Results 1 to 12 of 12
Thread: Mazak 510C Mystery Crash
06-05-2012, 08:11 PM #1
Mazak 510C Mystery Crash
So today, Our 2003 Mazak 510C with 640M (window 95 Based) decided to slam itself into the part on the table.
It had already run this program/Setup 7 times with no problems then out of the blue, going for a toolchange, instead of grabbing tool #7, it grabbed tool #0 and slamed down into Z- untill it hit the soft limit, smashing the Spindle nose with no tool (T0 is the spindle empty) into the top of the part.
Everything was set and running fine and this just RANDOMLY happened. This is very concerning as we have NO IDEA what caused the machine to do this.
Has anyone had something like this happen to them. It seems like it lost its mind for a little bit, but just long enough to cause chaos.
Any help in discovering the cause would be greatly appreciated!
06-05-2012, 08:39 PM #2
Was this in mazatrol or eia? I only run mine in eia, but if I post a program with only one tool I have to edit the post to T1 T0 M6 from T1 M6 or it will start the program with T0 and do as you described. With no next tool designated, T0 defaults and it doesn't even look at the one you told it to get. If I have more tools the post will add next tool like T1 T2 M6, but it won't do it with only one. A button had to be accidentally pressed or it did as you said and lost its mind. There is usually a reason though. I'll look at mine in the morning and see if anything stands out that could have caused this.
06-05-2012, 11:11 PM #3
Our machine accepts a T# M6 but allows you to stage tools in the carousel with just a T#.
It was an EIA program. I hate shit like this.
06-06-2012, 07:41 AM #4
It's the beginning of machines becoming self conscious. It's telling you mazak needs more money.
My fanuc has needed a reboot a few times too, don't know if its a little voltage thing that screwed it right up or what, last time it was telling me some end or record error and that my card was no good, as if a fanuc ever could play a tune. Thankfully no out of nowhere crashy smashy yet.
there's been a few mentions here before of machines out of nowhere even if running the same program for days just deciding to retire themselves with a bang.
Good luck figuring it out. I'd start by any loose wires.
06-06-2012, 08:31 AM #5
I was not able to recreate your situation on mine by pushing any buttons in between tools. Unless block skip somehow skipped the next tool I don't know. The only other thing I can think of that would cause this is tool life management. There are two different ways it looks at it based on parameter F94 bit 4. It almost seems as if the T7's life expired and it picked up T0 as the spare. I would check tool 7 and make sure there is no spare listed of 0, and also that if it has a life cycle listed and no spare is designated it may go grab 0.
06-06-2012, 09:16 AM #6It's the beginning of machines becoming self conscious. It's telling you mazak needs more money.
i blame most unexplainable wrecks on "no reset restart from feed hold after handwheeling machine"
no clue how you can work that into this wreck but something along that line (human error) is always a distinct possibility
06-07-2012, 11:18 PM #7
SND was right....Mazak needs more money.
I'd sell it ASAP, CHEAP, now that you've published that it's crash prone. I'm in the market for a cheap one.
All kidding aside, if you really want to know what happened, I believe you could have a Mazak factory guy come in and they can pull up every keystroke. At least I know they can on our horizontal which has a Nexus control, just for these instances...or more likely potential warranty crashes.
06-07-2012, 11:51 PM #8
Seen a brand spanking new DAEWOO not sure of actual model big 5 axis mill but it had newest fanuc controls in 2003, sounds eerily familiar. The company i worked for was blowing up buying machines literally 3 a week for bout 2 months and building a complex robot cell same time. They contracted the work out couldnt tell u company dont remember but there was actually 2 seperate contractors doing work.
End all be all the DAEWOO was installed ran test run fine had i think 8 production runs on it,the operation was pretty complex it was actually a cradle op in a hard iron housing that this happened on. WORST TIME TO HAPPEN!! The cradle tool was pretty much at max weight for spindle specs(very custom tool) and mid cut mind you this tool comes in housing closed then through macro was engaged and cutting mid cut when machine rapid out....smoked spindle....tried doing tool change....WITH HALF THE HOUSING STILL ON THE SMOKED TOOL!!! spindle..smoked pallet and fixture....smoked tool changer door and outer tool chain WAY SMOKED!! I was way far away huge shop didnt hear anything but the operator had to go home he was so distraught over it, good operator.
END OF STORY after all was said and done it was linked to a VIRUS in software communications. They believed it was one of there engineers used his personal laptop to upload some data to the new machine for an upcoming job he was working on!! Talk to ur best code breaker and backstep everything since u popped the crate!! The company paid over 2mil to rent machine for a month and ate it, junked it for parts.
Dont want to get u down just telling you a true occurance, think of everything possible that communicated with that machine
06-11-2012, 03:05 PM #9
Your comment is very good
06-11-2012, 05:26 PM #10
I had my Mazak VMC do something similar. I program using eia as well and I always ramp to the part with all 3 axis. After about 10 parts, the y axis failed to read anything after the tool change but the x and z still did. Needless to say the spindle came straight down at 100% rapid and crashed right through the y axis slides and into the casting below the slide ways. Mazak blamed it on the machine not properly being grounded. We had it grounded, just not with a copper bar into the ground. We now have a 12 foot bar into the ground. I nor Mazak cannot accurately say it was caused from this but they lean more towards it being the reason. As for the T0 thing, the mazatrol should not go anywhere in z except for up as there is no 0 offset? Unless it called out T0 and called up a different length offset.
06-11-2012, 05:32 PM #11
06-11-2012, 07:24 PM #12