Mechanical Properties and Performance of Engineering Ceramics and Composites IX, Volume 35, Issue 2 E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Serie: Ceramic Engineering and Science Proceedings
• Sprache: Englisch

Seitenzahl: 355

Veröffentlichungsjahr: 2014

Beschreibung

The Ceramic Engineering and Science Proceeding has been published by The American Ceramic Society since 1980. This series contains a collection of papers dealing with issues in both traditional ceramics (i.e., glass, whitewares, refractories, and porcelain enamel) and advanced ceramics. Topics covered in the area of advanced ceramic include bioceramics, nanomaterials, composites, solid oxide fuel cells, mechanical properties and structural design, advanced ceramic coatings, ceramic armor, porous ceramics, and more.

Leseprobe

Contents

Cover

Half Title page

Title page

Copyright page

Preface

Introduction

Creep, Fatigue, and Damage Characterization

Anisotropic Creep Behavior of a Unidirectional All-Oxide CMC

Introduction

Experimental Results

Anisotropic Creep Description with the Approach According to Hill

Numerical Simulation of Unit Cells

Conclusion

Acknowledgement

References

Indicators for the Damage Evolution at Intermediate Temperature under Air of a Sic/[Si-B-C] Composite Subjected to Cyclic and Static Loading

Introduction

Experimental

Acoustic Emission

Results and Discussion

Conclusion

Acknowledgements

References

Durability Results from Ceramic Matrix Composite with Differing Porosity Levels

Introduction

Procedure

Results

Discussion

Conclusion

Acknowledgments

References

Effects of Stress Concentrators on Damage Evolution in SiC/SiC Composites

Introduction

Experimental and Methods

Results and Discussion

Conclusion

References

Advancements in Acoustic Micro Imaging for the Non-Destructive Inspection of Ceramic Components and Devices

Introduction

Experimental

Case Study 1

Results and Discussion

Case Study 2

Results and Discussion

Conclusion

References

Effect of Specimen Geometry on Microstructural Fracture Behavior in Nano Composites under HVEM

Introduction

Model for Analysis

Interface Element

Results and Discussions

Conclusions

References

Processing and Properties of Carbides

Effects on Mechanical and Thermal Properties by Varying the Interconnectivity of SiC in a Si:SiC Composite System

Introduction

Experimental Procedure

Results & Discussion

Summary

References

Microstructure-Property Relationships in SiC/Diamond Composites as a Function of Diamond Content

Introduction

Experimental Procedure

Results & Discussion

Summary

References

Effect of SiC:B4C Ratio on the Properties of Si-Cu/SiC/B4C Composites

Introduction

Experimental Procedure

Results and Discussion

Summary

Acknowledgement

References

Plastic Deformation and Cracking Resistance of SiC Ceramics Measured by Indentation

Introduction

Experimental Procedure

Experimental Results

Discussion

Conclusion

References

Fabrication of SiC Fiber-Reinforced SiC Matrix Composites by Low Temperature Melt Infiltration Method Using Si-Hf and Si-Y Alloy

Introduction

Experimental Procedures

Results & Discussion

Conclusion

References

Processing and Properties of Non-Carbides

Development of Electrical Porcelain Insulators from Ceramic Minerals in Uganda

Introduction

Raw Materials, Sample Preparation and Experimental Techniques

Results

Discussions

Conclusion

Acknowledgements

References

The Mechanical Properties of Sandwich Structures Based on a Metal Ceramic Core and Fiber Metal Laminate Skin Material

1. Introduction

2. Experimental Procedure

3. Results and Discussion

4. Conclusions

References

Alkali Treatment on Sugarcane Bagasse to Improve Properties of Green Composites of Sugarcane Bagasse Fibers-Polypropylene

Introduction

Experimental Procedures

Results and Discussion

Conclusion

Acknowledgement

References

Characteristics of a Zirconia-Spinel Composite Processed by a Current-Activated Pressure-Assisted Densification Method

Introduction

Experimental Material and Procedures

Experimental Results

Discussion

Conclusions

Acknowledgement

References

Oxidation and Healing

Enhancement of Oxidation Resistance of Graphite Foams by SiC Coating for Concentrated Solar Power Applications

1. Introduction

2. Experiments

3. Results and Discussion

Conclusions

Acknowledgements

References

Spark Plasma Sintering of Ceramic Matrix Composites with Self-Healing Matrix

Introduction

1. Experimental Method

2. Results and Discussion

Conclusion

Acknowledgements

References

Advanced Ceramic Composite using Self-Healing and Fiber-Reinforcement

Introduction

Materials Design

Materials Preparation and Microstructure

Mechanical Properties

Next Trial for Shfrc

Conclusion

Acknowledgements

References

Delamination, Chipping, and Wear

Applying Fracture Mechanics Methods to Model Coating Delamination

Introduction

Methodology

Results and Discussion

Conclusions

References

A New Analysis of the Edge Chipping Resistance of Brittle Materials

Introduction

A New Approach

Materials and Methods

Results

Discussion

Conclusions

Acknowledgements

References

Tribological Background for the Use of Niobium Carbide (NbC) As Cutting Tools and For Wear Resistant Tribosystems

Introduction

Tribological Results

Conclusions

Acknowledgements

References

Index

Mechanical Properties and Performance of Engineering Ceramics and Composites IX

Copyright © 2015 by The American Ceramic Society. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone], or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data is available.

