Register To Comment
Page 1 of 9 123 ... LastLast
Results 1 to 10 of 90

Thread: New HW design - Omni Wheels and 30kW spinner

  1. #1
    TechAUmNu's Avatar
    Roboteer

    Status
    Offline
    Name
    Euan Mutch
    Robots
    Triforce(HW) (in progress) A450S (ESC) (in progress)
    Location
    Edinburgh, UK
    Posts
    175
    Post Thanks / Like
    Hi All,

    First post!

    I have been working on the design of my team's first heavyweight for the next season of Robot Wars (assuming it happens) and would like to see what people think of the wheels we are going to be using.

    We are a group of final year MEng computing and electronics students from Heriot-Watt University.

    To give some background on the robot.
    • 70cm 3000-6000rpm 15-25kg ring spinner attached using 170*25mm ball bearings and powered by 3* 150cc rotomax on 14S (friction drive to avoid shock and esc timing issues) (yes I know that is insane, but we want it to spin up in under 2s)
    • 3 Omni-Wheels with 50cc rotomax on 12S (not decided gearing yet)
    • 10mm 6082T6 aluminium shell with 5mm HDPE covering
    • 25Ah of LiPo
    • custom board between reciever and escs to handle orientation (Its quite hard to drive a 3 wheeled robot, especially upside down!)


    I have been trying to design the wheels so that if they get hit by an axe then they won't break, but its quite hard when there are 12 Axles per wheel!

    The wheel itself has 10mm silver steel and 4 bearings in each of the omni-rollers, with some nice grippy rubber to avoid the traction problem that sweeney todd had.
    It is made from 4 25mm ali plates bolted together with 12 M8*40 bolts. The rollers are fitted by removing the bolts and just slot into place making them easy to replace when bent.
    The main wheel axle is 25mm silver steel.

    wzTnkEl.jpg

    Any ideas for how to improve it? or should I post more nice renderings
    Last edited by TechAUmNu; 26th August 2016 at 10:18.

  2. #2
    Roboteer

    Status
    Offline
    Name
    David
    Location
    Blackpool
    Posts
    1,651
    Post Thanks / Like
    15-25kg isn't a weight category that currently fight

  3. #3
    typhoon_driver's Avatar
    Member

    Status
    Offline
    Name
    Gary Cairns
    Robots
    Hornet 2 - retired Boner - retired Smidsy - Sold 540 - Sold PP3D - Active
    Location
    Aberdeenshire
    Posts
    4,372
    Post Thanks / Like
    I'm assuming the spinner is 15 to 25kg.

    Nice idea but a few things that jump out at me as issues,

    - Ring spinners are a nightmare as your bearing supports have to be incredibly well machined (see MR speed square) to take the knocks. Not sure I understand what you mean by 170mm by 25mm ball bearings?

    - You will not get anywhere near a 2s spin up with friction drive. It's a method that is limited to the small contact area between the wheel and ring. Whereas belts and chains have contact right around the pulleys and cogs and transfer it far more efficiently.

    - What grade of ali?

    - Omni wheels are a nightmare as well. Each one of those pins has to contend with potential forces that will bend 20mm axles. Tried a couple of custom made ones in typhoon 2. They were binned after the qualifying match for series 7.

    Honestly, I'd junk the omni wheels and just go for 3 standard wheels with big axles set at 120 deg then use one of these to control it,

    http://www.robotmarketplace.com/products/0-OMX4.html

    No point re-inventing the wheel when it's easy enough to order up. The arena floor is slippy enough that 3 regular wheels should do the job fine. Driving with brushless is still very experimental. Here is where dragons lie.

    Oh and I think you will struggle to fit it all in a 70cm diameter shell. Typhoon 2 was 1 metre in diameter and it was tight on space. At 70cm diameter and 15 to 25kg I suspect you are only going for around 15mm thick steel? Might want to up that to 20mm plus at least.

  4. #4
    typhoon_driver's Avatar
    Member

    Status
    Offline
    Name
    Gary Cairns
    Robots
    Hornet 2 - retired Boner - retired Smidsy - Sold 540 - Sold PP3D - Active
    Location
    Aberdeenshire
    Posts
    4,372
    Post Thanks / Like
    Just realised I linked to the 4 omni wheel version of that controller. You want this one instead

    http://www.robotmarketplace.com/products/RL-OMX3.html

  5. #5
    TechAUmNu's Avatar
    Roboteer

    Status
    Offline
    Name
    Euan Mutch
    Robots
    Triforce(HW) (in progress) A450S (ESC) (in progress)
    Location
    Edinburgh, UK
    Posts
    175
    Post Thanks / Like
    Yes the spinner is 15-25kg, not the robot!

    Thanks for the great advice Gary!

    I am aware that ring spinners are a pain to get right. As shown by mr speed squared bearings are really not the way to go since the ring wants to deflect and will exert a huge point load on a single bearing at the point of impact. To solve this problem we are instead effectively making the entire spinner a giant bearing. (image to follow, currently rendering...)

    We did also plan on using machining a gear into the spinner so we could use something a bit more efficient, and then use something to handle the shock of impacts.

    All ali is 6082T6

    OK it is clearly a lot more slippy than I thought!



    The ring is made up from 3 rings with 9 sections each. The top and bottom sections are ali with the grove for the balls and the middle has 2 hardox blades. (Better shown in the picture to follow in next post!)


