.966g/mol = 0.64656grams of o2.

 

Using AFR 14.7 the fuel we would want would be 0.043984 grams of gasoline

 

Density of gasoline is 737.22kg/m3 and 1kg/m3 is 0.001g/cm3 = 0.73722 g/cm3

 

0.043984 g / 0.73722g/cm3 = 0.0596618cc of gasoline

 

0.0596618cc of gasoline in 6.4 ms = 0.00932215cc/ms * 1000ms/s * 60s/min = 559.329cc/min

 

Using that last line the required injector size using the following assumptions (3L, perfect volumetric efficiency, AFR 14.7, 40% valve open time, and a plethora of other assumptions)

 

7000RPM requires 651cc/min injectors at 80% duty cycle

6000RPM requires 560cc/min injectors at 80% duty cycle

3000RPM requires 280cc/min injectors at 80* duty cycle

1500RPM requires 140cc/min injectors at 80* duty cycle

750RPM requires 70cc/min injectors at 80* duty cycle

 

Note fairly massive injectors that would be absolutely overkill unless you were running boost at which point the calculation has another factor that the volume of air increases at a factor of (absolute kpa/100) or ((psi of boost/14.7) + 1)

Edited February 20, 2017 by seattlejester

[TimZ]

Oh I agree the benefit is minimal, and once you get past a certain RPM I imagine we become sloppy. Not much for power if I remember, but the benefit was fuel economy for timed sequential injection at very low load and engine speed (highway cruise at low engine speeds).

 

I was trying to work this out with math. I think I got it, took me a couple hours and some reference finding.

 

 

The benefit to max power is negligible, as I've been pointing out.  Max power was _never_ the point, though.  Doubling the required pulsewidth for increased control at low loads was always the point, and if you are using large enough injectors to require PWs in the <2ms range at idle, the benefit here is huge.  Being able to inject while the valve is open is nice, and directionally correct, but kind of beside the point.  People always seem to get way too hung up on this.

 

Injectors, even the new ones, get increasingly non-linear as you get near their min on-time. There is a zone (usually between ~1 and 1.5mS) where they still deliver fuel, but they tend to "glob" the fuel, so one cycle you get next to nothing and a few cycles later you get all the fuel that was missing the last few cycles, which is now way too much.  Getting out of that range and into the meat of the linear part of the curve is essential to getting good drivability.  

[Chickenman]

Actually Tim, my first reply was in regards to Max power. It was the opposite to what you were discussing ( Sequentail and olw rpm advantages ),  but I thought I made that pretty clear. My replay was for a caution to OP that the stock 188cc injectors may be adequate for batch fire at high rpm's,  but not if he goes to fully timed sequential.

 

Then we all got into the discussion of the different " types " of sequential injection strategy. The SDS link I supplied sums the situation up very eloquently to the situation I was warning about. So really, just a difference of what we were all visualizing in our heads. That's why discussions over Beers are great. After a few beers you forget what you were discussing and just start chasing the waitresses... LOL 

 

Edit: Yes, sequential has very littel effec5t on max power cfompared to sequential. In fact F1 engines in the V8 and V10 era used Multipoint injcetion with individual throttle body injectors. At 20,000 rpm it was better to jsut inject straight into the plenum and let use a wet plenum system. That all changed with the V6 turbo engines and  heavily restricted fuel flow regulation.  Direct injection was the only way they could make the power and fuel economy they needed.

 

I'll see if I can find the video of the Renault V8 at full chat. It's quite interesting the injector design.

 

Was also reading an article the other day that passenger car manufactures are now achieving AFR as low as 30 to 1 at cruise with Jet Flame fuel injection. Which is what Mercedes is running on their F1 cars and why they have had such a huge advantage the last few years. 

 

Good discussion 

Edited February 20, 2017 by Chickenman

[Chickenman]

Found it. Video of Renault V8 engine.  Circa 2004. These were the years of 20,000 RPM's!! Note the Individual Throttle body injectors.

 

https://www.youtube.com/watch?v=y2iBbwocYZw

 

Edit: Google Turbulent Jet Injection for some interesting reading: 

 

http://www.enginelabs.com/news/f1-tech-what-is-turbulent-jet-ignition/

 

The Mahle system is a developement of HCCI ( Homogenous Chage Compression Ignition ). HCCI systems were developed way back in the 70's and known as Stratified charge engines. One of the more succesful designs was Honda's CVCC engine. Back then electronic fuel injection was not yet developed. Honda relied on a mechanical method ( 3 valves per cylinder, two inlet tracts and a 3 barrel carburator ) to achive their golas.

 

As Direct petrol Injection engines have recently developed,  stratified combustion chamber ( HCCI )  designs were re-surrected, so to speak. Mahle has now taken that a step further with their Turbulent Jet Ignition system.

 

Piston top designs are very complex these days and Ferrari has recently annouced that it is developing 3D printed Steel pistons for their 2017 and 2018 F1 engines.

 

https://www.motorsport.com/f1/news/ferrari-eyes-innovative-piston-solution-thanks-to-3d-printing-870008/?utm_source=en&utm_medium=HomePage&utm_campaign=GiorgioPiolaWidget

Edited February 20, 2017 by Chickenman