Possible to automate swapping a long tool in and out of lathe?

[tonymor]

I'm pretty sure I remember seeing a video a few years ago where someone had set up a robot to put a really long drill/boring in and out of a horizontal lathe.
I've done some looking now and I can't find anything even mentioning using a robot to do something similar.

We are making some long rods for hydraulic cylinders that require 40mm holes drilled up to 500mm deep.
The problem is the long drill needs to be put in and removed mid cycle so that it does not crash into the back panels/chuck when doing the rest of the machining.

We will be installing a robot fed lathe to do this work. Right now the idea is to have two lathes, with one dedicated for drilling.

Does anyone have suggestions for automating the process of putting the long drill in and out?
We don't need some amazing contraption that can change all of the tools on the lathe, just this one long drill...

 

[boosted]

Several builders offer long tool changing/storage as an option. Typically the turret just drives over and drags the tool out of a special pocket in the sheetmetal.

You could probably accomplish the same thing (albeit less elegantly) with a robot tender and a capto interface on the turret. I don't see any capto blocks online that you can actuate via mcode, but it probably wouldn't be too hard to build your own, if you had to.

 

[Ox]

Actually, I am sure that seen Haas doo this on a #4 (?) lathe several years ago, but they were dooing it for a boring bar that was MUCH bigger than you are talking about, so they should be able to handle your app.

Others doo it as well, but I can't say who.


-----------------------

Think Snow Eh!
Ox

 

[Orange Vise]

The problem is the long drill needs to be put in and removed mid cycle so that it does not crash into the back panels/chuck when doing the rest of the machining.

A robot loaded drill adds many points of failure - potential misload, mis-unload, mis-clamp, or mis-unclamp.

Would be easier to just extend the main chuck out and add a drawtube extension. You'd lose rigidity but it wouldn't be horrible.

Or buy a twin turret.

 

[Orange Vise]

I don't see any capto blocks online that you can actuate via mcode, but it probably wouldn't be too hard to build your own, if you had to.

The issue is that capto uses very high drawbar force and a very shallow taper, so pneumatic is out of the question. It's gotta be double acting hydraulic to both clamp and eject the tool. Then on top of that then you have a plumb hydraulics to a turret that has infinite rotation.

I suppose you could use a robot to actuate a manual capto holder using an impact driving end effector. Not sure how reliable that would be though.

 

[Comatose]

By the time you have something custom automated and working right, you may well be spending as much as a B axis lathe that doesn't have a turret to worry about. Just something to think about.

 

[tonymor]

A robot loaded drill adds many points of failure - potential misload, mis-unload, mis-clamp, or mis-unclamp.

Would be easier to just extend the main chuck out and add a drawtube extension. You'd lose rigidity but it wouldn't be horrible.

Or buy a twin turret.

The rods we are running will be up to 3m long. Space inside the machine is already tight for front loading with the robot. Twin turret would be an option I suppose, but adding more cost than will be in the budget I would think.

 

[tonymor]

By the time you have something custom automated and working right, you may well be spending as much as a B axis lathe that doesn't have a turret to worry about. Just something to think about.

What's the price tag on a 2m between center b axis lathe with 4" through spindle capacity? The lathes we're looking at are in the 300-350k range.

 

[boosted]

I suppose you could use a robot to actuate a manual capto holder using an impact driving end effector. Not sure how reliable that would be though.

I don't see why you couldn't fashion up something to automatically actuate a manual capto holder. Worst case would probably be a little servo mounted to a torque multiplier. I'd be really surprised if a better solution is not already out there somewhere.

 

[tonymor]

I don't see why you couldn't fashion up something to automatically actuate a manual capto holder. Worst case would probably be a little servo mounted to a torque multiplier. I'd be really surprised if a better solution is not already out there somewhere.

Yeah I'm really looking for that better solution that's already out there .

 

[Orange Vise]

I don't see why you couldn't fashion up something to automatically actuate a manual capto holder. Worst case would probably be a little servo mounted to a torque multiplier. I'd be really surprised if a better solution is not already out there somewhere.

The main issue is the turret's infinite rotation. How would you wire such a thing?

 

[tonymor]

The main issue is the turret's infinite rotation. How would you wire such a thing?

