Motorcycle steering approach

[Chris Beeves]

Hey there people!
I'm something of a wallflower in here, but thought I'd share a little bit of what I am working on.

It's a steering approach on bike sims focusing on countersteering and realistic physics based feedback.
(I temporarily mapped the throttle to the rear brake pedal to be able to show what the actual input is, felt a little funny and looks a little funny )

What do you guys think, how would you steer a bike sim?

//Chris

 

[GeekyDeaks]

Love it! I'm a regular biker since I don't own a car, so I find the controls in the bike sims much more jarring than the car ones where I have very little real world experience.

I'm curious how you are managing the transition to countersteering and also if you have any kind of force feedback.

You planning on marketing this, or is it a hobby project you might want to share?

 

[Chris Beeves]

Yeah, it really is much more difficult to capture what a bike rider expects a simulator to do than a car driver. I have quite some way to go still, but as a biker myself I enjoy this concept most of any iv'e done so far.

The transition is a tricky thing. Since it's a race sim I just try not to spend so much time in that speed area.. My code brain has trouble delivering a neat solution there.
There is physics based force feedback, which you can probably spot in the video if you look closely. The handlebars follow the ingame handlebars.

No, I don't have a clear path where this will go. It is a hobby project so far, we will see what the future brings! I'd love if the bike sim HW community would start growing!

 

[GeekyDeaks]

The transition is a tricky thing. Since it's a race sim I just try not to spend so much time in that speed area.. My code brain has trouble delivering a neat solution there.

Totally pragmatic approach! This has piqued my interest though. I have a current project on the go, but I might see if I can dig out my old applied maths books and see if it would be possible to work out the gyroscopic torque and apply it in a sensible way. Which sim are you using?

There is physics based force feedback, which you can probably spot in the video if you look closely. The handlebars follow the ingame handlebars.

Oh neat. I didn't honestly know that any of the bike sims reported the forces felt at the bars!

No, I don't have a clear path where this will go. It is a hobby project so far, we will see what the future brings! I'd love if the bike sim HW community would start growing!

I'd love to see the details of your design if you are happy to share. I need something to keep me busy over the Xmas holiday!

 

[Chris Beeves]

Totally pragmatic approach! This has piqued my interest though. I have a current project on the go, but I might see if I can dig out my old applied maths books and see if it would be possible to work out the gyroscopic torque and apply it in a sensible way. Which sim are you using?



Oh neat. I didn't honestly know that any of the bike sims reported the forces felt at the bars!

I'd love to see the details of your design if you are happy to share. I need something to keep me busy over the Xmas holiday!

The sim is GPbikes. The only real bike sim with proper realtime physics. It has its quirks, but by far the best bike sim available.
It actually calculates most stuff, so it's mostly the job of turning it into good input/feedback.

I am the worst at documenting my work, so I can't say I have a lot to share other than the concept videos. I need to step my game up here though, so I'll keep you posted!

 

[stigs2cousin]

This simulator was pushed by Honda in the 2000s. (Don´t remember the exact dates.)

It was rather good for exploring dangerous situations, driving feel and physics were horrible.

With your simulator you seem "on track" with the "handle bar push".
(Sorry, I couldn´t find the right word for the german "Lenkimpuls")

I also think that slow speed physics are not that important for racers as you probably don´t want to simulate the exercises from the driving test.

(and t.b.h. racebikes suck royally below the "self stabilizing speed" with their riding position, suspension kinematics and mostly steering dampers. That´s why these days I prefer the GS-class of bikes. Which of course has nothing to do with getting old, fat and having bummed knees)

One thing I think that will add tremendously to your simulator is a tripple monitor setup.
I know for a "wrap around screen" you´ll need rather large monitors/TVs and a monster PC to drive them, but riding a bike without peripheral vision (not to mention looking into/behind the corner) would be really difficult for me.

 

[Chris Beeves]

The Honda riding trainer is another type of beast indeed! For what it is it’s a great tool. Doesn’t take much training to understand how to operate it. Not sure how I will make mine work better for beginners..

You mean the countersteering? Yeah, that’s the main idea of my steering.

I only use VR with the sim, the depth perception is something you don’t want to be without when you’ve tried it. Not a big fan of triples or ultrawides when it comes to bikes. You also need to be able to manipulate the angles, and that is much easier in a headset.

 

[stigs2cousin]

I "forgot" about VR, but it sounds logical to use that.

How do you measure "steering forces" ?

As the angle changes very little, do you use strain gauges in the handlebar?
It must be a complicated algorithm to account for angle change and the speed of change (impulse)

I rember (not ot fondly) the prolongued and heated discussions we had on this topic under colleagues.

In the end I was contend to stay on the surface of the topic
(It works, we can all do it, so s.t.f.u)

 

[Chris Beeves]

Yeah, I have experimented with strain gauges in all shapes and sizes. Right now I have single point beam type cells inside the handlebar tubes that measure torque. It works OK but I have ideas for improvement.

The devil is most certainly in the code. It has been a brain ache getting this far only having access to my very limited coding skills. Baby steps..

 

[Daniel R.]

Great stuff. Wanted to try a approach myself and this kinda motivated me again.
I’m really impressed what you built/tried so far (from your videos)!!
Go on with that.
Motorcycle motion simulation is more a challenge than for 4 wheels. And there is a need for unconventional thinking and experiments.

Can’t wait what this thread will bring up in the future.

 

[Chris Beeves]

I’ll be going on for sure! Currently working on beefing the steering up. Currently it’s MultiJetFusion printed and has some flex to it. Holds up but I want a more solid feel.

For sure it’s something different from car sims. Especially when approaching different riders.

 

[GeekyDeaks]

The devil is most certainly in the code. It has been a brain ache getting this far only having access to my very limited coding skills. Baby steps..

If you want to share the code you currently have, I'd love to take a look at it. I do a fair bit of tinkering with microcontrollers and I have plenty of half-baked projects, so don't worry about polish . I'm very curious how you are using the strain gauges and how everything maps to the inputs for GP Bikes.

This has prompted me to dig around the garage for spare bits to see what I can cobble together

 

[Chris Beeves]

The algorithm feels like something of my secret sauce. Not sure I will open source that. Haven’t even showed it to my wife (mostly because she is completely uninterested)
Maybe I can send a hex file if you decide to start building!

 

[Chris Beeves]

Working on adding load sensing to the upper platform now in order to measure rider weight transfer better than estimating it by head position. Interesting stuff!

 

[Chris Beeves]

Still haven’t gotten around to installing the load sensors in the platform, but I did have some aluminum clamps made and replaced the RC servos for a proper linear servo instead. Super solid!