Pneumatic safety

kolyur

Lifetime Supporting Member + Moderator
Join Date
Oct 2004
Location
Wooster, Ohio
Posts
1,602
I'd like to hear some opinions on how to handle safety of pneumatic actuators on machinery. Traditionally, most of our equipment would be surrounded by hard guarding. When a guard door is opened, the main dump valve would release--no safety issue. However, with some newer equipment we are using light curtains and don't necessarily want the air to dump every time the curtain is broken. Ideally I'd like the machine cycle to pause until the curtain is reset, then continue. We also use servo actuators and robots, but these have hardwired safety such that they are mechanically locked out and will immediately stop whenever the curtain is tripped.

The pneumatic devices are trickier, however, because at any point in the cycle, some of them will be actuated and power needs to remain on those valves to keep them in position (we typically used single solenoid spring return valves). So I can't just dump the air or drop power to the output card. What I've been doing is writing my PLC logic in such a way that the valves cannot shift if the curtain is tripped. This works but makes me a little nervous because I'm handling a safety function through a non-safety programmable device.

Is the solution to use double solenoid valves, so that the valve signal can be dropped without shifting the valve? What part should a risk assessment play in this decision? We do light assembly and most of our pneumatic actuators are small enough that they couldn't cause more than a swollen finger.

Thanks for your suggestions.
 
Ideally I'd like the machine cycle to pause until the curtain is reset, then continue
It should not be necessary to reset before continuing, if the light curtain is interrupted in normal operation, i.e. for inserting or removal of parts to be processed. Reset is required for E-stops.

Is the solution to use double solenoid valves, so that the valve signal can be dropped without shifting the valve?
No, the actuators would still be "armed".
You could have 5/3-way valves where the air is released in the middle position.

What part should a risk assessment play in this decision?
Everything. :)

We do light assembly and most of our pneumatic actuators are small enough that they couldn't cause more than a swollen finger.
You would have to verify this. Even a small diameter cylinder can have enough force to cause harm if there is a pin or edge.

I am guessing that the internal friction of the pneu cylinders will limit the speed so much that they cannot cause harm, if they move by their own weight when air pressure is completely removed.

I am also guessing that the real problem is that if the pneu cylinders are interrupted at the wrong time, it can be a problem that parts are not processed completely, thats a production issue not a safety issue.
Maybe you could solve this by some kind of mechanical brake that holds the cylinders in position even with air removed.
 
I don't know that I can help answer your question, because I always dump the air.

On machines where I want things to stay put, I'll use a combination of double solenoids, and pilot-operated check valves (I like Norgren's) at the cylinder. This does a great job of keeping things in place.

Of course, if the operators break the light curtain at random, you'll still have issues, as some cylinders may be at mid-stroke... :(

🍻

-Eric
 
You could have a traffic light to help operators from preemptively interrupting the sequence.
And if they do interrupt the sequence at the wrong time, it may be a question of having the right dialogue on the HMI. I.e. "Light curtain interrupted before part was finished. Continue processing part ?".
 
On machines where I want things to stay put, I'll use a combination of double solenoids, and pilot-operated check valves (I like Norgren's) at the cylinder. This does a great job of keeping things in place.
I am familiar with those check valves. The problem I have with them is that they retain the air in the cylinder even if the E-stop is triggered and the main air is dumped. So if someone gets trapped, the only way to release the pressure is to disassemble the fitting or hose. Or, power the machine back up and shift the valve manually (not likely when everyone's in panic mode).
 
the only way to release the pressure is to disassemble the fitting or hose
Festo sells a part that is placed between the cylinder and check valve. It has a little button that will release the air when pressed. It can be very handy to have a manual release option
 
I am familiar with those check valves. The problem I have with them is that they retain the air in the cylinder even if the E-stop is triggered and the main air is dumped. So if someone gets trapped, the only way to release the pressure is to disassemble the fitting or hose. Or, power the machine back up and shift the valve manually (not likely when everyone's in panic mode).
You could add a manual valve between the cylinder port and the check valve as a bleeder (I'm picturing a push-pull, maybe even a red, mushroom-head?). This would also be handy for times when you need to move the cylinder by hand for maintenance, etc.