ISBN: 978-1-119-03118-5ISSN: 0196-6219

Preface

This volume is a compilation of papers presented in the Mechanical Behavior and Performance of Ceramics & Composites symposium during the 38th International Conference & Exposition on Advanced Ceramics and Composites (ICACC) held January 26-31, 2014 in Daytona Beach, Florida.

This long-standing symposium received presentations on a wide variety of topics thus providing the opportunity for researchers in different areas of related fields to interact. This volume emphasizes some practical aspects of real-world engineering applications of materials such as oxidation, fatigue, wear, nondestructive evaluation, and mechanical behavior as associated with systems ranging from niobium carbide to metallic-ceramic sandwich structures to ceramic matrix composites. Symposium topics included:

Fabrication, Microstructure, and Properties

Creep and Fatigue

Oxidation and Self-healing

Delamination, Chipping, and Wear

NDE

Significant time and effort is required to organize a symposium and publish a proceeding volume. We would like to extend our sincere thanks and appreciation to the symposium organizers, invited speakers, session chairs, presenters, manuscript reviewers, and conference attendees for their enthusiastic participation and contributions. Finally, credit also goes to the dedicated, tireless, and courteous staff at The American Ceramic Society for making this symposium a huge success.

DILEEP SINGHArgonne National Laboratory

JONATHAN SALEMNASA Glenn Research Center

Introduction

This issue of the Ceramic Engineering and Science Proceedings (CESP) is one of seven issues published from manuscripts submitted and approved for the proceedings of the 38th International Conference on Advanced Ceramics and Composites (ICACC), held January 26-31, 2014 in Daytona Beach, Florida. ICACC is the most prominent international meeting in the area of advanced structural, functional, and nanoscopic ceramics, composites, and other emerging ceramic materials and technologies. This prestigious conference has been organized by The American Ceramic Society’s (ACerS) Engineering Ceramics Division (ECD) since 1977.

The 38th ICACC hosted more than 1,000 attendees from 40 countries and approximately 800 presentations. The topics ranged from ceramic nanomaterials to structural reliability of ceramic components which demonstrated the linkage between materials science developments at the atomic level and macro level structural applications. Papers addressed material, model, and component development and investigated the interrelations between the processing, properties, and microstructure of ceramic materials.

The conference was organized into the following 19 symposia and sessions.

Symposium 1

Mechanical Behavior and Performance of Ceramics and Composites

Symposium 2

Advanced Ceramic Coatings for Structural, Environmental, and Functional Applications

Symposium 3

11th International Symposium on Solid Oxide Fuel Cells (SOFC): Materials, Science, and Technology

Symposium 4

Armor Ceramics

Symposium 5

Next Generation Bioceramics and Biocomposites

Symposium 6

Advanced Materials and Technologies for Energy Generation and Rechargeable Energy Storage

Symposium 7

8th International Symposium on Nanostructured Materials and Nanocomposites

Symposium 8

8th International Symposium on Advanced Processing & Manufacturing Technologies for Structural & Multifunctional Materials and Systems (APMT), In Honor of Prof. Stuart Hampshire

Symposium 9

Porous Ceramics: Novel Developments and Applications

Symposium 10

Virtual Materials (Computational) Design and Ceramic Genome

Symposium 11

Advanced Materials and Innovative Processing ideas for the Industrial Root Technology

Symposium 12

Materials for Extreme Environments: Ultrahigh Temperature Ceramics (UHTCs) and Nanolaminated Ternary Carbides and Nitrides (MAX Phases)

Symposium 13

Advanced Ceramics and Composites for Sustainable Nuclear Energy and Fusion Energy

Focused Session 1

Geopolymers, Chemically Bonded Ceramics, Eco-friendly and Sustainable Materials

Focused Session 2

Advanced Ceramic Materials and Processing for Photonics and Energy

Focused Session 3

Rare Earth Oxides for Energy, Optics and Biomedical Applications

Focused Session 4

Ion-Transport Membranes

Special Session

2nd Pacific Rim Engineering Ceramics Summit

Special Session

3rd Global Young Investigators Forum

The proceedings papers from this conference are published in the below seven issues of the 2014 CESP; Volume 35, Issues 2-8, as listed below.

