e-stop behaviour on ATC!
04 Mar 2024 14:37 #295138
by amartel
e-stop behaviour on ATC! was created by amartel
Hi folks,
I'm currently retrofitting an 80's Leadwell VMC with LinuxCNC, and there is something I can't get my mind around, relating to the Toolchanger behaviour in the event of "mid-change" emergency stop. Let me explain:
The ATC is a carousel Geneva mecanism type. Once the tailstock is at home position, the ATC engages by shifting on the right, using a pneumatic cylinder. The drawbar then releases and the ATC pulls down, then proceed to shift the pocket with the Geneva mecanism, then pulls up. The drawbar locks the toolholder in place, then the ATC shifts on the left to clear the tailstock. No big deal here and HAL components for this kind of setup is well documented here.
However my concern pertains to the ATC behaviour if I would happen to hit the estop while the ATC is positioned right underneath the tailstock. See the solenoids that drive the pneumatic cylinders are all spring-loaded (dual-acting, single output) units. They have either a default position (no output) or an activated position (output = 1) , meaning if I were to cut main power using a proper harwired e-stop loop, BOTH solenoids would get back to their default position at the same time, leading to a catastrophic failure since the ATC would travel diagonally to its standby location. Don't know if it makes sense and I still haven't tested the Hal component so I might be mistaken in my assumptions. There are definitely other people out there who encountered similar challenge on a safety standpoint and I haven't found anything yet on the matter.
My question is then : Would it be possible to keep those outputs out from the e-stop latch and keep their state unchanged until the tool change can be "rescued" after re-enabling the controller? FYI I wanted to simply add a purge valve on the air supply for the machine but that would not solve the issue completely since the caroussel vertical motion is spring-loaded. It would not prevent movement since there is potential energy stored. I must be looking to just keep the ATC cylinders outputs unchanged in the event of an emergency stop...
How you guys would go about this one?... I really don't want to cheap on safety features.
Best regards!
I'm currently retrofitting an 80's Leadwell VMC with LinuxCNC, and there is something I can't get my mind around, relating to the Toolchanger behaviour in the event of "mid-change" emergency stop. Let me explain:
The ATC is a carousel Geneva mecanism type. Once the tailstock is at home position, the ATC engages by shifting on the right, using a pneumatic cylinder. The drawbar then releases and the ATC pulls down, then proceed to shift the pocket with the Geneva mecanism, then pulls up. The drawbar locks the toolholder in place, then the ATC shifts on the left to clear the tailstock. No big deal here and HAL components for this kind of setup is well documented here.
However my concern pertains to the ATC behaviour if I would happen to hit the estop while the ATC is positioned right underneath the tailstock. See the solenoids that drive the pneumatic cylinders are all spring-loaded (dual-acting, single output) units. They have either a default position (no output) or an activated position (output = 1) , meaning if I were to cut main power using a proper harwired e-stop loop, BOTH solenoids would get back to their default position at the same time, leading to a catastrophic failure since the ATC would travel diagonally to its standby location. Don't know if it makes sense and I still haven't tested the Hal component so I might be mistaken in my assumptions. There are definitely other people out there who encountered similar challenge on a safety standpoint and I haven't found anything yet on the matter.
My question is then : Would it be possible to keep those outputs out from the e-stop latch and keep their state unchanged until the tool change can be "rescued" after re-enabling the controller? FYI I wanted to simply add a purge valve on the air supply for the machine but that would not solve the issue completely since the caroussel vertical motion is spring-loaded. It would not prevent movement since there is potential energy stored. I must be looking to just keep the ATC cylinders outputs unchanged in the event of an emergency stop...
How you guys would go about this one?... I really don't want to cheap on safety features.
Best regards!
Please Log in or Create an account to join the conversation.
05 Mar 2024 12:31 #295225
by andypugh
Replied by andypugh on topic e-stop behaviour on ATC!
Look at specifying ON_ABORT behaviour.
linuxcnc.org/docs/stable/html/remap/rema...p:sec:error-handling
linuxcnc.org/docs/stable/html/remap/rema...p:sec:error-handling
Please Log in or Create an account to join the conversation.
Time to create page: 0.061 seconds