The fact that it did a belly flop then back to vertical even though it still didn't land successfully was still amazing to see! A huge amount of respect and props to you Mr. Benard!!! Keep it going
@@swamyolo6208 That's not a bad idea but he already ate up all of his mass margins and adding anything more such as more rockets and igniters would only add way more complexity and mass. sorry to be "that guy"
The 3 stages of control engineer grief all present in one video: 1. "I can add a filter to fix it." 2. "Filter didn't work. Need better hardware." 3. "It's all working! ... if it was actually where it thought it was." I think I've gone through this cycle on every single one of my projects lol
Its really enjoyable to see how you overcome the engineering challenges. When you stick the landing, you will join a very exclusive club of engineers that can reverse park a rocket (non destructively).
Closely Check 0:26 to about 0:32 - While the engine is not running, your control system is. This leads to a massive over correction when the second motor lights up. Think about shutting down gimbaling until the second motor is running. This would have made the flight MUCH smoother and easier to land. As you know the engines have a very high thrust right in the beginning and then the thrust sharply falls off. If that high amount of thrust is used for gimbaling, you'll have to use the entire rest of the thrust to correct it. - Alternatively you can use the high thrust for gimbaling but you have to dial the angle down to a fraction for this small amount of time
In a way that is correct, but the thing is, when igniting, the engine will point in the wrong direction and the rocket will be completly out of range for the landing pad. Letting the control system run is more appropriate as it doesn't exactly knows when the engine is gonna ignite... I might be completly wrong, and you completly right. But it is always nice to share opinions. Have a good day you all ☺️🚀
As my wife I suffering with cancer you are a bright spot in my life. Love you little jokes and am glad to see you got over the Covid blues. Keep up the great work.
A few thoughts. I used to be involved in quadcopter software development and the loop rate can make quite a difference. Some digital servos will run up to 400Hz. At that rate mechanical inertia may well help damp out spikes. It depends on the source and frequency of the spikes. To make a simple lightweight damper for the landing gear take two carbon tubes where one fits inside the other. Add a bit of grease on the sliding surfaces for damping. You can get silicone damping grease in a wide variety of viscosities so you can tune the damping rate. If you seal the ends of the tubes you can also use the trapped air as a spring. The grease will seal well enough for the short time you need to hold pressure. You can adjust the spring rate by adjusting the ratio of open to compressed air volume.
This is some of the most beautiful machine footage ive ever seen in my life and it was just for a 10 second clip in your video. Some of those shoots deserves to be presented in a museum
Keep it up, Joe! Your ability to analyze what went wrong during each particular flight is perhaps your greatest gift - every iteration is a new challenge, but it's always something new that you're tweaking / improving / trying. It has been a lot of fun to watch (along with your high-altitude, parachute-assisted-descent rockets!), and I'll be here to watch every video to see how this project progresses!
This is now 1 of the extremely, *_EXTREMELY_* few channels that have succeeded in making me wanna follow along on their journey enough to make me subscribe.
Joe, when you kick that ascent motor out of the way for the descent, it is resulting in the rocket pitching in the opposite direction, which is to be expected. You need to eject it just before burnout, not wait until it stops. That way, the rocket still has upward momentum and the kickout will go directly down, with the resultant pitch of the rocket nulled. That way, it will descend downward vertically, not pitched over before initiation of the descent motor.
But there are high chances that the motor will not eject because of the little fuel burning in it, the motor is powerful enough to launch a 1Kg rocket and if the motor if giving out thrust alone without a payload, even a minimal amount of thrust is enough to not let it out.
Part 2: Also if the landing motor is fired to push it out, the thrust will just add up to the upward motion of the rocket as you said "the rocket still has upward momentum". NOTE: I didn't make the vectoring and ejection system, I'm just stating what I think and me no professnol :)
the machined ''beams' you now use on your TVC mount look like you could still strip a lot of material from them. Milling through it to generate a U-shape should leave you with a part that is maybe 80% lighter, and probably just as stiff. weight that could go to something else, or just to a little more height.
