UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Composite sol-gel alumina ceramic/siloxane coatings Li, Gesheng 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-0534.pdf [ 15.98MB ]
Metadata
JSON: 831-1.0078714.json
JSON-LD: 831-1.0078714-ld.json
RDF/XML (Pretty): 831-1.0078714-rdf.xml
RDF/JSON: 831-1.0078714-rdf.json
Turtle: 831-1.0078714-turtle.txt
N-Triples: 831-1.0078714-rdf-ntriples.txt
Original Record: 831-1.0078714-source.json
Full Text
831-1.0078714-fulltext.txt
Citation
831-1.0078714.ris

Full Text

COMPOSITE SOL-GEL ALUMINA CERAMIC/SILOXANE COATINGS B Y G E S H E N G LI M. ENG., INSTITUTE OF METAL RESEARCH, CHINESE A C A D E M Y OF SCIENCE, 1994 B.ENG., D E P A R T M E N T OF MATERIAL SCIENCE, SHANGHAI J IAOTONG UNIVERSITY, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF MASTER OF APPLIED SCIENCE in T H E FACULTY OF GRADUATE STUDIES D E P A R T M E N T OF METALS A N D MATERIALS ENGINEERING W e accept this thesis as conforming To the required standards THE UNIVERSITY OF BRITISH COLUMBIA June 2004 © Gesheng Li, 2004 THE UNIVERSITY OF BRITISH COLUMBIA FACULTY OF GRADUATE STUDIES Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Gesheng Li 26/07/2004 Name of Author (please print) Date (dd/mm/yyyy) Title of Thesis: Composite Sol-Gel Alumina Ceramic/Siloxane Coatings .Degree: Master of Applied Science Year: 2£>€>H Department of Metals and Materials Engineering The University of British Columbia Vancouver, BC Canada grad. ubc.ca/forms/?form I D=THS page 1 of 1 last updated: 26-Jul-04 ABSTRACT This thesis relates to the novel , re lat ively l o w temperature process (160-300°C) for preparat ion o f chemica l ly bonded composite sol-gel ( C B - C S G ) coatings. Th is " w a r m " temperature process (160-300°C) overcomes some drawbacks o f the convent ional composi te sol-gel process, i n part icular m i n i m i z a t i o n o f stresses due to di f ferent ial thermal expansion coeff ic ient between the coat ing and the substrate. The biggest cont r ibut ion o f this research involves use o f si loxane bond coat between the C B -CSG coating and the m i l d steel substrate, w h i c h relaxes residual stress i n the coat ing, as we l l as protects the substrate f r o m damage (corrosion) dur ing the chemical bonding process. A method o f mu l t i - gun spraying to u n i f o r m l y distr ibute the phosphates throughout the coat ing dur ing the chemical bond ing process is also an or ig ina l contr ibut ion o f th is Thesis. The resul t ing coatings are free o f surface cracks, have med ium hardness (6.0GPa), moderate adhesion (42.0MPa) and good scratch resistance (17.0kgf) . E lectrochemical analysis shows that the mul t i - layer coat ing composed o f the siloxane bond coat and the C B - C S G " top coat" forms a physica l barr ier against wet corrosion. i i T A B L E OF CONTENTS A B S T R A C T T A B L E O F C O N T E N T S LIST O F FIGURES LIST O F T A B L E S N O M E N C L A T U R E A C K N O W L E D G E M E N T S C H A P T E R 1 INTRODUCTION , 1 1.1 Introduction to Sol-Gel 2 1.2 Advantages o f Sol-Gel Technique 3 1.3 Limitat ions o f Sol-Gel Technique 4 1.4 Poly(methylphenylsiloxane) 6 1.5 Objectives 6 C H A P T E R 2 L I T E R A T U R E R E V I E W 8 2.1 Composite Sol-Gel Ceramic Coatings 8 2.2 Aqueous Sol-Gel Ceramic Coatings on Steel 11 2.3 Functionally Graded Bond Coat 12 2.4 Chemically Bonded Phosphate Ceramics 14 2.5 Residual Stresses in Mult i - layer Coatings 15 2.6 Ceramic/polymer Composite Coatings 18 2.7 Adhesion o f Coatings 20 2.8 Chloride Corrosion 23 i i i 2.9 Corrosion Protection by Mult i -Layer Coatings 25 C H A P T E R 3 E X P E R I M E N T A L P R O C E D U R E 27 3.1 Substrate Preparation ^ 3.2 Deposition o f Coatings 27 3.3 L o w Temperature Process 29 3.5 Scratch Tests 30 3.5 Hardness Measurements 31 3.5 Adhesion Tests 33 3.6 A i r Permeability Measurements 34 3.7 "Wafer" Curvature Radius Measurements 34 3.8 Microscopy and X R D 3 5 3.9 Potentiodynamic Evaluations o f the Coatings J C H A P T E R 4 RESULTS AND DISCUSSION 38 4.1 Coating Structures 38 4.1.1 Group A Structure 38 4.1.2 Group B Structure 40 4.1.3 Group C Structure 42 4.2 Cracks and Air Permeability 44 4.2.1 Cracks 43 4.2.2 Air -Permeabi l i ty 49 4.3 Contact Measurements 51 4.3.1 Hardness o f CB-CSG Ceramic 51 4.3.2 Interface Fracture Toughness 56 iv 4.3.3 Nano Indentation and Sonic Wave Modulus 60 4.4 Scratch Resistance 62 4.4.1 Crit ical Scratching Force for Coatings 62 4.4.2 Scratch Hardness o f Coatings 66 4.5 Thermal Residual Stress 69 4.6 Adhesion 76 4.6.1 Group A Adhesion 76 4.6.2 Group B Adhesion 79 4.6.3 Group C Adhesion 81 4.7 Potentiodynamic Evaluation of Coating 83 CHAPTER 5 CONCLUSIONS 87 5.1 Summary on Coating Properties 87 5.2 Summary on Adhesion 88 5.3 Summary on Contact Measurements 89 5.4 Summary on Residual Stress and Potentiodynamic Evaluation 90 5.5 Conclusions 91 CHAPTER 6 FUTURE WORK 93 REFERENCES 95 Appendix I Determination o f Young 's Modulus by Sonic Waves 110 Appendix II Original Data o f the Radius o f Curvature vs. Thickness 114 v LIST OF FIGURES P A G E Fig. 1.1. Mo lecu la r structure o f PMPS siloxane 5 Fig.3.1 Schematic representations o f three basic mul t i - layer structures 28 Fig.3.2 Sett ing o f the p u l l - o f f tests 33 Fig.4.1.1 S E M image o f dual- layer coating 38 Fig.4.1.2 Ceramic/po lymer composite layer between bond-coat and top-coat 39 Fig.4.1.3. CB-CSG/s i loxane / C B - C S G tr i - layer structural coat ing (Group B ) 40 Fig.4.1.4 Tr i - layer Coat ing structure 41 Fig.4.1.5 Cross section image o f mono layer C B - C S G coat ing 42 Fig.4.1.6 The phosphorus E D X mapping o f Fig.4.1.5 42 Fig.4.2.1 Dry ing-shr inkage cracks 44 Fig.4.2.2 Web- l i ke phosphate cracks 45 Fig.4.2.3 The cracks due to the mismatch o f C T E 46 Fig.4.2.4 The surface crack-free coating by 160°C 4 days cur ing 47 Fig.4.2.5 The h igh magni f ica t ion o f surface crack-free coat ing 48 F ig.4.3.1. S E M indentat ion trace o f C B - C S G surface, V icke rs indent 300g load 52 Fig.4.3.2 S E M indentat ion trace on C B - C S G surface, 300g load 53 Fig.4.3.3. The S E M indentat ion trace o f C B - C S G ceramic coat ing, V ickers indent 54 300g load, Group C Fig.4.3.4 The S E M indentat ion trace o f C B - C S G ceramic coat ing, V ickers indent 55 300g load, Group C vi Fig. 4.3.5 Indentat ion trace o f substrate-polymer-coating cross section, indented at 55 300g load, Group A Fig.4.3.6 M o d e l for the ident i f icat ion o f interface fracture toughness 57 Fig.4.3.7. Opt ica l images o f typ ical coat ing detachment after R o c k w e l l C 58 indentat ion Fig.4.3.8 Detachment area o f Fig.4.3.7 by image analysis 58 Fig.4.3.9. A p lo t o f substrate surface roughness and interface fracture toughness 59 Fig.4.4.1. (a) Opt ica l micrograph o f a scratch groove; (b) curves o f norma l force, 63 acoustics, effect ive f r i c t ion and distance; Fig.4.4.2. The scattering o f scratch cr i t ical force and coat ing thickness (mono- layer 64 Group C C B - C S G coating) Fig.4.4.3. Opt ica l micro-graph o f scratching made on C B - C S G coat ing, 100X, 66 scratch m o v i n g d i rect ion f r o m r ight to left Fig.4.4.4. Schematic o f cohesive fai lure and interfacial fa i lure 67 Fig.4.5.1 The C B - C S G thickness Vs. the radius o f curvature 70 Fig.4.5.2. A typ ica l interferometr ic result f r o m a C B - C S G coat ing ( r = l .35m) 71 Fig.4.5.3 A typ ica l interferometr ic result f r o m a bare substrate ( r = l 7 . 2 4 m ) 72 Fig.4.5.4 30|j,m C B - C S G coat ing, containing 5-lOu.m si loxane bond coat ( r=14.47) 73 Fig.4.5.5 The thickness o f si loxane bond coat vs. curvature radius 74 Fig.4.6.1. Schematics o f fracture mechanisms in di f ferent coat ing structures 76 Fig.4.6.2. Fracture surface o f Group A after an adhesion strength test 77 Fig.4.6.3. Appearance o f the fracture surface generated i n adhesion test 78 vii Fig.4.6.4. A n appearance o f fracture surface after a p u l l - o f f adhesion test 78 Fig.4.6.5. Schematic o f fracture path for sample shown in Fig.4.6.4 79 Fig.4.6.6. S E M image o f Group B fracture surface after an adhesion strength test 80 Fig. 4.6.7 A crack was constrained in one layer by the si loxane B C 80 Fig.4.6.8. S E M image o f Group C fracture surface after an adhesion strength test 81 Fig. 4.7.1 Polar izat ion curves o f the mul t i - layer CSG/s i loxane coat ing (Group A - C ) 84 on the m i l d steel substrates compared w i t h that o f the bare m i l d steel ( p H = 5 , D I H 2 0 ) viii LIST OF TABLES P A G E Table 4.2.1 The air permeabi l i ty o f various structures 50 Table.4.3.1 The average values o f hardness and modulus measured by nano-indenter 61 Table 4.4.1 Scratch resistance o f various structures 62 Table 4.7.1 Corros ion protect ion ef f ic iency o f di f ferent structure coatings 86 Table.5.1 The summary o f characteristics o f various coatings 93 Table I I . 1 C B - C S G thickness vs. curvature radius for Group C 114 Table I I .2 Si loxane thickness vs. curvature radius for Group A 115 ix Nomenclature Abbreviations A S M C a lumina re inforced si loxane matr ix composite B C bond coat C B chemica l ly bond C B - C S G chemica l ly bonded composite sol-gel C R curvature radius or radius o f curvature C S G composi te sol-gel C T E coef f ic ient o f thermal expansion C P C ceramic/po lymer composite D I w a t e r de- ionized water D S I depth sensing indentat ion F E A f in i te element analysis F G B C funct iona l ly graded bond coat M A P mono a l u m i n u m phosphate M M T montmor i l l on i te M P S methy lphenyls i loxane M S m i l d steel P A phosphor ic acid P C S G phosphated composite sol-gel P M M A po ly (methylmethacry late) P M P S po ly(methy lpheny ls i loxane) P V D physical vapor deposi t ion Ref. reference S A M C si loxane mod i f i ed a lumina matr ix composite S E M scanning electron microscope SG sol-gel SS stainless steel T B C thermal barrier coat ing U T S u l t imate tensile strength Vs . versus X R D x-ray d i f f rac t ion L a t i n Symbo ls a contact radius (urn) A, B empir ica l correct ion factors i n sonic wave fo r modu lus determinat ion b scratch w i d t h ( u m ) d density o f sample (g /cm 3 ) Ds density o f substrate (g /cm 3 ) Dsc average density o f coat ing and substrate (g / cm 3 ) E Y o u n g ' s modulus (GPa) Ei elastic modulus o f C B - C S G coating (GPa) £2 elastic modulus o f substrate (GPa) £ j average elastic modulus (GPa) En Y o u n g ' s modulus o f fu l l y dense coating (GPa) Ec Y o u n g ' s modulus o f coat ing (GPa) ECb elastic modulus o f C B - C S G coating (GPa) Ed dynamic Y o n g ' s modulus (GPa) Ep elastic modulus o f si loxane (GPa) Es elastic modulus o f substrate (GPa) / ef fect ive f r i c t i on coeff ic ient FN normal load (N) F( transverse force ( N ) G shear modulus (GPa) Gc average shear modulus o f coat ing (GPa) Ga-c average shear modulus o f coat ing and substrate (GPa) hi, h2 thickness o f coat ing and substrate respectively (urn) hc, hp thickness o f C B - C S G layer and si loxane respect ively (urn) H V ickers hardness (GPa) hc thickness o f C B - C S G layer (urn) hp thickness o f si loxane (urn) i cor, icor corros ion current densities o f the bare and coated substrates respectively ( A / m ) Kc interface toughness (MPa. M 1 / 2 ) Kuc Interface fracture toughness (shear) (MPa. m ) x i i i k constant o f hydro lys is , molar ratio o f hydro lys is L d imens ion (m) m mass (g) n number o f samples N resonant f requency (Hz) P protect ion ef f ic iency (%) Pc porosi ty o f coat ing (%) r the radius o f curvature ( u m ) ri the radius o f indent ( u m ) Ri radius o f annular crack ( u m ) Ra the average surface roughness o f substrate ( u m ) Rc the radius o f curvature w i t h coating ( u m ) Rs the radius o f curvature w i thou t coating ( u m ) Sh scratch hardness (GPa) / thickness ( u m ) Tc thickness o f coat ing ( u m ) Ts.c total thickness o f coat ing and substrate ( u m ) Ts thickness o f substrate ( u m ) w w i d t h (\ivri) Y y ie ld strength (MPa) x i v G R E E K S Y M B O L S a constant o f porosi ty 8 def lect ion (urn) K di f ference o f curvature w i t h and w i thout coat ing ( u m ) cf thermal expansion mismatch stress (MPa) 27 elastic modulus rat io v average Poisson's rat ion o f coat ing and substrate Vi Poisson's rat io o f substrate V2 Poisson's rat io o f C B - C S G coat ing Vc Poisson's ratio o f coat ing vs Poisson's rat io o f substrate Acknowledgements I w o u l d l ike to express m y sincere gratitude to m y supervisor Dr . T o m Troczynsk i (Professor o f Ceramics, U B C ) , group leader Dr . Quanzu Yang , and other colleagues. M a n y thanks to Dr . H o w a r d Hawthorne for his suggestions on m y thesis. I w o u l d l ike to acknowledge the exper imental help f r o m Dr. Yongsong X i e and Dr. Zhao l i n Tang. I w o u l d also l ike to acknowledge the f inancial support o f the Nat iona l Science and Engineer ing Research Counc i l o f Canada, N C E - A u t o 2 1 , and Datec Coat ing Corporat ion. The Nat ional Research Counc i l o f Canada is acknowledged for the exper imental and equipment support. x i i i C H A P T E R 1: I N T R O D U C T I O N The processing and characterization o f ceramic coatings are impor tant issues in this research. For example, the space-shuttle Columbia disaster happened on February 1, 2003. I t is bel ieved that the left w i n g o f the shuttle was scratched by a piece o f f oam that fe l l f r o m the external tank dur ing launch, w h i c h destroyed the heat-resistant ceramic coating on the outer body and a l lowed the surface temperature to exceed its cr i t ica l value upon re-entry. I f the coat ing fai lure was an interfacial fa i lure, i t cou ld be labeled as a manufactur ing p rob lem. I f the coat ing fai lure was a cohesive fa i lure, however , i t cou ld be asserted to be a p rob lem related to coat ing design or to the coat ing mater ia l i tself. I n this research, a novel sol-gel processing method was developed to produce a composite sol-gel coat ing at " w a r m " temperatures (100-300°C) . The average hardness o f the coating is - 6 . 0 GPa, the average adhesion strength is ~42.0 M P a , and the average scratch cr i t ical force is - 1 7 . 0 kgf. The " w a r m " temperature process successful ly produces coatings free o f surface cracks. A l u m i n u m phosphate is used as the b ind ing phase i n the sol-gel coat ing. T o obta in suitable strength, i t is necessary to homogeneously distr ibute the phosphate throughout the coat ing through a method o f m u l t i - g u n spraying. A si loxane bond coat was developed to reduce the residual stresses due to the d i f ferent ia l thermal expansion o f the ceramic coat ing and steel substrate, w h i c h decreases the deformat ion o f the substrate. Another mer i t o f this research is that i t provides a method to produce a mul t i - layer ceramic/si loxane coat ing on m i l d steel ( M S ) w i thou t coat ing buck l ing . 