gear reduction

[ptonsparky]

We have a customer that uses some low speed conveyors for moving product through an enclosed cooler. Occasionaly a slat will break free and begin to tear the heck out of things. OC protection on the motors do no good for this problem because of the gear reduction. We have motion detection on the end roller but by the time it stops things are bad. I am leaning towards counting the chain links on each side of the conveyor so I at least know when one side finally breaks. How do others do it? Forget PM.

 

[cadpoint]

Mitubuishi (sp) pattened laser motion motion detection, I worked with the man that submitted it. Maybe you could shoot a beam down or across this product, and allow the beam detection to time out for a limited time.
His patten was for laser interuption, acting as gate around machines, break the beam the machine turns off.

Yes, you could use motion limit switches as your detection, but depending on the enviroment a simple maintaince of cleaning the lenses of the laser might just be a lot easier.

While I sudo missed exactly what the problem is, I'm not so sure a chain or some mechanical measurement device would be as good an over all fix verses the light detection, frankly these parts have a life cycle limit, weight and wear and tear might also be a big consideration.

In either case your product can not be shout down with product engaged on either future type of dection device engaged. Yes it can be done but it will have to be programmable.

Sounds like a fun project, enjoy and hope my thoughts help!

 

[Jraef]

So when one slat breaks, your chains move independently? I'm trying to envision how counting links would help. If a broken slat would cause the chains to change relative travel position, I like that idea.

Do the breaks happen in the cold area? If so, things like laser detection may not work (frost, condensation etc.) unless they can be heated from inside. Sick Sensors makes a laser line scanner that you could use to look at a vertical profile of the products passing by underneath on the conveyor and detect a change in product position or spacing. But that would be very expensive.

I think you were on the right track, look at the failure and think about anything that changes when it happens but doesn't happen otherwise, then solve for a simple way to detect that change.

 

[ptonsparky]

Cooling is a relative term. From roughly 185 deg to target of 10 over outside ambient. Product is dusty. I do not believe I can detect a broken slat in time to prevent at least one chain from breaking but shutting down at that point would be a plus.

Also having trouble with fire detection but that is another problem. Last heat sensor failed to shut things down.

Yup lots of fun.

 

[BJ Conner]

It's hard to tell what to do without a picture.
Here's link to some run-out switches. These or something similar might be used to pick up a dropped slat. There is a world of limit and proximity switches out there.

http://www.crouse-hinds.com/catalog/pdfs/00417.pdf

 

[LarryFine]

We have motion detection on the end roller but by the time it stops things are bad.

It sounds like a mechanical disconnect or clutch of some kind would be faster at minimizing the physical damage than shutting down the motor.

I think you should fly a few of us out there to see the machine with our own eyes.

 

[BJ Conner]

here is another path. I have not used these on motors as small as I assume yours to be. There are solid state relays than can detect over current that is just above normal load and generate an alarm or stop signal. They can tell you when your motor is running unloaded ( broken bolt or shaft).
I recommend you call your local AB, Eaton rep and tell them you want a solid state motor management relay.


http://www.abb.com/product/seitp329/831b464ea227488fc12572fe00302388.aspx

http://www.eaton.com/EatonCom/Marke...orInsightOverloadandMonitoringRelay/index.htm

 

[ptonsparky]

...I think you should fly a few of us out there to see the machine with our own eyes.

Absolutely. Be prepared to cringe and shudder. This is the real world.

 

[rattus]

It sounds like a mechanical disconnect or clutch of some kind would be faster at minimizing the physical damage than shutting down the motor.

I think you should fly a few of us out there to see the machine with our own eyes.

Older garage door openers used an adjustable slip clutch. Later models use an adjustable electronic system.

 

[eric7379]

A variable torque clutch would be the way to go.

We used to have a similar problem on one of the machines at the facility that I work at. If there was a product jam, the chain would break before the jam was detected by sensors.

We installed a variable torque clutch that is pneumatically operated. Depending on how many product streams were being used, the air pressure was increased or decreased. We tied the control for this in to the PLC and HMI so that the control was automatic (as the number of product stream was increased by the operator, the pressure automatically increased).

