What do you MPR and HPR people think of Thrust vector control

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
yeah that would be cool! I cant wait for some 20 second long burning motors to come out that would be really nice for tvc and everyone can enjoy some part of tvc in the future!

What if we could send this as a request for AT or CTI or maybe Estes?
 
Not really, thanks. You just agreed with me that aerodynamic stability is required for a rocket to coast, which was my point. So I don't see the discord you are referring to.

I think what he's referring to is an app used for gaming, called Discord. You can use it from a web browser, your phone or your PC. It has native clients for Windows, Mac and Linux.
 
You lost me at "tpa". FWIW, it is not hard to build an adaptive control system with aero control. You can command a dither signal in roll and measure the response to estimate q, and adjust gains accordingly.
Tpa is throttle position attenuation. It reduces the pid profile(gains) based on throttle position. Most fpv race drone flight controllers set the tpa break point at about 75% throttle and reduce up to 15% on a default tune. It’s a very good way to get rid of high speed oscillations.
 
What if we could send this as a request for AT or CTI or maybe Estes?

I read apost from AT on Facebook talking about the challenges of these long burn motors. he was discussing something about the burn temps being outside of certification requirements. And I believe he was saying that he was going to propose the change with the gov body who sets the standards, as he was on the board. This was sometime last summer/fall that I read this, and I have no idea what the update is. But I believe the g8, g12 has something to do with that temp rule that excludes them.
 
There was a discussion about the same issue with temp limits on the outer casing somewhere on these forums related to the long burn motors. Maybe on the aerotech motor thread?
 
Tpa is throttle position attenuation. It reduces the pid profile(gains) based on throttle position. Most fpv race drone flight controllers set the tpa break point at about 75% throttle and reduce up to 15% on a default tune. It’s a very good way to get rid of high speed oscillations.
Thanks. Can you pint me to any good engineering analysis models for quad-copter performance and control?
 
I read apost from AT on Facebook talking about the challenges of these long burn motors. he was discussing something about the burn temps being outside of certification requirements. And I believe he was saying that he was going to propose the change with the gov body who sets the standards, as he was on the board. This was sometime last summer/fall that I read this, and I have no idea what the update is. But I believe the g8, g12 has something to do with that temp rule that excludes them.
I think AT proposed to the NFPA that they raise the allowable casing temperature to 750 Deg. F. For the G8 and similar SU motors. I have not read any further discussion or movement on the issue. The G8 is an odd motor in that it is not designed to a common spec such as 120 or 160 N-S. I suspect that it was designed to BPS specs. I would have just reduced the burn time and total impulse, untill it met the existing NFPA requirements. I also like the 120 N-s mark.
 
Thanks. Can you pint me to any good engineering analysis models for quad-copter performance and control?

I don’t have an engineering analysis examples, I’m just a diy builder/racer. But betaflight is the top open source firmware in the industry, and Joshua bardwell has a ton of videos explaining features and how too’s on betaflight. You could search YouTube for Joshua bardwell tpa or something similar and probably find some better explanations. Idk that tpa would directly work in its current form for rockets as its throttle sensitive(rockets don’t have throttle) but I don’t see why someone couldn’t modify the open source code, so that the tpa reads airspeed/gps speed/ or acceleration And base its attenuation on one of those metrics. Inav isa fully autonomous firmware based on betaflight which has full gps features. You could likely build a rocket glider with inav and have it stabilize the flight, and then fly itself back home on burnout( idk if that’s even legal) but it’s entirely possible.
 
It's been awhile since anyone has posted on this thread. Since the last post, the AT H13ST has been certified. It's 29 mm and has a 15 second burn time. Looking at the thrust curve, I'm wondering about the large initial spike within the first 1/2 second of ignition. That would create a hammer on the mount. Also, I think longer motors are not good for thrust vectoring, although the length probably comes with the longer burn. I think long motors could create a longer "lever arm" on the TVC structure.