Mechanical Properties and Performance of Engineering Ceramics and Composites IX, CESP Volume 35, Issue 2 (includes papers from Symposium 1)

Advances in Solid Oxide Fuel Cells X, CESP Volume 35, Issue 3 (includes papers from Symposium 3)

Advances in Ceramic Armor X, CESP Volume 35, Issue 4 (includes papers from Symposium 4)

Advances in Bioceramics and Porous Ceramics VII, CESP Volume 35, Issue 5 (includes papers from Symposia 5 and 9)

Advanced Processing and Manufacturing Technologies for Nanostructured and Multifunctional Materials, CESP Volume 35, Issue 6 (includes papers from Symposia 7 and 8)

Ceramic Materials for Energy Applications IV, CESP Volume 35, Issue 7 (includes papers from Symposia 6 and 13)

Developments in Strategic Materials and Computational Design V, CESP Volume 35, Issue 8 (includes papers from Symposia 2, 10, 11, and 12 and from Focused Sessions 1, 2, 3, and 4); the 3rd Global Pacific Rim Engineering Ceramics Summit; and the 3rd Annual Global Young Investigator Forum

The organization of the Daytona Beach meeting and the publication of these proceedings were possible thanks to the professional staff of ACerS and the tireless dedication of many ECD members. We would especially like to express our sincere thanks to the symposia organizers, session chairs, presenters and conference attendees, for their efforts and enthusiastic participation in the vibrant and cutting-edge conference.

ACerS and the ECD invite you to attend the 39th International Conference on Advanced Ceramics and Composites (http://www.ceramics.org/daytona2015) January 25-30, 2015 in Daytona Beach, Florida.

To purchase additional CESP issues as well as other ceramic publications, visit the ACerS-Wiley Publications home page at www.wiley.com/go/ceramics.

ANDREW GYEKENYESIOhio Aerospace Institute, NASA Glenn Research Center, USA

MICHAEL HALBIGNASA Glenn Research Center, USA

Volume EditorsJuly 2014

Creep, Fatigue, and Damage Characterization

ANISOTROPIC CREEP BEHAVIOR OF A UNIDIRECTIONAL ALL-OXIDE CMC

Katia Artzt, Stefan Hackemann, Ferdinand Flucht and Marion Bartsch

German Aerospace Center (DLR) Cologne, Germany

ABSTRACT

This paper is intended to give an overview of the creep behavior of unidirectional porous all-oxide CMC including experimental results as well as numerical simulations. The creep behavior was investigated by means of creep tests proving a tension- compression asymmetry. For compression creep, stress and temperature dependencies were determined and described by power-law creep equations. Activation energy was similar for all fiber orientations whereas the stress exponent differed with respect to the loading direction. Thus, for modeling the creep behavior of CMC lamina anisotropic creep must be taken into account. Within commercial finite element software implemented anisotropic creep laws are rare. The anisotropic approach according to Hill was tested for compression creep with regard to the applicability and limitations for CMCs. Since the Hill-approach did not capture the experimental results sufficiently, compression creep was further investigated by simulations on a microscopic scale via unit cells. Thereby the effect of different creep parameters for fibers and matrix and the resulting lamina’s deformation rate could be investigated. It became apparent that additionally compaction of the porous matrix has to be included in the numerical description.

INTRODUCTION

Al-oxide CMCs based on alumina fibers and matrices are favorable materials for use at high temperature in oxidizing atmospheres. One possibility for CMCs is the application as combustion liner. As a major drawback oxide fibers show the lowest creep resistance of all ceramic fibers1 due to their predominant ionic bonds. Therefore, the knowledge of the creep behavior becomes important for construction and dimensioning of components undergoing creep deformation over the life time.

In this study, WHIPOX™ material (‘Wound Highly Porous Oxide CMC’) is investigated2. In the first processing step fiber bundles are heat treated to diminish organics, infiltrated with alumina slurry and deposited on a rotating mandrel. Due to the winding process various geometries of components can be achieved and the deposition angle of the fibers can be changed producing laminas with different fiber orientations (e.g. ±45°). The green body is dried and sintered whereby the special microstructure of WHIPOX™ evolves. The weak porous matrix (Figure 1) leads to the desired high damage tolerance due to crack deflection at fibers. The matrix porosity is 40-85 Vol.% and the composite has an overall porosity of 20-50 Vol.%. These materials are based on alumina fibers (Nextel™ 610) or aluminosilicate fibers (Nextel™ 720) and an aluminosilicate or pure alumina matrix. The aluminosilicate type shows higher creep resistance3, but the alumina variant reveals better mechanical properties and higher heat conductivity. The investigated material consists of 3000 Denier Nextel™ 610 alumina fibers embedded in a pure alumina matrix. The material is sintered at 1573 K for 1h dwell time. To gain an overview of the creep characteristics of fibers and matrix, creep experiments with a quasi-unidirectional composite (±2°) were conducted.

Figure 1. Microstructure of WHIPOX™ (SEM micrograph).

EXPERIMENTAL RESULTS

Tension and compression creep experiments were conducted on WHIPOX™ with quasi-unidirectional fiber architecture (±2°). An angle of ±2° was chosen to maintain a better handling in wet state before drying and sintering. Stress, temperature and fiber orientation were varied within the experiments. The definition for the fiber orientation with regard to the load axis used can be obtained from Figure 2.

Figure 2. Definition of the samples’ fiber orientation by the angle α between fiber and load axis. In this case, a compression load is indicated by the arrows.

Figure 3 and Figure 4 illustrate some examples of strain rate versus strain − curves for tension (Figure 3) and compression creep tests (Figure 4).

