28th International Lonterence

on Advanced Ceramics and

Composites: A

A collection of Papers Presented at the 28th International Conference and Exposition on Advanced Ceramics and Composites held in conjunction with the 8th International Symposium on Ceramics in Energy Storage and Power Conversion Systems

Edgar Lara-Curzio Michael J. Readey Editors

January 25-30,2004 Cocoa Beach, Florida

Published by The American Ceramic Society 735 Ceramic Place Westerville, OH 4308 I www.ceramics.org

02004TheAmerican Ceramic Society ISSN 0 196-62 I 9

28th International Lonterence

on Advanced Ceramics and

Composites: A

A collection of Papers Presented at the 28th International Conference and Exposition on Advanced Ceramics and Composites held in conjunction with the 8th International Symposium on Ceramics in Energy Storage and Power Conversion Systems

Edgar Lara-Curzio Michael J. Readey Editors

January 25-30,2004 Cocoa Beach, Florida

Published by The American Ceramic Society 735 Ceramic Place Westerville, OH 4308 I www.ceramics.org

02004TheAmerican Ceramic Society ISSN 0 196-62 I 9

Copyright 2004,The American Ceramic Society. All rights reserved.

Statements of fact and opinion are the responsibility of the authors alone and do not imply an opinion on the part of the officers, staff, or members ofThe American Ceramic Society. The American Ceramic Society assumes no responsibility for the statements and opinions advanced by the contributors to its publications or by the speakers at its programs. Registered names and trademarks, etc., used in this publication, even without specific indica- tion thereof, are not to be considered unprotected by the law. The American Ceramic society assumes no responsibility for errors or omissions in content, this responsibility lies solely with the author as all papers are prepared, formatted, and provided toThe American Ceramic Society as camera-ready copy.

No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or oth- erwise, without written permission from the publisher.

Authorization to photocopy for internal or personal use beyond the limits of Sections I07 and I08 of the US. Copyright Law is granted by the American Ceramic Society ISSN 0 196-62 19, provided that the appropriate fee is paid directly to the Copyright Clearance Centel; Inc., 222 Rosewood Drive, Danvers, MA 0 I923 USA, www.copyright.com. Prior to photocopying items for educational classroom use, please contact Copyright Clearance Center; Inc.

This consent does not extend to copying items for general distribution or for advertising or promotional purposes or to republishing items in whole or in part in any work in any format.

Please direct republication or special copying permission requests to the DirectocTechnical Publications,The American Ceramic Society, FIO. Box 6 I 36, Westerville, OH 43086-6 I 36, USA.

Series Editor: Greg Geiger Production Manager: John Wilson Director,Technical Publications: Mark Mecklenborg

Editorial and Circulation Offices PO Box 6136 Westerville, Ohio 43086-6 I36

Contact Information Editorial: (6 14) 794-5858 Customer Service: (6 14) 794-5890 Fax: (6 14) 794-5892 E-Mail: info@ceramics.org Website: www.ceramics.org/cesp

Ceramic Engineering & Science Proceedings (CESP) (ISSN 0 196-62 19) is published five times a year by The American Ceramic Society PO Box 6 I 36, Westerville, Ohio 43086-6 I36 www.ceramics.org. Periodicals postage paid at Westerville, Ohio, and additional mailing offices. The American Ceramic Society assumes no responsibility for the statements and opinions advanced by the con- tributon t o its publications. Papers for this issue were submitted as camera-ready by the authors. Any errors or omissions are the responsibility of the authors. Change of Address: Please send address changes to Ceramic Engineering and Science Proceedings, PO Box 6 136, Westerville, Ohio 43086-6 136, or by e-mail t o infc@ceramics.org. Subscription rates: One year $220 (ACerS member $176) in North America Add $40 for subscriptions out- side North America. In Canada, add GST (registration number RI 239946 18). Single Issues: Single issues may be purchased online at www.ceramics.org or by calling Customer Service at

Back Issues: When available, back issues may be purchased online at w.ceramics.org or by calling Customer Service at (6 14) 794-5890. Copies: For a fee, photocopies of papen are available through Customer Service. Authorization to photocopy items for internal or personal use beyond the limits of Sections 107 or I08 of the US. Copyright Law is granted by The American Ceramic Society ISSN 0 196-62 19, provided that the appropriate fee is paid directly t o Copyright Clearance Center, Inc., 222 Rosewood Dc, Danvers, MA 0 1923, USA (978) 750-8400: www.copyrightcom. Prior to photocopying items for educational classroom use, please contact Copyright Clearance Center, Inc. This consent does not extend to copying items for general distribution, or for advertising or promotional purpos- es, or to republishing items in whole or in part in any work in any format Please direct republication or special copying permission requests t o the Director,Technical Publications,The American Ceramic Society PO. Box 6 136, Westerville, Ohio 43086-6 136, USA. Indexinp: An index of each issue appears at www.ceramics.org/ctindex.asp, Contributors: Each issue contains a collection of technical papen in a general area of interestThese papers are of practical value for the Ceramic industries and the general public.The issues are based on the proceedings of a conference. Both The American Ceramic Society and non-Society conferences provide these technical papers. Each issue is organized by an editor, who selects and edits material from the conference proceedings.The opinions expressed are entirely those of the presenters.There is no other review prior to publication.Author guidelines are available on request. Postmaster: Please send address changes to Ceramic Engineering and Science Proceedings, PO. Box 6 136, Westerville, Ohio 43086-61 36. Form 3579 requested.

(6 14) 794-5890.

