The ease and ability to predict sintering shrinkage and densification with the Skorohod-Olevsky viscous sintering (SOVS) model within a finite-element (FE) code have been improved with the use of an Arrhenius-type viscosity function. The need for a better viscosity function was identified by evaluating SOVS model predictions made using a previously published polynomial viscosity function. Predictions made using the original, polynomial viscosity function do not accurately reflect experimentally observed sintering behavior. To more easily and better predict sintering behavior using FE simulations, a thermally activated viscosity function based on creep theory was used with the SOVS model. In comparison with the polynomial viscosity function, SOVS model predictions made using the Arrhenius-type viscosity function are more representative of experimentally observed viscosity and sintering behavior. Additionally, the effects of changes in heating rate on densification can easily be predicted with the Arrhenius-type viscosity function. Another attribute of the Arrhenius-type viscosity function is that it provides the potential to link different sintering models. For example, the apparent activation energy, Q, for densification used in the construction of the master sintering curve for a low-temperature cofire ceramic dielectric has been used as the apparent activation energy for material flow in the Arrhenius-type viscosity function to predict heating rate-dependent sintering behavior using the SOVS model.
Low temperature co-fire ceramic (LTCC) materials technology offers a cost-effective and versatile approach to design and manufacture high performance and reliable advanced microelectronic packages (e.g., for wireless communications). A critical issue in manufacturing LTCC microelectronics is the need to precisely and reproducibly control shrinkage on sintering. Master Sintering Curve (MSC) theory has been evaluated and successfully applied as a tool to predict and control LTCC sintering. Dilatometer sintering experiments were designed and completed to characterize the anisotropic sintering behavior of green LTCC materials formed by tape casting. The resultant master sintering curve generated from these data provides a means to predict density as a function of sintering time and temperature. The application of MSC theory to DuPont 951 Green Tape{trademark} will be demonstrated.
Software has been developed and extended to allow finite element (FE) modeling of ceramic powder compaction using a cap-plasticity constitutive model. The underlying, general-purpose FE software can be used to model even the most complex three-dimensional (3D) geometries envisioned. Additionally, specialized software has been developed within this framework to address a general subclass of axisymmetric compacts that are common in industry. The expertise required to build the input deck, run the FE code, and post-process the results for this subclass of compacts is embedded within the specialized software. The user simply responds to a series of prompts, evaluates the quality of the FE mesh that is generated, and analyzes the graphical results that are produced. The specialized software allows users with little or no FE expertise to benefit from the tremendous power and insight that FE analysis can bring to the design cycle. The more general underlying software provides complete flexibility to model more complicated geometries and processes of interest to ceramic component manufacturers but requires significantly more user interaction and expertise.
In the manufacture of ceramic components, near-net-shape parts are commonly formed by uniaxially pressing granulated powders in rigid dies. Density gradients that are introduced into a powder compact during press-forming often increase the cost of manufacturing, and can degrade the performance and reliability of the finished part. Finite element method (FEM) modeling can be used to predict powder compaction response, and can provide insight into the causes of density gradients in green powder compacts; however, accurate numerical simulations require accurate material properties and realistic constitutive laws. To support an effort to implement an advanced cap plasticity model within the finite element framework to realistically simulate powder compaction, the authors have undertaken a project to directly measure as many of the requisite powder properties for modeling as possible. A soil mechanics approach has been refined and used to measure the pressure dependent properties of ceramic powders up to 68.9 MPa (10,000 psi). Due to the large strains associated with compacting low bulk density ceramic powders, a two-stage process was developed to accurately determine the pressure-density relationship of a ceramic powder in hydrostatic compression, and the properties of that same powder compact under deviatoric loading at the same specific pressures. Using this approach, the seven parameters that are required for application of a modified Drucker-Prager cap plasticity model were determined directly. The details of the experimental techniques used to obtain the modeling parameters and the results for two different granulated alumina powders are presented.
Ceramics represent a unique class of materials that are distinguished from common metals and plastics by their: (1) high hardness, stiffness, and good wear properties (i.e., abrasion resistance); (2) ability to withstand high temperatures (i.e., refractoriness); (3) chemical durability; and (4) electrical properties that allow them to be electrical insulators, semiconductors, or ionic conductors. Ceramics can be broken down into two general categories, traditional and advanced ceramics. Traditional ceramics include common household products such as clay pots, tiles, pipe, and bricks, porcelain china, sinks, and electrical insulators, and thermally insulating refractory bricks for ovens and fireplaces. Advanced ceramics, also referred to as ''high-tech'' ceramics, include products such as spark plug bodies, piston rings, catalyst supports, and water pump seals for automobiles, thermally insulating tiles for the space shuttle, sodium vapor lamp tubes in streetlights, and the capacitors, resistors, transducers, and varistors in the solid-state electronics we use daily. The major differences between traditional and advanced ceramics are in the processing tolerances and cost. Traditional ceramics are manufactured with inexpensive raw materials, are relatively tolerant of minor process deviations, and are relatively inexpensive. Advanced ceramics are typically made with more refined raw materials and processing to optimize a given property or combination of properties (e.g., mechanical, electrical, dielectric, optical, thermal, physical, and/or magnetic) for a given application. Advanced ceramics generally have improved performance and reliability over traditional ceramics, but are typically more expensive. Additionally, advanced ceramics are typically more sensitive to the chemical and physical defects present in the starting raw materials, or those that are introduced during manufacturing.