1 1.1 Introduction to Sol-Gel A sol is a dispersion o f sol id particles ( -0 .1 -1 u m ) i n a l i q u i d where on ly B rown ian m o t i o n suspends the particles. A gel is a state i n w h i c h the l i q u i d phase is dispersed in the sol id and vice-versa, y ie ld ing a sol id ne twork conta in ing l iqu id components. The sol-gel coat ing process usually consists o f four steps: (1) The desired col lo idal particles are dispersed in a l i qu id to f o r m a sol. (2) The sol is deposited to produce coatings on the substrates by spraying, d ipp ing or spinning. (3) The particles i n the sol are po lymer ized through the removal o f the s tabi l iz ing components, w h i c h produces a gel . (4) The f ina l heat treatments pyro lyze the remain ing organic or inorganic components and f o r m an amorphous or crystal l ine coating [1 -4 ] . There are t w o dist inct reactions in the sol-gel process: hydro lys is o f a lcohol groups and condensation o f the result ing hyd roxy l groups. For instance, the isomorphous y - A l O O H precursor exists as the un-hydro lyzed species [A l ( O H 2 ) 6 ] 3 + be low p H 3, and can be hydro lyzed extensively w i t h increasing p H [1 ] : [Al(OH2)6f+ +kH20^[Al(OH)k{OH2)6_kf-k)+ +kH,0+ (1.1.1) kH30++kOH ~^2kH20 (1.1.2) where k is def ined as the mo la r rat io o f hydrolys is. 2 I t is general ly agreed that the preferred condensation sites are those that max im ize interactions between lone pair electrons on a bound hydrox ide l igand on one a l u m i n u m species w i t h a pro ton on a water-molecule bound to another a l u m i n u m species [1 ] . For example, t w o s ing ly -hydro lyzed monomers (k=\) condense to a d imer v i a an olat ion reaction [1 ] : 2[Al(H20f6+ -H+ - > Al{H20)5OH2+]-2H20 - > Al2{OH)2(H20)\+ (1.1.3) However , the condensation mechanism o f a l u m i n u m isopropoxide is more complex than the above descr ipt ion, and the hydrolys is o f the a l u m i n u m alkoxides is not very we l l understood [1 ] . 1.2 Advantages of the Composite Sol-Gel Technique Some non-hydrated f i l lers can be added into the sol-gel s lurry to decrease the shrinkage strain du r ing gelat ion, producing what is cal led a composi te sol-gel (CSG) . The composite sol-gel is a k i n d o f "ceramic paint" , and adheres easily to var ious metal l ic substrates l ike m i l d steel, stainless steel, a luminum, n icke l , and copper. The mono a luminum phosphate can be used as the binder i n this paint. I t fo rms a chemical bond w i t h almost everyth ing that can stand the processing temperature [3 ] . Because the phosphate is used as a chemical bond ing phase, a "chemica l l y bonded composi te so l -ge l " (CB-CSG) coat ing is produced. 3 The mater ial i n the gel state can be easily shaped into complex geometries, e.g. tube and wheel , by s imple air -spraying, d ipp ing or spinning. Ano ther advantage is that h igh pur i ty products can be produced. Because the organo-metal l ic precursor o f the desired ceramic oxides can be m ixed , dissolved i n a specif ied solvent, and hydro lyzed into a sol and subsequently a gel, the composi t ion is h igh ly control lable. I t can be sintered at l o w temperatures, usual ly around 400-700°C. General ly , the composi te sol-gel technique provides a s imple, inexpensive, and effect ive method o f p roduc ing h igh qual i ty coatings [5 ] . 1.3 Limitations of Sol-Gel Techniques There are some l imi ta t ions to the sol-gel technique, e.g. weak bond ing , porosi ty. I n part icular, the coat ing thickness is l im i ted since the mismatch o f the coeff ic ient o f thermal expansion ( C T E ) increases w i t h temperature. H igher temperature processing or service of ten generates stresses that increase w i t h coat ing thickness. W h e n the residual stresses exceed the adhesion strength, the coating w i l l break up. Part icular ly , the residual stresses i n complex geometr ic locations, e.g. edges and corners, are more severe than those i n a f lat shape, w h i c h also l im i ts w ide appl icat ion o f the sol-gel technique. Aqueous sol-gel processing typ ica l ly proceeds i n an acidic env i ronment o f p H = 2-4, w h i c h can corrode m i l d steel. A convent ional way to address this issue is to coat the substrate w i t h zinc or i ron phosphate. However , the resul t ing t h i n (less than a few microns) and usual ly micro-porous phosphate f i lms are unstable at the elevated process 4 temperature (>400°C) needed for chemical bonding. Th is can lead to the creat ion o f addit ional interfacial porosi ty , and w h i c h a l lows access o f corrosive species to the steel surface. Such a coat ing system does not provide long- term corros ion protect ion for the m i l d steel. Therefore, a non-porous and thermal ly stable "bond-coat " o f a lumina reinforced siloxane matr ix composite ( A S M C ) f i l m has been developed in this w o r k to coat the m i l d steel surface. Th is should protect the substrate dur ing C B - C S G processing at temperatures between 100-400°C as we l l as p rov id ing a si loxane membrane for corrosion protect ion in storage and service [5 ] . I n other words, when the si loxane is incorporated, the aqueous sol can be deposited on the surface o f m i l d steel and heat treated at 100-In order to overcome the l imi tat ions o f t radi t ional C B - C S G coatings, e.g. h igh residual stresses, l o w wear-resistance, and cracks, a " w a r m " temperature process has been developed. I n this system, a si loxane composite bond coat reduces the residual stresses by grading the thermal expansion. 400°C. C H 3 "I Fig.l.l. Molecular structure of PMPS siloxane [6] 5 1.4 Poly(methylphenylsiloxane) There are several reasons to choose po ly (methy lpheny ls i loxane) , P M P S , as a po lymer component i n the mul t i layer coating. P M P S has a re lat ive ly h igh service temperature (nomina l : 480°C , Cotronics Inc. U S ) due to its molecular structure. I t is an inorganic po lymer w i t h no carbon atoms in the backbone and h igh ly cross- l inked chains o f alternating s i l i con and oxygen atoms (F ig . 1.1), by means o f w h i c h i t can stand a relat ively h igh temperature [6, 7 ] . 1.5 Objectives The convent ional chemica l ly bonded sol-gel process includes an ind iv idua l deposit ion o f slurry and phosphate and f i r i ng at 400-900°C after each deposit ion. Therefore, i t was d i f f i cu l t for such a process to evenly distr ibute the phosphate throughout the coat ing. Moreover , cracks were of ten observed i n the convent ional method [5, 8 ] . Therefore, the f i rst object ive was to develop a low-temperature process (160-300°C) for the produc t ion o f a crack-free coat ing w i t h phosphate distr ibuted homogeneously throughout . I t was d i f f i cu l t to deposit the aqueous slurry on m i l d steel w i thou t buck l i ng o f the coating [9 ] . The second object ive was thus to develop a si loxane bond coat for the m i l d 6 steel substrate to prevent this buck l ing . The th i rd object ive was to characterize the C B -CSG coating and its si loxane bond coat through indentat ion tests, scratch tests, adhesion tests, and electrochemical analysis. Since the th in steel substrate of ten suffers deformat ion due to residual stresses, i t was bel ieved that the si loxane bond coat could reduce these by grading the thermal expansion. Therefore, the four th object ive was to gather evidence as to whether or not the siloxane bond coat reduces the residual stresses due to the d i f ferent ia l thermal expansion o f coat ing and substrate. 7 CHAPTER 2: LITERATURE REVIEW 2.1 Composite Sol-Gel Ceramic Coatings The sol-gel process has been successful i n the product ion o f hard, erosion resistant and corrosion resistant ceramic coatings and structural ceramics (e.g. AI2O3, Z r O i ) [1-5 ] . However , the sol-gel f i lms have tended to crack i f they are thicker than several microns [3 ] . Calc ined ceramic particulates dispersed i n a sol-gel mat r i x produce a composite sol-gel (CSG) coat ing. Th is CSG technique avoids the large shrinkage strain o f the sol-gel f i lms f o l l o w i n g heat treatment and densi f icat ion [9 ] . The th icker coatings (up to - 5 0 0 u.m) do not crack upon dry ing because the gel phase contains up to about 8 0 v o l % o f ceramic f i l ler . A n added advantage o f the chemica l ly bonded composite sol-gel (CB-CSG) is the possib i l i ty o f cont ro l l ing its strength by simple chemical reactions [5, 8, 10, 11]. I t was bel ieved that the composite sol-gel (CSG) coatings st i l l needed a re lat ively h igh temperature, t yp ica l l y i n excess o f 600°C, to gain enough strength and hardness, and to el iminate the porosi ty due to structural collapse [5, 10]. For most metal l ic substrates o f interest, inc lud ing steel, a l um inum and magnesium al loys, the m a x i m u m cur ing temperature must be, however , be low about 600°C. Th is p rob lem can be avoided i f the di f fus ional process (e.g. sintering) is replaced w i t h a chemical bond ing process [3 ] . 8 I n recent research, a paint- l ike aqueous sol conta in ing the ceramic oxide precursors and inert f i l lers has been deposited using a pressurized air-spray gun, fo l l owed by phosphat ing and cur ing at 400-900°C [8 ] . This is the composi te sol-gel ceramic technique. This technique includes dispersing f ine ceramic powders i n sol-gel solut ion, spraying the paint onto a substrate, and f ina l ly f i r i ng at 400-900°C [12 ] . The technique has successfully produced 25-200 u m coatings [4 ] , i f phosphor ic acid ( P A ) is used to init iate the fo rmat ion o f a phosphated a lumina sol-gel coat ing by subsequent heat treatment at 400-600°C. I t is general ly agreed that the l iqu id mono a l u m i n u m phosphate ( M A P ) is the best binder, w h i c h upon heat treatment is eventual ly converted to three phases, Ber l in i te and Cristobali te forms o f A I P O 4 or Var isci te, A K J H b P O ^ . The average hardness o f a C B -CSG coating is about 4.0 GPa [13] , wh i le the m a x i m u m hardness o f C B - C S G coat ing is reported to be ~ H V i o g 6.47+1.44 GPa [6 ] . Sometimes, there were some h igh hardness in isolated areas (~8.0GPa) on the C B - C S G coatings, usual ly w i t h i n smoother and shinier regions [40] . Church and Kun tson impregnated 8 5 w t % P A w i t h l o w concentrations o f some other inorganic chemicals, such as ZrCh, TiC>2, and Cr203, into the a lumina mat r ix to obtain 7.0-10.0 GPa hardness [10] . The p u l l - o f f adhesion was reported to be - 4 2 M P a [14 ] . I t is very susceptible to substrate interactions [12, 14], w h i c h included mechanical inter locks and chemical bonds at the interface, ind icat ing an oppor tuni ty to increase the adhesion further. 9 The convent ional composite sol-gel process is an ind iv idua l deposi t ion method that involves the po lymer iza t ion o f a luminum isopropoxide th rough the removal o f stabi l iz ing organic components. The coat ing is dr ied at 70°C to gel the layer and f i red at 400-900°C to pyro lyze the remain ing organic components, w h i c h fo rms an amorphous or crystal l ine layer [4 ] . The f o l l o w i n g is a wel l-establ ished process for p roduc ing composite sol-gel a lumina coatings [3 ] : 1. Heat at 550°C to f o r m a th in ox id ized f i l m on the sand-blasted M S substrate. 2. Spray a 10-60 u m composite sol-gel a lumina layer on the substrate. 3. Heat at 550°C for 20 minutes. 4. Spray composite sol to the desired thickness. 5. Heat at 550°C for 10 minutes. 6. Brush w i t h 8 5 % phosphor ic acid. 7. Heat at 250°C for 20 minutes and at 600°C for 20 minutes. There are some inevi table drawbacks to this process. First, w h e n the phosphoric acid (PA) is over- impregnated on the surface o f green CSG. I t fo rms pseudoboehmite or gelatinous boehmite, w h i c h has a l o w strength [3, 4, 12], and produces sol id a luminum phosphate at r o o m temperature, w h i c h often induces cracking. Second, the phosphate needs to reach the interface and f o r m a chemical bond between the coat ing and the substrate, but i t is d i f f i cu l t for the ind iv idua l ly deposited phosphate to penetrate through a CSG layer more than 20 u m th ick. Th i rd , the relat ively h igh temperature, 550-600°C, 10 sometimes affects the properties o f the substrate, e.g. A l or M g . I t is probable that phase transformations occur i n the substrate. F ina l ly , this process uses chemical reactions at 550-600°C [3 ] : Al203+2H3PO, -* 2Al(H2PO,)3+H20 (2.1) Al(H2PO,)3 <=> 2H3PO, + AlPO, (2.2) Al(H2P04)3+Al20, <=>3AlPO, +3H20 (2.3) These reactions are related to the compl icated t ransformat ions o f var ious phases, wh ich often generate large mismatches i n thermal expansion dur ing c o o l - d o w n f r o m 550-600°C. Therefore, i f the convent ional method is fo l l owed , cracks are inevi table. A viable w a y to produce crack-free coatings is to decrease the process temperature. Unfor tunate ly , there is no report on a " w a r m " temperature (<250°C) version o f the composite sol-gel a lumina coat ing process. 2.2 Aqueous Sol-Gel Ceramic Coatings on Steel M i l d steel is very d i f f i cu l t to coat w i t h the acidic aqueous sol-gel s lurry due to its easy corrosion dur ing heat treatment i n air. The convent ional w a y to t ry to prevent this is as fo l lows: (1) deposit the sol-gel der ived coat ing on phosphatized (zinc or i ron 11 phosphate) m i l d steel; (2) heat-treat i t i n an inert atmosphere; and (3) cover i t w i t h epoxy resin after heat treatment dur ing storage and service [15] . This method has some disadvantages: (1) the treatment i n inert atmosphere requires complex and expensive equipment [15, 16]; (2) the commonly -used phosphate layer cannot survive beyond 300°C; and (3) the epoxy resin on the surface cannot resist h igh temperatures and possesses l o w scratch resistance. Sometimes the sol-gel coat ing on M S needs to be heat treated i n air, w h i c h requires an inh ib i tor -doped sol, but these inhib i tors often decrease the adhesion strength o f the coatings [17 ] . Therefore, to deposit aqueous sol-gel coat ing on M S , the biggest challenge is to overcome the weak corrosion resistance o f M S , and to produce a coat ing w i t h h igh adhesion strength and good resistance to buck l ing . 2.3 Functionally Graded Bond Coat To address the p rob lem o f weak bonding, usual ly a bond coat ( B C ) is used to increase the adhesion strength o f a coat ing system. For example, the glue ut i l ized i n dental crowns is a good example o f a B C , i n this case a h igh l y adhesive po lymer connecting the top porcela in layer and the tooth substrate. Another example o f B C is the zinc phosphate used to bond an aluminosi l icate top- layer on an M S substrate [15 ] . 12 However , i n most cases, the var ia t ion i n composi t ion between the other layers and the B C is always the ma in reason for a large mismatch in thermal expansion. In other words, the t radi t ional ceramic/metal system common ly suffers fai lures due to the excessive residual stresses generated dur ing heating and coo l ing. The residual stresses have a substantial effect on the coating properties, w h i c h can give rise to either deformat ion o f the substrate or buck l ing o f the coating [18-21] . Thus, people have t r ied to use a funct ional ly graded bond coat ( F G B C ) , a B C w i t h a gradual var ia t ion o f compos i t ion , to solve the prob lem. The f i rs t concept o f F G B C was proposed to wi thstand the severe stress o f the C T E mismatch in ceramic engines. The unique idea o f F G B C was to prov ide a composite coat ing, where the coat ing composi t ion varies gradual ly f r o m coat ing to substrate [23] . The l i terature [21-28] showed that F G B C s cou ld grade the thermal expansion, and thus reduce the residual stresses f o l l o w i n g coo l -down f r o m a h igh temperature. Therefore, the F G B C has been an indispensable constituent fo r th ick coat ing systems to reduce the residual stresses [28-33] . A po lymer should be an ideal candidate fo r the bond coat, since po lymers have a lot o f advantages, e.g. excel lent resil ience, good adhesion, good corros ion resistance, and f lu id- impermeabi l i ty . Unfor tunate ly , there has not been any report on the sol-gel ceramic/si loxane F G B C , because most polymers begin to degrade above 150°C, wh i le 13 the convent ional sol-gel process temperature is often >400°C . Therefore, there has not been very much research on the polymer/ceramic sol-gel process. 