Our Sites

Fabricator adapts laser cutting to single-piece part flow

Programming automation and strategic batching play key roles

A cut sheet emerges from the laser at Percival Scientific. To make part removal and sorting easier, the company rarely tabs-in a part, unless there’s just no other way to keep the cutting process stable.

How does a fabricator keep up with the fiber laser? The conventional answer involves automation, from automated load/unload to material handling towers, all the way to sophisticated storage and retrieval systems, with shelves of active inventory spanning the shop floor.

Such automation can grow with the shop, as managers work to ensure the shop’s constraint operation—be it in bending, deburring, welding, or anywhere else—is never starved of work. The throughput of the constraint governs a fabricator’s output, and if a constraint is starved of work, the shop isn’t producing as much as it could be. Sure, the constraint moves depending on a custom fabricator’s order mix that day, week, or month. But one thing’s probably for certain: The constraint itself isn’t likely to be the cutting process.

Or is it?

The story of Percival Scientific’s sheet metal fabrication operation in Perry, Iowa, is a reminder of what drives technology investment: the demands of customers, both external and internal (that is, those at downstream operations like assembly). Customer demands at a large custom fabricator are far different from Percival’s, a high-product-mix, low-volume equipment manufacturer of environmental-control chambers.

“We build incubators from the size of a small dorm refrigerator to commercial walk-ins large enough for growing corn, and everything in between, said Jesse Smith, Percival’s director of manufacturing.

When he and other managers analyzed flow in the fabrication shop, they knew the cutting department would soon become a constraint. If things didn’t change, it would prohibit future growth.

Their approach involved more than simply buying a new machine. They instead took a holistic look at the cutting process, from order release, programming, and order sequencing to offloading and sorting parts for the next operation. In this sense, if parts aren’t ready for the next operation (usually bending), then laser cutting really isn’t finished.

Moving Away From the Shear and Punch

Five years ago Percival’s cutting department consisted of a shear and two turret punch presses. It seems old-school, but it wasn’t entirely, especially considering the company’s single-piece, or single-unit, part flow production.

Percival’s history, which goes back more than 125 years, is one of adaptation. The company began in the early 1900s making butcher knives. After the invention of refrigeration, the company adapted and expanded into butcher cabinets and refrigerated meat lockers for food producers and grocers.

In the 1960s Iowa State researchers reached out to the company about making a refrigerated cabinet with a more controlled environment. This in turn led to products that involved environmentally controlled chambers for the research market, both in universities and in private industry. Over the years the company pruned its product offerings and, in the 1980s, began to focus entirely on the research market, and that focus remains to this day.

Every part is etched with a part number to ensure proper sorting.

When Percival fabricates and assembles a unit, it likely won’t build the same unit again. It is within this context that the company’s fabrication department evolved.

Considering the product mix and business model, fabricating parts to stock just wouldn’t work. Some custom fabricators might build ahead for a large customer, hold some work-in-process (WIP) for components used on multiple products, or perhaps hold some finished goods for certain product lines. Percival’s product mix is just too unpredictable for this. Everything is made to order. So over the years the company has built its fabrication and assembly operation around a single-unit part flow.

“We were doing individual, custom jobs,” Smith said. “We would take the job orders and create bills of materials—all different sizes and types of materials—and create drawings.”

Once a customer order was released from engineering, the cutting operator received a cut list—again, for a single unit—which he’d use to draw material from raw stock. Product designs incorporate about six different thicknesses of material, most of which is carbon steel, including galvanneal, with some stainless and aluminum thrown into the mix.

Cutting, including the nesting and programming, essentially was manual and prone to human error. The operators sheared each sheet to size, then put that sheet on a cart and wheeled it over to the turret press, which cut only the parts needed for one unit and nothing else. Operators then did the same for the other material grades and thicknesses (mostly thin sheet, like 22 and 18 gauge). They then placed cut parts onto carts that held every piece for a single unit and wheeled them on to the next operation, usually bending on a press brake or a bidirectional folder, depending on the sheet size and bend geometries required.

The single-unit part flow simplifies part flow, but the old-school shear and punch arrangement just wasn’t very scalable. It wasn’t just the machine technology, it was the way jobs were scheduled. Adapting static nesting (in which all parts on a nest layout go toward a single job) to Percival’s product mix wasn’t easy. Operators spent their days managing remnants and switching out materials.

Smith uses sales history and forecasts to plan production, with the goal of having a certain amount of work related to a certain amount of revenue ship out the door within a certain period. “We know how many sales dollars go through our facility,” he said, adding that to reach a certain amount, he needs a certain amount of work flowing through each department. Considering this, Smith knew that in the long run, the cutting department just wouldn’t be able to produce enough to keep pace, given space and labor constraints—with workers retiring and qualified replacements extremely tough to find.

Keeping the Pace

In 2016 the company purchased its first laser, a 2-kW machine from Mazak Optonics Corp. (again, the company processes thin stock), along with a nesting and programming package from Radan.

But machine and software upgrades were just part of the story. With these new systems, Percival adopted an entirely new way to process cut parts. The method adapts the dynamic nesting practices common among custom fabricators to Percival’s single-unit part flow. In essence, Percival has added a little batch-style processing to its single-piece part flow.

