Nov 16 meeting notes:

Location: Long Range Systems

In attendance:

John Carmack

Phil Eaton

Russ Blink

Neil Milburn

Darin Smith

Next week's meeting will be skipped due to thanksgiving.   Are we back to Tuesday after that, or are we moving to Thursday permanently?

I have started the work to get us incorporated so we have some liability protection.

I have contacted Alpine Insurance (on JP Aerospace's recommendation) about getting launch insurance for us, since Tripoli experimental launches evidently won't cover liquid engines.

I wrote a couple differentially throttled flight simulators to get some feel for how our platform should behave.  I was dead wrong about possibly being able to fly something like this with completely manual controls.  With an attitude sensor and computer hunting to a requested angle, I think we will be able to do a tethered demonstration flight, but it looks like we are going to need some axial sensors and more computer control before flying it completely free.

I spent some time looking into flow regulating valves for our engines.

Relying only on metered jets to control peroxide flow is a significant problem with catalyst based engines.  If some of the peroxide fails to decompose, the chamber pressure drops, so even more peroxide gets shoved in at a faster rate, causing less stay time in the catalyst and a vicious circle.

A cavitating venturi is a type of valve that causes a fixed volume of liquid to flow over a range of pressure deltas.  I have been unable to locate any suppliers on the web, but I did find another type of flow regulating valve at: http://www.mcmaster.com/catalog/106/html/0348.html

This appears to use sliding plates to vary the flow aperture.  The pressure rating is a bit low, and it has a washer of dubious peroxide compatibility, but I have ordered a 0.5 gpm flow versions for us to try out.  The fitting sizes are odd, so we will have to adapt on both sides.

If we add chamber pressure sensors, we may just be able to adjust the PWM duty cycle based on chamber pressure and achieve the same goals, but a mechanical valve solution would be cheaper and less troublesome.

The PWM driver program now has integrated load cell logging, instead of forcing me to run a separate program in another window, which will be a big help next engine test.  I need to add an option to let it drive the latching remote bottle valve as well as the momentary solenoids.

Phil has the solid state relay driver board nicely laid out, with six channels, integrated LEDs, and connectors for both a parallel port driver or a micro controller ribbon cable.

Russ machined down a socket to make a catalyst hole punch, which is a big improvement over cutting the foam with a knife.  We compressed the pack in a press so that we got at least twice as much foam in.  We should count the number of discs, but it is probably difficult to get it apart now.

We should still try and machine a few injector plates with much smaller holes before we test again.  We can test with the old plate and the compressed catalyst pack first, then change the injector plate if we still aren't getting good results.

We have 10 liters of 90% peroxide on order from X-L Space Systems.  All future testing should be with peroxide from a common batch, which will remove a major variable from our results.

We are still waiting on the silver screen order from India.

We are still waiting on the larger load cells from Omega.

We are still waiting on the USB joystick for flight control from the laptop.

We are still waiting on literature for the magnetometers.

We are still waiting on Darin's basic stamps for the accelerometer testing.

We are still waiting on recommendations for leasing a permanent testing site.

We are still waiting on the components for the dumb rocket.

We built the necessary hoses and laid out all the fittings for the VTVL platform tonight.  Russ is going to make a distribution manifold for us with a quarter in and six eighth outs.  We will probably plumb everything together in parallel with engine development.  Even before we have four working engines for it, we can test all the telemetry and electronics with just water spraying out of the solenoids.

We have the fill system all loaded onto a cart now, and it is a lot easier to move around.  We still need to secure the catch tank and battery down on it.

Darin is going to prepare a lecture for us on guidance systems, based on his time with TI.

The cold weather cracked Russ's peroxide concentrator, so we didn't have any high test peroxide to fire today.  Instead, we did a lot of water flow tests to characterize the solenoids and plumbing we are using.

20hz pulse frequency, one second total.

        100psi  200psi

10%     15ml    15ml

20%     20ml    20ml

30%     27ml    30ml

40%     33ml    39ml

50%     40ml    48ml

60%     46ml    56ml

70%     52ml    67ml

80%     60ml    76ml

90%     68ml    82ml

100%    70ml    95ml

On the 200psi tests, we did notice that we had lost a bit of pressure at the 90% level, and topped it up, so the 70% to 90% numbers are probably a bit down.

The solenoid did not open at all with a 2ms pulse with 100psi behind it, and a 3ms pulse was not consistent.  2ms pulses works fine without any pressure behind it, but the pressure obviously effects the responsiveness a lot.

We noted that our water was very cloudy after being ejected from the engine.  Our fill system involves pressurizing with nitrogen after the water/peroxide is loaded, and it seems to dissolve a fair amount into the water.  We don't know if this is also an issue for peroxide.  Yet another reason to try and find a double ended tank, or plumb a second port in our bottle manifold for a siphon tube.

We removed the stock bottle manifold and remote bottle valve assembly and attached the straight shot bottle manifold that Russ had cut.  The not-very-scientific test of blowing through the valves had shown the remote bottle valve to be much more restrictive than the quarter inch solenoid, and our successful firing of the large motor seemed to show that it was flow restricted.

We ran 50% and 100% duty cycle tests, and got the same numbers as before.  This definitively shows that the small 1/8" solenoids we are using are more restrictive than the remote bottle valve and the manual manifold valve, so it isn't an issue for us.  We will run this test again when we move up to using the quarter inch solenoids on bigger engines.

We replaced the original manifold assembly, because the pressure gauge and blow off valve are good measures during testing.  However, the straight out manifold does make for an extremely clean and simple arrangement.  It is tempting to rigidly connect everything together directly onto the bottle for the dumb rocket, eliminating all flexible plumbing.

We then did a couple tests with the different metering jets we had.

.045" metering jet, 20hz pwm, one second total:

        100psi

50%     34ml

100%    56ml

.018" jet

        100psi  200psi

50%     14ml    19ml

100%    22ml    30ml

We then put the small engine with the compressed catalyst pack on to see how much more restrictive the foam bed is now.

Engine with catalyst pack, no jet, 20hz pwm, one second total:

        100psi  200psi

50%     38      46

100%    66      94

These are still only minor restrictions in flow.  We may still not have enough back pressure through the catalyst bed.

Our current full throttle engine design targets are roughly 50N thrust for each engine in the VTVL platform, and 700N for the dumb rocket engine.

If we assume 100s specific impulse, which may be a little conservative, and 1.4g/ml peroxide density, we need to flow 36 ml / second for the small engine, and 500 ml/second for the large engine.

We don't yet know what our chamber pressures are, so we don't know exactly how much pressure drop we will have into the engine, but it looks like the .045" jets are more than large enough for the small engines, and we might consider some in the .030" range.

Some extrapolation does show that we may have problems with the big engine.  The small solenoids we are using now (super powershot) are rated for 175hp of nitrous flow, while the largest size available (super big shot) are rated for 600+hp.  The remote bottle valve flows noticeably less, so it may only flow twice as well as the small solenoid, which would probably require extremely large pressures to get 500 ml/second of flow.

We should test the remote bottle valve without a solenoid at all, but we will need another solid state relay driver and some minor modifications to the driver program to control it.  We should have the completed driver board next meeting, so we can do this test before running any peroxide.

We can run 1000psi in our tank if we need to, but we should find out exactly what combination of pressure, solenoids, and manifolds we need to get the 500ml/second for the big engine.  If none of the standard parts can cut it, Russ wants to make a custom slide valve