Rocket Silo Launch - DIAMOND-X Flight Test 2

Jo up at 5 4 3 2 [Music] 1 hi everybody this video is about the second test flight of my actively controlled model rocket diamondx its first test flight is very successful and I have a dedicated video about it on my channel I'd recommend watching that one first if you're new to this project in this video we're going to talk about some modifications that have been made to the rocket and then watch the footage of its second test flight just to recap Diamond X is about 1 M long and about 1 kilo in weight and it has four actively controlled AF fins the the rocket is now in what I'm calling the block two configuration so let's take a look at the changes that have been made since the first test flight so here we have Diamond X as it's getting prepared for its second test flight after about here there are very little changes this new Raceway has been installed you can see that there's a connector there for a booster I'm just checking this out to make sure that it doesn't affect the vehicle aerodynamically it's not not really on there very straight but it's where the whole lined it up as in the nose there's been a little modification I'm sure you guys can tell we have a camera in the nose now so let's take this piece off this is the new Arrow shroud and the arrow shroud had to be modified with a camera Dome on it so flight One Flew with this front Arrow shroud with a pointy tip rounded tip now let's take this guy off and see what has changed under the hood so this is the avionics assembly of diamond X I didn't really show this off very much in the intro video so I'll go into a little bit of detail as to what's going on up here let's start off on this side here you see lots of cabling going on the first thing to note are these two circuit boards right here so this circuit board is the primary flight computer and this one back here is the power distribution system the primary flight computer has a massive IC right in the middle there this big chip in the middle is the actual flight computer itself the microcontroller and then around the periphery here we have lots of different sensors so inertial measurement units and uh barometric pressure sensors magnetometers on this side we have our micro SD card this holds our flight data and the USB port we use to program the system on this side here's a ribbon cable that goes from the primary flight computer to the power distribution system and finally we have Servo outputs and other power outputs um that go down the Raceway towards the tailfins the changes that have been made here's the old battery Mount you can see that we've just modified it with a camera and the camera control board these are just run cam these are split HDs um new power connector here to take data to it hello Mr Momo now I want to uh very explicitly cover one thing really quickly this is not a Seeker this is just a camera that looks forward and Records the data so there is no passing of data between this camera system and the flight computer you can see there are two wires that connect the camera to the flight computer there these ones 5 volts is red and black is ground so the the primary flight computer has no way of seeing any of the data that this camera records it just goes to this little micro SD card on board so it may look like this thing has a Seeker now because there's a camera it absolutely does not and you probably shouldn't develop such a system my country that's illegal and it's probably illegal in yours so when it comes ready to fly this thing everything will already be set up we'll already have parachutes packed um pyrochar installed not run down uh inside the vehicle everything will be all set except for plugging in these xt30 connectors right here that connects the main battery to the flight computer and then the last thing we have to do is slide this Arrow cover on and then all the setup including calibrating the inertial measurement system uh turning cameras on servos everything like that is all controlled via the link so I can control that from the ground station which is super helpful So in theory this is a very quick setup system I just plug the battery in put this guy in place four screws and ready to go in practice typically some part of that process goes wrong and I have to take the whole thing apart but we'll try tomorrow and see if this guy is easy to put together that was one of its primary design goals is to make it not super annoying to fly so let's talk about flight test 2's guidance system this time we're executing a pitch prog program so at liftoff we're going to try to point straight up at 3 seconds we're going to pitch down to 20° below the Horizon and then at 5 Seconds Point straight back up and hopefully we're traveling vertical when our parachute deploys the plot in the top right is a matb simulation of this and we'll compare this to flight data after the flight we're also going to be launching from the atlas Silo this is a hot launch Silo that I talked about more in the first video on diamond X Hot launch means that the main rocket motor is ignited inside of The Silo and in this case the rocket is held in place by pairs of aluminum rails sandwiching f a flame diverter is located at the base of the silo made of ceramic tiles and this redirects the Rockets hot exhaust gases up a separate channel so we should see our hot exhaust exiting out of the right Channel and then our rocket out of the left so let's watch the flight footage going up in five 4 3 2 1 nice oh that was gorgeous need a little bit longer [Applause] [Music] rail [Music] [Music] is no the flight went great we got a good parachute deployment there's the parachute deployment charge and it did go off only damage was one broken fin but that honestly might not be a bad thing CU it's better the fin brakes than the really expensive Fin and motor mount and the only other piece that was damaged was a little crack on this Raceway here not quite