As the design team releases orders and exports DXF files from CAD, Smith identifies enough orders for about two to three days’ worth of work—about five orders on average. (Though considering the company’s product mix, that average is based on an extremely wide range.) He weighs various factors to choose these orders, including the scheduled due date and, especially, the needed value of sales dollars on the floor to meet the expected sales numbers.

Workers place all the sheet metal components for a single unit on carts, which flow with the job through bending, welding, and assembly.

Using customized software, Smith pulls these orders from ERP and filters out sheet metal parts to be nested. “From there, I combine all of those parts into a spreadsheet that the [nesting software] is able to read and load from the bill of material generated by the ERP system.”

This tight “handshake” between the ERP and programming streamlines nesting significantly, but Percival needed one more element to ensure the cutting department could really scale up for the future: part identification.

Percival’s part mix is widely disparate, but it just so happens that many of those parts look alike. This is one huge benefit of single-piece part flow over batch production. A laser cutting machine can be incredibly productive, but if parts get lost or are sorted incorrectly and, due to mistakes downstream, must be recut on the laser, well, what good is all that laser cutting productivity?

For this reason, the company identifies every cut part that emerges from the laser. Still, it doesn’t rely on stickers. For one thing, stickers can leave a residue that needs to be cleaned off prior to powder coating. And someone of course has to apply those stickers, again opening the door to human error.

So the company relies on etching part numbers on each part. Does this slow the laser cutting cycle time? Sure, but it’s faster than having to recut a lost part. The nature of the company’s nest layouts also makes etching practical. Part sizes run the gamut, but the company rarely cuts a nest full of tiny parts, each requiring a laser-etched number. If the rare nest layout with hundreds of parts does come up on the schedule, it’s usually run last on the shift, available for sorting the next morning.

The real constraint for on-machine labeling is programming. Finding the best place for a part number on every nest would have been time-consuming and arduous, especially considering the shop never runs the same nest twice. So here Radan customized a part identification module to automate the task.

“The system will find the perfect location to etch the numbers on each part,” Smith said.

Next Steps

After the cutting process, workers denest and sort the cut parts, then manage the skeleton, which the laser usually scrap-cuts for ease of handling. Also for ease of handling and sorting, the operation avoids microtabbing parts in place. Most parts are large enough to remain stable throughout the cutting cycle; and if they aren’t, they’re nested across slats.

Reading the etched numbers, workers sort and place cut pieces onto carts, which are then rolled to bending—an area that may well be the company’s next area for improvement. Today the company’s press brakes and folders bend one unit at a time. Pieces for one product may be split between the folders and brakes, but immediately after forming they return to the cart to continue their “one-unit-flow” journey to welding, coating, and assembly.

One issue is that this flow adds setup time, particularly on the press brake. The company recently invested in a new electric brake with off-line bend simulation and programming, but the operator still needs to switch out tools many, many times a day.

Here, Smith’s long-term plan is to carry the batching practices in laser cutting to the bending department as well. So instead of pulling enough cutting jobs to fill two or three days of work, the scheduler would pull a certain batch of work to undergo the “cut-bend” process, with material due in their respective carts by a certain time and date, at which point they’ll flow downstream to the next process.

Smith and his team continue their holistic approach toward improvement, and it’s not just about looking at inches per minute on the laser or bend cycles per minute on the brake. Sometimes low-hanging fruit can be found by just spending a few minutes watching people work.

Smith recalled one walk through the shop where he noticed welders spending a lot of time peeling protective vinyl wrap off stainless steel—a wrap that itself slowed the laser cutting cycle time. “So I just started ordering stainless steel without it, and we just put a piece of parchment paper between the sheets in the stack.”

He added that this approach wouldn’t work for cosmetically critical parts, or if handling requirements caused major scratches. “But we’ve tested our stainless products made both with and without the vinyl wrap, and presented them to our sales and marketing teams, and no one could tell the difference. And it’s saved us a ton of time.”

Different Journey, Same Goal

With a history in batch production with hard tooling, most custom fabricators produce ahead to fill a sheet, batch similar orders together for maximum machine efficiency, and yet face mounting WIP as a result. So they work to reduce setup times, reduce batch sizes, and implement a combination of dynamic and static nesting to increase overall throughput and get closer to single-piece part flow—but not so close that they end up cutting just a few parts on each sheet and spending too much time moving material and managing remnants.

Percival is coming from the other direction, but it has the same destination. Years ago the company achieved single-piece part flow; now the fabrication department is refining that approach and batching strategically so that it can get more out of its available machines and people.

“To put this in perspective, prior to the laser, we had nine people in the cutting department, and operators were working 10-hour days and putting a half day in on Saturday just to survive and keep up with the rest of the plant. Now when we’re humming, we have four people and no overtime.”

He added that those extra cutting operators haven’t left the company, but instead have been retained and retrained to work in other areas of the plant. “We’ve gained a ton of labor efficiency during a time when qualified labor is incredibly difficult to find.”

Photos courtesy of Radan, www.radan.com.

Mazak Optonics Corp., www.mazakoptonics.com

Percival Scientific Inc., www.percival-scientific.com

About the Author
The Fabricator

Tim Heston

Senior Editor

2135 Point Blvd

Elgin, IL 60123

815-381-1314

Tim Heston, The Fabricator's senior editor, has covered the metal fabrication industry since 1998, starting his career at the American Welding Society's Welding Journal. Since then he has covered the full range of metal fabrication processes, from stamping, bending, and cutting to grinding and polishing. He joined The Fabricator's staff in October 2007.