For the motor mount hardware , I would suggest you look at the cyclic control of the Align 700E RC Helicopters. Maybe include thrust bearings in the gimble. Love watching the ingenuity and enthusiasm. It takes a dedicated man to launch at dawn.
I was briefly working with drone control systems on a student autonomous drone project. I concluded that instead of controlling a parameter directly it is better to control the derivative (or second derivative). So for example when the rocket angle is off you'd control the angular acceleration (or angular velocity if your sensors aren't precise/fast enough). You could still use PID controllers but they would be cascaded. The great thing about this is that you can now control how aggressively the set point is pursued, like setting max angular speed/velocity etc. I'm sure there are better options than cascading PID controllers, but this almost always performed better than a single first order PID loop. Faster (several kHz) loops like motor ESC and self balancing etc are a different story, they need to be fast.
Dude I think I'm speaking for everyone when I say that we all love watching you progress towards this amazing goal. I have been watching since the beginning and when I look back, I am amazed at how far you've gone. Keep making these videos, they are really inspirational to everyone who watches this channel, including myself. You have inspired me and motivated me to do crazy things that I wouldn't normally do and I thank you for that. Bounce be gone and godspeed Scout F!
“Motivation is the stuff that permeates your entire being when you have a clear, vivid picture in your mind of what you want to do and an intense, burning, all-consuming desire in your heart to fight for it.”
...and this is how you fall for sunken cost fallacy. Not that I would want him to give up, the opposite is the case. But this mindset in general is how you end up sticking to an erroneous approach for far too long, just because you've been at it for so long and invested so much time, effort and money into it.
Been watching your progress on and off for a while now. Always amazed just how difficult and complex what you are attempting to do is. I have no doubt whatsoever you will be successful. I enjoy watching your progress and look forward to your eventual and inevitable success.
I sometimes go out and shoot video with one or two mirrorless cameras and usually a couple of action cameras and sometimes running a dedicated audio recorder and it can be quite a handful by myself. The I see you running what seems like even more cameras while doing a rocket flight checklist and countdown AND running a livestream simultaneously. That's nearly as impressive as it's going to be when this things finally lands like you want it to. I can't imagine how much work all of this is and how you don't forget to hit a record button or check a battery level there is beyond me. Oh yeah and you're flying a drone. You are not human.
I actually stopped following your progress for some time cause I thought you was hitting a bottle neck. Then I saw you got the throttling working pretty well, then I was excited. I am ME graduate student and I really think you're very close. Nice work and I love it!
Keep up the good work Joe. Started following BPS the past month or so. The problem with software is that it takes time to code some good algorithms for computing the things it needs. Training Neural nets could be one idea, I mean for the ship to make the right decisions based on those nets.
I think the people who are suggesting liquid fueled engines are severely underestimating the complexities involved in that. I also think they don't realize that using liquid propellants kinda defeats at least part of the purpose of this entire project.
Liquid is too complex, yes, however, hybrid is even simpler than the current setup because it's throttleable. Granted, he's not doing this because it is easy, he's doing this because it is hard.
That was astoundingly good! It was vertical at touch down and looks like bounce was why it fellover (it wasn't on the landing pad). You are so so close to nailing a landing.
14:27 Stupid idea, but what if you cheat a little and add some sort of spikes on the landing legs that will hold rocket as it lands. A spike that just springs out as soon as the landing legs deploy would let you choose an appropriate length while being flush when the landing legs are retracted( when not deployed ). Better, if the spikes were angled away from the Y-axis, so if it lands on spike first, it will push the rocket in the opposite direction.