I put a couple of the BPS Space TVC gimbals together. I have not flown one yet. When I put a CTI G33 in one for static testing, the motor failed and destroyed the mount. But, that's an entire thread in itself. CTI is working on that problem.
When putting the TVC gimbals together and just dry running them, I noted a couple things.

Keeping in mind that this thread is about MPR and HPR.
The gimbal seems to have a lot more play than I think would be good. I think it's very important to choose servos for TVC gimbals. Inexpensive plastic gear servos seem to actually have less noise than inexpensive metal gear digital servos. I picked up a couple MKS (good) digital servos to compare and they seem like they would be much better. But, they are a lot more expensive.

The typical connection using Z pattern servo rods seems to create some of the play. It makes sense, because you must feed the rod through the hole in the horn at an angle. This means the hole is not tight on the rod. I suggest rod connectors on both sides of the rod. I tried this and it removed some of the play.

Again, long motors, say over 3 grain, don't seem like they would be good for TVC gimbals. I believe that more work needs to go into using TVC vanes for control particularly for HPR rockets. Again, there's going to be a lot more stress on moving parts like a gimbal with HPR motors.

For continued stability, I suggest considering some type of modified delay grain that may be able to supply enough thrust after main motor burn out to provide coasting stability. We already know that a delay grain burning can reduce base drag. For how much "thrust" you would need in some kind of enhanced delay grain or even if it could work, I don't know.

Unless you are trying to spin stabilize your rocket, spin (roll) is the enemy of control. Can't control spin with TVC gimbals. Could do it with TVC vanes. A TVC vane system might be able to be used as a add-on to the motor retaining system.

So, my 2 cents worth. I'm planning on flying my BPS TVC gimbal rocket three times (if I can get the G33's from stop blowing up), and then I'm doing something else. I decided that I'm more interested in controlling descent than up part. ;-)
 
Looking at the thrust curve, I'm wondering about the large initial spike within the first 1/2 second of ignition. That would create a hammer on the mount.
Given this motor is only a H13 the magnitude of this spike will likely be a lot less than the thrust or thrust spike of other higher thrust motors, perhaps even in lower impulse classes (guessing, I haven't trawled the curves), so not likely to be a problem from a strength POV.

Also, I think longer motors are not good for thrust vectoring, although the length probably comes with the longer burn. I think long motors could create a longer "lever arm" on the TVC structure.
Motor will have less maximum control authority, which could cause slower response with straightening the rocket after a perturbation (eg. wind shear). Probably not a showstopper most of the time. It can also produce lower gain the plant section of the control loop so if the flyer has set the gains too high, inadvertently, so it will be less inclined to suffer less from oscillatory or unstable behaviour.

The gimbal seems to have a lot more play than I think would be good. I think it's very important to choose servos for TVC gimbals. Inexpensive plastic gear servos seem to actually have less noise than inexpensive metal gear digital servos. I picked up a couple MKS (good) digital servos to compare and they seem like they would be much better. But, they are a lot more expensive.
MKS are great servos. Slop in the linkages is probably more of a problem than servo gears I suspect. I suspect any noise (jitter) will be dealt with by the control loop reasonably well. Faster servos will perform better, of course.

For continued stability, I suggest considering some type of modified delay grain that may be able to supply enough thrust after main motor burn out to provide coasting stability. We already know that a delay grain burning can reduce base drag.
This might be a consideration on some of the more high-performance rockets people might try this on, especially if they deliberately make them not statically stable. That being said the additional mass of the system effectively dials back the speed, as does the typical use of low-thrust motors for this system, so base drag will provide almost no assistance with stability for most flights IMHO.
 
The typical connection using Z pattern servo rods seems to create some of the play. It makes sense, because you must feed the rod through the hole in the horn at an angle. This means the hole is not tight on the rod. I suggest rod connectors on both sides of the rod. I tried this and it removed some of the play.

Again, long motors, say over 3 grain, don't seem like they would be good for TVC gimbals. I believe that more work needs to go into using TVC vanes for control particularly for HPR rockets. Again, there's going to be a lot more stress on moving parts like a gimbal with HPR motors.

