There are failsafe, failsafe and failsafes...

[Richard L]

I think we have discussed most of the systems available and their capabilties. It is a good moment to see what failsafe mechanisms are available from the simpest to the most complex. I will update the list when more contributions are posted, if any.

Direct detection:

1. Fluid tank level sensor :
Advance warning but only good for early warning but little use for system power failure or blocked nozzle.

2. Loss of system power:
If power is interrupted to the WIA system, it will de-engergise a relay and performs "Plan B", will not detect blocked jet or cut pipe or no fluid flow.

3. Inline pressure gauge:
This is very cost effective but require user paying attention.

4. Inline single pressure switch:
If the system looses fliud pressure during activation, an inline preset pressure switch can action a boost drop etc.

5. Two inline pressure switches:
Detects blocked nozzle (over pressure) and "lost of fluid pressure" (under pressure). This method is quite cost effective and within DIY capability.

6. Flow switch:
In inline flow switch, comprised of spring-load magnetic plunger inside a tube where the water flows through. The rate of spring determines the promixity of a reed switch again the magnetic pluger deflection. Output: on/off. - cost: low

7. Turbine flow sensor:
Good for progressive system but requires complex electronics circuitry to report a differnce between the "actual" and "planned" flow in real time.


More..?

In-direct detection:

1. Knock detection:
This a by far the simplest, comes free with your car. It will only work up to certain power upgrade. Beyond a few PSI from stock, the ECU's ability to wind back ignition would be heavily taxed and may not dial in enough retard to cope with heavy knock. Don't forget, retarding igntion will also drastically increase your combustion temperature - not good.

2. AFR tracking:
If your ECU is capable of regulating the afr at WOT, it will be very effective. Unfortunately factory lambda is not of the "wide band" type so without equipping yourself with a wide-band lambda probe system, it will not work very well. The reliability on this method may lie on the accuracy of you wide-band interfacing mechanism. Method is effective but no good for mix other than high concentration of alcohol.

3. EGT tracking:
This method is rarely considered. Reading EGT has many benifits, it can indirectly estimate your afr, you ignition timing and more importly, when is your pistons are about to melt. It is a pity that not many EGY gauge gives you an analogue output so that you can use it for "Plan B" if the EGT exceeds certain preset value. I believe the SPA digital gauges has a "hi/lo" detection circuit inbuilt and an output capable of switching a relay. Need some research.

3. PWM vs pump cycle-rate tracking:
This method can only be incorporated by the manufacturer of the water/alcohol injection systems. It looks at the duty cycle of the WI system (PWM type only) against the predicted flow relative to the pump speed. If it is out of the predicted envelope, it outputs a "fault signal" to switch a relay. Difficult to incorporate into a rotary pump due to the lag time caused by the rotating mass of the motor.

3. PWM vs flow sensor tracking:
Self-contain system that reads the actual flow against the WIA valve's PWN or a fixed flow system. Very reliable but require some electronic processing.

4. Inlet temperature tracking:
If two temperature probes are placed "before" and "after" the water jet, any temperature differential will indicate the presence of "latent heat" at work.

I leave it as it is until the this topic attracts more interest. It is important to know what how "safe" is a "failsafe" . If anyone and suggest more method, please post here and I will update this database of this failsafes.

Richard

[Ultimate Collision Center]

Very informative post

[bcblues]

Thank you! Excellent sum.

[Richard L]

This may help someone to visualize the less well known sensors:

A rotary sensor:



A piston flow switch :



Richard

[ejh25]

Originally Posted by Richard L

This may help someone to visualize the less well known sensors:

A rotary sensor:



A piston flow switch :



Richard

Thanks Richard. Very helpful!

[TheMadScientist]

Very nice post.
But you forgot the "BEST" failsafe. The loose nut behind the wheel.

TMS

[AaronB]

Nice post. What about adding which kits include which failsafes off the shelf? Might make one helpful reference post...

[Richard L]

Originally Posted by AaronB

Nice post. What about adding which kits include which failsafes off the shelf? Might make one helpful reference post...

I will make a post on the failsafes employed by each WI maker but need to do some research first, I will try to be as acuurate as I can.

If there is any inaccuracy in my assessment, please post correction.

As WIA become more and more popular, there will be occasions where tuning boundary is pushed beyond SOA "Safe Operation Area". The functions of the "failsafe " mechanism becomes paramount.

Please be patient, it will take some time to collate the information.


Richard

[ucbsti]

I think failsafes should never be singlepoint

I'd have two failsafes on my unit, I think I'd try something like this.