Copyright 0 2004,The American Ceramic Society, All rights reserved

Contents

28th International Conference on Advanced Ceramics and Composites:A

Preface .............................................................

Ceramics and Components in Energy Conversion Systems

Ceramic Components in Gas Turbine Engines: Why Has It Taken So long? .............................. D.W. Richerson

Development of the 8000 KW Class Hybrid Gas Turbine ....... T Sugimoto,Y Ichikawa, H. Nagata, K Igashira, S.Tsuruzono, andT, Fukudome

Development and Evaluation of CMC Vane

A. Kajiwara,T, Nakamura,T Araki, and H. Murata

Ceramic Combustor Design for ST5+ Microturbine Engine . ....

CMC Combustor l i n e r Design for a Model RAM Jet Engine .....

for NGSST Engine .................................

J, Shi,V.Vedula, E. Sun, D. Bombara, J. Holowczak,W.Tredway,A. Chen, and C. Fotache

T, Morimoto, S. Ogihara, H.Taguchi,T Kojima, K. Shimodaira, K. Okai, and H. Futamur

Burner Rig Test of Silicon Nitride Gas Turbine Nozzle .........

Materials for Advanced Battery and Energy Storage

A,]. Salkind

Effect of Ni-AI Precursor Type on Fabrication and

TN.Tiegs, F.C. Montgomery, and PA. Menchhofer

M. Ishizaki,T, Suetsuna, M.Asayama, M.Ando, N. Kondo, andT, Ohji

Systems (Batteries, Capacitors, Fuel Cells)

Properties of Tic-Ni,AI Composites .....................

................

V

MMCs by Activated Melt Infiltration High Melting

J. Kuebler, K. Lemster; Ph. Gasser; U.E. Klotz, andT Graule

Multifunctional Metal-Ceramic Composites by

R. Janssen, M. Leverkoehne, and I.]. Coronel,

Solid Freeform Fabrication of a Piezoelectric Ceramic Torsional Actuator Motor ...................... B.A. Bender; C. Kim, and CCm. Wu

Alloys and Oxide Ceramics ...........................

Solid Free Forming (SFF) ............................

Centrifugal Sintering ............................... Y Kinemuchi, K.Watari, and S. Uchimura

Alumina-Based Functionally Gradient Materials by Centrifugal Molding Technology ..................... C.-H. Chen,T. Nishikawa, S. Honda, and H.Awaji

Investigation of a Novel Air Brazing Composition for High

K.S.Weil, J.S. Hardy, and J. Darsell Temperature, Oxidation-Resistant Ceramic Joining ...........

Joining of Advanced Structural Materials

D. Singh, F. Guiterrez-Mora, N. Chen, K.C. Goretta, and J.L. Routbort

Mechanical Performance of Advanced Ceramic Parts Joined by Plastic Flow .......................... F. Guiterrez-Mora, D. Singh, N. Chen, and K.C. Goretta

by Plastic Deformation .............................

Physical Characterization of Transparent PUT Ceramics Prepared by Electrophoretic Deposition . .......... T. Nicolay and E. Bartscherer

Fabrication of Microstructured Ceramics by Electrophoretic Deposition of Optimized Suspensions ........ H. von Both, M. Dauscher; and J. HauOett

low Cost Process for Mullite Utilizing Industrial Wastes as Starting Raw Material ....................... K. Saiintawong, S. Wada, and A. Jaroenworaluck

vi

Low-Cost Processing of Fine Grained Transparent Yttrium Aluminum Garnet ........................... H. Lee,T-I. Mah, and TA, Parthasarathy

Gas-Pressure Sintering of Silicon Nitride with Lutetia Additive ............................... N. Kondo, M. Ishizaki, and T; Ohji

Use of Combustion Synthesis in Preparing Ceramic Matrix and Metal-Matrix Composite Powders .............. K.S. Weil and J.S. Hardy

Mechanical Reliability of Si,N, ........................ K. Sharma, l?S. Shankar; Jf Singh, and M.K. Ferber

Correlation of Finite Element with Experimental Results of the Small-Scale Vibration Response

S.R Short and S. Huo of a Damaged Ceramic Beam .........................

Macro-Micro Stress Analysis of Porous Ceramics

Y Ikeda,Y Nagano, H. Kawamoto, and N.Takano

X-Ray and Neutron Diffraction Studies on a

I.M. Low and Z. 00

by Homogenization Method ..........................

Functionally-Graded Ti,SiC,-Tic System ..................

Modeling of Transient Thermal Damage in

J.H. Underwood, M.E.Todaro, and G.N.Vigilante Ceramics for Cannon Bore Applications

Strengthening of Ceramics by Shot Peening

..................

............... W. Pfeiffer and T Frey

Solid Oxide Fuel Cells

DOE FE Distributed Generation Program ................. M.C. Williams

lanthanum Gallate Electrolyte for Intermediate Temperature ... S. Elangovan, B. Heck, S. Balagopal, D. Larsen, M.Timper, and J. Hartvigsen

vii

Solid Oxide Fuel Cell Development at

L. Blum, H.-P Buchkremer; L.G.J. de Haart, H. Nabielek J.W. Quadakkers, U. Reisgen, R Steinberger-Wilckens, R.W. Steinbrech, F.Tietz, I.Vinke

Forschungszentrum Juelich ...........................

Development of MOLB Type SOFC ...................... H. Miyarnoto, K. Mori,T. Mizoguchi, S. Kanehira, K.Takenobu, M. Nishiura, A. Nakanishi, M. Hattori, andY Sakaki

Development of Advanced Co-Fired Planar Solid

Z. Liu, G. Roman, J. Kidwell,T. Cable, R Goettlel; D. Larsen, J, Pike, and S. Elangovan

Electrophoresis: An Appropriate Manufacturing Technique for Intermediate Temperature Solid Oxide Fuel Cells ......... S. Kuehn and R Clasen

Oxide Fuel Cells with High Strength .....................

Microstructure-Performance Relationships in LSM-YSZ Cathodes ............................... J.A. Ruud,T Strikel;V. Midha, B.N. Ramamurthi, A.L. Linsebiglel; and D.J. Fogelman

Role of Cathode in Single Chamber SOFC ................. T Suzuki, PJasinski, F. Dogan, and H.U.Anderson

Morphology Control of SOFC Electrodes by

T Fukui, K. Murata, C.C. Huang, M. Naito, H.Abe, and K. Nogi

Improved SOFC Cathodes and Cathode Contact Layers ........ F-Tietz, H.-P Buchkremel;V.A.C. HaanappeLA. Mai, N.H. Menzler, J. Mertens,W.J. Quadakkers, D. Rutenbeck, S. Ulhenbruck, M. Zahid, and D. Stover

Mechano-Chemical Bonding Technique ...................

Characterization of Solid Oxide Fuel Cell Layers by Computed X-Ray Microtomography and Small-Angle Scattering ........................... A.J.Allen,TA. Dobbins, J. Ilavsky, F. Zhao,A.Virkar; J.Almer; and F. DeCarlo

Kinetics of the Hydrogen Reduction of NiONSZ and Associated M icrost ruct u ral Changes .................. M. Radovic, E. Lara-Curio, B. Armstrong, L. Walker; Piortorelli, and C. Walls

... Vl l l

Elastic Properties, Equibiaxial Strength and Fracture Toughness of 8mol%YSZ Electrolyte

M. Radovic, E. Lara-Curio, RTrejo, B.Armstrong, and C. Walls

Sintering of BaCeo~asYo~,sO,, WithMithout SrTiO, Dopant ..... F. Dynys, A. Sayir, and FJ. Heimann

Material for Solid Oxide Fuel Cells (SOFCs) . . ..............

High Temperature Seals for Solid Oxide Fuel Cells (SOFC) . . . . . . RN. Singh

Evaluation of Sodium Aluminosilicate Glass Composite Seal with Magnesia Filler .................... K.A. Nielsen, M. Solvang, F.W. Poulsen, and FH. Larsen

Durable Seal Materials for Planar Solid Oxide Fuel Cells ....... C.A. Lewinsohn, S. Elangovan, and S.M. Quist

Development of a Compliant Seal for Use in Planar Solid Oxide Fuel Cells ............................... ICS. Weil and J.S. Hardy

A Comparison of the Electrical Properties of YSZ Processed using Traditional, Fast-Fire, and

M. Ugorek D. Edwards, and H. Shulman

Enhancement of YSZ Electrolyte Thin Film Growth

Z. Xu and J. Sankar

Synthesis of Yttria Stabilized Zirconia Thin

Z. Xu, S.Tameru, and J. Sankar

Microwave Sintering Techniques .......................

Rate for Fuel Cell Applications ........................

Films by Electrolytic Deposition .......................

Sintering and Stability of the BaCe,.,_xZrxYo.,03, System ......

Conducting Complex Sr3(Ca I +xNb,J0,-8 Perovskite . .........

Z. Zhong, A. Sayir, and F. Dynys

Microstructure and Ordering Mode of a Protonic

M.-H. Berger and A. Sayir

ix

Nuclear Microprobe Using Elastic Recoil Detection (ERD) for Hydrogen Profiling in High Temperature Protonic Conductors ............................... f? Berger; A. Sayir, and M.-H. Berger

Ionic Conductivity in the Bi,O,- AI,O,-M,O, (MsCa, Y) System ........................................ Y-T Liu andT.4 Sheu

A Performance Based Multi-Process Cost Model forSOFCs ....................................... M. Koslowske, H. Benson, I. Bar-On, and R Kirchain

Development of a Tri-layer Electrochemical Model for a Solid Oxide Fuel Cell B. Ramamurthi,V. Midha, J. Ruud, and M.Thompson

Reduction and Re-oxidation of Anodes for Solid

1. Malzbender; E. Wessel, RW. Steinbrech, and L Singheiser

...........................

Oxide Fuel Cells (SOFC) .............................

Numerical Characterization of the Fracture Behavior of Solid Oxide Fuel Cell Materials by Means of

B.N. Nguyen, 6.1. Koeppel, f? Singh, M.A. Khaleel, and S. Ahzi Modified Boundary layer Modeling .....................

Chromium Poisoning of Cathodes by Ferritic Stainless Steel .... TD. Kaun,TA. Cruse, and M. Krumpelt

Effect of Impurities on Anode Performance ............... C.A.-H. Chung, K.V. Hansen, and M. Mogensen

X

Ceramics in Environment Applications

Comparison of Corrosion Resistance of Cordierite and Silicon Carbide Diesel Particulate Filters t o Combustion Products of Diesel Fuel Containing Fe and Ce Additives ....... D. O'Sullivan, S. Hampshire, M.J. Porneroy, and M.J. Murtagh

Overview of Ceramic Materials for Diesel Particulate Filter Applications ................................. W.A. Cutler

Soot Mass Limit Analysis of Sic DPF ..................... H. Sato, K. Ogyu, K.Yamayose,A. Kudo, and K. Ohno

A Mechanistic Model for Particle Deposition in Diesel Particulate Filters Using the l a t t i ce Boltzmann Technique ..... M. Stewart, D. Rector; G. Muntean, and G. Maupin

Development of Catalyzed Diesel Particulate Filter for the Control of Diesel Engine Emissions ................ Y Huang, Z. Dang, and A. Bar-llan

The Use of Transparent P U T Ceramics in a Biochemical Thin Fi lm lnterferometric Sensor .......................

