What's new
What's new

Automating a CNC drilling machine

garyhlucas

Stainless
Joined
Oct 17, 2013
Location
New Jersey
I have posted about my grandson working with me at my job and we are getting ready to start on a new project. The company purchased a 2 axis drilling machine with 5 spindles to drill 104 holes in polypropylene header pipe. It was here when I arrived and they asked me to get it up and running after another engineer had installed guards and such. It is a Chinese machine and basically junk made from scrap metal. However I did manage to get it to do a beautiful job and I cut the cycle time from almost 3 minutes to 1 minute flat. Because it was running at 3 minutes they had already ordered a second machine, double sided with 16 spindles! That is the one we are taking a whack at now.

I thought I'd write about it because I think I have figured out a way to automate it, eliminating 2 operators per shift over 2 or 3 shifts. We extrude our own pipe and the pipe must sit for 24 hours between being extruded and cut to length before drilling because the material spirals a bit as it crystallizes. So the pipes must be stored in large racks. So I felt the key to automating this process was finding a way to automate unloading the racks, and possibly loading them automatically.

What I came up with is two flat straps across the rack and under the pipes attached to an unloader that sits on top of the rack. A shaft on one side with slots for the straps like tie down straps is turned by a gear motor that rolls the straps up. So the whole 1000 lbs of pipes rise in the rack and flow over the fixed side and down a ramp into the machine. An eye senses the ramp is going empty and turns on the lift motor and it goes off when the ramp refills. There are two sets of clamps that are parallel for drilling. So the ramp leads to a shaft with slotted disks. Pipes fall in the slot and the disk rotates left or right and rolls the pipe into the loading station.

The pipes are sitting in the loading station on the left. I've added a gripper to both ends of the Z axis each with an air slide to drop them below the spindles. Th Z axis picks up the pipe in the load station with the left gripper and the finished pipe in the right gripper and travels to the right end of the machine. It puts the new pipe in the clamps and drops the finished pipe into one of the 6 carts my grandson just finished building. Then the machine drills its way back to the starting position using the 16 spindles.

We are fabricating the rack unloader right now. We have lots of racks already. The key is getting the rack unloaded, if that works everything else is fairly simple. No CNC controller, using a PLC that can handle the 3 axis motions, and 8 sets of air operated drilling clamps and the HMI. A second identical PLC on the Y axis carriage handles about a dozen sensors and limit switches along with 12 air valves. Only a 24vdc power cable and an Ethernet cable connect the two PLCs.

If this works out we will be building a second machine. This machine will be a lathe that picks the pipe off the extruder line again with a linear axis and double grippers. The new pipe gets placed on the expanding arbor by the first gripper and the finished pipe is picked and dropped into a vertical conveyor that drops it in the racks that feed the drill machine! You can't just drop pipe 5 feet into an empty rack. So the straps in the rack will be wound up tight and as the pile builds the straps will slack off lowering the whole bundle to the bottom. When the pile starts rising above the rack, it's full and the operation stops. A gate on the extruder lets it keep making pipe and dropping them into a manual bin for a short time.

The lathe does 3 operations on 4 lengths of pipe. So it has two servo operated 2 axis carriages. One by the chuck grooves the pipe and chamfers the end. The other end of the lathe has an expanding chuck on the tailstock. A second carriage grooves that end, trims the length, and chamfers it. The tailstock and both carriages run on the same linear rails. So the program moves the tailstock and second carriage as needed for the different pipe lengths. This lathe eliminates two more operators.

If you think we are getting rid of lots of employees, it is actually a small number, because our process is REALLY labor intensive and lots of jobs require thinking and sensing and handling that simply can't be done by robots or machines as the solar collectors we build are floppy and vary in size from 2' x 4' to 4' x 20'

I'll post some photos if people are interested.
 
I tried to think of a way to have a tooth pick dispenser for bundles of pipe. Having an indexing chain feed bunk with three spaces would drop the pipe the last bit into a set position.
 
Mr. Lucas
I was cleaning out my desk today, and found an email from you from 2004, helping me with a water flow test procedure.
You didn’t know me, but you took the time to type out a full page of your advise.
I want to thank you for being the type of person who makes the world a better place.
 
Damn,
That’s nice of you!

You know at my new job we make solar heaters for swimming pools. Not really making the world a better place in my mind. My last job I was inventing waste treatment plants that turn sewage into water so clean you could safely drink it. I felt I was making a contribution to the world and I miss that.

Last night I was speaking with a guy about building sensors that I came up with which make those plants turn out even cleaner water, using less power, far less expensive chemicals and are cheap and easy to maintain. I think I am going to persue this on my own time with my grandson.
 
I just ran across your post, and it's very interesting.....please post more as you progress! A couple years ago I read some of your posts and if I'm not mistaken you were between jobs?

Anyway, this thread is very informative! Jeff
 
Yes I lost my job with a wastewater treatment plant manufacturer because I made the mistake of unloading on the owner, when a slacker took a text out of context and sent it to my boss, who's response was that I had done something wrong, instead of asking me about the situation first.

The company has been posting on LinkedIn about the projects that are now reaching completion and of course the owner, the engineers, the contractor and the customers are all standing around with big smiles on there faces like they did something special. Actually the entire plant, the electrical, the PLC programming and every one of the unique features of that plant were done by me personally. I guess I should just be happy with the royalty checks for thousands of dollars that I get from each one they complete. It will be interesting to see how they fare over the next couple of years with the actual innovator gone.