sure why this one broke I imagine either one of the parachute lines or something grabbed the end of that screw and tried to tear it out the front of the rocket as the parachute deployed cuz these screws they don't like end in the paper they actually go through a little bit so it's kind of a snag Hazard maybe not a great design but we'll pull the flight video off and pull the flight data off and see how it looked okay so I've uploaded the flight data here to curve and I wanted to walk you through some of the plots that I found interesting so the first one here is the pitch of the rocket in green and the pitch set point in Orange so we can see here that there at the beginning this is as the rocket leaves the rail the pitch is pretty unstable this is when the rocket is rolling uncontrollably and our control system hasn't turned on yet so it turns on about here and then we can see we quickly arrest that rolling and then start to control our pitch and we start to bring our pitch steadily down down down towards our set point of 0 degrees 0° in this case being vertical just based on the coordinate system that I use at 3 seconds in a flight that set point changes to 20° and our rocket starts to bring the pitch back up to 20° this is pulling the nose below vertical to 20° and we get to about 18° before our set point changes again at 5 seconds in the flight at 5 Seconds we rapidly bring that pitch back down and we're slowly bringing it back down to zero and we get to about one degree here and our parachutes to play right about here so in terms of pitch our rocket was pretty well controlled this next plot shows pitch and Y this is the same profile in Orange as we saw in green above and the green line here is our yaw and the interesting thing here is that our yaw should be going to zero and it's pretty close but it isn't quite at zero we have this kind of bias at about5 or6 degrees for the duration of the flight which is pretty interesting I'm not quite sure why this happened could be an aerodynamic thing or control thing I'm not quite sure but I'll look into that this next plot shows our fin angles so you can see at the beginning here our fins are set to 0° which is just pointing straight up wherever I calibrated them and then right here is where the control system enables and that corresponds to chaotic motion here there's no fin motion and then our system starts to control itself and then our fins start to move around a bunch you can see right here our fins kind of go nuts there's lots of oscillatory behavior here that corresponds to the second Pitch set point change and I'm not quite sure why this happened this is another thing that I'm going to look into but we induced a lot of oscillation on our fins so I was like okay which axis is causing these fins to oscillate and most of it is occurring because of roll so this is the roll AIS command so our fins uh receive a Commands to control pitch and Yacht and roll and we can see each of these separately so we can see that in rule our fins are starting to kind of go nuts right at this time the other thing to note here is this section we see that our fins are commanding a roll of about .8 or1 de so this is the roll bias detection so I can never get the fins quite straight when I calibrate them on the ground so the rocket has to detect the roll offset that's how you know how far off I've calibrated the fins and then has to correct for it and you can see in this case I was about 8° off in calibrating the fins which isn't which isn't bad it detected it pretty quickly but um next time I can calibrate the fins a little bit you know further towards that side and then in theory uh I can reduce this offset I could also look at the pitch and yaw commands data we can see here the yaw deflection in Orange it looks like we're trying to correct for that yaw bias but it it isn't really doing a whole lot and then in Pitch here we are trying harder and harder and harder to to correct for pitch and this corresponds to a greater difference between our pitch and its set point compared to our yaw and its set point now this one is interesting so this is a torque based controller and we can compare the pitch commands torque in green to the acceleration we see on the pitch axis this is getting kind of nuanced but if we want to do acceleration control I need to be able to tell the rocket hey I'd like you to pull one g on the pitch axis and then have it achieve that by moving the fins a certain amount and this data here I haven't actually you know pared through it at all but this data will be really useful in trying to correlate a certain amount of fin deflection to a certain amount of acceleration generated on the rocket so this will be useful in analysis later but I did collect this data and so it'll be helpful later on lastly here we have a roll angle a roll set point for this entire flight was 0 de if you watched the flight test one video you saw that we had a uh roll program so we rolled to the left 45° and then back to the right this rocket it was supposed to go to zero and we can see that we had a little bit of a pitch bias or sorry correction a roll bias um as soon as the control system enabled but we steadily brought that back down to zero and then this oscillation here again corresponds to that second Pitch set point change so there's some problem in the control system that's causing the vehicle to roll whenever I try to change its pitch this could be a number of things still working through it but we can see that oscillatory behavior en roll that we saw on the fins there so overall collected a bunch of great data here particularly this data will help me um work through the acceleration control problems I've been having in simulations um but in theory this Rocket's only going to get better with time as we collect more data and we refine the simulation models so that I can do a bunch of test flights on the ground and then go out and fly it and hopefully the