What's new
What's new

Light out machining - what extra things do I need to do?

aarongough

Stainless
Joined
Oct 27, 2014
Location
Toronto, Canada
Hey guys!
I'm looking at transitioning the bulk of my production machining to 'lights out', mainly because this will free up the VMC for my use during the day to do prototyping and so on... It should make a lot better overall use of my VMC.

The program that will be running overnight is one that's machining multiple operations on a single part family on a single fixture. Mostly hard-milling with some composite machining at the end. The hardmilling is done dry with an air-blast, the composite machining is done under vacuum. Cycle time will be around 4 hours.

All my parts are run on quick-change fixture plates, so there's no concern about running the wrong offsets as they never change. My process is stable enough now that I never tweak offsets or anything, all I'm doing is basically changing to new tools every 9-10 parts. My programs are setup to change to an empty spindle at the end of the program to make sure tools don't get stuck in the spindle as it cools...

I'm planning on getting a internet accessible camera setup in the shop so that I can monitor the machine as it's an older one (Fadal VMC10) that doesn't have monitoring capabilities.

What other things should I be looking at changing/implementing to make the transition to 'lights out' smoother?

Thanks guys!
-Aaron
 
Hey guys!
I'm looking at transitioning the bulk of my production machining to 'lights out', mainly because this will free up the VMC for my use during the day to do prototyping and so on... It should make a lot better overall use of my VMC.

The program that will be running overnight is one that's machining multiple operations on a single part family on a single fixture. Mostly hard-milling with some composite machining at the end. The hardmilling is done dry with an air-blast, the composite machining is done under vacuum. Cycle time will be around 4 hours.

All my parts are run on quick-change fixture plates, so there's no concern about running the wrong offsets as they never change. My process is stable enough now that I never tweak offsets or anything, all I'm doing is basically changing to new tools every 9-10 parts. My programs are setup to change to an empty spindle at the end of the program to make sure tools don't get stuck in the spindle as it cools...

I'm planning on getting a internet accessible camera setup in the shop so that I can monitor the machine as it's an older one (Fadal VMC10) that doesn't have monitoring capabilities.

What other things should I be looking at changing/implementing to make the transition to 'lights out' smoother?

Thanks guys!
-Aaron

i have used a Mazak with afc or automatic feed control. each tool is set on tool page for a % of spindle turning load and spindle thrust so for example.
...drilling a hole, drill getting dull, machines senses high load and auto reduces feed to 50% gradually. if overload continues 5 seconds it does a feed hold and drill will turn but with no feed it just hovers.
.
same with facemills if overload continues 5 seconds after feed reduced to 50% it does a feed hold
.
my opinion it is the most needed thing you need.
 
i have used a Mazak with afc or automatic feed control. each tool is set on tool page for a % of spindle turning load and spindle thrust so for example.
...drilling a hole, drill getting dull, machines senses high load and auto reduces feed to 50% gradually. if overload continues 5 seconds it does a feed hold and drill will turn but with no feed it just hovers.
.
same with facemills if overload continues 5 seconds after feed reduced to 50% it does a feed hold
.
my opinion it is the most needed thing you need.

I actually don't use any large tools. Largest tool is the rougher I use for the composite which is 1/2"... None of the tools manage to register at all on the load meter even when running full clip, they all read the same as the spindle when cutting air (around 25% load).
 
A few things.. The goal now goes from as many parts as possible in as little time as possible
to as many GOOD parts as possible by morning..

Your QTY is limited by what you can get on the table, so there is your quantity..

Now you want a nice stable process where you won't have to worry about breaking tools..
Seems like you are confident on that front..

I'd say you are pretty good to go.. The one thing you may want to consider, since the
pallet doesn't have to run as quick as possible, it just has to be done by the time you
get back in the morning.. Can you tweak (as in slow down) your process to increase your
tool life?? Who cares if it takes longer, it just has to be done by the morning.. That
would be a win win, free parts off the machine, plus reduced tool cost.
 
Aaron, you might want to look at what comes out the serial port when the machine encounters a fault. The console is echoed to the serial port.

BTW, they use standard XOn/XOff flow control, which is implemented by the pyserial library.

You can use a Raspberry Pi as an internet accessible camera. That or an old Android cellphone with charger.
 
Sometimes, especially with an impeller which might have 20-30 hr cycle time, it's tempting to let it run overnight. Several years ago, I had my air compressor burst into flames. Fortunately I was there and snuffed it out. But to this day I can't get comfortable leaving the compressor on all night.
 
I used to run lights out on my bar fed lathe a lot. If you have a stable process, which it sounds like you do, just make sure your fluids are topped off so you don't alarm out due to lack of way lube or something stupid.