Figure 3. Tension creep rates for different stresses and fiber orientations (Left: 0°. Middle: 45°. Right: 90°.).

Figure 4. Strain rates vs. the absolute value of strain or time in compression creep. Left: different fiber orientations at 1473 K and −30 MPa. Right: Stress variation within one experiment (1448K, 0° fiber orientation) for determination of the stress exponent.

The tension experiments were conducted at stresses between 2 and 60 MPa depending on the fiber orientation due to the different strength of fibers and porous matrix (Figure 3). For tensile experiments of the 90° specimens stresses of 2-4 MPa were chosen. Damage emerged at low strains, indicated by the increase of strain rate, and fracture occurred at small strains of 0.4-1%. In contrast to the 90° orientation, the 0° samples exhibit a longer almost steady state creep regime (Figure 3, left). Experiments were terminated after 6 to 12% strain without fracture. The 45° samples show deformation characteristics of both the 0° and the 90° samples with a steady state creep regime and intermediate fracture strains between 2 to 3% (Figure 3, middle).

An example for orientation dependent compression creep at −30 MPa and 1473 K is illustrated in Figure 4 (left). The 0° specimen deforms with the lowest strain rate. With increasing angle between load and fiber axes, the absolute value of the strain rate rises. For all fiber orientations the negative creep rate is maximum at the beginning and decreases over the course of time. Only short-time (6-8 h) experiments were performed.

In case of compression tests, temperature and stress dependencies were determined according to the Norton power-law creep equation

(1)

with the creep rate , the constant A, the stress σ, the stress exponent n, the creep activation energy Q, the universal gas constant R and the temperature T.Equation (1) characterizes the steady state creep. If experimental results do not show secondary steady state creep, the minimum creep rate is chosen for creep evaluation. For compression creep no lowest or steady state creep rate could be observed. Therefore, the stress exponent (Table I) was determined by stress variation within a single experiment (e.g. Figure 4 right). For determining the activation energy experiments at different temperatures but at similar creep strains were evaluated.

For all fiber directions activation energies were found in the magnitude of about 700 kJ/mol. This is consistent with the results of an earlier creep study of WHIPOXTM 4. Contrary to the activation energy the stress exponent is not constant but varies from about 3 for fiber alignment parallel to the load (0°) to 1.9 (90°) (Table I). This indicates different stress exponents for the single components fiber and matrix. Fibers presumably dominate creep in 0° orientation with a stress exponent of 3. A similar value of 2.9 to 3.4 assigned to interface-reaction-controlled creep for fiber bundles and single fibers in tension experiments was already reported elsewhere 4-10. Considering the 90° orientation the stress exponent is about 1.9 indicating a larger influence of the matrix. Because the exact portion of matrix and fiber creep deformation in 90° has not been investigated yet, it can be only concluded that the matrix must possess a smaller stress exponent, e.g. between 1 and 2. Matrix compaction seems to play an important role, which was verified by fiber displacement visualization (Figure 5). Optical micrographs before and after creep were compared. Thereby fibers’ displacement within the matrix was evaluated. In the center of the sample, compaction led to fiber-approaching in load axis, but almost no transverse deformation could be detected.

Table I: Experimentally determined stress exponents for compression creep.

Figure 5. Optical micrographs of the cross section of a 90° sample before (left) and after (middle) compression creep. The load axis was horizontal. Superposition of the fibers (right) shows the matrix compression: Gray points after creep, black points before the experiment.

ANISOTROPIC CREEP DESCRIPTION WITH THE APPROACH ACCORDING TO HILL

The aim of creep characterization in this case is to determine deformation behavior at high temperature. It is feasible to describe deformation of complete laminas using anisotropic creep laws which will be discussed in this paragraph. As a supplement more information about the interaction between fibers and matrix can be gained by assigning isotropic creep laws to matrix and fibers on a microscopic scale (next section).

In commercial FE-software, models for anisotropic creep are rare. One material model uses the Hill potential theory, originally developed as an anisotropic yield criterion of metal plasticity11.

Analogous to the plastic potential in metal yield, a creep potential f is defined as

(2)

for the rate-dependent deformation with the effective Hill stress

(3)

F, G, H, L, M and N are the Hill parameters, σij the components of the stress tensor. The strain increment tensor results in

(4)

Thereby, the constant multiplier λ can be replaced and the creep rate expressed as12

(5)

is the equivalent creep strain rate. In this case, a slightly varied Norton approach (equation 1) is chosen with a parameter A dependent on the creep strain and temperature so that

(6)

σ0 is the applied tension or compression stress. With the assumption of isotropic behavior in the 2-3 plane (direction 1 is along the fiber axis), the 6 Hill parameters and the parameter A are determinable with measurements in 3 directions (0°, 45° and 90° fiber orientation), as described by Hyde et al.12. Afterwards the creep behavior of all other orientations can be calculated and compared to experimental results. With a determined stress exponent n and activation energy Q creep rates at other temperatures and stresses are also accessible.

