Compared with a screw-together design, a snap-fit assembly requires more engineering time. A snap-fit needs many features to work correctly, such as sufficient length for flexibility, space to operate in, and a ramp to deflect the snap during assembly. In addition, snap-fits almost always require an undercut, either for the hook or the receiver. To form these undercuts, complicated mold features, such as side actions or shutoffs, are often needed. Side actions, or lifters, not only add to the cost of the mold, but they often need space in the part to operate. Holes from shutoffs can be problematic if they let light and dirt inside a product, and they are not attractive on the outer surface of a product. Special features are required to keep the parts aligned while the snaps are being engaged, and extra features are often needed to allow the snaps to be released for disassembly.

All these factors must be addressed when designing parts to snap together, and it can be very time-consuming to correctly design and analyze a snap-fit. When designing a snap-fit, the question of moldability must be kept paramount. This greatly limits your options for designing the part. Conversely, a screw-together design needs only a pair of holes.

Most of the cost of a mold is the labor to make it. A mold that takes longer to make costs more than a mold that is easy to make. It has been stated that "every pull eliminated from a mold cuts about $5,000 off its cost." Moreover, complex molds often require more maintenance than simpler ones. And, because molds with side actions often cannot run as fast as molds with no actions, fewer parts can be produced per unit of time.

Another factor that can add significant time and cost to a snap-fit assembly is the inherent risk that the snaps will not work correctly the first time. Snap-fits are often designed to be "steel safe." This allows engineers to remove metal from the mold to tune the snap to give it the right feel. However, in doing this, engineers assume that at least one more trip to the mold maker will be needed to modify the parts.

Plastic assemblies do not behave simplistically in testing, particularly under impact. A plastic part is inherently elastic. Under an impact load-a drop test, for example-an assembly can often distort elastically to amazing levels, unhooking snaps that appear very secure. This means that after product testing, snap-fits often must be redesigned at a cost of time and money, significantly delaying a program. This happened at Polaroid during testing of the ComboCam, a combined instant and digital camera in which snap-fits turned out to be a major problem. During impact testing, the body panels separated and exposed the high-voltage strobe circuit, a condition that would prevent it from obtaining a UL rating. The team went through three iterations of snap redesign and tool changes to solve the problem, and still the camera failed. Finally, the snaps were replaced with screws, which passed the test the first time. Two months of program time were lost, delaying the product introduction. Of course, there were also additional engineering and mold shop costs to redesign and change the molds.

Comparing Costs

To quantify the comparative costs of assembly techniques, we designed a small, two-piece enclosure for a circuit board as a representative test case. Two different versions were designed: one with screws and one with snaps. The CAD models were then given to design engineers for them to estimate the design time required to create the models, including the time to design the snaps. Next, the models were given to a manufacturing engineer to estimate the tooling costs for the parts. Finally, the product assembly time and screw costs were estimated.

Five engineers estimated the time needed to design the parts as either a screw-together assembly or a snap-fit assembly. According to their figures, the snap-fit assembly would take three times longer to design than the screw-together assembly. It's important to note that their estimates do not include the time needed to conceive the snap-fit and analyze it to ensure it might work. This additional time would probably double the design time for the snap-fit assembly, making the snapped parts six times more costly to design than a comparable screw-together assembly. To design some of the sophisticated snaps used in the past at Polaroid, it sometimes took a week just to invent a way to make a snap that could do the job and be moldable.

Next, we estimated the time and cost to build the tooling for the two different assemblies in both the United States and China. According to our figures, the mold for the snap-fit cover would require 260 more hours to build than a mold for a screw-together cover. If the tooling were made in the United States, the molds for the snap-fit parts would cost $32,000 more than the molds for the screw-together parts. If the tooling were made in China, the price differential is smaller-$12,800-but still substantial.

These results beg the question: Can the higher cost of snap-fit tooling be offset by eliminating the cost of screwing the parts together and the cost of the screws themselves? Manufacturing engineers estimate that a typical assembler would need 15 seconds to nest the parts and install four screws. That equates to four units per minute or 240 units per hour.

If the cost for an assembler in China is $1 per hour, the labor cost to assemble our test product is less than $0.005 per unit. If the cost for an assembler in the United States is $16 per hour, the labor cost to assemble our product is $0.07 per unit. In China, we pay $0.002 per screw. In the United States, we pay $0.01 per screw. So, if we add $0.008 per unit for the screws in China and $0.04 per unit for the screws in the United States, the total cost of the screw-together assembly ranges from $0.013 to $0.11 per unit.

For simplicity, we will ignore the engineering costs and just look at the tooling cost to determine break-even quantities. In China, the tooling costs $12,800 more, so a break-even quantity for snap-fits would be 1,070,000 units. In the United States, the additional cost of tooling is $32,000, so the break-even quantity would be 290,000 units.

Another factor is the time from when the tools are started until the tools are ready to make parts. Remember, the tooling for the snap-fit assembly takes 260 hours longer than the tooling for the screw-together assembly, assuming no revisions are needed. This is a 160 percent increase in time over the screw-together design and represents a schedule difference of up to 6.5 weeks just for the tooling.

The longer engineering and tool fabrication time for snap-fit parts yields longer design cycles, and longer time-to-market. This added time should be considered as a cost, not only for the development costs incurred, but also for tying up working capital and delaying the return on investment of the product. Assuming these costs are amortized over the product run, the cost benefit of using snap-fits may be lost unless the number of units produced is very high. The value of a product delivered to the marketplace is a combination of the product's unit cost, development cost, and the timeliness of the product's delivery to the market. When all these factors are considered together, the true cost of using snap-fit assemblies may be higher than initially thought.

Snaps can be an inexpensive and effective assembly technique and can reduce the unit cost of a product, but their hidden costs can be more expensive than you can afford.