Roll & Attitude control

Discussions on high/extreme altitude and mach busting rockets.

Moderator: Moderators

Lamp
Aged like a good scotch
Aged like a good scotch
Posts: 1103
Joined: Fri May 08, 2009 9:28 am
Location: Melbourne

Re: Roll & Attitude control

Postby Lamp » Wed Jul 29, 2015 10:07 am

jase wrote:Nic, awesome man.

Be cool if you could use four internal reaction wheels to control attitude instead of the fins...

Or not, I dunno - nice work anyway 8)

Nearly ready for flight testing?


+1
Sometimes Nic you make me feel like I'm a caveman just chucking spears compared to what you do... ;-)
‘It takes sixty-five thousand errors before you are qualified to make a rocket.’ — Werhner von Braun
TRA 12286 L3
TAP member

User avatar
cryoscum
STOP... Hammer Time The Prodigy
STOP... Hammer Time    The Prodigy
Posts: 1802
Joined: Wed Oct 06, 2010 2:54 pm
Location: QLD, NSW & VIC

Re: Roll & Attitude control

Postby cryoscum » Wed Jul 29, 2015 1:32 pm

Lamp wrote:+1
Sometimes Nic you make me feel like I'm a caveman just chucking spears compared to what you do... ;-)


Mate, you certainly may look like a caveman, but your work is clearly post-industrial in nature :)
1.jpg
1.jpg (19.95 KiB) Viewed 2024 times
AMRS L3 | NAR L3 | QRS 089 | MDRA 224
AMRS Technical Advisory Group

Total impulse for 2016: 32,458 Ns (thus far)
Total impulse for 2015: 84,231 Ns
Total impulse for 2014: 40,757 Ns
Total impulse for 2013: 62,927 Ns

User avatar
jase
It runs on moonshine
It runs on moonshine
Posts: 1217
Joined: Wed May 11, 2011 11:06 pm
Location: 43° South

Re: Roll & Attitude control

Postby jase » Thu Jul 30, 2015 9:02 am

The link to the wind tunnel testing was interesting Nic.

http://ntrs.nasa.gov/archive/nasa/casi. ... 019688.pdf

It mentions that some of the problems you identified earlier can be overcome by addressing the span of the tail fins - who would have thought!

cryoscum wrote:In simplified terms:
- Canard fins cause vortices that affects the lift generated by the main fins at the rear of the rocket
- Usually the flow of air is disturbed in such a way that any roll force produced by the canard causes an additional & opposite roll force in the main fin behind it. This reduces the perceived effect of the roll control. The size of the opposing forces stay proportionate with an increase in canard AOA, hence you see an increased roll force with greater canard AoA, but nowhere near what you’d expect if there were no main fins.
- If the leading edge sweep angle of the canard is below the fin’s mach line, you may get more turbulence and even multiple vortices, making the problem worse
- Big aspect ratio fins also make it worse
- This causes the need for seemingly HUGE canard AoA to gain good roll control, in turn adding drag in large numbers and straining the servos
- Some missile systems have just given up on using canards alone for roll control, e.g. AIM9 that rather uses rollerons augmented by the canards for roll control.


Have a fin span ratio of 0.75 (tail compared to canards) or less is good, apparently :wink:

Your system would weigh a bit too? Having it all forward will no doubt help.
Me and my rockets run on Moonshine.

User avatar
cryoscum
STOP... Hammer Time The Prodigy
STOP... Hammer Time    The Prodigy
Posts: 1802
Joined: Wed Oct 06, 2010 2:54 pm
Location: QLD, NSW & VIC

Re: Roll & Attitude control

Postby cryoscum » Thu Jul 30, 2015 9:21 am

Very right, Jase. The problem we have with our hobby rockets is that we don't have a heap of high explosive in the nose that acts as great ballast, coupled with military grade hardware/electronics. For us to significantly mess with the tail fins (i.e. making them smaller to avoid interference with the canards) is likely to be fatal when doing high Mach flights.
So the correct answer here is, once again, compromise after compromise with no clear right or wrong to be had... Who would have ever thought THAT!?
AMRS L3 | NAR L3 | QRS 089 | MDRA 224
AMRS Technical Advisory Group

Total impulse for 2016: 32,458 Ns (thus far)
Total impulse for 2015: 84,231 Ns
Total impulse for 2014: 40,757 Ns
Total impulse for 2013: 62,927 Ns

User avatar
jase
It runs on moonshine
It runs on moonshine
Posts: 1217
Joined: Wed May 11, 2011 11:06 pm
Location: 43° South

Re: Roll & Attitude control

Postby jase » Thu Jul 30, 2015 9:35 am

Yeah mate, no explosives in model rockets (think I did read that somewhere) BUT, you do have some really cool electronics and well designed other stuff adding ballast fwd, right?