The Hill model was adopted for the compression creep using the 0°, 45°, 90° results at 1473 K, 30 MPa (Figure 6) and a stress exponent of 1.9 which corresponds to the exponent of the 90° orientation. Ideally, the Hill parameters are constant for a material. However, in this case parameters decreased with an increasing creep strain (Figure 6, left). For further calculations, mean Hill parameters were used (Table II). The determination of the average was conducted for strain from 2.4% to 3.2% because of the higher reliability of the measured values.

Figure 6. Determined Hill parameters for compression creep curves at 1473 K and −30MPa (left) and parameter A from equation (6) for averaged Hill parameters of Table II for a stress exponent of 1.9 (right).

Table II: Mean Hill parameters.

Due to averaging no definite solution for the parameter A from equation (6) exists. Hence A (Figure 6, right) was determined by a least square method. The Hill parameters and A were used to calculate creep rates in 15° and 30° fiber orientation as well as for another temperature and stress. In Figure 7 and Figure 8 the results are depicted.

Figure 7. Experimental creep rates and calculated creep rates according to Hill (dotted lines) for different fiber orientations at 1473 K and −30 MPa.

Figure 8. Conversion of Hill creep rates to other stresses or temperature with a chosen stress exponent of 1.9. Left: different compression stresses at 1423 K of 90° samples. Right: 0° oriented specimen at 1473 K.

The calculated creep rates according to Hill’s approach fit well to the experimental strain rates in case of constant stress even though mean Hill parameters were used (Figure 7). The greater the strain the smaller is the deviation found between experimental and calculated values. Strain measuring inaccuracy at the beginning of the 30° curve and stress redistribution could be the reason for the slightly shifted experimental curve. Nevertheless the basic creep behavior is gathered by the Hill calculation. This is only valid for 30 MPa, because the Hill parameters were calculated for this special load case. Conversion to other stresses cannot be performed in a straight forward manner: The stress exponent as well as the shape of the graph depends on the fiber orientation. In the Hill model, only one stress exponent and one total creep equation can be specified. Whereas the calculated Hill creep rates describe the curves of the 90° samples with a stress exponent of 1.9 quite well, the Hill model does not fit for the 0° sample. For small strains the absolute creep rate according to Hill exceeds the observed data, while in the case of 0° samples the absolute experimental creep rate is underestimated at higher strains. Consequently it is not possible to describe creep of all fiber orientations with high accuracy using a constant set of Hill parameters. The use of strain-dependent Hill parameters will certainly improve the results but the problem of varying stress parameters is not solved even by this refined approach.

Another shortcoming of using the Hill model is worth mentioning. Since the model is based on metal yielding, isochoric behavior is assumed with the hydrostatic pressure having no Influence on creep, only considering the deviatoric components of the stress tensor. For WHIPOX™, which has a porous matrix, hydrostatic pressure is likely to result in volume shrinkage. Another assumption of the model is the symmetric behavior under tension and compression apparent at the squared stress components in equation (3). This is not true for WHIPOX™ showing a significant creep asymmetry. A solution of the last problem is provided by Voyjiadjis13 who developed a model considering the deviating creep behavior in tension and compression. However, both models lack the possibilities for incorporating fiber orientation dependent stress exponents and activation energies or different creep equations for tension and compression tests. For compressible porous materials the Hill model has to be modified or an entirely new model must be developed which is beyond the scope of this work. Prior to that, it is first of all interesting to understand the creep mechanisms and their contribution on the overall creep deformation of the CMC. In order to learn what the necessary details of numerical models for capturing the creep behavior are, microscopic models explicitly describing fiber and matrix have been generated and analyzed.

NUMERICAL SIMULATION OF UNIT CELLS

Compression experiments (Figure 4 left) proved that creep rates strongly depend on the fiber orientation and also on creep properties of the single components fiber and matrix. Thus, numerical simulations on a microscopic scale with isotropic creep laws were conducted by means of the finite element software ANSYS14 to understand the interaction of fibers and matrix and to comprehend the experimental results in compression tests. A unit cell was chosen as geometry instead of a much larger representative volume element (RVE) which would deliver a more precise description of the real fiber distribution but would also increase the numerical effort. The observed unit cell for 90° orientation consisted of 4314 elements describing one fiber and the surrounding matrix. Dependent on the material laws a simulation with a resulting creep strain of 2% took 4.5 hours using four CPUs (Intel® Core ™ i7-2600 CPU @ 3.4GHz). According to Mansilla15 the ideal edge length of an RVE would be 50-times the fiber radius, thus 18-times the length of the unit cell, increasing the computational effort enormously to an unsustainable extent. Hence, only unit cells were further investigated. Owing to the smaller dimensions compared with an RVE the simulation results of unit cells cannot be transferred one-to-one to the experimental results but general conclusions concerning deformation mechanisms may be drawn. The unit cells for the different fiber orientations are depicted in Figure 9. According to the angle between force direction and fibers, the size of the unit cells had to be varied to preserve periodicity. The fiber diameter accounts 12 μm and the fiber volume content 40%. The models were meshed with quadratic 20-node solid elements (SOLID 186). To enable periodic boundary conditions, 3 master nodes were introduced (MASS21 elements) for building up the constraint equations. A constant force was applied on the master node coupled with the faces normal to the z-direction.

