Motion Simulator System with Pneumatic Muscle

Hi Everybody,

Allow me first to introduce myself: I am a senior mechatronics student in Turkey and doing my internship in FESTO-Turkey. My boss here asked me to do a project which is also necessary for me so that my internship is accepted by my college. The project is the design and application of a motion simulator for racing. It is already made by a customer of FESTO Germany but we can't have the details of that project so I must see through the design process. However,I do have a video of the motion simulator working which I will be putting on my web-space later this evening (I will inform you when I do). As you will be seeing from the video the motion platform is actuated through pneumatic muscles (another example using the pneumatic muscles is the x6-simulator as you may already know). In order to drive these muscles we must use proportional valves of FESTO,details of which are attached to this post as a catalog (I attached both German and English catalogs). MPES-3-1/2-6-010 and MPES-3-1/2-6-010 are the two types of valves that can be used to drive the pneumatic muscles and you can find the detailed specifications in the catalogs. So,after giving you the basic informations my question follows: There are several interfaces that are suggested in the main website of x-sim simulator to use to control the valves of any pneumatic system, but from what I understand all examples in the forum are built with pneumatic cylinders. Is any one of the cards usable for this type of actuator (pneumatic muscle) and these types of valves? If yes, which one? I asked this question also to r.hermanau and I got an answer from him acknowledging that the veleman interface or their self developed interface card can be used but I wanted to be sure before purchasing the card and therefore I am asking the forum as him suggested as well. I am waiting excited on your answers because I am in short notice and because I believe (from what I've read in the forum) that there are many users with great knowledge about the subject (unlike me) in the forum. Thank you all because of great information until now and in advance for your help about my project.

festo_catalogues.rar (1.48 MB)

edit: the attachments are corrected

Hi Refugee

Welcome to our forum!

For normal cylinders you need 4 valves. 2 air-intake valves, and 2 exhaust valves. I never used these pneumatic muscles, but I think they need just 2 valves at all, because they go into home position automatically (because of the expander muscles). As you can assign movement profiles in the force profiler (valve matrix), you can use manifold valve configurations.

Its not a question of the right interface for the cylinder/muscle configuration. Its a question of assignment in the valve matrix.

But for now, prop. valves arent supported in the current Force-Profiler. This is a feature of Force-Profiler 2.0, wich isnt released, still. When do you have to start this project? I assume it wont take that much time, until the Profiler 2.0 will be released.

But besides that, only the Velleman VM129/K8055 will be able to support proportional valves for now, as it has an analog outputs. The others, mentioned on the site, are just I/O.

There would be a way to realise a 3 speeds mode with different pressures for the pneumatic muscles with EDIT: 4 single valves per muscle, if you cannot wait.

BTW, please do not use spaces in your file-names, as the linux server cannot handle them. You could use under-lines instead.

ego

BTW: How do you want to attach a way measurement system? Whith our interfaces only potentiometer type systems or hall sensors are supported. Another member tried to use inclinometers, but they seem to have probs with fast direction changes.

Perhaps you could use some tackle potis!?

Do you have some construction drawings yet? As the current Profiler supports a CAD visualisation interface, all CAD simulator drawings are wellcome, in oder to animate them within the X-Sim software.

ego

Hi ego,

Thank you for your interest in my project. I want to apologize in advance for all my further posts because my questions could be very trivial and maybe also silly. I am fairly new to pneumatics (I became familiar with it for the first time during internship which is in its 3rd week now) and also motion simulators. So I want to say that I'm sorry if I waste your time with senseless questions in the future.

Anyway, moving on to the comments of you: I didn't see the page about the valve matrix until I read your post so now I checked and read it,but I still have questions. From the very little that I've understood, the matrix controls any type of valve combination for simulators, but it only provides an open/close control (which coincides what you've said about the control of proportional valves). Not directly related to my project (because I must use pneumatic muscles) but, how do you achieve middle positions with normal pneumatic valves and cylinders? From what I understand Force Profiler extracts data from both the game and the joystick/wheel (G-Force,Collision etc.) and uses these data to calculate the necessary position of the chair(Sollposition). And from what I know, the normal pneumatic cylinders can only move to end positions when they are used with normal valves. So,again, how do you achieve placing the chair to the Sollposition? Or am I completely lost here?

I think I must start working on the project next week at the very latest because I have 6 weeks left on my internship and the project must be at least very close to finalized until the internship ends. Do you think that Force-Profiler 2.0 will be released in such a short time? (I doubt it to be honest).

Is this the method mentioned as Soft-Pneumatic on the website? From I understand this method achieves a slower movement of the actuators and the chair but does it achieve middle positioning? (which is what I need) If it does can you elaborate on the details of this method?

I didn't talk about this with my head engineer in the department. I will inform you when I know more. But, are you suggesting that potentiometer type systems can't be used with pneumatic muscles?