Low Cost Synthesis of Alumina Reinforced Fe-Cr-Ni Alloys . ..... T. Sekhert, R. Janssen, and N. Claussen

T Nicolay

High Temperature Behavior of Ceramic Foams from SilSiC-Filled Preceramic Polymers .................. J. Zeschky,T Hoefner; H. Dannheim, M. Scheffler; P Greil, D. Loidl, S. Puchegger; and H. Peterlik

Stabilization of Counter Electrode for NASICON Based Potentiometric CO, Sensor ......................

Using Polyethylene Glycol-Mixed Sols ....................

Y. Miyachi, G. Sakai, K. Shirnanoe, and N.Yarnazoe

Microstructural Control of SnO, Thin Films by

G. Sakai, C. Sato, K. Shirnanoe, and N.Yamazoe

xi

Mixed-Potential Type Ceramic Sensors for No, Monitoring .....

Electrode Materials for Mixed Potential Nox Sensors D.L. West, F.C. Montgomery, and TR. Armstrong

Study of High Surface Area Alumina and Ga-Alumina

S.M. Zemskova, J.M. Faas, C.L. Boyer, PW. Park, 1. Wen, and I. Petrov

Development of Strong Photocatalytic Fiber and Environmental Purification ........................... H.Yamaoka,Y Harada,T Fujii, S. Otani, andT lshikawa

B.G. Nair, J. Nachlas, M. Middlemas, C.A. Lewinsohn, and S. Bhavaraju

.........

Materials for Denox Catalyst Applications .................

Processing of Biomorphous SIC Ceramics from Paper Preforms by Chemical Vapor Infiltraiton and

D.A. Streitwieser, N. Popovska, H. Gerhard, and G. Emig Reaction (CVI-R) Technique ...........................

Formation of Porous Structures by Directional Solidification of the Eutectic ......................... F.W. Dynys and A. Sayir

High Surface Area Carbon Substrates for

K.P Gadkaree,TTao, and W.A. Cutler Environmental Applications ..........................

Development of High Surface Area Monoliths for Sulfur Removal ................................ L. He, L.K. Owens, W.A. Cutler, and C.M. Sorenson

Processing of Porous Biomorphous TIC Ceramics by Chemical Vapor Infiltration and Reaction (CVI-R) Technique ....

Charge Transport Model in Gas-Solid Interface for Gas Sensors ................................... S.P Lee and Y-KYoon

N. Popovska, D.A. Streitwieser, C. Xu, and H. Gerhard

Corrosion Resistant Refractory Ceramics for Slagging Gasifier Environment ......................... E. Medvedovski and RE. Chinn

xii

Influence of the Dopants and the Metal Electrodes on the Electrical Response of Hematite Based Humidity Sensors ................................. 1.-M.Tulliani, f? Palmero, and P. Bonville

Light Weight Ceramic Sandwich Structure from Preceramic Polymers ............................... T. Hoefner, 1. Zeschky M. Schefflel; and I? Greil

Selective Catalytic Reduction and No, Storage in Vehicle Emission Control ........................... E.N. Cokel; S. Hammache, D.A. Peiia, and J.E. Miller

Ceramic Armor

Ballistic Impact of Silicon Carbide with Tungsten Carbide Spheres .................................. M.J. Normandia and B. Leavy

Toughness and Hardness of LPS-Sic and LPS-Sic

K.A. Schwetz,T. Kempf D. Saldsiedel; and R.Telle Based Composites .................................

Indentation Testing of Armor Ceramics .................. E. Medvedovski and f? Sakar

Metallic Bonding of Ceramic Armor Using Reactive Multilayer Foils ................................... A. Duckham, M. Brown, E. Besnoin, D. vanHeerden, O.M. Knio, and TPWeihs

Strain Rate Effects on Fragment Size of Britt le Materials .................................. F. Zhou, J.-F. Molinari, and KT. Ramesh

xiii

Preface

The 28th International Conference and Exposition on Advanced Ceramics and Composites was held in Cocoa Beach, Florida, January 25-30,2004. The conference was organized in conjunction with the 8th International Symposium on Ceramics in Energy Storage and Power Conversion Systems. More than 600 participants from 23 countries registered to attend the meeting, which had a record-breaking number of presentations with 447 oral and 93 poster presentations.

Dr. Jithendra P. Singh of Argonne National Laboratory presented the James 1. Mueller Memorial lecture, and received the James I. Mueller award, which is the most prestigious award granted by the Engineering Ceramics Division of The American Ceramic Society. The title of his presentation was Residual Stresses in Composites and Coatings. Professor Roger Naslain of the University of Bordeaux, France was the recipient of the 2004 Bridge Building Award, which recognizes researchers outside the United States who have made significant contributions to engineering ceramics. The title of his Bridge Building Award lecture was Sic-Matrix Composites: Non-Brittle Ceramics for Thermostructural Applications.

The success of this conference, which has become one of the premier international meetings on ceramic materials and composites, is the result of the quality of the papers presented by the participants and the dedication of session chairs and volunteers who champion the organization of symposia and focused sessions. The success of this con- ference is also in great part due to the contributions of the staff ofThe American Ceramic Society. In particular, we are indebted to Chris Schniuer, Greg Geiger and Marilyn Stoltz for their professionalism and dedication.