1. Loss of power. A simple relay could solve this problem. Have the relay only powered when in boost ranges where you will be spraying (pressure switch), when you are in boost, power SHOULD be going to the pump. If its not, have the relay send a audio/visual signal or cut fuel/boost

2. A dual pressure switch for the conditions of no flow or clogged nozzle. There are many pressure switches (digital) out there that have dual output, so you can have 2 pressure setpoints. Once again, id have the relay required for this be boost activated as well, so the pressure switch isnt sending out signals when you arent even in boost and not using meth/water

Materials needed
2 pressure switches-1 single setpoint for boost, one dual set point
Automotive relays, 2 if you dont care to know if its low flow or clogged nozzle, 3 if you do.
some wires/solder/etc
some electrical know-how

[Richard L]

Originally Posted by ucbsti

I think failsafes should never be singlepoint

I'd have two failsafes on my unit, I think I'd try something like this.

1. Loss of power. A simple relay could solve this problem. Have the relay only powered when in boost ranges where you will be spraying (pressure switch), when you are in boost, power SHOULD be going to the pump. If its not, have the relay send a audio/visual signal or cut fuel/boost

2. A dual pressure switch for the conditions of no flow or clogged nozzle. There are many pressure switches (digital) out there that have dual output, so you can have 2 pressure setpoints. Once again, id have the relay required for this be boost activated as well, so the pressure switch isnt sending out signals when you arent even in boost and not using meth/water

Materials needed
2 pressure switches-1 single setpoint for boost, one dual set point
Automotive relays, 2 if you dont care to know if its low flow or clogged nozzle, 3 if you do.
some wires/solder/etc
some electrical know-how

This is a good solid start to "failsafe construction". Cost is low and very effective. It will work very well with on/off system. May be further thought on how the same idea can be used for a progressive system or a partial clogged nozzle.

Richard

[Richard L]

I have put the failsafe details on each of the WI kit manufacturers, re: "whose, who, what...." post.

This thread deal mainly with different kinds of failsafes. Some failsafe is more safer than others. If you can think of any other safe way to detect WI failure, please contribute - I will put them in on the first post.

I am trying to urge people to consider the importance and effectiveness of a failsafe system before tuning your engine too aggressively.

Richard

[AaronWRX]

Originally Posted by Richard L

I am trying to urge people to consider the importance and effectiveness of a failsafe system before tuning your engine too aggressively.

Richard

Yes this message is lost on some people. I've seen too many posts recently where people are running 100% meth on smaller turbos to get big turbo dyno numbers using reflash EM with no visible knock detection.

over on the other sti forum it seems that people think WI in in the same bolt-on category as a downpipe.

[Richard L]

Originally Posted by AaronWRX

Yes this message is lost on some people. I've seen too many posts recently where people are running 100% meth on smaller turbos to get big turbo dyno numbers using reflash EM with no visible knock detection.

over on the other sti forum it seems that people think WI in in the same bolt-on category as a downpipe.

One day when WI has all the failsafe inbuild, it will be as simple as a bolt-on
tailpipe. May be?

Richard

[TROLL]

ok just interested in covering all the bases when i do go ahead and install my Aquamist 2D kit and i dont yet fully understand the failsafes integrated into the kit in base form, and the add-ons for monitoring the system and adding additional methods of failure detection.

[Richard L]

Originally Posted by Wrinkleboi

ok just interested in covering all the bases when i do go ahead and install my Aquamist 2D kit and i dont yet fully understand the failsafes integrated into the kit in base form, and the add-ons for monitoring the system and adding additional methods of failure detection.

I will try to explain how the 2d's failsafe works and now much grounds it covers.

In order to detect a flow fault in a constantly changing system, it is almost impossible using the conventional method such as a pair of pressure switches or a flow sensor.

The 2d diagnostic system reads the incoming PWM signal from the fuel injector and compares it against the PWM signal from Aquamist piston pump. If there is any inbalance between the signals, the comparator will throw a error flag and switch the orange wire to ground.

Error detecting capabilities of the 2d are listed below:
1) Blocked jet or partically blocked jet (error detection sensitive can be increase on request)
2) Cut pipe between the HSV and jet during injection.
3) Cut pipe between the pump and HSV during non-injection period.
4) Water pressure drop during injection (caused by clogged filter)
5) Loss of system pressure during non-injection period (such as trapped air in the system)
6) Water pump failure
7) Water pump fuse blown
8) Sweat on your forehead during racing (on special request only)

One thing you must do to ensure the system is in working order:
Upon turning on the ignition key, the fault led will light up for 5-10 seconds and goes out. If the led doesn't go off, the system checking routine has located a fault.

I am sorry to be a bit vague on the design of the diagnostic routine, it is one of the rare system on the market that is able to score quite well on the "failsafe" department.