2.4 Chemically Bonded Phosphate Ceramics I t is technologica l ly important to fabricate a lumina ceramics at " w a r m " temperatures (<300°C) . The phosphate sol-gel a lumina has a l o w fo rmat ion temperature o f ~150°C, at w h i c h the ceramic is l i ke ly to be stress-free. Moreover , the phosphate a lumina ceramic (Ber l in i te) has a h igh compressive strength o f ~110 M P a . The mono a luminum phosphate can chemical ly bond to many materials [34 ] , except some organic polymers. K ingery f i rst used the phosphoric acid to bond a lumina at 250-300°C [35] . Later, mono a luminum phosphate ( M A P ) hydrates were produced by the react ion between a lumina and orthophosphor ic acid at 100-150°C. M A P can convert to Ber l in i te above 150°C as a b ind ing phase [36-38] . A wel l -crysta l l ized AIPO4 (Ber l in i te) was even synthesized at 150°C by react ing boehmite and phosphor ic acid [37, 38-42] . The hardness values o f Ber l in i te and Var isc i te are ~6.5 GPa and - 4 . 0 GPa respectively [43-46] . I t is general ly agreed that the l i qu id mono a l u m i n u m phosphate ( M A P ) is converted to three phases upon heat treatment: Ber l in i te , Cr istobal i te forms o f AIPO4, or Var isc i te, A1(H2P04)3 [36 ] . On ly the Ber l in i te was fo rmed w h e n the M A P was cured for 4 days at ~150°C [34-36] . The Ber l in i te was fo rmed by the f o l l o w i n g reaction [34 ] : 14 Al203 + 2AIH3(P0,)2 • H20 -> AAIPO, +4H20 (2.4) Ber l in i te is the phase that bonds ind iv idua l particles and forms the Ber l in i te-bonded a lumina ceramic [34] . The cur ing t ime decreases w i t h increasing temperature [40-43] . The mono l i th ic a lumina gels can be fo rmed by hydrolys is and condensation o f a l u m i n u m alkox ide, and the phosphorus acts as the " b r i d g e " in the network o f the phosphate ceramic [39] . F r o m the earlier l im i ted research [44-49] , i t is seen to be possible to produce a wel l -crysta l l ized composi te sol-gel ceramic coating at ~160°C, w h i c h has h igh adhesion strength, h igh compressive strength, h igh hardness, and even h igh modulus. 2.5 Residual Stresses in Multi-layer Coatings The most suitable method to determine residual stresses is f in i te element analysis ( F E A ) , w h i c h can quant i tat ively ident i fy the stress d is t r ibut ion i n complex mul t i - layer structures. However , as a so-cal led "c losed- fo rm analyt ical m e t h o d , " F E A is l im i ted to case-by-case studies [50-56] . Latt ice structure measurement techniques, such as X- ray d i f f ract ion, are d i f f i cu l t to implement. A viable stress " ind ica to r " is the curvature radius (CR) o f a coat ing system [13] using the so-called " w a f e r " method. For example, a "wa fe r " curvature method has been used to measure the residual stress (140 M P a ) in an aluminide bond coat (elastic modulus « 1 lOGPa) [15] . 15 The residual stresses in the C B - C S G coat ing can be related to the f o l l o w i n g : (1) the stress due to chemical reactions; (2) phase t ransformat ion- induced stress; (3) geometr ical ly induced stress, e.g. at edges and corners; (4) thermal expansion coeff ic ient (CTE) mismatch- induced stress; (5) the external service stress. Th is research is focused on the C T E mismatch- induced stress. W h e n the C T E o f the substrate is smaller than that o f the coat ing, the coat ing is in tension. W h e n the C T E o f substrate is larger than that o f coat ing, the coat ing is in a state o f compression [8, 32, 57, 58] . The bending moment or the residual stress increases w i t h the increment o f coat ing thickness, and correspondingly the curvature radius (CR) o f the coating decreases w i t h the increase o f bending force [23, 33 ] . For the same thickness, the residual stress o f a si loxane bond coat ing should be smaller than that o f a 100% C B - C S G coating, since the visco-elast ic si loxane relaxes the stresses more easily than br i t t le ceramics [12] . The curvature radius (CR) can prov ide a context for analysis o f the relat ionship between the stress and the coat ing thickness. The salient characteristics are as fo l lows: the bending moment increases w i t h the increment o f the coat ing thickness, and tends to approach a f i x e d value, w h i l e the curvature radius o f the coat ing decreases w i t h monotonic increase o f the coat ing thickness [58] . 16 W h e n the thickness and elastic modulus o f the bond coat were m u c h smaller than those o f the C B - C S G top layer, the CB-CSG/s i loxane could be regarded as a single layer system in order to use the bi-mater ia l model for calculat ing the residual stress. Thus, the thermal mismatch stress f a 7 ) i n the coating was calculated according to the equation be low (2.5) [32 ] : 6£2(l + £)(l-v,)o-T K = — —  2— (2.5) E2h, [(zZ{2-l)2+4zZZ(l + {)2] Where K is the di f ference o f curvature w i t h and w i thou t coat ing ( — - — ) respectively hi and h2 are the thickness o f coat ing and substrate respectively, g=hi/h2 E2 and V2 are substrate elastic modulus and Poisson's ratio o f substrate respectively Ei and v\ are C B - C S G elastic modulus and Poisson's ratio o f coat ing respectively The def lect ion 8 o f equat ion (2.5) is 8 = KD 2 / 8 [32 ] , where D is the substrate diameter or length. Elastic modulus ratio o f coat ing to that o f substrate, 27 = - ^ ' ^ V ' ^ E2/(l-v2) The average elastic modulus (Es) o f the dual layer coat ing was estimated by, 17 [59-61] (2.6) hc and hp are the thickness o f the CB-CSG layer and the siloxane layer respect ively Ec and Ep are the elastic modulus o f the CB-CSG layer and the siloxane layer respectively There have been many debates over the accurate determinat ion o f the residual stress in a bi layer system. Equat ion (2.5) was bu i l t upon a " f u l l y elast ic" model [32 ] , w h i c h is also dependent on the relat ive thickness h\/h2 and the relat ive modulus ratio Z. However , this equat ion can st i l l help i n understanding the rat ionale o f the sol-gel process. 2.6 Ceramic/polymer Composite Coatings Histor ica l ly , i t was bel ieved that po lymers and ceramics were i rrelevant to each other because po lymers were ma in ly used for l o w temperature appl icat ions wh i le ceramics were often used for h igh temperature applications because o f their in t r ins ical ly strong bonds. Consequent ly, there has not been very m u c h research on the ceramic/polymer composi te system, except some studies o f l o w temperature polymer/ceramic composites, since the dif ference in the processing temperatures o f ceramics and polymers is too large. W i t h the development o f h igh temperature po lymers and the low-temperature processing o f ceramics, e.g. the chemical ly bonded sol-gel technique, i t is n o w possible to heat-treat po lymers and ceramics together and to develop the ceramic/polymer 18 composite system, w h i c h may promise a combinat ion o f the advantages o f bo th polymers and ceramics, e.g. good elast ici ty and h igh hardness. Recently, ceramic/po lymer composite (CPC) coatings have evoked intense research interests due to their unique characteristics [60-83 ] , e.g. thermal barriers [66 ] , compl iant layers [5, 81 ] , mechanical strength [82 ] , stress tolerance [54 ] , molecular barriers [82] and f lame retardant properties [83] . The development o f C P C coatings can be tracked back to the research done by Toyota i n 1990 [79, 81] . CPCs were c o m m o n l y a sort o f low- load ceramic re inforced po lymer matr ix composi te, where the ceramic was usually montmor i l lon i tes ( M M T s ) , and the po lymer mat r i x cou ld be n y l o n [79 ] , po ly (methy l methacrylate) ( P M M A ) [84 ] , epoxy resin [85 ] , polystyrene [86 ] , polyurethane [87 ] , po lyani l ine [ 8 8 ] , etc. Y e h and coworkers found that the incorporat ion o f M M T into a polyani l ine mat r i x led to an ef fect ively enhanced corros ion resistance o f po lyani l ine after a number o f measurements o f electrochemical corrosion [89 ] . Messaddeq et al. dispersed P M M A into a z i rcon ia sol and f i red at 200°C for 10 minutes, increasing the l i fe t ime o f stainless steel [90 ] . Masalsk i and colleagues [91] bel ieved that the CPC coat ing had to be f i red above 400°C to achieve h igh adhesion strength, and that thus the ma in drawback o f the ceramic/polymer composi te was the temperature l im i ta t ion . The industry a lways demands high-temperature systems. Unfor tunate ly , there has not been very m u c h research related to the high-temperature (>450°C) co-sinter ing process o f ceramics and polymers. Therefore, i t is necessary to do more intensive research exp lor ing this u n k n o w n f ie ld . 19 2.7 A d h e s i o n o f Coa t ings The adhesion is one o f the most important issues in sol-gel coat ing characterization. Unfor tunate ly there is not very m u c h in fo rmat ion about the assessment o f coat ing adhesion because the correlat ion between fai lure modes and adherence is st i l l poor ly understood despite the widespread use o f adherence tests. For example, D u et al. [70] reported that they st i l l do not f u l l y understand the di f ferent adhesion performance among the sol-gel coatings after peel testing. Sexsmith and Troczynsk i [92] proposed a model o f a peel test to calculate the exper imental ly determined stress-strain relat ionship. L i n and Berndt [93] summarized the most widely-used measurements on adhesion: A S T M - C 6 3 3 or D I N - 5 0 1 6 0 . They agreed that these tests are s imple but do not promote any deep understanding o f coat ing performance. Scratch was a pract ical w a y to assess coating adhesion [48] . Basical ly , there are t w o types o f scratch fa i lure, cohesive fai lure and interfacial fa i lure [50] . Cohesive fai lure indicates good interfacial adhesion, but interfacial fa i lure is o f ten generated by defects, e.g. pores and cracks, that are created dur ing the processing [51 ] . X i e and Hawthorne [13, 94] summarized the advantages and disadvantages o f the dif ferent ways o f assessing coat ing adhesion: p u l l - o f f tests, peel tests, four -po in t bending tests, indentat ion tests and scratching tests. They attested that mic ro-c rack ing could release the thermal stresses, and that the C B - C S G coat ing interface was strengthened 20 w i t h increased processing temperature. They also cited that they st i l l d id not completely understand the reasons o f the measured interfacial toughness and residual stress data trends in their sol-gel coat ing research [40] . But they bel ieved that the interfacial fracture toughness was a reasonable w a y to characterize the adhesion o f sol-gel coatings. I n p u l l - o f f tests, the cr i t ical normal load, at w h i c h the coat ing detachment init iates, is of ten used to determine the adhesion strength o f a coat ing. General ly, i t is bel ieved that the use o f cr i t ica l normal load (pu l l -o f f ) is adequate fo r a semi-quanti tat ive routine mon i to r ing o f the adhesion [50] . However , this is on ly suitable for compar ison between coating systems w i t h close properties. For di f ferent coatings, the comparison o f coating/substrate adhesion could not be achieved by s imply compar ing the cr i t ica l normal load. For example, i n p u l l - o f f tests, the br i t t le sol-gel a lumina coat ing of ten broke up at the interface, but the detachment o f the soft si loxane bond coat of ten occurred in the midd le o f the coat ing rather than at the interface. In this case, i t was not suitable to use the p u l l - o f f adhesion test to compare the C B - C S G coat ing and the si loxane po lymer coating. Another p rob lem i n the understanding o f coat ing adhesion was the lack o f rel iable models, even though many researchers have proposed a number o f models for adhesion bonds, such as chemical bonding [95, 96 ] , d i f fus ion [97-99] , mechanical in ter locking [100] , and residual stresses [81] . When look ing at the l i terature related to the adhesion o f sol-gel coatings to metal l ic substrates, clear relations cou ld not be found, e.g. 21 substrate roughness, heat-treatment temperature and coat ing adhesion. Th is was always due to incomplete in fo rmat ion about processing condit ions. For example, Masa lsk i and coauthors [102] reported that the sol-gel AI2O3 coat ing fired at 500°C had a better fa i lure resistance than that fired at 850°C, but they d id not give any reason w h y the higher temperature cur ing led to a better adhesion. Bergmann [104, 105] bel ieved that the surface roughness o f substrate increased the adhesion strength, and the adhesion increased w i t h the increment o f preheat ing temperature o f substrate for a g iven surface roughness. He also observed that a lumina and z i rconia had poor contact w i t h the substrates when they were sprayed on a smooth (roughness, Ra ~ 0.05 u m ) stainless steel substrate [106] . Accord ing to the w o r k o f R ichard et al . [109] , the residual stresses near the interface decreased w i t h decreasing surface roughness. The adhesion increased l inear ly w i t h the increment o f co ld (<100°C) substrate roughness, and the adhesion increased w i t h the increment o f substrate temperature, where the best adhesion could be obtained between 300 and 500°C for stainless steel [96 ] . However , a clear quant i f icat ion o f coat ing adhesion vs. substrate roughness was not g iven. O n the other hand, the substrate surface condi t ions were also important factors contr ibut ing to the adhesion strength. The metal l ic-substrate surfaces in air were covered w i t h a layer o f metal ox ide. The density and thickness o f this metal ox ide layer var ied depending on the metal-substrate, and on h o w the surface was treated. For example, phosphoric acid anodiz ing o f steel generated a hard porous i ron oxide layer w i t h a thickness o f about 50 n m [110, 111]. The i ron oxide layer had a s igni f icant popula t ion o f 22 hydroxy l groups i n a h u m i d environment. These surface hydroxy ls cou ld part icipate in the sol-gel condensation react ion to f o r m a chemical l inkage, M-O-Fe . Th is chemical bond fo rmat ion produced a strong interact ion o f the sol-gel layer w i t h the M S surface in the in i t ia l stage [70] . Th is research can give an explanat ion as to the rat ionale o f the ind iv idual deposi t ion o f convent ional sol-gel process. I t is possible to f o r m weak bonds between the non-phosphate sol-gel coat ing and the steel substrate, but these weak bonds can degrade at elevated temperatures and thus y ie ld a decrease in the adhesion strength [70] . Therefore, the question o f h o w to contro l the substrate-surface condi t ions was another important issue in sol-gel coat ing research. On the who le , i t was bel ieved that there was not enough in fo rma t ion to promote a deep understanding o f the relat ionship between coat ing adhesion and sol-gel processing [112-120] . The standard o f adhesion is st i l l poor ly understood [121-128] . Therefore, i n order to understand the abovementioned problems to the necessary extent, further invest igat ion o f adhesion is required. The interfacial f racture toughness is a basic parameter o f the interface. Therefore, i t is a potent ial candidate for character izat ion o f the adhesion o f the sol-gel coat ing. 2.8 C h l o r i d e C o r r o s i o n The f o l l o w i n g electrochemical equations govern the corrosion o f i ron in a sodium chlor ide so lut ion [113 ] : 23 Fe^>Fe3+ +3e' (2.7) 02 + 2H20 + 4e~ -+40H~ (neutral/alkal ine) (2.8) 2H+ +2e~ - > # 2 t (acid) (2.9) The contact o f i ron w i t h water and oxygen or hydrogen fo rms a galvanic cel l leading to the ox ida t ion o f i ron. Since i ron possesses anodic and cathodic sites, e.g. Carbon in the constant electric contact, the only way to inh ib i t corros ion is to el iminate the contact between the electrolyte and the reactants [113] . The si loxane bond coating suppresses the cathodic react ion by l i m i t i n g the d i f fus ion o f the electrolyte, i.e. oxygen and water, to the substrate. I t a l so .b locks the transport o f electrons to the i ron surface. I n a sod ium chlor ide solut ion, the chlor ide is one o f the most deleterious factors in terms o f steel corrosion. The antagonistic nature o f the Cl" ions is due to their abi l i ty to absorb on the steel surface, where h igh current densities are generated at the Cl" adsorption site. Hydro lys is o f i ron ions f r o m the anodic react ion causes a decrease i n p H , w h i c h discourages oxide film repair and accelerates attacks. Even in h igh-pur i ty water, in wh ich the level o f Cl" is as l o w as a f e w mi l l ig rams per l i ter, the attack o f Cl" ions on m i l d steel results i n u n i f o r m rather than local ized corrosion [113] . 