gyroscope jet simulator - concept brainstorming thread

A while ago I was visiting an old military hangar / museum near Götheborg in Sweden.
There were some simple flight simulators and my family and I had a lot of fun in this simple simulators.
I instantly thought of how cool it would be to have a simulator with free tilt and roll features and I came across these things:

http://www.erlebnismanagement.com/files/spacetrainer_gyroscope.jpg

Now before I start building I decided to build a less difficult 2DOF simulator to get familar with one and the other thing. However meanwhile I´d like to read some of your thoughts about a gyroscopic simulator.

There are several things to keep in mind when it comes to free rotation.

- First thing is it doesn´t make much sense with a monitor or beamer setup so a VR system like the oculus rift is a must.

- It is important to keep the g-forces under control. In a jet simulation you could easily touch your body limits. However I saw the spacetrainers and they go really fast, faster than a motor driven system probably will and at the end you decide if you do rolls and loopings in a flight sim. Anyway, sort of an emergency button to stop the simulation would be nice to have.

- A lot of space is required. However I did some calculations and came to the conclusion that it should be possible to build a roll/tilt system that fits into a room with a height of 2,30m. It won´t be too comfy but it isn´t comfy in a jet either ;-)

- The main issue will be to find small but powerfull motors for the inside ring. AC motors are too large and
heavy so I thought of using 2 DC motors per ring. My first idea was to use one motor per direction having
the motors set up on free-wheel gear from a bicycle. However as I just learned by starting with a 2DOF Sim
it is pretty difficult to find small powerful gear motors that are fast enough so probably two motors per
ring need to do the job.

I would only use two rings instead of 3 in the spacetrainer thing due to the space a 3rd ring consumes.
I know this leads to a lack of free movement experience but if required a powerful motor could rotate
the whole platform which isn´t the same as a third ring but better than nothing.

- Then you need to stay in balance when you want to enter the simulation. This means the motors need to have a good holding torque or it requires an electronic gyroscopic balance system. Doing the balance with the motormovement will probably kill the motors more sooner than later unless AC motors are used.
This leads back to the idea having one motor per direction. This works on the outter ring because weight and size don´t matter that much there but for the inside ring it is far more challenging.

So what do you think, apart from that I am mad ? ;-)

EDIT1: Now as @BlazinH mentioned the required rpm for rolls might be higher than those for pitch I rotated the 3D model and then thought why doing an inside *ring* at all. It could be useful for the balance doing a ring but for weight, space and even costs it would be smarter to just have some sort of a bridge. Like the rough sketch on space4.jpg and to go even further the outer ring could be a smaller rectangular
shape. This will reduce the required space, it can also be much lower. There is no need to split the rings and it will fit easily in a normal room height. Both rings need motors with good holding torque so winch motors would be a good idea as long as they are fast enough. On the other hand there is enough space to place them now so even an extra transition could be made.



EDIT2: Oh well, now when I see the 3D sketches I remember that I saw something similar a long while ago made by http://motionsim.blogspot.de/
If I remember right he used some plastic pipes for it but it wasn´t 360°. So kudos to him for putting that anchor in my brain ;-) There is about 50-60cm space when you sit in a jet, with this setup I calculated about 75cm. Now the required space is about 2,40m*1,60m*1,85m. Perfect.



EDIT3: Some calculations, the winch motors have a drum with a diameter of about 32mm and do 5m/minute. This means they are somewhere between 40 and 50 rpm. That means they are fast enough for pitch and roll. The weight for the outter ring with stabilizing diagonals is about 3.39kg per meter steel. Which means the outer ring will have a weight of about 24kg with rectangular pipes of 70x40x2mm the 80mm pipes are already 3mm which is too much and too heavy. The motors are 5.5 kg each so another 11kg when I have two of them on the outer ring. So that will be 35-40kg for the outer ring. The inner "ring" will be close which means the roll motor(s) need to rotate an overall weight of about 200kg. Considering that the weight is not balanced means the motors need to be able to rotate about 300 to 400kg. Maybe I should consider using other materials than steel. At least the inside ring could be aluminium but I am not a fan of mixing materials, this usually will break the weaker parts.

EDIT4: Thinking about the weight optimization I pulled out some tables and formulas. Coulnd´t find any formula about the torsion yet. But at least for the bending:

The long sides are about 2m and the force is about 2000N (rough for 200kg).

for steel:

area moment = outside width*height³ - inside width*height³ / 12
bend = force * length³ / 192* 206.000 N/mm³ (steel) * area moment

This means a 70*40*2mm rectangular pipe bends about 1,440 mm.
Which is hardly noticable when you are in the simulator.