I'm actually gonna start making the CAD drawings tomorrow with the head engineer so I will post them here when they are done (or close to done). From what you said, I understand that the drawings can be used to animate the effects in the X-Sim software without actually having the simulator? If I'm correct, it would be great and I'd appreciate it if you can give me some further information.

I'm sorry about the long post but I want to understand everything correctly and I'm a little short on knowledge about the subject. Thank you very much for your help.

Can

Thats right

Its related to your project, because on software side its equal if you use cylinders or muscles.
Its just a different way to assign the valve profiles in the matrix

Without seeing a drawing, I cannot give you hints, how to attach a way measurement system. Generally the Force-Profiler gets the current position (Ist-Position) data from potentiometers attached to the actuators, and decides, depending the current position, how long the valves have to be open until the desired position (Soll Position) is reached. When the Soll Postition is reached, the air-intake valve closes, as well as the exhaust valve, and the actuator rests in fixed position.
That way you can drive the actuators to middle postition.

No, I dont think so.

No, on the website the pulsmode softmotion is mentioned. The pulsmode needs, in your case, only 2 valves per muscle and works with 2 valves. But you will notice the plusing valves while riding the simulator.
The smoother method is to use different pressures.

I didn't talk about this with my head engineer in the department. I will inform you when I know more. But, are you suggesting that potentiometer type systems can't be used with pneumatic muscles?[/quote]
Think about it perhaps you'll find a way

Thats right. It would be helpful for you and for us. We'd need the files in stl or vrml format. Ask me for other file-types.
All moving parts have to be a seperated file.
There are some informations as well in the CAD section

ego

Oh, I should have informed myself better about the muscles before

ego

EDIT:

Do you have a source for this information? I cannot find anything.

Hi ego and sirnoname,

Sorry for the late reply but I was trying to gather more knowledge about the information and answers you gave me using the forum and asking people around the company because I want to take minimum of your time with silly questions (well,they still happen anyway but I hope you would excuse me ).

Now I understand it ego,thanks. It is useful for me to improve myself about the subject. However, due to my need of using Pneumatic Muscles, I can't really use this information for my project but thanks again anyway.

We will most probably design the simulator in SolidWorks. I will check out the CAD section to see the possible applications of SolidWorks file formats.

To answer ego's question as well, I'm not sure if they have driving control included, but I know that they can move to any desired position just by having the according input voltage,the values of which vary between 0-10 Volts. I will be having more information by tomorrow or later this afternoon on technical details and I will be posting them here as well. Once again, excuse my inadequate knowledge.

Well,I am pretty much lost about this point. I wanted to have more information reading the x-simulator web page about USO but I didn't understand much to be honest. If you can be a little more explanatory about your suggestion or give me any sources to gather basic information about USO, I'd appreciate it a lot. My specific questions are: What do you mean by receiving serial packets? What is the function of the DAC chip,will it be working as the controller microchip? Do I have to build it or can I have readily built ones? (I am still trying to understand these myself by reading through the net,so it'd be also OK if you just give me sources instead of trying to explain yourself) You also said that you can add an analog output for velleman so you mean that if you do so,I can use the Velleman card and a microchip (DAC??) and I would have my solution? I haven't bought any card yet but if it's really necessary I am willing to purchase the Velleman card right away because I don't have a long time as I said before.

Thank you both again...

Refugee

Please post as many informations as you can, as we are new into this type of actuators too.

The USO is an software interface which is implemented in the Profiler, and is able to to provide Soll-Position to an external hardware interface.
This external interface is the DAC chip, Sirnoname mentions. He could implement a mod into the Profiler 1.8, which would enable you to receive analog data via the serial port of your PC.
But as you seem to be quite uninformed about this matter, I wish you good luck to get yourself informed about it in 6 weeks

I need more informations about the muscles if you want my help!

ego

Well I understand the basics that USO is a software interface that must be connected to a hardware chip. But I have the following questions: Does USO give the Soll-Position of the chair as the output? Does it do this in 2 axes because sirnoname mentioned something about 2 outputs? Will the serial port be working for receiving or transmitting information or both? If it is used for receiving what kind of information would be received (is it the current position of the chair)? I know I'm uninformed but I am very eager to learn and I won't be all alone in the project as I will receive help from the engineers in the firm later on. But I want to learn as much as I can by myself so any additional source or information would be greatly appreciated .

ok, just to know if we are talking about the same model of actuator.
I mentioned this type:


And this muscle doesnt seem to have any integrated way measurement system nor any controlling devices. Its just a tube.

A way to control them via the X-Sim Software could be to connect them to 4 single valves this way:

BTW, this should be your job

As it seems that you have much knowledge about the electronic part (serial-port, PIC etc) it would be better perhaps, if you go this easier way.
Just my humble opinion. Because 6 weeks is not that much to get into the electronic part AND assemble a simulator.