24 2.9 Corrosion Protection by Multi-Layer Coatings A ceramic coat ing can reduce corrosion o f a metal substrate, since the ceramic material is usual ly nobler than the metal. Composi t ions, interfaces, defects, thickness, and structures are the impor tant parameters af fect ing the corrosion resistance o f a coat ing. The chemical compos i t ion o f the coating and the microstructure o f the substrate exhibi ts a direct inf luence on the electrochemical behavior o f the system [129] . The corrosion rate is no rma l l y proport ional to the corrosion-current density [113] . The corrosion resistance o f the mul t i - layer coatings is s igni f icant ly d i f ferent f r o m that o f the single-layer coatings [130] . The mul t i - layer coatings have a m u c h smaller corrosion current density than the mono- layer coatings and the bare M S because i n the mul t i - layer system, the pinholes can be fu l l y b locked by corrosion products, e.g. i ron oxides, w h i c h prevents further transport o f oxygen to the steel substrate [131] . The protect ion ef f ic iency, P can be represented by : P(%) = 100(l-icor/i°cor) (2.10) where i°cor and icor denote corros ion current densities o f the bare and coated substrates respectively [126] . 25 Coat ing defects act as pathways that a l low the corrosive species to reach the interface [132 ] . The pores and micro-cracks prov ide a direct path fo r corrosive species and lead to rapid local ized galvanic corrosion o f the metal substrate [132, 133]. The corrosion current o f the sol-gel coatings decreases to ~10" 4 t imes that o f the m i l d steel substrates w i t h the presence o f homogeneous macro-pores [134] . The ceramic part icles incorporated into the methy lphenyls i loxane ( M P S ) hybr id coating enhance the corros ion resistance o f steel: the corrosion resistance increases w i t h increasing coat ing thickness, and acts as a physical barrier, w h i c h ef fect ive ly separates the anode f r o m the cathode [5, 126]. The corrosion-resistance o f the interface is increased w i t h the incorporat ion o f the po lymer component, a fact at t r ibuted to the fo rmat ion o f chemical bond ing at the interface [126] . The po lymer bond coat not on ly improves the adhesion, but also increases the corrosion resistance o f the system [136] . The corrosion protect ion o f the bond coat increased w i t h increasing coat ing thickness [137] . I n summary, i t is bel ieved that the mul t i - layer coat ing fo rms a better barrier against wet corros ion than the single-layer coating. The po lymer bond coat can increase the corrosion resistance o f the coat ing system, i n addi t ion to i m p r o v i n g bond strength and decreasing residual stresses i n the system, this supports the or ig ina l object ive o f this thesis. 26 CHAPTER 3: EXPERIMENTAL PROCEDURE 3.1 Substrate Preparation A I S I 1010 m i l d steel was used as a substrate. The substrates (3.8x3.8 cm) were sandblasted by 70-80 psi air pressure us ing 220 gr i t b r o w n fused-a lumina part icles, result ing i n a surface roughness o f 0.5-5 u m . The average surface roughness (Ra) was measured by a W y k o ® opt ical interference surface pro f i le r (Veeco Instruments Inc., U S A ) . 3.2 Deposition of Coatings The or ig inal a lumina sol was prepared by adding 306 g (1.5 m o i ) a l u m i n u m isopropoxide ( A l ( i O C H ( C H 3 ) 2 ) 3 , 9 8 % , A ld r i ch ) to 3 L hot (85°C) d is t i l led water. I M ni t r ic acid was used to adjust the p H o f the sol to 3. The mix tu re was stirred v igorous ly for 16 hours at 85°C. The excess solvent was s low ly evaporated f r o m the sol unt i l its molar i ty was 1.5 M . 42 g calc ined ct-alumina (0.3-0.5 u m , A 1 6 S G A l c o a Industr ia l Chemicals, U S A ) was added to 100 m L o f the or ig inal sol as the f i l ler . The sol was ba l l -m i l l ed for 24 hours to break apart part ic le aggregates and to ensure un i f o rm i t y o f the solut ion before the spraying o f the coating. 27 The sol was sprayed on the m i l d steel substrate, f o l l owed by the spraying o f mono a luminum phosphate ( M A P ) and methylphenyls i loxane ( M P S , Dureaseal 1529H Cotronics Corp. , U S A ) . The M A P and M P S were deposited by using t w o spray-guns i n a back-draft booth. Subsequently, the sol was deposited again f o l l owed by the impregnat ion o f 2 5 w t % mono a luminum phosphate ( M A P ) , cur ing at 121 °C for 4 hours, and f ina l ly bak ing at 300°C for 30 minutes. A two-gun method is used to distr ibute the M A P throughout the coating. I t consists o f one gun spraying the sol and another gun l igh t ly spraying 2 5 w t % M A P occasionally. I t is bel ieved that this is the only way to distr ibute phosphate throughout the C B - C S G coat ing w h e n the coat ing is th icker than 2 0 u m . C B - C S G 30-40 urn Siloxane 5-10 um Group A Substrate C B - C S G 30-40 u m Siloxane 5-10 u m C B - C S G 10-20 u m Substrate Group B C B - C S G 30-40 u m Substrate Group C Fig.3.1. Schematic representations of three basic multi-layer structures 28 Three variants o f this basic process were used to produce coat ing types A to C, as shown in F ig .3 .1 . Group A was the dual- layer system conta in ing the si loxane bond coat and the C B - C S G top coat on the M S substrate. Group B was a t r i - layer system, containing a C B - C S G pr imer, a si loxane bond coat and a C B - C S G top coat. Group C was a mono- layer C B - C S G coat ing on the M S substrate. The M S substrates used i n the mono-layer Group C were heated to 500°C to oxid ize their surfaces so that they cou ld be treated w i t h aqueous slurry af terward. The samples were kept i n a glass container w i t h a desiccant to prevent contaminat ion. 3.3 Low Temperature Process This process uses low-temperature cur ing to produce a C B - C S G a lumina coat ing free o f surface cracks, inc lud ing 1 day o f cur ing at 121°C and 5 days o f cur ing at 160°C. A f te r the 1 day o f cur ing at 121 °C , the M A P reacts w i t h hydrated a lumina and forms AIPO4. Chemica l bond ing occurs w h e n the AIPO4 is converted to Ber l in i te as the bonding agent after 5 days o f cur ing at 160°C. The residual phosphate was washed away w i t h w a r m (45°C) water. The f o l l o w i n g is the detailed procedure for this low-temperature process: 1. Heat at 550°C to f o r m a th in ox id ized f i l m on the sand-blasted m i l d steel. 2. Spray sol and 2 5 % M A P w i t h two guns to the desired thickness. 3. Cure at 1 6 0 ° C f o r 1 day. 29 4. Spray 5 0 % M A P onto the coat ing surface. 5. Cure at 1 6 0 ° C f o r 5 days. 6. Wash away the residual phosphate w i t h water at 45°C. This l o w temperature process includes a mu l t i - gun spraying technique to distribute the phosphate throughout the coat ing and prevent cracks. I t can produce the same coatings as the convent ional method (referred to pp.10). I n this l o w temperature process, mono a l u m i n u m phosphate reacts preferent ia l ly w i t h the smal l , reactive particles, e.g. y-hydrated a lumina, dur ing the 160°C cur ing, and provides a bond ing phase for the coarser f i l le r part icles, e.g. ct-alumina. This coat ing is gradual ly consol idated w i thout generation o f strains dur ing the 5 days o f cur ing at 160°C. 3.4 Scratch Tests The scratching cr i t ical force at w h i c h a coat ing fai ls was used to represent its scratch resistance. A l l measurements are per formed on a Romulus I V apparatus (Quad Group, U S A ) at a changing vert ical load o f 0-45 k g and a 2.0 c m distance. The coat ing moved at a constant rate o f t ravel (0.04 mm/s) and the force increased at a constant rate o f loading (0.1 kg/s) un t i l the coat ing fai led. A sensor near the scratch t ip detected the acoustic signals f r o m the coat ing fracture and conf i rmed the cr i t ica l force values. 30 I n this research, a hemispher ical d iamond t ip w i t h a radius o f 533 u m was used. The scratch tests were conducted on the f ine ly pol ished coat ing surfaces. The surface o f the coating was pol ished to a roughness o f less than 1 u m w i t h 2400 gr i t sandpaper. Scratch hardness (S/,) is def ined b y the f o l l o w i n g equat ion: Sh=^t (3-D where F^is the normal load and b is the scratch w i d t h [40 ] , w h i c h is the distance between the scratch ridges. is the projected load bearing area based on the assumption that 8 the coating behaves perfect ly p last ical ly and the load is carr ied on l y b y the f ront ha l f o f the spherical t ip . Four samples per group were used to determine the average cr i t ical force and the average scratch hardness. 3.5 Indentation Hardness Measurements The average hardness o f 30 samples was measured b y a micro-hardness tester (M ic romet 3, Tech-met , O N , Canada) using a Vickers indenter at a load o f 300 g and a durat ion o f 15 seconds. A nano- indentat ion system (Fisherscope H I 0 0 , Germany) was used to measure the elastic modulus o f the coat ing. N o r m a l load on the B e r k o v i c h indenter was ranged 31 f r o m 0.4 to 10 m N in 25 equal increments at a rate o f one per second, and then held at the peak load for another 50 seconds. The 5 m m th ick substrates were used to prevent the indenter f r o m af fect ing the deformat ion o f the coat ing system, because there was always a plastic zone under the indent f ie ld. The depth o f this plastic zone was - 1 . 5 t imes the indent diagonal- length. Each successive indent was isolated at a distance o f - 1 m m to avo id over lapping o f plastic zone onto ne ighbor ing indents. A R o c k w e l l C hardness tester (Buehler Inc., U S A ) was used to determine the coating's interfacial fracture toughness. Each sample (Group C) was cut into three coupons for measurement o f the interfacial fracture toughness. The R o c k w e l l C indenter was a spherical d iamond w i t h a 0.2 m m t ip radius. The test procedure was conducted according to A S T M standard E l 8 - 0 2 . A s the h igh- load indenter (150 k g f ) was appl ied on the coat ing and caused the plast ic de format ion o f the under ly ing substrate, the severe deformat ion o f the substrate forced the coat ing to be displaced radia l ly and induced a compressive radial stress in the coat ing w h i c h decreased w i t h the increase in distance f r o m the indenter. The cr i t ical crack extension force was determined by measur ing the size o f the detachment area [13, 40, 48 ] . The measured crack size was the average o f three indentations. The calculat ion o f interfacial fracture toughness was per formed according to the method o f Ref. [13 ] . 32 3.5 A d h e s i o n Tests The adhesion strength o f the coatings to the steel substrate was tested by a pu l l - o f f test, A S T M standard C-633-79. Five samples per group were tested to determine the average tensile strength. The coatings were glued to t w o ident ical rods by a 3 M 2214 regular epoxy adhesive. The adhesion measurements were made using an Inst ron universal testing machine ( Instron Corp. U S A ) w i t h 4,450 kg load cel l at a cross-head speed o f 1 mm/minu te . The m a x i m u m force at w h i c h the two rods were separated was determined (see Fig.3.2). A f t e r the test, the cross-sections o f the fractures were observed by S E M . F ig.3.2. Se t t ing o f the p u l l - o f f tests 33 3.6 Air Permeability Measurements The average air permeabi l i ty was measured by a " V a c u p e r m " permeabi l i ty tester (Univers i ty o f M issour i -Ro l la , U S A ) . The coatings were deposited onto the porous t i les, w i t h a permeabi l i ty o f ~5.0 mi l l idarcys. The average value o f f ive samples was taken as the air permeabi l i ty o f the coat ing. I n the permeabi l i ty tester, i t pumps the air into the chamber and then forces the air to permeate the coat ing. A sensor simultaneously detects the pressure i n the chamber and sends i t to a computer. The computer calculates the permeabi l i ty o f the samples by a model between the pressure and the t ime. 3.7 "Wafer" Curvature Radius Measurements 200-300 u m th ick 316 stainless steel was used as the substrate. A l l the samples were cooled d o w n f r o m the 300°C f i r i ng before the radius o f the surface curvature was measured. The f i rst group o f samples (Group C) inc luded bare substrate, l O u m , 20 u m , 3 0 u m , 4 0 u m , 60 u m , 80 u m , and 100 u m th ick C B - C S G coatings respect ively. Another six Group A coatings contained a 30 u m C B - C S G top coat and di f ferent thickness values for the si loxane pr imer , e.g. 0 u m , 5 u m , 8 u m , 10 u m , 15 u m , and 20 u m respectively, 34 and each sample was measured 4 t imes at di f ferent locations. There was approximately ±2 u m deviat ion o f the thickness measurement for each sample. A n inter ferometr ic surface imaging system ( W Y K O N T - 2 0 0 0 , Veeco Instruments Inc. U S A ) was used to measure the radius o f curvature o f the bare substrates and the coatings. Each measurement was focused on a 3.7x4.8 m m area. The f ina l curvature radius R„ o f each area was the average o f absolute Rx and Ry\ therefore, the radius o f curvature o f each sample was calculated by, r = -I.\Rn\ (3.1) "„=y The radius o f curvature o f each sample was taken as the average value o f four measurements. The highest and lowest values were not included. 3.8 Microscopy and XRD The preparat ion o f the samples used for S E M observations compr ised sectioning w i t h a d iamond saw, vacuum mount ing w i t h a low-v iscos i ty resin ( Industr ia l Formulators, Inc. Canada), f o l l owed by pol ish ing w i t h 50-1200 gr i t sandpaper, 5, 1, and 0.5 u m d iamond slurries, and si lk c loth on the grooved metal platens. The samples were pol ished to a flatness <0.1 u m w i t h nano-size col lo idal si l ica. 35 For the fracture observat ion samples, the coupons were cut ver t ica l ly f r o m the substrate towards the coat ing w i t h a high-speed saw to obta in a V - g r o o v e close to the coating.. The sample was placed in l i qu id n i t rogen for approx imate ly 1 minute , and then broken immediate ly after being taken out. 600 and 1,200 gr i t sandpaper and 1 urn d iamond paste were used to po l ish the faces normal to the fracture plane o f samples. The m o v i n g d i rect ion o f the po l ish ing wheel was f r o m coat ing to substrate, to prevent cracks f r o m propagat ing due to the external load wh i le po l ish ing. The epoxy moun t ing also could have protected the cracks f r o m deformat ion to some extent. The samples for the crack observations were prepared in the same way. The specimens were carbon coated ( - 2 0 angstrom) for B S E (backscattering electron) detect ion o f the cross-sectional morpho logy coupled w i t h E D X (Energy Dispersed X - ray ) . The cross-sections o f the si loxane layer sputtered by go ld were detected at a h igh magn i f i ca t ion by SE (secondary electron) w i t h a H i tach i S-2300 S E M instrument at a w o r k i n g vol tage o f - 2 0 . 