• Welcome to PiBoSo Official Forum. Please login or sign up.
 

Full motion simulator

Started by PiBoSo, November 19, 2014, 05:11:00 PM

Previous topic - Next topic

HornetMaX

Quote from: Furious on November 19, 2014, 08:42:41 PM
Where is the usage of gravity to simulate acceleration?
You'll never be able to simulate acceleration unless, of course, you accelerate for real.

As already stated multiple times, I'd be already pretty happy with:

  • Bike frame with usual commands for throttle, brakes, clutch.
  • Handlebars controlling the GPB steering (with our without force feedback). Direct steer could be configured, but I'd tend to think usual steer (target ean angle) with reverse input (nadlebar left to turn right) would be OK.
  • Bike frame lean angle dictated by GPB (i.e. not dictated by the player).
  • Bike frame pitch not necessary (if present, only passive: GB dictating the bike pitch).

Having something that properly simulates the lean of the bike would be a hell of a nightmare.

MaX.

matty0l215

November 19, 2014, 10:43:59 PM #16 Last Edit: November 19, 2014, 10:53:15 PM by matty0l215
Quote from: HornetMaX on November 19, 2014, 10:12:06 PM
You'll never be able to simulate acceleration unless, of course, you accelerate for real.

That would be the hardest thing to simulate

Quote from: HornetMaX on November 19, 2014, 10:12:06 PM

  • Bike frame with usual commands for throttle, brakes, clutch.
  • Handlebars controlling the GPB steering (with our without force feedback). Direct steer could be configured, but I'd tend to think usual steer (target ean angle) with reverse input (nadlebar left to turn right) would be OK.
  • Bike frame lean angle dictated by GPB (i.e. not dictated by the player).

Already doable or have been done just not very refined.

Quote from: HornetMaX on November 19, 2014, 10:12:06 PM

  • Bike frame pitch not necessary (if present, only passive: GB dictating the bike pitch).

Do you mean like wheelieing (If that's even a word :P), if so also very achievable.

Quote from: HornetMaX on November 19, 2014, 10:12:06 PM
Having something that properly simulates the lean of the bike would be a hell of a nightmare.

Air rams would be your best bet. Or a very elaborate gear mechanism through the back of the bike, would have to be strong to hold a bike and a person moving about on a single pivot.

The real problem would be the cost, it would all be bloody expensive.

Edit- Have a look at this maybe http://m.instructables.com/id/Arduino-Pneumatic-Flight-Simulator/?ALLSTEPS
For faster responses, please visit the discord server- HERE

Furious




thats what me and my friends are coming up with.
Oculus rift.
Biker possition tracking.
Acceleration simulation ( upt to 80 % of the real bike acceleration)
corner simulation ( leaning in different direction than turn)
handlebars feedback
all inputs
etc.

also the part with servos drivers and plugin for GP Bikes already done.

HornetMaX

November 19, 2014, 11:29:12 PM #18 Last Edit: November 19, 2014, 11:30:53 PM by HornetMaX
Quote from: matty0l215 on November 19, 2014, 10:43:59 PM
Quote from: HornetMaX on November 19, 2014, 10:12:06 PM
Having something that properly simulates the lean of the bike would be a hell of a nightmare.

Air rams would be your best bet. Or a very elaborate gear mechanism through the back of the bike, would have to be strong to hold a bike and a person moving about on a single pivot.

The real problem would be the cost, it would all be bloody expensive.
The real problem is not cost, is that we don't know how to combine a lean input and a steering input.
It has nothing to do with the mechanics of the controller, it's a conceptual problem.

Quote from: matty0l215 on November 19, 2014, 10:43:59 PM
Edit- Have a look at this maybe http://m.instructables.com/id/Arduino-Pneumatic-Flight-Simulator/?ALLSTEPS
Smart !

Quote from: Furious on November 19, 2014, 11:18:19 PM
thats what me and my friends are coming up with.
Pretty cool but ...

Quote from: Furious on November 19, 2014, 11:18:19 PM
Acceleration simulation ( upt to 80 % of the real bike acceleration)
If the bike in GPB accelerates forward at 1g, how do you simulate that ?!?!  Even at 0.8g ... :o
Same for a lateral acceleration. There's no way you can create that ...

Quote from: Furious on November 19, 2014, 11:18:19 PM
corner simulation ( leaning in different direction than turn)
handlebars feedback
all inputs
etc.

also the part with servos drivers and plugin for GP Bikes already done.
Which axes on your rig are inputs to GPB and which ones are only output ones (i.e. GPB dictates the value) ?

