Thanks everyone for the ideas and suggestions. Some comments:
Spring return valves would be nice but I selected servo controlled ball
valves because 1) I was able to make a lightweight 1/2 inch full flow
cryo valve with low torque, and 2) once the vehicle takes off, I don't
want a power glitch to shut it down prematurely. I'll have a backup
system for parachute deployment if the flight computer fails.
After turning on the CDI to the spark plug, I wait 500 ms then check for
igniter pressure and turn off the CDI. The igniter propellant solenoids
stay on until the main valves open and it checks for chamber pressure.
EMI with the CDI used to be a problem but after I switched to miniature
1/4-28 spark plugs and an RC airplane CDI module, I haven't seen any
noise on the analog transducers. However, I am running it from a
separate offboard power supply though as I wasn't able to sufficiently
filter the spikes on the power rail.
I always clean everything from the tank outlet tubes downstream after
each run just to be safe.
-Bob
On 02/07/2018 09:23 PM, Robert Watzlavick wrote:
I already found a bug. There's no reason the vents shouldn't be immediately opened:
1. Close the main propellant valves, open the vent valves, and close the helium valve (they may have to be staggered for current limit reasons though).
2. Wait the expected time for the main valves to close, about 1 second.
3. Close the igniter solenoid valves and turn the CDI off.
4. If the main valve closed and the vent valves opened, re-open the helium valve to vent it.
-Bob
On 02/07/2018 09:10 PM, Robert Watzlavick wrote:
I'm coding up the ground abort logic in the flight computer for my liquid biprop rocket and I'm trying to determine the best abort sequence. The main fuel and oxidizer valves are on a single servo actuator, each tank has a separate servo controlled vent valve, and the helium valve is servo controlled. There is no way to vent the helium tank except through the regulators, into the tank manifolds, and out the tank vents. The augmented spark igniter has separate solenoid valves for each propellant and separate control for the CDI used by the spark plug. There are a variety of different conditions that can generate an abort (valve failure to open/close, low chamber pressure, no CDI sense, etc.) but I was thinking of the following general sequence:
1. Close the main propellant valves.
2. Wait the expected time for the main valves to close, about 1 second.
3. Close the igniter solenoid valves and turn the CDI off. The idea behind waiting a second before closing the valves is to encourage any propellants currently being injected a chance to burn off in a controlled manner. If the main valves got stuck and only partially opened, the main chamber pressure might not reach the expected value so that would trigger an abort. There is a single pressure transducer on the igniter that I'm using to verify igniter operation as well as main chamber operation. The expected igniter pressure is 130 psia and the main chamber pressure is 250 psia so I can tell when the main chamber starts up.
4. Close the helium valve and open the vent valves. Leaving the helium valve closed doesn't leave the vehicle as safe as dumping it through the tank vents but if there is a failure of the main valve to close, then it seems that shutting off the high pressure supply would help keep the fire from being fed by propellants under pressure. They still would all leak out, just more slowly. But if there is a fire that eventually destroys the ability to control the helium valve, then the tank will remain pressurized so it seems it would be better to go ahead and depressurize the helium tank. All tanks have burst discs on them including the helium tank.
Suggestions? I could also tailor the sequence based on the pre-abort conditions. For example, if I know the main valves successfully closed and the vents opened, leave the helium valve open. On my test stand, the abort would simultaneously close the main valves, close the igniter solenoids, open the vent valves, and shut off the helium valve (it was supplied from a cylinder through a solenoid valve).
-Bob