Verification of hardware clearances

    Hello,

    Do you have any proven methods for verifying the wear of individual axes in a 6-axis robot? I have a robot that moves a detail under a laser head fixed in place. Despite the fact that the endpoint of the path is the same as the starting point (only rotated 360 degrees around the Z-axis), it can be seen that it does not hit the same position.
    I would like to understand the cause of this phenomenon and estimate to what extent it is due to the worn manipulator and to what extent it is due to play in the gripper.
    Attached is a photo where a rotation of 359 degrees was performed as part of the tests.

    Sharing the image of the test with a 359-degree rotation adds clarity to your inquiry enabling others to better visualize and possibly offer more targeted solutions or insights. Your transparency in providing visual evidence is highly appreciated.

    Quote from doomedblue

    how old is your robot? the cut line itself looks like it is within the accuracy tolerance of a robot arm (but obviously idk for sure) did this problem just start?

    The robot is from November 2015, and the exact model is YR-MH00050-J00. After complete cutting, the deviation between the beginning and the end is approximately 0.4mm. In the robot's specification, the only tolerance parameter I found was repeatability at a level of +/- 0.07mm, and indeed, each time the final error is about 0.4mm with a tolerance of +/- 0.07mm. The positional error always occurs in the same direction - when looking at the picture, the cutting end is always lower than the beginning, regardless of whether the movement is clockwise or counterclockwise.
    Until now, the laser has been roughly cutting the material, and the detail was finished on a CNC machine, so no one paid attention to the cutting quality after the laser itself. Recently, an idea emerged to cut the finished parts with the laser alone, and only then did we notice that there is a deviation that disqualifies the product. As an interesting fact: On another, similar machine with an ABB robot, it was possible to achieve the desired effect. There, smaller details are cut, and the robot is newer. Therefore, we are considering replacing the manipulator with a newer one, but first, we would like to determine unequivocally where the cause of the Yaskawa inaccuracy lies.

    Quote from Dawid Firlej

    I found was repeatability at a level of +/- 0.07mm, and indeed, each time the final error is about 0.4mm with a tolerance of +/- 0.07mm. The positional error always occurs in the same direction.

    Off of this, your robot is repeatable. There is a difference between repeatability and accuracy. You can program around repeatability not so much with accuracy.

    I know a thing or two, because I’ve seen a thing or two. Don't even ask about a third thing. I won't know it.

    Quote from doomedblue

    if the cut is off .4mm every time and its the same way every time why would you think the manipulator is wrong or inaccurate.

    Since after a 360-degree rotation around the Z-axis, the TCP should be exactly in the same position, but it appears to be shifted by 0.4mm. The programmed rotation looks as follows:

    IMOV P010 TF

    IMOV P020 TF

    IMOV P020 TF

    IMOV P020 TF

    Where:

    P010 (0,0,2,0,0,0)

    P020 (0,0,0,0,0,120)

Advertising from our partners