0 k V . A coat ing thickness gauge (Positector 6000, DeFelsko Corp. , N Y , U S A ) was used to measure the thickness o f the sol-gel coatings. A l l the coat ing thicknesses were conf i rmed by the measurements f r o m the S E M images o f their cross-sections. X - ray d i f f rac t ion ( X R D ) was per formed on an X - r a y generator p l a t f o r m (Phi l ips P W 1830). The surface morpho logy o f the coatings was observed by a N i k o n E P I P H O T 300 opt ical microscope. The a lumina vo lume percentages o f f i ve samples per group were 36 assessed by an image analyzer (Clemex version PE 3.5, C lemex Technologies Inc., Canada). 3.9 Potentiodynamic Evaluations of the Coatings A l l Potent iodynamic evaluations were carried out by a Solart ron 1286 potentiostat (Solartron Group Companies, U K ) using the W i n d o w s ™ X P operat ing system in conjunct ion w i t h a D e l l ™ P H I PC. The Potent iodynamic evaluations were conducted as fo l lows: • A piece o f copper w i re was welded onto the back surface o f the substrate. • The exposed surface o f the copper w i re and the edges o f the samples were stamped w i t h epoxy. • The cut edges and epoxy crevices were sealed w i t h a wa te rp roo f T E F L O N ® tape and an acetone d i lu ted s top-of f lacquer ( 5 0 % Micros top, Pyramid Plastics, Inc) , w h i c h was a insulat ing and wate rp roo f paint, i n order to prevent galvanic actions a long the edges and defects between the epoxy and the coatings. • The sample was f i xed i n a quartz glass tube whose end was sealed by waterp roo f tape. A 0.7x0.7 m m section w i t h i n the center o f each sample was exposed to so lut ion dur ing the testing and faced w i t h the lugg in capi l lary as close as possible. • N i t rogen (+10 psi) was connected to dr ive o f f the oxygen i n the water, and the electrolyte was magnet ica l ly stirred wh i le the experiment was runn ing . 37 C H A P T E R 4: R E S U L T S A N D D I S C U S S I O N 4.1 Coating Structures 4.1.1 G r o u p A S t r u c t u r e Figure 4.1.1 shows a typical dual-layer microstructure of the Group A coating. The coating is comprised of a siloxane bond coat and a CB-CSG top coat, where the CB-CSG top-layer is well anchored to the siloxane bond coat. Fig.4.1.1 S E M image o f d u a l - l a y e r c o a t i n g The OC-AI2O3 particles or the CSG aggregates penetrated into the siloxane primer and acted as reinforcing phases. These reinforcing phases form mechanical interlocks between the siloxane and the CSG. In Fig.4.1.2, there was a siloxane/A^Os composite area in the siloxane bond coat near the top coat. Some of the CB-CSG aggregates even 38 penetrated into the bo t tom o f the siloxane bond coat, f o r m i n g a CB-CSG/s i loxane composite pr imer. Ac tua l l y , the siloxane f i l m is i n a discont inuous state at some locations. This m igh t be w h y the Group A coat ing had a higher adhesion strength ( -10.0 MPa) than that o f the 100% si loxane (4-7.0 MPa) . x £ . Q k Q 0 8 8 2 8 k V 28j-'m Fig.4.1.2 Ceramic/polymer composite layer between bond-coat and top-coat. Since the poly(methy lpheny ls i loxane) (PMPS) does not adhere w e l l to ceramics, it was necessary to simultaneously spray the monomer M P S and the C S G sol before cur ing them together. Th is process a l lows the a lumina part icles to penetrate into the siloxane mat r ix and to f o r m good anchor ing points. 39 Therefore, the si loxane composite layer could be considered as a funct ional ly graded bond coat w i t h a graded composi t ion f r o m top coat to bond coat, reducing the thermal expansion mismatch among the dif ferent coat ing layers. 4.1.2 Group B Structure Figure 4.1.3 is a S E M picture o f the tr i - layer microstructure o f the Group B coating, consist ing o f a ~5 u m siloxane layer interleaved w i t h a 30-40 p m C B - C S G top coat and a 10-20 p m C B - C S G pr imer, result ing i n a sandwich structure coating. Figure 4.1.4 is a higher magn i f i ca t ion o f Fig.4.1.3; i n some places, the siloxane bond coat presents a discont inuous si loxane film. The purpose o f developing the tr i - layer Group B coating was to increase the strength o f the bond coat. Substrate x 8 8 G 0 0 0 8 2 8 H V 5 0 . u r n Fig.4.1.3. CB-CSG/siloxane /CB-CSG tri-layer structural coating (Group B) 40 1 Fig.4.1.4 Tri-layer Coating structure The s t i f f C B - C S G layer is to absorb external loads and protect the soft si loxane under-layer i n service. The si loxane under-layer i n turn prevents the spread o f cracks into other layers and relaxes the stress to the substrate. Some cracks or iginate at the coating /substrate interface and develop through the coat ing, hur t ing its adhesion properties. Therefore, the si loxane bond-coat can also act as a crack inh ib i tor , thereby possibly increasing the fracture resistance o f the coating. 41 Substrate x l . Q k 00 0kV 50JLim Fig.4.1.5 Cross section image of mono layer C B - C S G coating • • 1 i L*« • » .p^ - . j " • \ . • •* " i J '"••*«" ^i Fig.4.1.6 The phosphorus E D X mapping of Fig.4.1.5 4.1.3 G roup C Structures Figure 4.1.5. shows a mono-layer CB-CSG coating on an oxidized mild steel substrate. There is no crack observed through the cross-section. Figure 4.1.6 is a 42 phosphorus mapp ing by E D X , and shows evidence o f the u n i f o r m phosphate d is t r ibut ion throughout the coat ing achieved v ia the two-gun preparat ion method. I n most cases, such phosphorus mappings by E D X show that the phosphorus concentrat ion gradual ly increases f r o m the interface to the surface. The surface has the highest concentrat ion o f phosphate, and the interface has the lowest concentrat ion. The f inal step o f the process has t radi t ional ly been to deposit a large vo lume o f phosphate on the coating surface, and this is the reason w h y the surface usual ly has the highest concentrat ion o f phosphorus. Th is phenomenon also shows that the phosphate cannot penetrate through the th ick C B - C S G layer, and thus that on ly t w o - g u n spraying can distribute phosphate evenly throughout the coat ing. 43 4.2 Cracks and A i r Permeability 4.2.1 Cracks in Coatings The conventional composite sol-gel method involves multi-step spraying and individually firing at 550-600°C, which often generates cracks, e.g. shrinkage cracks, phosphate cracks and thermal expansion-mismatch cracks. Fig.4.2.1 Drying-shrinkage cracks 44 D r y i n g shrinkage of ten induces cracks when the water evaporat ion occurs too quickly. The shrinkage cracks are of ten o f small size and inv is ib le to the eye (F ig. 4.2.1), and can be easily obscured by po l ish ing w i t h 1200 grit sandpaper. Freeze-drying is also a viable way to decrease the shrinkage strain and avoid shrinkage cracks. Fig.4.2.2 W e b - l i k e phosphate c racks To increase the hardness and adhesion o f the sol-gel coat ing, i t is necessary to impregnate enough phosphate. However , brushing o f 8 5 % phosphoric acid onto the green-body o f composite sol-gel a lumina of ten generates the phosphate cracks even at room temperature. Figure 4.2.2 is an example o f "phosphate cracks". The phosphate cracks often appear i n a web- l i ke shape on the coating surface. 45 Essential ly, phosphate cracks are generated by the chemical reactions o f the phosphoric acid and the hydrate alumina. When the phosphor ic acid ( P A ) is over-impregnated on the surface o f green body, i t forms pseudoboehmite or gelatinous boehmite, w h i c h has a l o w strength (referred to pp.10). The gelatinous boehmite reacts w i t h the residual phosphoric acid ( P A ) , as shown in equation (4.2.1). H3PO, + AlOOH => AlPO, i +2H20 (4.2.1) Fig.4.2.3 T h e c racks due to the m i s m a t c h o f C T E 46 This react ion produces sol id a luminum phosphate at r o o m temperature, wh ich often induces strains i n the green network. W h e n the stresses accompany ing the strains exceed the strength o f the green network, they w i l l induce phosphate cracks. Usage o f mono a luminum phosphate ( M A P ) can decrease the probabi l i ty o f p roduc ing phosphate cracks by means o f retarding the chemical reactions at r o o m temperature, since M A P is less chemical ly active than PA. W h e n the green sol-gel is deposited w i t h 8 5 % phosphor ic acid, the phosphoric acid w i l l react w i t h a lumina to f o r m M A P . When fired at 550-600°C, the M A P w i l l qu ick ly convert to Ber l in i te , Cr istobal i te or other phases, w h i c h causes vo lume changes over a short t ime. W h e n the coat ing is cooled down , the di f ferent ia l thermal expansion o f the di f ferent phases w i l l generate mismatch cracks. Fig.4.2.4 The surface crack-free coating after 4 days of curing at 160°C 47 On the other hand, the di f ferent ia l thermal expansion o f the sol-gel coat ing and the metal l ic substrate also induces strains i n the convent ional sol-gel process, w h i c h lead to large cracks v is ib le to the eye. Figure 4.2.3 is a heavi ly phosphatized coat ing. The coating surface induces cracks due to the brushing o f 8 5 % P A and C T E mismatch. This k ind o f cracking is the biggest challenge in the tradi t ional chemica l ly bonded sol-gel process. The C T E mismatch cracks are very d i f f i cu l t to prevent, since the thermal expansion differences are inevi table w i t h changing temperature. Fig.4.2.5 The high magnification of surface crack-free coating 48 One viable w a y to prevent the phosphate cracks and the C T E - m i s m a t c h cracks is to decrease the processing temperature. Cur ing at 160°C for 5 days a l lows some stress relaxat ion and produces crack-free coatings. Th is l o w temperature cur ing a l lows the M A P to gradual ly t rans form to Ber l in i te w i thou t induc ing large strains. A f te r the phase t ransformat ion is completed, the coating is almost stress-free. Th is was evidenced by " W a f e r " experiment, w h i c h w i l l be discussed in chapter 4.5. A f te r long- te rm cur ing, the coating obtains a strength that a l lows i t to be f i red again at a h igher temperature (>300°C) and cooled d o w n w i thou t generat ion o f cracks. Fig.4.2.4 shows a surface crack-free coating after 5 days o f cur ing at 160°C. There are no cracks even at the corners and the edges. Fig.4.2.5 is a higher magni f ica t ion o f the surface crack-free coat ing by the l o w temperature process. I t is reasonable that l o w temperature (160°C) cur ing reduces the possibi l i ty o f crack ing due to di f ferent ia l thermal expansion. 4.2.2 Air-Permeability Table.4.2.1 shows that the C B - C S G coating is air permeable at a thickness o f - 4 0 urn. The air permeabi l i ty o f t radi t ional C B - C S G coatings is h igher than that o f l o w temperature C B - C S G coatings, i m p l y i n g that the l o w temperature cur ing produces fewer defects, e.g. pores, cracks and fissures, than the t radi t ional sol-gel method. The air permeabi l i ty after cur ing at 160°C is - 0 . 2 3 mi l l idarcy , ind icat ing that the sol-gel coat ing is relat ively dense. 49 Table 4.2.1 the air permeability of various structures Tradi t ional N e w A i r Permeabi l i ty (m i l ida rcy ) 300-500°C f i r i ng 160°C cur ing Dual- layer Group A N M Non-detectable Tr i - layer Group B N M Non-detectable Mono- layer Group C (40 p m ) 0.45 (0.08) 0.23 (0.03) *Values in parentheses show the standard deviation * N M : not measured General ly, the air permeabi l i ty o f C B - C S G coatings decreases w i t h increasing thickness. The ai r -permeabi l i ty o f the si loxane incorporated C B - C S G mul t i - layer coat ing is non-detectable (Table 4.2.1). Plausibly, the hermetic monomer sealed the voids o f the gel-network dur ing spraying and prevented the substrate f r o m contact ing corrosive species such as oxygen, and the po lymer islands decreased the number o f C B - C S G channels after the po lymer iza t ion o f methylphenyls i loxane ( M P S ) . W h e n these islands connect together and f o r m a cont inuous layer, i t is d i f f i cu l t for the oxygen to contact the substrate i n the mul t i - layer coat ing system. The corrosion protect ion o f the mul t i - layer coating w i l l be discussed in chapter 4.7. 50 4.3 Contact Measurements 4.3.1 H a r d n e s s o f C B - C S G C e r a m i c The average hardness o f the C B - C S G surface was measured as 6.24 GPa, close to the hardness o f Ber l in i te . Th is could be explained by the processing, where, after the f ina l coating step (spraying o f 5 0 w t % M A P onto the C B - C S G surface), the m o n o a lum inum phosphate ( M A P ) f o r m e d a chemical bond w i t h the C S G and gradual ly t ransformed to Ber l in i te or Cristobal i te dur ing cur ing at 160°C for 5 days (referred to pp.14) . Figure 4.3.1 is an indentat ion trace on the top surface o f a Group A coat ing, the hardness o f w h i c h was measured as 6.05 GPa. S imi la r ly , the hardness o f the top surface o f the Group C coat ing (Fig.4.3.2) was 6.26 GPa. The C B - C S G coat ing (Group C) is bel ieved to be a Ber l in i te-bonded a lumina network ; however , there are some micro-pores, channels for the evaporat ion o f water, on the surface o f the C B - C S G coat ing, causing the measured hardness to be lower than that o f Ber l in i te. The hardness o f Ber l in i te is - 6 . 5 GPa, and the average hardness o f the Group C coating is - 6 . 2 4 GPa, homogeneously distr ibuted across the coat ing surface. Therefore, i t is bel ieved that the Ber l in i te AIPO4 is distr ibuted homogeneously across the surface o f the C B - C S G coat ing. I t was accepted that on ly the Ber l in i te was fo rmed after the M A P 51 was cured at ~ 1 5 0 ° C fo r 4 days, where the mono a lum inum phosphate hydrate reacted w i t h a lumina at ~ 1 5 0 ° C and formed Ber l in i te by the f o l l o w i n g react ion [34 ] : Al20, + 2AlHi(POA)2 • H20 -> AAIPO, + 4H20 (4.3.1) Fig.4.3.1. SEM indentation trace of CB-CSG surface, Vickers indent 300g load Ber l in i te is the phase that bonds ind iv idual a lumina part icles and forms the ber l in i te-bonded a lumina ceramic. However , i n this research, a wel l -c rys ta l l i zed AIPO4 network was not found by X R D . I t is bel ieved that on ly the vi t reous Ber l in i te (hardness 6.5 GPa) was fo rmed o n the surface and inside the coat ing [65] . The cross-sectional hardness measurements were conducted o n samples mounted w i t h i n a 25 .0mm diameter cast epoxy-resin b lock, w h i c h could prevent the error caused 52 by the coat ing thickness. Add i t i ona l l y , the 40-100 urn C B - C S G coat ing decreased the inf luence o f coat ing thickness to a tolerable extent, since, i n th in f i lms , the cross-sectional hardness was affected by the substrate. Figure 4.3.3 shows an image o f the mono- layer Group C coat ing. A hardness o f 6.0 GPa was measured at the interface, p rov id ing further evidence o f phosphate d is t r ibut ion to the interface by the m u l t i - g u n spray technique. F rom observation o f Fig.4.3.3, the Ber l in i te phase fo rmed at the interface wi thout generating a crack along the interface. I n Fig.4.3.4, the indentat ion was conducted r ight on the interface between coat ing and substrate, but crack onset d id not occur at the apexes o f the indent contour, demonstrat ing the h igh adhesion strength o f the C B - C S G coating. However , as seen in Fig.4.3.5, there may have been an interfacial crack induced along the interface, w h i c h was obscured by the elastic si loxane. The hardness o f siloxane was measured at 0.12 GPa. Fig.4.3.2 S E M i n d e n t a t i o n t race on C B - C S G sur face , 300g l oad 53 For some samples, the cross-sectional hardness o f the C B - C S G coating was between 3.0-6.0GPa, i.e. between that o f Varisci te and Ber l in i te , ind icat ing that the bonding agent could be a mix tu re o f the two . Occasional ly, however , some o f the 3.0GPa values could have been related to the heterogeneity o f micro-structure, i.e. porosi ty or agglomerat ion o f y -a lumina or Var isc i te, even though the 45°C water washes away most o f the Varisci te and hydrated y-a lumina on the surface. I t is bel ieved that long- term cur ing w o u l d eventual ly convert the rest o f the Var isci te to Ber l in i te . Fig.4.3.3. The SEM indentation trace of CB-CSG ceramic coating, Vickers indent 300 g load, Group C I n some extreme results, a surface hardness o f - 1 0 . 