Overcoming the punching challenges of a high-product-mix perforation operation
August 1, 2013
Carter Day International punches thousands of holes to make specialized screens. One screen can take several hours to punch. Problem is, the company doesn’t have enough volume to warrant using a specialized perforation press. So it uses a traditional turret punch press, but with highly durable tooling.
At Carter Day International in Minneapolis it’s all about separation, in a literal sense. A 100-employee company, the equipment manufacturer has been making agricultural and petrochemical processing machines for more than a century. Barley, wheat, and anything else harvested from the fields for processing contain certain parts that have to be separated from other parts, and all that separation occurs in machines with screens with thousands of holes, between 5,000 and 100,000 per sheet (see Figure 1). Punched, rolled, seam-welded, and installed into specialized machines, Carter Day’s cylinders tumble and separate harvested material.
These manufacturing steps don’t seem too far out of the ordinary except for one challenge: Carter Day doesn’t have a dedicated perforating press—and for good reason. “We are a low-volume, high-variety operation,” said Dan Vesledahl, manufacturing engineer. “We’ll run three screens in the morning, and then process something completely different for the rest of the day.”
If the company had a dedicated perforating system, workers would spend much of the day changing out large tools, and the system would sit idle for much of a shift. Being a make-to-order operation, Carter Day would find it impractical to make screens to stock. The fabrication department might punch a screen one day and then may not punch that product for another three years.
The company has considered cluster punches, but here again the operation’s high variety presents complications. The fabrication department must cut numerous hole patterns, which in turn would require a large library of cluster punches. Finding and maintaining all those tools further complicates an already complicated operation.
“Say we process 15 different products,” Vesledahl explained. “We may have 15 different hole spacings and center-to-center differences.”
The high variety also makes it less cost-effective to outsource. The company does send out certain screens during peak times, but it really can’t take advantage of economic order quantities. Aside from raw material, every part in the shop has an order attached to it. Nothing is made to stock. Vesledahl recalled one recent situation in which he was thinking about outsourcing a job involving one hole-intensive screen. Carter Day could have purchased one screen for $1,200, bought two screens for the same price, or opted to make it in-house for $400. The last option won out, of course.
If the fabricator did take advantage of economic order quantities when outsourcing, it might be sitting on inventory for years. Eventually workers would be spending more time looking for the parts they needed and less time actually fabricating and assembling the product. That’s why Carter Day produces most of its screens in-house with a turret punch press.
This process worked for decades, but about four years ago the company began scrutinizing it. Sure, if it’s not broken, don’t fix it. But how exactly do you define what’s broken?
When Vesledahl arrived at Carter Day four years ago, he took himself out of the production process, at least in one sense. For years work orders had been sent to the manufacturing engineer, who programmed every job and sent that job to the floor, where sheet was sheared to length and then loaded onto the turret punch press.
Programming wasn’t easy or quick, especially when it came to jobs with tens of thousands of holes. Often the computer spent 20 to 30 minutes churning through all the computations. The procedure worked OK, though, and, because of the company’s make-to-order environment, it seemed logical. The next time a program needed to be run could be months or years away.
These days, even in a pure job shop environment, it makes sense to save a program in a library. Storage space on computer hard drives, servers, and machine controls isn’t a constraint anymore. Though the program may not run again for years, chances are it will run again someday, so why not save it? So for the past four years Vesledahl has built an extensive program library.
“I’ve built a database, and for the majority of the programs we do in-house, the program is already at the machine,” he said. The routing shows the operator the job number, which he can pull up on the machine control. “The job goes [from the front office] right to the lead-man in the [punching] area,” he said. “It cut myself out of the picture.”
Design changes don’t occur often, but when they do, Vesledahl retrieves the part program in the database. An engineering change order (ECO) may call for, say, a 12⁄64-inch hole to be changed to a 12.5⁄64-in. hole. “I now just pull the old part up, which has all the location-center marks and starting points, then just change the hole size.”
Carter Day’s intensive holemaking has several inherent challenges, the most obvious being tool life. Several years ago the company used conventional punch tools that often had to be changed out after every job, which is somewhat understandable considering one sheet can have tens of thousands of holes. The company does resharpen some tools, but because the punching applications are so intensive, it’s often been more cost-effective to simply replace the tool after it’s used.