Don't the capto holders on a lathe have the one bolt that needs rotated to clamp and unclamp the tool? I'm guessing he's talking about a servo on the robot.

 

[tonymor]

I guess I was thinking there would be something out there with a connection similar to 5th axis rock lock. The robot would supply the air to release it. There would be a base mounted to the turret and the drill would be held in the mounting plate. But have not been able to find anything like it.

 

[Orange Vise]

Don't the capto holders on a lathe have the one bolt that needs rotated to clamp and unclamp the tool? I'm guessing he's talking about a servo on the robot.

Yea but the robot's also holding the drill.

The robot would supply the air to release it.

This could work if rather than changing a Capto drill, the robot changes a tool block holding the drill, and the tool block is adapted to be mounted to a ZPS mounted to the turret. Not sure what the stack height is going to look like though.

Again, lots of points of failure. I would just budget a twin turret and call it a day...

 

[EmGo]

Do you have enough room in z to where you could just slide it ? A holder that bolts to either the slide or the turret base, where the drill slides forward to drill, then back to get out of the way ?

 

[Orange Vise]

What's the price tag on a 2m between center b axis lathe with 4" through spindle capacity? The lathes we're looking at are in the 300-350k range.

750K-1MM.

Could definitely fare much better on the used market.

 

[EmGo]

You know ... if you're installing robots and all, this is probably not a onesey-twosey job ? In which case, a gun drill for first op is pretty trouble-free and could end up being cheaper and more reliable in the long run anyhow ?

 

[Strostkovy]

I'm way out of my element on stuff this big, but aren't pneumatic drawn collets cheap? No idea if it is strong enough, but may be an easy option. Large diameter slip rings and pneumatic swivels (even with hollow IDs) should be pretty easy to get your hands on if needed. Or just code carefully to avoid full rotations and have the hoses breakaway.

My real input is to put sensors and air blast on the robot arm. Either a touch point sensor or a beam break sensor, so you can do go/no go gauging with the robot to prevent catastrophic failures from a part or tool misload. Airblast is great for making sure mating and measuring surfaces are clean.

We do a related trick with our welding robot. Some parts are tight access and the fixture clamps absolutely need to be locked down or the robot will hit the clamps. I program a quick glancing blow around all clamps and parts that could be fixtured incorrectly. Out robot has a spring loaded welding schnozz so a glancing blow will stop it, but a direct feed-in style hit is a bad crash that sucks to recover from because it bends the gooseneck.

I would be very amused to see a huge drill head mounted on one of those giant robots that can just reach into the lathe and drill the part itself.

 

[tonymor]

You know ... if you're installing robots and all, this is probably not a onesey-twosey job ? In which case, a gun drill for first op is pretty trouble-free and could end up being cheaper and more reliable in the long run anyhow ?

Correct. We're building a production line from scratch to produce thousands. Aiming for a 6 minute part to part cycle time. The current plan is to have one lathe doing all the machining apart from the deep hole, then hand off to the second machine for drilling. I think we will be tight on the cycle time for the first machine, and not 100% of the rods will be drilled as a second op (70-80%). So we're not fully utilizing the 2nd machine.

We are planning on buying 2 identical lathes though to allow some redundancy. In case one machine goes down we could machine rods fully on the one, we would just need someone popping the drill in and out.

I had this brain fart on Friday thinking if we could automate the drill removal then we could use both machines in parallel, therefore allowing ourselves 12 minutes part to part on each machine.

 

[Orange Vise]

Correct. We're building a production line from scratch to produce thousands. Aiming for a 6 minute part to part cycle time.

At 240 pcs/day, how are you planning to schedule insert changes?

The current plan is to have one lathe doing all the machining apart from the deep hole, then hand off to the second machine for drilling. I think we will be tight on the cycle time for the first machine, and not 100% of the rods will be drilled as a second op (70-80%). So we're not fully utilizing the 2nd machine.

I think this is still the better plan.

If you space your chuck forward far enough you can install four drills in the same machine at stations 1, 4, 7, and 10 (assuming 12 station turret). That'll keep this machine uninterrupted for 4X longer.

We've automated four lathes and 10+ mills. The common bottleneck on all of them is always tooling.