0 GPa was observed, w h i c h migh t have been caused by the hard aggregation o f alpha a lumina and strong l igands o f chemical bonds, or possib ly by wel l -crysta l l ized Ber l in i te or Cristobal i te (referred to 54 pp.14). I t is possible to f o r m a network w i t h the combinat ion o f Ber l in i te binders and cc-alumina f i l lers, w h i c h have an average hardness over 10 GPa. Therefore, the relationship between the contro l o f impregnat ing chemicals and the hardness o f the coating needs further research. 4.3.4 The SEM indentation trace of CB-CSG ceramic coating, Vickers indent 300 g load, Group C 4.3.5 Indentation trace of substrate-polymer-coating cross section, indented at 300 g load, Group A 55 4.3.2 Interfacial Fracture Toughness The interfacial fracture toughness is a basic mater ial parameter o f the interface. The higher the inter facial fracture toughness, the higher the adhesion strength o f the coating. Therefore, i t is bel ieved that the interfacial fracture toughness is a reasonable indicator o f coat ing adhesion and that its measurement is a good w a y to quant i tat ively assess the interface. W h e n a h igh- load R o c k w e l l C indenter is appl ied to a coat ing surface, the coat ing experiences interfacial cracks between coat ing and substrate r=a. The interfacial cracks spread radia l ly toward a cr i t ica l po in t where the energy release rate was unable to propagate the cracks. The value for interfacial fracture toughness was ident i f ied w i t h the value o f energy release at w h i c h the cracks arrested, as schematical ly depicted i n Fig.4.3.6, showing a mode l o f the detachment by a h igh load ax isymmetr ic indenter. The detachment occurs as fo l l ows : (1) the interfacial crack f ront is forced to advance and to break up the coat ing by displacement o f the substrate; (2) the detachment occurs w i t h an unbuckled annular plate o f coat ing, w h i c h remains intact; (3) the annular plate o f coat ing buckles. 56 Fig.4.3.6 Model for the identification of interface fracture toughness The R o c k w e l l C indentat ion could induce some radial cracks i n the coat ing that were not caused by displacement o f the substrate. Opt ica l image analysis avoids the obscuring o f micro-cracks due to go ld sputtering or carbon coat ing for S E M sample preparation, and therefore an opt ical microscope was used instead o f S E M . Figure 4.3.7 57 depicts the trai t o f the interface between C B - C S G coat ing and substrate where the radial cracks resulted f r o m the tensile hoop stresses when the coat ing was displaced radial ly. The detached coat ing adjacent to the indentat ion was carefu l ly discerned through the use o f image analysis software (Clemex version P E ™ 3.5, C lemex Technologies Inc., Canada), w i t h w h i c h the contact radius (a) and the radius o f the indentat ion- induced annular cracks (R,) were measured, as shown in Fig.4.3.7 and Fig.4.3.8. Fig.4.3.7. Optical images of typical coating detachment after Rockwell C indentation Fig.4.3.8 Detachment area of Fig.4.3.7 by image analysis 58 Since the delaminat ion is due to mode I I (shear) interface crack ing on ly , the interfacial fracture toughness is presented by Knc. The models and calculat ion o f interfacial fracture toughness are taken f r o m the l i terature [13 ] . 4.5 re Q_ E 4 3.5 H 3 JLZ O) 2.5 -3 O 1-<D 2 -k. 3 O 1.5 -re LL 1 -Q) O re terf 0.5 -0 -2 3 4 5 Substrate Roughness (micron) Fig.4.3.9. A plot of substrate surface roughness and interfacial fracture toughness Figure 4.3.9 shows a p lo t o f interfacial surface roughness and interfacial fracture toughness, where the inter facial fracture toughness increased w i t h the surface roughness o f the substrate; the rougher the substrate surface, the higher the adhesion strength. However , after the surface roughness reached a value higher than 4 p m , the interfacial fracture toughness tended to become stable at 3.5-4.0 M P a . m 1 / 2 . Therefore, the o p t i m u m surface roughness o f the substrate should be -4 .0 -5 .0 p m , at w h i c h the coat ing possessed a good capabi l i ty to wi ths tand shatter or to absorb impact energy. The sand-blasted steel substrate resulted i n a higher interfacial fracture toughness than the gr inded steel 59 substrate [13] . However , the mode l o f the interfacial fracture toughness was based upon the " fu l ly-e last ic m o d e l " [13] . The drawback o f using opt ical image analysis to measure the radius o f annular cracks is that i t depends on the experience and ski l ls o f the operator. As seen i n Fig.4.3.9, typ ica l data scatter for the K n C is 0 . 5 - 0 . 7 M P a . m 1 / 2 or up to 3 0 % o f the average. 4.3.3 Nano Indentation and Sonic Wave Modulus Table 4.3.1 lists the average hardness and elastic modulus o f the C B - C S G top coat and the si loxane bond coat, measured by a depth-sensing indentat ion. The hardness and elastic modulus o f si loxane were m u c h lower than those o f the C B - C S G ceramic. These values w i l l be used to determine residual stresses i n a later chapter. The Poisson's ratio o f C B - C S G was 0.25 as measured by sonic waves (see Append ix I ) . F rom Table 4 .3 .1 , the dynamic Young ' s modulus o f the coat ing as measured by acoustic method (sonic waves) was - 2 2 5 GPa, w h i c h seemed remarkably d i f ferent f r o m that measured by the nano-indenter. I t is probable that the acoustic method neglected the presence o f defects, e.g. cracks and pores, and was averaged v ia the v ibrat ional frequency o f the coatings conta in ing 8 0 w t % a-a lumina (elastic modulus: - 3 7 0 GPa). O n the other hand, the surface defects, e.g. open-pores and surface fissures, affected the values o f the indentat ion method, so the values measured by acoustic methods were of ten higher than those measured by indentat ion. 60 Table.4.3.1. The average values of hardness and elastic modulus measured by nano-indenter Hardness (GPa) Elastic M o d u l u s - ^ 7 (GPa) 1-V Elastic Modu lus by Sonic Wave (GPa) Siloxane B o n d Coat 0 .12(0 .05) 5.0 (0.4) N M C B - C S G Top Coat 7.65 (2.7) 122.0 (30.0) 225.45 (40.0) *Values in parentheses show the standard deviation * N M : not measured 61 4.4 Scratch Resistance 4.4.1 Critical Scratching Force for Coatings A s measured by paral lel scratching, the average cr i t ica l force values fo r Groups A , B, and C were 4.7 k g , 10.3 k g and 17.0 k g respectively (Table 4.4.1). These values depend on the vo lume percentage o f reinforcement i n the si loxane, referred to chapter 4.6. Table 4.4.1 Scratch resistance of various coating structures Groups Structure Critical Force (kg) Group A ~~ Dual- layer 4.7 (0.5) Group B Tr i - layer 10 .3 (1 .0 ) Group C Mono- layer 17 .0 (1 .4 ) * Values in parentheses show the standard deviation The image analysis showed that the percentage o f re inforcement i n the siloxane was ~ 1 0 v o l % and ~ 2 5 v o l % in dual- layer Group A and t r i - layer Group B respectively. The cr i t ical force increased w i t h the increment o f coat ing thickness. Tr i - layer Group B had a C B - C S G pr imer - 5 - 1 0 p m thicker than that o f dual- layer Group A . Therefore, t r i -layer Group B had a higher cr i t ical force than dual- layer Group A . A l t h o u g h mono- layer Group C d id not have the po lymer constituent referred to Fig.4.4.1 (a), i t possessed the highest scratch resistance. 62 As shown in Fig.4.4.1 (b), the distance traveled by the stylus under an increasing load (0-45 kg) at the onset of coating failure was -0.82 cm. This result also was verified by the assessment of effective friction, which was defined as the transverse force (Ft, diamond drag force) divided by the normal force (Fn). r Transverse t n n F f = x 100 = — x 100 (no unit) Normal F. (4.4.1) -10 (a) Hormal 10 12 Distance (mm) (b) Fig.4.4.1. (a) Optical micrograph of a scratch groove ( at 100X); (b) curves of normal force, acoustics, effective friction and distance; 63 A t a magn i f i ca t ion o f 100X, the scratching artifacts o f the coat ing were clearly discerned. A n abrupt change i n the effect ive f r i c t ion , i n Fig.4.4.1 (b) , signaled the coat ing fai lure, w h i c h is thought to be caused by one o f the f o l l o w i n g cr i t ica l events: (1) the C B - C S G is torn f r o m the substrate; ( 2 ) the substrate fai ls and the d iamond dislodges substrate mater ia l ; ( 3 ) the adhesive strength o f the coat ing is exceeded. I n the case shown in Fig.4.4.1 (b), the acoustic signal record ing indicates that the surface o f the C B - C S G coat ing cracked in i t ia l l y at 0.18 c m and the cr i t i ca l force at the onset o f C B - C S G surface spal lat ion was - 2 . 5 kg . 20 40 60 80 Coating Thickness (micron) 100 120 Fig.4.4.2. The scratch critical force vs. coating thickness (for mono-layer Group C CB-CSG coating) 64 The scratch cr i t ica l force at w h i c h the coating fai ls increases w i t h an increase in coating thickness: the th icker the coat ing, the higher the cr i t ical force. F igure 4.4.2 shows the cr i t ical force vs. the coat ing thickness. W h e n coat ing thickness was be low 60 u m , the cr i t ical force tended to increase w i t h the coating thickness. However , the average residual stresses in the C B - C S G coat ing also increased w i t h increasing coat ing thickness (e.g. compare Fig.4.5.1 i n Chapter 4.5). Therefore, after a certain thickness, the cr i t ica l force begins to decrease w i t h increasing coat ing thickness. W h e n the coat ing is be ing scratched, the stresses inc lude the residual stress f r o m processing and the bending stress f r o m scratching. F r o m Fig . 4.4.2, i t is apparent that: (1) when the thickness o f the Group C coatings was be low 20 p m , the stresses could be considered to be negl ig ib le. The cr i t ical force increased almost l inear ly w i t h increasing coating thickness; (2) i n the 20-60 p m thickness range, the effects o f the stresses could not be ignored, the rate o f change o f the cr i t ical force w i t h thickness decreased due to the stresses; (3) after 60 p m , the cr i t ical force decreased w i t h fur ther increase i n coat ing thickness because o f the residual stresses. However , i f the residual stress were relaxed, the cr i t ical force w o u l d l i ke ly continue to increase w i t h increasing o f the coating thickness. Stress re laxat ion w i l l be discussed i n chapter 4.5. General ly, the C B - C S G mono- layer coatings had good scratch resistance when their thickness was b e l o w 100 p m . 65 4.4.2 Scratch Hardness of Coatings There are t w o types o f scratch fai lure: cohesive fai lure and interfacial fai lure. The f o r m o f cohesive fai lure is usual ly part ial cone cracking or con formal cracking. Partial cone cracking of ten occurs around the perimeter o f the t ra i l ing edge o f the scratch t ip. Conformal c rack ing fo rms in f ront o f the m o v i n g indenter at the leading edge rather than at the t ra i l ing edge. Figure 4.4.1 (a) shows the onset o f some conformal cracking at the leading edge o f the scratch. This k i n d o f cracking is of ten due to the cohesive properties o f the coat ing network . Fig.4.4.3. Optical micro-graph of scratching made on CB-CSG coating, 100X, scratch moving direction from right to left 66 Interfacial fa i lure of ten presents as a ch ipp ing o f the coat ing i n f ront o f the scratch t ip , w h i c h is related to fai lure at the interface. Th is type o f fa i lure develops w i t h increasing scratch load. N o t al l o f the chipping is direct ly correlated w i t h the interface; some failures related to the pores or micro-cracks can also be considered to be interfacial . Interfacial fai lures of ten manifest as a r ing crack i n the coat ing and lead to chipping, w h i c h is essential ly the k i n k i n g o f a delaminat ion crack back in to and through the coat ing, as schematical ly shown in Fig.4.4.4. General ly, inter facial fa i lure is the l i ke ly cause when a ch ipp ing event is observed. Partial Cone Cracking —^ J Cohesive Failure Conformal Cracking Chipping Interfacial failure Fig.4.4.4. schematic of cohesive failure and interfacial failure For example, in-s i tu observation o f Fig.4.4.3, w h i c h was a C B - C S G coating produced by the convent ional method, indicated that the obvious ch ipp ing event dominated the scratching procedure, and no cone cracks or conformal cracks appeared. 67 I n consequence o f this observation, the defects at the interface, e.g. fissures and f laws, could not be considered negl ig ible. Interfacial fa i lure dominated the who le scratching procedure, even under a l o w load. I n other words, no cone cracks were found at the t ra i l ing edge or at the scratching head. Therefore, interfacial fa i lure is the dominant mode. Interfacial fai lures are due to the defects, e.g. pores and cracks, w h i c h are of ten generated dur ing h igh temperature processing. For example, the h igher the process temperature, the faster the water evaporat ion and the greater the porosi ty . Therefore, l o w temperature cur ing is an effect ive way to prevent interfacial fa i lure. The scratch hardness was independent o f the surface roughness o f the coating. Literature [13] shows the l inear trend o f the scratch-hardness values across the specimen obtained f r o m the exper imental data. The scratch hardness is between 5.5-6.0 GPa, and i t is a l i t t le lower than that measured by V ickers indentat ion because the scratch hardness measurement includes the effects o f porosi ty. Even though there was some hardness var ia t ion across the coat ing, i t st i l l reflected the homogeneous d is t r ibut ion o f hardness over the coat ing surface and thus the homogeneous d is t r ibut ion o f the chemical bonds. 68 4.5 Thermal Residual Stresses The steel substrate is always subject to residual compression i n a ceramic coating/metal system f o l l o w i n g coo l -down f r o m a h igh temperature, and as a result suffers deformat ion. One purpose o f using the si loxane bond coat ( B C ) is to decrease the effect o f the residual stress to the substrate. I n this chapter, the residual stress was estimated by a " w a f e r " curvature method. This approach revealed whether or not the siloxane bond coat cou ld relax the residual stresses due to visco-elast ic deformat ion o f the B C . I t is hypothesized that the incorporat ion o f si loxane relaxed the interfacial di f ferent ial thermal contract ion stress dur ing cool ing f r o m process temperature through the visco-elastic de format ion o f the siloxane. When a coating/substrate composi te is at an elevated temperature and cooled d o w n to room temperature, the mismatch o f thermal expansion coeff ic ient between the coat ing and the substrate results i n the residual stress. The direct ion o f warp ing is always the same: the coat ing warps towards the substrate, w h i c h indicates that the thermal expansion coeff ic ient o f the C B - C S G coat ing is lower than that o f the steel substrate. The plots i n F ig . 4.5.1 show that the bending moment increased w i t h increasing coating thickness. The or ig ina l data are l isted in the Append ix I I . I n this result, the 69 curvature radius decreased as.coat ing thickness increased, ind ica t ing that the average stress increased w i t h increasing coat ing thickness. 20 40 60 80 CB-CSG Thickness (micron) 100 120 Fig.4.5.1 The CB-CSG thickness vs. the radius of curvature Figure 4.5.2 is a typ ica l interferometr ic result for the 60 p m th ick Group C coating w i thou t si loxane bond coat, where the average curvature radius is 1.35 m. Figure 4.5.3 is the result for the bare substrate, where the measured curvature radius is 17.24 m. Figure 4.5.4 is the result for Group A , i.e. for the 60 p m thickness top coat and the 5-10 p m siloxane B C . The average curvature radius i n Fig.4.5.4 is 14.47 m, m u c h larger than that o f a C B - C S G coat ing o f the same thickness. 70 a o a. X (mm) X (mm) Fig.4.5.2. A typical interferometric result from a Group C coating (r=1.35m) 71 Fig.4.5.3 A typical interferometric result from a bare substrate (r=17.24m) Consider ing that the C B - C S G Poisson's ratio is 0.25 and that the Young ' s modulus o f si loxane as measured by a depth-sensing indentat ion was 5.0 GPa, the calculated average residual stress for the Group C coating i n Fig.4.5.2 is calculated using 72 equation (2.5) to be -8.0 MPa, and that for the Group A coating in Fig.4.5.4 is -0.30 MPa. a o E >> -10 w 3.59 X ( m m ) X ( m m ) Fig.4.5.4 30pm CB-CSG coating, containing 5-10pm siloxane bond coat (r=14.47m) The residual stress for the Group A coating system was much lower than that for the Group C system, indicating that the incorporation of siloxane substantially decreased the thermally induced stress. The siloxane bond coat significantly decreases the thermal 73 stress between the CB-CSG and the substrate. The siloxane BC acts as a compliant layer, which relaxes the stress caused by the thermal expansion coefficient mismatch. The siloxane bond coat thus significantly decreases deformation of the substrate. 16 -4 T 2 -I 1 1 1 1 0 5 10 15 20 25 Siloxane Thickness (micron) Fig.4.5.5 The thickness of siloxane bond coat vs. the curvature radius The trend of the experimental results in Fig.4.5.5 showed that, for the Group A coating, the curvature radius rapidly increased when the siloxane thickness was below -10 um. At a thickness higher than 10 um, the CR stabilized at 14-15 m, which appeared to indicate that the 10 um siloxane BC was enough to relax the residual stresses. It thus seems unnecessary to increase the siloxane thickness beyond 10 um. 74 I t is not easy to contro l the thickness o f the bond-coat by hand-spraying when its thickness is be low 10 p m , in part icular be low 5 p m . The bond coat was often discontinuous w h e n i t was thinner than 10 p m , since the surface roughness o f the substrate was around 5 p m . A t a thickness greater than 10 p m , however , the bond coat became a cont inuous layer, and the residual stress and the curvature radius tended to be stable. Therefore, the compl iant bond coat can release the energy o f thermal stress, accommodate the bending severity o f the coating or substrate, and decrease the deformat ion o f the substrate. A plausible analysis shows that the natural warp o f the substrate dur ing f i r i ng does not greatly affect the f ina l curvature radius o f the coating. Since the steel substrates are isotropic, the heat treatment and ambient coo l ing can also be made homogeneous, showing imp l i c i t l y that the warp ing and de format ion o f the substrate can be neglected dur ing the heat treatment. I t can therefore be concluded that the si loxane pr imer decreases the thermal stress to w h i c h the substrate is subjected. The si loxane BC acts as a compl ian t layer, grading the thermal expansion coef f ic ient mismatch by releasing the energy and re lax ing the stresses. 75 4.6 Adhesion 4.6.1 Group A Adhesion Figure 4.6.1 gives a schematic of the fracture mechanisms of the coatings Groups A to C. CB-eSG Siloxane Substrate C B - C S G Siloxane eB-C$G Substrate; (a) (b) Group B -CB-CSG -Substrate (c) Fig.4.6.1. Schematics of fracture mechanisms in different coating structures In dual- layer Group A , the C B - C S G ceramic layer general ly provides the wear resistance, and the si loxane bond coat is able to undergo visco-elast ic deformat ion in service. The average bond ing strength o f Group A is - 1 0 . 0 M P a ( 1 0 % alumina) , a higher adhesion strength than that o f 100% siloxane. Figure 4.6.2 is an S E M image o f a Group A coat ing after an adhesion test, as schematically depicted i n Fig.4.6.1 (a). It shows that break-up seemed to occur at the interface between the bond coat and the substrate. It appears, f r o m Fig.4.6.3, that the siloxane adhesion to the substrate was stronger than that o f the C B - C S G to the si loxane; but under the opt ical microscope, i n Fig.4.6.4, i t was found that the fracture occurred inside the siloxane bond coat and not at the interface between the C B - C S G and the siloxane. Figure 4.6.5 indicates that the ul t imate tensile strength o f si loxane is lower than its adhesion strength, because the adhesion o f si loxane is compr ised o f t w o components, mechanical inter locks and po lymer ic adhesion. S E 2 0 K V 5 0 u m Fig.4.6.2. Fracture surface of Group A after an adhesion strength test 77 Fig.4.6.3. Appearance of the fracture surface generated in adhesion test Stud GSG Coating siloxane: Substrate ; . I Stud I ; 1 Before adherence testing After adherence testing Fig.4.6.5. Schematic of fracture path for sample shown in Fig.4.6.4 4.6.2 Group B Adhesion The t r i - layer Group B ( 2 5 % alumina) had an average ul t imate tensile strength for the siloxane bond coat o f - 1 3 . 0 M P a . The bonding o f the C B - C S G pr imer to the sandblasted m i l d steel consisted o f mechanical inter locks and chemical bonds prov ided by the a luminum phosphates. Figure 4.6.6 shows the fracture o f a tr i - layer Group B coat ing, w h i c h started at the weakest point o f the bond coat and ended at the outer surface o f the top coat, where the fracture was contained w i t h i n the top coat by the bond coat, referred to Fig.4.6.1 (b). The weakest po in t o f B C strength is usual ly the area o f lowest a lumina concentrat ion. Spurious cracks, generated by the pu l l i ng force, were sometimes found to occur i n the C B - C S G top coat, but they were un l i ke ly to penetrate into the sub-layer unless they occurred in those areas where the si loxane was thinnest. F igure 4.6.7 shows a crack that was prevented f r o m develop ing into the pr imer by the bond coat. 79 Fig.4.6.6. SEM image of Group B fracture surface after an adhesion strength test Fig. 4.6.7 A crack was constrained within one layer by the siloxane BC 8 0 The t r i - layer Group B coatings had a higher y ie ld strength than that o f the dual-layer Group A coatings because the strength o f the bond coat appears to be dependent on the vo lume percentage o f a lumina, increasing w i t h vo lume percentage o f alumina. Compared w i t h Group A , Group B had a higher percentage o f a lumina, g i v ing i t a higher strength. However , the adhesion o f Group A was as l o w as 10.0 M P a , ind icat ing that the strength o f 100% si loxane was lower than 10.0 M P a . Fig.4.6.8. SEM image of Group C fracture surface after an adhesion strength test 4.6.3 Group C Adhesion Figure 4.6.8 shows a br i t t le fracture i n adhesion test for a mono- layer C B - C S G coating (Group C) , as schematical ly depicted i n Fig.4.6.1 (c). The fracture occurred at the interface and developed along the coat ing, indicat ing that the tensile strength o f C B - C S G is higher than the strength o f its adhesion to the substrate. The average adhesion strength o f Group C was measured as 42.0 M P a . The adhesion strength increased w i t h increasing vo lume percentage o f si loxane [5 ] . 81 I n some cases, an adhesion strength as h igh as ~70.0 M P a for Group C was observed. There are several factors that affect the measurement o f p u l l - o f f adhesion strength: (1) the adhesion is l im i ted by the strength o f the epoxy glue used ( -70 .0 M P a ) ; (2) the effect o f pores i n the br i t t le C B - C S G coatings is sensit ive to the a l ignment o f the appl ied force; (3) the prevalence o f crack propagat ion in br i t t le C B - C S G coatings reduces the cr i t ical force o f detachment; (4) the adhesion strength is dependent on the surface roughness o f the substrate. Hence, the actual adhesion m igh t be h igher than the p u l l - o f f results for a certain coating/substrate combinat ion, since the adhesion strength could be detr imental ly affected by the above defects. 82 4.7 Potentiodynamic Evaluations of Sol-Gel Alumina Coatings I t was anticipated that the composite mul t i - layered coat ing w o u l d improve the corrosion resistance o f the metal l ic substrate due to: ( i ) the statistical possib i l i ty o f the number o f through-coat ing defects (e.g. pores, cracks) decreasing w i t h increasing coat ing thickness; ( i i ) the decrease i n the oppor tuni ty for through-pore fo rmat ion , due to the mechanical penetrat ion or anchorage o f di f ferent layers; ( i i i ) the part ia l f i l l i n g o f ceramic pores by the po lymer dur ing cur ing. The results o f the electrochemical evaluat ion, F ig .4 .7 .1 , revealed that the mu l t i -layer coatings fo rmed an effect ive physical barrier against wet corros ion. There was a siloxane layer i n dual- layer Group A and i n t r i - layer Group B, w h i c h increased the ohmic resistance o f the who le system, and decreased the corrosion rate s igni f icant ly . Tr i - layer Group B had a lower corros ion current density (2 -3x10" 7 A / m 2 ) than dual- layer Group A 6 2 (5-7x10" A / m ), because the Group B coat ing was th icker (~10 p m ) than the Group A coating for the same thickness o f si loxane bond coat and C B - C S G top layers. The corrosion currents o f the coated m i l d steel were around one order o f magni tude smaller than those o f the bare m i l d steel. Mono- layer Group C had a higher corrosion current ( -10" A / m ) and a smaller corrosion resistance than Groups A and B because there were more permeable channels i n the mono- layer Group C coat ing, associated w i t h physicochemical changes i n both the siloxane and C B - C S G mater ia l - e.g. shrinkage, 83 water vapor, and organics-burnout dur ing the cur ing i n both the si loxane and the C B -CSG material . r - " 3.50--! i ^ . r 5 0 _ j Current Density log 1 0l (A/cm2) Fig.4.7.1. Polarization curves of the CSG/siloxane coating (Group A-C) on the mild steel substrates compared with that of the bare mild steel (pH=5, DI H2O) Fig.4.7.1 shows that the dual- layer Group A coat ing fo rmed an ef fect ive physical barrier against wet corros ion. The shape o f the polar izat ion curve o f the coated substrates was not very di f ferent f r o m that o f bare substrate: there were not passivat ion regions present, w h i c h impl ies that the coat ing indeed prov ided a physical barr ier for b lock ing the electrochemical corros ion process, but that the electrochemical behavior o f the substrate also affected the trend o f polar izat ion at this thickness. B y p lo t t ing the potent ia l versus the logar i thm o f the current for the various thicknesses o f CSG/si loxane bond coatings, the relat ionship between current density and coating thickness was determined for a Group A coating system inc lud ing a 5-10 u.m 84 siloxane bond coat, and a 10-100 urn C B - C S G coat ing [5 ] . The corros ion current density 9 2 • decreased w i t h increasing thickness, and was constant at - 1 0 " A / c m for the coatings thicker than ~ 5 0 p m [5 ] . Th is m igh t be because the l im i ted number o f through-pores had been fu l l y b locked b y corros ion products, prevent ing fur ther transport o f oxygen to the steel substrate w h e n the thickness o f C B - C S G was over 50 p m . Table 4.7.1 Corrosion protection efficiency of different structure coatings Group A Group B Group C Structure Dual- layer Tr i - layer Mono- layer C B - C S G D I H 2 0 9 7 % 9 9 % 9 2 % l w t % N a C l 9 4 % 9 6 % 8 4 % The corros ion current densities were obtained f r o m the intersect ion o f the anodic and cathodic Tafe l l ines for D I water and l w t % N a C l solut ion. The corrosion protect ion ef f ic iency o f the Group A coat ing i n D I H2O was calculated to be - 9 7 % according to equation (2.10). The corros ion ef f ic iency was 9 4 % in l w t % N a C l so lu t ion (pH=5) . Table 4.7.1 shows the corrosion protect ion ef f ic iency o f the other groups i n both the D I H2O and the l w t % N a C l solut ion. The CB-CSG/s i loxane bond coating had a better corrosion protect ion ef f ic iency than the tradi t ional non-si loxane C B - C S G coating in D I H2O because the insulat ing siloxane decreased the permeabi l i ty o f the coating to l i qu id corrosive m e d i u m . 85 In NaCl solution, the siloxane bond coat offers better corrosion resistance, since the PMPS has two advantages in terms of corrosion resistance to cathode ions, e.g. OH" or Cl": (1) The hydrophobic nature of the PMPS repels water molecules being absorbed onto its surface. (2) The siloxane component separates the metallic substrate from the corrosive solution, i.e. because of the polymer barrier there is no anodic reaction caused by hydrolysis of iron ions and changing of localized pH values. 86 C H A P T E R 5: C O N C L U S I O N S 5.1 S u m m a r y o f C o a t i n g P rope r t i es A " w a r m " temperature (160-300°C) process has been developed to produce C B -CSG coatings free o f surface cracks. I t overcomes the drawbacks o f the t radi t ional process for p roduc ing chemica l ly bonded composite sol-gel coatings, e.g. cracks, l o w thickness, l o w hardness, and poor phosphate d is t r ibut ion. The use o f mono a luminum phosphate leads to fewer phosphate cracks than the use o f phosphor ic acid. The l o w temperature (160°C) cur ing alleviates the problems associated w i t h the thermal expansion coeff ic ient mismatch. A m u l t i - g u n spray technique was developed to distr ibute the phosphate homogeneously throughout the coating. The technique results i n a crack-free chemical ly bonded composi te sol-gel ( C B - C S G ) a lumina coat ing w i t h m e d i u m hardness (6.0 GPa), moderate adhesion (42.0 M P a ) and good scratch-resistance (17.0 kg f ) . The novel si loxane/ceramic funct ional ly gradient mul t i layer structural coatings were successful ly fabr icated us ing mu l t i - gun deposi t ion methods. I t is bel ieved that mu l t i -gun spraying is cr i t ica l for the process, as i t a l lows the deposi t ion o f an aqueous sol onto m i l d steel w i thou t r isk o f coat ing buck l ing and interfacial corrosion. The mu l t i -gun spraying method disperses the C S G particles throughout the si loxane and the phosphate throughout the CSG, w h i c h s igni f icant ly increases the mechanical performance o f the coating and distr ibutes u n i f o r m l y the phosphate bonds throughout the ceramic coating. 