    Another thing, is it better to go with 10mm ali + 5mm HDPE or 10mm ali + 3.2mm hardox or some other combination?

  6. #6
    TechAUmNu's Avatar
    Roboteer

    Status
    Offline
    Name
    Euan Mutch
    Robots
    Triforce(HW) (in progress) A450S (ESC) (in progress)
    Location
    Edinburgh, UK
    Posts
    175
    Post Thanks / Like
    This is the render of it as current. The ring is a bit too thick currently.
    gallery.0.jpg

  7. #7
    typhoon_driver's Avatar
    Member

    Status
    Offline
    Name
    Gary Cairns
    Robots
    Hornet 2 - retired Boner - retired Smidsy - Sold 540 - Sold PP3D - Active
    Location
    Aberdeenshire
    Posts
    4,372
    Post Thanks / Like
    Nice render.

    Ringmaster tried to turn their spinner into a giant bearing. It didn't work out so well. One hit and it generally deforms everything enough that the bearing jams up. I could see this being an issue with a giant aluminium bearing as well. The only ring spinner I've seen that can take knocks and keep running is warrior from Team Whyachi in the US. Even that has trouble at times. Your bearing races should really be steel.

    I also think you are gearing it to go far too fast. Granted I don't know the ins and outs of the design but a very quick and nasty calculation using the team cosmos kinetic energy calculator puts your kinetic energy above 100kJ at 3000rpm and over 500kJ at 6000rpm. These numbers are insane (for reference these are larger than carbide and tombstone). All this energy has to come from somewhere and your motors will only delivery so much in a short period of time. At 6000rpm the current draw on the motors would be enormous as the disc fights with the air resistance at that speed.

    At 3000rpm your cutter speed will be approx 245mph with a 70cm dia disc. At 6000rpm your cutter speed will be approx 491mph.

    You also have to consider cutter bite with a spinner. This is essentially the time taken between one tooth leaving a position and the next taking up the same position with how much forward movement the robot has had during this time. I've posted about it in the past but essentially working out the length of time at a given rpm for one tooth to rotate and another take up the position. Then I tend to use this time with a forward robot verlocity of 1m/s (conservative but realistic at times) and calculate the distance the robot will have travelled. This is your cutter bite. It is independent of disc radius and is only dependent on forward motion and rpm. Bigger cutter bite is good as it means more of your tooth can get into your opponent and transfer energy. Too little and you end up grinding away on your opponent.

    Your plans are ambitious which is good but I'd bring it back just a little bit and slow down the disc a bit. No need to go so fast as you will in all likelihood either never reach those speeds or reach them once, hit something and then never work again.

    As to the question regarding ali etc I would recommend going with the hardox over the HDPE for a heavyweight

  8. #8
    adamclark's Avatar
    Member

    Status
    Offline
    Name
    adam clark
    Robots
    VOA, Armadillo, Corporal Punishment, CP2, Mangulator, Twister, 259, 259MW, Wowot, The Skinner
    Posts
    95
    Post Thanks / Like
    Mechanum wheels will slide around like a air hockey puck, I doubt you'll achieve any grip so negating anything powerful driving them. Instead of matching a gear, why not simulate a chain. Easy I would imagine with CaD. Lots of vertical holes, drop in silver steel pins and put chain sprockets on your motors.

  9. #9
    TechAUmNu's Avatar
    Roboteer

    Status
    Offline
    Name
    Euan Mutch
    Robots
    Triforce(HW) (in progress) A450S (ESC) (in progress)
    Location
    Edinburgh, UK
    Posts
    175
    Post Thanks / Like
    Yeah, I was thinking along the same lines, that's why the ring has to be so thick.
    I was planning on adding some steel for the bearings, clearly hardened steel vs ali is not a contest!

    I might have been over exaggerating a bit. I think 1500 rpm is a bit more realistic as a normal speed and be able to run it up to 2000rpm if we need the extra power.

    We did make a spreadsheet with some crude way of working out the stats for other spinners:
    https://docs.google.com/spreadsheets...it?usp=sharing
    Although it is based on energy calculations not torque so isn't very accurate.

    It seems that most people put lots of weight on their discs and forget that the energy mostly comes from speed. Hence why we want a lighter ring. I know that bite is important too, which is why we want such a powerful drive system. However I think that having too much bite can be your worst enemy, since you put a lot of the energy back into your own robot (which would be bad since as you said it would probably deform the ring).

    The main strategy is going to be dodging them as they try to ram/flip/hit us and then hit them in a weak spot.
    Last edited by TechAUmNu; 26th August 2016 at 13:10.

  10. #10
    TechAUmNu's Avatar
    Roboteer

    Status
    Offline
    Name
    Euan Mutch
    Robots
    Triforce(HW) (in progress) A450S (ESC) (in progress)
    Location
    Edinburgh, UK
    Posts
    175
    Post Thanks / Like
    Quote Originally Posted by adamclark View Post
    Mechanum wheels will slide around like a air hockey puck, I doubt you'll achieve any grip so negating anything powerful driving them. Instead of matching a gear, why not simulate a chain. Easy I would imagine with CaD. Lots of vertical holes, drop in silver steel pins and put chain sprockets on your motors.
    I don't get how that is simpler than just machining a gear on the inner surface of the ring? Since that would require some really wacky way to get the pins to stick out. I might be misunderstanding.

Register To Comment

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •