temp based water injection
 

Is there a reason none of the kits out there used temp based injection? It seems to make much more sense to me than using boost, as it takes ambient air temp into the equation.

I'm currently programming a raspberry pi to use as my gauges; so far I've got it reading boost and soon AFR.

I'm thinking while I'm in there I can rig up a few temperature sensors and an output for water injection as well. The first sensor will read pre-injection temp, and the 2nd will go after the injector; this way the first temperature can be used as reference for turning the pump on, while the 2nd sensor will let me know that the system is actually working. It probably wouldn't be hard to make it somewhat progressive by adding in extra injectors/ solenoids and turning them on at preset points. More expensive than simply varying the voltage to the pump, but easier to setup and also guarantees proper atomization.

What advantage does it have over boost controlled ones? I'm all for improving things, but why re-invent the wheel if the wheel works fine in the first place?

because some people despise simplicity, effectiveness, lack of failure points, and general ability to function properly.....

..if it works, its too mainstream and cookie cutter

couple of reasons...

1- if ambient temp is low enough to not need the water injection while in boost, you won't be wasting fluid.

2- if ambient temp is high, you might need the injection earlier than the preset boost/ IDC values.

3- having the 2nd sensor in place acts like a built-in monitor. If the delta between them is low when the injector should be firing, it can be used to trigger a failsafe/ warn user that something's wrong.

I'm not saying anything's wrong with using boost/IDC, but at this point in the project it would be very easy to add a few sensors and outputs just to play around with. I just want to make sure there isn't some inherent problem with using temp as the trigger.

if your reasoning is "in the name of science" then by all means, do it and post results.

I think its trying to solve a problem that's not there, tbh.

Your ambient wont' shoot up unless in boost. If it does, its heat soak, and you don't really wanna be spraying water into the engine at idle/cruise rpm. And adding sensors just to save a couple drops of water by not spraying when not hot is obviously not productive.

only time I can see this beneficial is if you're towing something or at moderate load on the highway for extended periods of time and generating heat that way while still not tripping the switch (say 2-3 psi or something)

Using IAT alone does not make more sense than using MAP. A big intercooler and an efficient turbo doesnt mean youre engine is not going to detonate when youre throwing positive pressure at it. Meanwhile high IATs at low load do not call for water injection because the engine either will not detonate, or if it is detonating it is not hurting anything (like with lean burn tuning).
Now if you are using MAP and IAT, that would make sense, even if it is probably unnecessary.

Search for articles about "internal cooling" in the NACA archives and you will find lots of data about the factors influencing detonation and how water injection (internal cooling as they call it) inhibits it.

I agree with the proper turbo and intercooler statement, but if I remember correctly this guy plans to install a mini pooper intercooler in his engine bay.......lol

Speaking from gobs of personal experience with this . . . .

Water injection, despite all the "chemical intercooler" advertising, does not significantly cool the intake charge. It does introduce liquid droplets into the combustion chamber that phase change during the combustion event imparting a great deal of detonation resistance.

In addition, temperature sensors have lag. This would be an extremely difficult control problem -- a small, variable target with lag.

Your MS can do a great job of controlling WI. That's how I do it. Look at my build thread for electrical diagrams and failsafes if you feel like tinkering.

also check out my old thread:
https://www.miataturbo.net/engine-pe...g-w-i-*-62002/

cool stuff guys. I've already got a map sensor feeding it the boost data, it can be setup to trigger the pump as well. I'd still want to keep a post injector temp sensor in there to monitor the system.

MS triggers based upon MAP and RPM.

Monitoring post injector temp is worthless. Given how little the temperature changes and the lag of the sensor, you would have a blown engine before it reliably revealed a problem. You should monitor line pressure and have a way to IBIT for nozzle clogging.

But, whatever (Secretary Sebelius impersonation).

For the sake of argument, what about using EGT as the trigger rather than IAT?

This idea makes more sense.

I would think it would still be based on MAP, but possibly with some kind of EGT or IAT compensation.
The way MS does it sounds like its really the best possible way. You need to inject water based on load, so thats just a VE table! From there you could add compensations, although my guess is that they are totally unnecessary.

I only used 1/2 of the MS method. The water-compatible injector proved to be made of unobtanium, so I just use the MS to trigger a simple, single-stage WI system with failsafes to drop to wastegate-only boost if there is an issue. KISS. Works.

If anyone knows of a cheap, reliable water-compatible injector . . . .

Reverant had 4 EGTs hooked up when he blew an engine. According to him there is no connection between EGT and knock. I can look for the thread but I'm pretty sure it was discussing ways to do ignition autotune (and why we don't yet).

what about using humidity as a monitor?

Solely relying on IATs would lead to the possibility of pooling water at idle depending upon nozzle location. Especially a problem with the heat soaked IAT sensors.

Sitting in traffic - temp based spray at idle

subzero temps and boost - no spray

If you want to use it as an "and" and not an "if", go for it. Spray only works when there is ample flow. You can heat up an intercooler by parking on blacktop.

I considered doing it with a water temp gauge to keep the spray out of a cold engine. I was not tuned for it so there wasn't an issue.

This is as complicated as it ever needs to be:

https://www.miataturbo.net/attachmen...ine=1384015838

this also works just fine:

https://www.miataturbo.net/attachmen...ine=1384015838