odonnell

As I get closer to starting my new 951 motor, I'm investigating my tuning options. Something of interest to me is X-tau tuning, which is basically a model for how much fuel actually makes it into the combustion chamber versus being pooled up in the runners or head and not being drawn in until later. Most aftermarket ECUs have this tuning option built in and it's mainly a complement for existing fueling tables - the main job of this "correction" is for throttle transience. Normally people just use the TPS rate of change.

I'm wondering if anyone has at least experimented with it on a 951 and if so, what came of it. It seems like a better approach to throttle transition smoothness with a speed density setup, but that's on paper and may not even be realistic for casual tuning guys. I run TPS-based transient enrichment on my '83 with a Megasquirt ECU and it works fine but has never been perfect across the wide temperature range we get here in TX. Too many stupid curves to adjust, each one affecting the other, etc. The 951 has a MS3 Pro ECU and full sequential, so I'm willing to try a little harder to dial everything in as best I can.

ealoken

I use TPS, IAT, MAP, EGT and lamda.

The Emu then calculates and helps with the table, the Autotune for cruising is wery good, all high boost is dyno material.

steven74

http://www.megamanual.com/ms2/xtau.htm I found this on the megasquirt website. On a sequential injected turbocharged engine, it would be tempting to see how much reversion and wall wetting there is. Would the surface condition, i.e. smoother vs. rough finish, of the intake port walls and using undercut intake valves provide a mechanical means to reduce “wall-wetting” in the first place? On a direct injected engine, the lack of a fuel injector before the combustion chamber is theorized to be the cause for carbon buildup. I would be inclined to think some wall wetness may be useful to prevent oil buildup, especially as the ring condition changes, like during initial break-in and then as they wear further with increased mileage. As far as a transient change setting, I wonder if a modification to the blow-off valve setting could help maintain a very low positive manifold pressure to prevent reversion.

odonnell

Reversion is mostly an issue with batch fired cars (stock) where the fuel "waiting" for the intake valve to open will begin to evaporate off the hot valve but some will splash (or have the mist sucked) into an adjacent cylinder during its intake stroke. Hence why get the best idle seemingly richer than stoich. More fuel has to be injected to compensate for that reversed fuel. With sequential it would maybe be less, but because of the 1/4 & 2/3 cylinder pairs sharing TDCs, there would generally be an adjacent cylinder with an intake stroke occurring during injection on the upcoming cylinder (because we aim to inject on a closed valve). It's not really an issue once the tune is set up either way but it's another thing to account for nonetheless.

For the wall wetting, I would imagine that some pools around the intake valve port because it's not straight. And even then some would form a film on the port walls especially when cold. The nice thing about X-tau is that you set your parameters and it approximates the real-time growth and decay of the puddle/film. It also accounts for the MAP instability at those small time steps - we are 180* apart for firing during a full cycle. The MAP signal will actually be quite shaky if you zoomed in on it with no filtering applied at the ECU. MAP also shoots up with load demand, but by the time the ECU has this fuel quantity calculated and ready to be injected on the next cycle, you already needed the fuel. So I think a TPS based X-tau approach is the best overall enrichment design.

Practical to tune? Not sure yet.

mikey_audiogeek

Strange thing is that if we were controlling the fuel and metering the air (instead of the usual controlling the air and metering the fuel) the natural latency just about dials out the fuel pooling factor.
Sorta expecting one of the manufacturers to go this way given that transmission strength is usually the limiting factor therefore maximising airflow is less relevant. Any takers?

odonnell

So.... you're diesel swapping your 944?

mikey_audiogeek

lol no, just musing that it's easy these days on something like an ls3 to set fuel quantity linear with throttle pedal position and modulate the fly-by-wire throttle butterfly to set air/fuel ratio.
So mapping would consist of setting throttle plate position for each value of fpw and rpm. You lose some airflow compared with the conventional approach but meh.

962 kid

Haven't done it on 944, have on one or two other platforms. It's definitely doable. Your comments about MAP instability aren't really an issue for accel enrich tuning as the correction is usually based on TPS accel rates anyways not MAP accel... if the MAP signal were a viable option you wouldn't need the enrichment anyways haha.

This video helps to visualize what is going on with the "puddle." There is no real airflow here so you can't see the puddle expand or contract but still neat.

Good video on X-tau model:

Adaptronic ECUs are generally set up to use an X-tau model, lot's of reading/viewing material if you search for that.