What happened to your RCS thruster system I think I saw you show off in an Everyday Astronaut video? Could that theoretically work in a situation like this?
As someone who just finished their mech eng final project that (to my own detriment) involved a truck load of ANSYS simulations, I don't envy the CFD that you ended up having to do 😂. I'll look forward to the next video with the details.
I know playing with terminal guidance is a taboo of sorts. But is it actually illegal? And how come there is so little on that topic available online outside of the US?
@@MoritzvonSchweinitzit's not illegal per se, but doing it without license is problematic and posting about it can be illegal. But that's US. It is illegal in many other countries.
“And as always, they were mounted with about thirty pounds of hot glue” there is something profoundly wrong with this man and that is why i am subscribed to his channel
Man, I'm an automation and Control engineer in Brazil, I have a masters degree in robótica and I have been teaching engineering and mechatronics for 8 years. I have no idea how on earth you deal with so many complex details in these projects. It seems like it is done for a team of 10 senior engineers. You are crazy! How do you tune the controllers before the launch? Do you have a detailed mathematical model for the whole structure? You are insane man, keep up with the work. You're going to heaven.
It's fun seeing you go through the same problems I encountered about 15 yrs ago. I built a rocket with canard driven by high-powered rc servos. I put together a controller with an $1100 Analog Devices IMU and a custom microcontroller with SD card parameter control and data recording. I initially wanted to control roll. had 9 flights without success. At first I thought I had a control system bug, then I thought I might have a hardware problem. Finally I mechanically fixed the canard at a constant angle. I had roll reversal in every flight. I finally installed a camera to verify the data collection. Unfortunately the rocket came in ballistic and the 8 foot rocket was compressed into 8" long wad. I researched this problem and ran across a group in the NW that was having the same problem and had identified AIAA research that identified the problem many years ago.
My dad helped develop a few rockets for General Dynamics, they used something called “dithering” which made the fins vibrate(?) and when they wanted to steer in a particular direction they just made the fins spend more time on one side by interrupting the occultations, favoring the direction they wanted to go
Yah pretty sure Salvatore pais scaled this technique up to make the tic tac vehicles.. wingless UFO can only be explained via vacuum creation via high frequency vibration. Likely powered by compact fusion reactor.
Which is not a very new concept as we used similar methods for early R/C aircraft control before we came into the wonderful world of Proportional servo control, in the late 60's.😁
Nice ... use of random control noise, and then removing some randomness to influence control. Makes me wonder if the "dithering" was a technique to disrupt shockwaves traveling across the control surface.
When you started in on canards, I started muttering "elevons". When you started talking about talking about rotation the whole fin, I nearly yelled "elevons" at the TV as well as grumbling in 737 Max overcontrol. Then when you showed your solution, I was like, DUH, elevons. Yes, I know the shuttle used a separate rudder, but honestly you should be getting the point since you had four "wings" instead of two. This is how supersonic aircraft control their flight. And if you see my name and picture, and recognize them, LONG LIVE THE PID. Glad I found your channel, we did know each other. And to really induce swearing you need to botch a Kalman filter.
Joey B out here controlling an aircraft over a WIDEEEEE envelope on his f̶i̶r̶s̶t̶ second try, and I hope ya'll know just how hard that is! Aircraft control derivatives scale with dynamic pressure (combo of airspeed and air density)... This basically means an aircraft behaves very differently at different speeds/altitudes with the same control deflection. The successful flight is a testament to Joe's math all collapsing down and "un-coupling" the rocket from its environment, from aero modeling to control system design to sensing. Next video is gonna be good! TL;DR: 😍🚀😍
Well no, a missile traditionally carries a payload of explosive compositions. A missile would be a destructive device. He simply engineers delivery vehicles. @@CIWS-Goalkeeper
@@OnlyFacts11b welp, have you heard of a concept of kinetic energy? I mean an explosive would be great but if you add enough speed it would work. Also rocket propellant is not so much different from explosives, tweak the burn rate and you get something that goes kaboom
The fins extra thickness didn't seem to be an issue, but if you do want to reduce there thickness in the future one option is to mount the servos inside the fuselage and have a thin shaft go up into the fin to control the tab. This could also allow almost all the cables to be inside the fuselage as well, further reducing drag.
The amount you talk about how much you love making videos about this stuff is so refreshing. So many TH-camrs get so caught up in creating content that they start to hate it, and while that isn't always controllable, I'm sure that as you grow, the donations and money from TH-cam will make it possible to have access to better parts and more prototypes. Keep up the grind.
The simulations at the end are outstanding and it demonstrates how your channel not only inspires others to take on engineering projects, but perhaps even more importantly the channel builds a community of very talented aerospace engineers who come together and make meaningful contributions. Great job! Kudos to Gabriel, Matt, Andrew and Dan.
I always love the sheer attention to every detail Joe puts into everything. Everything from the rocket's functionality, the camera shots, and the video edits are all phenomenal quality. Keep up the good work.
As a computer scientist, I would be interested if you could make a wind tunnel to gather enough data on the effect of the vortices and build a controller to compensate for them. No idea how hard it is to build a wind tunnel though, and I'm sure I'm underestimating the difficulty of that control problem. Good luck! Great video as always
Actually, building a supersonic wind tunnel in itself is not that difficult. Gathering the relevant amount of currency to do so, while building it in a place that is about 500 miles from your nearest neighbour........ different story.
