Liquid Injection Thrust Vector Control (aka Senior Design Project)

Soo....I've finally gotten a chance to actually revise and edit my paper for my senior design project....only 4 months after I handed it in. I got an A anyways, and I only found one or two blaring mistakes ;). But, there are just way too many amazing pictures that tell the story of making this thing beyond simply a report on the design, procedure, and results. So, I'm going to move ahead and get some of these great pictures of "real engineering" (as stated by Prof Maxworthy during our presentation).

I just thought that I'd post the end result first...

So this piece down here was essentially the basis of my project. This is the aft bulkhead and the most difficult part of the project to machine and get just right. The internal structure of this has two steps in it, one for the nozzle to rest in and another for the fuel to rest in. I had never machined anything like this before, so it was an experience and definitely a time to get it right (each disc of alumnium cost $30.00). I managed to get the two sets machined to within 5 thousandths of an inch in diameter and depth, within the tolerances that I needed. More importantly, the nozzle and the grain fit in perfectly. Seen below is a tapped 1/4-20 NPT hole for a NPT to AN fitting for the injection into the nozzle. The holes around the collar are for mounting the motor to the thrust stand.

Here is the fuel grain (plexi) along with the aft and forward bulkheads ready for integration. This was the first time I was going to put this together and I hadn't machined the stainless steel injector, so there is a piece of G-10 in it's place (bulkhead to the right). The flats are there because it was much easier to tap nice orthogonal holes into these bulkheads with a flat to zero against rather than attempting to use a rotary table. And it just looks slick. But again, something I never did was to actually mill something. It's a lot of fun to mill aluminum 5 mils at a time.

So here is everything, at least all of the motor hardware, when it was all assembeled. My real goal was to be able to get some movement forward or backwards and side to side. That was a completely unrealistic goal, but it made for an amazing looking thrust stand (in my opinion). I was really excited to have this part of the project down and looking like something....but then again, the hardest part of the plumbing was coming.....plumbing.

Luckily, David and Sarah had already bought most of the things that I needed to plumb my system. At the beginning of the semester I was all excited to put this sweet looking automated test set up together that would control the oxygen flow and the ignition system and the valves and the data acquisition. I had this whole thing planned out in my head how it would work and it was going to be great! Time was my enemy. Or maybe my eyes were too big for my stomach, well, I know that my eyes were too big for my stomach, but not too big. I managed to get all the pieces together from what was available in the lab and two trips to Norton Sales (coolest rocketry place on the face of the Earth) where I nabbed myself a little Marotta Valve for $300 and a few other pieces. This was the whole plan. I'd have LabView start an igniter to burn (I hadn't even thought hard about that part) then open the first valve (left side, picture below) to get the motor to fire. The second valve (right) would be connected to the injector point, where I'd inject oxygen gas (hoping that it'd further combust and really change the direction of the flow). I could control all of the timing with LabView, and I would use two load cells to measure the axial and side force of the motor so I could do all sorts of neat analysis.

The Sunday after Thanksgiving (project due in 3 weeks) was the first motor test fire, and only because me and Josh, one of my partners, really pushed and got everything ready. We went thought a cleaning procedure with an organic solvent to get any contaminants out of the plumbing, then set up the motor and got everything ready with the exception of an igniter. I ran all over the lab trying to think up something that would work and would burn (who thought that creating fire was difficult). I thought a small piece of solid propellant would work (we had so many scraps that it was easy to find) and then thought that using it could be too much for my motor. I finally ended up finding some Estes motor igniters. I took a small piece from a paper towel, doused it in WD-40, wrapped it around the igniter, hooked some leads to the igniter, and put it in the motor. This is the firing that you see below. In the video you hear Josh say, "I see smoke." I was so focused on the motor (I was controlling the ball valve for the oxygen) I ask him what he said. I turned it on before I finished talking and I was completely surprised. All I thought afterwards was "shit, when things go bad they really go bad fast, and that could have gone really bad way too fast".

So...the plan was that I would start the pressure off at something like 20 or 30 psi, and then increase the pressure on the regulator until I choke the flow. Well, I saw the smoke and opened up the ball valve and the motor just went, and it screamed. All I can think of as I'm watching this thing is "please don't blow up, please don't blow up" and I'm seeing this big bubble near the injector get bigger and bigger. That 7 seconds of burning felt more like 20 and I just turned it all off because I didn't know what to expect at all. Great learning experience and I was stoked after this fired. I actually made a rocket motor, from scratch, it ran for a whole 7 seconds., and I still had all 10 fingers. Just watch the video below and you can hear my excitement.

