Machine safety issue 1

[itsmoked]

I can't figure out where to post this so here it is.

I've got a machine that has a cantilevered two inch diameter axle six feet long. To support its load at a couple hundred RPM it needs to have BOTH ends supported. But to unload the product one end needs to have all support on one end go away frequently.

To do this they use a pneumatic cylinder to drive the support arm thru about 90°.

This is a looking 'straight down' view.

The problem is that this nasty arm is about 16 inches long and so sweeps a pretty substantial area. You don't want anyone or thing to be in the swept area when it does its thing.

Currently they have an optical retroreflector light beam switch that looks parallel to the closed (engaged) arm at a corner reflector screwed to the wall. If it's blocked the arm can't swing.

This is lousy because someone can safely be standing there well out of the arm's path and the process can be blocked causing confusion. Furthermore the arm sweeps over a table surface. Stuff can be placed there that isn't seen by the 2mm light beam with predicted results. Also someone can actually stand to the right just far enough to still get batted on the sweep though unseen by the beam.

In a perfect world I'd love some sort of light curtain that would flesh-out the entire swept area in 2D and yea or nae the cycle.

In a less than perfect world I'd like something to do at least a better job than one tiny beam path at protecting people. A straight look-down could be a problem because the operator is all over above the arm during product unloading and blank loading.

There could be something -not too heavy- four feet above the arm on a support?

Let's see what you've got.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[BrianPetersen]

This is my line of work.

Area scanner, scanning the floor underneath the hazard area (and extending out some distance).

https://www.keyence.ca/products/safety/laser-scanner/sz/

The older-tech way of doing this is a safety mat.

https://www.rockwellautomation.com/...ts/safety-presence-sensors/440f-matguard.html

I wouldn't recommend these unless the optoelectronic device is absolutely not feasible.

Or, if the area in question is rectangular or can be made so, and you can have a place to put both the sender and the receiver, a light curtain, horizontally above the floor.

https://www.digikey.com/catalog/en/partgroup/c4000-select-safety-light-curtains/73864

And you will of course need to shut down the air supply to that cylinder if the safety device is tripped.

https://www.smcusa.com/products/VP-X555-Residual-Pressure-Release-Valve-ISO13849-1~136476

And the relevant safety relays and control circuitry to make this all work.

 

[BrianPetersen]

The nice thing about the area scanner solution is that it doesn't have to be physically directly underneath the thing that you are trying to protect. It could be off in one corner of the scan zone.

 

[LittleInch]

Mechanical engineering general might have been a place to put it, but this sounds like an incident waiting to happen.

A photo or other drawing might be needed here to understand the limitation, but why isn't this physically caged off?

I think you need multiple safety devices here, not just one sensor.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[controlsdude]

I think you need to put it back into mechanical area. Can you not get mechanical to have a swing door around the arm that is rotating out by 90? Something that is a a box with a door that slides up and down to close the swing arm into its box.

Ask yourself if you don't at least do below would some operator or maint get hurt?
LS1 estop enable - where door is closed and axel is ready to rotate
LS2 estop - where the swing arm is swung out at 90 degrees and the door slides down and a locking Limit switch with key.
LS3 estop enable - where if the swing arm is in 0 degree position estop is enabled.
Air on or enabled - estop enable when air is unlocked but in series with LS1 and LS3.

 

[itsmoked]

Brian; Thanks for the links. As I mentioned there is a table about a foot below the swing arm. I think we may be able to put a scanner that sweeps the area under the arm. Scan a sector of maybe 80 degrees. That would be a large improvement.

Controls; I'm not seeing it. They have ten of these that have been running with the mickey-mouse single beam sensors since about 1950 and never had more than a bruise out of the setup. It's normally not a possibility at all because the operator has to be out of reach of the swing arm to activate the process that leads to the swing occurring. Only when someone is 'visiting' the process is there a hazard. Operations require a large mass to be manually horsed off the axle thru where the swing arm is. Trying to box that all in adding more mechanical apparatus is probably not going to fly.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[Compositepro]

Having a shaft end support that needs to be removed to unload a roll of material is not uncommon. In the examples that I have seen, the end support is moved manually by the operator. If the operator must manually remove the roll anyway, there is no advantage to powering the end support movement, and you remove the safety hazard.

 

[tygerdawg]

This would be a useful case study for my industrial safety class.
Steps I would present to my students:
[ul]
[li]OSHA directive and guidance on Machine Guarding is your friend, not your enemy[/li]
[li]Perform a Job Hazard Analysis on this. It will probably resolve to a SEVERE rating or even worse a "What the heck were you thinking, Bubba?" rating . Depending upon your company's policy response to ratings, actions required.[/li]
[li]Modern safety admin approach:[/li]

[li]Eliminate the hazard[/li]
[li]Substitute actions to reduce the worker exposure to the hazard[/li]
[li]Administrative controls (operator instructions & training)[/li]
[li]PPE (last resort)[/li]​

[/ul]

Elimination could be achieved by physical barriers, safety cages, laser scanners + control systems, etc. Modern laser scanners are all over the market and provide a very cost-effective solution despite their initial cost.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[BrianPetersen]

Yup. Risk assessment. Kinda what I was implying by suggesting components that are structure category 3 or 4, PL d or e per ISO 13849-1.

 

[Sparweb]

This problem starts as a mech engineering problem.
With the info I have now, it looks like you're allowing the mechanical problem to exist, and adding another ME problem, plus an EE solution.

Would a "cantilevered six (6) inch diameter axle six feet long" make the swing gate problem go away?
Note: not all axles are solid cylinders
Note: not all tubes have thin walls

The swing gate is misaligned with the end of the axle once it isn't exactly 90 degrees to the axle. The ability to release and open the gate may be a limitation to the support it can offer. It also becomes a potential cause of binding the axle after it's been opened and slammed shut 1000 times by the highest-paid employee who uses the machine.

Access to the end of the axle (apparently a worker will be accessing it every time the gate is opened) is still limited by the gate. Without knowing what the operation is or what's being manufactured, I suggest it's possible that a slight increase in access to the work will increase productivity, paying for changes by increased output.

Are there any other ways to offer access to the parts on the axle?
Retract the axle axially to the left,
Remove the right-hand support by retracting it axially, rather than hinging,
Split the right-hand support in half like tongs,
Extend/retract the right-hand support in halves from top and bottom (or two sides)

http://www.sparweb.ca

 

[3DDave]

Larger version of the problem?

https://www.youtube.com/watch?v=JwVLHDcVtXM