Author Topic: Marvins Creations  (Read 477 times)

MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Marvins Creations
« on: November 30, 2016, 08:40:35 pm »
Twin rotor helicopter 'Gemini' version 1.0

This is my latest attempt at building a working helicopter. Previous iterations used a fully mechanical cyclic to overcome limitations in code speed. The result of this was a large assembly under each rotor and the helicopter was barely controllable when it ever managed to generate lift for long enough to get off of the ground.

This version is much more light weight and uses a software cyclic which works much better. The system to synchronise the interleaved twin rotors is also massively improved and completely prevents collisions even with asymmetrical loading on the two rotors. The result is a stable platform that replicates all the key functionality of a real helicopter.

It turns out that the tail isn't really necessary, and actually negatively impacts the performance when not flying forward. The bare V shaped rotor assembly would probably happily fly on its own.

Operation/Controls:
C/T- Throttle: This sets the speed of rotation of the rotors, and should be set to full while the helicopter is in operation. The currently set operating speed is by no means optimal, but provides a satisfactory compromise between generating enough lift and not outpacing the cyclic control code.
G/H- Collective: Alters the pitch of all of the rotor blades to increase/decrease the amount of lift generated.
Y/I- Cyclic pitch
F/P- Cyclic roll: The cyclic control individually varies the pitch of each of the rotor blades over the course of each revolution. This causes one half of the rotor to generate more lift than the other, creating torque that alters the attitude of the helicopter.
D/U- Yaw: Acts as a rudder to change the heading of the helicopter by increasing the pitch of one rotor while decreasing the pitch of the other, creating an unbalance in the torque between the two rotors.

Blueprint



And here's what makes it tick. The two big blocks are the phase shifts and sine functions to calculate the correct angle for each blade. On the left are a couple of basic controllers to keep the helicopter roughly level, and the code going to the bottom of the screen is to sync the two rotors.


Close, but no cigar. Really, I'm just impressed that it managed to lift something so huge that far without falling out of the sky.
« Last Edit: November 30, 2016, 09:51:17 pm by MarvinMan »

Sleeping_Owl

  • Member
  • **
  • Posts: 18
  • Likes: +1/-0
    • View Profile
Re: Marvins Creations
« Reply #1 on: November 30, 2016, 11:13:56 pm »
Impressive. Getting things properly synchronized is usually a problem.
To avoid problems with code speed you can increase rotor's lift - put more fins on each blade or make blades longer so that fins have more linear speed at the same angular speed.

cupid_the_conqueror

  • Global Moderator
  • Experienced Member
  • *****
  • Posts: 89
  • Likes: +0/-0
  • A filthy casual
    • View Profile
    • the truth of rawbots
Re: Marvins Creations
« Reply #2 on: December 01, 2016, 05:36:26 am »
ladies and gentlemen the marvelous marvin never fails to impress! unlike me.. god the goblinABRT was bad..
At the end of the day, love conquerors all.

z26

  • Experienced Member
  • ***
  • Posts: 108
  • Likes: +0/-0
    • View Profile
Re: Marvins Creations
« Reply #3 on: December 01, 2016, 08:23:48 am »
Wow, lifting that spherical wheel thing from pressureline... the power and control to carry something like that (that probably wants to oscillate all over the place) is impressive



ladies and gentlemen the marvelous marvin never fails to impress! unlike me.. god the goblinABRT was bad..

At least YOU did something :p unlike someone else that hasn't even entered that acl thing

tob.s

  • Member
  • **
  • Posts: 24
  • Likes: +3/-0
    • View Profile
Re: Marvins Creations
« Reply #4 on: December 02, 2016, 06:45:29 am »
Impressive, it always amaces me what people can do with maths, some hundred years ago people with my math-knowledge would have burned you for this though  ;D still not sure if theres no magic involved

MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #5 on: December 02, 2016, 08:04:03 am »
I'm not entirely sure there isn't actually some witchcraft going on, as the software cyclic outright refused to work at all last time I tried it, hence the use of a mechanical version.

Doubling the rotor speed and increasing the limit on the blade pitch a little increased the lifting capacity significantly, allowing it lift well over 3/4 of that bot while staying controllable. I did use two fins per blade on my big helicopter, but the weight of that plus the component in the cyclic linkage were probably a major contributor to the poor performance.

BTW, that particular spherical wheel bot is mine, though I wouldn't be surprised if pressureline has done a better working one. Blueprints for that and the old mechanical twin rotor are available if anyone wants them.

z26

  • Experienced Member
  • ***
  • Posts: 108
  • Likes: +0/-0
    • View Profile
Re: Marvins Creations
« Reply #6 on: December 02, 2016, 10:57:20 am »
I just saw pressureline post an image of it one time so i assumed.

Also, your bot has weird input keys.  If it works for you thats all that matters.

MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #7 on: December 02, 2016, 04:47:57 pm »
The weird keys are a result of using a  Dvorak keyboard and rawbots not liking punctuation keys.

Tried turning the helicopter into an autogyro yesterday, but the rotors barely turned.

MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #8 on: January 13, 2017, 11:31:12 pm »
Single propeller plane

For some reason propeller planes are surprisingly difficult to make work in rawbots. It is especially odd when you consider the availability of infinite power, perfect aerofoil surfaces and disproportionately low gravity. Also, the worst possible propeller designs seem to work best, hence the high thrust version fitted to the blueprint. A standard twin blade prop works too, and permits takeoffs from the ground under its own power.

Operation/Controls:
G/H- Throttle.
Y/I- Pitch.
F/P- Roll. Some trim is automatically applied.
D/U- Yaw. Release set to zero to trim out most of the remaining weird torques.

Blueprint



Single prop racing/time trials could make for an interesting challenege seeing as speed is fundamentally limited unlike trying to do anything like that with jet aircraft.

MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #9 on: March 11, 2017, 09:59:02 pm »
Propeller plane V3.1

This is a much more refined version of the original propeller plane. The structure is much stronger and more aerodynamically stable. The main landing gear is now retractable, and a steerable tail wheel has been added. Pitch control has been moved to the tailplane, and the main lifting surfaces can act as flaps, permitting low speed takeoffs. The two blade fixed pitch propeller currently fitted is fully capable of vertical and interplanetary flight and produces relatively little torque when up to speed. I also started to add a cannon, which is somewhat synchronised to the propeller, but needs some work.

Operation/Controls:
G/H- Throttle.
Y/I- Pitch.
F/P- Roll. Some trim is automatically applied.
D/U- Yaw.
O/E- Yaw trim. Mostly redundant since adding automatic yaw control.
C/T- Lift(flaps) trim.
A- Toggle flaps standard/takeoff.
L- toggle landing gear.
Q- Cannon.

Blueprint


MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #10 on: March 13, 2017, 11:07:45 pm »
Bird bot V0.5

This has taken many, many attempts to get to a state that could be called working, which really is a testament to how well Trooks flapping glider worked back in the days of old rawbots. This flapping bot is capable of vertical takeoff and generating forward thrust. Each wing is split into a front and rear section, whose phase and amplitude can be controlled ta move the balance of lift forwards and backwards and provide forward thrust. When the wings are not flapping, the twisting lets them work as control surfaces to some extent, although the pitch ends up being inverted. It's actually quite realistic how the flapping can be cut in mid air and it will continue to glide before turning the flapping back on. The bot isn't especially polished, and the roll control is kinda tricky, but I thought people would like to see it.

Operation/Controls:
G- Toggle flapping.
H- Toggle deploy tail fin. Probably not a good idea.
Y/I- Pitch.
F/P- Roll.
C/T- Adjust forward thrust. Release set to 0. Can usually be left set to full forward, and setting to reverse probably kills a lot of the lift.

Blueprint


MarvinMan

  • Automation Expert
  • Experienced Member
  • ***
  • Posts: 236
  • Likes: +4/-0
    • View Profile
Re: Marvins Creations
« Reply #11 on: April 05, 2017, 09:40:34 pm »
Engine powered car

There isn't really much point to this other that it looks cool. It's a 3 cylinder plasma cannon engine fitted with a propeller. Produces a kinda useful amount of power from 12 caps. The dream is to put an engine like this into a plane that can fly under its own power, which may be feasible with a double acting (two cannons per 'cylinder') design and more caps per cannon. For now it's just fun to crank it over by hand and watch it spin up to speed.

Operation/Controls:
C/T- Trottle. Release set to 0. Controls capacitance, but the caps don't get a chance to fully charge so this doesn't have much effect.

Blueprint


Sleeping_Owl

  • Member
  • **
  • Posts: 18
  • Likes: +1/-0
    • View Profile
Re: Marvins Creations
« Reply #12 on: April 08, 2017, 01:17:25 pm »
Looks great! It is even unreliable, as the first piston engine should be :)
I like the idea, maybe I'll build another piston engine soon. Currently I have a one cylinder engine that stops after several revolutions.