Robot loading a lathe question

Contemplating setting up an old AdeptSix 300 6 axis robot (3.3kg cap) to load parts in a lathe for a 2nd OP. Im wondering what kind of work holding on the lathe is most economical?

Right now, I hand feed the parts one at a time, into a 5C collet with a stop (bout 2" inside of collet). There is a matching diameters that end up meeting, (part starts with the diameter turned from a bar pull op, that diameter goes in the collet, the other end gets a profile cut that ends up meeting at the diameter that was cut in the first op) so I feel like a collet is the way to go. A little mismatch (maybe .001-.002" max) is acceptable. The 5c isnt perfect but is acceptable.

Im just not sure what to expect in terms of a robot being able to get the part in a 5C collet reliably (aka accuracy - used robot so Im not relying on new machine specs). I dont know that I can afford a different collet nose and collets to go with something that opens up further to provide a larger target for the robot... which sort of leaves jaws and the 3jaw chuck (Kitagawa B208).

Am I worrying about nothing? This is all new to me, so Im learning and flying by the seat of my pants. Just not wanting to spend a bunch of time on something that is destined to fail if I know I cant afford the necessary stuff to start with.

I know Ill also have to worry about a way to feed the parts to the robot, getting the door opened and closed, safety stuff, oh, and learning how to work the robot to start with :P... but figured I would start with devising my plan before I jumped in too deep too fast. But I do tend to bite off more than I can chew and then work my way through it - slllooowwwly.

Thanks for any thoughts or opinions.
Wade

The biggest thing I'd worry about is how much the collet opens (.010?) and if the robot is repeatable enough to get the part right in there. Generally the shorter the robot the more accurate so you may be fine. But if that fails you could probably just switch to a three jaw chuck. Have you ever run that job wth a three jaw? Is it good enough?

For loading you may want to put a spring pusher in the bottom of the robot fingers that gets compressed as you approach the part before gripping. That way when the grippers open the spring will push and hold the part up against your stop while you wait for the chuck/collet to close. If you can't put a spring in the grippers you could put one in an available tool spot and 1) load part in chuck 2) close chuck 3)open grippers and get robot out of the way 4) index to spring pusher 5) go to, say, Z-1.0 with the spring pusher 6) open chuck so the spring pops the part against your stop 7) close chuck.

If possible you could put an air seat detect in the back of your chuck to make sure the part is there before you actually start machining.

Be warned, though....integrating these robots is a big job. make note of all the little things you're doing (blowing out chips, etc) that have to be completely controlled and made entirely standard because the robot will only do what you tell it to do! Good luck!

You need a good lead chamfer in the collet - multi-angle lead in. I typically do 45°x3mm transitioning (with radius) to a 30° x 2-3 mm depending on how much actual clamping area you have available. You also need something to push the part in and hold it independent of the robot. The spring would probably work, but I would suggest a pneumatic cylinder if you have a place to put one that isn't in the way of the turret - or if you can program your turret to not exceed 360° rotation, you can mount it to a tool block spot. Use plastic gripping jaws (UHMW) on the robot fingers to grip the part and just use the robot to get it good and started into the hole (as far as you can without causing all kinds of hang-ups with the robot) then let the cylinder drive it home slipping in the gripper jaws (the UHMW parts are consumables). Remember to always open the robot gripper before you clamp the part or you can influence position of the part in the collet with the robot.

This kind of application has always seemed like a good excuse for a 3d printer. Bang out a set of grippers that match the part you want to hold. Multiple iterations until you have it right. Make replacements as they wear out.

Sounds like the hard way to compete with a sub-spindle lathe.






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

Think Snow Eh!
Ox

Thanks for the input!

I have never run the parts in a 3 jaw, so dont know there... but cant hurt to try. Probably going to be the best bet there if its centered well enough.

Not sure what will happen to the part rigidity if I remove part support that 5-6mm back from the face of the collet. Part is already a little prone to chatter due to stick out.

But good things to think about... and cant hurt to try a couple things to make sure that they will run so that I know if its worth the time playing with the robot.

Thanks again
Wade

I agree with the chamfer on the collet (and also the part if possible). I use a spring plunger on the robot hand to push the part in all the way (a chuck in my case).

Also, picking up the part in a repeatable fashion is important. The way I pick up my parts is a bit sloppy and would not work with a collet but works fine with a chuck.