As someone that studies Engineering Informatics (mixture of computer science and electrical engineering) I often have to work with computer science that had never heard any physics or mechanical engineering lectures and they are most times better at algorithmic and similar thinking stuff but don't let have mechanical or electronics ideas. Wind tunnel up to a couple of hundred km/h no problem use a fan big and fast enough and you are good to go. You still need to think about how to reduce the effect of the turbulance of the fan, maybe build a pull configuration the expected wind speed for the same input power is slightly slower but should still work. But now the trans and supersonic Problems. These speeds are not possible with a fan alone. And the pull configuration does nothing. If your fan is to powerfull you suck a vacuum but still don't get mach 1 and in the push configuration you still would only near to mach 1 and than produce a high preassure zone. So you need a nozzle the same type as the one on the bottom of a rocket engine. It solves the Problem of converting high pressure gas to high speed gas. BUT the gas goes from very high pressure (tens or hundreds bar (atmospheres)) to 1 bar or below every expanding gas cools down. And this kind of depressurization would lead at least to freezing water and CO2, but depending on the pressure drop even to liquid nitrogen or oxygen (unlikly in home build variants). So you don't need to start with a high pressure gas. You need to start with a high pressure high temperature gas. Problem is most materials get weaker under high temperature but the high pressure is really high so the whole thing is near its failure before you even have moving gases. Second Problem you don't want to see effects of pressure waves reflecting of the tunnel walls because that would cause similiar uncontrolable feedback loops as the canards to the fins earlier in the video. So the wind tunnel should be at least factor 5 to 10 bigger to the test object (model rocket) now we test a small model with a fin can that is only 5 cm in with (to be honest to tiny to have good estimates on how a big one would react). The wind tunnel has at least to have a diameter of 25 cm. 0.25m*0.25m*pi/4=pi/64 m² this times the desired speed of 300 m/s (yes I know speed of sound is faster under normal pressure and temperature but I will not estimate which temperature we could manage to reach in our tunnel) at 0.5 bar is pi/64 m²*300 m/s*0.5=7.36 m³/s. So you would need every second of testing 7.4 kubic meters of air delivered to your wind tunnel, forget it. This is for each second 4 full (and not the smallest ones) air bottles that divers use. Because there is no for not that rich individuals available solution that would produce this preassure and quantity of compressed gas on the fly (exeption rocket motors but they burn to hot to hold a normal test object in exhaust).
I'm working on aerodynamics of rockets myself. From what I've heard from a people who works directly with that kind of aerodynamic tunnels, just the cost of preforming a single set of tests on subsonic tunnel capable of fitting rocket this size is at minimum few thousand dollars. Not to mention building a supersonic tunnel. You can get good results just by simulations alone tho, I basicly work with that, Solidworks is a pretty good tool for obtaining such data and is extreamly easy to use. However I would questions it's results for trans sonic flows.
I was sitting here waiting for you to mention the rollerons on the sidewinder cause its like, one of my absolute favorite engineering solutions ever. So elegantly simple and yet works incredibly well Edit: im not an engineer, just a mechanic, but i think a combination of servo tabs and entirely moving fins is the way to go. This is how the trim system works on just about all modern jet aircraft. Little tab moves in opposite direction of where you want the surface to go, and generates a moment on the major surface. Allows for precise correction without having to put an enormous load on whatever you're gonna hinge the tab with. Much easier to design a fin hinge that can withstand that sort of load than a tab hinge that can. Just my $0.02 though, supersonic aerodynamics is super weird, and aviation and rocketry are only marginally similar
21:15 Aside from the first couples seconds or so of this, you could've absolutely fooled me into thinking this was on-board flight footage. Well done to the community for helping here. Also would love a video on transsonic flight.
Before I finish watching the video; there is a neat trick that the first AIM-9 Sidewinder missile used to avoid needing to compensate for things like density altitude: it used force servos on the fins rather than position servos. The result is that, with a few assumptions about the center of force on the fin staying more or less the same distance behind the pivot, a given turning force on the fin created the same side force on the missile body more or less regardless of speed or altitude. The entire guidance and control system on the first marks of the AIM-9 are a fascinating cases study in how a spec project with almost no budget can force better designs.
For sure please do a video on control over transonic flight regimes! And I'd love to see one (maybe combine them) about CFD you mentioned, what specifically was the effort required to succeed, the process, and the modeling. We love the theory
transonic is really hard (it's why jetliners which have the power to exceed about 0.85 mach just don't). Best bet here is just to get through to the other side as fast as possible.
Only found this channel a month or so ago. Have been binging it basically non stop. Somehow I'd skipped over this video and the amunt I laughed at 15:14. The best BPS foreshadowing of all time.
Thru the whole length of the vid, one thing kept popping into mind was this: For control on the X-Y axis, a single, tapered spike on the nose, able to be pointed off-axis for 'steering'. This would eliminate downstream vortices from canards. Your tiny tabs on the back of the fins would take care of roll.
I've been watching your last 3-5 years of videos over the last month or two, this is pretty much the solution I've been yelling at the screen through all your roll control adventures. looking forward to the rest of the story.
Good show. There's also inversion layers reflecting sound and sound reflecting upwards then reflecting back down at a further distance having a bland detection somewhere between. Artillery sometimes does that.
Transonic video needs to happen for sure. Also great job on figuring this one out, I know it was a big step and took a lot of hard work. Every day you are getting closer to space and I love to see it!
The care you put into your these videos is amazing. Like in 13:07. If you are a professional video editor like me, you know you've had to shot multiple takes saying various minutes (maybe 20+ of them) and then choose the one that matched in post!
Im 28 years old, former police Sgt., college grad, and father of 3....I look up this man. What an inspiration you are. Weather permitting, I am about 1 week from getting my PPL as I now pursue aviation full time! Hope to meet you some day!
There is a reason for all moving fins other than extreme manouvrability. At the end of WW2 aircraft became uncontrollable due to the elevator being behind the shockwave (where it looses effectiveness) just when they were very much needed because of the onset of Mach tuck (which is a nose-down moment on the lifting surface in transonic flow). The problem of controllablility was solved on the Bell X-1 with an all moving horizontal stabilizer. On supersonic missiles this is the all moving fin. Your trim tab seems to function more as a spoiler (so to create drag) than as an aerodynamic surface to direct the airflow. It definitely gets credits for a shoestring approach. Hope you don't mind me making a suggestion: spoilers don'' t necessarily have to be in the fin section, they might as well be in the aera where you would expect the canard fins or anywhere else on the vehicle. The obvious advantage of moving the spoilers away from the fins is the available space for the servo's. Now your fins can be thin again to reduce the drag. I hope this suggestion will help you past Mach 2 and eventually into space in a controlled manner.
I'm 50, forced to retire last year. been doing some embedded system engineering works for more than 2 decades. Watching your videos, I have to say they made the right decision and that involuntary retirement was the best thing that has ever happened to this man. Keep up a good job as you have been! Good Luck too.
Don't feel too bad, Lockheed made a very similar connector mistake on the THAAD program. Missile went about 100 yard before burying itself into the desert. The connector was to load the guidance program, but when the wires pulled out they shorted the guidance system. So, no guidance. D'oh!
Silo launched rockets, guiding eggs to specific places on the ground, supersonic aerodynamic control, and soon to be a space shot? You are definitely on a watchlist somewhere.
10:36 that reminds me of the stabilizer-stabilator thing with airplanes. When you have a stabilator the stabilizer is the elevator, so basically the stabilizer itself rotates the change the altitude. When you have a stabilizer, there is a separate elevator that moves on the back of the stabilizer. This is like the stabilizer.