For continued stability, I suggest considering some type of modified delay grain that may be able to supply enough thrust after main motor burn out to provide coasting stability. We already know that a delay grain burning can reduce base drag. For how much "thrust" you would need in some kind of enhanced delay grain or even if it could work, I don't know.
My fix for reducing slop in Z-bend pushrods in my models with servos, often is to use a small drop of thin CA into the hole of the plastic servo arm, then move back and forth to loosen it to rotate freely. It worked nicely for the typical plastic servo arm, would not work quite as well for say a plywood servo arm.

Multi-grain motors are not endburners. TVC works best with long burn low thrust motors. If the rocket is going to use a multi-grain motor and fly fast, it needs to use aerodynamic control.

A low thrust slow boosting model is not going to coast more than 2-3 seconds or so. The amount of thrust needed to keep it pointed the right way would be far more than a modified time delay could produce.

Unless a person is doing a SCALE model of a vehicle without fins, a TVC rocket ought to have some fin area so it won't be aerodynamcally unstable. Joe Barnard often has small fins on his models (well, at least some flights he made this year did). That's the K.I.S.S. fix for a small problem some want to make into a bigger more complex and harder to solve issue.

Also, I've seen some finless TVC flights, slow boosting. That yes, after burnout the model tumbled. But NOT A PROBLEM, because it still coasted "up", to apogee anyway. And also, it proved how well the TVC was working, that it really did make an unstable rocket fly vertically.
 
Last edited:
Here's the curve of the H13 from thrustcurve.org
Scan_20200929.jpg

The spike at the start would be great for aerodynamically stable rockets, not so much for something like a TVC gimbaled rocket. The 29 mm gimbal from BPS isn't that strong. Did you see the video from BPS Space's flight using this motor? Didn't go too well.
I'm curious what experiences others will have with these motors.

It can also produce lower gain the plant section of the control loop so if the flyer has set the gains too high, inadvertently, so it will be less inclined to suffer less from oscillatory or unstable behaviour.

Sorry, since I don't program for TVC gimbals, I don't see what you mean here. Perhaps it's the wording?

I wanted to get good video of one of the BPS Space gimbals on a static test stand using the CTI G33 motor. However, the motor failed almost immediately and destroyed the GoPro camera. I wanted to see the movement with the servos on but not moving (stalled). Hopefully, I'll see this someday. Anyway, I think something should be done about the play I'm seeing in these gimbals. I still think Z bends in push rods even with added glue are not good for this. And, choose your servos carefully. Plastic gears might be OK if they are strong enough for the application.

I wasn't too clear about what I meant for a modified delay grain. What I meant was a delay grain lighting really what would be a small low thrust motor. If this is even possible it would take some development. Just talking here, because I'm not going to pursue that.

I think that TVC gimbal rockets have a very limited potential for HPR solid motors. TVC vanes would have a lot more potential because you could use a lot more motors, and larger motors.
Can you imagine the size of the gimbal and rocket airframe for a 54 mm long burn gimbaled rocket? I've witnessed a 38 mm J impulse motor gimbaled rocket. It was a fairly long burn commercial motor, and it was a lot of sky writing. And, I believe the 3D printed gimbal ended up failing from the heat.

TVC gimbaled rockets are cool and I appreciate the thought and work necessary to get one to work well. But, the question was "what do you MPR and HPR people think of Thrust vector control" and this is what I think. Also, I'm assuming this was about TVC gimbaled mounts and not others. I don't think I've seen a thrust vane system on this forum and wonder how different the control system would be.
 
Here's the curve of the H13 from thrustcurve.org
View attachment 433327

The spike at the start would be great for aerodynamically stable rockets, not so much for something like a TVC gimbaled rocket. The 29 mm gimbal from BPS isn't that strong. Did you see the video from BPS Space's flight using this motor? Didn't go too well.

Didn't have time to check out other BPS flights, but here's Sprint flight 10 from 2 months ago on an H13.