Throw up a video camera so you can re-play the event if things go wrong. Good while developing the system, and also good for monitoring when it's out on its own.

Bonus points for making the robot push the go button, and hit E-stop if it gets confused.

Chip

Whatever you do, don't document all of it....it's our ace in the hole that gets played just when AI thinks its ready to take over the world

Don't think you've mentioned what lathe this is on but here's a couple links to a couple types of parts loaders for 5c work.

OmniTurn GT-75 with shuttle loader - YouTube

OmniTurn GT-75 with track loader and cross drilling - YouTube

Chip Chester said:

Throw up a video camera so you can re-play the event if things go wrong. Good while developing the system, and also good for monitoring when it's out on its own
Chip

Click to expand...

YES! 10 characters

If you have .010 in your collet that robot should have no problem loading that part. You just have to be sure the parts to be loaded are grabbed by the robot the same every time. And you are very accurate in you robot programming. ( in other words rubbing on the way in is not going to be acceptable). I've been installing robots and automation for the last 15 years.

Machine - presently it will be an Okuma LNC-8
I would love to have a sub-spindle machine! Oh how good life would be... but wants and budget are having a fight right now... Add in the factor that I really seem to like to jump in over my head, and learn my way out the hard way - that just wouldnt make any sense.

I wasnt aware that any 5C collets would open up that much over listed size (ie collet open to .635" when unclamped for a .625" collet)

As far as the loader idea... I have been mulling that around in my head a TON as well! Watched those exact videos and others many times... main issue I see presently is in terms of actual part loading with the loader like in the video is the part is about 4" long - usually 2.0-2.2" of it is turned .625" but the remainder is un-turned (at bar diameter - rough finish 1.125"+) and the bars are seldom straight - so turning the .625 and then pulling and turning again will end up with the two diameters not being exactly in alignment and potentially off a good bit (no tail stock on the machine that cuts the 1st op). But Ive been thinking along this lines as well. The part config has what has me thinking about this little AdeptSix I have here - thinking it could grab on the machined diameter get it most the way in and then let go, and move to the front to add some pressure as the collet closes.

I wont lie - Im the type that always tries to do more with less - way too much more with way too little and no budget to support the idea... Hell I stole the tail stock control to run hi a pressure coolant pump for this stuff... course that means no tail stock when I want to use Hi Pres Coolant... But did that because I had no available M-codes. Im sure Im well over my head - and will find out just how far under I am at some point. My biggest thing with figuring this out is not that I have 1000s to run... but that Im a one man band, and feeding a part every 2-4 minutes (part depending) just kills me and of 100 different parts I could run - 95% of them will have the same starting features for the 2nd Op. Even if I only was able to have it feed 10 per loading - I would get freed up a bunch.

The video is a good idea for sure! Things can happen so fast you miss seeing something important without it being in slow motion. Thanks!

Im sure this will be a huge thing if I ever get into it... and honestly, Im excited to just learn about stuff - even if I cant make it work in this situation... And I really appreciate the info from everyone! It may talk me out of it, or it may start a fire I cant put out til I either succeed or fail... but one way or another - Ill come out the other end with some experience.

Thanks again!
Wade

Wade C said:

Machine - presently it will be an Okuma LNC-8
I would love to have a sub-spindle machine! Oh how good life would be... but wants and budget are having a fight right now... Add in the factor that I really seem to like to jump in over my head, and learn my way out the hard way - that just wouldnt make any sense.

Wade

Click to expand...

You already own and robot that is parked in a dark corner somewhere?


For small turned parts, I would think it much cheaper to build an automated slide - likely like the vids posted above that I didn't watch - and just tie in a stand-alone PLC to run the rest. But I walk away from 2-4 minute cycles all the time. So I put on a lot of miles in a day? My feet don't mind walking, but they hate standing still, so ....

Besides, 2-4 minutes on a part in a 5C, and that is a secondary opp? Really?


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

Think Snow Eh!
Ox

Ox said:

You already own and robot that is parked in a dark corner somewhere?

Click to expand...

Yup... got it for a song and dance and was going to try and use it for something else that I ended up deciding was not worth the hassle for.

Ox said:

For small turned parts, I would think it much cheaper to build an automated slide - likely like the vids posted above that I didn't watch - and just tie in a stand-alone PLC to run the rest. But I walk away from 2-4 minute cycles all the time. So I put on a lot of miles in a day? My feet don't mind walking, but they hate standing still, so ....

