A series of ten slicing demonstrations on a multi-wire slurry saw, manufactured by Yasunaga Engineering Company of Japan and distributed by GEOS Corporation of Stamford, Connecticut, was made to evaluate the silicon ingot wafering capabilities. The results revealed that the present sawing capabilities can provide usable wafer area from an ingot 1.05 m/sup 2//kg (e.g., kerf width 0.135 mm and wafer thickness 0.265 mm). Satisfactory surface qualities and excellent yield of silicon wafers were found. One drawback is that the add-on cost of producing wafer from this saw, as presently used, is considerably higher than the systems being developed by Varian and Crystal Systems for the Low-Cost Silicon Solar Array Protect (LSSA), Task II, primarily because the Yasunaga saw uses a large quantity of wire. The add-on cost can be significantly reduced by extending the wire life and/or by reuse of properly plated wire to restore the diameter.
As part of the Jet Propulsion Laboratory's Low-Cost Solar Array Project, test methods have been evaluated and procedures developed for testing photovoltaic flat-plate solar cell modules for resistance to impact by hailstones. Testing has included the use of simulated hailstones (frozen ice spheres projected at terminal velocity), steel balls, and other projectile types applied with three loading methods: pneumatic gun, gravity drop, and static loading. Results are presented that compare the advantages and disadvantages of the three test methods. Dropped-steel-ball tests are shown to exhibit little correlation with high-velocity ice-ball tests, whereas statically-loaded steel balls show a somewhat better correlation with ice-ball tests. Results are also presented on the hail impact strength of 16 flat-plate photovoltaic modules. The module designs tested have been shown to be capable of withstanding as large as 1-1/2-inch diameter and not capable of withstanding as small as 1/2-inch diameter simulated hail. The top surface material of the modules has a dominant influence on the hail impact resistance of the modules. In order of increasing impact strength for a given thickness, the top surface materials encountered in the modules tester were: clear silicone rubber, annealed glass, tempered glass, and acrylic sheet. The critical failure mechanism of each module type is explored and means for improving the hail resistance of future modules are described.
A preliminary version of a workbook for SAMICS, a uniform costing methodology for manufacturing processes, was published. Continued experimentation with the fluidized bed reactor for production of Si by Zn reduction of SiCl/sub 4/ showed that 78 percent of theoretical conversion efficiency from SiCl/sub 4/ to Si was attainable using a hemispherical bed support, versus 68 percent when using the concise bed support. The EFG process is now producing 2-inch wide ribbon at 2 to 3 inches per minute; however, conversion efficiencies remain low at 2 percent to 3 percent. Work on the CAST process has stopped pending contract negotiation. Materials other than SiO/sub 2/ are being evaluated for the die used in the inverted Stepanov process. Development of low-cost encapsulation materials and systems is continuing. Production process emphasis was shifted to the cost-effectiveness of processes still being considered, and to the sensitivity of these processes to commercially realistic tolerances. Engineering work centered in two primary areas: the finalizing of a substantially updated module design specification, and expanded activity in the development of environmental design requirements. (MHR)