Ceramic-metal composites can be made by reactive penetration of molten metals into dense ceramic performs. The metal penetration is driven by a large negative Gibbs energy for reaction, which is different from the more common physical infiltration of porous media. Reactions involving Al can be written generally as (x+2)Al + (3/y)MO{sub y} {yields} Al{sub 2}O{sub 3} + M{sub 3/y}Al{sub x}, where MO{sub y} is an oxide that is wet by molten Al. In low Po{sub 2} atmospheres and at temperature above about 900{degrees}c, molten Al reduces mullite to produce Al{sub 2}O{sub 3} + M{sub 3/y}Al{sub x}, where MO is an oxide that is wet by molten Al. In low Po{sub 2} atmospheres and at temperatures above about 900{degrees}C, molten al reduces mullite to produce Al{sub 2}O{sub 3} and Si. The Al/mullite reaction has a {Delta}G{sub r}{degrees} (1200K) of -1014 kJ/mol and, if the mullite is fully dense, the theoretical volume change on reaction is less than 1%. A microstructure of mutually-interpenetrating metal and ceramic phases generally is obtained. Penetration rate increases with increasing reaction temperature from 900 to 1150{degrees}C, and the reaction layer thickness increases linearly with time. Reaction rate is a maximum at 1150{degrees}C; above that temperature the reaction slows and stops after a relatively short period of linear growth. At 1300{degrees}C and above, no reaction layer is detected by optical microscopy. Observations of the reaction front by TEM show only al and Al{sub 2}O{sub 3} after reaction at 900{degrees}C, but Si is present in increasing amounts as the reaction temperature increases to 1100{degrees}C and above. The kinetic and microstructural data suggest that the deviation from linear growth kinetics at higher reaction temperatures and longer times is due to Si build-up and saturation at the reaction front. The activation energy for short reaction times at 900 to 1150{degrees}C varies from {approximately}90 to {approximately}200 kJ/mole.
New characterization and computational techniques have been developed to evaluate and simulate binder burnout from pressed powder compacts. Using engineering data and a control volume finite element method (CVFEM) thermal model, a nominally one dimensional (1-D) furnace has been designed to test, refine, and validate computer models that stimulate binder burnout assuming a 1-D thermal gradient across the ceramic body during heating. Experimentally, 1-D radial heat flow was achieved using a rod-shaped heater that directly heats the inside surface of a stack of ceramic annuli surrounded by thermal insulation. The computational modeling effort focused on producing a macroscopic model for binder burnout based on continuum approaches to heat and mass conservation for porous media. Two increasingly complex models have been developed that predict the temperature and mass of a porous powder compact as a function of time during binder burnout. The more complex model also predicts the pressure within a powder compact during binder burnout. Model predictions are in reasonably good agreement with experimental data on binder burnout from a 57-65% relative density pressed powder compact of a 94 wt% alumina body containing approx. 3 wt% binder. In conjunction with the detailed experimental data from the prototype binder burnout furnace, the models have also proven useful for conducting parametric studies to elucidate critical material property data required to support model development.
The properties and performance of a ceramic component is determined by a combination of the materials from which it was fabricated and how it was processed. Most ceramic components are manufactured by dry pressing a powder/binder system in which the organic binder provides formability and green compact strength. A key step in this manufacturing process is the removal of the binder from the powder compact after pressing. The organic binder is typically removed by a thermal decomposition process in which heating rate, temperature, and time are the key process parameters. Empirical approaches are generally used to design the burnout time-temperature cycle, often resulting in excessive processing times and energy usage, and higher overall manufacturing costs. Ideally, binder burnout should be completed as quickly as possible without damaging the compact, while using a minimum of energy. Process and computational modeling offer one means to achieve this end. The objective of this study is to develop an experimentally validated computer model that can be used to better understand, control, and optimize binder burnout from green ceramic compacts.
An investigation was made into the effect of microstructure on the peak toughness and shape of the crack growth resistance curves for two ceramic-metal composites. An Al{sup 2}O{sup 3}/Al composite formed by Reactive Metal Penetration was used along with an AlN/Al composite formed using a reactive infiltration technique. The results indicate that the toughness increases with an increase in the volume fraction of the metal phase for a particular composite composition, and the peak toughness and shape of the R-Curve also depend on the composite microstructure and metal composition.