Richard L

iTrader: (5)

A gear shift takes between 0.2 seconds to 0.6 seconds when you are racing. Normal DD driving may be doubled or trippled. This is the period when the dribble becomes a nuisance. Boost drops and fuel flow drop. The engine continues to ingest air through the throttle by-pass route.

The un-metered meths from dribbling will be drawn into the engine and causes some unpredictable combustion. In some serious cases, it will cause pre-ignition for the next intake/compression cycle, especially on cylinder one. Less serious conditions will just be a few pops and bangs. Sound a bit like the anti-lag system used on on rally cars.

Looking at the video page 1, the 16-feet tubingm, bleed-down took about 2-3 seconds and 2-feet tubing about 0.5 to 1 second. These are only estimates - you can time it yourself. Using a bigger nozzle will only shorten the bleed-down duration but the dribble quanity will be the same.

I will post a log chart for a 0-130mph run at WOT. You can see the gaps. Unfortunately I don't have a time stamp on the log.

3gEclipseTurbo

iTrader: (12)

Sorry if this was covered but why not use and injector instead of a nozzle? That way the injector closes and no dribble.

Richard L

iTrader: (5)

Yes it is an alternative but carry its own set of problems.

Fuel injector is design for fuel only and all wetted part are made of iron based alloy. No matter how pure methanol is used, over tome, it will aborb enough water to cause internal corrosion.

Valve pinto and seat are precision ground, without the lubricating properties of gasoline, it will wear quickly. As a result, it will start leaking. Further more, injectors are designed to have effective seal upto a certain line pressure. This is no guarantee that it is leak-proof beyond 100-200psi. So you have to operate the differential line pressure to around 30-40psi.

If you can overcome the all the problems above, you will have a good system, but will be costly.

Richard L

iTrader: (5)

I have created a chart from a log file of a run on 1-4th gear (up to 130mph), covering the signals of boost. rpm, road speed and fuel IDC. Max Boost ~15psi (100kpa).

With the help of those data, we can zoom into each section of the chart and study them in finer details. I will post more charts later and perhaps we can discuss how a wmi copes with those fast changing signal signals during lift-off and turbo ramp.



Notice the boost spool-up (blue trace) started to taper downwards from ~5000rpm compared the Fuel IDC (orange trace) continues to climb to ~7000rpm.

Richard L

iTrader: (5)

I have managed to add a trace (yellow) to the chart showing what the run-on/dribble might look based on the videos. In between during gear changes, the boost decreased rapidly but the meth flow only dropped slowly.

Until I have a true real time log, this is only an assumption.

Talonboost

Richard - detail question - what software did you make this chart in? Usually I find multiple plots on the same chart annoying as hell and headache inducing. But this one is very nice.

Richard L

iTrader: (5)

I use ms excel. I haven't done anything special other than selecting different backgrounds and line colours.

I only picked the columns I need from the log. It is just a drag and select graph.

Richard L

iTrader: (5)

I have been thinking quite a bit on the crack piston thread by Mike. What do you guys think if the knock was caused by pre-ignition and not by bad methanol distribution under over-lean conditions?

C6C6CH3vo

iTrader: (9)

I believe anything resulted in parts flying through the hood or block, or breaking off is from preignition. The knock sensor might say 1 count, but that one count is from the sound of breaking a rod in half.


Too many factors to conclude the cause of these former events with the available data. All we can say is too much tune for the cyl that went.

For exp:

1 and 4 cyl will see end gas reversion 1st typically from the larger scroll runner volume (esp 1). cyl 4 might be the 1st to heat up from excess egr flow, but cyl 1 probably takes on the most PVC oil (oil has an equivalent octane of about like... zero!) Even a single spark plug not torqued down to spec can cause preignition.

Richard L

iTrader: (5)

I tend to agree in most parts, especially the frame reversal. I really am not aware of cylinder #1 & cylinder #4 is prone to this, especially with all those race cam profiles. Learning something new today, thank you.

The bit I have some doubt is, why cylinder #1 & 2? Is it due to the residual meth migrating their way into those cylinders during lift off periods? This is before some bits broken off. One ring land goes, oil induced knock just takes over.

Engine mangement will not be looking for pre-ignition (wrong window) nor able to do anything about it.

JR From AEM

Really? so, what, are you just making stuff up now Richard?

I would think that before you create such a blatant falsehood as this graph you would take a few minutes to test it for yourself. You seem to now be basing your entire advertising claims on the "dribble" yet by your own statement above you have no real data.

According to the first shift point the downward slope of the PP graph predicts a 45 second dribble time. FORTY-FIVE SECONDS?. Yet in your own posts earlier in this thread you admit it could be as short as half a second. Why did you make such a joke of a picture? It destroys any credibility you have, at least with anybody who actually looks at it closely.

You guys have 15 psi check valves, injectors, pumps and flow gauges in house. How about taking a few minutes and actually getting the data BEFORE you fabricate graphs and publish them, OK?.

Test, then report. not the other way around.

wreckleford

iTrader: (2)

Wouldn't pulse width be a better parameter to tie meth. delivery to instead of duty cycle?

Duty cycle will normally tend to increase with rpm while pulse width, i.e. actual fuel demand may not be increasing.

wreckleford

iTrader: (2)

Your point may still be valid, but I don't think the x axis is in seconds. Those would be some really slow shifts. I think RIchard even mentioned this somewhere.

jo evo

iTrader: (2)

Im with u wreckleford,if the x axis where in seconds what car takes 160 seconds to reach 130mhp?

JR From AEM

It depends on if you are doing discrete injection events with the water/meth kit (i.e. variable frequency depending on engine speed like the fuel injectors itself) then pulsewidth would be better for the reasons you mentioned. If you are doing non-discrete injection events (constant flow, fixed injection period with variable duty cycle) then injector duty cycle would be better than pulsewidth.

If the unit if time in this graph is not seconds then the graph is even more useless and misleading than I first thought. Remember, the graph is supposed to be showing "during gear changes, the boost decreased rapidly but the meth flow only dropped slowly." How can there be anything rapid or slow without the X-Axis being some meaningful unit of time, i.e. seconds.

On top of the misleading time scale the drawing seems to suggest the flow only dropped about 15% over the course of a slow shift (that's what the first one looks like to me) so even if that was a 1 second shift his 0-130 MPH LOG tries to depict run on lasting 7 seconds or more, easily 7X or more than reality.

Richard, If you don't have the data don't just make stuff up to support your marketing line. You have Aquatec pumps, 15 psi check valves, high pressure line, 400cc injectors and flowmeters. Why did you post this misleading "log" when you could have the real data in 30 minutes??