This proceedings contains I86 peer-reviewed manuscripts on the topics of solid oxide fuel cells; ceramics in environmental applications; mechanical properties; advanced ceramic coatings; biomaterials and biomedical applications; ceramic armor; nanomateri- als and biomimetics; and ceramics and components in energy conversion.

We hope you will find these manuscripts both interesting and useful. We look forward to seeing you in Cocoa Beach next year.

Edgar Lara-Curzio Michael J. Readey

xiv

28th international Conference

on Advanced Ceramics and

Composites: A

Ceramics and Components in Energy Conversion Systems

CERAMIC COMPONENTS IN GAS TURBINE ENGINES: WHY HAS IT TAKEN SO LONG?

David W. Richerson, Richerson and Associates and the University of Utah, Salt Lake City, Utah

ABSTRACT

Extensive efforts have been in progress worldwide since the late 1960s to develop ceramic turbine components, particularly since the introduction of high-strength silicon nitride and silicon carbide materials. This paper identifies the difficult challenges that have been encountered and highlights some of the key milestones in overcoming these challenges. Issues of improvements in material properties (especially long term durability at high temperature), propexty measurement and standards, component fabrication, quality assurance, design methodology for brittle materials, life prediction codes, and accumulation of rig and engine testing are all discussed.

INTRODUCTION

Remarkable progress has been made since the invention of the gas turbine engine, especially during the past 20 years, to increase thermal efficiency. Early engines had well under 30% efficiency. Modem utility scale gas turbines are now being commercially introduced with thermal efficiency greater than 60%. This has been accomplished through a combination of design and advanced materials to increase turbine inlet temperature, to increase pressure ratio, and in some cases to reclaim waste heat by use of heat exchangers. Ceramics have already played important roles as thermal barrier coatings [l] and as cores, molds, molten metal filters, and other refractories for investment casting of sophisticated superalloy blades and vanes with intricate internal cooling passages [2]. But the big hope has been to take advantage of the high temperature properties of ceramics to replace the complex cooled metal components with simple uncooled ceramic components.

Gas turbine engine designers explored the use of ceramic components in turbine engines intermittently since the middle of the 1940s [3]. The thermal and mechanical propehes of oxides, carbides, cermets, and other compositions were studied and some prototype components were tested in engines [4]. None of these ceramic-based materials survived because they did not have the right combination of properties to withstand the severe conditions inside a gas turbine engine. The oxides and other monolithic (non-composite) compositions could not tolerate the thermal shock conditions, while the cermets exhibited too high creep at the turbine engine operation temperature (which was much lower than modem turbine engines).

Indeed, the hot section of a gas turbine engine is a challenging environment for materials to survive. The materials must withstand high mechanical andor thermal stress at high temperature for thousands of hours, resist oxidation and corrosion by high velocity gases (often entrained with particles or condensed phases), avoid chemical reaction with adjacent components, be stable in high-cycle and low-cycle vibration conditions, and have acceptable creep and stress rupture life. They must also withstand handling during inspection and assembly, localized contact stress

To the extent authorized under the laws of the United Slates of America. all copyright interests in this publication are the propeny of The Amencan Ceramic Society Any duplication reproduction, or republication of this publication or any part thereof. wlthout the express written consent of The A ~ C I I C M Ceranuc Society or fee pad to the Copyright Clearance Center, I S prohibiled

3

concentrations at points of attachment or contact with adjacent components, and occasional events of impact.

During the late 1960s some new silicon nitride and silicon carbide materials were developed that appeared to have the right combination of properties to justify a renewed effort to implement ceramics in turbine engines [S-1 11. This led to considerable enthusiasm, optimism and hype [12,13] worldwide and initiation of a series of major well-funded programs during the 1970s and 1980s and continued effort to the present time [14,15]. But in spite of all the effort and enthusiasm, ceramic gas turbine engine components have not yet achieved broad commercial success. Why? Because the hot section of a gas turbine engine is such a severe environment and because we have been forced to address a wide spectrum of challenges:

generating a detailed property database for each new ceramic material modification to support design and guide fkther material improvement

0 improving material properties (especially long term durability at high temperature) 0 learning to design with brittle materials (establishing and experimentally verifying design

and life prediction methods and codes) 0 developing fabrication technology to reproducibly fabricate at reasonable cost the

advanced ceramics into the required complex shaped turbine engine components, 0 establishing quality assurance procedures (including dimensional measurements, non-

destructive evaluation techniques, and proof test protocols) 0 conducting iterative rig, engine and field test trials to identify problems such as impact,

contact stress, enhanced rate of oxidation in a moist high pressure turbine hot section, and hot corrosion

The remainder of this paper briefly discusses each of these challenges and the evolution of technology to progress in solving each challenge. A more detailed discussion is available in references [ 161 and [ 171, from which this was extracted.

Generating a Detailed Property Database

Very little property data were available for silicon nitride and silicon carbide candidate turbine materials in 1970, and there was little understanding of the relationships of chemistry, processing, and microstructure to the properties. Scanning electron microscopy, energy dispersive x-ray analysis, transmission electron microscopy, and Auger spectroscopy were just emerging as analysis tools available to industry, but had not yet been applied to the new silicon nitride and silicon carbide materials. Strength testing for ceramics was done primarily in 3-point bending, but there were no standards for sample preparation (such as surface grinding procedure and surface finish), sample size, test fixtures, or methods of data analysis or fracture surface analysis. Use of Weibull statistics had not yet been established or validated for brittle material design. The significance of fracture toughness was just beginning to be realized. In addition to all of these deficiencies and limitations in knowledge, we didn’t even have computers on our desks or user-friendly software.

