My project : Zero-gravity, 4dof

Hi,

My name is Fern and I started building my sim.

Some parts have been fabricated and many will end being just prototypes to a more refined sim. Although it is in early stage I decided to post informations / concepts (no photos or videos yet) about it since I read some of you are also trying to attain this goal : Zero-gravity.

What do I mean by Zero-gravity ? It is having the seat like floating in the air and be able to raise it to the highest level with one finger, leave it there and it stays there then lower it down still with one finger and it stays there to. It is also being able to lean it like in a curve with one finger and bring it back with the same ease.

IMPOSSIBLE ???

I do not think so.

The solution is in 2 steps : first the static weight and AIR is the solution.

Here is the concept :
If myself and my seat with pedals, steering everything I want to be in a Zero-gravity state weights 315 lbs, I need a cylinder of 2 in (50mm) of bore at 100 psi to hold everything to the highest position. It has to be fixed right under the center of mass/gravity. Now I want to add just a few pounds and see everything go down to the lower position.

If my cylinder stroke is 8 in (200 mm) I will have to compress 393 cc of air. The solution is quite simple, I need an EXPANSION tank. So my cylinder is connected with the largest hose possible (½ or 12mm) without any valve to the tank that must be about 50 times more volume than the volume I want to displace (393 cc). So I connect a tank that is 393 * 50 = 19650 cc (19.65L). Now I have a volume expanded of 19.65L + 0.4L = 20.05L that I want to compress into 19.65L. The ratio is 1.02/1 and I have to add 6.3lbs to see the seat go down or the actuators/wiper motors will only have 6.3lbs to pull IF THE SYSTEM IS PROPERLY CALIBRATED that is having the right pressure but it does not have to be absolutely exact because the actuators can move more than that.

I know that I attain my first goal with that concept.

If you have a tripod, you may have only one cylinder in the middle or one to each corner in which case they must be independant, each one having its own expansion tank.

The solution for the second case, weight shift is also very simple.

When the seat leans, it does the same as a real car does in a curve so the solution is the same as the car manufacturer use : stabilizer bar/anti-roll bar.

If you have a 2 dof, you may use only one for the roll axis or you may also use another one for pitch too. If your sim is a tripod or hexapod, use three.
Where did I find a bar for this ? From a Honda Civic. They use small torsion bars to hold the trunk cover. It is only ¼ (6mm) in diameter but it can be bent to suit and adjusted. Be aware that heating will change properties of the spring.

I hope this gives ideas to you.

I will post more details later but do not expect the project to be finished before fall. Snow is melting now and motorcycle season is coming soon. I enjoy virtual racing but I am not sure that it can be as pleasant as riding my BMW. I also have to take care of the business.

Thanks for reading me

like to hear how ya plan to control this because air is tricky bugger

Let's see some drawings!

4dof? maybe someone here can help you..I learn a lot of stuff here.

Hi

@eaorobbie

Air is only a support and will not be used for actuators. Spings are more difficult to use for this.

@Frakk

The first post was mainly to help you think of solutions on the subject. I am not very good at drawing, specially with CAD, it is not even basic. I am designing most parts and brackets with emachineshop software.

http://www.emachineshop.com/

This bracket was designed using their software and they can fabricate parts I design with it. This bracket still needs refinement.



I will post a video as soon I can make one with the test rig.

@kubing

Thanks, I am sure I will find lot of usefull info and input here.

My project will have 4 actuators: 2 in the back that will mainly control heave(bumps)/sway(traction loss) and 2 others in front of the seat to control mainly roll/pitch. However all 4 will work together to control movements.

There will NOT be a u-joint and the seat will be positioned by a single reversed A-arm. That means that the pivot point will be virtual just like a hexapod.

I do not have much knowledge about profile now and how difficult they can be made.

Fern

Hi,

Here is the best drawing I could draw.



The location arm will be attached to each side of the seat to prevent spinning and to a single point in front at about this height to have less than 1 mm of fore/aft movement on total up/down travel. This will allow roll of course and also side movement to simulate traction loss .

In the rear the 2 actuators will be attached close to the roll axis to the seat and at the base they will spread about 18. They will provide the side to side and the up/down movement.

The place of the CG was calculated with the test seat and will be close to there with the final setup. The air cylinder will support everything right under the CG. Once pressurized, it will just ease the work of the actuators. Nothing else. Two stabilizer bars will also help when the seat rolls and pitches.

The front actuators will be located straight up on each side of the seat to be the least affected by the side movement.