As an Aerospace Control Engineer, this is absolutely amazing! Yes we would actually love to see more videos on control architecture and if you had implemented gain scheduling for pre and post mach flight!! Thanks a lot these video! All the best for future control missions
Outstanding video production Joe! Very interesting discussion on the control issues and the importance of doing good post flight analysis of what worked and what didn't.
I think you bumped into the problem they had breaking the sound barrier with the oldschool split control surfaces. You've gone smart and made a teeny tiny control surface, but the split control surface becomes basically useless in the transonic region, which is why they moved to all moving control surfaces. You've probably already come to this conclusion, but if you put a teeny all moving contol surface at the tip of your fin, you'll retain control authority at speed. Having the surfaces that far out will make them more potent in roll as they have a longer lever, but hey, make smaller movements. Keep it up, it's an instant click whenever I see one of your videos!
You could alternatively literally make ‘ailerons’ (on aircraft wings/tails) on the bottom side of the fins if you need something a touch more commanding. Then the servo is still housed in the fins, you can use simple linkages for mechanical leveraging to use small servos with ‘big force for small movement’ and it’s simple to model, print and attach.
In the past, only a couple of objects could reach supersonic, the goal being INCREIDIBLY difficult, challenging, and made by tens of engineers, but now, even a random guy with a youtube channel can do so. How humanity advances so quickly is just too beautiful
Maybe you need a simple analysis tool that checks all the connections are in place. ReadyToLaunch and return a green or red. Checks the values are where they should be. As you go higher with more cost then you might be glad you have this simple analyser. Good video. Keep going, you'll get there.
Is it worth simulating the glue and bumpy lumpy surface? Perhaps someone can make you a camera mount and design a groove in your rocket tube to fit the cable too. Total guess that imperfect surface might be affected by fluid dynamics at supersonic speed and above. A strong fluttering pocket of fluid could possibly move the rocker or damage it. Maybe even make it's fuel usage above calculation.
Supersonic flow is fascinating. I have been working on my own rocket to test control methods in a compressed flow flight envelope and can say the work you are doing is really amazing. I can't wait for the next video!
Dude, I super love nerding out with you. I can pretty confidently say that every video you've already put out and all the future ones that you haven't yet made. Just do what you love. I'm pretty sure we'll all like it.
Undergrad aerospace engi here. Glad to see how feedback control of dynamic systems and aerodynamics courses may be applied within a small scale in a workshop! I Hope I'd be able to construct any advanced sounding rocked before graduation inspired with your channel!
Your channel has a some of the most well made videos on this whole site! Thank you. I’d love to see a video about why it’s particularly hard to manage flight at that subsonic range of Mach 0.8 - 1.2.
I perceive through my experience of supersonic/near supersonic flight control (16 years around F-111 Avionics, :-)), your early on hunting may have been an effect of the boundary layer flow (BLF) across the fins going through changes due to the sonic transitioning shock wave generation. Once supersonic flow was stable, the roll surface could 'play' in a 'quiet' space. We had multiple static pressure sensors and localized pitots around our jet airframe that would measure Boundary layer pressure to, then help, fine tune surfaces such as flight controls but also engine inlet shaping to keep inlet air subsonic. I would suggest looking into the design changes done, over the early years, to both aileron placement along the trailing edges of wings to eliminate/minimize aileron blanking and things like spoiler tabs, etc. Also, early supersonic aircraft went from horizontal stabs w/elevator tabs, but had to switch to all moving tail planes for elevator control, as the elevators could also blank, but some of that effect had to do with the REALLY turbulent stuff coming of those wings (and their shock waves in front of the always rear mounted elevators/elevons. Definitely a lot more of a dilemma than just pointing the surface/tab where you want to 'steer'. Excellent work sir!!! 😁
You know you're one of the most brilliant people I have seen on TH-cam in this area and you're doing this on your own with help from Friends of course your education and your commitment consistency drive and absolute refusal to quit. If there's anyone that people would give money to to support the humanity's Space Program where the data comes to humanity not filtered through some space agency!
I think this is the best video you've ever made, keep upping that production values! This is amazing! Also, yes please, we all want to see the video about transonic flight regimes.
Way cool! Maybe this is in the next script already, but when you want to add in pitch and yaw, will the tabs needs to be substantively larger, and will that put a greater strain on your control software (and maybe your servos) to not overcrank the roll correction?
great job on making that rocket! this is so awesome!! honestly that CAD model of the data from the flight is pretty cool, ive not seen POV like that before
I would watch a video titled "10 sexy tops to keep it going steady (how to prevent your missle from rolling)." Joking aside, the production value of these videos just keeps climbing- love it!
My first thought on seeing the secret alternative control option (10:45) was of all the x-planes failing to break the sound barrier prior to the Bell X-1. They all had hinged control surfaces, where-as the X-1 has an all-flying stabilizer. This because during transition from transsonic to supersonic, the shock wave made the rear hinged part of the control surface unpredictable and ineffective. Great coverage of why this an issue (form 19:44), and the importance of control-surface sizing (10:59). The X-1's actuators become over powered by supersonic forces, but trim control saved the day for maintaining stabilized flight. But that's a different part of the "Right Stuff" story. Do you have a target velocity range for "stage sep" on the "space shot"? This seems to be a critical design concern to avoid the transonic region. Thinking this over gave me a 💡🤯 awe moment when recalling time in flight most strap on boosters separate on commercial launches. BTW: So-called "Naked" camera's are a common with FPV drones (whoops), and can be purchased disassembled. One major drawback with naked cameras is lack of heat sinks, so can prematurely shut down. Could be avoided if not powered up til critical time in the launch sequence. (if this turns out to have been the issue)
You are right, y’know… this _is_ huge! I am impressed, again, and always! As for “let me know if you would like….” yes, how many times, we love your videos, make! :)
Great news Joe, about the test flights. So amazing to view the simulated flight using telemetry, great work of friends continual on with learning and working with the Mission. :)
The ending was super interesting. You should always do that anyway, even if you have footage. I think it's brilliant. Those graphs didn't speak to me at all, but the simulations are super ililluminating. Great video, as always.
10:10 "I'm curious how you would solve it": Tiny spring loaded solenoids. "Digitally" excite solenoids between the fixed fins. They're either excited (proturde outside the skin a fraction of an inch) or unexcited (return flush with the skin). The plan is to control with 4 degrees of freedom on drag. As you know, at those speeds only a tiny amount of drag would be needed to get the effect. And being digital you can use PWM (pulse width modulation) or simply PFM (pulse frequency modulation) to effect adjustments. Put the solenoids in a negative feedback loop. With controls similar (i.e identical) to motor speed controls on helicopters you could maintain virtually perfect control with minimal weight and power penalty assuming fast reacting sensors (gyro like in a smartphone) and feedback. By using a "hinged mechanism" the solenoids could be mounted in line with the rocket, and the hinged fin could come out through a tiny slot when activated. It could even be just a small length of spring wire protruding through a tube at maybe a 15 degree angle.