First of all as an automatic engineer....respect to ur work chapeau....the problem of drifting is obviously the period between the first and the second burnout i think u need to add tow "2" motors very little impuls to recompensate on the pitching disturbance like the Russian missile system
I like you man and I have been following you for years. I just think the TVC craze has flown and landing it is an attempt to revisit the hype of when your channel blew up. Its good that you do really well at these attempts but dont give in to the hype of your channel and feel like you need to keep the audience happy. It might take 10 years but you will get there.
One thing that might help with landing on a pad is separating land vectoring into multiple stages, for example you could have one stage aiming the rocket towards the pad, a second which focuses on stopping the rockets lateral movement, a third making the rear point at the pad, and a fourth which focuses on the soft landing. Obviously this would work easier with longer engine burns or stabilizing fins, but it's at least a start.
Hey Joe, I have been watching your videos now for a while and am absolutely impressed with your tenacity. I have a suggestion that may be late to the party but may give you an easier time controlling the oscillation issues you are have on the gimble. 1. Instead of angular servos adopt a NEMA 8 or 11 stepper attached to either a gear rack and pinion or lead screw. Your stroke is roughly 5~7mm if I can gage from the video. The stepper has a holding torque to help stop the bounce but adding a lead screw will stop the bending moment and loss of leverage at the end effect. I recommend a lead screw vs gear rack due to the ability of the lead screw to add back feed counteraction. Use a fine pitch of about .5 and 8mm D. You will need to add a stepper driver of course and then in C create a simple PWM to step/direction algorithm to convert the PWM angular to steps. Not hard and there are several examples. This will add weight but not so much that you couldn't add grain size to compensate. 2. I see you try to use screws as pivot points. I highly recommend using bronze bushings and shoulder bolts giving you excellent tolerance at you pivots and an easier way to fine tune the fits as the shoulder bolt is precision ground. Use bronze flat washers and polish to fit between faces perpendicular to the pivot axis to remove lateral and bending slop within the pivot joints. This approach is similar to how space x controls the motor gimbles on the Falcon. How I know that is NDA protected so.... I wish you luck and low winds brother.
Problem I see is that with a finless/gimbled engine rocket, is that it is obviously unstable when the thrust stops. The rocket immediately went horizontal and at that low altitude never had a chance for the booster motor to settle in and stop the pendulum. Also the boost takes it off line from vertical and the return motor tries to correct this and get it back to the pad. Keep it up. You will get it. Exciting watching your progression.
Thanks for giving us a glimpse into an alternate reality where SpaceX decided to develop propulsive landing with solid rockets instead of liquid Merlins!
Never give up, never surrender, and maybe go faster on your control system. When I was doing control systems design (servos on very large robots in nuclear reactors) we were always shooting for 1kHz. Your slow servo rates are bound to encourage oscillation. I assume your rates are slow just because you have small processing horsepower. If they are slow because that better matches your dynamic systems that's another thing entirely, but I'm thinking your systems are small and light, so they could benefit by faster control loops. Good luck!
When using solid rocket motors for a vertical rocket landing, one of the problems is the inability to adjust the thrust of the motor. However, it is possible to get an adjustable vertical thrust component by deflecting the motor axis from the rocket axis. For symmetry it is necessary to take two engines and deflect them simultaneously in opposite directions. If you deflect the engines by 60 degrees, the force along the rocket's axis will decrease by half, if 90 degrees, then to zero.
