1. Do not share user accounts! Any account that is shared by another person will be blocked and closed. This means: we will close not only the account that is shared, but also the main account of the user who uses another person's account. We have the ability to detect account sharing, so please do not try to cheat the system. This action will take place on 04/18/2023. Read all forum rules.
    Dismiss Notice
  2. For downloading SimTools plugins you need a Download Package. Get it with virtual coins that you receive for forum activity or Buy Download Package - We have a zero Spam tolerance so read our forum rules first.

    Buy Now a Download Plan!
  3. Do not try to cheat our system and do not post an unnecessary amount of useless posts only to earn credits here. We have a zero spam tolerance policy and this will cause a ban of your user account. Otherwise we wish you a pleasant stay here! Read the forum rules
  4. We have a few rules which you need to read and accept before posting anything here! Following these rules will keep the forum clean and your stay pleasant. Do not follow these rules can lead to permanent exclusion from this website: Read the forum rules.
    Are you a company? Read our company rules

Thinking of finally pulling the trigger on a 2 DOF + Traction Loss Build

Discussion in 'DIY Motion Simulator Projects' started by Map63Vette, Mar 6, 2020.

  1. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    I've lurked here for a while now on and off and am finally thinking of pulling the trigger on building something myself. To tell the full story, I've been interested in a racing cockpit setup since I learned about Assetto Corsa a long time ago. However, after playing on a friend's triple monitor setup (with what I would guess was a Fanatec setup, but don't know for sure), I wasn't all that impressed. I still lacked a sense of speed and it made it tough to feel like I was really driving and hard to judge corners. Then VR happened and I bought an Oculus CV1 with the hope that it would solve that issue as well as a Thrustmaster TX/T300 wheel with their T3PA Pro pedals and TH8A shifter. That really helped with the sense of speed and the quest began to then build a cockpit to mount everything.

    The first edition was just made with scrap lumber my parents had sitting around and cost a grand total of something like $5 for the few bolts I didn't have to mount the wheel and pedals. I used a karting seat I had sitting around and eventually added some Aura bass shakers to add a little more immersion. It was likely way overbuilt from a strength standpoint and was hard to get in and out of, but worked great. I pulled the general dimensions from my car (a 2004 Viper), so it actually felt pretty natural when seated (first attachment below, shows the wrong kind of seat, but I don't have a model of my kart seat).
    It was fairly unwieldy though and hard to move around and fit through doors, so I later cut off the pedal portion and revised it where it could slide backward between the main outside rails toward the seat. I also removed some of the extra bracing as I thought it was a bit overkill.

    Cockpit V1.jpg

    Then I stumbled on the DOF Reality website and got to thinking about the design some more. My wife also complained that it was hard to get in and out of the few times she tried it, and I didn't disagree one bit, so I worked up a new design that was more of a central spine (the idea I got from the DOF Reality design). I also switched to an old office chair as a seat as it is way easier to get in and out of, more comfortable, and accepts drivers of various sizes much easier. The wheel "hoop" is a bit flimsy, but works fine for the wheel I have and otherwise feels pretty solid (next attachment).

    Cockpit V2.jpg

    I had originally wanted to build a motion rig, but I wasn't sure how motion compensation would work with the Oculus and in the earlier days it seemed to be more of an issue, so I dropped the idea. When I found the DOF reality website and the mention of motion compensation software, the thought perked up again. I also have a box of large wiper motors from industrial trucks in my basement that I acquired with the intent of using for a motion rig, so that was one hurdle and potential major expense covered. At this risk of this post being way too long, I'll continue on with my plan in the next one.
  2. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    Having found the DOF Reality website, I was really considering just going the commercial "plug and play" route, but there were a few things that kept hanging me up. Initially I wasn't sure if I wanted to go the full 6 DOF route, but the more I read on it the more I think a 2 DOF with traction loss probably made more sense for me with the VR setup. That being said, seeing as I already a handful of motors, that was another reason I was hesitant to buy "more", even though the commercial ones might be a bit better suited for it (mine are plastic gears and maybe a touch slow). I also don't know that I was a huge fan of the Arduino and MM setup. It's a bit ironic since I consider myself the tinkerer and like to have a lot of control over stuff, but the failure rate of MM cards put me off a bit and I really like the idea of JRKs better, though that added to the cost. That led to the next design I had put together, which is basically a DOF Reality clone, but using the motors I have already and maybe a tweak here and there based on some ideas I had.

    Sim Rig.jpg

    So this leads me to my main questions. I think my motors are up to the task. They are ~35 Nm stall torque and 60 rpm unloaded speed (see curve below). The two lines are for the two different speeds, so basically ignore the bottom line in each pair. One line is amp draw per torque, the other is rpm per torque. They cap out around 20 amps, so should be well within the capabilities of a JRK and it puts them around 240W each. Lever arms are going to be a bit trickier as they are a tapered splined shaft, so not as easy to make new arms and the ones already on them are pretty short (~1 inch). I might just take the existing arms off, cut off just the tapered end, and weld a longer section to that as I need the offset kink to clear hardware.

    Motor Curve.jpg

    The layout I have uses 1.5" CTC arms and gives around 7.5 degrees of front to back tilt and ~8.5 degrees of side to side with my current estimated pivot points. They are about 8" forward of the pivot point and roughly 7" to the side of the pivot. I've estimated the top frame weight with me at around 270 lbs, which should give some overhead for either a heavier driver or faster movements to return from full tilt. Do those numbers sound reasonable to people? The frame is mostly 1" x 1" x 0.085" square tube with the center spine being 1" x 3" x 0.120". The same goes for the bottom sections. I did some very simple FEA work on it and it appears to be fairly safely within the yield limit of the metal for sure and hopefully also under fatigue limit.