I for one would absolutely love to see a video on the transonic region! The better I understand what problems your facing the more engaging and interesting the videos are :)
OK Ace Studsville.... i FOUND THAT TO BE SUPER-DUPER INTERESTING !!!!! I was right about the ammount of control surface throw. As you said nothing about servo-speed or micro sample rates, I assume no problems there. I have to subscribe now, because what you are doing here is radically AWESOME!!! Great video. Can't wait to see what's upcoming. That tiny "trim tab" you ended up with for control, is about what I expected to see with those insane velocity's. Great work.
I've been following the Sugar Shot to Space project for what seems like 20 years. I dare say you're going to beat them to space. Subscribed. Good Luck!
I feel smart! When you asked for "how would you solve this", i guessed you would integrate controllable surface into fins. I did however envision the controllable surface in the front of the fins; I wonder if the slot created by rotating the fins would have an impact on induced drag/vortex shedding since the vortexes would be close to the boundary layer on the fins and/or would the main-fin having a flow cleansing effect. Anyways...going to keep watching the rest of the video. Absolutely love how engaged I feel watching your content.
It should also be noted that more modern variants of the aim-9 (namely the aim-9x as shown) does not use canards. The aim-9x uses fixed forward canards with all-moving tail fins coupled directly to thrust-vectoring fins inside of the motor nozzle.
When you asked the question i immediately thought about embedded control surfaces like on an rc airplane. I would not have immagined that small of a tab though.
You can also place the control surface closer to the center to keep the fins thinner and s8nce you need more control authority on pitch abd yaw it would automatically normalize the pid gains and can even make the control surface bigger
13:00. Why not 3D print the fins in halves, upper surface and lower surface? With the outer surfaces smooth and with desired airfoil shaped; and the inside with a structural grid, into which you design a mounting pocket for the servo and actuating rod(s). The wires concealed in fin root extensions up to where they can penetrate the body tube under the root extensions. Fasten the two halves of each fin together with small metal screws and nuts. Or a fin structure containing the sharp leading edge with a precise fitting side-surface to be able to access servo, wires, linkages and movable control surfaces for easy installation and replacement. Then use small movable canards up front to control pitch and yaw, not roll. Find the size that gives the necessary control authority at all flight regimes. However, what do you use for control above 70,000’ or so when aerodynamics become ineffective? Can you employ very tiny gas thrusters on the 2nd stage? A CO2 cartridge from a pellet pistol as a gas source with a tiny sevro-actuated gas control valve box? Just pitch and yaw. Then you might be able to keep your rocket inside of a small recovery zone on the way back down. Coming back at or above Mach 1, at about 5,000’ AGL, pitch the rocket as far as possible, maybe even back upwards, until speed bleeds off to where the recovery chute can be deployed under light aerodynamic loads. What do you think? Will that work?
Thanks to Incogni for sponsoring this video! To get an exclusive 60% off an annual Incogni plan, use this link: incogni.com/bpsincogni
What happened to your RCS thruster system I think I saw you show off in an Everyday Astronaut video? Could that theoretically work in a situation like this?
Looking for an intern?
I sent a email to you a email a few months back about rollerons.. nice to hear you talk about them
Please do not use scam ads.
Thumb down.
As someone who just finished their mech eng final project that (to my own detriment) involved a truck load of ANSYS simulations, I don't envy the CFD that you ended up having to do 😂. I'll look forward to the next video with the details.
“Supersonic aerodynamic control” is in no way the same thing as “terminal guidance” two dudes in the van out front of my house.
That's not his department. He just wants to go up.
@@ChucksSEADnDEAD wink wink nudge nudge, say no more
That's just me I'm homeless not FBI. The other guy is FBI tho and hes in my van questioning me about your TH-cam comment.
I know playing with terminal guidance is a taboo of sorts. But is it actually illegal? And how come there is so little on that topic available online outside of the US?
@@MoritzvonSchweinitzit's not illegal per se, but doing it without license is problematic and posting about it can be illegal. But that's US. It is illegal in many other countries.
13:07 cracked me up, i just wasn’t expecting that mouth movement
"Alright, we're BABABABA into the video..."
his sense of humor is awesome. hilarious genius!
Whenever you say “ should I do a video on it“ no matter what it is I immediately want to see it
Same dude, same
Aww he's a boy scout...
Its 1 page... But it exists
@@wapiti4680 ???
“And as always, they were mounted with about thirty pounds of hot glue” there is something profoundly wrong with this man and that is why i am subscribed to his channel
not sure if your quallified to comment
BPS: "The good news is we aren't doing any of that math today..."
Me: " :( "
BPS: "...That's the topic of the next video."
ME: " :) "
This. 😂
So true
Same
@@everettgihring8099 That
Don't be sad. You've got just enough time to study Quantum Field Theory before the next video.
The neat thing about unplugging the servo on the first flight is it gave you a baseline for comparison and illustration of future flights.
You could have convinced me the animation looking down along the airframe was real footage. Holy moly the little tabs work so well!
Collab????
@@reuellouwrens9853 I mean, Lafayette Systems already uses this type of control on their DIAMOND-X rocket.
Man, I'm an automation and Control engineer in Brazil, I have a masters degree in robótica and I have been teaching engineering and mechatronics for 8 years. I have no idea how on earth you deal with so many complex details in these projects. It seems like it is done for a team of 10 senior engineers. You are crazy! How do you tune the controllers before the launch? Do you have a detailed mathematical model for the whole structure? You are insane man, keep up with the work. You're going to heaven.
JOEY B BACK WITH ANOTHER SLAMMER OF A VIDEO
🆙🆙🆙
Was that a reference to the informal nickname of the AIM-120 AMRAAM, or just a coincidence?
@@mortlet5180AGM-84 SLAM-ER
It's fun seeing you go through the same problems I encountered about 15 yrs ago.
I built a rocket with canard driven by high-powered rc servos. I put together a controller with an $1100 Analog Devices IMU and a custom microcontroller with SD card parameter control and data recording.