It took off GREAT on the H13. That spike is incredibly brief, less than 1/2 second. It helps to get the model moving. Ideally for a TVC model it'd be nice to have less of a spike for a really slow liftoff. But in any case, the TVC mount held up fine to the spike, and the control system kept it under control for launch. So it launched NICELY So why do you say "it didn't go too well", as you fixated over the spike?

Now, it did get moving faster and faster. And began a small wobble which got worse the faster it got. Until finally it lost control. That's part of the problem with doing a finless rocket with a relatively small moment of inertia and a control system that can't be fast enough to stay on top of it. I mean let's say the control system could be upgraded to keep it under control at 200 mph, would it be expected to keep it under control at 400 mph? 4,000 mph? The aerodynamic loads increased with the square of the velocity, so for example flying at 400 mph has 8 times more aerodynamic forces than flying at 200 mph.

Which again loops back to what I said about adding some small fins to help keep it stable enough without being over-stable.

Keep in mind here that Joe was going for ALTITUDE. This model flew far faster on the H13 than anything else he's flown. So basically it flew faster than the control system could keep up with for a FINLESS model (either mechanically, or the programing as-tuned, or sensor rate, or whatever). So, it;'s not a "thrust spike" issue. If he added enough ballast (even better, noseweight), to make it fly slower, it would have been able to avoid going out of control near the end like that (or if it was a programming issue he may have bene able ot solve it that way.) Or, rather than adding weight, if he'd added just enough fin area to make it slightly stable, that likely would have fixed it.

But see how easily you totally dismissed the H13 for TVC, when it's a damned near perfect motor for TVC above G power? (To me, the best motor for TVC is the G12 reload, closely followed by the F10).
 
Last edited:
It took off GREAT on the H13. That spike is incredibly brief, less than 1/2 second. It helps to get the model moving. Ideally for a TVC model it'd be nice to have less of a spike for a really slow liftoff. But in any case, the TVC mount held up fine to the spike, and the control system kept it under control for launch. So it launched NICELY So why do you say "it didn't go too well", as you fixated over the spike?
Didn't go to well... at the end of the flight. I was watching when he launched this. I don't think he ever elaborated later about what happened. He has just moved on to something else. I appreciate your thoughts on what might have occurred George. It's good to hear the perspective of someone that has done this. And, I am not fixated on the spike or dismissing the H13 for TVC gimbal. I did say, there is a spike, I don't know what the spike is caused by, the spike would not be something particularly good for a TVC gimbal, and I've built two of BPS gimbals and evaluated them for strength. That's all, and I think you are coming down a bit hard on me and I'm wondering why.

The topic is MPR and HPR and I've heard different options on MPR but I think a H-impulse is MPR. F's and G's are low power.

For the H13, I did say...
I'm curious what experiences others will have with these motors.
The spike is one thing, the thrust and burn time of this motor is another. It's a MPR motor in a rocket (that has small fins btw) that launched great and did sky writing at the end of the burn.

I think that TVC gimbal rockets have a very limited potential for HPR solid motors. TVC vanes would have a lot more potential because you could use a lot more motors, and larger motors.

Well, I'm going to stop now, actually bummed I posted in the first place. What a waste of my time.
 
Sorry, since I don't program for TVC gimbals, I don't see what you mean here. Perhaps it's the wording?
Control system theory. Basically if the entire system, including the things we move to control it (plant) have too much gain it will oscillate. Think of a rudder that can only go hard left or hard right as an example, based on a small amount of airframe movement. If the gain is reduced, either in the plant or the electronics, it becomes less oscillatory and can behave nicely. I think the TVC system has programmable gains at some level. If you double the length of the motor for the same sideways servo movement the gain is reduced by approximately half for that part of the plant.
https://en.wikipedia.org/wiki/Control_system#Proportional_controlCheck out the graph here. Overdamped is what you get with low gains. As gain increases so does control performance, until it is too high and the results go oscillatory.

Well, I'm going to stop now, actually bummed I posted in the first place. What a waste of my time.
Don't think you have wasted your time. Asking questions and putting things out there are a way for everyone to learn and gets people thinking.
 
Last edited:
Back
Top