As you can see, the seat is positionned by the reversed A-arm and the 2 rear actuators. If only the rear actuators move up, the pivot point is the location of the front actuators upper attach points. But if at the same time the front actuators retract the roll center will be somewhere between the attach points.

Fern

Salut Fern,

Beau projet! On va suivre ça de près...

Project completed and racing

Hi, it's been some time,

I was busy building but now I am happy to announce that the project is operational and I am very :yes: satisfied with it.



The front motors attachment and the stabilizer bar (20 in/lb) that ease the work of the motors.

Rear motors in the low positions

A view of the way they are positionned and the Softpot Rotary Potentiometer that I find very convenient. Also note the self made polyurethane bushings.

The steering tilt to ease boarding and is held up with gas struts. A micro-switch reports the status to the microcontroler and the seat is automatically set to a boarding position with nose partially down.

Simply lift the front of the pedals and slide it on the track to adjust in a snap. This way my grandchildren can race easily when they come home.

When not racing the monitors can be rotated and lowered easily since their support is an old air cylinder. I do not need 6 monitors in that room.


Everything has been hand made except the pedals track and most have been done twice or more.

It was a very exciting learning experience and it is not completely over yet. A 4.3 touch screen will take place as a console for the system and communicate with the Arduino Mega which runs without a PID but a much simpler formula that was very easy to create and quickly calculated. The compiled sketch is less than 12K and the delay from the Profiler is 8ms. I use Sabertooth motor drivers.

Some might ask about motor temperature when racing and my answer would be that after 2 hours the highest reading was 41.2°C

The micro-controler also monitor the air pressure of the supporting system and adjust even while racing. All I notice is a click from the solenoid valve.

A video is being edited and will be posted to YouTube asap. The link will be shown here.

Thanks

Awesome.
that looks 100% professional.

What do you use for the microcontroller? Did you code the micro controller yourself?

Hello,

Thats really nice... looks cool. Keep it up. I hope you dont run into too much problems to find our perfect solution.

Regards
MSK

My project : Zero-gravity, 4dof VIDEO

Hi

Here is the video

Thats a great job there! Did you factor in vertical force ( bumps) in the motion as well?

I started with the Velleman VM140 but was completely unsatisfied with it. I considered Thanos solution but without the code it was not interesting enough and it was good only for 2 motors. Then in a post Thanos wrote that the Arduino Mega 2560 would be a good choice so I started looking at everything I found on it. I never programmed a microcontroller before, I do pretty good with Delphi 5-2010 but I never wrote a line in C since Turbo C from Borland.

Ok, I created a class for the motors, download a class for PID I found at Arduino playground but when I tried it it was jerky exactly like the VM140 when it was supposed to keep it's position. I remembered the creator of the PID class wrote : If you have a simpler solution that works, do not bother with a PID.

So I wrote a very simple and short formula that works. It made all the motors stable and controllable. Then I change the delay in the Profiler to 8ms from 33ms. I guess the refresh rate must be aorund 150/sec

I am willing to share the code with anyone that consider using the Arduino and compatible. It should work easily with 6 motors.

Fern

Well, all the motors have the same values in the Profiler. It is one axis per motor and the g-forces are in this order : lateral, longitudinal and vertical. The minimum are all set to 0, the maximum at 2000000 and I only change the % to 30-40 for lateral, 20 for longitudinal, 20-30 for vertical. They all add up. Of course some values are reversed and cars in a game can not all share the same profile.
But all the motors/axis are profiled this way.

great to hear! you share a similar story to me on the programming side.

I have had alot of experinece in C# and 1 uni beginners class in C. But im completely new to microcontrollers.
Im also using an Arduino to code my solution so I would love to see your source if you are willing to post it.

Im using the PID class that atmel produced so Im hoping its better than the Arduino PID, if its not I might adopt your solution.

Fair enough, however, did you try to use maximum adjustment to find a motion number?
I used to do the same thing, just put in a number, now I use max adjust, then untick the box , then I fiddle with that number and use collision effect to softens big jolts.
Just a suggestion. :thbup:

This is totally awesome and perfect A+, 10 out of 10

This is in my eyes is the best sim I have seen here.

Can we have a better quality video please and then one of the parts as it is moving.

You say 4dof but I only see 3dof.

It really looks great! Want to consider sharing your drawings and make a how to?

Wonderful job :clap: :thbup: ... says Barts, one of the best ever!
That engine is using to move the structure ?

Excellent, congratulations!

Regards

wow what a great job, very professional!!! :adore: