Preturbo 335i Strange Results

[Richard L]

I think we can now focus on the job in hand.

Correct me:

Meth off:
no timng pull, iat inceased with load

Meth on:
timing pull, iat increased considerably. (edited)

[rudypoochris]

Originally Posted by Richard L

I think we can now focus on the job in hand.

Correct me:

Meth off:
no timng pull, iat inceased with load

Meth on:
timing pull, iat drop considerably.

Close. More like this (have logs if you want me to email them):

Meth off:
No timing pull, iat increase only 5F (normal)

Meth on:
Timing pull, iat increase 22F (abnormal)

Basically the opposite of what I hoped for. Maybe the liquid is not atomizing? Condensing on the IC walls? Since I have a check valve to the meth tank, when meth comes on at 8psi manifold, the tank is still seeing 18psi or so from the previous run. This is a problem for atomization. I will remove the check valve and have a pressure regulator on order.

[Richard L]

This a very interesting. I like to go a bit deeper.

If you can email me the log richard-at-aquamist.co.uk. The charts you posted are not easy to read, some of the colours are very similar. Please made some simple comments with the log attachment.

Question did you do the windshield observation test? If so, did the spray looked good?

[reid-o]

I'm interested in the outcome as I have the same air assisted nozzle plumbed into an aquamist hsf2. Testing flow without air pressure on the nozzle and with the flow valve turned to minimum I get 400cc min.spraying into a gallon jug by activating the hsf test procedure. I have not verified for atomization but assumed that atomization should be fine by keeping flow down.

I also have a solenoid in the air line going to the air assisted nozzle that opens at the hsf activation.

I don't see how you have it connected but you may want to add an air solenoid in line going to each nozzle so as to avoid over spinning the turbos before 8 psi.

Why do you have a check valve if you are running a water valve? When closed it should prevent dripping.

Is your system maf based? Although the air returning through the system, I would be worried about the effect it had on metering.

I'm worried now about the pressure and atomization.

[Richard L]

I believe the set up is purely mechanical. The manifold pressure is fed to tank (via a checkvalve) as well as to the nozzle.

[reid-o]

I would assume that 2 lines from the charge pipe to open nozzles is a fairly large leak even if the air is being circulated past the maf sensor. I suspect that's one issue for increased iat.

If there's an 8 psi check valve on the water tank wouldn't that mean reduced water pressure as it has to overcome the check valve?

[reid-o]

Originally Posted by Richard L

I believe the set up is purely mechanical. The manifold pressure is fed to tank (via a checkvalve) as well as to the nozzle.

Yes but there are 2 lines from the charge pipe directly to the nozzles unless he has a second check vslve there too.
Seeing as how I see 3 ports on the charge pipe, I think it goes directly to the nozzles...ie constantly bleeding pressure

[Richard L]

You could be correct, I am not absolutely clean how the setup is routed. There are a few blocks, not sure if they are valves or checkvalves.

He will be posting more, hope to have a diagram to see how the whole thing is connected up.

[reid-o]

I did a boost leak test once and forgot that I had the air assisted nozzle connected. I was surprised how big the orifice was on the nozzle. It's likely a bigger problem since he has 2 of them.

I could be wrong though. Either way, I would switch over to solenoid to avoid the pressure drop from the check valves.

[Richard L]

Do you have the same car, a 335i? I believe he has a solenoid valve as well as a checkvalve. The water line is triggered by boost (post throttle). Not sure he if has the same arrangement on the air line.

If his car does not have MAP in between, it may not have mattered. I think the 335i is a speed-density fueling system.

Reducing the water pressure relative to the air pressure can only improve the atomisation.

In any case. the setup should be bench tested before installation, baseline data helps to trouble shoot should if things are not going according to plan, as happened in this case.

He tests were done on a ~45F environment. Water/methanol droplet will condense when it passes through the intercooler. I expect the cylinders would receive varied amount of water/methanol. The DME was unable to trimmer fuel flow to each cylinder as there are only two lambda sensors, one for each bank of three. This could be the reason why the reason was not good.

I think one should only injection just enough to cool the turbo. He is injection around 650-700cc/min.