Click to expand...

I can walk away... and do - lord knows the waist line needs the extra miles, but I get a lot of idle machine time because once I walk away to load another machine or start on something else, by the time that is done, its been sitting. Or if Im trying to do something else that requires much of a train of thought and focus - walking away, feeding a part, come back, realign my brain to the task at hand, and start again, to stop in a minute and repeat is quite frustrating for me.

I also fight limited work time presently - as I got "nominated" to be the stay at home dad (aka a working trophy husband) so I only get 2 days a week to work what used to be a full time gig. When there is work its scramble time, and when there isnt, its time to figure out how to be faster with what little time I have. Once the lil booger can start school Ill get opened up a little but I imagine then comes being a 'soccer dad' and so forth and likely wont have any more time than I do now. This is definitely more of a want than a need... I wont kid you...

Ox said:

Besides, 2-4 minutes on a part in a 5C, and that is a secondary opp? Really?

Click to expand...

Yup... plastic, aesthetic portion of a functional part, half the time transparent, almost always polished, 3"+ deep blind hole, tapered section in the ID, and so on. First op is maybe a minute... but its just the diameter cut for the collet. Drilling itself between the two drills needed is 1.5 min of that add in a taper with a minor ID of .250" and anywhere from 1.0" to 1.6" deep - and chatter becomes a fun one... lots of times I find I have to run at 300-500 RPM to get it out which certainly doesnt help cycle time either.

Yeah I like the loader idea for sure - just have not figured out exactly how to handle the issue of varying size and lack of center between the rod dia and the turned dia.

Thanks again for the input!
Wade

Wade C said:

Yup... got it for a song and dance and was going to try and use it for something else that I ended up deciding was not worth the hassle for.

Click to expand...

BTDTGTTS

I was looking real seriously at a huge job dooing finish work on some fairly large alum castings back around 2008. We did not "have" the job _ yet - at least, but we were producing samples, and I found a pr of nice 50kg Motoman's at auction for 1/2 what I would have had to pay otherwise - should I land the job. The job would have needed prolly 6 units all told, but this would be enough to at least get the job rolling and easier to bank roll.

I figgered that if we got the job, we saved ?? maybe $30G? And if we didn't land it, I should be able to offload them for at least what I had in them eh? So, other than tieing up funds, it shouldn't "cost" me much of anything...

Well, we didn't get the job (may have been a good thing) and then - let me introduce you to fall of 2008! Try selling a robot after Sept 2008! Good luck!

So there I was, going into the Great Recession with $40K in robots that are only getting older while setting there, and very little work. Thank God that when 2010 rolled around that the robot manufacturers didn't have enough stock or a full pipeline for when we came out the other side, and Motoman actually bought them from me.

I had them for sale for 2 &^%$#! years, no calls. Then I have 3 guys chompin at the bit in one week!
I prolly could'a asked for more $, but I wasn't gitt'n greedy at this point! I got my $ back, less 2 years interest.

Wade C said:

I can walk away... and do - lord knows the waist line needs the extra miles, but I ....


I also fight limited work time presently - as I got "nominated" to be the stay at home dad (aka a working trophy husband) .......


Thanks again for the input!
Wade

Click to expand...

Not exactly the mental image that I would have guessed for a "trophy Husband".

???





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

Think Snow Eh!
Ox

Ox said:

BTDTGTTS

Click to expand...

SITBDF (sent it back, didnt fit) hehehehe

Ox said:

I was looking real seriously at a huge job dooing finish work on some fairly large alum castings back around 2008. We did not "have" the job _ yet - at least, but I found a pr of nice 50kg Motoman's at auction for 1/2 what I would have had to pay for one - should I land the job. The job would have needed prolly 6 units all told, but this would be enough to at least get the job rolling and easier to bank roll.

I figgered that if we got the job, we saved ?? maybe $30G? And if we didn't land it, I should be able to offload them for at least what I had in them eh? So, other than tieing up funds, it shouldn't "cost" me much of anything...

Well, we didn't get the job (may have been a good thing) and then - let me introduce you to fall of 2008! Try selling a robot after Sept 2008! Good luck!

So there I was, going into the Great Recession with $30K in robots that are only getting older while setting there, and very little work. Thank God that when 2010 rolled around that the robot manufacturers didn't have enough stock or a full pipeline for when we came out the other side, and Motoman actually bought them from me.