Mr. Benard, there are 4 supports only in your rocket for landing. Your idea is very proud to me. But if 1 support is unstable, the whole rocket will tilt, and fall to the opposite side and finally drop as only 1 support in the opposite side. You can investigate in very slow motion of how the rocket land in less than 1 second. As the plan area of the rocket is a circular shape, but your rocket using 4 truss supports as square shape. it is difficult to form it as a whole structure. I would suggest to modify it to have 8 supports which is a little bit similar to a circular shape in order to to improve its stability. If it won't have adequate area around the rocket perimeter to install 8 supports, try to install it as 2 layers which have 4 supports in each layers but with different orientations. Another important point is its support leg length, whether it will form as 8 supports with same radius on plan view after landing or even longer length in upper supports in rocket by forming a truss to reduce its slenderness ratio to minimize its buckling or twisting and make 8 supports to form 4 corners as inner supports and anther 4 corners as outer supports. Moreover, "1 outer support together with 2 inner supports" or "2 outer supports together with 1 inner support" can also form itself as a large triangular stable support if horizontal members or even say extended tie, spring, elastic material as rubber band, etc. are installed between them and finally forming 4 bigger triangular 3D truss supports and its stability will increase much. I am a Structural Engineer, the rocket with 4 landing supports are similar to a tall building rest on a small pad footing foundation. Such building will become more stable if bigger size of pad footing foundation could be constructed to increase the safety factor for Overturning Checking as the moment arm from one side to the opposite side is increased. Please take it into consideration.
Would it be possible to add a lip under the nose cone to increase drag on descent while keeping the drag in the right spot for the accent? If that would even work with it's small size though
The fact that it did a belly flop then back to vertical even though it still didn't land successfully was still amazing to see! A huge amount of respect and props to you Mr. Benard!!! Keep it going
It’s like a mini starship lmao
@@johndavid360 yeahh
Add maybe rockets to the top of the rocket so when it lands to give a little kick so it doesn't bounce
@@swamyolo6208 That's not a bad idea but he already ate up all of his mass margins and adding anything more such as more rockets and igniters would only add way more complexity and mass. sorry to be "that guy"
The 3 stages of control engineer grief all present in one video:
1. "I can add a filter to fix it."
2. "Filter didn't work. Need better hardware."
3. "It's all working! ... if it was actually where it thought it was."
I think I've gone through this cycle on every single one of my projects lol
And then you realise you didn't need step 1 and 2 if you fixed 3 imideatly.
@@someonespotatohmm9513 hello mr potato
Its really enjoyable to see how you overcome the engineering challenges.
When you stick the landing, you will join a very exclusive club of engineers that can reverse park a rocket (non destructively).
He’ll legit be the third person/group to do it😂
Closely Check 0:26 to about 0:32 - While the engine is not running, your control system is. This leads to a massive over correction when the second motor lights up. Think about shutting down gimbaling until the second motor is running. This would have made the flight MUCH smoother and easier to land. As you know the engines have a very high thrust right in the beginning and then the thrust sharply falls off. If that high amount of thrust is used for gimbaling, you'll have to use the entire rest of the thrust to correct it. - Alternatively you can use the high thrust for gimbaling but you have to dial the angle down to a fraction for this small amount of time
yep - I think "integral windup" is the actual search term. Seems like this is a huge part of the instability.
@@reid-dye Interesting. Integral windup....
Nice analysis
Good catch!
In a way that is correct, but the thing is, when igniting, the engine will point in the wrong direction and the rocket will be completly out of range for the landing pad. Letting the control system run is more appropriate as it doesn't exactly knows when the engine is gonna ignite... I might be completly wrong, and you completly right. But it is always nice to share opinions. Have a good day you all ☺️🚀
That TI-84 shirt goes crazy
It’s going to work eventually. I can’t wait to see one successfully land.
Not enough to stick one landing, it needs to stick *all* landings. That's much more difficult, because you can't just get lucky.
As my wife I suffering with cancer you are a bright spot in my life. Love you little jokes and am glad to see you got over the Covid blues. Keep up the great work.
A few thoughts. I used to be involved in quadcopter software development and the loop rate can make quite a difference. Some digital servos will run up to 400Hz. At that rate mechanical inertia may well help damp out spikes. It depends on the source and frequency of the spikes.
To make a simple lightweight damper for the landing gear take two carbon tubes where one fits inside the other. Add a bit of grease on the sliding surfaces for damping. You can get silicone damping grease in a wide variety of viscosities so you can tune the damping rate. If you seal the ends of the tubes you can also use the trapped air as a spring. The grease will seal well enough for the short time you need to hold pressure. You can adjust the spring rate by adjusting the ratio of open to compressed air volume.
Won't the launch forces compress the legs?
Oh man so close, it looked really good until the landing part...! Awesome shirt btw
Guys he landed it!!!!!! New Twitter post!!!!!
That heavy metal TI-84 calculator is my new favorite thing
Our boi tony pepperoni back with another bepis spase video
This is some of the most beautiful machine footage ive ever seen in my life and it was just for a 10 second clip in your video.
Some of those shoots deserves to be presented in a museum
“It’s the summer of Scout.” Bring it on!
Keep it up, Joe! Your ability to analyze what went wrong during each particular flight is perhaps your greatest gift - every iteration is a new challenge, but it's always something new that you're tweaking / improving / trying. It has been a lot of fun to watch (along with your high-altitude, parachute-assisted-descent rockets!), and I'll be here to watch every video to see how this project progresses!
This is so cool to watch, joe - keep up the great work!
Joe, you're my boi, this is top notch building and editing content
Agreed
This is now 1 of the extremely, *_EXTREMELY_* few channels that have succeeded in making me wanna follow along on their journey enough to make me subscribe.
Joe, when you kick that ascent motor out of the way for the descent, it is resulting in the rocket pitching in the opposite direction, which is to be expected. You need to eject it just before burnout, not wait until it stops. That way, the rocket still has upward momentum and the kickout will go directly down, with the resultant pitch of the rocket nulled. That way, it will descend downward vertically, not pitched over before initiation of the descent motor.
But there are high chances that the motor will not eject because of the little fuel burning in it, the motor is powerful enough to launch a 1Kg rocket and if the motor if giving out thrust alone without a payload, even a minimal amount of thrust is enough to not let it out.
Part 2: Also if the landing motor is fired to push it out, the thrust will just add up to the upward motion of the rocket as you said "the rocket still has upward momentum".
NOTE: I didn't make the vectoring and ejection system, I'm just stating what I think and me no professnol :)
the machined ''beams' you now use on your TVC mount look like you could still strip a lot of material from them. Milling through it to generate a U-shape should leave you with a part that is maybe 80% lighter, and probably just as stiff. weight that could go to something else, or just to a little more height.
Hey! Been watching you for years! Im happy to see how far you've come!
The third flight was really, really good, amazing. That was probably the best landing attempt I’ve seen so far.
For the motor mount hardware , I would suggest you look at the cyclic control of the Align 700E RC Helicopters. Maybe include thrust bearings in the gimble. Love watching the ingenuity and enthusiasm. It takes a dedicated man to launch at dawn.
This progress is so insane to watch! I can't wait to see it land for the first time
Just what I needed to watch when taking a break from coding my own much simpler flight controller for PI Pico with FreeRTOS
I'm obsessed with this channel.
I think it's incredible you've done even this much! That 3rd flight was so close! Awesome job
I was briefly working with drone control systems on a student autonomous drone project. I concluded that instead of controlling a parameter directly it is better to control the derivative (or second derivative). So for example when the rocket angle is off you'd control the angular acceleration (or angular velocity if your sensors aren't precise/fast enough). You could still use PID controllers but they would be cascaded. The great thing about this is that you can now control how aggressively the set point is pursued, like setting max angular speed/velocity etc. I'm sure there are better options than cascading PID controllers, but this almost always performed better than a single first order PID loop. Faster (several kHz) loops like motor ESC and self balancing etc are a different story, they need to be fast.
I like the fact that after 5 years you keep trying and improving.
You're a legend, never give up.
Dude I think I'm speaking for everyone when I say that we all love watching you progress towards this amazing goal. I have been watching since the beginning and when I look back, I am amazed at how far you've gone. Keep making these videos, they are really inspirational to everyone who watches this channel, including myself. You have inspired me and motivated me to do crazy things that I wouldn't normally do and I thank you for that.
