Printable View
say 80-85%DC, big power cars have fuel just pooling on the valve, there's no way around it, they're injecting at least half the time on a closed valve.Quote:
Originally Posted by brusk
it may evaporate (some?) due to the heat on the intake valve, but still.. when you get that big, just throw octane at it and feed it fuel? haha
Time to bring this thread back to life. Im wondering if it's true the maximum you can hold the injector is 2ms under full load.....
will have to see what the limits inside the DME are, if they're restrictive at all.Quote:
Originally Posted by Abacus38
Where are you getting the 2ms number from?
This post here but at 8k rpm the injection window is 3.75ms on the intake stroke. If we continue spraying into the compression stroke we could gain and additional 2-3ms. The dynamic flow rate of the injectors are 14.5 mg @ 10% duty cycle and .4ms pulse width. If that's true then we can deliver .18125 g of fuel holding the injectors for 5ms. That's alot of fuel considering the amount of e85 needed for 1000hp is .13xxxx g at a 12.5(8.333333 E85) AFRQuote:
Originally Posted by Flinchy
The equation to for airmass per cylinder and fuel mass per is(from Designing and Tuning High-Performance Fuel Injection Systems By Greg Banish)
Volume*VE*kPa/.28705*Temp(in kelvin) x(#of cylinders)
Airmass per cylinder/afr
Input data
AFR ratio at full load is 8.3333333 to 1(Using e85)
Volume 3L,
VE 95% assuming built head to rev to 8k
Using the Garrett boost advisor 32.25psi is needed to make 1000hp converting that to kpa you get 222.355923kpa
Temp is 140F convert to kelvin 333.15
Airmass per cylinder is 1.1044471612 g/cyl is needed for 1000hp
1.1044471612/8.33333333 = 0.1325g of fuel of needed
ah ok - well it'd be 3.75MS for both the intake and compression (60/8000/2) - since both intake+compression = 1 rev... so that'd make it 135mg (.135g) for a full intake>compression stroke of 3.75msQuote:
Originally Posted by Abacus38
pretty much dead on 1000hp at 8k RPM if you can use both strokes?
awesome thanks for that, never knew about that formula before, interesting.Quote:
Originally Posted by Abacus38
sounds like awfully low boost for 1000 ponies though lol. also, the formula confuses me a little, does changing between say race gas and ethanol not change things, given the energy density?
Remember we are revving to 8K not 7K. HP equal torque over time if you can hold torque to higher rpm you'll make more power. If we stuck at 7k redline you'll need 38-40psi if I remember correctly. Cycle time is 120,000/rpm. A cycle time is that it take for the motor to compete 2 revolutions. The cycle time at 8000rpm is 15ms. Now I'm assuming the each stroke takes the same amount of time so dividing 15ms by 4 you get 3.75ms. 3.75ms is how long the intake stroke would be. Now I don't know how long we could inject into the compression stroke but with a mere 1.25ms more into the compression stroke would allow the injector to be open for a totally of 5ms which would deliver .18125g of fuel which is way more then 0.1325g of fuel need to run e85. This is just and educated guess on what Ive learned by reading up on the piezo injectors and Greg book on tuning. I'm no expert but hopefully the smart guys will figured out
ah right my bad, i worked out the timing completely back to front, sorry hah.Quote:
Originally Posted by Abacus38
yeah wow PLENTY of time assuming the injector can be set to inject whenever we want at whatever flow (maximum) we want. I'm sure even if it can't straight away, people will work out how, given the desire and skills to do so.
it's making sense to me so far, but takes the skilled/experienced tuners taking a look at the table, making changes, checking dyno to be sure.
Tony'll probably like this w/his dual shotgun and twins ;)
exciting times for the N54!
@jyamona stated that the tables involves crank angle, injector opening duration in ms, energy applied to the piezo stack, 1st and 2nd pulse (there's code for a 3rd too, not sure if / when it is used), and spark time.
We can def control how long and when the injectors are going to be open. Its just putting the pieces together. We wont need port injection IMO until we start pushing the motor to 9k rpm or some stupid whp level. Each injector can flow 277lb/hr or 3000cc of fuel each. The only problem is the injection window gets stupidly small at 9k so youll need PI unless you got some the same injectors they run in F1 lol
OK this is getting pretty cool. I've read about multiple pulse per event DI strategies, as well as injecting during the compression and combustion strokes. Did not know N54 had capacity for the former already, and maybe the latter could be tried out now. Can't wait to see some results from all this.Quote:
Originally Posted by Abacus38
the other thread has the image showing when each pulse is used. 3 pulses was.. 0-2000RPM high torque (except a little gap at lower RPM that is 1 injection), 2000-4000 is 2 injections, 4000+ is one.Quote:
Originally Posted by Abacus38
SO basically all that concerns us is controlling the single injection i guess? given multiple injections, you lose time in opening/closing.
9000rpm isn't much smaller, it's only 12% more revs/minute, still get 3.3333MS/stroke. So if you could, like your above example, inject 1/3 of the way into the compression stroke too for a total of 4.4444ms, that's still what.. .16111g of fuel, more than enough for 1000hp of flow. (yes i know, would have been easier just to take 12% off .18125g, but i wanted to be sure lol)
really, assuming these assumptions are all even remotely accurate, the flow figures are accurate, and you could inject safely to 1/3 of the compression stroke+all of the intake stroke, there's still enough flow at even 10k RPM, for 1000hp given your numbers above.