The same "failsafe" circuitry is applied on our system2s as well. I hope this write-up has more or less answered your question. Please let me know if I didn't explain it well enough.


Richard

[jblaine]

Thanks for this thread, RL.

[TROLL]

great idea for a thread... i'm interested to follow the discussion of failsafes and their pros and cons.
i'm also interested to outline the different failsafe options readily available either as individual components or as part of popular kits from companies like aquamist, snow performance, smc when it is released, and so on... did you say you wanted to save that discussion for a different thread or that you are just in the midst of compiling that info? if so i'm sure you could outline the failsafes from aquamist at least right now .
thanks, i really enjoy reading the water injection forum lately with all of the good info that is posted in such a cooperative (not competitive) nature.
bryan

[Richard L]

Originally Posted by Wrinkleboi

... did you say you wanted to save that discussion for a different thread or that you are just in the midst of compiling that info? if so i'm sure you could outline the failsafes from aquamist at least right now . bryan

This thread is originally intended for "failsafe" discussion, I would like this thread to continue if possible, the title is very direct.

it was originally posted in March but hasn't attracted many replies, may be the topic is now becoming more important with one or two reported WIA problems.

I will outline the aquamist failsafe in as much details as possible even it means other WI makers copying some of the ideas, so be it. I just like to see the WI injection concept grows especially the cost of fuel in going up each year.

I do need more people to contribute to this topic - however far fetched.

Richard

[scooterforever]

Richard --

I am running the 2d and am very pleased with it (it's a KTR special, with the tank in the back).

One idea: right now I use the fault wire (the orange one if I recall correctly) to light a warning LED. What I would like to do is also use this pull-to-ground to disconnect my stock boost control solenoid, which will drop the turbo down to wastegate boost (I think I am breaking this down correctly).

I could just use a 'normally closed' relay in line with the power lead to the boost solenoid. When triggered, the relay would sever (open) this connection.

The problem I have: I want this 'failsafe action relay' -- for lack of a better term -- to ignore the first trigger of the warning LED, which of course is the normal "system check" on startup.

I am attempting to build a relay system that ignores the first LED warning by using two DPDT relays, but... I'm not that good with electronics. I am sure I'll figure it out.

So here's an idea for the next iteration of Aquamist products: an additional warning wire that pulls to ground after the initial startup routine (but not during) would be great. The regular orange wire should stay, of course, but a second "only when there's a problem" wire would be helpful. In the alternative, you could market an add-on double-relay product that would use the orange warning lead but ignore the startup ground.

Great thread by the way. Satisfied customer
-- scooter

Edit: the reason I don't want to just have the boost solenoid disconnected on startup is b/c my understanding is that that will throw a CEL.

[bcblues]

scooter: Are you trying to avoid the CEL during the startup routine? I assume that as soon as the warning light circuit is closed, the relay would also close and it would be business as usual, except perhaps for throwing the CEL at startup.

[Richard L]

Scooter,

You can put a 39ohm 25W resistor between +12V and the n/o contact of the 5-pin relay. This way, the ECU will be fooled in thinking the boost valve is still there during the 5-10 second period.

Richard

[ucbsti]

I see you left out flow meters. They are used for fault/flow control in precision chemical delivery systems

They can be ultrasonic (super expensive, but highly accurate), or vortex based (more common, less accurate, but much cheaper)

A vortex unit could be adapted to be used with an alky kit. Most have 0-10v or 4-20ma output and can be powered by 12V

[Richard L]

We have designed quite a cost effective flow sensor, signal conditioned.
0-5V output, ranger 100-450cc/min. Can be extended to 1 litre/min. when used with DDS3 (Dash Display System) flow monitor dash gauge. Suitable for water/methanol/mix.



If anyone is interested to know more, I will post more information.

Richard

[ucbsti]

Originally Posted by Richard L

We have designed quite a cost effective flow sensor, signal conditioned.
0-5V output, ranger 100-450cc/min. Can be extended to 1 litre/min. when used with DDS3 (Dash Display System) flow monitor dash gauge. Suitable for water/methanol/mix.



If anyone is interested to know more, I will post more information.

Richard

Very nice

Id like to know if its integrated into the controller to make sure the system is flowing what it should be at all times

[Phil Jr.]

Originally Posted by Richard L

We have designed quite a cost effective flow sensor, signal conditioned.
0-5V output, ranger 100-450cc/min. Can be extended to 1 litre/min. when used with DDS3 (Dash Display System) flow monitor dash gauge. Suitable for water/methanol/mix.



If anyone is interested to know more, I will post more information.

Richard

How much is this part? And do you think it could be wired to simple flash a light when a clog is detected?