For example, your frame pitch (wheeling) is dictated by GPB: it's an output axis.

Now what about your lean axis and steering axis ?

MaX.

Furious

Pitch axis and roll axis are dictated by GP Bikes however not the actual orientation of the bike in space but two parameters combined. Main part of the motion is to use part of the gravity to imitate accelerating. when we use up to 60 degrees in roll, you can easily calculate how much G you get from it. But we also use the roll orientation to provide feedback about bike state. For example, take a look on 2 situations:
-Bike is accelerating but keeps horizontal
-Bike is not accelerating but you are doing wheelie

the senses feel almost the same thing, so we combine both informations with proper scaling to provide feedback

input for the GP bikes are all the standard controls of the bike. Everything that has part in real riding. Your position ( hard to track the bikers center of the gravity so we use head tracking) handlebars force ( or axis depends what will be working with GP Bikes direct steering) and some obvious like throtle, brake etc.

on pitch axis we want to move the bike according to the force vector.
If you are turning and keep straight on the bike the force vector is parallel to your lean angle. So to simulate it we don't have to roll to the sides at all.
But if you move your body of the bike to the center of corner you move your center of gravity so the force vector wont be paraller any more ( as it still goes trough the center of gravity of the whole setup).
So as follow, we won't change the value of the force, but, at least keep the right direction of it. That's why we will pitch the bike in opposite direction of turning if the biker moves.

HornetMaX

Quote from: Furious on November 20, 2014, 12:17:40 AM
Pitch axis and roll axis are dictated by GP Bikes however not the actual orientation of the bike in space but two parameters combined. Main part of the motion is to use part of the gravity to imitate accelerating. when we use up to 60 degrees in roll, you can easily calculate how much G you get from it. But we also use the roll orientation to provide feedback about bike state. For example, take a look on 2 situations:
-Bike is accelerating but keeps horizontal
-Bike is not accelerating but you are doing wheelie

the senses feel almost the same thing, so we combine both informations with proper scaling to provide feedback
Very unclear: you mean  that if the bike in GPB stays horizontal and accelerates, your rig will wheelie ?

Quote from: Furious on November 20, 2014, 12:17:40 AM
input for the GP bikes are all the standard controls of the bike. Everything that has part in real riding. Your position ( hard to track the bikers center of the gravity so we use head tracking) handlebars force ( or axis depends what will be working with GP Bikes direct steering) and some obvious like throtle, brake etc.
OK.

Quote from: Furious on November 20, 2014, 12:17:40 AM
on pitch axis we want to move the bike according to the force vector.
If you are turning and keep straight on the bike the force vector is parallel to your lean angle. So to simulate it we don't have to roll to the sides at all.
But if you move your body of the bike to the center of corner you move your center of gravity so the force vector wont be paraller any more ( as it still goes trough the center of gravity of the whole setup).
So as follow, we won't change the value of the force, but, at least keep the right direction of it. That's why we will pitch the bike in opposite direction of turning if the biker moves.
Totally lost. Which force vector ?
And this: "pitch the bike in opposite direction of turning" ?!
Pitch angle is the angle that varies when (for example) you do a wheeling. How can you pitch the bike in the "opposite direction of turning" ?!

Did you actually mean yaw (bike spinning around a vertical axis) ?

MaX.

Furious

November 20, 2014, 12:52:38 AM #21 Last Edit: November 20, 2014, 12:56:24 AM by Furious
QuoteVery unclear: you mean  that if the bike in GPB stays horizontal and accelerates, your rig will wheelie ?
yes. Why? Cause i simulate forces. Keep the force vector on the same direction. Imagine : gravity acceleration vertical and bike acceleration horizontal vector. Sum them up and you have the angle . We Pitch the bike to make that new vector parallel to gravity vector.

Sorry. I missmatched the words. I meant pitch when writing roll and the opposite.

And the most important thing. We are using oculus rift. So we can "cheat" with forces even more. We tested it and it really works. Just hope Piboso will provide oculus support for our tests.

HornetMaX

Quote from: Furious on November 20, 2014, 12:52:38 AM
QuoteVery unclear: you mean  that if the bike in GPB stays horizontal and accelerates, your rig will wheelie ?
yes. Why? Cause i simulate forces. Keep the force vector on the same direction. Imagine : gravity acceleration vertical and bike acceleration horizontal vector. Sum them up and you have the angle . We Pitch the bike to make that new vector parallel to gravity vector.
So at best you're just rotating the bike frame in order to align the gravity with the overall force acting on the simulated bike. That's very different from saying "Acceleration simulation (up to 80 % of the real bike acceleration)".

