Printer Friendly

Computer-assisted fab shop management.

Computer-assisted fab shop management

Way back in the late 1960s, early generations of NC turret punch presses often were programmed "by hand" using an animal known as the Friden Flex-o-writer. It cranked out a basic ASCII eight-channel punched paper tape.

Over the years, the computer has played a greater and greater role in NC programming. Today computer-assisted can mean a lot more than just an easier way to generate an NC program.

To fully appreciate benefits the computer can provide, we need to go back and look at the evolution of what might be a typical modern fabrication shop. At some point in time, maybe five years ago, or seven years ago, or ten years ago--it really doesn't matter how long--the owner put in his first NC turret punch press. He liked what it could do, so he bought a larger one. Then, perhaps, a new generation came along and he bought a faster one. Now we're in the hypothetical situation where the guy has $250,000 or more tied up in each one of maybe three or four machines, and he's scrambling.

He's spending all his time programming; the machines are sitting idle half the time; he's not getting anywhere near the maximum utilization out of this equipment. He's ready to throw up his hands and scream. Now the question is: Where does he go from here? What does he have to do now to solve his problems?

Carter West comments, "That's exactly the scenario that's happening on a daily basis. And he's driven by a thought process that's been burned into his brain for years and years--possibly generations--on how to achieve the end point of making product. There's another way to do things. There is a way out of this. You see he's in a squirrel cage. He buys one machine. He thinks that to get more work out he's got to buy another machine; and to get still more work out he's got to buy another machine.

"What's he doing? He's digging a hole that he himself can't get out of because every machine he buys means he's quoting more; that he's taking on more work; that he has to buy more material; that he has to review more quotations; that he has to do more programming; that he has to hire more helpers. So all he's doing is perpetuating, on a three-fold scale, the problem he has with machine number one.

"He's still not getting any more product out. What he has to do is become more efficient. And how does he become more efficient? By standing back, taking a look at it, letting the computer do the operation for him. Now he can look at the hole that he dug. Now he can fill that hole. But, I think the first thing that has to happen is he has to learn that there is another way.

"It's not just Amada's system. It's nice that we talk about it. But there are so many systems out there. So it's going to be a matter of finding just the right one. They all have different functions, different features. It depends on what you want. Do you want fast or do you want slow, do you want easy or do you want complicated? What do you want it to do for you?

"The basic thing people buying these products should do is make absolutely sure that they are compatible with the biggest share of systems in use. If you buy an Atari, or something like that, you may not be compatible with the major line of vendors that you're working with. Then you won't be able to go online with them--you won't be talking the same language. And, although you will have the technology, it can't be shared to benefit anybody outside of your little world."

The point, of course, is that the fabricator ought not look just at his own operation. He also should look at compatibility within his vendor and customer network.

Internal networking

According to West, one of the major things that should be considered is internal. Integrating and networking the various fabricating machine tools--handshaking them--is the most important thing to be more efficient. Typically, the job shopper will have three or four shearing machines in one area, several press brakes all lined up in a row, and a couple of punch presses in another department.

Carter asks, "What's wrong with that picture? What's wrong is that's old technology. New technology is a central computer control room and one, two, and three or more fabricating cells. Inside each one of these cells, depending on what work you're going to send down to them, you have an NC turret punching machine, a bending machine, possibly a small shear, and maybe some hand tools such as a drill press, sanding/deburring device, or some such thing.

"Some cells could be smaller equipment. Others could be larger equipment. Others could be very large equipment. This would relate both to the material (thickness) and the type (size) of parts you want to process.

"So now networking the system comes into play. The master computer in the control room enables a couple of operators or programmers to program the punching and bending. Also, there are other software packages out there that can do production controlling, material review, material requests, quotations--on and on and on. It's all being done in one room.

"Now data can come down through the DNC links into each cell, with production requirements and even schedules. There is scheduling software that enables the computer to give you that data. For example, let's say you need six different parts, all of the same material (16-ga CRS). The computer, not the person, will lay out the appropriate number of each part to get the desired end results. Say you need 35 of one, 50 of another, 100 of a third, 12 of another, 80 of the fifth part, and 10 of number six. The computer will give you the ratios so that, after the last sheet is done, you'll have all your parts. This is called automatic part nesting.

"Now what have we done? We've run six jobs through in the time that it would have taken us to do one with manual programming. Now, while the cell is in operation, because we have a CNC operator, he can double as a press brake operator. As parts come off the punching machine, he can be processing them through the press brake. What does that give us? Say there are seven subassemblies. The old way we can't do anything until that seventh part is completed. Now we have all the parts combined. So as soon as the very first one comes off, we're making assemblies. So now we're amortizing much less time over the part run, giving a much better cost ratio. Again, we're being more efficient and competitive.

Picking the right size

Focusing again on our hypothetical job shop with three or more machines, we asked Carter about optimization of blank size. If we have a part that's roughly 1 ft x 2 ft, are we better off to run it one up? Or, are we better off to run them four up--to go to a bigger machine? Who makes this kind of decision? Is there something in the program that helps us do this?