This is what I was thinking. If you consider part of the compression stroke available, suddenly the time constraint at high rpm becomes less of a limit. One thing not discussed in these back of the envelope flow calculations is the effect of higher cylinder pressure on flow.Quote:
Originally Posted by Flinchy
What I gathered from reading is some DI systems are limited to intake stroke only due to the fact that it is harder to open the injector and to push the fuel into the cylinder. I do not have a feel for whether that matters in N54 case. But consider 30 PSI boost situation, at the halfway point you're at roughly 60 PSI and of course towards the top of the stoke @ 10.2 CR you're at over 300 PSI. That assumes uniform pressure ignoring all the pressure waves bouncing around in there. But even if flow is restricted somewhat you now have whatever % of the stoke duration you're going to use extra to put more fuel in, much larger window.
I don't understand why you'd want to inject during the combustion stroke, but they said it was used for power and only after ignition on the way down. However it eats exhaust valves in certain conditions and of course cylinder pressures are going to be pretty high.
isnt the the fuel pressure something like 3000psi? It seems the injector would still be happy to open?Quote:
Originally Posted by ajsalida
One would think so yes but who knows. There are (or may be) two things going on. The force needed to actuate the piezo element against cylinder pressure, and the force needed to push the fuel. If stock peak boost is say 12 PSI and you're looking to put fuel in at 60 PSI during the middle of the compression stroke under 30 PSI of boost, you're already at 5 times vs 12 during just intake stroke only. May be a concern I don't know enough to say.Quote:
Originally Posted by xbox_fan
The peizo injectors operating range is 50-200bar so its more the capable of injection during the compression stroke but the main concern would be cylinder washed
some DI engines can operate fully stratified mode - Where the injector injects as late as possible in the compression stroke, for fuel economy reasons (and less detonation etc)Quote:
Originally Posted by ajsalida
N54 operates homogenous though, not strat as is stated in the other thread.
so yeah maybe if you were looking for max flow all the way to ignition, it might be a concern you have to calculate for? But that's likely significantly more flow than anyone will ever need for any reason whatsoever still, since even injecting 1/3 of the way into compression is loads of power anyway.
Yes I read that N54 can operate in strat mode however NOx emissions are higher in that mode so you have to read a different type of cat but we in the US can't run this cat since our gas has sulfur in it.
I think that spec is for the fuel pressure side and I doubt there's any issues there. For example the injector has to be able to stay closed under that range or it will leak. What I meant was the piezo element forcing the valve open under ~5x higher than anticipated cylinder pressures on the other side. There was some indication in what I read about some systems not being able to open in the higher pressures during compression and/or combustion strokes. I can't find any spec for that though. May not even be an issue.Quote:
Originally Posted by Abacus38
I used to work with piezo actuators a lot in a former life, for acoustic applications and also vibration control. Very cool stuff but very quirky.
I think what a lot of people are forgetting here is. With the stock HPFP in place alone, its not the injecting, or injectors causing the issue. The HPFP nose dives meaning it cannot even deliver the fuel asked of it for the small injection window it is already using. Trying to lengthen the injection window is only going to compound the situation, if the pump already cannot deliver enough fuel. For this to work we will 100% need the HPFP solutions we have pioneered to use the DI system only to make big power. The main hurdle is, we still are not 100% happy with the control scheme using the double barrel, we got it work, and work well. But I would prefer if the DME had complete control of both, but when you give it control of both, you lose all consistency. There is still work to be done on that front. Terry was going to integrate dual pump control into the JB, but it looks like he has abandoned that in favor if PI control. Maybe one day we will have both.
Subbing to this thread. Starting to get some good info in here :)
I will add some info of what I'm looking at on the DME table side a bit later, trying to figure out units on the table axis
maybe now the tables have been found and it's looking like high power DI only is realistic, it might be a motivator for terry to work on it?Quote:
Originally Posted by Tony@VargasTurboTech
even needing the dual HPFP+JB4 controlling it, it's still a cleaner solution than PI imo.
also, what issues are you running in to with controlling the second pump? does the DME just not have a clue what do with having 2?
I do not understand the DME enough to know why it doesn't like it. But one run you have perfect fuel pressure, you turn around, and do another run and fuel pressure is in the tank. Do another run, and its back fine again. Our control system uses an RPM switch to activate the second pump with a constant regulated voltage, and it works perfectly every time, but controlling isn't quite as precise, as if the DME could adjust it on the fly.Quote:
Originally Posted by Flinchy
How did you have them hooked up when the DME was controlling both? Like a Y connector?Quote:
Originally Posted by Tony@VargasTurboTech
As far as fuel, I am not going to get into how the fuel is routed, it took many many revisions to get a fuel supply scheme that would work properly. My guess is now that these tables are open other DI solutions will surface, and I am not trying to give them a head start. For wiring, the wiring is simply ran to both pumps meaning the DME has control of both FCV's. So in theory if pressure is dropping, the DME should just adjust voltage to both pumps and bring it up. In practice its obviously not that simple.Quote:
Originally Posted by jyamona