More water / meth injection

RICE RACING
Posts: 448
Joined: Mon Feb 24, 2014 1:08 am

Re: More water / meth injection

Post by RICE RACING »

Lqqk wrote: Tue Feb 01, 2022 12:02 am Any recommendations on a flow sensor. I’ve been researching and currently haven’t found anything that looks perfect
Ah only thing I can contribute is have 4 different cars on longer term test *running since ~2017* using the Aquamist combine twin HSV and flow meter manifold unit. I cant vouch for accuracy of the flow meter but they seem repeatable,reliable, and durable so far.
impdnkkdikogmfhn.jpg
impdnkkdikogmfhn.jpg (183.58 KiB) Viewed 186975 times
http://www.riceracing.com.au
https://www.youtube.com/riceracingdonmega
Real turbo road cars fast > reliable > durable
Water Injection Specialist
stevieturbo
Posts: 1321
Joined: Sun Dec 07, 2008 2:04 pm

Re: More water / meth injection

Post by stevieturbo »

I had bought some cheap flow meters ages ago to try...but that came to an end when I had no frequency inputs.

They might have worked on the input side of the pump, as I doubt they'd like the high pressure side. But they came from a coffee machine or something like that. I just got them off ebay, but they were available in various flow rates and operated off 5v
Flow varies with frequency.

As to how stable flow would be on that side of the pump, no idea.

But in lieu of a flow sensor, pressure close to the nozzles is a good alternative ( combined with pump current...or pump current perhaps even in lieu of the pressure sensor )
nmerdan
Posts: 30
Joined: Mon Jan 27, 2020 7:40 am

Re: More water / meth injection

Post by nmerdan »

Some other useful features that I can think of:
1. Multiple solenoids control with multiple maps available for each (now there's 4 maps available, if that's the limit we can make 2 for each). Use case for this is, for example, to have one solenoid for the intake charge cooling which you will use regularly and then another solenoid for the DI jets which are only used on max effort calibrations.

2. Flow meter improvements - this is a little more complex as the flow is affected by many things, but it would be great to have a linearization based on pressure and solenoid duties above. Maybe add a temperature multiplier in the mix and support a temp sensor for people that want to be even more accurate. We'd need two tables, each for one solenoid (if the support for two is added) and then the final flow would be a sum of those two. Then the limp would be be triggered by a percentage error of the final flow expected - this would be useful in detecting things like blocked jets. I'm not sure if that's entirely clear, so to give an example:

To set the linearizations, you would run your solenoid #1 from 0% to 100% duty at pressures you expect to run the system on on the bench and record values. Do the same for solenoid #2. Then let's say the flow through solenoid #1 is 200cc/min at 50% and the flow through solenoid #2 is 150cc/min at 30% (as the duties are separately configurable based on load). And now you install the system and set the allowable error to 10% (this could also be a 2D map based on expected flow, so we can allow larger % errors where the flow is lower). You drive the car and at the point in the maps above, your expected flow is 350cc/min, so limp would be triggered if the flow exceeds 385cc/min or drops below 315cc/min (+- 10%). First would detect a leak somewhere in the system and the 2nd would detect blockage.

If there's a better place to submit these feature requests/ideas, please let me know :)
stevieturbo
Posts: 1321
Joined: Sun Dec 07, 2008 2:04 pm

Re: More water / meth injection

Post by stevieturbo »

There are no strategies in place to deal with a regular injector flowing too much or too little...so any expectation of it on a WI system...just seems way OTT
nmerdan
Posts: 30
Joined: Mon Jan 27, 2020 7:40 am

Re: More water / meth injection

Post by nmerdan »

The WMI systems are often less reliable and durable than fuel injector systems, since a lot of them are self-made and/or often use inferior fittings/lines/filters/etc. which is why additional safeties are always a good thing.

Aftermarket WMI controllers already offer this functionality (example https://www.aemelectronics.com/products ... g-software) and it doesn't look like something too complicated to implement. I don't think it's OTT to be able to detect blockage or a line that broke/popped in part of your system.

The current flow sensor configuration will only detect pump failure and main feed popping off, making it not much different than the pressure sensor.
stevieturbo
Posts: 1321
Joined: Sun Dec 07, 2008 2:04 pm

Re: More water / meth injection

Post by stevieturbo »

If you use quality parts, screwed hose fittings, the system should be very reliable.

The only problems I've had over the years have been from the push fittings most suppliers offer which over time, will fail. Either that or simply a pump failing to operate.

If really getting fanatical.....go for lambda per cylinder then you can instantly keep check for any anomalies on any given cylinder.

Although don't really think that option is fully covered with the ecu's either.
nmerdan
Posts: 30
Joined: Mon Jan 27, 2020 7:40 am

Re: More water / meth injection

Post by nmerdan »

Sure, but you're not only tuning the cars you build, you are tuning remotely, etc. so we're again going back to the fact that additional safeties are always a good thing. With the current way of how the flow sensor operates, I see very little reason to use it and instead run a simpler pressure-based safety.

For my own car I run Peter's fast acting EGT per cylinder, which while not as accurate as lambda, they should provide an additional level of safety.

Anyway, I laid out my suggestions on what I'd like to see based on my personal preference and what other products in the market offer. I already ordered the Syvecs WMI controller, so I have vested interest to see the product improving as I plan on using it extensively.
Ryan.g
Syvecs Staff - Caretaker
Posts: 498
Joined: Fri May 30, 2008 4:05 pm

Re: More water / meth injection

Post by Ryan.g »

nmerdan wrote: Sun Feb 06, 2022 11:52 am
2. Flow meter improvements - this is a little more complex as the flow is affected by many things, but it would be great to have a linearization based on pressure and solenoid duties above. Maybe add a temperature multiplier in the mix and support a temp sensor for people that want to be even more accurate. We'd need two tables, each for one solenoid (if the support for two is added) and then the final flow would be a sum of those two. Then the limp would be be triggered by a percentage error of the final flow expected - this would be useful in detecting things like blocked jets. I'm not sure if that's entirely clear, so to give an example:


If there's a better place to submit these feature requests/ideas, please let me know :)
I was making a change for another dealer who needed some custom PWM output options, while in there i made the Flow Limp based on WMIFinalDuty

Firmware - viewtopic.php?p=7868#p7868

Default Cal - viewtopic.php?p=7832#p7832
RICE RACING
Posts: 448
Joined: Mon Feb 24, 2014 1:08 am

Re: More water / meth injection

Post by RICE RACING »

Here is a 'ghetto' set up doing with just control by Syvecs S7 ECU.
There are many ways to do WMI so if anyone is interested and does not want to do it themselves etc hit up the squirting specialist 8-)
Ghetto set up 1.jpeg
Ghetto set up 1.jpeg (266.76 KiB) Viewed 186850 times
Ghetto set up 2.jpeg
Ghetto set up 2.jpeg (130.15 KiB) Viewed 186850 times
http://www.riceracing.com.au
https://www.youtube.com/riceracingdonmega
Real turbo road cars fast > reliable > durable
Water Injection Specialist
nmerdan
Posts: 30
Joined: Mon Jan 27, 2020 7:40 am

Re: More water / meth injection

Post by nmerdan »

Ryan.g wrote: Tue Feb 08, 2022 12:43 pm I was making a change for another dealer who needed some custom PWM output options, while in there i made the Flow Limp based on WMIFinalDuty
Thanks, that's great! Now all that is left is multiple solenoid support and higher-than-normal flow limp :)
Post Reply