"That's an important observation," West muses. "Let me go back in time to where we're talking about the job shop owner or the production manager toiling over quotations, toiling over machine problems, toiling over low production. He is now given the opportunity to make a better, more rational decision as to how to process the part. He can ask himself: Do I put it on the little machine, or do I make twenty at a time on the bigger machine? Because he now has both the mind set and the time to think about those things, he can be more competitive using the right machine tool.

"But if he's down there digging the hole, he can't see the machines up on the surface. He's just hurrying away. He's working, working, working. He looks up and he sees that they're making one part at a time over on the 357; it goes up, makes two holes--thunk thunk--and comes back. So it's punching but it's not doing it very efficiently. Maybe it's costing about $1.50/hole. What you want it to do is run about $0.02/hole. So you have to make about 40 of these parts on one run in about two minutes.

"The guy with the computer now has the ability to get out of that hole and give his mind the time required to look at the big picture and make the right decisions."

Even in the early days of programming with a PDP-8, there were always considerations over whether you were better off to make an X, Y move or whether you were better off to make a turret rotate. Of course, this was because the turret rotated a lot slower than the servos could drive the sheet. So, in general, you were better off to take a tool and use it all over the sheet and then change tools--but not always. There were exceptions--particularly if you could execute an X, Y, move while you were rotating the turret; but that kept a lot of people awake at night too--especially if the sheet was between the turrets. There were a lot of these value judgments, including how to use the equipment itself. Are these now available as part of the software?

"Yes, those are actually part of the entities that are being taken away from the human mind so the man can do other things," West explains. "Those responsibilities are being given to the computer, more and more as the computer goes through generations of upgrading.

"You know it used to be a powerful thing that the programmer would make a judgment call. He would say, `well it's quicker to make a turret rotation than go to a tool; or its quicker to change the tool location in the turret; or it's quicker to cut it all out rather than stop the machine to pull out the pocket.'

"We've done vast studies on actual time cycles, and on what is relative. If the guy says it's quicker for the part program to have the operator change the tools and put them into sequential order--that's relative. It might be quicker on a part basis, but then you're incurring the setup downtime. What's more important--the machine punching holes or the machine sitting there getting tools exchanged without turret management? We've proved time and time again that, for cash flow reasons, we need the machine punching holes--not sitting idle having punches changed.

"Something as simple as cutting out a pocket versus punching out the entire area--that's a judgment call. The computer would elect to punch it all out if it's larger than the trap door--because the trap door can be used without stopping the machine. Or it calculates the percentage of time. If the pocket is so large that processing it out is more than 10 percent longer than stopping the machine and having the operator intervene, it will decide the operator must intervene. And it will calculate that time into the schedule.

"The computer is using cycle time studies that we've fed it based on various standards--sheet sizes, sheet thicknesses, weight--which determine how long it's going to take a person to handle that weight. These are things that a human programmer can only guess at."

Sizing up the need Again we asked West about the guy with three machines who now realizes he has a problem. What does he need, initially, in the way of computing power? Is this broadly flexible?

"Concerning hardware, it all depends," explains Carter. "It depends on what you want to process through the computer. The key is compatibility. Most of the computers today are using PCs that operate on MS-DOS or PC-DOS. The speed of the computer is relative to what you are running. For CAD/CAM we prefer the 386 machine, primarily because the matrix of numbers going through the processor at one time is quite intensive. If we are going to run just our G-code editor, which is basically just a fancy word processor with graphics, you only need a 286 unit because it processes fast enough.

"Now, if you run the nesting software that I described, with all the different parts put onto one sheet, the computer is playing a six-dimensional chess game. So what we recommend is running the 386 with 25 MHz clock speed and 3 Mb RAM. You need a 32-bit processor because it can do this operation much, much quicker."

Does our guy want to do that?

"It depends," Carter replies. "If you're a fab shop just starting out, No you wouldn't want to. You have to learn how to crawl before you can walk, and walk before you run. I guess what I'm saying is that there are natural progressive steps that you must take to get to the end point of being a substantial job shop--one with 200,000 sq ft or more of floor space.

"The smaller guy would be better advised to start out with what we call the CAM or 1E editor. It has all the ingredients needed to make G-code tapes running off CAM system tapes without any design aspect to it. He can look at the part, he can manipulate the part, but he can't design the part.

"Once he learns that system, he can get the CAD module. From then on it's just buying disks and plugging them in. From there you get automatic nesting, next you get 3-D unfolding for your press brake, and then you get press brake software.

"Eventually, a centralized computer is controlling all the machines on the floor. The importance is that the computer knows if a machine is not working, and which one has some time to take the overflow. Without this knowledge, even the smartest foreman or production controller cannot foresee that a certain press brake is going to be done in 15 minutes and it's going to have another 20 minutes before any more parts get to it. The computer has the power to tell you that--through networking.

