01-02-2014 10:26 AM
Dear Rahul;
You are still giving advice about computing/calculating torques.
I repeat, one more time, I dont need a VI to calculate the torque.
I have already calculated them.
I can not apply the torque (which I have calculated) to the motor.
Please, I kindly request you to analyze the projects which I have attached with my post.
I share them with you again.
Please run them.
The yellow link will not move as desired.
Because of link mass, it oscillates.
I need an idea about how to apply a torque to a motor on the simulator,
I could solve this problem.
The link you shared is useless;
Because the simulator had not been used!
It uses a 3D Picture Control to simulate the skeleton of puma robot.
I need to apply arbibrary torques to the motors in simulator!
Please I repeat my request, check out the projects which I have shared the links.
Open XML files, search for "mass" by Ctrl-F combination.
If mass is 1kgs, the yellow link will move at desired speed.
But If it is 10kgs, the yellow link will not move because the torque is not enough.
If you still dont understand what the problem is, I will really record a video for it.
Thank you
Happy new year
Dinçer Saygılı
Acrome Robotics
Istanbul Technical University
ARI-3 SciencePark
01-03-2014 03:39 PM
Hi dncr,
It looks to me as if the motor in the simulation cannot supply the torque needed. Even if you were able to apply torques, the motor may not be able to handle it. If the motor could handle the torque then it would be able to provide the velocity you desired. But the behavior displayed in the simulation is similar to the real world situation of a motor not supplying enough torque and oscillating at a certain point.
01-03-2014 04:35 PM
Dear Paolo;
Now, I am sure you got what the problem is.
At first thank you for analyzing the relevant files.
Yes, its clear that the torque is not enough to rotate the yellow joint.
But if the torque can be controlled, it is possible to give a desired veloticy profile.
This method is called Computed Torque Control of Robots.
For revolute joints, torque is computed and for prismatic joints the force is computed.
In this situation, the torque of motor applied is unknown.
So, we can not be sure that even the velocity of the yellow link is equal to desired velocity.
The torque must be controlled. It is the only way that for realistic simulation.
Please show me the way to control the torques.
Thank you
Dinçer Saygılı
Acrome Robotics
Istanbul Technical University
ARI-3 SciencePark
01-03-2014 05:49 PM
Paolo;
I want a similar VI with that Simulink Block in Matlab,
It is called Joint Actuator
Here is official information about it.
http://www.mathworks.com/help/physmod/sm/mech/ref/jointactuator.html
With this block, I am able to control the torque(or force for prismatic joints)
so I can simulate the system for real masses.
Thank you
Dinçer Saygılı
Acrome Robotics
Istanbul Technical University
ARI-3 SciencePark
01-07-2014 02:39 PM
Hi dncr,
I was able to make the simulation you sent work by changing the Fmax parameter in the xml file on line 1043. I changed the number from 2.5 to 25 and the 10kg simulation seemed to work the same as the 1 kg simulation. Changing the Fmax seems to increase the limits to allow for it to be driven to higher forces.
Please let me know if this helps.
01-07-2014 04:44 PM
Dear Paolo;
Thank you for your interest in my problem.
I have just tried what you offered, and yes it worked!
But unfortunately, I think you should have also predicted that, this is not enough.
I should control the torque with a time-dependent function, so I will need a VI (G Code) to take the control of the motor.
This example is not important. In real simulation, all of the torques must be controlled.
Here I want to give an advice to NI Developers of Robotics Module.
Please, take a look at robotics textbooks.
(I offer Introduction to Robotics by John J. Craig)
At first, all of examples and simulations of this book should been able to realized by robotics module
And second, unfortunately, robotics module is mostly designed for mobile robots(only wheeled).
If I can not control the torque or force, the simulation becomes unnecessary.
I have lost too much time for a simulation and I have begun to think that give up about LabVIEW and go through its alternatives for simulation.
The precious developers of your team can do better than that.
Thank you again
Dinçer Saygılı
Acrome Robotics
Istanbul Technical University
ARI-3 SciencePark
01-08-2014 02:36 PM
Hi dncr,
You are right that we do not have inherent VIs that accomplish what you want.
Though the simulation in this community example may be of help in your application.
If you do end up switching platforms, I wish you good luck.
01-08-2014 03:17 PM
Dear Paolo;
Thank you for your interest.
I hope the relevant VIs will be added to Robotics Module as soon as possible with updates/upgrades.
If you recognize a VI that I can usei please send me an e-mail: saygilinu@itu.edu.tr
I will take a look the example you offered.
Thank you again
Dinçer Saygılı
Acrome Robotics
Istanbul Technical University
ARI-3 SciencePark