Metal fabrication’s albatross: Poor material quality

Modern automation technology in metal fabrication can accomplish some incredible feats, even in high-product-mix operations. Still, material quality remains the wild card. Slobodan Miljevic / iStock / Getty Images Plus

In recent years I’ve visited more than a few operations dealing with a wild card that seems to be getting wilder by the year: poor material quality. It remains the hidden cost that prevents many operations from being as productive as they could be. A lot of it comes from siloed operations where purchasing departments are laser-focused on getting the best deal they can. The actual laser in manufacturing (in another isolated silo) sits idle as personnel fuss with subpar material.

Subpar material isn’t new, but it’s getting more attention now than ever. Why? It might be because, as metal fabrication technology advances and more turn to automation to boost throughput, the true cost of cheap material skyrockets.

Some of that cost has always been there, of course, though it hasn’t always been tracked that way. Consider material that, due to distortion, needs to be microtabbed within the sheet metal nest in such a way that slows part removal and shake-out. Costs within the machine might be low, considering the speed and efficiency of modern equipment, but what about the pile of cut parts sitting in the offload station or stored in a tower? A shop might have an amazing laser that runs lights out, but people still might be using hammers or other hand tools to remove stubborn parts from the skeleton. And then comes the deburring to knock off all those tabs and clean up the edges.

The root causes don’t all stem from bad material. Material geometry might have some tight corners, and some strategic toolpaths (like cutting extra around certain geometries) could have helped. Some modern lasers can even shape their beams to widen kerfs and facilitate easier part removal. Variability with slat conditions can play a role too: the gunk on them raising material slightly, which in effect changes the laser beam cutting condition (different focus point and assist-gas flow characteristics); continually cutting directly over slats; failure to clean slats properly and regularly. The same goes for regular laser maintenance, from cleaning the bellows to maintaining the chiller.

But when you have bad material, it’s tough to work around physics. If material is stressed more in some areas than others, cutting is bound to relieve that stress enough to cause material movement. Significant distortion requires more tabbing or, at best, strategic cut sequencing to prevent head crashes.

All material releases some stress when cut, but what if that release of stress didn’t cause significant distortion? Fewer tabs might be needed, and programmers could place them strategically for easy part removal. No distortion also opens the possibility to run with no microtabs whatsoever. Some nests could be cut with common lines, which can send material yield through the roof and, in essence, reduce the true cost of that more expensive stock.

Good material also opens the door to part-removal automation, either with a robot or cartesian system. These help reduce variability even further as automation stacks parts neatly onto pallets. When the cutting machine cycle ends, the parts are truly complete, ready to be taken to the next operation. That can’t be said for most North American laser cutting operations today. Even in operations with a host of material handling towers, laser cut parts sit tabbed into skeletons—either in a pile or stored in a tower, waiting for someone to shake them out and deburr them. Some phenomenal levels of laser cutting automation is often followed by someone knocking a microtab off part edges with a grinding wheel or, at best, manually feeding workpieces into a deburring machine.

With part removal automation, on the other hand, parts physically can’t be tabbed in place (mini-nests aside). That eliminates at least one complication for operations downstream.

As with any technology, tradeoffs abound. Material yield might fall with automated part stacking since parts need to be oriented in certain ways for the suctions to securely grasp them. Some small parts might need to be grouped in mini-nests, which need to be snapped apart manually. Oily surfaces on some material cause issues of their own. But often, if material quality (consistent chemistry and thickness) isn’t on par, part removal automation just isn’t a reliable option.

The same can be said for operations downstream, including bending. A material might have a minimum yield that’s specified, but what about the maximum yield strength and ultimate tensile strength? The more those levels vary, and the narrower the window you have between yield and tensile, the more challenging the bending operation becomes. The same goes for material thickness tolerances.

The whole point of automation is to reduce variability, but the opposite can happen when you mix automation and poor material together. Yes, some machines do have intelligence to sense material surface imperfections and adjust accordingly. The same goes for adaptive bending technology on press brakes. Overall, though, feeding subpar material to incredible manufacturing technology is a bit like driving a luxury car on a dirt road versus the Audubon. Yeah, the car can make it through, but the Audubon (topnotch material) gives a far smoother, and superior, ride.