|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
By Larry Forsythe, ITW Camtex, Arlington, Texas
The electronics industry utilizes strapping media to hold bundles of trays together for handling, packing and transportation. Since the packing process may seem remote and less important to specialists concerned with IC packaging, the logistics of packing and shipping are often not monitored as closely as the manufacturing operations within the assembly arena. The materials and equipment performance, as well as the monitoring processes utilized in this last step of the operation may not be closely scrutinized until catastrophe strikes. Equipment Selection Selections vary from the least expensive hand ratchet metal-banding manual configurations through fully automatic systems. These systems are capable of handling highly pliable polypropylene and polyethylene banding materials. Focusing on the more automated models and limiting material selection to two types of strap material still leaves numerous considerations for variations in grade, thickness and width before purchasing. Conformance capability, flexibility, thermal endurance and annealing requirements are as much driven through equipment selection as product processing requirements, such as "dry bake." Material properties, which deserve consideration, include elasticity, tensile strength, surface finish, shrink rates and ESD characteristics. All of these considerations suggest that the strapping and the materials involved in this process should be given a higher status and level of control than a "commodity mentality" may dictate. Before embarking on material selection, a brief discussion of equipment capability is in order. For simplicity's sake, equipment considerations will be limited to semiautomatic and automatic strapping configurations, the types routinely used in the electronics industry. Critical Elements Strap tension and throughput rate are the two most critical elements involved in selection. Defining the significance and criticality of these elements provides the basis for material and process definition. Most current packaging operations do not require a fully automated system. As automation continues to migrate into all aspects of this technology-driven business, however, automation may be an item to be considered during the capital investment cycle. All of the automatic systems (and some of the semiautomatic systems) impact a brick (stack of several trays) with at least two tension cycles. Generally, executing a completed strap subjects the brick to a multiple cycle process, which typically starts with take-up of free strap material. The second phase is commonly referred to as the "crush cycle" and occurs as either a mechanical clutch or electronic counter pulls the loop snug. 'Peak Tension' A "peak tension," often achieving a threshold of 3-5x greater than the final tension, is created by design to distort the material undergoing strapping, with the presumption that there will be some level of memory recovery in the crushed material that will allow recovery to the original shape-thereby providing a controlled snug tension. This cycle is achieved with a "hot shoe" in place, which serves to reflow the banding material, creating a welded seam. The final cycle occurs when the residual strapping material is cut and the hot shoe is removed. These steps reduce the tension to a final tautness, and the banding and heater head is initialized while waiting for the next "start" cycle. Banding material selection is defined by equipment capability. Most equipment systems were designed to crush the parcel during the peak tension phase of the process, so many of the systems utilized require heavier grades of banding than is called for in the specs.
Crush Cycles The size of the parcel for which the machine was designed and the size of the banding material contribute the greatest impact to the amount of "crush." Large parcel equipment capable of inducing 600 lbs. of tension or more during the crush cycle is not ideal for containing rigid plastic trays. Calibrating this type of equipment to run without passing into the crush stage results in a high level of inconsistency, incapable of developing the desired final results. Semiautomatic systems reduce the number of cycles by hand feeding the strap. These systems generally avoid the sharp, almost knee-jerk, and peak tension cycle. If not designed for small parcel handling, however, they are also limited in their ability to provide consistent final tension control. Bundling There are several manufacturers of strapping systems designed for small parcel bundling. These systems typically are extremely consistent and lend themselves to utilization in strapping applications involving bricks of relatively rigid, brittle plastic. Typical crush tensions in these application-specific systems were found to be in the neighborhood of between 2-3x the final tension. A year-long study demonstrated that some systems operated within the constraints defined with extremely repeatable performance. The natural variation was demonstrated at ± 3 lbs. from the nominal set point of 15 lbs.* Automated systems are available that can handle extremely pliable, very thin strapping, which makes for consistent performance. These systems may require the strap material to be annealed, so that it will feed through the looping tower, making this consumable item more expensive-but negligible in comparison to the value of the cargo. The material selection for the strapping equipment, as mentioned above, may be defined by the size and automated constraints of the equipment. In most cases, smaller is better. Preventing Breakage The ideal banding should conform to the brick at the corners to prevent breakage in these areas, while still achieving a snug loop around the bundle. Larger (wider and thicker) straps tend to create a "logging chain" effect, where the tendency is to crush the brick while still maintaining some slack in the center spans. In a worst-case scenario, the corners and side rails may be broken. In a more controlled scenario, however, the extreme tension on the edges of the trays tends to cause the brick to bow-in the top two and bottom two trays. This action changes the highly engineered pocket area that was so carefully designed for protecting the IC. The result may be catastrophic if that particular group of products is mishandled during its shipping life. Dry Bake A consideration of "dry bake" and the temperatures employed in this process must also be weighed against the strapping material and size selections. As the industry continues to migrate toward lower processing temperatures, the transition to thinner, more compliant band becomes plausible. Potential savings in tray material cost (and cycle time improvements through dry bake) may provide an extremely fast payback against any increases in cost associated with process improvement or capitalization in this area. Small parcel systems and compliant banding (thin and narrow) have been tested through the drop testing of product, followed by lead or ball scan inspection. The results indicate that an adequate tension for a conforming strap resides in the final tension of from 7-30 lbs.* Higher tensions create distortion and may result in some coplanarity degradation on leaded devices. Peak crush tension, when targeting 15 pound final tensions, have been recorded at 20-30 lbs. per square inches on automated equipment, with excellent consistency from brick to brick for loops applied both in the length and width directions of the brick. As in most areas of control, more is not better, as tensions exceeding these guidelines were found to distort the trays, which ultimately impacted the pocket capture integrity. Tray suppliers, who are under pressure to reduce costs, strive to create lightweight, tight tolerance designs utilizing engineering-grade plastics. Alternative material resins must continually be developed to achieve cost points driven by the market, with temperature capability and strength becoming the tradeoff. Those assembly sites embracing the challenge and choosing to control the strapping area will be rewarded with the ability to employ lower-cost tray materials, especially in newer package designs, where recycle programs are unable to provide sufficient trays. Targets are attainable for form, fit and function, but the necessity to survive excessive strapping strengths without bowing or breakage is close to the greatest challenge for tray suppliers. Dry Bake Impact Shrinkage studies across a vast assortment of strap banding materials seem to indicate that regardless of the pre-tensioning conditions, a post-bake band strap will relax or tighten to a final tension of between 10 and 30 lbs. (see table). Using this concept in reverse implies a loose strap at 2 to 3 lbs. prior to bake will shrink and conform to the stack with a final tension ideally around 10 lbs. The selection and development of the equipment, banding material and processes should be analyzed together as a system. Conclusion Yield losses associated with carrier trays and the processing surrounding the final steps in the pack and ship area are extremely difficult to isolate. While some cases are catastrophic in nature, it would not be difficult to imagine that a field failure rate of 0.1% or more might be realized through the joint contributions of strap, bake, pack and transport handling. Potential reductions in dunnage along with the realization of tray material savings should serve to quickly offset any investment in this area. * ITW Tech Center, Strapping, Dec. 5, 1998.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