I initially wanted to control roll. had 9 flights without success. At first I thought I had a control system bug, then I thought I might have a hardware problem. Finally I mechanically fixed the canard at a constant angle. I had roll reversal in every flight. I finally installed a camera to verify the data collection. Unfortunately the rocket came in ballistic and the 8 foot rocket was compressed into 8" long wad.
I researched this problem and ran across a group in the NW that was having the same problem and had identified AIAA research that identified the problem many years ago.
My dad helped develop a few rockets for General Dynamics, they used something called “dithering” which made the fins vibrate(?) and when they wanted to steer in a particular direction they just made the fins spend more time on one side by interrupting the occultations, favoring the direction they wanted to go
Yah pretty sure Salvatore pais scaled this technique up to make the tic tac vehicles.. wingless UFO can only be explained via vacuum creation via high frequency vibration. Likely powered by compact fusion reactor.
So, duty cycle control? I've seen it work with rocket motors (see: kinetic kill vehicle used for the DART) but never with fins. Interesting!
Which is not a very new concept as we used similar methods for early R/C aircraft control before we came into the wonderful world of Proportional servo control, in the late 60's.😁
@@paulholmes672, well it was the early 1960’s when he was doing it, so yeah
Nice ... use of random control noise, and then removing some randomness to influence control. Makes me wonder if the "dithering" was a technique to disrupt shockwaves traveling across the control surface.
When you started in on canards, I started muttering "elevons". When you started talking about talking about rotation the whole fin, I nearly yelled "elevons" at the TV as well as grumbling in 737 Max overcontrol. Then when you showed your solution, I was like, DUH, elevons. Yes, I know the shuttle used a separate rudder, but honestly you should be getting the point since you had four "wings" instead of two. This is how supersonic aircraft control their flight.
And if you see my name and picture, and recognize them, LONG LIVE THE PID. Glad I found your channel, we did know each other. And to really induce swearing you need to botch a Kalman filter.
Joey B out here controlling an aircraft over a WIDEEEEE envelope on his f̶i̶r̶s̶t̶ second try, and I hope ya'll know just how hard that is! Aircraft control derivatives scale with dynamic pressure (combo of airspeed and air density)... This basically means an aircraft behaves very differently at different speeds/altitudes with the same control deflection. The successful flight is a testament to Joe's math all collapsing down and "un-coupling" the rocket from its environment, from aero modeling to control system design to sensing. Next video is gonna be good! TL;DR: 😍🚀😍
Awww thank you so much dude! That means a lot :)
100% ♥ ... x-plane level engineering!
Thanks! An awesome video, I just loved model rockets when I was a kid. Turned that into a EE/computer programmer.
Imagine the thickness of FBI's folder on this guy
😂😂
he uses Incogni, he's safe
I think
Bro is watching his every move to insure he does not make a missile because a missile, is just a rocket crashed in the right place
Well no, a missile traditionally carries a payload of explosive compositions. A missile would be a destructive device. He simply engineers delivery vehicles. @@CIWS-Goalkeeper
@@OnlyFacts11b welp, have you heard of a concept of kinetic energy? I mean an explosive would be great but if you add enough speed it would work. Also rocket propellant is not so much different from explosives, tweak the burn rate and you get something that goes kaboom
The fins extra thickness didn't seem to be an issue, but if you do want to reduce there thickness in the future one option is to mount the servos inside the fuselage and have a thin shaft go up into the fin to control the tab.
This could also allow almost all the cables to be inside the fuselage as well, further reducing drag.
We leaving the ground with this one 🗣️🔥🔥🔥
YOU KNOW IT 🆙
we goin UP 📈
We’re going ⬆️⬆️⬆️⬆️⬆️⬆️📈📈📈📈📈📈📈📈
The amount you talk about how much you love making videos about this stuff is so refreshing. So many TH-camrs get so caught up in creating content that they start to hate it, and while that isn't always controllable, I'm sure that as you grow, the donations and money from TH-cam will make it possible to have access to better parts and more prototypes. Keep up the grind.
"I am not proud of everything I do." is my life motto.
I hear you.
We have NO regrets.
@@glennllewellyn7369I do
Sometimes… life SUCKS and sometimes, the thing you make are… disturbing
Thanks! You can use the support. I do industrial control robots, so I know just how touchy the modeld are.
The simulations at the end are outstanding and it demonstrates how your channel not only inspires others to take on engineering projects, but perhaps even more importantly the channel builds a community of very talented aerospace engineers who come together and make meaningful contributions. Great job! Kudos to Gabriel, Matt, Andrew and Dan.
I always love the sheer attention to every detail Joe puts into everything. Everything from the rocket's functionality, the camera shots, and the video edits are all phenomenal quality. Keep up the good work.
As a computer scientist, I would be interested if you could make a wind tunnel to gather enough data on the effect of the vortices and build a controller to compensate for them. No idea how hard it is to build a wind tunnel though, and I'm sure I'm underestimating the difficulty of that control problem. Good luck! Great video as always
The problem is that funneling air into a supersonic flow isn't exactly for the home gamer.
building a wind tunnel is actuly kind easy. building a supersonic wind tunnel however, less easy.
Actually, building a supersonic wind tunnel in itself is not that difficult. Gathering the relevant amount of currency to do so, while building it in a place that is about 500 miles from your nearest neighbour........ different story.
As someone that studies Engineering Informatics (mixture of computer science and electrical engineering) I often have to work with computer science that had never heard any physics or mechanical engineering lectures and they are most times better at algorithmic and similar thinking stuff but don't let have mechanical or electronics ideas.
Wind tunnel up to a couple of hundred km/h no problem use a fan big and fast enough and you are good to go. You still need to think about how to reduce the effect of the turbulance of the fan, maybe build a pull configuration the expected wind speed for the same input power is slightly slower but should still work.
But now the trans and supersonic Problems. These speeds are not possible with a fan alone. And the pull configuration does nothing. If your fan is to powerfull you suck a vacuum but still don't get mach 1 and in the push configuration you still would only near to mach 1 and than produce a high preassure zone. So you need a nozzle the same type as the one on the bottom of a rocket engine. It solves the Problem of converting high pressure gas to high speed gas. BUT the gas goes from very high pressure (tens or hundreds bar (atmospheres)) to 1 bar or below every expanding gas cools down. And this kind of depressurization would lead at least to freezing water and CO2, but depending on the pressure drop even to liquid nitrogen or oxygen (unlikly in home build variants). So you don't need to start with a high pressure gas. You need to start with a high pressure high temperature gas. Problem is most materials get weaker under high temperature but the high pressure is really high so the whole thing is near its failure before you even have moving gases.