How far is it to your shop? Load up, go home, come back 4hrs later and load up again. Instant 2nd shift.
 
Sometimes, especially with an impeller which might have 20-30 hr cycle time, it's tempting to let it run overnight. Several years ago, I had my air compressor burst into flames. Fortunately I was there and snuffed it out. But to this day I can't get comfortable leaving the compressor on all night.

If you weren't there, how bad would it have been?
 
If I had the means to watch a process, I would want a way to stop it, via remote E-stop.
Wouldn't mind having a push-button fire extinguisher, too.
There's probably a multi-relay cell or internet-enabled doo-dad that can take care of those items.

Chip
 
At a shop I used to work at we ran lights out all the time on one machine. There was a macro that checked tool height (for tool breakage) and the program would stop if the tool was broken. A camera is a must. Once a coolant line sprung a small leak and basically emptied 60 liters of coolant on the floor.
 
having one guy at night keep a eye and ear on 2 or more running machines is fairly cheap. also guy can be setting up next jobs, taking done parts off and deburring. changing tooling etc.
.
most night shift guys are older (any kids are grownup) and they usually get $1 or $2 more per hour working the night shift. some work 3am-3pm and 3pm-3am or rather day shift guy comes in early and evening shift guy stays late.
 
having one guy at night keep a eye and ear on 2 or more running machines is fairly cheap.

Also defeats the point of "lights out" manufacturing.

If you start with an 8 hour shift where the machines run for 8 hours, your spindle-hour:labor-hour ratio is 1:1.

You could have a day-shift do nothing more than 8 hours of parts changing and setup, with the red lights on all shift, but hit the green buttons on the way out the door, and if they run all night, you get 16 hours run time, or 2:1 ratio. You've already doubled production. Use the labor time to ensure the late hours don't /need/ labor time.
 
We run lights out on occasion. On most of our long run jobs we are using a bunch of small tools that will break if you look at them funny, its nice to be able to run broken tool detection. I'm close to the shop and usually stop by on my way home from being out at night.

If your tools are as stable as you say I wouldn't worry about it, at most you'll come in to $150 in scrap tools and lost time? Taking the risk of breaking all your tools is paid off pretty quickly in a free second shift.

The air compressor concern is a legitimate one, I have heard stories that make me cringe. Maybe take some preventative measures with where you locate it and inspect it routinely.
 
I saw mention of broken tool check methods - Renishaw probes and tool setters are great for preventing SNAFUs. Before tapping a hole, make sure there isn't metal where you expect a hole (just an illustrative example, not something likely to be needed)

I didn't see anyone mention tool groups or whatever your machines would call it. Using a 1/2" EM on the part? What's the tool life? If you have a good idea how many parts it's good for, set it up to switch to another tool after 'x' parts on the counter. Set up three or four 1/2" EMs and don't touch it for a FEW days, even. Great for taps, too. That way you don't run an end mill /until/ it breaks, but you switch it before. You can also use that in conjunction with the broken tool detection in case of breakage.

Nothing like leaving work at 4pm, coming in at 7am, and finding out the machine alarmed at 4:30 pm and sat waiting.
 
Also I'm a one man shop... So there isn't even a 'day shift', it's just me and one machine. The idea behind moving to lights-out is to make the machine more available to me during the day basically. And also to reduce the amount of time that I spend in a really noisy environment!
 
I saw mention of broken tool check methods - Renishaw probes and tool setters are great for preventing SNAFUs. Before tapping a hole, make sure there isn't metal where you expect a hole (just an illustrative example, not something likely to be needed)

I didn't see anyone mention tool groups or whatever your machines would call it. Using a 1/2" EM on the part? What's the tool life? If you have a good idea how many parts it's good for, set it up to switch to another tool after 'x' parts on the counter. Set up three or four 1/2" EMs and don't touch it for a FEW days, even. Great for taps, too. That way you don't run an end mill /until/ it breaks, but you switch it before. You can also use that in conjunction with the broken tool detection in case of breakage.

Nothing like leaving work at 4pm, coming in at 7am, and finding out the machine alarmed at 4:30 pm and sat waiting.

I actually just changed that 1/2" tool for the first time since I started machining that part... Over 100 hours of tool life :) Only possible because I'm machining composites in that case.

The rest of my tools have very predictable lifetimes, the shortest being about 10 parts and longest being about 50 or so right now. I've actually never broken a tool in this process so I'm not terribly worried about that. Even if I did and then broke the follow-up tool the worst it would cost me is like $50-$60 or so, probably less.

All my processes are optimized for producing very fine surface finishes, so relatively small stepovers and so on, which again reduces the risk of breaking a tool.
 








 
Back
Top