Device developers put a lot of time and energy into the design of a new product. However, they may forget that all design needs to consider the next stages of manufacture, including assembly. The assembly stage represents a place to improve a product's quality, time to market, and cost-effectiveness. The best way to achieve those improvements is to begin thinking about assembly early in the process; that is, during the design stage.

Working on design for automation up front is vital in medical devices because of the approval process involved. Once regulatory go-ahead is secured, it can be difficult to go back and make even minor modifications to a design, whether to reduce changeovers or any other purpose.

Automated assembly is a common manufacturing process that can be applied to medical devices. High-speed assembly and continuous motion assembly are processes that, as the names imply, create parts quickly.

In continuous motion, multiple processes occur without interruption for every cycle, effectively overlapping. Tooling never loses contact with the individual components, enabling part alignment during assembly. As a result, the process is smooth and unlikely to damage components. Continuous motion enables speeds from 400 to 1000 ppm.

For complex parts, it may be better to use intermittent motion. Intermittent motion systems, unlike continuous motion, pauses the process periodically to perform particular actions. Intermittent motion tops out at around 250 ppm.

For either case, the design of the component must be a focus. There are a variety of factors that must be taken into consideration to ensure that a product design is optimized for automated assembly. However, it takes expertise to translate a product's design into one that can be output through automated assembly. This article explores those characteristics and discusses the benefits and pitfalls found during design for continuous motion assembly.

Manufacturing, Shipping, and Storage Challenges

One important aspect of preparing for assembly is tracking parts between the time they come out of the mold and when they go into an automated assembly system. Tracking the details of this period helps ensure an efficient automation process. For example, when components are piled on top of each other in storage containers for several days, weight and heat in the container can cause warpage or other damage.

A precision continuous motion assembly machine is designed to handle parts of very specific dimensions. However, an automation designer must be able to calibrate the system to accommodate the widest range of possible misshapen components to maximize the efficiency of the assembly process.

For example, the sight chamber on an IV set is a small, flexible, plastic part that's usually shown in designs as a very straight tube. However, the straightness of the chamber part isn't critical to the part's function, and the reality of high-volume manufacturing is that no molded part like this is perfectly straight. The connection points at the ends of the chamber are what really matter. So even if the chambers become slightly bowed either in manufacturing or shipping, as long as the connection points are properly formed so that fluid in the IV will remain sterile and stable, then the bowed chambers are still usable. The machine should be designed so that it accommodates noncritical variances and that acceptable parts aren't thrown away unnecessarily.

Minimizing Changeovers

Equipment changeovers can be costly and time-consuming. Minimizing the need for changeovers should be a top priority in designing for continuous motion assembly.

Changeovers are often required to assemble different sizes of the same product line. Automation designers (in-house or contract designers with automation expertise) can nearly always find ways to implement common features—say, the same height or diameter for a certain component of different size products—to reduce the number of changeovers required or the amount of time spent doing so.

In the case of highly aesthetic medical devices, the product design and automation teams should work together to determine a common feature that both increases compatibility among different products and meets the visual requirements. In some cases an assembly feature can be made small or located out of view so that it does not affect product quality and is unnoticed by the product's end-user.

Feeding and Sorting

In storage, prior to being fed into an assembly system, various components of a product may become stuck together. It is key to ensure these parts come apart easily for a smooth, efficient assembly process.

An automation designer should look at each part and identify every possible configuration to consider whether they might stick together and cause system backups. To prevent parts from wedging together, a possible option may be to incorporate internal or external ribs in the components. For potential static issues, ionizers or special coatings can be added to the material.

Also, designers should consider any parts with irregular shapes. Most of these types of parts require some kind of buffer step for sorting and feeding. A continuous motion machine is not equipped to handle such parts. For example, many noninvasive medical devices are created with sweeping contours to provide visual appeal. For such a part to be handled in a high-speed automation system, there must be some subtle feature built in for sorting, such as a small tab that's invisible to the user and inconsequential to the use of the device. Otherwise, sorting may be the limiting factor in achieving the desired production rates.

External part shapes can also be a problem during the high-speed transfer from the feeders to the machine. Triangular, oval, and other asymmetrical parts can pose a challenge because they're prone to wedging and locking together during the machine-feeding stage. Often, fixing such problems can be as simple as adding a small orientation feature, such as a notch, to avoid jams. Such automation-enhancing features must be designed so that they facilitate the process but don't alter the functionality and appearance of a product.

Another challenge an automation designer can help address is matching the output of the manufacturing process to the continuous motion assembly system. The quest to meet market demands sometimes means that OEMs push the molds a bit (e.g., they try to get one more cycle in, even if the mold is finished), resulting in occasional short-shot parts, warpage, or other potential defects. An automation expert can compensate by determining which parts may be susceptible to short shots, whether such defects are acceptable, and how to inspect for such problems.

Problems with Flash

The locations and sizes of mold flash (excess material escaping the mold), gating, and other aspects that affect aesthetics should also be considered. Aesthetics are important, but it should be kept in mind that all designers (the product designer, mold designer, and automation designer) have different and often competing requirements. The product designer wants to hide flash and gate marks. The mold designer, meanwhile, is primarily concerned with how the mold opens and closes, as well as leakage issues. The automation designer needs to make sure flash and gate marks don't interfere with the transfer and feeding functions of the machine. For medical devices, visible defects may be unacceptable. Early and ongoing negotiation is often necessary to satisfy all of these concerns.

No-Touch Parts

In automation design, it's critical to identify and discuss areas of parts that are highly sensitive or cannot be touched at all (fluid paths, for example). Medical devices often carry such limitations and these components could be inappropriate for automated assembly. A certain amount of machine contact is unavoidable, particularly in the pickup area of an automation system. The pickup is a transition from sorting and feeding, where the system inevitably touches a part in two different places at the same time.

