Sunday, August 09, 2009

Thats why we test...

We tested again Saturday. This was the first time I was responsible for the FAA notification and not just piggy-backing on the FAR standard waivers. The Airspace at FAR is interesting. From ground from ground level to 1200 feet AGL we are in uncontrolled airspace. From 1200 ft AGL l to 18000 ft MSL we are under the FAA jurisdiction, from 18000fto to 50K ft we are under the jurisdiction of the Edwards 2508 restricted area. Since I’m under 1200 ft AGL for tethered (or free) flights I don’t need a waiver or authorization to enter controlled airspace. Since I’m not entering controlled airspace. Yet the new FAR 101.27 requires I notify the nearest ATC facility between 24 and 72 hours before the event. A lot of the stuff flown at FAR goes above the 18K limit so when I asked FARS primary Pyro op who to notify I was given the 2508CCB contact. They did not want the notification because we aren’t in their airspace. When I called the loacal High Desert ATC facility

ATC: how high,
ME: 45ft
ATC: Why are you telling me this.
ME:Can I have your Fax number.
ATC:Here it is
ATC:What am I supposed to do with the document your faxing?
ME:I don’t really know I just know I have to send it to you. It does not effect your airspace were chained to a 10000lb forklift.
ME: FAR 101.27 says I have to tell someone. I’m crossing my t’s and dotting my i’s.
ATC I see.

So I faxed the notification required of 101.27 to the High desert ATC facility.

A few weeks ago our FAA representative asked if he could come out a watch us fly so Wynn came out to observe our flights and flight preparation. We were all setup by 9 am and were attempting to do a 90 second Tethered hover. We had never successfully flown at that weight. We didn’t do it on Saturday either. You can see the heavy takeoff video here our vehicle is VERY light without payload or propellant we weight less than 100 lbs. Since we are using a low performance propellant so we need a lot of it so our weight with payload and full fuel is about 400lbs. This means that we have a 3:1 weight change in flight. From the video its obvious that with a 3:1 weight change we need to do some gain scheduling as we had plenty of thrust, plenty of control authority we just weren’t stable. We continued to do short hops until we had dissipated enough propellant to fly stably. We eventually got an absolutely perfect 65 second flight but that happened after the one remaining video camera ran out of battery. So no video. We will try again soon.

I think that Wynn (our FAA rep) was impressed with how remote the FAR site is. You just can’t get a feel for how remote things are by looking at google earth. Standing on site and seeing the total desolation in every direction gives you a better feel. I think Wynn enjoyed the show, I’m glad we got at least one successful flight off for him.

Thoughts on the stability issue: With a gimbals system the forces generated by the thrust vector scale directly with the motor thrust. With vanes the relationship is not so clear as one increases thrust one increases the density of the gasses within the cone that the vanes act in. So its not clear if the vector forces are linear with thrust (thrust is a good stand in for weight and a marginally good stand in for rotational inertia) or if they increase or decrease? Looking at the video this looks a lot like earlier problems where the differential gain was too high. I don't really know if I should just lower all the attitude gains or if I should vary the P,D terms differently.

I have not yet reviewed the data, I hope to do so this week.

I also added another wiggles video from onboard.


David said...

Testing is so important and there is lots to learn here.

Observing it would appear at first its a small oscillation that procedures to increase until it is over your flight control limits.

It is a factor of just the added weight or is it a factor of fuel slashing hence baffles would be needed to stop the weight transfer affects?

The positives are that the craft has more than enough power to lift itself so your still in the game. Hopefully your can simulate what is going on so improvements can be made before the next test.

Paul Breed said...

The vehicle has pretty extensive slosh baffles.

David Masten said...

There is one good thing about testing at the airport, the ATC folks we need to notify know what is going on and what to do with the info. Which is mostly - expect a request for clearance and keep that corner of the Class D clear. ;)

C. Scott Ananian said...

I wonder if you can determine the answer to your vane question experimentally? Putting an orthogonal load cell on a motor test platform and waving the vane through its range of motion while varying the thrust should net you lots of Real World Data you could curve fit. That would probably give you accurate information for *your* exhaust and vane geometry, which would surely be better than an analytic guess (especially near the endpoints).

Thomas said...

Presumably, at higher thrust, the jet of rocket exhaust is wider, so more of the length of the vanes is in the jet?

David said...

The baffles are extensive - nice work actually. So this leaves the issue is a control v reaction problem. This could be similar to a road car over correct a car when the rear end gets loose. You over correct and the back slides out the other side in usually ever increasing oscillations until it spins. It a factor of not being able to reduce the correction fast enough once the return the center is about to occur hence it goes over the center line while control is still in the previous position. The reaction speed of the driver / control is not fast enough for the change of position and angle of the vehicle. Would a gimbaled motor be an improvement rather than vanes?