While in the process of fitting a 7 port head to my Mini, I've been looking around for a method that would give the correct injection timing on a sequential injection setup. Well, I haven't found a definitive answer, but many pieces of information that I tried to put together for a “best guess”.
When the car will be up and running again, I will fine tune the settings on a rolling road. But before that, I’d like your comments to see if we can “engineer it right” on a sheet of paper before hitting the road!
Important notice: sequential timing is relevant for emissions and low RPM torque/correct fueling/acceleration response, it is not supposed to increase max power. So basically its timing is only relevant at low/partial throttle, below 3000rpm, when your injectors work at 10%-20% duty cycle. Since the intake stroke corresponds to 25% "duty cycle", where your injectors are working above this threshold, part of the injection process will take place with your inlet valve closed, or with very low inflow (since the valve is opened before the cylinder intake stroke, and closed during compression period), working almost the same way as any "wet manifold" system, injected or not.
The basic idea of sequential injection is that, to get a nicely running engine, with low emissions, lower consumption and better response, all the injected fuel needs to be burned during the power stroke. It seems trivial but the consequences are the following:
- since any cam has some overlap between the closing of the exhaust valve and opening of the intake valve (about 20-25º for OEM, 40-50º for mildly tuned ones, 50º-70º for “trackday”, 70º-110º for “race”), you don't want to inject too early otherwise some of your injected fuel will exit through the exhaust valve without ever being burned, not good for MOT purposes.
- you must time the injection so that the fuel as time to vaporize and be sucked in the cylinder without any fuel "left behind" in the manifold, since, again, some of this “left-over” fuel would exit through the exhaust valve during overlap or evaporate through your air filter.
In my case, I’m using Calver’s RE13OT cam. It’s intake period starts 34,5º before TDC, while the exhaust closes 32,5º after TDC, which gives a quite significant overlap of 67º.
Basically the best way to handle the calculations seems to be going backwards. The good news is that many ECU systems allow to set “end of event” timing for sequential injection: you’re basically saying, all the fuel must be sprayed before XXXº after TDC, and the ECU will calculate when to start opening the injector to do so, given the engine RPM and throttle position.
There is another annoying factor that must be taken into account: travel time. Since your injector is about 10cm from inlet port, the fuel mix needs some time to travel this distance.
I found a website that estimates air flow velocity in a cylinder head (sorry, it’s in French): http://www.flat4ever.com/?page=gaz
The result for my setup (1380cc engine, 40mm throttle bodies, 36mm inlet valve) is 13,7m/s at “throttle body”, and 56,8m/s at inlet valve, but unfortunately those results are for full throttle/ 6000 RPM. At idle or partial throttle, I considered that the average air speed will in fact pretty much correspond to the speed at the valve. Using the calculator, you get this result if the manifold had a minimum diameter of 20mm, which is the case if you consider “part throttle situations”. And there is another “simplification good news”. At any RPM, air speed variations are such that the time taken to travel the 10cm between the injector bung and the inlet valve remains constant in terms of crank rotation! It always corresponds to about 65º of crank rotation… So the last drop of fuel coming from your injector will travel for 65º of crank rotation before getting into the cylinder.
Lastly, we need to consider that there is very little “sucking” at the inlet port after the piston has traveled more than say 80% of its stroke (piston speed is maximum at half course and quickly decreases afterwards), which corresponds to 145º after TDC, so the last drop of fuel should reach the inlet port around 145º after TDC if it doesn't want to take the next train!
All this gives the following result:
The end of the injection sequence should be at 145º-65º= 80º after TDC !
Applying this to my setup, here is the injection sequence for 3 situations:
(My Mini will be fitted with 4x IWP042 Pico injectors that flow 250cc at 3bars. Since I currently run 2x490cc IWP069 injectors in a wet manifold setup, I can easily extrapolate the fuel flow needed for my engine and the resulting injection timings.)
At idle, the injectors should remain open during 1,5ms, corresponding to 9º of crank rotation at 1000RPM.
At a cruise speed of 2500-3000 RPM, the injectors are open about 3ms corresponding to about 40º to 55º of crank rotation, so they will start injecting at 25º to 40º after TDC just after the exhaust valve is fully closed. So we are still in a optimal sequential injection configuration.
At 6000 RPM and full throttle, the injectors will remain open for about 7ms, corresponding to 250º of crank rotation, starting to open 170º before TDC, which corresponds to 135º before the inlet valve starts opening: just as good as any wet manifold setup.
So now, please shoot your comments and/or experiences!!!
Thanks a lot!
Diogo.
Edited by diogoteix, 18 September 2013 - 08:53 AM.
