Among the issues to be addressed when designing ceramic composite components are anisotropy, brittle constituent behavior, statistical strength characteristics, unique time-dependent behaviors, and failure modes that are only becoming clear as the materials and test methods mature.
In a collaborative project with Virginia Tech, Cleveland State University, NASA Lewis Research Center, and Hibbitt, Karlsson & Sorensen, Inc., Babcock & Wilcox (B&W), through the CFCC program, is assembling a design methodology for fiber-wound tubular components.
This is one of the simpler CFCC forms that must address these unique design issues, and the tubular configuration is useful in power generation and other applications.
Fiber-wound tubes can be roughly approximated as two-dimensional laminated structures, somewhat simplifying the analytical modeling, but the design procedure is still complex.
The design methodology approach combines commercially available numerical design (finite element) tools with state-of-the-art university and federal laboratory technology to predict the life of laminated tubular ceramic composite components.
Design Approach
The B&W design method uses a macromechanical approach that accommodates statistical strength-behavior while capturing time-dependent damage throughout the model.
The macro approach provides economical design tools that are engineer friendly, and the time-dependent damage analysis, which can accommodate combined damage modes.
Ultimately these design tools will need to be extended to address more complex architectures, but the nature of the analytical development and failure mode characterization will be addressed in the laminated tubular models.
The analytical tools presented are new developments that are being applied to component analysis for the first time. Significant experimental work will be required to characterize the material behavior and verify the design tools.
The Methodology
The issues of scatter in strength, anisotropic material, and nontraditional, time-dependent damage behavior require a departure from the design philosophy prevalent in the analysis of metallic and even nonbrittle composite structural components.
Statistical design approaches must be used that adequately represent the damage behavior of the different modes of failure and the interaction among modes. In most cases, the methodology must track the accumulation of damage with time in a manner that results in life prediction of the component being designed.
Nontraditional Design
The development of such a nontraditional design philosophy is faced with many problems that must be overcome to encourage industrial acceptance of CFCC components.
To help ease this transition, the B&W team has chosen widely used, commercially available structural design tools for the fundamental predictions of states of stress with capabilities for modeling the time-dependent material behaviors that are important to the component life.
These available finite-element tools can then be interfaced with the statistical reliability-prediction tools and methods of capturing and tracking damage that are evolving in the university and national laboratory environment.
Design Methodology
The CFCC design methodology philosophy for fiber-wound cylindrical
composites is shown schematically in Fig. 1 .
This figure illustrates that the traditional finite-element tools maintain their identity and perform the static or time-dependent stress predictions that characterize the thermo-mechanically loaded structure.
The MRLife code from The Materials Response Group at Virginia Polytechnic Institute uses these stress results along with experimentally determined mechanistic damage representations to estimate changes in the material stiffness and durability.
The NASA Lewis Research Center code, C/CARES, predicts component reliability in the current damage state by utilizing a stochastic failure criterion, which also reflects the anisotropic nature of ceramic composites. By repeatedly performing this analysis as the deformation/damage state accumulates, component life is characterized through its time-dependent reliability.
Design Procedure
As an example of applying this methodology, a recent analysis of an advanced heat exchanger tube will be briefly reviewed and conceptually extended, where necessary, to illustrate application of a completed methodology. The approach still needs much work, so the actual analysis of the heat exchanger was not carried to a full time-dependent reliability prediction. Some of this limitation is a result of the need for further analytical development, but it is also a result of the need for further material development as well as needed characterization of material properties and performance. The B&W advanced heat exchanger, developed in a DOE-sponsored program, is a component of a waste heat recovery system that operates in an industrial furnace downstream of the flue exhaust. A schematic of the system and a picture of a CFCC tube are shown in Fig. 2 (graphic, 59k).
Waste Heat Recovery
The waste heat recovery system is an array of nine bayonet-type heat exchanger tubes where the bayonet consists of two concentric tubes. The outer tube consisted of a zirconia matrix and an alumina-zirconia fiber. Ceramic insulation surrounds the top of this outer tube and serves as an expansion joint between the tube and the plenum. An inner steel tube connects to an upper plenum which serves as the air inlet. The air enters the upper plenum, proceeds down the inner tube, exits the base of the inner tube, reverses direction and moves up between the steel and the ceramic composite tube.
As the clean air travels up the bayonet, it is heated from the flue gas passing along the outside of the ceramic composite tube. The heated air collects in the lower plenum located at the top of the ceramic section of the bayonet. At this point, the preheated air begins the return path back to the combustion section of the process. The analysis discussed here was performed as a part of the DOE Advanced Heat Exchanger Program. It is concerned with the outer CFCC tube and the life/reliability prediction of that component.
Stress Analysis
The ABAQUS finite-element program was used for the stress analysis of the heat exchanger tube. The laminated CFCC tube was modeled along with the surrounding insulation that provides the physical support for the tube and control of the tube temperature distribution.
The stresses during operation are controlled by the temperature distribution that changes from the inside to the outside of the tube and along the length of the tube. These temperature distributions were applied to the model and the resulting thermal stresses were calculated. The reliability of the tube was then predicted with the C/CARES code.
Strength Characterization
The Weibull data for statistical strength characterization in
this analysis were determined from first matrix damage indications
from a number of C-ring specimens exposed to the flue gas environment.
Full implementation of the design methodology will require a significant
amount of data to characterize fatigue, thermal exposure and other
time-dependent damage behavior in the principal directions along
and transverse to the fiber. If the data were available, the state
of damage could be calculated at each element in the model using
the MRLife Code, and the damaged material could be reanalyzed
for component reliability with the C/CARES Code. Through reiteration
of this process, reliability can be predicted as a function of
time as shown in Fig. 3 .
This figure shows a typical reliability plot for four different
thermal loads. To make design decisions based on such data, an
acceptance criterion must be established. For example, if the
acceptance criterion is for a reliability of 0.999 for 300 time
units, the lowest load level shown in this figure is an acceptable
condition.
Sophisticated Tools
As can be seen from the analysis, this design philosophy will require an increased sophistication of design tools that address all pertinent failure modes--many of which are not yet understood for these developing materials.
Different design codes/acceptance criteria, and new material testing standards for characterizing the unique materials will also be needed. As CFCC materials become more complex with fabrics and woven structures, the analytical tools and failure modes will become more complicated and require new development.
Industrial Acceptance
With advanced materials and a new design philosophy, much work is needed to hasten industrial readiness for commercial designs with ceramic composite components.
The most important issues are timing of the material development and manufacturing methods, design methodology development, and the hastening of a codes and standards infrastructure to stimulate industrial acceptance of CFCC designs.
The precompetitive work that is needed to bring design issues along with material and manufacturing development is the type of activity that the federal government wants to support to stimulate paths to commercialization. DOE has recognized this, and through the CFCC materials development program has supported the planning and initiation of such design activities.
Supporting Development
Babcock & Wilcox and Solar Turbines engineers are working together to discuss these issues with other industrial, university and government organizations. The goal is to encourage establishment of a design culture that can provide timely support for commercializing CFCC products.
Return to CFCC News No. 4, Table of Contents
Return to CFCC News Home Page
Comments to: mgc@ornl.gov
Revised: July 5, 1995