🍻

-Eric

EDIT: I see AJZ beat me to it!... :D
 
You could add a manual valve between the cylinder port and the check valve as a bleeder (I'm picturing a push-pull, maybe even a red, mushroom-head?).
Actually that's exactly what we did on the existing ones. The mushroom head was originally white but we painted it red. :)
 
HUMOR ON
Kolyr said
"We do light assembly and most of our pneumatic actuators are small enough that they couldn't cause more than a swollen finger"
OK we will let you try it out with YOUR finger.
HUMOR OFF

I think a major component of your problem is unstated ie why do you have this concern and under what conditions?

IF maintenance work then machine should be locked out and tagged out, all cylinders bled down where possible or jammed with a mechanical block.

The problems come about when doing troubleshooting with the machine running or doing adjustments either by maintenance or operators. Even here I prefer not to rely on automatic controls and sometimes install a manual valve that I can control to make it (one component or a subsystem) do what i want and not have to worry about rest of machine.
 
HUMOR ON
Kolyr said
"We do light assembly and most of our pneumatic actuators are small enough that they couldn't cause more than a swollen finger"
OK we will let you try it out with YOUR finger.
HUMOR OFF
Good point, and the answer is YES I would try it with my finger. Consider the following hypothetical:
* Pneumatic actuators on machine have injury potential as mentioned above.
* If E-stop is triggered: all air is dumped.
* If light curtain is tripped: air remains on, PLC outputs remain on (or off), PLC logic has input for light curtain and prevents valves from shifting in logic. Essentially, safety is handled in the PLC.

The whole root of my question is: when is this setup acceptable? Personally, in the above situation, I would trust that the PLC logic would not fail when I'm sticking my hand under a cylinder. HOWEVER, if the cylinder was a 4" bore with a knife attached, I'd have a different mindset. For those of you who would say that this control scheme is never acceptable: What if the cylinders were so small that they could only cause some discomfort but no real injury? And is one man's discomfort another man's injury? o_O

I guess I could phrase my question this way: How do you use the risk assessment to determine the appropriate safety controls?
 
The way I addressed a similar problem a few years back was to have a safety guard that closed just behind the light curtain. Basically the light curtain kept the operator from being hurt by the guard. The hard guard protected the operator from the actuators that could not be rendered fully safe because they may be in motion. The customer was obviously happy enough with the machine that they have been talking to me about adding another one. It'll be a nice little side job because all the control system design and programming are already done. I like those.
 
The pneumatic devices are trickier, however, because at any point in the cycle, some of them will be actuated and power needs to remain on those valves to keep them in position (we typically used single solenoid spring return valves). So I can't just dump the air or drop power to the output card. What I've been doing is writing my PLC logic in such a way that the valves cannot shift if the curtain is tripped.

If the danger/trap is removed by the cylinder being returned I see no problem with this. It depends on the way it is piped/wired. However, there may be a secondary accident waiting to happen with the cylinder return, finger trap may be. You may consider shrouding/ guarding on the the cylinder return. Why do the cylinders need to remain in place can you not return them sequentially. You must do everything within reason to prevent possible accidents. You have not stated how far the light curtain is from the moving part, is it possible to break the curtain and interfere with the moving part my guess it is, so given the scenario is a light curtain adequate for the job. Maybe as recently suggested a guard door may be appropriate. You may also find that sufficient training for the operator will negate the need for said door and light curtain providing it is locked out for other users. There is a whole load of scenarios that can cause accidents from life threatening to finger cutting, when they happen we try to put prevention in place, trick is, put prevention in first. Regards Badger.
 
My opinion - and it is only that - is that as long as the cylinders cannot change state when the light curtain is triggered, it will be correct.

I work with several robots and the last thing you want is products falling from the pickup head - especially when they are 50kg sacks.

I would also try to make tripping the light curtain a PITA to persuade the operator to press the control stop rather than using the curtain "Because it's easier".
 

Similar Topics

I have an actuator operating a clamp that requires PLd/SIL3, I’ve seen some SMC double valves with feedback, but they are described as dump...
Replies
9
Views
2,360
Hello, According to ANSI B11 and any other applicable standards in the US, is there a written standard on collaborative safe force of an air...
Replies
5
Views
2,347
I'm looking for some input on a problem I'm having with some long air cylinders and two hand safety. Right now I'm using a two hand safety relay...
Replies
13
Views
5,644
Any thoughts on this one? I'm designing the controls for an operator-loaded, light-guard protected hydro-pneumatic press (pneum approach...
Replies
3
Views
4,385
Dear experts, i have pretty basic/ dump question, can someone help me to figure out how to read/ what is this? i assume it is solenoid vavle...
Replies
4
Views
1,062
Back
Top Bottom