Figure 9. Unit cells with different fiber orientations (from left to right: 0°, 30°, 45°, 60°, 90°).

Table III lists the material properties of the components. Fibers and matrix were considered as isotropic, and in a first step the Norton-equation for steady state creep was assumed. Because of missing parameters for compression creep, tension creep data for the fibers were adapted from elsewhere4. The data for the matrix was estimated as follows. Based on the experimental results for the 0° and 90° fiber orientation the overall creep rate for the matrix was chosen higher and the stress exponent lower than for the fibers.

Table III. Material data for simulation − elastic and Norton-creep-constants, (equation 1).

The resulting deformation curves are depicted in Figure 10.

Figure 10. Resulting strain rates of the unit cell simulations at a compression stress of 50 MPa at 1473 K. For both fiber and matrix a Norton creep law was used (parameters Table III).

Although the stationary Norton creep law is used, all curves exhibit a short primary creep regime. This is caused by the different elastic and creep constants and the consequently developing stress redistribution until a stress equilibrium is reached, also discussed elsewhere17. The 0° sample has the lowest creep rate which coincidences with experimental results. Apart from that the sequence of the deformation rates is different. Whereas in experiments the 90° sample shows the fastest length variation, in the simulation the 45° sample has the greatest negative creep rate. Compared to the experiments, the 30° and 60° unit cells have higher deformation rates, too. This can presumably be explained by the emerging deformation behavior. Figure 11 illustrates the resulting deformations of the unit cells. Whereas edges and faces of the 0° and 90° oriented unit cells stay flat, a sliding process between fibers and matrix accompanied by a change of the fiber orientation with respect to the applied load resulted in uneven surfaces of the unit cells in all other cases.

Figure 11. Deformation of the unit cells for different fiber orientations (from left to right: 0°, 30°, 45°, 60°, 90°).

To have a better visual impression of the deformation, as an example the unit cell of the 45° sample is translated in sequence periodically with the result of bigger segment of the fiber-matrix structure (Figure 12).

Figure 12. Sequence of unit cells (45°) for a better visual understanding of the arising fiber-sliding in the simulation with a scaled deformation by a factor of 100. The creep strain after 2000s creep at 50MPa in load direction is shown.

Quite likely the sliding process is mainly responsible for the high creep rates of the 30°, 45° or 60° unit cells. Yet, a Norton creep law was assigned to both fibers and matrix. Assuming that the matrix dominates creep in the 90° samples, its creep behavior can be better described by a primary creep law with regard to the experimental results (90° curve, Figure 4 left):

(7)

A Norton creep law seems still appropriate for the fibers. After a short transient regime, experiments with fiber-dominated creep (0° curve Figure 4, left) showed an almost constant creep rate. Therefore new simulations with different creep parameters for fiber and matrix (Table IV) were conducted. Figure 13 shows the resulting creep curves for the chosen parameters.

Table IV: Simulation creep constants for equation (7): Fibers are assigned to a strain-independent creep law, the matrix to a primary creep law.

fiber

matrix

A

1.17E13

4.44E17

n

3

1

b

0

-0.7252

Q[kJ mol−1]

710

710

Figure 13. Simulation deformation rates at 1473 K and −30MPa. For comparison, experimental results of the 0°, 45° and 90° oriented samples are also plotted.

Both the 0° and the 90° curves fit well to experimental results whereas the absolute strain rate of the 45° sample is still too high. The 45° curve ought to lie between 0° and 90° as experiments showed. Thus changing creep laws from pure steady state creep to a combination of primary and secondary creep improves the simulation results only slightly. The lacking compressibility in numerical models seems to have great influence on the results. In reality compaction due to matrix compressibility emerges (Figure 5). Therefore, non-isochoric behavior will be further included.

CONCLUSION

Experimental results of a quasi-unidirectional CMC revealed different creep behavior in tension and compression. Due to the porous matrix, the samples with 90° fiber orientation showed the largest creep deformation rates. In tension creep the fracture strain was 0.4-1% for 90° samples whereas in 0° fiber orientation creep strain above 8% was feasible. At the beginning of compression experiments, the absolute creep rate was highest and decreased continually. Creep parameters, i.e. temperature and stress dependencies, were determined in case of compression stress. Thereby, the activation energy averaged to 700 kJ/mol and the stress exponent varied with the fiber orientation from 3 to 1.9. The variation in the stress exponent was presumably caused by different stress exponents of fibers and matrix and quite likely due to the effect of matrix compaction. Latter one could be visualized by optical micrographs.

The description of anisotropic creep via an equivalent stress according to Hill was tested for compression creep of WHIPOX™. It allows a fast numerical estimation of the CMCs’ deformation behavior. The Hill model consists of few parameters which can easily be determined with a couple of measurements. One further advantage is the built-in implementation in commercial finite element software, e.g. ANSYS. However, the reliability of the results is limited. For WHIPOX™, the most important limitations are the assumed isochoric behavior of the material, tension and compression symmetry as well as the fact that only one stress exponent and creep equation can be specified.

