Servomotor Losses and Torque Utilization

[kamenges]

I have an application where I am camming a rotary axis to a master axis driving a web material. The cam profile is fairly simple. The axis is a tucker so it interacts with the web in one location on the web in a given cycle and repeats indefinitely. It needs to interact with the web at web speed but can them slow down and speed up as necessary to tuck in the correct location the next cycle based on the product length. For those of you who know, this is very similar to a rotary knife application.

My question has to do with motor heating on the longer product lengths. Once the cycle length it longer than roughly twice the tucker circumference I have a choice to make. I can either stop the tucker axis 180 degrees away from the tuck position to allow web to pass and then restart the axis to engage the web at the desired location. However, I could also let the axis overshoot the 180 degree position, back up an identical distance on the other side of 180 degrees then continue on with the move tuck for the next cycle. This would work for lengths up to just shy of 4 times the tucker circumference, which gets me past the longest length I have to deal with.

Is there a motor heating benefit to be gained by using one profile over the other? In the end the motor is managing the same amount of energy in either case. In one case it is using higher torque but has a near zero torque during dwell. In the other case it is using lower torque but it never stops. I think the only way it matters is if servomotors experience increasing or decreasing percentage losses as a function of torque.

Any ideas would be appreciated.

Keith

 

[Steve Bailey]

On the rotary knife applications I've had to deal with, the worst case was always when the product length was less than the knife circumference. The maximum speed for a cut at 75% of the circumference was significantly slower than the maximum speed for a cut length of 125% of the circumference due to the torque required to accelerate and decelerate the knife.
I think the best answer to your question would come the motor manufacturer. I believe servo motors are designed to be able to maintain continuous stall torque indefinitely at zero speed without overheating. It is hard to imagine that the RMS torque for a cycle in which you reversed direction of rotation would be less than for a cycle in which the motor simply stopped.

 

[kamenges]

You are correct. The worst case is the shorter lengths. But I am pretty much locked in on that profile. There aren't as many ways to skin that cat.

The one I do have a choice on is the longer sheets. The RMS calcs between the two profile methods come up very close to each other. The overshoot profile doesn't have the dwell time for cooling but the lower torque of the overshoot profile compensates for that almost perfectly. I would expect this. We are ultimately talking about the same total energy regardless of how I do this. In the end I have to get a given mass to a given speed; same energy either way. It all really comes down to whether a motor's heat generation would be any different delivering that energy based on the profile and whether that difference is enough to make me choose one profile over the other.

Keith

 

[Steve Bailey]

If the RMS torque comes out the same for both profiles, and you're within safe running conditions with both, then it comes down to which profile gives you lower losses. If the motor manufacturer can't offer any insight, then I'd suggest running for a shift using one profile and measure the motor temperature. Then run another shift using the other profile and compare the temperature. As long as its relatively easy to switch between profiles.