70*30*2mm = 1,724mm
70*20*2mm = 2,148mm
60*20*2mm = 3,215mm

These numbers are so small that calculating for 1m isn´t even worth doing the math.
Even the 60*20*2mm should still be strong enough so what about the weight advantage of smaller steel:

(6m frame + 1m for diagonal torsion stabilizers)

7m * 70*40*2mm ~ 23,73kg
7m * 60*20*2mm ~ 16,70kg

The value for aluminium is 70.000 N/mm³

area moment = outside width*height³ - inside width*height³ / 12
bend = force * length³ / 3* 70.000 N/mm³ (aluminium) * area moment

That means aluminium bends 100 times more or even more using the same geometry as steel.
So only a cage concept could work for aluminium which will result in comparable weights.
Considering that aluminium costs way more and I cannot weld it I will stick to steel and will use
60*20*2mm or a comparable round pipe. We even have two bars on the outer ring so the force is half of that which means a bending of 1.6075mm while in "standby" adding a max g-force of 5 we are at 8,0375 mm bending. And seriously I don´t think I want to blackout in the sim at 5g

I hope I haven´t made any mistakes with that calculation.
However these 8mm bending needs to be calculated for the roll axes. Every centimeter less in length of the roll ring counts. If everything fails some 45° triangular bending stabilizers like on a bridge could help but on the cost of weight. At the end I hope the inside ring won´t be 200kg with a person like me it will be more like 90 to 120kg.

EDIT : 2017. march 7th



Alright I made a cage of cardboard just to check out if 1m³ would be enough. It would if there wouldn´t be the diagonal edges so I am going to extend the dimensions to 1,20m³ but to save weight I´ll consider doing just the octagon. There just needs to be some sort of safety belt at the end but that can be solved.
About 12m (15m for more stability) steel pipes are required to weld this inner thing.

Yes software is another thing. If everything fails with VR a set of small monitors close to the "pilots" head would work too but are way less immersive plus you risk hitting them while doing a loop or roll ;-)

And anything cable-bound inside the system leads to problems with the wiring. So it might be a good idea considering hollow axles from the beginning.

Anyway, this is what I really do not care about right now. There are solutions and there will be better solutions in the future. We are just at the beginning of useful VR sets.

The mechanical part needs to be solved first. The model with the stepper motors is nice but it can´t be translated to lifesize, it will get too large and steppers with that massive power aren´t easy to find or are way too expensive. However for software development it is of course a nice idea and I appreciate that someone is doing that effort.

If everything fails with reading the data from the game you could still calculate with the joystick movements. That ain´t so nice and probably tough to calibrate with each object you fly but it should be possible. Though it doesn´t help with movements caused by collisions and such.

In my eyes the most complex thing between software and mechanic parts is the required speed. It probably is easy to do uncontrolled spinning in a flight simulation but you can´t translate this 1:1 to the sim.
So the question is what will the sim do if the game does too fast movements. A roll is a roll, it doesn´t help if you are turned upside down and the game turned your plane 360° ;-)

What is a reasonable rpm for a flight sim ?
I wouldn´t suggest doing more than 20 rpm in sim even though that spacetrainer goes like 30-40 turns per minute.
On a rollercoaster you do 8 loopings in 125 seconds and thats already feeling very fast.

So anything that goes too fast should be detected by the software and just put you back in balanced position until the movement is over.
For example if you are trying to start a roll and the first 20° are too fast it should just stop there and put you back in position.

Oh well, there we are with VR and motion cancellation.

However maybe we should consider this sim first to be used with the physica of a Fokker Dr.I instead of a Mig29 until we know how much rpm still feel good ;-)

Several have considered this type of sim, me included, but I have yet to see a project completed. There are several "challenges" that must be overcome with this type of design and the cost of implementing is considerable. I don't wish to discourage you, just warn you.

Btw, from my experimentation, a roll rate of around 30rpm and a pitch rate of around 10rpm should be sufficient to keep up with just about any aircraft under normal flight conditions.

Here is a little 3D Sketch of what actually is in my mind + the proof that such a sim could fit into a room with 2,30m ceiling height.

The frame most likely needs more stabilisation in the roll direction. So maybe at the end it will be more like a cubic cage with diagonal bars.
However with these dimensions it requires 2,20m x 2,60m x 2,30m. The space inside currently measured by a person which is about 1,80m like me.
It could be made even a little smaller so it fits in rooms with exactly 2,30m height.

If someone considers rotating the whole thing for a 3rd axis this would require a room of about 4x4m.