Bounce be gone and godspeed Scout F!
Dang! She's close to landing! Love it. You will get it soon!
“Motivation is the stuff that permeates your entire being when you have a clear, vivid picture in your mind of what you want to do and an intense, burning, all-consuming desire in your heart to fight for it.”
Never give up Joe. You've been at this too many years to quit! I would know, I've watched along with you! 😂 You got this. 🚀💪🏻
...and this is how you fall for sunken cost fallacy.
Not that I would want him to give up, the opposite is the case.
But this mindset in general is how you end up sticking to an erroneous approach for far too long, just because you've been at it for so long and invested so much time, effort and money into it.
Aside from all the awesome rocketry, I just have to say that the sound design on this video is out of this world amazing!
I doubt I will ever fly a model rocket myself, but it's so enjoyable to watch you progress in this challenging hobby. Good luck on your next flight!
You can start model rocketry way easier than what Joe is doing! It's cheap and super fun to get into!
Been watching your progress on and off for a while now. Always amazed just how difficult and complex what you are attempting to do is. I have no doubt whatsoever you will be successful. I enjoy watching your progress and look forward to your eventual and inevitable success.
Also, damn close!! Especially after such a hard yaw I’m amazed it came anywhere near correcting that but it did! Well done
It's so cool to see it do an entire starship belly flop and almost make it.
You’re at a place where an $83 improvement shouldn’t be flinched at. Keep it moving, your work is gooood.
That TI-84 T-shirt is actually epic and badass
Man been following you for a few years I see the evolution the skill set improving always moving forward!
Keep it up, Joe. I know you'll stick the landing soon. Also, the CNC build montage in this video was really cool.
LETS GOOO JOE, YOU FINALLY DID IT
(actually heard from Matt lowne
I sometimes go out and shoot video with one or two mirrorless cameras and usually a couple of action cameras and sometimes running a dedicated audio recorder and it can be quite a handful by myself. The I see you running what seems like even more cameras while doing a rocket flight checklist and countdown AND running a livestream simultaneously. That's nearly as impressive as it's going to be when this things finally lands like you want it to. I can't imagine how much work all of this is and how you don't forget to hit a record button or check a battery level there is beyond me.
Oh yeah and you're flying a drone. You are not human.
I actually stopped following your progress for some time cause I thought you was hitting a bottle neck. Then I saw you got the throttling working pretty well, then I was excited. I am ME graduate student and I really think you're very close. Nice work and I love it!
It feels like you are getting closer and closer to the dream landing!
That would look even better with rounded corners and being de-burred.
I think you did a great job
Keep up the good work Joe. Started following BPS the past month or so. The problem with software is that it takes time to code some good algorithms for computing the things it needs. Training Neural nets could be one idea, I mean for the ship to make the right decisions based on those nets.
This thing should be renamed to: BURNING DESIRE
Very inspiring indeed.
Thanks Joe!
The wiggle is mostly because of actuator placement. Put it far away from the hinge and use longer actuator arms with maybe additional linkages.
The irony if Scout 4 makes it through!
I think the people who are suggesting liquid fueled engines are severely underestimating the complexities involved in that. I also think they don't realize that using liquid propellants kinda defeats at least part of the purpose of this entire project.
I feel like making a liquid engine in the same size as a model rocket engine would be nearly (if not) impossible.
@@helloeverything496 Meh, I don't think it would be too dissimilar to an RCS thruster. I think it's possible, just way more complicated and heavy.
Liquid is too complex, yes, however, hybrid is even simpler than the current setup because it's throttleable.
Granted, he's not doing this because it is easy, he's doing this because it is hard.
That was astoundingly good! It was vertical at touch down and looks like bounce was why it fellover (it wasn't on the landing pad). You are so so close to nailing a landing.
It looks like you're so close! Keep going, you're almost there!
14:27 Stupid idea, but what if you cheat a little and add some sort of spikes on the landing legs that will hold rocket as it lands. A spike that just springs out as soon as the landing legs deploy would let you choose an appropriate length while being flush when the landing legs are retracted( when not deployed ).
Better, if the spikes were angled away from the Y-axis, so if it lands on spike first, it will push the rocket in the opposite direction.
First of all as an automatic engineer....respect to ur work chapeau....the problem of drifting is obviously the period between the first and the second burnout i think u need to add tow "2" motors very little impuls to recompensate on the pitching disturbance like the Russian missile system
Good to see you're still trying to stick a landing.
I catch these videos every now and again. But wow!!! They are awesome!!! Keep going! Almost there!!
same here ! it's amazing !
That looks like Lucerne dry lake.
Welcome to my old launch site.
Watching your progress with some fascination.
this video deserves much more than it has
Great!
Keep doing.
We all are with you.
I like you man and I have been following you for years. I just think the TVC craze has flown and landing it is an attempt to revisit the hype of when your channel blew up. Its good that you do really well at these attempts but dont give in to the hype of your channel and feel like you need to keep the audience happy. It might take 10 years but you will get there.
Incredible progress so far! So close.
One thing that might help with landing on a pad is separating land vectoring into multiple stages, for example you could have one stage aiming the rocket towards the pad, a second which focuses on stopping the rockets lateral movement, a third making the rear point at the pad, and a fourth which focuses on the soft landing. Obviously this would work easier with longer engine burns or stabilizing fins, but it's at least a start.
This continues to be awesome. Enjoy watching you improve and move onto solving the next problem. Keep it up!!
Although he's critiquing the landing & find mistakes, but that launch was real smooth!! 😍
No more bounce! Excited to see the solutions. Love the progress, love the shirt
Парень-ты гений! Тебе пора заниматься большими ракетами! Удачи тебе в твоих свершениях!
I love your vids man! Much more praise I could give but it sums up as I get super happy when I see you have posted. Thank you kind sir!
Hey Joe,
I have been watching your videos now for a while and am absolutely impressed with your tenacity. I have a suggestion that may be late to the party but may give you an easier time controlling the oscillation issues you are have on the gimble.
1. Instead of angular servos adopt a NEMA 8 or 11 stepper attached to either a gear rack and pinion or lead screw. Your stroke is roughly 5~7mm if I can gage from the video. The stepper has a holding torque to help stop the bounce but adding a lead screw will stop the bending moment and loss of leverage at the end effect. I recommend a lead screw vs gear rack due to the ability of the lead screw to add back feed counteraction. Use a fine pitch of about .5 and 8mm D. You will need to add a stepper driver of course and then in C create a simple PWM to step/direction algorithm to convert the PWM angular to steps. Not hard and there are several examples. This will add weight but not so much that you couldn't add grain size to compensate.
2. I see you try to use screws as pivot points. I highly recommend using bronze bushings and shoulder bolts giving you excellent tolerance at you pivots and an easier way to fine tune the fits as the shoulder bolt is precision ground. Use bronze flat washers and polish to fit between faces perpendicular to the pivot axis to remove lateral and bending slop within the pivot joints.
This approach is similar to how space x controls the motor gimbles on the Falcon. How I know that is NDA protected so....
I wish you luck and low winds brother.
Been a long time coming! Good work, Joe!
ooohh man, you almost there!....amazing progress mate. Congrats!
Supercool video again! I really like the detailed explanation of everything that happens during the flight. Thanks for that
Love seeing these updates man! So cool to see how this project has grown!
I love all of your projects, and scout is probably one of my favorite of the entire TH-cam ! Good luck !
Getting so close, cant wait for the next upload
I'm so excited with this project!
Problem I see is that with a finless/gimbled engine rocket, is that it is obviously unstable when the thrust stops. The rocket immediately went horizontal and at that low altitude never had a chance for the booster motor to settle in and stop the pendulum. Also the boost takes it off line from vertical and the return motor tries to correct this and get it back to the pad. Keep it up. You will get it. Exciting watching your progression.
Thanks for giving us a glimpse into an alternate reality where SpaceX decided to develop propulsive landing with solid rockets instead of liquid Merlins!
that was perfect!! no one else can land solid fuel rockets! beautiful! c'mon space community give this man a CEO!
?????????? This is his company lmao and he didn’t land one (in this video at least)
Never give up, never surrender, and maybe go faster on your control system. When I was doing control systems design (servos on very large robots in nuclear reactors) we were always shooting for 1kHz. Your slow servo rates are bound to encourage oscillation. I assume your rates are slow just because you have small processing horsepower. If they are slow because that better matches your dynamic systems that's another thing entirely, but I'm thinking your systems are small and light, so they could benefit by faster control loops. Good luck!
When using solid rocket motors for a vertical rocket landing, one of the problems is the inability to adjust the thrust of the motor.
However, it is possible to get an adjustable vertical thrust component by deflecting the motor axis from the rocket axis.
For symmetry it is necessary to take two engines and deflect them simultaneously in opposite directions.
If you deflect the engines by 60 degrees, the force along the rocket's axis will decrease by half, if 90 degrees, then to zero.
Can't wait to see you nail the first landing!!
Inspirational. Keep it up, you are close!
Impressive work Joe! Keep it up!
Thank you for the video. It's really great. Looking forward to the next. Sky and winds and that stuff to you.
After 7 years he has finally done it
I respect you and your work so much
Awesome. Next video you are going to land it sure. The road has not been easy since you started, but the end is in sight
"Cleaner than the last one." is always a step forward.
Mr. Benard, there are 4 supports only in your rocket for landing. Your idea is very proud to me. But if 1 support is unstable, the whole rocket will tilt, and fall to the opposite side and finally drop as only 1 support in the opposite side. You can investigate in very slow motion of how the rocket land in less than 1 second. As the plan area of the rocket is a circular shape, but your rocket using 4 truss supports as square shape. it is difficult to form it as a whole structure. I would suggest to modify it to have 8 supports which is a little bit similar to a circular shape in order to to improve its stability. If it won't have adequate area around the rocket perimeter to install 8 supports, try to install it as 2 layers which have 4 supports in each layers but with different orientations. Another important point is its support leg length, whether it will form as 8 supports with same radius on plan view after landing or even longer length in upper supports in rocket by forming a truss to reduce its slenderness ratio to minimize its buckling or twisting and make 8 supports to form 4 corners as inner supports and anther 4 corners as outer supports. Moreover, "1 outer support together with 2 inner supports" or "2 outer supports together with 1 inner support" can also form itself as a large triangular stable support if horizontal members or even say extended tie, spring, elastic material as rubber band, etc. are installed between them and finally forming 4 bigger triangular 3D truss supports and its stability will increase much. I am a Structural Engineer, the rocket with 4 landing supports are similar to a tall building rest on a small pad footing foundation. Such building will become more stable if bigger size of pad footing foundation could be constructed to increase the safety factor for Overturning Checking as the moment arm from one side to the opposite side is increased. Please take it into consideration.
Hard pays.. huge improvement...best of luck ...
I love your infectious excitement!
HELL YAY idea make a detachable insterment filled ball thing with an independent parachute
idk if anyone has said this yet or if you all ready use it but you could use a crumple zone in the legs to take all the energy from landing
I enjoy your videos, the learn as you go. Looking forward to those successful powered landings!
Flight 3 was a new shepherd landing burn moment
Would it be possible to add a lip under the nose cone to increase drag on descent while keeping the drag in the right spot for the accent? If that would even work with it's small size though
Awesome video, cant wait for the rest.
Thanks.