BPS.space
patreon
Throttle Control with Solid Motors?
🕑 Added 2022-01-08 23:55:21 +0000 UTCYes! Well... sorta...
For the last few weeks (and a few weeks in the fall too) I've been trying to get throttle control with these Estes F15 rocket motors that Scout(the landing rocket) uses. That's the last step toward closing in on landing this thing after a billion years of trying.
First I tried messing with something called the Krushnic effect: https://www.nar.org/pdf/TCR1.pdf
This is where you push a sleeve past the exit plane of the nozzle, and as you start to mess with the airflow surrounding the exhaust, you reduce the effective thrust of the motor. Equations surrounding how this works don't really exist, and measured data for it is sparse. It's a chaotic and turbulent process so the relationship between sleeve diameter, length, thrust reduced, etc is going to vary for each different class of motor. In the fall of 2021 I conducted several static fire tests on the stand at fixed sleeve lengths and compared the results (shown in a plot).
In simulations on the full Scout flight profile, I found that to effectively deal with the variation between different F15 motors and and ignition come-up times, I needed at least 20% total throttle range. That would put our nominal throttle at 90% which would give us up or down 10% in range. One of the things I'm trying to do more often is actually *listen* to the data and let that drive design decisions. This is relevant because during testing I was only able to achieve roughly 10% total throttle range.
Could an actuated sleeve increase landing reliability? Sure! Does it close all the edge cases? Nope :( On to the next attempt at throttling!
This one is fairly nutty, but it looked good in simulations, so hear me out. In the absence of tight tolerances in your build, what if you could just turn the rocket motor on and off really fast? This is how kinetic kill vehicles for ICBM interception work: https://youtu.be/KBMU6l6GsdM
So the idea was that we'd quickly actuate a paddle or deflector in front of the nozzle to block all axial exhaust flow and divert out in equal directions to the the side. In theory, this was a cool idea and again, it did work in sims! In practice, I don't have a CNC machine that can cut anything harder than aluminum, and in hot fire tests the rocket motor chewed right through that aluminum blocker. I also had trouble getting tolerances with the 3d printed part tight enough that we didn't divert a majority of the exhaust in one direction, which literally changes the vector of the thrust. Basically, we'd introduce a bunch of side force when the blocker is engaged, which is no good for this application.
Finally, I came across this post on Reddit: https://www.reddit.com/r/rocketry/comments/kvhxfd/solid_rocket_motor_throttle_opinions_if_you_want/
I had actually commented on this a while back, but didn't think much of it at the time. This person used ceramic blockers to variably impinge the exhaust, which gives some degree of control on *how much* throttle you want. So I checked with them to see if they were alright with me trying a version of their design(they were!) then I got to work on my own spin on it.
I used ultra high temp easy-to-machine ceramic from McMaster for the blockers, a servo on either side, and spun up a little variable throttle program to run during firing and IT WORKS. We need a few more firings before flight, but here's what I've found. Exhaust diverts pretty evenly out to the side, so undesired side-forces seem minimal, though it goes everywhere so this setup needs to stick out the aft of the rocket a bit.
Axial thrust seems to correlate pretty well with throttle commands, though we still need more firings to dial in the relationship between the two. Erosion on the ceramic is much lower than I thought! We fired with a hairline fracture that I accidentally created during assembly, and it spread during firing, but other than that I'm surprised how well the material held up. Without that fracture I'd be fine firing it a few more times with these ceramic blockers.
One thing to note is that exhaust from black powder motors like this won't be as hot as exhaust from APCP motors. I really want to make a little Aerotech G8 long burn hopper this year, and I'm skeptical of how well the ceramic will deal with an 18 second burn, but that's a problem for future me.
I figured I'd put all this info in a Patreon post along with some photos since I need a few weeks to get a video about it together. I'm excited to be experimenting with this stuff and it gives me more confidence that maybe we can stick the landing this year? Every time I say that it doesn't happen but maybe *this* time it will???
Blue skies, and happy 2022!
Joe
Comments
Yes of course! https://discord.gg/bgf3nxt3xj
Joe, can you send me the link to the discord, my other account got deleted.
Joe, there a couple of places that you might find some wisdom and interest from very old aircraft information . First the first airplane engines had nothing approaching a carburetor …. Just a pipe that dripped fuel on a plate in the intake of the engine !! They basically operated at one speed with no or little throttle excepting to shut the fuel off which you could realistically do just once !! The answer to being able to land without killing the engine was a “ blip “ switch which was used to momentarily kill the ignition to loose power without ( hopefully ) killing the engine !! The same is true with your attempt at roll control with the fins ! Many of the earliest airplanes pivoted control surfaces ( frequently the rudder ) on a pivot as you did !! The problem is that with increased velocity the input needed to deflect the control surface increases exponentially and if it deflects too far gets ripped clear off ( or broken loose on the control shaft !! The answer then and now is to add a much smaller control surface to the permanently mounted stabilizer !! The needed inputs are much smaller because the amount of deflection needed for control is decreased with velocity !! As C.S. Lewis would say if it helps grab it and if not let it go !! :-) Tony Watkins
Alexander Scheeben
Very interesting approaches. Looking forward to all the details in the video ^^
This is super exciting! Even with inconsistencies of the motors this landing series is one of my favourites! Keep up the great work Joe
This is awesome. Looks like a great direction to get a reliable landing.
This is some really solid scientific work here Joe! Great characterization and looks really promising for augmenting the while landing capability. I bet with some carbon coating you can get the duration up more. The tile material seems to have taken the previous burn really well even with stagnating the flow. Your highest heating is going to be through the gap between the ceramics so in the end you may just be able to adjust your servo rods some after each flight and close the gap. Odds are you can pull 2 to 3 firingings in the current setup. I would also totally steal this and out it on BEPIS