But if you want to try it the hard way. Let's go on! It would be helpful for us
Afaik the USO would just provide your interface with Soll-Position. 2 axes.
The USO is an output interface.
Afaik, your interface would have to manage the way measurement

ego

And what about position feedback? How could he center his simulator?

Then please tell me wich interface you mention. Because the muscle, I posted before, dont have an interface. Just a tube, nothing more...

EDIT: I think 5V is the center position of the valve.

That seems logical

Well,yo got it exactly right sirnoname. The pneumatic muscle works exactly like that. Following that,we come to the crucial point:

I don't think that I'm gonna need 4 axes because the simulator would be for shows only, we don't need that much of reality. So I understand that with your addition to the profiler the Sollposition can be transfered through COM ports to an external chip that should convert these signals to according values and transmit them to our valves. Once again coming back to your solution:

What confuses me here is that you mention USO and Velleman. If I need just the serial port of my PC and a chip,why do I need Velleman interface? Does this mean that the Velleman interface will be working as the needed microchip? (I don't think so) If not why do we need an analog output for Velleman in profiler? Also, would you be so kind and post some material about DAC chips if you've got any? Would I have to build them or can I buy some readily built ones?

I know I'm asking a lot of questions but as ego said,I'm going the hard way on a short time because I have to do so. Even if it isn't finished I have to be beyond a certain point in all aspects of the project. So thanks again for your kind and detailed answers.

Refugee

So, as I understand it,all I need now is the new version of x-sim, the velleman interface and op-amp to amplify the signal of Velleman. I don't see any need for USO and a programmed chip if it can be solved as easily as this. Thank you very very much! The only question that remains is how I can order the Velleman card but I believe the obvious and only answer to that is their own web-site,right?

And if I need 4 axes I'm gonna need two control PC's and 2 Velleman cards,right? Or can one Velleman solve it?

In addition to that I wanna ask another (maybe silly) question: As we are talking x and y Sollpositionen are we talking about positions in the translational or rotational axes? From what you are saying (or the way you're saying it) I'm understanding that they are translational,but isn't the simulator supposed to move in rotational axes to simulate curves,G-Forces,accelerations,brakes etc? Are we gonna need 4 axes then? Or are you talking about rotational axes when you say x and y Sollposition anyway?

Thanks a lot again (I know that I thanked too much but actually I can't thank you enough )

Refugee

I just checked the site for Velleman card and saw that there are two outputs and got my answer which was obvious anyway

However, I didn't quite understand your other answer. Couldn't I put the question clearly? Whatever names we give them, is the Profiler giving the positions in x and y rotational axes as output to Velleman? Or is it giving regular x and y positions in space with respect to center position? Did I make it clear? Maybe I'm not explaining myself well enough,let me know if it's so and I will try to make myself clearer.

Also,the question about purchasing the velleman... I'm right about purchasing it through their website,right? The x-sim website does not provide any help on that I suppose.

BTW, we are starting the design process of the simulator today and I will post the drawings when they are ready to be shown (even if they aren't finalized)

Whats the point? I dont get it. The axes are the strokes of the actuators. The actuators are translational, and if you build a pivot point simulator, the translational movements turn to rotational movements. But it doesnt affect the centerposition. Center is in your case 5V.

BTW: I'd try it here: http://www.conrad-electronic.com -> 191137 - 62
If you would assemble it by yourself you could save 16Euros -> 191003 - 62

Yes,yes, after I read the posts again I understood that. Sorry for the extra questions. I also found that website but thanks anyway. I am now looking to order the cards and although it seems hard, my hopes have risen up that we can finish the project in 6 weeks. Most of that due to your help

Hi again,
I haven't been writing much lately,because there was no progress in the project until today. But,today or tomorrow we are ordering the Velleman card and starting the design of the simulator. Also,we will consult a firm that can be more experienced in the matter but I need to collect and present them the data we have so far. So, I wanted to check with you again and confirm if my knowledge is correct so far.

So,here goes:

I will have one Game PC that will have the racing game and Yoda open (BTW,is there a full list of games that Yoda supports?). Yoda will extract data from the game and send it to Force-Sender which will be open in Game PC (?) and also gather telemetrie data from game controller (joystick,wheel etc.)

Then, all of the collected data will be sent to Force-Profiler which will be operating on another Control PC. Well,here comes the tricky part. sirnoname added analog output for Velleman (thanks again,BTW) but without Valvematrix. Well,what does not having a valvematrix for me? Will the Force-Profiler be calculating the positions in X and Y axes (for the two actuators) according to collected Telemetriedata or will this be my job? After this part is complete,the rest should be easier because the Force Profiler will be giving the X and Y positions as two analog outputs through Velleman interface (0-5 Volts which should be increased to 0-10 Volts with an op-amp) and this should be connected to the Pressure regulator of the two pneumatic muscles. If I understand correctly, this should be it theoretically. What is the reality about calculation of X&Y positions and are the rest of my assumptions correct? I am excitingly waiting for your answers.

Thanks....

Refugee