Publications

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

To date, all varistors used in ferroelectric neutron generators have been supplied from a single, proprietary source, General Electric Corporate Research and Development (GE CR&D). To protect against the vulnerability of a single source, Sandia initiated a program in the early 1980's to develop a second source for this material. A chemical preparation process for making homogeneous, high purity ZnO-based varistor powder was generated, scaled to production quantities, and transferred to external suppliers. In 1992, the chem-prep varistor program was suspended when it appeared there was sufficient inventory of GE CR&D material to supply ferroelectric neutron generator production for many years. In 1999, neutron generator production schedules increased substantially, resulting in a predicted exhaustion of the existing supply of varistor material within five years. The chem-prep program was restarted in January, 2000. The goals of the program were to (1) duplicate the chem-prep powder synthesis process that had been qualified for WR production, (2) demonstrate sintered billets from the chem-prep powder met requirements, (3) develop a process for rod fabrication and demonstrate that all component specifications could be met, and (4) optimize the process from powder synthesis through component fabrication for full-scale production. The first three of these goals have been met and are discussed in this report. A facility for the fabrication of production quantities of chem-prep powder has been established. All batches since the restart have met compositional requirements, but differences in sintering behavior between the original process and the restarted process were noted. Investigation into the equipment, precipitant stoichiometry, and powder processing procedures were not able to resolve the discrepancies. It was determined that the restarted process, which incorporated Na doping for electrical stability (a process that was not introduced until the end of the initial program and had not been investigated for processing effects), was responsible for the differences. Rod components fabricated since the restart have met requirements and have performed at a level comparable to chem-prep rods from the original program and GE CR&D rods currently in production.

Chemically prepared zinc oxide powders are fabricated for the production of high aspect ratio varistor components. Colloidal processing in water was performed to reduce agglomerates to primary particles, form a high solids loading slurry, and prevent dopant migration. The milled and dispersed powder exhibited a viscoelastic to elastic behavioral transition at a volume loading of 43-46%. The origin of this transition was studied using acoustic spectroscopy, zeta potential measurements and oscillatory rheology. The phenomenon occurs due to a volume fraction solids dependent reduction in the zeta potential of the solid phase. It is postulated to result from divalent ion binding within the polyelectrolyte dispersant chain, and was mitigated using a polyethylene glycol plasticizing additive. Chemically prepared zinc oxide powders were processed for the production of high aspect ratio varistor components. Near net shape casting methods including slip casting and agarose gelcasting were evaluated for effectiveness in achieving a uniform green microstructure achieving density values near the theoretical maximum during sintering. The structure of the green parts was examined by mercury porisimetry. Agarose gelcasting produced green parts with low solids loading values and did not achieve high fired density. Isopressing the agarose cast parts after drying raised the fired density to greater than 95%, but the parts exhibited catastrophic shorting during electrical testing. Slip casting produced high green density parts, which exhibited high fired density values. The electrical characteristics of slip cast parts are comparable with dry pressed powder compacts. Alternative methods for near net shape forming of ceramic dispersions were investigated for use with the chemically prepared ZnO material. Recommendations for further investigation to achieve a viable production process are presented.

The electrical properties of lead zirconate titanate ceramics near the 95/5 composition are extremely sensitive to the chemical composition and processing conditions. To precisely control the lead stoichiometry in a solid solution has been a challenge because of lead volatility during high temperature sintering. In this study, we investigated the effect of the amount of lead in the solid solution on crystal structure, dielectric behavior, and phase transformation characteristics for chemically prepared niobium modified PZT 95/5 ceramics. Implications are important for process control and assurance of material performance.

The hydrostatically induced ferroelectric(FE)-to-antiferroelectric(AFE) phase transformation for chemically prepared niobium modified PZT 95/5 ceramics was studied as a function of density and pore former type (Lucite or Avicel). Special attention was placed on the effect of different pore formers on the charge release behavior associated with the FE-to-AFE phase transformation. Within the same density range (7.26 g/cm3 to 7.44 g/cm3), results showed that ceramics prepared with Lucite pore former exhibit a higher bulk modulus and a sharper polarization release behavior than those prepared with Avicel. In addition, the average transformation pressure was 10.7% greater and the amount of polarization released was 2.1% higher for ceramics with Lucite pore former. The increased transformation pressure was attributed to the increase of bulk modulus associated with Lucite pore former. Data indicated that a minimum volumetric transformational strain of -0.42% was required to trigger the hydrostatically induced FE-to-AFE phase transformation. This work has important implications for increasing the high temperature charge output for neutron generator power supply units.