All of these things needed to be developed and put in place concurrent to and often as a prerequisite to making progress in development of ceramics for gas turbine engines. As shown in Figure 1, materials characterization became a center point of virtually every program. Industry, government organizations, research institutes, and universities collaborated to develop

4

standardized tests for each key property, to design and build reliable test fixtures and equipment, and to gather and validate a wide range of property data [18-40]. Strength testing evolved from 3-point bending to 4-point bending to uniaxial tension (which was considered necessary by engine designers). Test methodologies were established and implemented for creep, stress rupture, fracture toughness, thermal conductivity, thermal shock resistance, and other properties.

Data to Guide Materials Improvement Iterations

Data to Guide Component

MATERIAL CHARACTERIZATION

and Elastic Properties

Figure 1 Development Requirements for Materials Characterization to Support Design and Life Prediction, Guide Materials Improvement, and Guide Component Failure Analysis [ 171

Silicon nitride and silicon carbide materials available in the early 1970s proved to not have adequate properties (especially at high temperature) to provide acceptable life and reliabifity. As data was generated (from characterization as well as component rig or engine testing) that identified deficiencies, companies that developed the materials worked diligently to develop improved versions. Every new version or modification required extensive characterization before the material could be considered for fabrication into turbine engine components. Each iteration of development, characterization, and testing took years. As a result, technology of ceramics for turbines progressed slowly from one plateau to another. One plateau would be reached that solved one challenge and allowed more extensive engine testing, but this testing then would identify a further limiting challenge that would require more material development to reach a new plateau. The subsequent section identifies some of these plateaus.

Improving Properties

By 1970 four types of ceramic materials had been identified as important candidates for turbine components:

5

1. Reaction bonded silicon nitride (RBSN) 2. Reaction sintered silicon carbide (RSSC) 3. Hot pressed silicon nitride (HPSN) 4. Lithium aluminum silicate (LAS)

As gas turbine engine companies began to evaluate these 1970-vintage materials, key deficiencies were identified that limited the stress and temperature at which they could be used and the time that they would remain functional. Initial RBSN materials had about 30% porosity and were thus low strength and vulnerable to oxidation degradation. Early HPSN materials had excellent room temperature strength, but the strength, creep resistance, and stress rupture life were inadequate above about 1000° C. Furthermore, HPSN could not be fabricated directly to the complex shapes required for turbines, necessitating extensive and costly diamond grinding. !970-vintage RSSC looked promising for low stress combustor components, but did not have adequate strength for other high stress components. LAS performed well as a honeycomb rotary regenerator in laboratory tests, but failed during on-road tests due to cracking caused by ion exchange of lithium ions with sodium ions (from road salt) and hydrogen ions (from sulfuric acid condensate from sulfur impurities in the fuel).

In spite of the deficiencies, the early 1970-vintage ceramic materials were good enough to demonstrate that ceramic turbine components could be designed to survive at least for a short time under gas turbine start, steady state, and shutdown conditions. However, it was clear that improvements in properties and reduction in fabrication cost would be required before ceramics would be suitable for significant application in gas turbine engines. Since 1970 extensive worldwide efforts have been conducted that have resulted in large improvements in the baseline cenunic materials and in development of new categories of silicon nitride and silicon carbide ceramics that are fabricated by net-shape fabrication processes such as pressureless sintering, overpressure sintering, and hot isostatic pressing (HE'). This progression is illustrated in Figure 2 and reviewed in subsequent paragraphs.

Figure 2. Progression of Ceramic Materials Development for Turbine Components [ 171

6

Reaction-bonded Silicon Nitride

RBSN is fabricated by forming a compact of silicon particles into the desired shape and reacting with nitrogen in a high temperature fiunace (about 1400' C) to convert to silicon nitride [6,41- 481. Because this requires a gas phase reactant, the resulting RBSN contains interconnected porosity. 1970-vintage RBSN typically had a density of about 2.2 g/cm3, which means that the material contained about 30% porosity. This porosity limited the strength and also allowed i n p s s of oxygen to cause material degradation at high temperature. The typical strength of the 2.2 g/cm3 RBSN in 3-point bending was under 100 MPa.

Through process development iterations during the 1970s, the density and strength of RBSN were progressively increased. By 1979 several RBSN materials were commercially available with a density of 2.7-2.8 g/cm' and an average strength in 4-point bending greater than 350 h4Pa [49-521. These materials, along with hot pressed silicon nitride and reaction-sintered silicon carbide, allowed engine companies to reach a first plateau of rig and engine testing. However, the properties were not adequate for reliability or long-term life.

Reaction-sintered Silicon Carbide

The second category of ceramics available in 1970 was RSSC, which was also referred to as siliconized silicon carbide. RSSC was fabricated by preparing a shaped perform h m a mixture of silicon carbide powder and a source of carbon such as graphite powder. This preform was then infiltrated with molten silicon in a vacuum above 1405" C. The molten silicon reacted with the carbon to form additional Sic and also filled all of the remaining pores. The resulting RSSC had near-zero porosity, strength comparable or better to RBSN, high thermal conductivity, and excellent oxidation resistance [53-581.

1970-vintage RSSC typically had flexure strength under 200 MPa. Improvements in density, uniformity, microstructure, and processing resulted in substantial increase in strength [59-641 during the early 1970s and 1980s as illustrated in Figure 3.