My rooms in the second floor are 2,36m in height so there is a gap of 6cm. Should be enough for the movement.

In this example the rings are planned in 80x40mm rectangular pipes.
Maybe even too massive to heavy and consuming too much space. I will calculate the weight and stability for different materials but from my experience
60x40x2mm steel pipes should do the job as well, at least for my weight.

Aluminium may be better for the weight but usually the torsion stability isn´t worth the try plus it is difficult to weld... at least for me.

There is another thing that should be considered. The diameter of the rings should fit through a door. In this concept the outer ring is 1,95m this fits through most doors
but I guess I´d have problems getting it up the stairs so it will be better to have half rings connected with bolts.

@BlazinH :

Yes, no doubt there are several challenges but you never know when you don´t try, right?

Okay, 30 and 10 rpm sound doable. It might be wise to swap the rings then, that way a
slower, smaller motor (or two of them) on the inner ring could do the pitch.

Hi!
That looks very interesting project and i put thump's up for you. I was thinking that rig and wondering is there any metalworking place near you. Because they often have possibility to bend round tube. Then you don't have to weld so much and that makes it stronger.

@mrbeginner about a year ago I wanted to get some bent pipes for a 2Dof sim and I asked 12 companies if they can do that. It was either extremely expensive or they just refused to do it :-(
Don´t ask me why and whats wrong with these guys. Usually this a job done in some minutes if you have the bending machines. So it looks like I need to weld them on my own.

That's the problem in this world. If someone can help he wants money for even a little help. Have you ask metal working school if they could help?

I´ll do further research once I have a good overall concept. Today I spoke to a neighbor he said he knows someone that does steel works.
The outer ring made of 50mm bend round steel would be great, that would be about 5 to 6kg less weight.

Today I was thinking about the best position for the "pilot" flying the simulator. When I started this "project" I thought the best position would be what best fits the available space.
But I think I was wrong. Considering that we are used to a pivot between our legs and our torso as we know it from gymnastics on a bar this is a way better position. For the sim mechanics the perfectly balanced center would be the better choice but due to changing pilots this will be hard to achieve.

Conclusion: The base position for the seat should be as close to the pivot as possible but still be configurable. Sitting too much backwards or to the front will cause the pilot to move a larger diameter at loopings increasing the feeling of falling plus drawing way more force to the geometry.

I am not quite sure but I think the pivot on jets is way behind the pilot, this is no real option for the sim so the movement will not be as realistic as in a modern jet.
However it will be hard to simulate an afterburner and such things anyway ;-)



(Sorry for mistreading the human 3D model, I was too lazy to add bones so I chopped his limbs .)

Look up the SEGA r360 arcade...

It might be easier to do a sphere...

Hmm well, probably not easier but I like the idea putting a sphere into another cage sphere or sort of rings.
I think the SEGA r360 is still different, haven´t seen it live yet but it looks like it uses 2 axes?

Having four motors in a sphere in sphere system could rotate the inner sphere freely...
Just like these stone spheres on a fountain.

Though it could be difficult to power the gaming system inside.
Best bet is to have it run on accumulators.

Now, how to build a perfect and stable sphere ?

It sounds like you were already thinking about that tronicgr ?


http://domhaus.at/images/fotos/vision_01.jpg
http://domhaus.at/images/fotos/vision_02.jpg

I have a cnc for building a small sphere like that but not a 2m sphere.
Besides it will be very difficult to get that one through a door so it even needs to be chopped into at least 3 segments.

Anyway, it is really worth thinking about it this way.

Looks great but no loopings/rolls with that.
About the concern for missing bump effects, these could still be simulated directly at the seat inside the sphere.
Adding a compact 2DOF system to the seat. Of course quite complex and still the question where the power comes from then.
Maybe some vibration feedback will add already some live.

Sigh, thanks for the sphere suggestion now I am thinking about methods to build a GRP (GFK) sphere haha.
Actually I have an idea but I will move this to version 2.0.
A GRP hemisphere still needs a big mould to be created and I don´t have that space right now, plus there are other obstacles that make it more complex than the frame version from above.
It would require a seamless door.

We are in the 21st century, this could be a great job for a large 3D printer

Actually was thinking to retrofit an existing R360... easier than build one from scratch. Think of it like the "DeLorean DMC-12" of simulators Its just cool looking!

If you can get your hands on one. It is a big tall isn´t it ? You need a playing garage for this

For the retro-reffect you also need the chiptune music then.

It happens to have a large and tall garage, yes. By retrofit I mean stripping it from existing electronics and using modern methods of realtime motion control.

I know, just kidding game over *insert coin* ;-)