Second Problem you don't want to see effects of pressure waves reflecting of the tunnel walls because that would cause similiar uncontrolable feedback loops as the canards to the fins earlier in the video. So the wind tunnel should be at least factor 5 to 10 bigger to the test object (model rocket) now we test a small model with a fin can that is only 5 cm in with (to be honest to tiny to have good estimates on how a big one would react). The wind tunnel has at least to have a diameter of 25 cm. 0.25m*0.25m*pi/4=pi/64 m² this times the desired speed of 300 m/s (yes I know speed of sound is faster under normal pressure and temperature but I will not estimate which temperature we could manage to reach in our tunnel) at 0.5 bar is pi/64 m²*300 m/s*0.5=7.36 m³/s. So you would need every second of testing 7.4 kubic meters of air delivered to your wind tunnel, forget it. This is for each second 4 full (and not the smallest ones) air bottles that divers use. Because there is no for not that rich individuals available solution that would produce this preassure and quantity of compressed gas on the fly (exeption rocket motors but they burn to hot to hold a normal test object in exhaust).
I'm working on aerodynamics of rockets myself. From what I've heard from a people who works directly with that kind of aerodynamic tunnels, just the cost of preforming a single set of tests on subsonic tunnel capable of fitting rocket this size is at minimum few thousand dollars. Not to mention building a supersonic tunnel.
You can get good results just by simulations alone tho, I basicly work with that, Solidworks is a pretty good tool for obtaining such data and is extreamly easy to use. However I would questions it's results for trans sonic flows.
I was sitting here waiting for you to mention the rollerons on the sidewinder cause its like, one of my absolute favorite engineering solutions ever. So elegantly simple and yet works incredibly well
Edit: im not an engineer, just a mechanic, but i think a combination of servo tabs and entirely moving fins is the way to go. This is how the trim system works on just about all modern jet aircraft. Little tab moves in opposite direction of where you want the surface to go, and generates a moment on the major surface. Allows for precise correction without having to put an enormous load on whatever you're gonna hinge the tab with. Much easier to design a fin hinge that can withstand that sort of load than a tab hinge that can. Just my $0.02 though, supersonic aerodynamics is super weird, and aviation and rocketry are only marginally similar
21:15 Aside from the first couples seconds or so of this, you could've absolutely fooled me into thinking this was on-board flight footage. Well done to the community for helping here. Also would love a video on transsonic flight.
Before I finish watching the video; there is a neat trick that the first AIM-9 Sidewinder missile used to avoid needing to compensate for things like density altitude: it used force servos on the fins rather than position servos. The result is that, with a few assumptions about the center of force on the fin staying more or less the same distance behind the pivot, a given turning force on the fin created the same side force on the missile body more or less regardless of speed or altitude.
The entire guidance and control system on the first marks of the AIM-9 are a fascinating cases study in how a spec project with almost no budget can force better designs.
For sure please do a video on control over transonic flight regimes! And I'd love to see one (maybe combine them) about CFD you mentioned, what specifically was the effort required to succeed, the process, and the modeling. We love the theory
transonic is really hard (it's why jetliners which have the power to exceed about 0.85 mach just don't). Best bet here is just to get through to the other side as fast as possible.
Only found this channel a month or so ago. Have been binging it basically non stop. Somehow I'd skipped over this video and the amunt I laughed at 15:14. The best BPS foreshadowing of all time.
15:15 🤣🤣🤣🤣🤣. I'm very glad for your sake that you got that on camera!
Thru the whole length of the vid, one thing kept popping into mind was this: For control on the X-Y axis, a single, tapered spike on the nose, able to be pointed off-axis for 'steering'. This would eliminate downstream vortices from canards. Your tiny tabs on the back of the fins would take care of roll.
I've been watching your last 3-5 years of videos over the last month or two, this is pretty much the solution I've been yelling at the screen through all your roll control adventures. looking forward to the rest of the story.
The missile knows where it is at all times. It knows this because it knows where it isn't by subtracting where it is from where it isn't.
joey out here innovating cost-effective aerospace telemetry and guidance solutions with servo-in-wing-tabs
Good show. There's also inversion layers reflecting sound and sound reflecting upwards then reflecting back down at a further distance having a bland detection somewhere between. Artillery sometimes does that.
You are the perfect mix of youtuber and engineer, these videos are great and inspire people. Thank you... can't wait for the math video :)
This edit hits different! Loving it 💯
KEG ROCKET NO WAY, bro can you please tell me if your gonna keep uploading in TH-cam I need to know!, I’m a big fan bro!
So excited for your project!
Transonic video needs to happen for sure. Also great job on figuring this one out, I know it was a big step and took a lot of hard work. Every day you are getting closer to space and I love to see it!
I love watching the process over the years
Thanks!
We leavin the earths atmosphere with this one 🗣️ 🔥🔥🔥
The care you put into your these videos is amazing. Like in 13:07. If you are a professional video editor like me, you know you've had to shot multiple takes saying various minutes (maybe 20+ of them) and then choose the one that matched in post!
You’re both a good rocket scientist and also a good storyteller
Im 28 years old, former police Sgt., college grad, and father of 3....I look up this man. What an inspiration you are. Weather permitting, I am about 1 week from getting my PPL as I now pursue aviation full time! Hope to meet you some day!
There is a reason for all moving fins other than extreme manouvrability. At the end of WW2 aircraft became uncontrollable due to the elevator being behind the shockwave (where it looses effectiveness) just when they were very much needed because of the onset of Mach tuck (which is a nose-down moment on the lifting surface in transonic flow). The problem of controllablility was solved on the Bell X-1 with an all moving horizontal stabilizer. On supersonic missiles this is the all moving fin. Your trim tab seems to function more as a spoiler (so to create drag) than as an aerodynamic surface to direct the airflow. It definitely gets credits for a shoestring approach. Hope you don't mind me making a suggestion: spoilers don'' t necessarily have to be in the fin section, they might as well be in the aera where you would expect the canard fins or anywhere else on the vehicle. The obvious advantage of moving the spoilers away from the fins is the available space for the servo's. Now your fins can be thin again to reduce the drag. I hope this suggestion will help you past Mach 2 and eventually into space in a controlled manner.
I love all the positive on subject comments here, so nice to see from real space enthusiasts.
What awesome work, keep it going from the UK.
There’s nothing better than trigonometry
REAL
What about laminar flow?
😮😭
*Quadratic Equations left the chat*
there is one and it is called calculus
@@Physicsguy-r7k Hell yeahh calculus the GOAT
I'm 50, forced to retire last year. been doing some embedded system engineering works for more than 2 decades. Watching your videos, I have to say they made the right decision and that involuntary retirement was the best thing that has ever happened to this man. Keep up a good job as you have been! Good Luck too.
Don't feel too bad, Lockheed made a very similar connector mistake on the THAAD program. Missile went about 100 yard before burying itself into the desert. The connector was to load the guidance program, but when the wires pulled out they shorted the guidance system. So, no guidance. D'oh!
Thanks
Silo launched rockets, guiding eggs to specific places on the ground, supersonic aerodynamic control, and soon to be a space shot? You are definitely on a watchlist somewhere.
10:36 that reminds me of the stabilizer-stabilator thing with airplanes. When you have a stabilator the stabilizer is the elevator, so basically the stabilizer itself rotates the change the altitude. When you have a stabilizer, there is a separate elevator that moves on the back of the stabilizer. This is like the stabilizer.
19:36 They are nice graphs. What software is it in?
I know i am 7 months late, but in case you still want the info. It's graphana
As an Aerospace Control Engineer, this is absolutely amazing! Yes we would actually love to see more videos on control architecture and if you had implemented gain scheduling for pre and post mach flight!!
Thanks a lot these video! All the best for future control missions
> "Switched over to fancier onboard cameras, GoPros"
> 2 minutes later
> "None of the GoPros worked"
You got the full GoPro experience
Outstanding video production Joe! Very interesting discussion on the control issues and the importance of doing good post flight analysis of what worked and what didn't.
I think you bumped into the problem they had breaking the sound barrier with the oldschool split control surfaces. You've gone smart and made a teeny tiny control surface, but the split control surface becomes basically useless in the transonic region, which is why they moved to all moving control surfaces.
You've probably already come to this conclusion, but if you put a teeny all moving contol surface at the tip of your fin, you'll retain control authority at speed. Having the surfaces that far out will make them more potent in roll as they have a longer lever, but hey, make smaller movements.
Keep it up, it's an instant click whenever I see one of your videos!
You could alternatively literally make ‘ailerons’ (on aircraft wings/tails) on the bottom side of the fins if you need something a touch more commanding. Then the servo is still housed in the fins, you can use simple linkages for mechanical leveraging to use small servos with ‘big force for small movement’ and it’s simple to model, print and attach.
I just watched past where you’re asking for help💀😂
BRO ITS YHE GUY BEPIS JOHN OUT HERE ROCKET? a good start to the day :)
That telemetry video simulation is super impressive. I love the amount of data collected and the effort taken to reenact the flight video.
Joseph bizzlington back with actual missiles now 😂
those visual representations of the camera data are nothing short of absolutely incredible! Man, that is exceptional work. Bravo all.
I am 13 year old boy. You have truly inspired me to pursue my passion in rocketry.
Be prepared for lots of failure, and you best damn well make sure you get up and try again. Do. Not. Give. Up.
Me too I was 12 when I saw this guy and now I’m 16 sending my first high power rocket
@darkmetal20 Me too!!! I already tried to build a sugar - kno3 rocket but it failed
This summer I will try again!
Internet safety: 📉📉📉
In the past, only a couple of objects could reach supersonic, the goal being INCREIDIBLY difficult, challenging, and made by tens of engineers, but now, even a random guy with a youtube channel can do so. How humanity advances so quickly is just too beautiful
YES HE UPLOADED
Bro I love been checking like once a day for like 3 months
@@JackDalfino I just turn on notifications
We are all excited but why are you yelling?
@@dr4d1swhy are you so mad, and I’m not yelling dumbass 💀
Maybe you need a simple analysis tool that checks all the connections are in place. ReadyToLaunch and return a green or red. Checks the values are where they should be.
As you go higher with more cost then you might be glad you have this simple analyser.
Good video. Keep going, you'll get there.
Is it worth simulating the glue and bumpy lumpy surface?
Perhaps someone can make you a camera mount and design a groove in your rocket tube to fit the cable too.
Total guess that imperfect surface might be affected by fluid dynamics at supersonic speed and above. A strong fluttering pocket of fluid could possibly move the rocker or damage it. Maybe even make it's fuel usage above calculation.
20:04 i want that video
Supersonic flow is fascinating. I have been working on my own rocket to test control methods in a compressed flow flight envelope and can say the work you are doing is really amazing. I can't wait for the next video!
AIM 120 AMRAAM has left the chat
You do all this to put something in orbit is something. You do it for the latter ,and filming for entertainment is something++ . Bravo
STOP MOVING THE GRAPHS!
MOVE THE GRAPHS!
EAT THE GRAPHS!
@@gamerbossharmonFUCK THE GRAPHS!
Dude, I super love nerding out with you. I can pretty confidently say that every video you've already put out and all the future ones that you haven't yet made. Just do what you love. I'm pretty sure we'll all like it.
he looks exactly like elon musk
Ikr😂
lmao elon needs to see this🤣
Undergrad aerospace engi here. Glad to see how feedback control of dynamic systems and aerodynamics courses may be applied within a small scale in a workshop! I Hope I'd be able to construct any advanced sounding rocked before graduation inspired with your channel!
Your channel has a some of the most well made videos on this whole site! Thank you. I’d love to see a video about why it’s particularly hard to manage flight at that subsonic range of Mach 0.8 - 1.2.
I perceive through my experience of supersonic/near supersonic flight control (16 years around F-111 Avionics, :-)), your early on hunting may have been an effect of the boundary layer flow (BLF) across the fins going through changes due to the sonic transitioning shock wave generation. Once supersonic flow was stable, the roll surface could 'play' in a 'quiet' space. We had multiple static pressure sensors and localized pitots around our jet airframe that would measure Boundary layer pressure to, then help, fine tune surfaces such as flight controls but also engine inlet shaping to keep inlet air subsonic. I would suggest looking into the design changes done, over the early years, to both aileron placement along the trailing edges of wings to eliminate/minimize aileron blanking and things like spoiler tabs, etc. Also, early supersonic aircraft went from horizontal stabs w/elevator tabs, but had to switch to all moving tail planes for elevator control, as the elevators could also blank, but some of that effect had to do with the REALLY turbulent stuff coming of those wings (and their shock waves in front of the always rear mounted elevators/elevons. Definitely a lot more of a dilemma than just pointing the surface/tab where you want to 'steer'. Excellent work sir!!! 😁
You know you're one of the most brilliant people I have seen on TH-cam in this area and you're doing this on your own with help from Friends of course your education and your commitment consistency drive and absolute refusal to quit. If there's anyone that people would give money to to support the humanity's Space Program where the data comes to humanity not filtered through some space agency!