Also, if the bike in GPB accelerate forward at 1g, the overall, force acting on the bike (roughly) is pointing down (1g gravity) and forward (1g acc) at 45 degrees: if you want to keep your rig's frame oriented so that the gravity is aligned as the overall force acting on thew virtual bike, you should pitch your bike forward (45 degrees !!), not backward. Which would be very weird: when you accelerate the rig frame would pitch forward and when you brake it would pitch backward (wheelie).

As you said you're doing the opposite (i.e. bike accelerating = rig pitching backward), you're not keeping the vectors aligned.

I don't think trying to simulate the forces is a good idea, because basically you can't and, in trying to do so, you mess up the bike attitude (your rig pitches while the bike in GPB is horizontal).

@Admins: you may want to split the discussion between Furious and me to a separate topic. But keep it, it's very interesting.

MaX.


Gibbon

November 20, 2014, 09:56:21 AM #24 Last Edit: November 20, 2014, 12:21:17 PM by wkp
Quote from: HornetMaX on November 20, 2014, 07:50:01 AM
Quote from: Furious on November 20, 2014, 12:52:38 AM
QuoteVery unclear: you mean  that if the bike in GPB stays horizontal and accelerates, your rig will wheelie ?
yes. Why? Cause i simulate forces. Keep the force vector on the same direction. Imagine : gravity acceleration vertical and bike acceleration horizontal vector. Sum them up and you have the angle . We Pitch the bike to make that new vector parallel to gravity vector.
So at best you're just rotating the bike frame in order to align the gravity with the overall force acting on the simulated bike. That's very different from saying "Acceleration simulation (up to 80 % of the real bike acceleration)".

Also, if the bike in GPB accelerate forward at 1g, the overall, force acting on the bike (roughly) is pointing down (1g gravity) and forward (1g acc) at 45 degrees: if you want to keep your rig's frame oriented so that the gravity is aligned as the overall force acting on thew virtual bike, you should pitch your bike forward (45 degrees !!), not backward. Which would be very weird: when you accelerate the rig frame would pitch forward and when you brake it would pitch backward (wheelie).

As you said you're doing the opposite (i.e. bike accelerating = rig pitching backward), you're not keeping the vectors aligned.

I don't think trying to simulate the forces is a good idea, because basically you can't and, in trying to do so, you mess up the bike attitude (your rig pitches while the bike in GPB is horizontal).

@Admins: you may want to split the discussion between Furious and me to a separate topic. But keep it, it's very interesting.

MaX.

Hi, I'm also working on a full motion simulator based on GPB and Oculus Rift (I'm not the one who sent the mail).  ;)
Furious, you did a really nice job but I totally agree with HornetMax, it will be very hard (well, almost impossible) to simulate the gravity and simulate forces will not a give you a "instinctive" approach wich is in my opinion quite important in simulation.

Also I have a question about your plateform and how the rider will move his body around the bike...it seems that there is no lot of space.

@Piboso: Do you have more information about the company? What country ?

Furious

As I said. I messed up pitch and roll words at some point.

We are going to pitch the bike forward while breaking and pitch it backawrd while accelerating.

We do roll the bike in the opposite direction of the turn if the biker moves to the center of corner.

And as i sad. We can simulate ACCELERATING. When we pitch the bike 60 degrees backwards one vector is 0.85 G and the one to simulate gravity is 0.5 G

I'm sure you are wrong with your directions and I'll show that on the picture.

There is gonna be more space for biker to move around.

h106frp

Found another link for multi DOF (2,3,6) platforms that claim to support GPB already.
http://www.motionsystems.eu/software/platform-manager

lluisete

Champion of Summer Cup 2012 and SRC125cc.<br />Subchampion of SRC500cc.<br />Admin of @gpbikes_fans<br />Personal instagram @lluiscabedo

PiBoSo

Quote from: h106frp on November 20, 2014, 01:57:57 PM
Found another link for multi DOF (2,3,6) platforms that claim to support GPB already.
http://www.motionsystems.eu/software/platform-manager

Ah, perfect, at last!
Thank you.
"La perfezione non è il nostro obiettivo, è la nostra tendenza".

TFC

Surely the biggest problem for a bike motion simulator is the corners? IRL when you lean a bike right over to corner you're being held in the seat / on the bike through the g-force generated by the bike traveling away from its previous trajectory - the same way the water stays in the bucket when swung upside down - centrifugal?

All full motion bike simulators I've seen so far look odd. Looks like you would be literally hanging on to the bike when cornering - feeling gravity's pull vertically, instead of being held in the seat through g-force horizontally.