    Traction loss is my next big question though. My plan was to use a tapered roller bearing on the front pivot as they put up with both radial and axial loading well and I've got plenty of leftover ones laying around from various car work I've done over the years. However, I'm not really sure how much angle to shoot for here and what kind of forces I'd be dealing with. Rolling forces are going to be way lower than gravity (at least with good bearings), so I think I can get away with a lot longer CTC arm on the traction loss motor to get that angle up, but I don't really know what to target. I was aiming for 10 degrees each front and back for the main seat frame. I didn't quite get there, but from what I hear high angle isn't as important with VR. Any suggestions on a good angle target for traction loss? I could also potentially move the motor forward to get more angle with less throw, but it's 6 of one and a half dozen of the other.
  3. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    So to condense my thoughts a bit after all that rambling, I'm basically leaning the DIY route over the commercial one for a few reasons. The first one is I already have more than enough motors, so that's a fairly significant cost savings. The second is that I really prefer the idea of JRKs over the Arduino + MM setup based on what I've read. I also have what I expect will be all the necessary tools (welder, saws, etc.) to build it. I also might tailor the layout a little more to reflect my car, but that's pretty minor.

    At this point I'm hoping/estimating it will cost around $500 to finish. I got a great deal on a pair of never used JRKs for $150 and am estimating steel material costs at around $200. Add in another $50 for a pair of power supplies and I think I'm pretty well on my way. I'm planning to use some skateboard wheels for the traction loss at the rear, but need to double check the weight rating and bearing capabilities. I know I'll need another JRK at some point, but I'm not in a huge hurry to build this either, so I might just keep looking used for a lot of stuff and maybe see if I can do some digging through scrap piles to get some of the metal.
  4. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,463
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,614Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
  5. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    I had looked at SimCalc some, though I'm not sure it plays with my setup quite as friendly. It seems more built around the shoulder mount actuators as I'm not getting any roll response out of it when I try to put in my values. It does give me forces though, and I think they seem to line up with what I was expecting.

    I'm a mechanical engineer by trade, so that's another reason going the DIY route has been appealing to me. I've figured up what I think the forces and such should be, but the one thing that seems to be missing in a lot of the builds, designs, and specs I've read is a full motor curve. A stall torque and RPM really don't tell the whole story, though that's generally the only data you get on generic motors. You won't get full torque at full rpm, so I take the mm/s numbers that people throw out with a decent grain of salt. Based on my motor curve, weight, and mechanics, I think at max tilt I only need around 15 Nm to at least keep me there. Granted you need more than that to accelerate (F=ma after all), so it's a bit harder to really predict what the motor has left, but say 20 Nm is enough to accelerate me back to the resting position. The motor can only do 20 rpm at that load, so the mm/s number at full tilt is going to be different than the mm/s number at the resting position.

    So all told I think it would be interesting to see a chart of what a full simulator build can do over it's full range of motion, though I think that's a bit more complicated to produce. I guess at any one moment the motor can put out stall torque to get you moving, it would then just run out of torque as the speed rises, but if your mass is already set in motion with the initial impulse it would take less force to keep it going. Dynamics are just a tough thing to simulate and wrap your head around sometimes. Statics are much easier to calculate.

    Regarding the traction loss, wow, that's quite a bit of motion. I think it probably makes sense though. The pivot is quite a bit longer, so you've got to sweep pretty far to get a noticeable angle. I'll have to check out those builds and see how long the pivots are. 200 mm is an easy number to use, but without knowing how far from the pivot that is, it could mean any angle. I also did some reading up on skateboard wheels and bearings as that was my plan on what to do at the back. I had some initial concerns that they might not like the load that much as they typically only put up with a person's weight when rollerblading or skateboarding (same bearings), but that's over 4 wheels and 8 bearings. Turns out a single bearing is rated for around 300 lbs radial load, so I'm way less concerned about using only two skateboard wheels in the back now. I might still consider a third just for longer life, but will have to see how I can integrate them.
  6. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,463
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,614Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    Use the Sim Geometry tab in SimCalc to match the design you have in mind.
  7. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    I'll have to check that out. I might have gotten an older version because it only had the one option.
  8. Map63Vette

    Map63Vette Member

    Joined:
    May 19, 2016
    Messages:
    127
    Occupation:
    Mechanical Engineer
    Location:
    Lawrence, KS
    Balance:
    430Coins
    Ratings:
    +31 / 0 / -0
    So I just opened and ran SimCalc again and it worked, so I must have put in some number wrong or misinterpreted some measurement before. Have also tried running a simulation on some dynamic motion software (ADAMS if anyone is interested). It's easy enough to build a generic model with main linkage and pivot points, but I'm not super confident on incorporating a motor in the software. I have something in there now with the speed and torque curve I have to try to better understand the motion of the platform through the whole travel range, but I think the motor controllers themselves play a big part in it and that's not something I have any experience trying to simulate.

    So basically I can run a simulation with my motor properties from a starting static position, but it will just ramp the motors up to their max rpm and keep going. I'm not the most experienced user with the software either, but I don't think I can necessarily make a stopping point, I just limit the simulation time frame to the angle sweep on the motor arms that I'm interested in. The reality is a PID controller is going to target a set angle movement and will ramp the motors up and down accordingly to reach it and even reverse torque on them to hold the position. I think ADAMS actually has the ability to simulate control systems, but that's way over my head at this point.

    Long story short, for a random interesting number of the day, it takes my motors approximately 0.12 seconds to hit max rpm when doing a full side tilt from the static zero position. The platform starts off with 0 deg/s movement and ramps up to just under 70 deg/s. Not sure if any of this is useful or relevant, but I find it a bit interesting at least. It at least helps reinforce the fact that max speed and torque of motors really doesn't tell the whole story.