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

[Strostkovy]

These are chromed rods for cylinders and we have to be relatively careful handling them. 3m is the max length. They will be anywhere from 1-3m long. Diameters every increment of 5 between 50 and 75. Have all kinds of fun nesting lengths to avoid big drops.

Front loading with the robot has a lot of benefits for quick changeovers from one diameter to another and different length combinations.

A long time ago I wrote a program to optimize tube cutting for minimizing waste. It also matched common angles wherever possible. It wasn't user friendly and I don't have it anymore, but I'm sure someone somewhere has a decent program for it, if you are still calculating it all by hand.

 

[CarbideBob]

Big red flashing light and a human somewhere around?
Of course it can be done with a robot. With enough money and time one can do lots of things.
ROI

 

[tonymor]

A long time ago I wrote a program to optimize tube cutting for minimizing waste. It also matched common angles wherever possible. It wasn't user friendly and I don't have it anymore, but I'm sure someone somewhere has a decent program for it, if you are still calculating it all by hand.

Yeah we do it all manually but when you're typically getting 2/3 PCs per length on the longer stuff and 5/6 on the shorter it's not the hardest thing to do. And we'll only have a few weeks of orders at any one time so you're often limited in the lengths you can choose from. Now that we're getting into bigger volumes it might be worth coming up with something a bit slicker.

 

[YoDoug]

MD Tooling makes a live tool motor actuated turret mounted steady rest. The live tool motor running in one direction engages steady, other way disengages. My guess is they could do the same for some type of tool mount, Capto/HSK/etc. I would reach out to them.

 

[Ox]

Live tool runs a hydro pump eh?
Interesting idea!


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

Think Snow Eh!
Ox

 

[Orange Vise]

Hmm, the more you describe it, the more complicated it gets.

How is the robot going to handle the different lengths and diameters? How many different lengths and diameters are there? How many of each size will you be running per batch?

Keep in mind that robots are quite flexy compared to machine tools. A 3m long, 75mm OD rod is going to exert a significantly different moment force than a 1m rod of the same or smaller OD. Your options here are to program every sized rod differently to account for different moments, or oversize your robot by a significant amount.

What operations are you performing other than the drilling?

 

[DanielG]

I think a cartesian gantry robot is going to make way more sense than an arm for loading. 3m of 75mm round bar only weighs about 230 pounds, but it's long and spindly. A standard six axis robot is going to want to grab it in the middle to minimize torque. It then has to line up the end 1.5m away with whatever you want it to go into. Keep in mind that the natural frequency of the beam, even if you can hold it completely rigidly, is about 1Hz.

A gantry built with modular off-the-shelf parts would handle it very easily. I'll use lathe coordinates for this explanation. Let's say your material rack is left-aligned. Robot gantry runs in Y, trolley runs in Z. The left gripper is fixed to the trolley. The right gripper is on a Z' axis (on the trolley) to allow it to adjust to pick up different length bars. Z' is cheap and easy, because you don't care at all about accuracy (+-1" would be fine); belt drive actuator maybe.

You rack your material flat. With a six minute cycle time, you need about a four or five foot wide infeed/outfeed table to store an hour's worth of production. If you want longer, you either make the table wider and do drawers.

Cell controller tells the robot how long the part is, it adjusts Z', picks up the part and loads it into the lathe for turning. Once that's done the robot picks it up and loads it into the drilling machine*. When that's done, it goes back into the same slot in the rack, maybe with a quick stop at a laser engraver to put the part number, etc on one end. The laser obviously isn't necessary, I've just always liked machines that do this, it eliminates a lot of screwups later down the line.

I'm with EmGo on the lathe not being the correct machine for drilling the hole if you've already done the turning work. A 2m lathe is almost 5m long, so a machine to hold a 3m piece steady while it drills the end isn't going to be any longer. You can securely clamp the part at both ends without needing 3-jaw chucks, steady-rests, etc.

I'm not sure how you're getting the blanks into the cell and the finished parts out; I assume on carts. The gantry loader will be happy to load/unload from the cart to the feed table for you, so it covers the material handling for the cell as well.

 

[triumph406]

I might be missing something, but a machine with a Turret style like a Mori Seiki SL8 would be the ticket. The rotation of the turret is normal to the Z-axis, rather then parallel with the Z-axis

 

[Ox]

Sure is the long way around the block to simply not use a bar feeder and sub-spindle machine.


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

Think Snow Eh!
Ox

 

[EmGo]

Sure is the long way around the block to simply not use a bar feeder and sub-spindle machine.

I still think a lot of the long-tube endworking jobs would be better off done in either a double-end, stationary-part machine or in one of the skinny-chuck-in-middle lathes, that still spin the part but both ends are free to be worked on simultaneously and the part doesn't have to move 5' in both directions.

That whole 3' deep hole is so awkward, doesn't make sense to me to do it in one operation, sometimes you can take this "Do it all in one setting" thing too far.

