A good product, one that people will want to buy, needs to be greater than the sum of its parts. But, what happens if one or more of those parts are out of spec? You may find yourself with excessive rework costs, excessive scrap or, worst of all, unhappy customers.
Unfortunately, even in this era of Six Sigma and digital assembly technologies, the occasional defective part, or parts, is a fact of life. One of the downsides of using foreign suppliers in low-cost countries like China and India is that quality can sometimes be an issue. Then again, given the high-tech nature of many of today’s products, the margin for error is smaller than ever. The automotive industry continues to be plagued by recall and warranty problems. Even Toyota, the gold standard in terms of quality, has its issues.
In practice, manufacturing engineers can attack the problem of parts quality in two broad areas-in the supply chain or on the shop floor. Depending on a company’s particular needs, it may choose to focus its efforts in one area or another. But, in most cases, it makes sense to address parts quality throughout the value stream.
As Gary Kone, president of automated assembly machine manufacturer FTT Manufacturing Inc. (Geneseo, NY), puts it, there’s no one practice that will ensure only good parts go into your assemblies. “It takes a series of little things to be a quality manufacturer,” Kone says.
Getting What You NeedIn many ways, the key to ensuring your suppliers provide you with what you need is thinking long and hard about what you are paying for and what exactly you require. At the most basic level, this means employing suppliers who are reliable and will work with you in the event things go wrong. Unfortunately, this will not always make your accounting department happy, because it may require saying no to the lowest bidder.
For example, the decision to outsource to low-cost providers, either domestically or abroad, may cause problems when the time comes to actually go into production. Are they really all they claim to be? Are they truly a “Six Sigma” company, or is that just marketing? Do they have the will and the technical ability to back up all the promises they made when the two of you sat down to sign your purchase agreement?
“Cheapest price is not lowest cost. Never has been, never will be,” says Larry Stockline, president of electromechanical press and process-monitoring equipment manufacturer Promess Inc. (Brighton, MI). “Offering a cheaper price because you do not have the capital investment for inspection equipment...will come back to haunt many users.”
In electronics assembly, in particular, it pays to know exactly what you are getting and whom you are working with. This is especially true given the current “gray market” of undocumented or counterfeit goods that is especially prevalent on the other side of the Pacific.
“[The] reasons for receiving bad parts are numerous. These include human error, bad parts lots at the manufacturing stage, and also, at times, parts [that] remain on the shelf too long and consequently may not perform optimally to the specification,” says Zulki Khan, president of electronics manufacturer NexLogic Technologies Inc. (San Jose, CA). “A contract [manufacturer] is always advised to buy components from known and reputable sources. That can include some brokerage houses, but mostly from VARs (value-added resellers) and authorized distributors.”
Khan notes this practice can be especially important when procuring high-tech, active components, such as integrated circuits, field-programmable gate arrays, digital signal processors and static random access memory units. “Contract manufacturing buyers should specify date codes for these components. What this does is...help ensure that contract manufacturers are buying components which are reliable and can be traced back in case of problems,” he says.
No matter what the product, Kone advises those buyers considering foreign vendors to remember that one of the hidden “costs” of doing business with, say, Vietnam, is distance. It can take weeks for bulkier items to make their way from Asia to a final assembly plant in the Western Hemisphere. As a result, there may already be additional shipments in the pipeline by the time a problem is discovered.
Then there is the simple fact of having to deal with multiple languages, contrasting cultures and different time zones. “It’s nice to be able to just reach out and touch someone,” Kone says, regarding the advantage of domestic suppliers.
Knowing What You WantOf course, a supply chain, like a good marriage, is very much a two-way street, and it pays to try to anticipate assembly issues in advance. For example, is there some way you can design your product so that its constituent parts will be less likely to cause problems when it comes time to put them together? Then again, are there ways you can frame your needs in such a way that your suppliers are that much more apt to meet with success?
In the electronics sector, Khan suggests that design engineers strive, whenever possible, to use components that have stood the test of time and are available through multiple sources. That way, if your initial supplier lets you down, or gets behind schedule, you can go with someone else. Granted, Khan says, it is not always possible to design away all future problems. Nonetheless, it can’t hurt to try.
With regard to other types of operations, manufacturers and machine builders should do their best to anticipate any problems that could prevent smooth assembly. For example, Kone whose company also manufactures components for other assemblers-which means he is actively involved on both sides of the assembler-supplier equation-points out that when working with plastic components, some variability is a fact of life. Plant temperature, plant humidity, the time that will elapse between when a part is made and when it will be processed by an OEM: These and other factors will inevitably affect product quality and final assembly.
Similarly, Boris Baeumler, applications engineer at screwdriver and automation company DEPRAG Inc. (Lewisville, TX), says that by selecting the right threaded fasteners, manufacturers can avoid difficulties when they ramp up to full production. For example, manufacturing engineers can identify whether a particular fastener will have a tendency to jam an automated feed system. They can also recommend design changes that will help ensure successful final assembly.
“That’s a key role people like us play,” Baeumler says. “We can recognize issues and say, ‘hold off a minute, the direction you’re going in will cause problems.’”