The system can be designed to avoid these no-touch areas. For example, in a typical syringe and piston assembly, one end of the piston can be touched—where the thumb is applied to administer the shot. The other end, usually a rubber seal, must be avoided. Again, the key point is the machine-feeding stage. Engineers would design the system to grab the thumb end of the piston, letting the no-touch rubber seal dangle as it enters the assembly machine.

However, no-touch areas aren't always obvious. That's why it's important to identify all such zones early in the process. Automation engineers can nearly always find a way to avoid these areas. But in the rare cases in which there are simply no good zones to touch in the machine, then adjustments in the product design may be necessary (such as extending a part to add a “sacrificial zone” that serves no other purpose than to be handled by the assembly machine).

In addition, the material involved may require a different manufacturing path than that offered by the assembly machine. For example, handles for medical parts and other devices often incorporate silicone rubber to make them easier to grip. The material is added to the plastic part while still in the mold, providing a tackiness that makes human handling easier, but automated handling more challenging. In such a case, the product designer might want to add a feature in other areas of the part—such as small tabs— to make sure the machine can handle the areas. The part should be oriented in the feeding stage so that the problematic handle is avoided.

The Right Fit for Automated Assembly

When assembly is done by human hands, people can cheat a little bit. They vary the axis, making adjustments as two parts are put together. Consider how one puts a lid on a plastic storage container. You don't slap the lid on all at once; you start at one edge and then go around the rim applying varying pressure.

Automated assembly machines, on the other hand, are more exact and less flexible during assembly. They don't offer the rotating orbital motion or on-the-fly adjustment of human hands. Instead, they put components together horizontally or vertically at a specified force. Variations from part to part can therefore cause problems in putting the parts together correctly and consistently.

However, with early data on the potential for noncritical part variations, an automation designer can calibrate the assembly system to account for the widest expected tolerance range in putting parts together. This requires looking beyond the tolerance variation on print to consider warpage, flash, and other common problems that are likely to affect how the parts will actually be received in bulk. No one wants to throw away parts due to noncritical variations, so it pays to design the assembly system to accommodate all expected variations.

Parts should also be considered in terms of how easy they are to assemble. For example, self-alignment features can make assembly between parts more forgiving. Consider tapering a part that requires a cap on one end. The tapering can facilitate the fit. Proper alignment within a tight tolerance is critical to minimize the amount of spring-loaded force required for the machine to put the parts together.

Designing an assembly process to a specified force applied by the machine, as opposed to designing to the specified dimensions of a finished product, is another way to improve the assembly process. If a system is designed to assemble to specific dimensions, with unlimited force applied by the assembly machine, parts that are even a tiny fractions of an inch larger than specified will be smashed by the machine as it tries to make the finished product conform to the proper dimensions.

A better way is to determine optimal force to perform the assembly successfully with the widest range of expected noncritical part variation. This optimal force can be determined through testing very early on, using prototype tools and parts. The force applied by the machine is precise and consistent, meaning that the system will assemble parts within the expected variation correctly every time (instead of crushing them, the way the system might do if assembling to specified dimensions). Meanwhile, the parts that fall outside the expected range of variation simply won't go together at all.

For example, consider a common margarine tub. Whether the tubs vary in height by 1/32 of an inch doesn't matter; what matters is whether the lid can be snapped on properly every time. If the machine is designed to assemble to a specified dimension, it could fall short or crush the parts with size variations. But if the tub and lid are assembled consistently with just enough force to snap the lid on (without crushing the tub), it yields a much greater rate of success and a lot less scrap. This arrangement typically requires that the parts are designed with their own assembly stop (i.e., the lid and tub have ridges that align so that assembly is complete when the lid snaps into place).

All of these factors influence the ability to produce parts at a tolerance close enough to maintain a proper fit in assembly every time.

Defense against Defects

Rapid inspection capability, along with the seamless elimination of defective parts, is one of the biggest advantages of an automated continuous motion system. To make the most of this function, however, OEMs must begin by clearly identifying exactly what separates a good part from a bad one for the product. Given today's competitive pressures, manufacturers typically want to minimize scrap and use as many parts as possible. A high-speed automated assembly system can be designed to inspect within very strict parameters or much looser specs. It's up to the OEM to decide what those specifications must be based on their application requirements.

With this understanding in mind, the objectives should then include identifying expected upstream defects and determining whether and to what extent variations would be acceptable. For those determined as defects that must be eliminated, an appropriate inspection method for each problem must be devised. By knowing the locations and nature of variations that may occur, the automation team can more easily place inspection probes that quickly and reliably identify the problems.

The goal is to make defects impossible for the machine to miss, and the right combination of part design and assembly techniques can help do that. For example, an OEM could adjust the designs so that the defective parts simply won't go together at a given assembly force. Either the components engage, or they don't. That makes the job of the probe charged with checking alignment simpler and more dependable.

Other Considerations

It's important for the part designer, automation designer, and the mold designer to work closely early in design. One reason to do so, beyond those mentioned earlier in this article, is the shift to lean manufacturing techniques. The principles of lean, which make good business sense, often result in fewer people working on more projects. Experienced automation designers can sometimes uncover ways to improve fit, dimensions, and tolerance of a product. In addition, automation machine engineering typically must happen concurrently with part design.

Conclusion

Continuous motion assembly is a process that can help medical device manufacturers achieve a quick product cycle. But such benefits can only be achieved when all parties understand the goals of the product and are in on the design up front. To ensure an efficient and effective assembly process, when considering any part change, the automation designer should be contacted immediately to work on any required adjustments to the assembly system.

Pat Phillips is engineering manager for Haumiller Engineeering (Elgin, IL).

Copyright ©2009 Medical Device & Diagnostic Industry