Are two injectors necessary?

AD_ward9

The fault tolerant architecture is shown in the General One_Set_of_Solutions_to_CCR_Safety.pdf.

One question, is "Are two injectors necessary/desirable?"

There are two sub-issues:
1. Use of cheap solenoid type injectors is undesirable: they fail too easily. For example, put on a different first stage with higher interstage pressure, and they fail straight away, usually open. The injectors in common use are for machine automation and were never conceived as life critical components. A much more suitable injector is a needle valve or pinch valve, controlled by a stepper motor or DC motor, in each case with position feedback.

2. Even the best injector cannot achieve 1 billion hours MTBF. Two can. BUT, when they fail, they tend to fail open, so this means the valves should be series or parallel, controlled by a third valve. Complexity goes up a big notch.

Debate welcomed.

Cheers,
Alex

MichaelS.

I have been thinking about this post for a while, and have a question ho much does a injector cost and what kind of complexity are we talking about with 2 or even 3 injectors?

MichaelS.

Another question is it possible for either prediction software or hardware to be added for detection for possible failure, or is it just better for good old fashioned common sense testing prior to use?

PeteS

If:-

a/ you have two valves in series, and one fails open,

or

b/ you have two valves in parrallel and one fails shut,

how do you know? When the second valve fails? In which case why have the complexity of two valves.....

__________________
Pete S.

http://www.steggle.co.nz

Every silver lining has a cloud

CCR900

Alex

Most (if not all) ECCR's have two injectors. If the primary electronic valve fails the diver is taught to use the manual bypass (and isolate the primary electronic valve). The easy and speed of isolation varies from machine to machine.

Placing a secondary solenoid would imply a seperate batteries, gas supply/1st stage & controller. All that means complexity which given the unfriendly environment most of us dive in implies a greater risk of equipment failure/aborted dives.

One issue might be to plug the manual bypass into a seprate O2 source.

F

Duncan Price

I've been thinking along the lines of a needle vavle controled by a stepper motor. I'm not sure how much feedback you require. Personally I've have an autotuning PID loop dirven by the O2 cells. Effectively you'd end up with a KISS-needle valve system like I use except its driven by the CCR rather than the wearer. Should be more stable in operation.

Not sure about series injectors. Might work - be worried about synchronisation and power consumption.

Are you proposing to incorporate manual O2 and diluent injection as well?

Duncan

EBT

Echoing petes thought trail really. Surely if you have a manual injector in a parallel route, you've got the best of both worlds without additional fault risk?

__________________
Eagles May Soar, but weasels don't get sucked into jet engines!

Rebreather World Terms of service
Real diving t-shirts for real divers

Nad

I second this. It would have more electronic, it need more space etc.

Nad

__________________
The Impossible is often the Untried

Genesis

A stepper-driven needle valve would be very nice.

Its also quite large, and hermetically sealing it becomes a problem. Further, I'm not sold on this being any more stable over time than a solenoid valve. The failure modes are different for them, but not necessarily better. For example, power loss or an internal fault in a stepper motor results in the valve remaining wherever it was last.

Solenoid valves are trivial to seal (the coil and valve portions are usually separate, the valve is already sealed, and the coil frequently is available as a potted, sealed unit), they're reasonably cheap and they work. They also close if power is interrupted for any reason (e.g. controller or battery failure, wiring failure, etc.)

The advantage that a stepper has is that with positional feedback in normal operation the software can determine where it is (assuming closed-loop operation), which is not something most solenoid valves have (e.g. a secondary switch to indicate open or closed.)

I see arguments both ways on this, and I'm not sold on either being better than the other. Both have failure modes and reaching the desired MTBF without two inline is not likely with a single unit of either.

The big advantage I can see to a stepper drive is that that it can continually vary the delivery. However, since you usually want to inject the O2 before the scrubber and sense after, in order to promote mixing of the O2 with the loop gas (and preventing a "slug" of high FO2 gas from being delivered to the diver) it would seem that a stepper drive might not actually produce any better results "in real life" than a solenoid does in terms of PO2 control..... that would require experimental validation.

PeteS

Feedback from the valve is not required. You already have feedback from the loop PPO2.

Surely pushing more gas with a consequently lower PPCO2 through the scrubber would reduce the ability of the scrubber to remove that CO2.

And "slugs" of O2 don't seem to happen on the Inspiration. The graph I get from the vision software reflect this as well.

I think that although the stepper/DC motor controlled needle valve may be a very elegant soloution, I do think that you are adding an unrequired and unneccesary level of complexity - Software, Wiring, Sealing, Maintenance, etc.

__________________
Pete S.

http://www.steggle.co.nz

Every silver lining has a cloud