Pneumatic press safety systems 1

[RyreInc]

It has recently come to our attention that a competitor is making presses that, while including a light curtain and emergency stop along with guarding, use single (i.e. no redundancy) un-monitored valves to actuate their presses. These are in the range of roughly 500-1000lb press force.

In the pursuit of cost saving some of us would like start building at least some of our machines with this valving scheme. I know that from a technical and regulatory standpoint this is probably unsafe, but what about from a practical standpoint? What is common practice in industry? What other considerations should we entertain?

What does a safety system on your machine/press look like?

Thanks for your advice!

 

[Compositepro]

OSHA has regulations specific to presses. These are available on the internet. You had better read them if you are in the business.

 

[RyreInc]

Yes I am familiar with OSHA regulations. I was hoping to survey contributors here to see what common practice is.

 

[John2025]

I'm just guessing at your configuration, but if your system depends on the light curtain and electro-pneumatics to protect a worker during production, then I would absolutely require redundancy. I've never had the courage to trust that kind of system. I expect physical guards that comply with Table O-10 of 1910.217 and an effective Lockout-Tagout program, but maybe I'm just a coward.
John

 

[itsmoked]

Often seen on presses are dual hand initiation controls so both hands are needed outside the press to have it cycle.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[RyreInc]

Typically these presses would be operated by a single cycle start button and then rely on the light curtain during press movement.

Even a 2-hand control would not solve this issue. My go-to example is if a 3-way spring-return valve had one of its springs break. The valve will shift and the press may suddenly advance regardless of the state of the 2-hand control (or any safety sensing system for that matter).

Redundant valving would greatly reduce the risk, and monitoring would call attention to the partial failure. Without BOTH of these, the press is not safe.

What about a rodlock or separately-valved PO checks? does this count as redundancy?

 

[BrianPetersen]

Depends how they are actuated. Depends whether there is a single point of failure.

Cross-piloted PO checks are easy to install but are still actuated by the same directional valve ... which is a single point of failure.

Separate PO checks or rod locks actuated by an independent directional valve require two directional valves to actuate for the machine to move (provided that the system is properly arranged). If done in this manner, that independent directional valve would have to be monitored somehow to make sure it is changing state properly in both directions (not just a switch in the reset circuit - that alone won't prove that the monitoring switch is not shorted) because a failure of the pilot valve would not be self-revealing (the machine could still seem to be working normally).

This will not prove that there is not a piece of dirt stopping your PO check from sealing. This will not prove that your rod lock's friction surface is not worn out (I've seen it happen). PO checks cause stored pneumatic energy - even if someone switches off the air to the machine! - which can be a troublesome issue on its own account.

I'm interested in seeing what thoughts others have on this matter also. I have been finding that attempting to apply ISO 13849-1 to pneumatics in which you have to stop something from crashing down by gravity is a nightmare with no completely bulletproof magic solution. But the inability to achieve perfection is no excuse to do nothing. Just cutting solenoid power to one plain ordinary directional valve will provide no protection whatsoever against the spool failing to return.

A risk assessment should address whether mid-stroke stopping of the pneumatic function is a critical requirement. If the spool fails to return, the cylinder will carry through to the end of its stroke. If that's buried someplace deep in the machine with no normal operator access to it, maybe that's not an issue. If the full stroke of a short-stroke cylinder happens so quickly that there's no way someone could get to it in time, perhaps it's not an issue, either. But if the operator is inserting their hands into the tooling on every cycle to insert and remove parts in the fixture and it's big enough and accessible enough to be a hand-crusher ... ! ! !