After those seven seconds of sheer terror, we got everything disconnected to check out the damage. Amazingly, everything looked great. We realized that the throat for the nozzle was too large to support the flow rate that we could supply, or that may have simply been the pressure that we were running at. There was an odd place in the fuel where near the injector that burned much faster than at other points in the injector, and we deposited a good layer of carbon on the injector face. Other than that, the motor performed amazingly well and worked.

The next day I ran over to the fabrication shop to use the band saw there and see what the cross section actually looked like near the injector. You can see that there was increased burning at all of the injeciton points, but more on the lower left hand corner. When I was making the injector, I broke a bit on the last hole and really didn't want to go through the process of machining such a small piece on our lathe and re-drilling everything again. I flipped the injector over and took a countersink as close to the injector hole on the other side and tapped the broken bit to get it out. It succeeded, however the small amount of non-comformity at that injection hole caused the bubble, which was evident during all of the tests.

But now, we had a working motor and knew how it ran, kind of. The nozzle was redesigned to ensure that we attained a choked flow at the throat (now 0.125" rather than 0.25") and a much smaller angle at the nozzle exit (3 deg rather than 17 deg). This produced a very nice straight flow from the nozzle, and it looked awesome (first picture in the post). Getting back to the testing though, the whole purpose of this project was to try to test liquid injection thrust vector control. Well, from my entire grand plan at the beginning of the process, the only thing I had managed to do was fabricate the motor and put together the minimal plumbing necessary to fire the motor. There was no load cell set up, no data acquisition system put together in LabView and the plumbing for using oxygen for the LITVC injector wasn't cleaned. We bit the bullet and said "hey, we can film some results, have amazing movies, and just use nitrogen gas for the injection fluid not oxygen." It simplified the entire thing and gave us the opportunity to finish the project. That was Tuesday night, the project was now due Friday at 5:00 PM. Soo....as you can see below is our set up. We borrowed a nice non-flamable metal table to mount our motor on. The box there was a water tank for doing fluids experiments, but we decided having two pieces of 0.5" plexi between us and the motor was better than nothing. In the background on the motor side is the K-bottle of oxygen and in the foreground to the right is the nitrogen bottle.

This is how the motor was integrated into all of this. In all honesty it was really simple from a plumbing and design standpoint.
And here are the three videos that gave my group and me A's.

Those three test fires were all from that Tuesday night before the project was due. There was about an hour recycle time to get everything disassembled, cleaned, check out the video, get everything prepped for firing again and firing. The biggest pain in the ass was the igniter. For what ever reason we would get it put together, test the continuity and everything would check out. We'd get it in the motor and have everything ready to go, short out the igniter and it wouldn't work. We took it all apart again, put it back together and got it back in the motor and voila, it'd work. Never worked the first time but always worked the second time. So, we got our motor burning with a head end pressure of 100 psi, hopefully a motor pressure of ~50 psi, and let it burn as long as we felt comfortable. Originally this was going to be 18 seconds. The first test fire was 7 seconds. Those three motor tests ran for about 30 seconds plus/minus 2 seconds. Having being the one operating the oxygen....those thirty seconds were actually about 3 minutes, maybe 10 in my mind. I don't know if you can hear me screaming in those videos asking for a time count, but the motor was so loud you couldn't hear anything. I was amazed at the scream a little 1/8" diameter hole could make, and it was awesome. I was more surprised that we didn't get the attention of the campus cops for making that much noise at 11 PM, 12 AM and 1 AM, cause it was loud.

So, all of our data was taken (those three videos) and managed to say from those (the last video around 28 seconds after ignition) that we manipulated the flame, but didn't necessarily produced a sustained thrust vector and definitely didn't have complete control authority over that thrust vectoring. All in all my group managed to write the papers from Thursday night at 6 PM to Friday morning at 6 AM, got those sent off to Kinkos to be printed and bound, burned a CD with the videos and handed it all in. I think I slept 15 or 16 hours that week, but got everything finished. It was one of the coolest projects that I've worked on so far, but I'd like to make this my baseline for my future projects.