No one is saying make the tail fins smaller, just the span ratio, I suppose that is why the tail fins on the AIM9 are longer not wider...

...anyway, I think what you are doing is awesome and I'd love to see it work :D

[edit - the testing in the article I posted above negates the need for rollerons?]
Me and my rockets run on Moonshine.

User avatar
jase
It runs on moonshine
It runs on moonshine
Posts: 1217
Joined: Wed May 11, 2011 11:06 pm
Location: 43° South

Re: Roll & Attitude control

Postby jase » Wed Aug 19, 2015 2:12 pm

Found some test footage along these lines - very cool.



The guy who has posted the footage (Jiminaus50) has 5 test flights, check out his channel for more.

Bloody love rocketry, cool factor out of this world! 8)
Me and my rockets run on Moonshine.

User avatar
cryoscum
STOP... Hammer Time The Prodigy
STOP... Hammer Time    The Prodigy
Posts: 1802
Joined: Wed Oct 06, 2010 2:54 pm
Location: QLD, NSW & VIC

Re: Roll & Attitude control

Postby cryoscum » Thu Aug 20, 2015 8:11 am

It's Jim Jarvis ("aus" refers to Austin Texas). Nice guy.
Last I heard they ran into oscillation problems at higher speeds due to using off-the-shelf hobby stabilisation electronics. Exactly what we're trying to avoid...
AMRS L3 | NAR L3 | QRS 089 | MDRA 224
AMRS Technical Advisory Group

Total impulse for 2016: 32,458 Ns (thus far)
Total impulse for 2015: 84,231 Ns
Total impulse for 2014: 40,757 Ns
Total impulse for 2013: 62,927 Ns

User avatar
OverTheTop
It's only money...
It's only money...
Posts: 2658
Joined: Sat Jul 21, 2007 2:20 pm
Location: Melbourne, Australia

Re: Roll & Attitude control

Postby OverTheTop » Thu Aug 20, 2015 8:26 am

It will be interesting to see how my version of this works, since I am deliberately trying to use off-the-shelf parts :D

I am just working on the servo head and fins in 3D CAD currently. Almost ready to hit the "print" button.
TRA #13430
L3
"Everybody's simulation model is guilty until proven innocent" (Thomas H. Lawrence 1994)

strud
Engineer
Engineer
Posts: 917
Joined: Thu Oct 02, 2008 7:37 am
Location: Central Tablelands, NSW

Re: Roll & Attitude control

Postby strud » Thu Aug 20, 2015 5:29 pm

As Cryoscum has stated and I have previously, if you don't have appropriate gains for the individual control loops for the given instantaneous conditions (ie velocity, air density etc) then the control performance will not be consistent and likely disappointing.

User avatar
OverTheTop
It's only money...
It's only money...
Posts: 2658
Joined: Sat Jul 21, 2007 2:20 pm
Location: Melbourne, Australia

Re: Roll & Attitude control

Postby OverTheTop » Thu Aug 20, 2015 5:48 pm

I don't expect earth-shattering flys-like-its-on-rails performance. Just keep it upright as best as possible. Gentle guidance will stop it going oscillatory and probably conserve some of the kinetic energy along the way.

If it is one thing I have learned about control systems over the years, if you keep the response slow enough and keep the transport delay down you can get some pretty good outcomes. Not fast performance, but good, solid accuracy.

My input specification is to use "plug and play" systems. That is the challenge I have set myself. So that is what I'm trying to work with. I hope your system with scheduled gains works well. I don't have the resources at the moment to pour into such a project.