I will keep posting about this project. It is actually kind of neat. Currently we have torn the whole machine down to its basic frame and will shortly start putting it all back together.
 
I don't want to argue against building custom automation, but it seems like you could rig an LNS bar feeder or similar for a lathe to grab a pipe from the hopper and feed it into position.

Then all you'd need would be a couple kicker cylinders to push the finished pieces from the machine to a "done" pile on the front.

LNS already has indexing individual rounds from a hopper and pushing them to a servo-controlled position handled. Why re-invent? Worst case you'd need to extend the length of the hopper and make a custom tip for the pusher, but the pipes are light.
 
I’ll have to post some pictures of what I am doing. I think when you see it you’ll be amazed at simple this automation is. I’ve done lots of automation projects and have a good track record so I think it will work. I always have a Plan B if it doesn’t.
 
Drill.jpg
A SolidWorks model view of the machine. The rack on the left will hold about 300 pipes 4' long and the straps will pull them to the top to overflow into the machine. The gripper on the left end of the carriage grabs the pipe to load and the gripper on the right end grabs the pipe to unload. Carriage moves to the right and dumps the finished header into the cart, while placing the next one in the clamps. It then drills its way back to the start. The whole mechanism is duplicated on the back of the machine.

Lots of I/O, 12 air valves on the carriage and 13 valves on the base. 16 sensors on the carriage and 16 more on the base. So it will have 3 Do-More BRX PLCs, one in the main cabinet for motion and spindle drives, one in the base, and one on the carriage connected via ethernet. Just two control cables and no junction boxes. The air valves are in the PLC enclosures and premolded sensor cables easily reach the local PLCs.

I should have all the parts for the whole upgrade in about 2 weeks. Worse part is I have lost my chaffeur/assistant to High School!
 
So today I finally got to test my prototype Auto Rack feeder. It works! Needs some revisions but this was built from scrap I scrounged from around the plant and a gearbox from a machine I built 25 years ago.7B10B731-D63C-4E29-B858-E789866791C5.jpg6CF4DFD4-4576-43A8-9193-3BC63DFE6ED4.jpg
 

Attachments

  • 4DF22575-D7BB-4280-98C9-B7C31601A507.jpg
    4DF22575-D7BB-4280-98C9-B7C31601A507.jpg
    13.7 KB · Views: 123
Well, the sideways pix are not that clear but I can't figure out where the Z axis moves the pipes. How does that happen?
 
That hydraulic schematic is an oops, not part of this project.

The pipes get moved by grippers on each end of the Z axis that drop down to grab the pipes. Then X axis moves to the right to bring one in and send one out.
 
An update on the drilling machine automation. Going slow only because of having almost no help and the need for me to keep a whole lot of china junk machines and worn out machines running so product can go out the door.

I have re-evaluated the control system I am going to install using an Automation Direct BRX PLC, and will now be using 3 of them talking by Ethernet. One in the main cabinet does the high speed motion control and runs the spindle drives using all of the onboard IO plus a couple expansion modules.

A second one on the base of the machine controls the 8 sets of pipe clamps and 10 sensors to insure the parts are in and properly clamped. It also will control the automatic doors, autoloader, and the sensors for the discharge bins. The two jog pendants for setting up will connect here as well so it uses all the IO and a couple of add on modules too.

The third PLC rides up on the X axis and controls the 6 tool slides, and 4 grippers that load the machine, plus the X, Z1, Z2 limit switches, and sensors to ensure the pipes are properly in the grippers while moving. Virtually all the IO is used up again. Since all three PLCS are now within cable length of all the valves, sensors, and drives the wiring has been radically reduced. In fact all the 32 air valves now share the enclosures with the PLCs and the wire lengths are about 12".

We recently hired a very good maintenance guy with great electrical skills and I thought my prayers were answered. Then he missed a couple of days with no call in, and just disappeared for the past five days so my prayers went straight to hell!

However my grandson has been working with me for the past two weeks as he is on winter break AND I have a second young student going to NJIT also working with me over winter break. I am now more professor than engineer, their skills at 17 & 19 are about what mine were at 10, but they both work hard and you have to start somewhere. I figure a couple of engineers that actually might know how to drill holes, tap, and cut metal might better appreciate the guys they may wind up supervising!

On a side note, my grandson just found out he got a 4 year scholarship to NJIT for $14.5K per year for 4 years! However he likely won't get to use it all, because has already taken so many college courses he will enter NJIT as a junior, in the same class as my other helper that is 2 years older. The good news is he already has two jobs, working with me, and working at a Target store where he is quickly working his way up. He has a good chance of graduating college with NO debt!
 
I am a fan of the lite bus setup - many small computers controlling semi discrete systems. Our drill and angle line run with win ce boards for every punch or spindle, another for clamps and hold downs. The mother ship pc controls the conveyor. Rarely any hiccups, and if they happen more than twice a shift easy to narrow it down.
Our coper runs a pc to run a single plc to run ethercat cards - and always hiccups and impossible to deal with maintance wise. granted when it runs right it is nice with no dwells, but a moth across the shop can change it.
 








 
Back
Top