Flying Shear Position Control

  • Thread starter Hooshang Farhadi
  • Start date
H

Thread Starter

Hooshang Farhadi

Dear Colleague,

I want to control Position, Cut Cycle and Braking of a flying shear in Bar rolling mill factory.

I can't use PLC for this purpose, So my question is "Is there any independent controller for this purpose?".

Best Regard
 
P

Peter Nachtwey

Perhaps the best PLCs can but in general

1. PLCs are too slow with their many millisecond scans.

2. PLCs do not have deterministic scans.

3. Often the inputs are not updated synchronously with the PLC scan.

4. That analog output is slow.

5. PLCs do not have the motion algorithms built in. A Control Logix might but it will still require a M02AE card. Simple move here and move there commands will not do.

6. Most PLCs have PIDs with time constants expressed in minutes not milliseconds.

7. A PLC will require reading the updates at the beginning of the scan and updating the output at the end of the scan. A good motion controller can read the inputs and update the control outputs in a few microseconds.

The control will require a PID with feed forwards. Do PLCs have feed forwards? The best do but to use feed forwards you need to calculate the target velocity and acceleration at every scan. This means that a motion profile generator is required. Writing one takes a lot of time. If the PLC has a cam table then that can be used if it does cubic interpolation between the points but will the cam table also provide velocity and accelerations? Now the problem shifts to how to generate the cam table.

The killer issue gearing to the line encoder that measures the speed of the material that is being cut. One can't simply gear positions. One must match position, velocity and accelerations. If you look at the www.plcs.net forum you will see that estimating velocity with a PLC is not easy and no one even considers matching the acceleration. If the line encoder does not provide a smooth position, velocity and acceleration the slave or shear cannot be tuned as aggressively and this limits accuracy. The derivative gain should be used on high performance system. The derivative gain is multiplied by the error between the target and actual velocities but often estimating the error between the target and actual velocity is difficult because of sample jitter and low resolution quantizing. This limits the use of the derivative gain.

Finally, if you buy a PLC that is what you are getting a dumb box. A good motion controller will let you simulate the motion so you can simulate the timing, velocities and accelerations so that the components can be properly sized.

Even if the PLC can do the job can the programmer?

I have seen too many PLC programmers try to do the impossible with a PLC and waste a lot of time.

Peter Nachtwey
President
Delta Computer Systems, Inc.
 
H

Hooshang Farhadi

Because this factory has very old technology, and there isn't any PLC system, so, they don't want to expend for purchasing expensive PLC system.
 
As a wise man once said:

You want to make and omelet, you have to break some eggs...

I agree with the people in the prior posts, buy a real motion controller: Delta Tau, Galil, etc. Some PLCs have motion controllers that you can use with them, but they end up costing similar prices to a better suited "pure" motion controller.

By the way, how are you going to implement a new system without spending any money?

KEJR
 
C

curt wuollet

You might even check with Lenze etal. They have drives that are dedicated to synchronous motion.

Regards
cww
 
A
I have used the following controls for a flying shear for a speed of 66 cuts per minute using a GE Fanuc PLC and an axis positioning module.
I have also worked with some of the reliance electric control and also a German company named EMG (I don't know is they still exist.
The application you are doing sounds like a close call for EMG's controllers.
 
Hello:

Before you spend any time investigating any motion control solution, perform the following calculation: Tolerance of the shear position (i.e. .010 inches) divided by the linear speed of the product (i.e. 10 inches per second) The result will be a time in seconds, .001 seconds in the above example. Choose no motion controller that does not close the position loop in less time than calculated above. If the start position for the shear must be determined by the program you will write, choose no controller that will have a loop execution time greater than the calculated time. If both the position loop and the loop execution time will be in play, then the sum of these must be less than the time you calculated. Some controllers support interrupts to reduce response times for such functions. Some controllers and drive support position latches and provide special functions for executing flying shears and other functions where two axes must meet in space and time, both being precalculated.

Employ a servo system that provides sufficient bandwidth from position through torque loop and motion and the motor to achieve the desired result. Approximate the FFT of the velocity profile you need to perform the required motion and take 3 to 5 terms if necessary to determine this.

Stay away from any controller that outputs a current (torque) command to the servo amplifier, using the encoder to close the velocity loop unless the encoder had a minimum resolution of 100,000 lines per rev and the servo loop in the controller solves the velocity loop at least as fast as every 500 microseconds. Better yet, simply stay away from such controllers. All quality drives close the velocity loop far better than the controllers that output a torque command and tuning will be much easier. (Please do not argue, affected manufacturers, or I will post the derivation of the equations and the results of the calculations)

The above is based upon executing many flying shear systems that are similar in operation.
 
W

William Sturm

Larry Said:  "All quality drives close the velocity loop far better than the controllers that output a torque command and tuning will be much easier.  (Please do not argue, affected manufacturers, or I will post the derivation of the equations and the results of the calculations)"

I am not arguing at all, I have had success with both methods.  But, I would be very curious to see your equations and results that support your preference for velocity mode amplifiers.  It would make for some interesting discussion, in a new thread perhaps...

Bill Sturm
 
hi dear LARRY;

may i ask you to help me find the calculations. i want to know more about basic calculations and equations of flying shear.

so i'm waiting for your help eagerly
 
If you can provide information on:

Max Line Speed
Min Cut Length
Shear Carriage Weight
Desired Carriage Motive Power Source
Available power

We could develop a much better set of recommendations for you
 
P

Peter Nachtwey

---- snip ----

>We could develop a much better set of
>recommendations for you

Yes, but I would add the cut time because the longer the cut time the short the time the shear has to get back to position and this minimizes the maximum line speed.
 
hi dear all may i have your email addresses so that i could send you some pictures in this case?

as you know there is no attachment facilities,in here.
 
hi dear ken brown;

the line speed varies between 10-50(m/min),the length of work piece changes between 1-12(m)and the gauge(thickness)of work piece varies between 6 - 16 (mm).

i know no more thing about the rest,please tell me how to find them .
just one more thing:it's my first project on flying shears,so if possible please tell me what do you mean from the last three ones(shear carriage weight,desire carriage motive power source,available power source).

do you mean motors power(kw)?
 
Thread starter Similar threads Forum Replies Date
A Motion Control 3
P General Automation Chat 0
M Motion Control 6
M General Automation Chat 0
I Computing 5
Top