I think this is the best video you've ever made, keep upping that production values! This is amazing!
Also, yes please, we all want to see the video about transonic flight regimes.
Way cool! Maybe this is in the next script already, but when you want to add in pitch and yaw, will the tabs needs to be substantively larger, and will that put a greater strain on your control software (and maybe your servos) to not overcrank the roll correction?
I was driving the other day across the desert and had a thought that I hadn't seen anything new from you in awhile. Good to see a new video!
great job on making that rocket! this is so awesome!! honestly that CAD model of the data from the flight is pretty cool, ive not seen POV like that before
I would watch a video titled "10 sexy tops to keep it going steady (how to prevent your missle from rolling)." Joking aside, the production value of these videos just keeps climbing- love it!
This is awesome! And your videos were already amazing, but the increase in animations, editing, etc took it through the roof!!
That red curve...always speechless watching your videos. Such a brilliant young man.
My first thought on seeing the secret alternative control option (10:45) was of all the x-planes failing to break the sound barrier prior to the Bell X-1. They all had hinged control surfaces, where-as the X-1 has an all-flying stabilizer. This because during transition from transsonic to supersonic, the shock wave made the rear hinged part of the control surface unpredictable and ineffective. Great coverage of why this an issue (form 19:44), and the importance of control-surface sizing (10:59).
The X-1's actuators become over powered by supersonic forces, but trim control saved the day for maintaining stabilized flight. But that's a different part of the "Right Stuff" story.
Do you have a target velocity range for "stage sep" on the "space shot"? This seems to be a critical design concern to avoid the transonic region. Thinking this over gave me a 💡🤯 awe moment when recalling time in flight most strap on boosters separate on commercial launches.
BTW: So-called "Naked" camera's are a common with FPV drones (whoops), and can be purchased disassembled. One major drawback with naked cameras is lack of heat sinks, so can prematurely shut down. Could be avoided if not powered up til critical time in the launch sequence. (if this turns out to have been the issue)
You are right, y’know… this _is_ huge! I am impressed, again, and always!
As for “let me know if you would like….” yes, how many times, we love your videos, make! :)
Great news Joe, about the test flights. So amazing to view the simulated flight using telemetry, great work of friends continual on with learning and working with the Mission. :)
The ending was super interesting. You should always do that anyway, even if you have footage. I think it's brilliant. Those graphs didn't speak to me at all, but the simulations are super ililluminating. Great video, as always.
10:10 "I'm curious how you would solve it": Tiny spring loaded solenoids. "Digitally" excite solenoids between the fixed fins. They're either excited (proturde outside the skin a fraction of an inch) or unexcited (return flush with the skin). The plan is to control with 4 degrees of freedom on drag. As you know, at those speeds only a tiny amount of drag would be needed to get the effect. And being digital you can use PWM (pulse width modulation) or simply PFM (pulse frequency modulation) to effect adjustments. Put the solenoids in a negative feedback loop. With controls similar (i.e identical) to motor speed controls on helicopters you could maintain virtually perfect control with minimal weight and power penalty assuming fast reacting sensors (gyro like in a smartphone) and feedback. By using a "hinged mechanism" the solenoids could be mounted in line with the rocket, and the hinged fin could come out through a tiny slot when activated. It could even be just a small length of spring wire protruding through a tube at maybe a 15 degree angle.
I for one would absolutely love to see a video on the transonic region!
The better I understand what problems your facing the more engaging and interesting the videos are :)
6:10 your face after the cut back screams "why did I even think that was a good joke in the first place" in the most hilarious way.
OK Ace Studsville.... i FOUND THAT TO BE SUPER-DUPER INTERESTING !!!!! I was right about the ammount of control surface throw. As you said nothing about servo-speed or micro sample rates, I assume no problems there. I have to subscribe now, because what you are doing here is radically AWESOME!!! Great video. Can't wait to see what's upcoming. That tiny "trim tab" you ended up with for control, is about what I expected to see with those insane velocity's. Great work.
I've been following the Sugar Shot to Space project for what seems like 20 years. I dare say you're going to beat them to space. Subscribed. Good Luck!
I feel smart! When you asked for "how would you solve this", i guessed you would integrate controllable surface into fins. I did however envision the controllable surface in the front of the fins; I wonder if the slot created by rotating the fins would have an impact on induced drag/vortex shedding since the vortexes would be close to the boundary layer on the fins and/or would the main-fin having a flow cleansing effect. Anyways...going to keep watching the rest of the video. Absolutely love how engaged I feel watching your content.
It should also be noted that more modern variants of the aim-9 (namely the aim-9x as shown) does not use canards. The aim-9x uses fixed forward canards with all-moving tail fins coupled directly to thrust-vectoring fins inside of the motor nozzle.
When you asked the question i immediately thought about embedded control surfaces like on an rc airplane. I would not have immagined that small of a tab though.
You can also place the control surface closer to the center to keep the fins thinner and s8nce you need more control authority on pitch abd yaw it would automatically normalize the pid gains and can even make the control surface bigger
Gabriel Yamato is the best when we talk about simulations with Ansys. Tu é foda!
I definitely would be interested in video(s) about the transonic regime. Not to mention subsonic/supersonic/hypersonic
13:00. Why not 3D print the fins in halves, upper surface and lower surface? With the outer surfaces smooth and with desired airfoil shaped; and the inside with a structural grid, into which you design a mounting pocket for the servo and actuating rod(s). The wires concealed in fin root extensions up to where they can penetrate the body tube under the root extensions. Fasten the two halves of each fin together with small metal screws and nuts. Or a fin structure containing the sharp leading edge with a precise fitting side-surface to be able to access servo, wires, linkages and movable control surfaces for easy installation and replacement. Then use small movable canards up front to control pitch and yaw, not roll. Find the size that gives the necessary control authority at all flight regimes.
However, what do you use for control above 70,000’ or so when aerodynamics become ineffective? Can you employ very tiny gas thrusters on the 2nd stage? A CO2 cartridge from a pellet pistol as a gas source with a tiny sevro-actuated gas control valve box? Just pitch and yaw.
Then you might be able to keep your rocket inside of a small recovery zone on the way back down. Coming back at or above Mach 1, at about 5,000’ AGL, pitch the rocket as far as possible, maybe even back upwards, until speed bleeds off to where the recovery chute can be deployed under light aerodynamic loads.
What do you think? Will that work?