Of course, some engineering would be needed. You would need to figure out what the "breaking point" would be for the chains or whatever else is at risk of being damaged. On ours, it is set to release well below that threshold. We also have a proximity sensor looking at the clutch plate so that if the clutch ever "releases" during production, the drive is given a stop command.

This setup has saved a lot of downtime since it has been installed.

Of course, it is kind of hard to come up with a viable solution without seeing what it is that you are referring to.

 

[hurk27]

Back in the "70's" I worked for a company (Fiber Bond) that made non woven fiber products, the material was basically glued together by a binder that was for the most part just a paint, after each application of the binder (2) it would pass through a large oven for curring (3 passes 3 conveyors), this conveyor used nothing but thin metal tubes about 1" in diameter flattened on the ends with a hole drilled in each end that bolted to the drive chains, chains were keyed to the drive shaft so they wouldn't change position wit the other side, well we have the same problem, one of the tubes would break loose and destroy half of the conveyor before it could be stopped.

Because this conveyor was operated at different speeds for different products, trying to use some kind of load detection was fruitless, we tried slip clutches, a spring loaded tensioner with a limit switch, all failed to stop the line, so we turned to optical, because the tube would fall down between top and bottom of each pass, it was possible to mount the IR transmitter and reflector inside of the conveyor, we put one on each side about a foot from the side, and aimed them in a cross fire pattern to the reflectors on the other end, if anything was to drop in their path it would trigger an EMG stop and light a conveyor fault light we set up.

The drive motor was DC so dynamic braking was employed to bring it to a stop as fast as we could, it would still sometimes take out a couple more bars, but at least they wouldn't get all wrapped around everything, as it turned out it also solved another problem that if the fiber web product broke (operator set speeds wrong to other conveyors) it would stop the whole line without the fiber web jambing up and doing more damage.

The line was a continuous run and speed sinked line so if one part had to be shut down it all had to be shut down.

With an air flow type system, I would presume the space inside of the conveyor would be open similar to the drying ovens I had?

 

[ohmhead]

We have a customer that uses some low speed conveyors for moving product through an enclosed cooler. Occasionaly a slat will break free and begin to tear the heck out of things. OC protection on the motors do no good for this problem because of the gear reduction. We have motion detection on the end roller but by the time it stops things are bad. I am leaning towards counting the chain links on each side of the conveyor so I at least know when one side finally breaks. How do others do it? Forget PM.

Well cant solve your issue but if you think counting the links of that chain would be needed think Hall effect sensor they work in the sub temps or high heat range and fast or slow speed they can work directly with your logic controller with a digital output signal it will count the links one at a time .

 

[ptonsparky]

You guys are making the wheels turn. Good info. Thank you.

 

[ptonsparky]

I think I have talked them into a small PLC & HMI to track temperatures for fire shutdown, and control VFDs at 2 locations (a previous thread). Another input or two plus a couple more HMI screens. Love it.

 

[wptski]

A variable torque clutch would be the way to go.

We used to have a similar problem on one of the machines at the facility that I work at. If there was a product jam, the chain would break before the jam was detected by sensors.

We installed a variable torque clutch that is pneumatically operated. Depending on how many product streams were being used, the air pressure was increased or decreased. We tied the control for this in to the PLC and HMI so that the control was automatic (as the number of product stream was increased by the operator, the pressure automatically increased).

Of course, some engineering would be needed. You would need to figure out what the "breaking point" would be for the chains or whatever else is at risk of being damaged. On ours, it is set to release well below that threshold. We also have a proximity sensor looking at the clutch plate so that if the clutch ever "releases" during production, the drive is given a stop command.

This setup has saved a lot of downtime since it has been installed.

Of course, it is kind of hard to come up with a viable solution without seeing what it is that you are referring to.

Being a retired machine repairman from the auto industry I have seen much machinery with setups like that. It's the way to go.

I remember one machine years ago that had a ball detent clutch but no provision to shut the motor off. The machine wouldn't wreck bu it destroyed itself in no time.

 

[billsnuff]

how about a brake on the motor / gear box and a sensor to detect the presence of the chain. if the chain goes slack when the slat breaks free, a light beam might work seriesed in with the stop.

 

[bpk]

We would mount two proximity switches, 1 on each side of the slat, and set them as separate inputs into our PLC, and had them programmed to a timer so if one is made without the other for a determined amount of time it would shut down