Niobium doped PZT 95/5 (lead zirconate-lead titanate) is the material used in voltage bars for all ferroelectric neutron generator power supplies. In June of 1999, the transfer and scale-up of the Sandia Process from Department 1846 to Department 14192 was initiated. The laboratory-scale process of 1.6 kg has been successfully scaled to a production batch quantity of 10 kg. This report documents efforts to characterize and optimize the production-scale process utilizing Design of Experiments methodology. Of the 34 factors identified in the powder preparation sub-process, 11 were initially selected for the screening design. Additional experiments and safety analysis subsequently reduced the screening design to six factors. Three of the six factors (Milling Time, Media Size, and Pyrolysis Air Flow) were identified as statistically significant for one or more responses and were further investigated through a full factorial interaction design. Analysis of the interaction design resulted in developing models for Powder Bulk Density, Powder Tap Density, and +20 Mesh Fraction. Subsequent batches validated the models. The initial baseline powder preparation conditions were modified, resulting in improved powder yield by significantly reducing the +20 mesh waste fraction. Response variation analysis indicated additional investigation of the powder preparation sub-process steps was necessary to identify and reduce the sources of variation to further optimize the process.

Chemically prepared zinc oxide powders are fabricated for the production of high aspect ratio varistor components. Colloidal processing was performed to reduce agglomerates to primary particles, form a high solids loadingslurry, and prevent dopant migration. The milled and dispersed powder exhibited a viscoelastic to elastic behavioral transition at a volume loading of 43-46%. The origin of this transition was studied using acoustic spectroscopy, zeta potential measurements, and oscillatory rheology. The phenomenon occurs due to a volume fraction solids dependent reduction in the zeta potential of the solid phase. It is postulated to result from divalent ion binding within the polyelectrolyte dispersant chain and was mitigated using a polyethylene glycol plasticizing additive. This allowed for increased solids loading in the slurry and a green body fabrication study to be presented in our companion paper.

The Materials Chemistry Department 1846 has developed a lab-scale chem-prep process for the synthesis of PNZT 95/5, referred to as the ''SP'' process (Sandia Process). This process (TSP) has been successfully transferred to and scaled-up by Department 14192 (Ceramics and Glass Department), producing the larger quantities of PZT powder required to meet the future supply needs of Sandia for neutron generator production. The particle size distributions of TSP powders routinely have been found to contain a large particle size fraction that was absent in development (SP) powders. This SAND report documents experimental studies focused on characterizing these particles and assessing their potential impact on material performance. To characterize these larger particles, fractionation of several TSP powders was performed. The ''large particle size fractions'' obtained were characterized by particle size analysis, SEM, and ICP analysis and incorporated into compacts and sintered. Large particles were found to be very similar in structure and composition as the bulk of the powder. Studies showed that the large-size fractions of the powders behave similarly to the non-fractionated powder with respect to the types of microstructural features once sintered. Powders were also compared that were prepared using different post-synthesis processing (i.e. differences in precipitate drying). Results showed that these powders contained different amounts and sizes of porous inclusions when sintered. How this affects the functional performance of the PZT 95/5 material is the subject of future investigations.

Abstract not provided.

The Materials Chemistry Department 1846 has developed a lab-scale chem-prep process for the synthesis of PNZT 95/5, a ferroelectric material that is used in neutron generator power supplies. This process (Sandia Process, or SP) has been successfully transferred to and scaled by Department 14192 (Ceramics and Glass Department), (Transferred Sandia Process, or TSP), to meet the future supply needs of Sandia for its neutron generator production responsibilities. In going from the development-size SP batch (1.6 kg/batch) to the production-scale TSP powder batch size (10 kg/batch), it was important that it be determined if the scaling process caused any ''performance-critical'' changes in the PNZT 95/5 being produced. One area where a difference was found was in the particle size distributions of the calcined PNZT powders. Documented in this SAND report are the results of an experimental study to determine the origin of the differences in the particle size distribution of the SP and TSP powders.