Hi,
this is what I found in KUKA manual:
High axis torques arise when the stationary robot starts to move. For this
reason, the axis torques are not monitored in the starting phase (approx.
700 ms).
Does it mean that first 0,7s (!) after stop, the robot can move with max load (without stopping) even though the torqmon is set low?
What can I do to protect a fragile product when robot picks the product from a fixture? I dont want to use a super slow movement as we need a fair cycle time.
Thanks.
TORQMON - 700ms delay after stop
-
jirikk -
October 30, 2017 at 4:55 PM -
Thread is marked as Resolved.
-
-
every time robot accelerates, current through motors is increased.
if you make robot stop-and-go then you will have longer cycle time and current spikes.
torque monitoring (collision detection) ignores those spikes but that also means during those periods, collision detection does not work.idea is to use approximated motion and smooth paths so that current does not spike.
-
Yes, I use the approx motion as much as possible, but when picking up the part, I have to stop fully, close the grippers and then back out. In this motion I have low ACC and TORQMON but still sometimes it breaks the part, even moves the whole conveyor if the part gets stuck in the fixture on the conveyor.
I am just surprised the the TORQMON inactive period is that loooong
I would assume, the motion model would count with these acc spikes.
Is there any other trick besides slow acc and speed ?
Thanks. -
I don't think that 700ms is fixed. It should be dynamically determined from spike... but why would you use this in place when speed and acceleration are needed?