Unit cells gave a qualitative impression of the interaction and stress redistribution between fibers and matrix in compression creep. The effects of creep laws were investigated. As a result, fibers can be assigned to a Norton creep law whereas the matrix must be at least described with a primary creep law. For 0° and 90° samples the use of different creep laws for fiber and matrix was sufficient to capture the experimental results. However, the unit cell simulation resulted for the 45° sample in a higher deformation than for the 90° sample, which is in contradiction to the experiments. It is likely that the lacking compressibility in the numerical models has great influence on the results. Thus, non-isochoric behavior seems to be an essential detail for porous CMCs.

ACKNOWLEDGEMENT

The authors would like to thank B. Kanka and P. Herzog for manufacturing of WHIPOX™ material.

REFERENCES

1 Tressler, R. E. and DiCarlo, J. A., 1993, High temperature mechanical properties of advanced ceramic fibers, in “High Temperature Ceramic Matrix Composites (HAT-CMC-1)”, eds. R. Naslain, J. Lamon, and D. Doumeingts, Bordeaux, France, pp.33–49.

2 Kanka, B., and Schneider, H., “Aluminosilicate fiber/mullite matrix composites with favorable high-temperature properties.” Journal of the European Ceramic Society 20.5 (2000): 619–623.

3 Wilson, D. M., and Visser, L. R., “High performance oxide fibers for metal and ceramic composites.” Composites Part A: Applied Science and Manufacturing 32.8 (2001): 1143–1153.

4 Hackemann, S., Flucht, F., and Braue, W., “Creep investigations of alumina-based all-oxide ceramic matrix composites.” Composites Part A: Applied Science and Manufacturing 41.12 (2010): 1768–1776.

5 Pysher, D. J., and Tressler, R. E., “Tensile Creep Rupture Behavior of Alumina-Based Polycrystalline Oxide Fibers.” Proceedings of the 16th Annual Conference on Composites and Advanced Ceramic Materials, Part 1 of 2: Ceramic Engineering and Science Proceedings, Volume 13, Issue 7/8. John Wiley & Sons, Inc., 2009.

6 Pysher, D. J., and Tressler, R. E., “Creep rupture studies of two alumina-based ceramic fibres.” Journal of materials science 27.2 (1992): 423–428.

7 Wilson, D. M., Lueneburg, D. C., and Lieder, S. L., “High Temperature Properties of Nextel 610 and Alumina-Based Nanocomposite Fibers.” Proceedings of the 17th Annual Conference on Composites and Advanced Ceramic Materials, Part 1 of 2: Ceramic Engineering and Science Proceedings, Volume 14, Issue 7/8. John Wiley & Sons, Inc., 1993.

8 Hammond, V. H., and Elzey. D. M., “Comparing the creep response of alumina tows and single filaments.” Scripta materialia 46.4 (2002): 287–291.

9 Wilson, D. M. “New high temperature oxide fibers.” High Temperature Ceramic Matrix Composites (2001): 1–12.

10 Armani, C. J., et al. “Creep of Nextel™ 610 Fiber at 1100° C in Air and in Steam.” International Journal of Applied Ceramic Technology (2012).

11 Hill, R. “A theory of the yielding and plastic flow of anisotropic metals.” Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 193.1033 (1948): 281–297.

12 Hyde, T. H., et al. “Anisotropic creep and fracture behaviour of a 9CrMoNbV weld metal at 650 C.” Anisotropic Behaviour of Damaged Materials. Springer Berlin Heidelberg, 2003. 295–316.

13 Voyiadjis, G. Z., and Zolochevsky, A., “Modeling of secondary creep behavior for anisotropic materials with different properties in tension and compression.” International Journal of Plasticity 14.10 (1998): 1059–1083.

14 ANSYS Inc., http://www.ansys.com, ANSYS Mechanical APDL 14.5.

15 Mansilla, D.T. “Analysis and Simulation of Transverse Random Fracture of Long Fibre Reinforced Composites”. Diss. Universitat de Girona, 2005.

16 “http://solutions.3mdeutschland.de/3MContentRetrievalAPI/BlobServIet?lmd=1269270794000&locale=de_DE&assetType=MMM_Image&assetId=12585637171208rblobAttribute=ImageFile,” [Online]. [Accessed 11 11 2013].

17 Holmes, J. W., and Wu, X., “Elevated temperature creep behavior of continuous fiber-reinforced ceramics.” High Temperature Mechanical Behavior of Ceramic Composites, ed. Nair, SV and Jakus, K (1995): 193–259.