Then there is the question of variation and problems that may result even when two or more parts are technically “in spec.” Let’s say, for example, that pin “A” needs to fit into component “B.” If part A is at the high end of its tolerance range and B is at the low end, a slip fit may suddenly feel an awful lot like a press fit-despite the fact that you are working with “good” parts. Then again, if A is at the small end of the range and the hole in B is just on the acceptable side of too large, the final assembly may experience an unacceptable amount of slop.
“It’s a situation where you have plus and minus tolerances,” says Stockline. “If you have a plus on one side and a negative on the other, you may have problems.”
With this in mind, Kone suggests that whenever possible, assemblers should outsource multiple, high-tolerance parts to a single supplier. That way the supplier will have a better feel for the big picture and how its particular product, or products, function as part of a larger whole.
Obviously, in many cases this approach simply isn’t possible. For example, if a product calls for the use of a number of different materials or technologies-say, a composite filter, a plastic cap and a metal tube-then more likely than not you are going to be working with a number of different companies. However, by bringing production engineers into the picture early on, a company may be able to nip any future compatibility issues in the bud.
“In high-tolerance assemblies...we take the tolerances and ‘bend’ them in such a way that they fit better,” Kone says. “We look at how those tolerances overlap from part to part, and how they go together.
Better still, assemblers can make the issue of compatibility somebody else’s responsibility by outsourcing production of an entire subassembly to a single manufacturer. This approach is becoming increasingly common in everything from aerospace to the automotive sector.
For example, Boeing, which once fabricated everything from wings to landing gear in-house, now receives entire wings and fuselage sections preassembled at its final assembly facilities in Renton and Everett, WA. It then simply “snaps” together these prefabricated pieces to make up a complete airplane.
No matter what you are assembling-or how you are assembling it-Baeumler warns that manufacturers need to make sure the sample parts they use when setting up a production process are the same as those that will be used during actual production.
For example, Baeumler says that when engineers at his company are building an assembly machine, they insist on being given actual design specifications and not just sample fasteners. That way they can be confident their equipment will perform successfully on the shop floor.
Kone agrees, saying he’s seen plenty of instances in which there were all sorts of differences between sample and actual parts. This can be especially vexing with plastic components, he says.
Ensuring QualityOf course, no manufacturer can rely on its supply chain to ensure that every part it receives is in spec. Therefore, assemblers need to police their own processes as much as possible to weed out bad parts before they unduly affect the bottom line.
Increasingly, this is being done via in-line testing, so that companies can flag defective assemblies before they have a chance to travel any further than necessary down the value stream. For example, many automated assembly machines now include one or more test stations that do everything from check for leaks to evaluate electronic or mechanical functionality. Similarly, today’s DC electric screwdrivers provide instant feedback in the event a threaded fastener fails to perform as specified.
It may also make sense for an assembler to inspect at least of some of its parts before they arrive on the assembly line, or implement some kind of system to screen problem components before they are installed. For example, Stockline says some of his customers now employ probes to determine the exact height of all bearing shoulders before pressing together their bearing assemblies. This approach not only allows a manufacturer to weed out defective parts; it also means a “smart” press can use the information to ensure it presses to the correct depth for each individual bearing.
Along these same lines, Kone cites the example of an automated assembly machine his company built that included a pass-fail mechanism to ensure that all the flat seals going into a certain product were truly flat. In this instance, the problem parts-seals with an unacceptable amount of curvature in their material-were not so much compromising product quality as causing expensive machine jams.
Similarly, Baeumler says his company sometimes incorporates optical sensors into its screw-feeding systems to detect any parts that have incorrect lengths or head heights. He warns, though, that these systems can be expensive, and it may be more cost-effective to have suppliers do the screening, even though it may make the parts slightly more expensive.
With regard to electronics components, Khan says contract manufacturers should employ a rigorous inspection system to ensure that all incoming components’ footprints, values, tolerances and date codes are acceptable. “It is important to have a strong incoming inspection to prevent bad parts from leaving the audit area and into the assembly floor,” Khan says.
Granted, these kinds of measures all carry a cost. But, it’s important not to be penny-wise and pound-foolish. As Kone puts it, “Engineers don’t do cost accounting, and accountants shouldn’t run engineering.” In other words, don’t be afraid to incur some additional expenses up front to avoid even bigger headaches down the road.
Finally, assemblers might want to explore the possibility of using today’s process-monitoring technologies to ensure performance-based quality in each assembly, irrespective of variability in that assembly’s constituent parts.
For example, Stockline says he had a customer who was having trouble meeting increasingly rigorous standards for the hydraulic check valves he manufactured for antilock braking systems. To ensure the valves functioned correctly, this customer needed to use ultra-high-precision bores, springs and balls. After assembly, each valve underwent a pressure test. If it didn’t maintain a pressure within 175 psi of a 1,000-psi target, the valve would be scrapped.
To solve the problem, Stockline’s company built a system that turned the entire process on its head by first filling the valve with fluid and then pressing in the ball and spring to the correct pressure-1,000 psi. Overnight, getting the check valves to perform correctly became a piece of cake. They now regularly fall within 5 psi of the target pressure. The valve maker has also been able to save money on parts, because it no longer needs them machined to such precise specifications.
“We twisted the whole design concept into a function,” Stockline says. “In the old days it was all about dimensions, that was easy to understand...this is about function, how it is that you use the product.... I wanted to get inside the tolerances, zero [variation] should be the goal.”