 

[Ox]

His hole is 20" deep, and it can all be done up inside the spindle.
Meaning that as long as he has 1" sticking out to face and chamfer, the drill can start there and go in from there.
The part does NOT need to stick out 20" to drill a 20" hole!
It does NOT need a steady rest.

Now, we don't know of any OD requirements that may need sticking out, but with a sub, it can be pulled out, and the sub used as the tailstock.

Only catch is that you need to keep 3" (?) collets in the sub to allow it to pass through, so you can't turn the end down to 2" and grab on that. You need to grab a little further down where it is still 3".

If he needs a tailstock, he can put a center in the lower turret, and then you still git tailstock and sub-spindle both.
The sub is of very little value if you cannot deep through the whole bar through.

You would need some sort of run-in table/fixture. Whether that is a commodity bar feeder, or some custom built unit, whatever...

Would also need some sort of run-out table as well.


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

Think Snow Eh!
Ox

 

[EmGo]

His hole is 20" deep, and it can all be done up inside the spindle.

The rods we are running will be up to 3m long.

Which is a total pain in the ass in any lathe with less than a 10' long bed ... even if you bore the hole with only 6" sticking out, you still have to remove the finished part.

Is there a subspindle lathe with 10' of Z ? If so, I bet a 2' long drill won't be a problem. Just use a steady rest to get the end far enough from the chuck so the drill doesn't crash anything.

This whole "change the drill" thing sounds like unnecessary complexity. Complexity = bad.

 

[Ox]

You don't need 10' of daylight - if you pass through the sub.

I have ran parts that were prolly up to 8-10' long through my machine with only around 20" of daylight.

Unless you have seen this done, I understand that it is a hard concept to git'cher head around.
You git stuck in the "between centers" part as a requirement for your part to fit.


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

Think Snow Eh!
Ox

 

[EmGo]

You don't need 10' of daylight - if you pass through the sub.

You git stuck in the "between centers" part as a requirement for your part to fit.

You're right, I was fixated on the dumb robot. Do it your way, don't even need the damn robot, just let the bars clang out the back end onto a conveyor.

And while we're at it, ditch the chuck and use collets, save a tornado in the machine.

And then, cut the sheet metal out behind the collet chuck, built a 20" deep box, and just run the drill normally, just make sure it's in a rear position when using any turning tools.

Easy peasy. We don' need no steenkin drill changer and no robot, neither

 

[Ox]

More likely - you would leave the part out as needed to clear any turning opps.
Your turning tools could maybe be spaced off in Z+ a bit to aid in this as well.

Robot or not, no need for a 3 jaw chuck for "bar" work.


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

Think Snow Eh!
Ox

 

[MotoX]

Why not leave the drill in the turret and cut a hole in the sheet metal for clearance of the drill when the turret is close to the chuck? How many positions of the turret are going to be occupied?

Box it in to keep the chips and coolant where they belong. If youre feeling fancy, plumb a coolant line for washdown.

Most builders usually have plenty of dead space behind the chuck anyway. Make some use of it.

 

[tonymor]

Hmm, the more you describe it, the more complicated it gets.

How is the robot going to handle the different lengths and diameters? How many different lengths and diameters are there? How many of each size will you be running per batch?

Keep in mind that robots are quite flexy compared to machine tools. A 3m long, 75mm OD rod is going to exert a significantly different moment force than a 1m rod of the same or smaller OD. Your options here are to program every sized rod differently to account for different moments, or oversize your robot by a significant amount.

What operations are you performing other than the drilling?

We have a number of robots feeding lathes with a similar setup back at HQ.
With a self centering gripper - OD is not a problem anymore. Otherwise, you can calculate the change on center height based on the geometry of the gripper.

The robot gets 3 pieces of info - Length, diameter and qty. It picks up in the center and makes a z offset as required when loading. Also adjusts for diameter if necessary.
Ideally, I will get my 3m bed machine and always be able to pick up in the center of the rod to help avoid any issues with moments.

Rough turn, finish turn, drill/tap face, starter drill, bore, internal groove, holes and slots drilled in x/ milled in y... Plenty of other operations.

 

[tonymor]

Why not leave the drill in the turret and cut a hole in the sheet metal for clearance of the drill when the turret is close to the chuck? How many positions of the turret are going to be occupied?

Box it in to keep the chips and coolant where they belong. If youre feeling fancy, plumb a coolant line for washdown.

Most builders usually have plenty of dead space behind the chuck anyway. Make some use of it.

Too many operations other than the drilling to always have the drill lined up with the extra deep cut out. We have made a few of these parts before and we are probably using 10 of the 12 stations on our turret. Orange vice pointed out we could space the chuck out to gain the needed clearance, but then we lose space for loading the longer rods.

 

[Orange Vise]

Orange vice pointed out we could space the chuck out to gain the needed clearance, but then we lose space for loading the longer rods.

4 meter machine.

But as Murphy's Law would have it, someone's gonna order even longer piston rods to exceed the capacity of the 4m machine.

 

[Ox]

You can run all 40' through a subspindle machine...


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

Think Snow Eh!
Ox