How can I do this ?

faehigkeit1

Member
F
Join Date
Jul 2007
Location
virginia
Posts
77
Hello,
I just had a project dumped in my lap that I need some help with.
We have a grinding machine that has two grinding wheels. One wheel is in a fixed position and the other wheel (called the regulator wheel) indexes inward as it's surface wears down.
This machine has an infeed table that is feeding round bars of steel into the grinding wheels. As the bars are being ground down, the regulator wheel gets smaller from wear so it must index inward (about 3 thousands) every so often. This is controlled by a measuring device on the outfeed table (zumbach laser mic) that in turn pulses a hydraulic solenoid to inch (ever so slightly) the regulator wheel inward. The amount of travel is controlled by the human operator(hydraulic valve adjustment).
How can I track the regulator wheel movement and accordingly adjust the infeed table to match this position? I only need about +/- 5 thousands accuracy . The infeed table has a 110 volt ac motor (63rpm) with an attached gear box (40:1). No brake.

Not sure how to tackle this problem.
I would really appreciate any suggestions. We don't have an EE on the payroll so I'm on my own with this one.
Thanks for any suggestions
Dean
 
This is easy enough

When the laser micrometer detects the bars are getting too wide that is because the grinding wheel is getting to small. If the laser feedback detects an error of 0.001 inch the set point for the hydraulic actuator must be moved in by one. You should be able to get much better accuracy than 0.005 inches.

I would user a hydraulic valve with a servo quality spool and the valve should be small. The valve does not need to be a servo valve, a proportional valve will do but the spool quality is a must. I would use a MDT rod, Temposonic or Balluff, for the hydraulic actuator feedback. 5 micron resolution is easily obtainable. Now what kind of PLC do you have? Rockwell has a M02AS for the ControlLogix that would do the trick. My company, Delta Computer Systems Inc, has very power hydraulic control modules that a programmable and can probably do the whole job without the PLC and much better because the update rates are much higher.

What is the feedback from the laser measuring device? This is key.
What is the feed rate of the metal rod? If there is a lot of distance between the grinding wheel and the laser sensor this could complicate things because we may need to take into account for the fact the laser position is always delayed. If the feed rate is relatively constant we can predict ahead.
What is the rate of the grinding wheel wear per foot of rod?

All this can be taken into account. 0.005 inches should be easy unless there is a lot of mechanical slop.
 
This sounds like a ceterless grinder. It it?

We've got a few automated centerless grinders. The controllers are PLC/2s and even with those dinosaurs we get considerably better accuracy than .005. One of Peter's controllers would do an even better job.
 
I have done some centerless grinders before. We did a manual type control where the operator could just tell the wheel to move in what ever amount they needed. We did this with a Automation Direct 05 plc using the step & dir output to run a stepper. This worked but even with the gear box you can miss steps. Most of these machines don't have linear bearings just dove tail type slides. These can be hard to position. The next one we did we used Baldor Mint drives and servo motors. This worked much better but was more expensive.

I have done a total of 3-4 of these grinders but we never did one with a laser mic. The company I worked for did purchase some that were "closed loop" from Koyo. These were using a Fanuc controller. I know at the time they were very $.

One of Peters controllers would work great for this type of application. I bet you could use an RMC70 with a Red Lion HMI and not need a plc at all.

Hope this helps.
 
many years ago I used to fix grindong machines that were automated. The bars were measured by caliper transducer type instruments and the grinding wheel moved with a stepping motor.
 
Thanks for all the suggestions.
Peter, where can I learn more about the programmable hydraulic control modules that your company manufactures?
Thanks again
Dean
 
Fyi

Peter Nachtwey said:
If the laser feedback detects an error of 0.001 inch the set point for the hydraulic actuator must be moved in by one.

Actually, an error of .001 would translate to an adjustment of .0005.................. .0005 in from BOTH sides of the part..........=.001 total material removed.
Click to expand...
 
Thanks for all the help.
The regulator wheel is tracking satisfactory with setup we have at present time. As the product is detected (by the laser mic) as not being in tolerance (getting bigger)it pulses a hydraulic valve that moves the regulating wheel "in" by a couple of thousands of an inch . The amount of travel on each pulse is adjusted by the human operator (hydraulic regulator adjustment). This works just fine but the operator must also mechanicaly "bump" the infeed table "in" every so often (about every 15 thousands of error between the two). I need to track the position of the regulator wheel and , say, when the difference in position between the regulator wheel and the infeed table equals 10 thousands, move the infeed by ten thousands..short motor pulse from relay..not realy critical positioning. Just needs to be + or - 5 thousands or so with machanical slop. I think we can tune this in using an off delay timer...maybe a logo relay...that I happen to have on hand. This machine is about 30 yrs old so anything would be an improvement over manually bumping the table into position every so often..sorry that I did't explain this application very well from the first post. I am new at these type of problems.. Only been with this company since last july and they keep throwing these projects at me (job security- if I can come through:)
Thanks again everybody for all your patience with me. It is a comforting feeling to know that you guys are out there!
Regards
Dean
 
Back
Top Bottom