1991 NO*. NT-730 Ro8c 1, 1

Tunptrahut ('C)

Figure 3. Progress in Improvement of RSSC Materials [17]

7

Hot Pressed Silicon Nitride

A second category of silicon nitride that was just becoming available by 1970 was HPSN [4,65- 681. HPSN was fabricated by adding an oxide such as MgO to silicon nitride powder and applying pnssure through graphite punch and die tooling at about 1750" C. This resulted in a multiphase ceramk With near-zero porosity and very high strength (>700 h4Pa as measured in 3- point flexure) at mom temperature, as illustrated by the L u m HS-110 HPSN in Figure 4. In spite of the excellent morn temperahue strength, though, the strength dropped rapidly as the use temperature was increased above loOO-llOOo C.

Studies during the early 19708 determined that the high temperature strength and rupture life were controlled by the chemistry and crystallinity of the grain boundary phase (typically a complex silicate). A non-crystalline (glass) phase would soften above a critical temperature and lead to deformation and slow crack growth by grain boundary sliding. Extensive "grain boundary engineerin#' studies and process control studies were conducted worldwide that resulted in substantial improvements in the properties of HPSN materials [69-921. Figure 4 shows some examples of the level of improvements that were achieved.

k

I 0 200 400 600 800 lo00 1200 1400

Temperature ("C)

140

100 B m

m

0

Figure 4. Impmvements in the Strength of Hot Pressed Silicon Nitride Materials [ 171

8

Lithium Aluminum Silicate (LAS)

The fourth material available in the early 1970s was lithium aluminum silicate. LAS had near- zero thermal expansion and was studied extensively as a honeycomb rotary regenerator core, as a thick-walled flow separator housing, and later as a matrix for Sic fibers. The honeycomb regenerator was tested successfully under a laboratory environment, but proved not to be successful during on-road testing. Ion exchange occurred between the lithium and sodium (from road salt) and hydrogen (from sulfuric acid condensate from sulfur impurities in the fuel) and resulted in cracking. A technique was developed to leach out the lithium from the LAS to form an aluminosilicate that was more stable. Considerable development also was conducted with cordierite-based compositions similar to those later developed for catalytic converter honeycomb catalyst substrates.

Hot Isostatic Pressing (HIP)

Hot pressed silicon nitride was an important material during early turbine engine developments, especially for various rotor configurations. However, the high temperature properties, especially creep and stress rupture life, were still limiting. Also, complex shapes could not be achieved without extensive, expensive diamond grinding. Asea in Sweden began in about 197 1 to explore an alternative, hot isostatic pressing (HIP) of silicon nitride. By 1977 Larker and his co-workers at Asea reported direct fabrication of complex shapes of silicon nitride using a glass encapsulation process and achieving 3-point flexure strength of about 550 MPa at 1370’ C [93- 951. Other organizations also conducted development of HIP silicon nitride, and some companies ultimately licensed the Asea technology for complex shape fabrication [96-1121.

Figure 5 Compares the strength versus temperature of some of the HIP silicon nitride materials with 1975-vintage (NC-132) hot pressed silicon nitride. The CSN 101 was an Asea material, the NT154 and NT164 were Norton materials, and the GNlO was an Allied-Signal Ceramic Components material. Even more dramatic improvements are illustrated in Figure 6 for the static fatigue life based on tensile creep tests at 1370 OC and 100 MPa. The property improvements were made possible by the much higher pressure that could be applied during HIP than by uniaxial hot pressing. The higher pressure allowed full densification to be achieved with decreased sintering aid or alternate sintering aid (such as Ytterbium) and thus resulted in increased control of the composition and characteristics of the grain boundary phase.

Although HIP silicon nitride fabricated using glass encapsulation had excellent bulk properties, chemical interaction during the HIP cycle between the silicon nitride and the glass resulted in lesser properties at the surface. The overall probability of survival for candidate turbine components was decreased unless the surface material could be removed by machining. In most cases this was considered not to be cost effective, so the silicon nitride HIPed with glass encapsulation lost favor during the mid 199Os, especially for long life applications. However, “sinter-HIP” was still used whereby the ceramic was first densified by sintering to the point that the porosity was all internal (closed porosity) rather than connected to the surface. The component could then be HIPed without glass encapsulation to remove much of the closed porosity and achieve significant increase in strength. Even in this case, though, the HIP step represented an undesirable added cost.

9

0 1977 Aesa HIP Si,N,, 3-pt flexure I 1 9 3 N O W NCX-5102 HIP Sl& tsnsiOn i

140

20

o ~ ' " " ' ~ ' ' ' ~ " ' " ' ' I ' ' ' l ' " ' B ' 0 0 200 400 600 800 loo0 1200 1400 1600

Temperature ("C)

Figure 5 . Evolution of Hot Isostatically Ressed Silicon Nitride [ 171

Sintered Silicon Nitride

The ideal low cost approach for fabrication of ceramics is conventional pressureless sintering. Pressureless sintering was demonstrated for silicon nitride by the mid-19708, but the propedes were not much better than those of RBSN or RSSC. For example, the fvst reported sintered silicon nitride by Terwilliger [ 1301 in 1974 was less than 77% of theoretical density. This material, which contained MgO as the sintering aid, exhibited high weight loss due to decomposition of silicon nitride and volatilization of MgO.

Many developments conducted over many years would be required to achieve properties in sintered silicon nitride that would approach those of HIP silicon nitride. Some of the key directions of development included (1) control of silicon nitride decomposition, (2) development of processes for synthesis of high quality, fine particle, reactive silicon nitride powders, (3) refinement of sintering aids and the resulting grain boundary chemistry, (4) increasing fracture toughness by achieving elongated, fibrous microstructure, (5) sintered RBSN. International efforts resulted in progress in each of these areas [ 114-129,521. Examples illustrating the evolution of improvement in room temperature and high temperature strength are shown in Figure 6. The strength improvements were accompanied by large improvements in static fatigue life (Figure 7) and fracture toughness.