Wish me luck :wink:
TRA #13430
L3
"Everybody's simulation model is guilty until proven innocent" (Thomas H. Lawrence 1994)

vance2loud
Rocket Crew
Rocket Crew
Posts: 283
Joined: Wed May 14, 2014 12:44 pm
Location: Devonport, Tasmania

Re: Roll & Attitude control

Postby vance2loud » Thu Aug 20, 2015 7:02 pm

Would using a good accelerometer mixed to reduce control rate at higher velocity help.
A Raven altimeter's TX port could be useful if you can separate the barometer and accelerometer signal from it.
Just trying to think of off the shelf devices with a high G accelerometer and an output signal.
TRA #15165 AMRS #50
I do it because i can, I can because I want to, I want to because you said I couldn't. - Unknown
I have not failed. I've just found 10,000 ways that won't work.
- Thomas Alva Edison

strud
Engineer
Engineer
Posts: 917
Joined: Thu Oct 02, 2008 7:37 am
Location: Central Tablelands, NSW

Re: Roll & Attitude control

Postby strud » Thu Aug 20, 2015 8:26 pm

OTT, I was commenting more on the performance of the system in the video.

Maybe a good approach would be to simply return the surfaces to a null position if velocity gets above the point where the control loop would be unstable.
Hopefully by the time you reach silly speeds the trajectory error from launch could be righted......

Your point about transport delays is a very good one, an indeed even with an ideal gain schedule for high speeds, the latency in the servos is likely to cause issues. This can be modelled pretty easily, but it is well known that actuator latency is often the limiting factor for systems with high dynamics.

User avatar
SpaceManMat
Astronaut
Astronaut
Posts: 2094
Joined: Mon Dec 15, 2008 9:56 pm
Location: Brisbane

Re: Roll & Attitude control

Postby SpaceManMat » Fri Aug 21, 2015 9:40 am

strud wrote:OTT, I was commenting more on the performance of the system in the video.

Maybe a good approach would be to simply return the surfaces to a null position if velocity gets above the point where the control loop would be unstable.
Hopefully by the time you reach silly speeds the trajectory error from launch could be righted......

Your point about transport delays is a very good one, an indeed even with an ideal gain schedule for high speeds, the latency in the servos is likely to cause issues. This can be modelled pretty easily, but it is well known that actuator latency is often the limiting factor for systems with high dynamics.


I think that the most important time to have control is at very slow speeds ie just after getting off the rail, this is where the flight trajectory is set. You need to get the rocket upright as soon as possible after launch. The longer it is off the more energy is lost to adding horizontal velocity rather than vertical and the more energy is required to correct it. I think that once it does get moving that for the most part you should lock out adjustments. Otherwise you will be reacting to noise and minor disturbances.
QRS: 124
AMRS: 32 L2 RSO
Highest Altitude: 13,647 feet
Fastest Flight: Mach 1.55
Largest Motor: CTI 1115J530 IM
Current Project: X Wing

User avatar
OverTheTop
It's only money...
It's only money...
Posts: 2658
Joined: Sat Jul 21, 2007 2:20 pm
Location: Melbourne, Australia

Re: Roll & Attitude control

Postby OverTheTop » Fri Aug 21, 2015 9:56 am

I think that the most important time to have control is at very slow speeds

The problem with that is that when using small control surfaces they are fairly ineffective at lower velocities. I have seen one set of tests that showed little effect under 250fps. That is partly dependent on the throw of the system at the time but in line with my expectations. If you make the fins bigger, static stability is compromised and dynamic control is more twitchy and harder to control.

As to noise and minor movements when traveling fast, these can be dealt with appropriately by the control loop(s). That is what they are there for. Remember, system noise (externally and internally generated) is part of the input variables too, and needs to be dealt with in the system design. :D

I guess controlling a motor gimbal would be a way to get good low speed control of the airframe, but you pay that weight penalty for the rest of the flight after burnout.
TRA #13430
L3
"Everybody's simulation model is guilty until proven innocent" (Thomas H. Lawrence 1994)

User avatar
jase
It runs on moonshine
It runs on moonshine
Posts: 1217
Joined: Wed May 11, 2011 11:06 pm
Location: 43° South

Re: Roll & Attitude control

Postby jase » Fri Aug 21, 2015 1:58 pm

strud wrote:Maybe a good approach would be to simply return the surfaces to a null position if velocity gets above the point where the control loop would be unstable.
Hopefully by the time you reach silly speeds the trajectory error from launch could be righted......


Loving this thread 8) Trying to get my head around what would happen, Strud this is pretty much it isn't it.

All the work would be done in the first couple of seconds of flight, then by the time you are travelling at 'Nic' type velocities any corrections would be minute - or, as you've suggested, the system just goes quiet because all the stabilization work is done.

Hey Stewart, just curious, are you using the Eagle Tree gear or have you found something else?
Me and my rockets run on Moonshine.


Return to “Higher, Faster, Further”

Who is online

Users browsing this forum: No registered users and 4 guests