two match up pretty well so now let's actually compare the mat lab model that I have to our actual flight data so the plot on on top is the mat lab model where I started at about a 50° pitch angle kind of like um the rocket was at when its control system enabled and then the plot on bottom is that same pitch profile data from the actual flight itself so if my mat lab model is 100% accurate then these two should basically line up one for one the things that I'm looking at are one the slope of the lines when it's trying to change its pitch and then two the behavior of the rocket after these step responses so after the set point changes and overall these are pretty good the slopes are almost identical and we can see basically the same behavior of the rocket there's an underdamped response as we change that pitch set point they're not identical though so there's a couple changes that I need to make to my mat lab model to have it more accurately represent the flight data that I have so the first thing that I'm noticing is that the mat lab model has more oscillations in Pitch than the actual flight data so this is showing that the rocket itself has a higher damping ratio than the model would suggest uh right now my model doesn't actually give the rocket any natural damp and it should have a little bit so I'll add that and then try to kind of pick some coefficients so that the model lines up with the flight data more accurately the second thing that I'm noticing is after the second Pitch set point change the rocket changes its attitude much quicker so that initial pull-up after we change our set point back down to zero gets us pretty close to zero in flight but it only gets us to about 8° in the uh the simulation so there's something else that's wrong it could be the mass moment of inertia is incorrect or it could be that the rocket in Flight is actually less stable than the model would suggest and there's a couple reasons I think that might be so I'll continue to tweak this model in every single flight I will compare the model data to the flight data and so hopefully I can get a really really accurate simulation and if I have an accurate simulation then it's really easy to kind of test the rocket on the ground and I don't have to spend time and money burning Rocket Motors to go test the rocket so what's coming up next as I talked about in the first diamond de video I've been working on a series of boosters that can fly with this rocket the first one with a fixed 38mm motor is is pretty close to being done I'm a little concerned to fly it because this rocket doesn't have a dual deploy system and even on this flight you know on stage flight it landed pretty far from the launch site so I'm a little concerned to fly it with a booster where it flies even higher and then it lands even further away so maybe I just limit those flights to low winds days or I work on some other system of having the rocket fly up wind not quite sure yet but flying it with the booster would give me more flight time so I could collect better and more accurate flight data the second thing that I'm doing is I'm still working on that acceleration control piece that I talked about when we looked at the curve data acceleration control is very difficult to test on the ground because it relies on feedback from the rocket itself so I can't you know pick the rocket up and kind of simulate the accelerations so I need to work on a system where inside of the flight software when we're simulating on the ground the rocket kind of computes the accelerations it should be seeing and then you know applies those after the fact and that way the rocket can kind of internally test itself so that's know Hardware in the loop simulation Now The Silo itself has definitely seen better days as you can see in this clip the pressure that built up inside of the main Silo was so high that it burst the inner wall of The Silo and then that wall hit the outer wall of the flame trench and broke both of them and they were only held in place cuz there were some screws holding them in at the top so that is one of the reasons we didn't see a ton of exhaust gases coming out of the flame trench because it just blew out the side of the silo that probably helped the rocket a little bit but it didn't look as cool um and that also means that the silo in this case is basically unflyable at this point I would need to rebuild and frankly redesign parts of it to strengthen it against the pressures that built up um but for the meantime I'm just going to fly this thing off of a rail it's way easier and I can still collect uh all of the flight data that I need doing that in terms of video the next one that's going to release after this will be a diamond X build video so I've recorded a bunch of footage of me building a second's diamond X and upgrading it to a block three standard where it can fly with a booster I've got a ton of great footage there so I'm hoping you guys will enjoy that I also talked about a shock Deep dive video that I've been working on U before I finished that one up I wanted to fly shock on a sixth test flight and I did and it didn't go very well so the Rocket's dead so I'm going to incorporate some of that footage into the video and then release that um coming up a little bit later and instead of being kind of the culmination of a successful test program I'll just talk about the lessons that I learned in that and hopefully you guys don't have to relearn all those lessons yourself if you want to get into actively controlled model rockets well that's all I have for now thank you everybody who's been interested in the project and followed along so far got lots of interesting stuff coming up so stay tuned for [Music] more

Channel: Lafayette Systems

Views: 143,340

Keywords: rocket, control, rocketry, guidance, controls, navigation, fins, silo, hot launch, tube, flight

Id: UeJZToWH-C4

Length: 18min 23sec (1103 seconds)

Published: Wed Dec 27 2023