"If you can achieve this, when you buy more machines you're creating profit centers. Once you get efficient at what you are doing, once these machine tools are running 80 to 90 percent of the work period, you can start to add machines. Then, you can stand back and watch them crank out money, be more competitive, and take a bigger market share.

"I think it's very important to bring out that the cure is not just buying an automatic punching machine to save labor and get more parts expedited through production. Because, as I said earlier, we could be mashing out thousands of parts a day or we can mash out two parts a day--it matters not. All we've done is transfer the bottleneck from the punch press to the press brake area. So you always have to consider the two machines as being part of an integrated, interlinked system.

"The common ratio is two press brakes to one punching machine, because the punching machine in the flat stage can put out that much work. I used to set up job shops at the consulting level. And, one of my recommendations was, depending on the size and the range of material they were processing, if they were to buy a punching machine--with a 50" x 72" table, then I'd recommend they buy not only a 6-ft press brake but also a 12-ft press brake. Why? Because the punching machine has the ability to reposition the sheet and go 12 ft. So if you do not use a press brake that can accomodate the punching machine, you've accomplished nothing."

What about post processing?

"Running two files simultaneously, the computer can take the geometric code and post process it to any machine or all machines," Carter explains. "Once the manager gets freed up, he has the time to be able to select the proper machine, Maybe a part was run on one machine the first time but, later on, that machine is either not available or is running more important parts. So now he needs to run it on a different machine. With the click of a button, he can post process to that machine. The computer is taking the same information, just regenerating the code for that specific machine.

"What I advocate is, rather than clogging up the machine memory and slowing down the memory processing time, buy a network download terminal--a cheap little clone will do. All the information is given electronically to this unit. All it does is emulate storage management of an individual machine. So in a sense you have two machines. One is doing the CAD/CAM and post processing and the other is replacing the metal filing cabinet with all the little blue boxes of tapes.

"That way everybody is sharing the same information. And, you can take it a step further. The download terminal could have a post processor in it.

"Then, raw geometric files (the GMT files) can be downloaded to it. So, if the operator that day is running say a Pega 357 with the 58 autolimits turret, rather than just dumping information to it, he first post processes it to that machine. That way, you don't have half a dozen different post-processed part programs cluttering up the memory. You just have one geometric file that came in from CAD, and then you post process it to your machine, whichever one you're running."

Using acquired knowledge

Of course, there are other things that become factors in determining which machine to run. These include material thickness, sheet size, and a lot of other things. Beyond cell load, putting a certain job into a particular cell might enable the operator to avoid tooling changes.

West stresses, "Yes, and I want to re-emphasize that it's acquired knowledge. But, once you acquire the knowledge, if you don't have the time to apply it, it's wasted. For instance, take myself. I spent 22 years in the industry. If I were to go back, it would take me a little bit of time to get back up to speed. Then, I would do it totally differently than I used to. This is because, until I came to work at US Amada, I did not understand that I was in that hole. I had dug a big hole, and I had blinders on. I could not see around me because I was so burdened down with detail. I had to get this done. I had to get that done. I never had the chance, or the process time available in my brain, to look and see I was doing it all wrong.

"You get squirreled away like that. You don't know it until you get removed completely from the situation, take a step back, turn around, and look at it. I could have been making ten times the amount of money--by just getting back and looking at it. This is what I'm trying to tell your readers. Keep an open mind. Give it a chance. Step away from it. Take an unbiased look back at your operation, and admit that you do need more time to develop a better thought process. By doing computer linking, computer networking, and getting your automatic machine tool operation to where you don't have a labor problem. you can release your brain to more powerful functions.

"In fact, once you get into that scenario, what happens then is your workers can then use their brains to better potentials. So your benefit is amplified by being automated: less learning, less craftsmen. Let the computer do more thinking for you."

PHOTO : CAD drawings from engineering can be copied into Amada CAD by floppy, telephone modem, or networking. The machine program can be generated in a little as five to ten minutes. Amada CAD/CAM accepts CAD drawings in two standard geometry file formats--DXF and IGES.

PHOTO : The engineer can review the CAD program before generating the G-code program file.

PHOTO : CAM software specified the correct tooling for the NC turret press operator.

PHOTO : An on-site terminal provides continuous update of schedule changes for maximum operator productivity.

PHOTO : Computer nesting of parts provides optimum sheet utilization. Parts for coin box can all be nested on one sheet. This simplifies scheduling and speeds assembly operations.
COPYRIGHT 1989 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Dobbins, Donald B.
Publication:Tooling & Production
Date:Jun 1, 1989
Words:3441
Previous Article:Presetting for tool-management systems.
Next Article:SPC software: ready when you are.
Topics:


Related Articles
Sheet-metal artistry for hire.
CAD/CAM - capturing the master's expertise.
The untended fab shop - can we achieve it?
Work team co-location works for foundries and customers.
James E. Bushman: From Student to Teacher at Cast-Fab Technologies.
LEBANON - Feb. 2 - Shopping Month Fails To Revive Economy.
Engineering Victory with agile CAM software: tooling-up for the first all-new American-made production motorcycle in nearly 60 years required...

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters