Automated clamping on a 2 pallet VMC

[Scott_bob]

Here is the problem:

We use vacuum clamping to hold our parts down on our custom fixtures.
A "seal" is installed into the fixture to maximize our surface area of the clamping force, and to accommodate material variation. This "seal" is compressed as the vacuum pulls the part down into contact with the fixture. Locating pins are always used, either along an edge or in holes to prevent the part from shifting.

This all works fine except when the vacuum fails due to a leak in the seal, or someone neglects to connect the vacuum hose, or turn on the vacuum system.
Murphy's Law is just waiting to keep us all on our toes. Ever see a piece of material get thrown around inside a machining center? It happens...

The question I am asking here is this:
Has anyone set up a pressure (in this case vacuum) sensing circuit on your CNC that uses an M-Code to activate this system?

Basically, I want the CNC to NOT move or STOP moving "If" no pressure is "seen" in the vacuum clamping system. I am sure it would be the same for hydraulic clamping systems... It is a simple safeguard system that can save a lot of injury to personnel, equipment and tooling...

If you have any experience with this, let me know.

[Geof]

Assuming your equipment has microswitches to prevent operation with the doors open you could just get a simple vacuum sensing switch and wire it in series with the door switch. No vacuum is then equivalent to the door being open. It depends on how your machine responds but with my equipment opening the door with the machine running gives a feedhold and stops the spindle.

[DareBee]

Great idea and simple too

[bradleync]

scott_bob,

Only problem is the alarm message you get from opening the door would be displayed if clamping is not active. This would have nothing to do with the cause of the alarm. Could be confuseing.
Are you going to be using this to activate the vacuum or just moniter the vacuum force?
Not sure if one M-code would be enough...

[Geof]

scott_bob,

Only problem is the alarm message you get from opening the door would be displayed if clamping is not active. This would have nothing to do with the cause of the alarm. Could be confuseing.
Are you going to be using this to activate the vacuum or just moniter the vacuum force?
Not sure if one M-code would be enough...

Just monitor the presence of vacuum; if the operator forgets to flip the valve you get a door open message when cycle start is pushed. If you have only two choices; door or vacuum it is not difficult to sort out which is the cause of the alarm.

I don't see how you would do it with an M-code unless you used the M-fin which can close a relay to activate an external operation and then sits waiting until it detects a contact closure indicating that the operation is complete before continuing running the program. Or at least that is the Haas sequence. The problem I see with doing it this way is that once the M-fin sequence is complete the controller stops looking at those contacts so if you then lose vacuum for some reason after the program has started running it will not stop it.

As far as I can see your choices are tie into either the door circuit or E-stop; I don't like E-stop because you have to reset and start from the beginning. With the door alarm you have the chance of being able to restore the vacuum and continue; if the problem is not obvious and cannot be corrected to continue then hit reset.

[Scott_bob]

I Think we were going to use an M-code to activate the Vacuum force.
That way until the M-fin signal we cannot get motion.

But after that, I like the idea of tieing into the door open circuit because we would get like Geof says, motion hold and spindle off, if vacuum clamping is lost. E-Stop is not desirable except for safety, because it is not as easy to recover from. Thanks Geof!

Keep up the comments! We all benefit...

[fpworks]

It might be easier to use part detection sensors, such as a microswitch that verifies the part is fully seated against the fixture. This could also be used to detect a part that is not properly installed in the fixture.

Problems with this idea: that you'll need a switch for every position on the fixture, you'll have to modify your fixture for the switches, and the operator will have to connect/disconnect a cable for every pallet change (or go wireless)

Just throwing around another idea...

[Geof]

I Think we were going to use an M-code to activate the Vacuum force.
That way until the M-fin signal we cannot get motion.

But after that, I like the idea of tieing into the door open circuit because we would get like Geof says, motion hold and spindle off, if vacuum clamping is lost. E-Stop is not desirable except for safety, because it is not as easy to recover from. Thanks Geof!

Keep up the comments! We all benefit...

It took a few hours but eventually I realised it would not work to combine both.

You have to use either M-code/M-fin OR sensor in door switch circuit. If you use both your cycle will never start; the door hold will be active at cycle start because there is no vacuum so you will never get to read the M-code to activate the vacuum.

[Neal]

Another possiblity might be to use the I/O macro I 1 thru 6 inputs. (see Macro Section 18 of the User's Manual) This could be hooked to a vacum switch and when the vacum switch activates it can be checked via an "If... Then" conditional statement. Not quite sure how it would work but on the surface it looks promising.

Neal