A second challenge is the cost of a setup error. Small screens, like those 3 by 1 ft., take about 20 minutes to punch. But some of the larger screens—60 by 40 in. or even larger—may take up to two hours, and that’s just one sheet. So what if a punch is set to the wrong height?
Set the punch too high and it doesn’t punch through the metal; set it too low and it can stretch the web material around the perforation. If the operator doesn’t screw the punch into the holder just right, the punch-height position can change ever so slightly. This isn’t the end of the world if the error is caught early. But these applications have hours-long cycle times. Even if the perforation error were caught immediately after being offloaded, the machine just spent two hours producing a bad part.
To overcome this second challenge, the shop invested in quick-change punch technology, dubbed EXP™, from Wilson Tool International®. It uses a universal punch holder, so instead of switching from an A-type holder to a B-holder (the two types Carter Day uses), the operator now uses one type of holder for every job. And instead of screwing the punch into the holder, the operator gives a key a quarter-turn one way to unlock the punch from the holder to pull it out; he then pushes in the new punch and gives the key another quarter-turn to lock it in place (see Figure 2).
To overcome the first challenge, Carter Day now uses tool sets made of Ultima®, a proprietary tooling material from Wilson. Some of the less-used tooling remains conventional—working with such a high product mix, certain tooling isn’t used but once a year, if that—but the rest has gone to the new tooling material. According to Vesledahl, for one common screen the shop now can punch up to 18 before the tool wears out—and each screen has more than 18,000 holes.
Such perforation isn’t suited to the company’s laser cutting system, which spends its day cutting sheet metal pieces, including enclosures and brackets. Tying up the laser with a hole-intensive job would impede throughput.
But the company does perform some perforation work outside its turret punch press (see Figure 3). Some screens are outsourced, depending on order volume and capacity levels. Some of the thicker material, greater than 0.25 in., is sent to the company’s AKS plasma cutting table with an upgraded, high-density plasma system using Hypertherm’s True Hole™ technology, which cuts holes down to a 1-to-1 diameter-to-depth ratio. Such thick perforation would put excessive wear on the punch press.
The punch press concentrates on the lion’s share of screens, most of them thin and made of aluminum or deep-draw cold-rolled steel. The shop does process some stainless screens, but those are either outsourced or sent to a press brake retrofitted with a gang punch setup. The press brake ram isn’t fast, of course, but each hit produces about 100 holes, which helps reduce cycle times. The rare stainless job requires unique tools and an entirely different setup than the mild steel and aluminum, Vesledahl said, which is why it makes sense to run the work on a separate machine.
Vesledahl is the first to admit that running such hole-intensive operations on a conventional punch press isn’t ideal, though it’s the best for what’s available off-the-shelf. Dedicated perforation systems are designed for higher-volume situations. That said, Vesledahl added that he is starting to work with a firm to build a gang-punch press designed for Carter Day’s high-mix, low-volume environment. Still, the custom machine would be used only for frequent perforation jobs, not every screen. The variety of hole sizes and orientations is just too great.
Fabricators know the challenge of high product variety all too well, and Vesledahl is tackling the challenge like many others have. He identifies product families frequently produced and develops ways to optimize that production.
Vesledahl added that the company does not increase throughput for the sake of increasing throughput. Downstream demand drives the pace of manufacturing, and the punch press itself usually hasn’t been a bottleneck. Like in any high-product-mix operation, demand rises and falls for various products. And if demand does rise significantly in a short period, it usually comes from a large-volume order—say, for 50 screens. At that point it becomes more cost-effective to outsource.
“We’re a small company, and we don’t have a lot of time to experiment,” Vesledahl said. “If it’s not broken, we don’t try to fix it.”
All the same, defining what’s broken can be a matter of perception. For Carter Day, enhancing in-house capability—with changes in programming, new tooling, and perhaps a custom gang-punch machine down the road—has increased its flexibility, reduced changeover times, and sped overall response time. The faster and more consistent (that is, less variable) an upstream operation such as cutting can be, the less work-in-process buffers are needed downstream.
The sheet metal cutting department produces about 80 percent of its products in-house, but the amount that’s outsourced isn’t as important as what is outsourced, and this often depends on a job’s quantity. As Vesledahl sees it, the better the sheet metal cutting department can respond to various levels of demand, the more choices the company has when it comes to outsourcing, contracting only when there is a clear financial advantage (that is, sufficient volume) in doing so.