I had them for sale for 2 &^%$#! years, no calls. Then I have 3 guys chompin at the bit in one week!
I prolly could'a asked for more $, but I wasn't gitt'n greedy at this point! I got my $ back, less 2 years interest.

Click to expand...

Yeah 2008 took care of a lot of things... I was in the restaurant industry in a small town at the time (this whole machining thing was a side job at the time - hoping to make it full time someday). I saw a lot of the down turn and read about a lot of it here as well. I think my favorite part of it was the radio and TV commercials telling people "dont eat out, stay home and cook for yourself, you'll save money"...

Glad you were able to get clear of the unneeded robots... and werent completely screwed by the whole ordeal! Id have been in full on freak out mode... we'll at the time I was, but for a whole different reason.

I have a 1/10 that in this lil guy here, and it was more of a swap than a buy anyway... hoping some day to find out it was a good trade of my time. But if it never pans out, it will be an affordable lesson learned... I hope Its been here for a couple years now - and I keep telling myself I need to play with it but just havent had the time.

Ox said:

Not exactly the mental image that I would have guessed for a "trophy Husband".

???

Click to expand...

Hehehheheee Touche`
Yeah I call myself a trophy husband more for my benefit... and the comedy. If I dont laugh, Ill probably cry. Wife has the benefits with her job, and with the cost of health insurance and the wonderful requirement that we have it, and that she makes more than me - it was a losing battle to even try and argue that we could make it if she stayed home... hell wouldnt even be an argument - just a display of my own subborness and stupidity. So, now I spend the week wiping butts, making PBJ's, watching Seasame Street, practicing the alphabet, and repeatedly saying things like, "no, dont throw the kitty..." or "I dont think the cat likes being on the trampoline." or "Why is there bologna in the DVD player?" or my personal favorite, "STOP pushing buttons!"... At least she is good about not pushing buttons when the machines are on... but she loves the control panels for sure... High alert around here when machines are on and she slips out of the house to "come see dad" on my work days.
PS - most of that is true... but I may have embelished a little bit... She hasnt put bologna in the DVD player... yet.

Wade

The fact that you are wanting to shove clear finished plastic in a collet that is known to barely open past nominal at all, really makes this sound tough with the robot. At least not if you have the part held rigid in the bot. I just see lots of drag marks all up and down one side of the part - even with big lead on the collet.

Maybe - if you was able to hold it loose - like in the rack, line it up slightly lower than the collet - with big lead, and then shove it in (slowly) with a small air cyl. This way the lead (chamfer) can actually lift it off of the fixture that your bot is holding and not have rigid drag marks.

Could you rig an air cyl to eject as well? The stop in the collet could actually just be the retracted position of the cyl. With a spring loaded retract cyl, you could get away with a single port rotary union for the eject cyl. Thus, no rigid drag marks for ejection either.

???



EDIT:


Also, by the sounds of it, you wouldn't need to complicate it by even closing the door?
Does your bot have much I/O?

Glue a L/S to your tailstock (not in use eh?) and set it so that when your turret is all the way to Z+, that it trips the L/S. This is the last line in your program. Your aux PLC is triggered, and it will open your collet, eject the part, retract the ejector/stop, then tells your bot to move in, push part in, close collet, retract bot, and trigger the CYCLE START button.

You can place a relay between the lathes PLC output and the collet closer coil, and trigger it off the lathes 24v source. The machine would never even know the collet was opened.

Also a relay in the CYCLE START button too.

Not really that hard to doo...


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

Drainin' the Swamp!
Ox

Parts arent polished going in, just machine finish - polishing is the 3rd op.... then on to the mill. Im hoping that drag marks would be superficial enough to come out in the process of polishing. Honestly, I just figured that a 5C wouldnt open up enough, nor the robot be accurate enough to even get it in there... that I had been focusing on searching the auction site for another style of collet nose that opened more that I could afford... But no luck there led me to asking here about what I might expect.

I was thinking for removal on the part I would either figure a way to "reverse" the loading procedure or I also though a spring eject would work well enough as long as I had a way to catch the part on its way out. I like the air cyl idea for ejection as you could control the ejection speed as well which would make catching it a little easier.

Lots to think about for sure!
Wade

I was adding to my post above after while you posted.
Check above.


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

Think Snow Eh!
Ox