10

NQK InruLron SN-8% 1986

5000-

4000

n

3000 r, Q) !e

P E 2000 0

'Oo0

0 200 400 600 800 loo0 1200 1400

Temperature ("C)

. . . . . . . . . . . . . . . . . . . . . . . Tensile Creep at

1370°C and 100 MPa -

-

1

NC-132 . 0.5 h

3 o--m. I " ' I ' ' ' ' '

Figure 6. Improvements in Strength of Sintered and Over-pressure Sintered silicon nitride [ 171

11

Learning to sinter in a moderate overpressure atmosphere of nitrogen (about 1-4 atmospheres) to minimize silicon nitride decomposition during sintering was one especially key advance. This allowed sintering at higher temperature to achieve close to theoretical density (typically less than 2% porosity), but also resulted in enhanced growth of elongated beta silicon nitride grains during sintering, as illustrated for AS 800 silicon nitride in Figure 8. These grains intertwined in such a way that fiacture toughness was substantially increased [119-1211. 1975-vintage hot pressed silicon nitride had fiacture toughness of about 4.5-5.0 MPa.mln. Some of the overpressure sintered silicon nitride materials by 1990 had fracture toughness greater than 8.0 MPa.mln. This unusually high toughness for a ceramic has been an important factor in the reliability and successful use of silicon nitride for turbine components and for other applications.

Figure 8. Microstructure of Over-Pressure Sintered AS 800 Silicon Nitride Showing the Elongated Grains That Result in High Fracture Toughness (Photo courtesy of Honeywell Engines, Systems, and Services)

Two other important technologies evolved. One was referred to as “sintered RBSN” and the other was “sinter-HIP”. For sintered RBSN, sintering aids were mixed with the initial silicon powder, the silicon was nitrided to achieve partial densification, and then the RBSN containing the sintering aid was overpressure sintered. This allowed fabrication of dense silicon nitride parts using a less expensive starting powder (silicon versus silicon nitride) and also reduced the total shrinkage during sintering. For sinter-HIP, overpressure sintering would be conducted with silicon nitride powder or sintered RBSN to achieve a high enough density that only closed porosity remained. Then the part could be placed in a HIP unit without requiring glass encapsulation to reduce the porosity to close to zero and thus achieve further improvement in strength.

Sintered Silicon Carbide

Sic evolved similarly to silicon nitride. RSSC and recrystallized S ic were the primary S ic materials available up to the mid 1970s, followed sequentially by hot ptesstd SIC 1130-1321,

12

pressureless sintered Sic [133-1451, and HIP and sinter-HIP Sic. Hot pressed Sic became commercially available in the early 1970s, but never really caught on. In tests that provided a direct comparison, it did not have comparable thermal shock resistance to hot pressed silicon nitride. By the mid to late 1970s sintered Sic, especially from Carborundum Company, became an important candidate for turbine components and was evaluated extensively throughout the 1980s. 1977-vintage Hexoloy SA (sintered alpha-Sic from Carborundum) had 4-point flexure strength of 300 MPa at room temperature and about 350 MPa at 1450’ C. By 1980 several sintered Sic materials were available with 4-point flexure strength between 400 and 500 MPa.

Ceramic Matrix Composites (CMCs)

An alternate approach to monolithic Sic was pursued starting in the late 1970s when high- strength Sic-based fibers became available. SEP in France developed a composite consisting of Sic-based fibers (Nicalon fibers from Japan) in a chemical vapor infiltration (CVI) matrix [ 1461. This general category of composites became known as CVI SiC/SiC. Later General Electric Company developed a melt-infiltrated reaction sintered composite material (MI-SiCtSiC) and Dow Coming developed a polymer infiltrated and pyrolyzed (PIP-SiC/SiC) composite. Extensive development has been conducted on these ceramic matrix composite materials particularly since 1990 [ 147-1561.

The advantage of the continuous fiber reinforced CMC materials has been their fracture behavior. Rather than failing in a brittle mode like the monolithic ceramics, these CMC materials are able to tolerate much larger material defects and to sustain substantial accumulation of structural damage prior to hcture. The result is increased strain-to-failure and a non- catastrophic failure mode. However, studies during the 1980s and 1990s determined that the SiC/SiC ceramic matrix composites were susceptible to degradation due to interaction with the application environment (especially oxygen and water vapor) at high temperature [157]. Extensive developments on interface layers, fiber improvements, matrix modification, and surface coatings have been conducted during the 1990s and are continuing. Parallel efforts have focused on oxiddoxide ceramic matrix composites [ 1581, although current versions have lower strength and lower use temperature than the SiC/SiC composites.

Learning to Design with Brittle Materials

Design methodology for brittle materials was at a very early stage prior to about 1970. Ford Motor Company initiated a program in 1967 to design an automotive gas turbine engine with an all-ceramic hot section (their 820 ceramic engine design). At that time computers were still cumbersome and there were no established design tools and guidelines for the complex 3D analyses that would be required for brittle materials. Ford and other early pioneers adopted by necessity a “learn-by-doing” approach that required substantial data gathering, many iterations, and experimental verification at each step of the way [159]. This involved a systems approach that became the model for future programs. It consisted of time-consuming iterations of the following sequence of activities:

0 Define a platform for development, e.g. a specific engine or a generic engine concept. Select one or more conceptual component designs.

0 Gather a material property database suitable to support analytical design requirements.

13