INDICATORS FOR THE DAMAGE EVOLUTION AT INTERMEDIATE TEMPERATURE UNDER AIR OF A SIC/[SI-B-C] COMPOSITE SUBJECTED TO CYCLIC AND STATIC LOADING

Elie Racle1,2, Nathalie Godin1, Pascal Reynaud1, Mohamed R’Mili1, Gilbert Fantozzi1, Lionel Marcin2, Florent Bouillon3, Myriam Kaminski4

1 MATEIS, INSA-Lyon, F-69621 Villeurbanne, France

2 SNECMA − Groupe SAFRAN, rond-point René Raveau, F- 77550 Moissy Cramayel, France

3 HERAKLES − Groupe SAFRAN, Les Cinq Chemins, F- 33185 Le Haillan, France

4 ONERA, 29 avenue de la Division Leclerc, F- 92322 Chatillon, France

ABSTRACT

The low density and the high tensile strength of Ceramic Matrix Composites (CMC) make them a good technical solution to design aeronautical structural components. To fully understand damage mechanisms and be able to design components, its behavior has to be analyzed during fatigue tests. The aim of the present study is to compare behavior of this composite under static and cyclic loading. Tests are realized under the same conditions of temperature and maximal load levels in order to determine the effects of cycles on the sequence of damage mechanisms. Hence the evolution of mechanical parameters is analyzed. Nevertheless the complexity of mechanisms and duration of tests make the use of complementary damage indicators necessary. Different approaches based on acoustic emission can be taken into consideration in order to quantify damage along the tests. In this case the analysis of acoustical energy is studied by comparing to the evolution of strain energy. This method enables to point out different damage levels during tests.

INTRODUCTION

To optimize design of parts, the mechanical behavior has to be fully understood. The aim of the study consists in analyzing and comparing material behavior under cyclic and static fatigue loadings, at the same temperature and under air, to determine effects of a cyclic loading on damage and lifetime. CMCs seems to be promising material for aeronautical applications, even if its constituent materials are brittle, the strain at failure is rather high due to considerable matrix cracking and cracks deflection at interfaces1. However, these materials behavior is affected by oxidation of interphases and fibers and the ultimate failure is governed by slow crack growth in fibers2. Self-healing material has been developed to protect fibers against oxidation, which increased largely the lifetime of material. Nevertheless, under air and for temperature above 550°C, self-healing is not significant enough to fully protect the material. This is why it is important to understand the material behavior for those temperatures.

As specified above, the lifetimes under these types of loadings are rather long which makes it hard to realize tests on laboratory equipment. This kind of studies needs to be done with a limited number of tests, thus the use of different techniques to monitor the damage in real time is mandatory. Acoustic emission (AE) appears to be a good candidate in this case. It consists of recording transient elastic waves on the material created by damage mechanisms. There are several studies referring to this kind of method for different types of CMCs under tensile tests3,4,5. For fatigue tests, damage can be analyzed from different points of view, first by linking each acoustical event to the damage mechanism which generated it6. This process needs clustering algorithms7. Another approach consists in considering the evolution of released energy8-10. It is generally accepted that the energy of an AE signal is related to the energy released by the source. Consequently, AE energy gives information about material damage; it is then possible to point out precursory elements to ultimate failure or to simulate AE energy evolution with a power law to determine lifetime.

In the current study, the global AE is analyzed and compared to the mechanical energy during cyclic and static fatigue tests. This process is based on studies realized by Minak11,12 on organic matrix composites. The goal of this process is to determine new damage markers for CMCs.

EXPERIMENTAL

The material is composed of Nicalon SiC fibers coated with PyC and a self-healing [Si-B-C]. The fibers reinforcement is composed of several layers of 2D satin fabrics linked together by strands of fibers in the third direction. In this study all the specimens have a dog-bone shape with a thickness of 4 mm and a gauge section of 60 mm × 16 mm.

Tests are realized at a temperature of 450°C. This temperature is critical for the material since SiC can be degraded by oxidation without any self-healing effect. Three different types of tests are performed: tensile tests, static fatigue tests and cyclic fatigue tests. Cyclic fatigue tests are made on hydraulic tensile test machine whereas tensile and static tensile tests are made on a pneumatic tensile machine which has been designed to allow a long constant load. Strain is measured using extensometer. In the case of static fatigue tests, in order to determine what stress level creates damage on the material because of oxidation, the imposed stress increases every Ti (time for 18% of Nf, number of cycles to failure) of 6% of tensile strength. At the same time cyclic fatigue tests are realized with imposed stress oscillating at a frequency of 2 Hz between 0 and an incremental maximum value which is increasing of 6% every 18% of Nf (Figure 1)

Figure 1. Applied stress for a. Static fatigue test b. Cyclic fatigue test

ACOUSTIC EMISSION

Acquisition System

Two piezoelectric sensors (Micro80, Mistras Group) are maintained on the specimen surface. Medium viscosity vacuum grease is used to ensure a good coupling between the specimen and sensors. Each sensor is connected to the data acquisition system (PCI2, Mistras Group) via a preamplifier with a 40 dB gain and 20-1200 kHz bandwidth (Mistras Group). For each detected signal, with an acquisition threshold of 45 dB on the pneumatic tensile machine and 55 dB on the hydraulic tensile machine, the data acquisition system records waveform descriptors such as amplitude or duration and mechanical information (stress, strain).

Location of AE sources

During tests, AE is recorded with 2 piezoelectric sensors, one on each side of the reduced section (Figure 2). The position of a detected source can be determined linearly knowing the wave velocity in the material using the formula described in eq. (1)

Figure 2. AE setup on a composite specimen

(1)

where v is the wave velocity in the material, tsens1 and tsens2