87 The result ing s i loxane/CB-CSG mul t i layer coatings were found to be u n i f o r m , adhesive, and relat ively dense. Three groups o f coatings ( A , B, C) were processed. For the Group A and B coatings, the thermal ly stable siloxane-based "bond coat" f i l m was deposited onto the m i l d steel to protect the metal surface dur ing the C B - C S G processing. The C B - C S G " top coat" protected the si loxane-based bond coat (and the meta l l ic substrate) f r o m wear damage. The si loxane-based bond coat prov ided corrosion resistance and damage tolerance through its l o w permeabi l i ty , its good adhesion, and its elasticity. 5.2 S u m m a r y o f A d h e s i o n The adhesion o f the coatings was dependent on the s i loxane/a lumina composite strength in the mul t i - layer structure. D u r i n g processing o f the coatings, the alpha a lumina particles were m i x e d into the si loxane, w h i c h increased the strength o f the who le coating system. The t radi t ional mono- layer Group C ( 4 0 p m ) had a 42.0 M P a adhesion strength, a 17.0 k g scratch cr i t ical force, and a - 0 . 2 mi l idarcy permeabi l i t y to air. The adhesion strength o f the coatings increased w i t h the surface roughness o f the substrate: the rougher the substrate surface, the higher the adhesion strength. However , after the surface roughness reached a value higher than 4 p m , the rate o f increase o f interfacial fracture toughness tended to become stable at 3 .5-4.0MPa.m . Therefore, the 88 op t imum surface roughness o f the substrate is - 4 . 0 urn. The sand-blasted stainless steel substrate resulted i n a higher fracture toughness than the gr inded stainless steel substrate. 5.3 Summary of Contact Measurements The average hardness o f the mono- layer C B - C S G ceramic coatings was 6.0 GPa. The mono a l u m i n u m phosphate ( M A P ) fo rmed chemical bonds w i t h the C S G and converted to ber l in i te or cristobali te after 5 days o f cur ing at 160°C. The average scratch cr i t ical force values for Groups A , B, and C were 4.7 k g , 10.3 k g and 17.0 k g respectively, and these values were dependent on the vo lume percentage o f si loxane. The scratch cr i t ical force increased w i t h increasing coat ing thickness: the th icker the coat ing, the higher the cr i t ical force. W h e n the thickness was be low 20 urn, the stresses were considered to be negl ig ib le. The cr i t ica l force increased almost l inearly w i t h increasing coat ing thickness. For thickness values f r o m 20-60 u m , the effects o f the residual stresses could not be ignored, as the rate o f increase o f scratch cr i t ical force vs. thickness decreased due to the cont r ibut ion o f the residual stresses. A f te r 60 u m , the cr i t ical force decreased w i t h increasing coat ing thickness due to the effect o f these stresses. The interfacial scratch fai lures were due to defects, e.g. pores and cracks, w h i c h are often generated by the h igh temperature processing. Therefore, l o w temperature 89 cur ing is an ef fect ive w a y to prevent interfacial fa i lure. The C B - C S G mono- layer Group C coating general ly had good scratch resistance when its thickness was less than 100 p m . 5.4 Summary of Residual Stress and Potentiodynamic Evaluations A n impor tant mer i t o f this w o r k is the use o f the inorganic (si loxane) po lymer bond coat (pr imer) at the interface between the metal l ic substrate and the ceramic coating. I t is bel ieved that such a coating system has been proposed and explored for the first t ime in this work . The " W a f e r " experiments to determine the curvature radius show that the si loxane pr imer can s ign i f icant ly decrease the thermal ly induced stress i n the C B - C S G coat ing. The si loxane pr imer can be used as a compl iant layer, w h i c h relaxes the stress caused by the thermal expansion coeff ic ient mismatch. The si loxane bond coat can also increase the bond strength o f the coat ing by decreasing the number o f coat ing defects. The potent iodynamic and permeabi l i ty experiments showed that the C B - C S G ceramic coatings w i t h si loxane s igni f icant ly decreased ceramic poros i ty and increased corrosion resistance o f the m i l d steel substrates. The corrosion current density o f the C B -CSG coating decreased w i t h increasing overal l coat ing thickness (> 100pm) , and f ina l ly tended to stabil ize at ~10~ 9 A / c m 2 . The inf luence o f the substrate decreased w i t h increasing coat ing thickness. The corrosion protect ion ef f ic iency o f the t radi t ional C B -CSG is - 9 2 % . Mu l t i l aye r coatings f o r m effect ive physical barriers to enhance corrosion 90 protect ion o f the m i l d steel substrate. I n part icular, the incorporat ion o f si loxane increases the corrosion resistance o f coated m i l d steel i n l w t % N a C l solut ion. A summary o f the coating characteristics for Groups A to C is g iven in Table 5 .1 . 5.5 Conc lus ions A " w a r m " temperature process (160-300°C) has been developed to m o d i f y the tradit ional sol-gel coat ing process. I t overcomes the drawbacks o f the t radi t ional process for producing composite sol-gel coatings, e.g. cracks, l o w thickness, l o w hardness, and poor phosphate d is t r ibut ion. A mu l t i -gun spraying technique was developed to more ef fect ively distr ibute the phosphate throughout the coat ing. The new process employs l o w temperature cur ing (160°C) , w h i c h results in a sol-gel a lumina coat ing free o f surface cracks w i t h a m e d i u m hardness (6.0 GPa), a moderate adhesion (42.0 M P a ) and good scratch resistance (17.0 kg f ) . This process can be used to produce re lat ive ly th ick (40-300 u m ) coatings. The air permeabi l i ty o f the C B - C S G coat ing is - 0 . 5 mi l idarcy at a thickness o f 40.0 u m . Th is research also explains w h y C B - C S G coatings have m e d i u m hardness and good scratch resistance. Another mer i t o f this research is the use o f po lymer ic si loxane as the bond coat for C B - C S G coatings, w h i c h acts to relax the residual stress between the C B - C S G coating and the steel substrate, and therefore decreases substrate deformat ion. The siloxane and the aqueous sol were deposited onto the m i l d steel th rough a mu l t i -gun 91 spraying technique, w h i c h successfully prevented the coat ing f r o m buck l i ng on the non-oxid ized m i l d steel substrate dur ing the heat treatment at temperatures up to 300°C. Table.5.1. The summary of characteristics of various coatings Properties Group A Group B Group C Siloxane Coating Structure Dual- layer Tr i - layer Mono- laye r N M Adhesion strength (MPa) 13.0 18.0 42.0 < 1 0 M P a Scratch Critical Force (kg) 4.7 10.3 17.0 N M Hardness (GPa) N / A N / A 6.24 0.12 Poisson's Ratio N / A N / A 0.25 0.35 Air Permeability (milidarcys) Non-detectable Non-detectable - 0 . 2 impermeable Residual Stress at 40pm (MPa) <1.0 N / A - 8 . 0 0 Interface Fracture Toughness N / A N / A 3.5 N M (MPa.m1/2) Elastic Modulus (Sonic, GPa) N / A N / A 225.5 5.0 Protection Efficiency (DI H 20) 9 7 % 9 9 % 9 2 % N M Protection Efficiency (1% NaCl) 9 4 % 9 6 % 8 4 % N M * N M : not measured 92 C H A P T E R 6 : F U T U R E W O R K One mer i t o f this w o r k was the development o f a " w a r m " temperature process for producing a crack-free composite sol-gel a lumina coat ing, w h i c h includes a mu l t i -gun spraying technique to distr ibute phosphate throughout the coat ing. Future w o r k should therefore focus on develop ing a further understanding o f the issues related to the " w a r m " temperature process, to increase the adhesion, hardness and scratch resistance o f the coatings. The f o l l o w i n g subjects should be addressed: 1. Ad jus t ing the " w a r m " temperature process to homogeneously distr ibute the phosphate th rough the coat ing. 2. M o d i f y i n g the " w a r m " temperature process to f o r m a wel l -c rys ta l l i zed phosphate a lumina coat ing. 3. Increasing the adhesion and the hardness o f the coatings up to 70 M P a and 10 GPa respectively. 4. M o d e l i n g the behavior o f the C B - C S G coating at h igh temperatures. Another cont r ibu t ion o f this w o r k is the development o f inorganic po lymer bond coat for residual stress re laxat ion and corrosion protect ion. However , there are some uncertanties i n the product ion o f the si loxane bond coat. Since the adhesion o f the siloxane bond coat is dependent on mechanical in ter lock ing, the mechanical anchor ing is 93 stronger and thus the adhesion strength o f the bond coat is h igher at points where there is greater penetrat ion o f a lumina particles. Therefore, future w o r k should also include further study o f the adhesion o f the si loxane bond coat to metals and ceramics, the effects o f secondary part iculate phases dispersed i n the si loxane bond coat, the characteristics o f the viscoelast ic deformat ion o f the bond coat, the effects o f chemical modi f icat ions o f the bond coat on its performance in residual stress relaxat ion, and the model ing o f the stress re laxat ion due to the bond coat. 94 References: [1] C.J.Brinker and G.W.Sherer, "So l -Ge l Science," Academic Press, San D iego, 1990 [2] C.J.Brinker, A .J .Hurd , P.R.Schunk, C.S.Ashely, R.A.Cairncross, J.Samuel, K.S.Chen, C.Scotto and R.A.Schwartz , "So l -Ge l Der ived Ceramic F i lms—Fundamenta ls and App l ica t ions , " i n : K.Stern (Ed.) , Meta l lu rg ica l and Ceramic Protect ive Coatings, Chapman & H a l l , London , pp .112-151, 1996 [3] T .Troczynsk i and Q.Yang, "Process for M a k i n g Chemica l l y Bonded Sol-Gel Ceramics," U.S. Pat. No.6 ,284,682, M a y , 2001 [4] T .O ld ing , M.Sayer and D.Bar row, "Ceramic Sol-Gel Composi te Coatings for Electr ical Insu la t ion, " T h i n Sol id F i lms, 398-399 581-586 (2001) [5] G .L i , H . K i m , M.Chen , Q.Yang and T .Troczynsk i , "Process Engineer ing o f Ceramic Composite Coatings for Fuel Cel l Systems," Ecomaterials & Ecoprocesses Sympos ium C O M 2003, 4 2 n d Conference o f Metal lurg ists, M . Mostaghaci (eds.), Vancouver , B C , Canada, I S B N 1-894475-44-5, Copyr ight by C I M , pp. 236, August , 2003 [6] ht tp: / /www.psrc.usm.edu/macrog/s i l icone.htm [7] "Silanes and Si l icones for Creative Chemists," Un i ted Chemica l Technologies, Inc. product advert isement brochure, 2002 [8] C.H.Hsueh and P.Miranda, " M o d e l i n g o f Contact- Induced Radia l Crack ing In Ceramic B i layer Coatings on Compl iant Substrates," J. Mater . Res., 18 [5] 818-821 (2003) 95 [9] J.Lawrence and L . L i , " A Laser-Based Technique for the Coat ing o f M i l d Steel w i t h a Vi t reous Enamel , " Surf. Coat. Technol . , 140 238-243 (2001) [10] P. K .Church and O.J.Kuntson, " M e t h o d o f Impregnat ing Porous Refractory Bodies w i t h Inorganic C h r o m i u m C o m p o u n d , " U.S. Pat. N o . 3,789,096, January 1974 [11] P. M i randa , A . Pajares, F. Guiberteau, F.L.Cumbrera and B . R . L a w n , "Contact Fracture o f Br i t t le B i layer Coatings on the Soft Substrates," J. Mater . Res., 16 [1] 115-126, (2001) [12] D. A . Bar row, T. E. Petroff, M . Sayer, " M e t h o d for Producing T h i c k Ceramic F i lms by a Sol Gel Coat ing Process," U.S. Pat. No.5,585,136, December 1996 [13] Y . X i e and H .M.Hawthorne , "Measur ing the Adhes ion o f Sol -Gel Der ived Coatings to a Duct i le Substrate by an Indentat ion-Based M e t h o d , " Surf. Coat. Technol . , 172 42-50 (2003) [14] Q.Yang, "Compos i te Sol-Gel Ceramics," U B C , Ph.D thesis, M a y 1999 [15] A . R . D i Giampaolo , M .Med ina , R.Reyes and M.Veles , " Z i n c Phosphate Interlayer for So l -Gel -Der ived A luminos i l i ca te Coat ing on A I S I - 1 0 1 0 Carbon Steel ," Surf. Coat. Technol . , 89 31-37 (1997) [16] H. L i , K. L iang , L. M e i , S. Gu , S. Wang , " O x i d a t i o n Protect ion o f M i l d Steel by Z i rcon ia Sol-Gel Coat ings, " Mater. Letts., 51 320-324 (2001) [17] M . Fallet, H. Mahd joub , B. Gautier, J. P. Bauer, "E lec t rochemica l Behavior o f Ceramic Sol-Gel Coat ings on M i l d Steel," J. Non-Cryst . Sol ids, 293-295 527-533 (2001) [18] S.Wi lson, H .M.Hawthorne , Q.Yang and T .Troczynsk i , " S l i d i n g and Abras ive Wear o f Composi te Sol-Gel A l u m i n a Coated A l A l l o y s , " Surf. Coat. Technol . , 133-134 389-396 (2000) 96 [19] T. Sugama and N . Carc ie l lo , "Hydro the rma l l y Synthesized A l u m i n u m Phosphate Cements," A d v . Cem. Res., 5 [17] 31-40 (1993) [20] M.Watanabe, D . R . M u m m , S.Chiras and A.G.Evans, "Measurement o f the Residual Stress in a P t -A lum in ide B o n d Coat , " Scripta Mater. , 46 67-70 (2002) [21] C.H.Hsueh and A.G.Evans, "Residual Stress and Crack ing in Meta l /Ceramic Systems for Microe lect ron ics Packaging," J. A m . Ceram. S o c , 68 [3] 120-127 (1985) [22] F. Lang, Z. Y u , " T h e Corros ion Resistance and Wear Resistance o f T h i c k T i N Coatings Deposi ted by A r c I on P la t ing , " Surf. Coat. Technol . , 145 80-87 (2001) [23] K .A. K h o r , and Y . W . G u , "Ef fects o f Residual Stress on the Performance o f Plasma Sprayed Funct ional ly Graded Z r 0 2 / N i C o C r A l Y Coat ings," Mat . Sci. Eng. , A 277 64-68 (2000) [24] Y . W . G u , K .A . K h o r , Y . Q . F u , Y . W a n g , "Func t iona l l y Graded Z r 0 2 - N i C r A l Y Coatings Prepared by Plasma Spraying Us ing Pre -Mixed Spheroidized Powders, " Surf. Coat. Technol . , 96 305-312 (1997) [25] M . N i i n o , S. Maeda, "Recent Development Status o f Funct iona l ly Gradient Mater ia ls , " ISIJ Int., 30 [9] 699-711 (1990) [26] J.D. Lee, H .Y .Ra, K.T. H o n g , S.K. Hur , "Ana lys is o f Depos i t ion Phenomena and Residual Stress in Plasma Sprayed Coat ings," Surf. Coat. Technol . , 56 27-31 (1992) [27] C.S. Richard, G. Beranger, J. L u , J. F. Flavenot, "The Inf luences o f Heat Treatments and In terd i f fus ion on the Adhes ion o f Plasma-Sprayed N i C r A l Y Coat ings, " Surf. Coat. T e c h n o l , 82 99-109 (1996) [28] A . V . V i rka r , "Strengthening o f Ox ide Ceramics by Transformat ion- Induced Stresses," J. A m . Ceram. S o c , 70 [3] 164-170 (1987) 97 [29] W . D . K i n g e r y , " In t roduc t ion to Ceramics," John W i l e y & Sons, Inc., pp.605, 1960 [30] P.Pantucek and I. Kvernes, " W e a r Resistant & Thermal Barr ier Coat ings in Diesel Engines, Methods for Op t im iza t ion o f Strength and Service L i f e , " i n C.Coutsouradis et al. (eds.), Mater ia ls for Advanced Power Engineer ing, Part 1, K l u w e r Academic , Dordrecht, pp.741-764, 1994 [31] M . M e l l a l i , P.Fauchais, A . Gr imaud, " In f luence o f Substrate Roughness and Temperature on the Adhes ion/Cohes ion o f A l u m i n a Coat ings," Surf. Coat. Technol . , 81 275-286 (1996) [32] A.G.Evans and J. W . Hutch inson, "The ThermoMechan ica l In tegr i ty o f T h i n F i lms and Mu l t i - Laye rs , " A c t a Meta l l . Mater. , 43 [ 7 ] , pp.2507-2530, (1995) [33] S.J.Bull , " M o d e l i n g o f Residual Stress in Ox ide Scales," Ox ida t ion o f Meta ls , 49 112-117(1998) [34] A .S .Wagh and S.Y.Jeong, "Chemica l l y Bonded Phosphate Ceramics: I I , W a r m Temperature Process for A l u m i n a Ceramics," J. A m . Ceram. S o c , 86 [11] 1845-1849 (2003) [35] W.D .K inge ry , "Fundamenta l Study o f Phosphate B o n d i n g in Refractor ies: I I , Co ld -Setting Propert ies," J. A m . Ceram. S o c , 33 [8] 242-247 (1950). [36] M.J .O 'Hara , J.J.Duga and H.D. Sheet Jr., "Studies i n Phosphate B o n d i n g , " A m . Ceram. Soc. B u l l , 51 [7 ] 590-595 (1972) [37] J. V . Bothe Jr. and P. W . B r o w n , "Low-Tempera tu re Synthesis o f A 1 P 0 4 , " Ceram. Trans., 16 6 8 9 - 6 9 9 ( 1 9 9 1 ) [38] F. J. Gonzales and J. W . Hal loran, "React ion o f Orthophosphate A c i d w i t h Several Forms o f A l u m i n u m O x i d e , " A m . Ceram. Soc. B u l l . , 59 [7] 727-731 (1980) 98 [39] B. E. Yoldas, " A Transparent Porous A l u m i n a , " A m . Ceram. Soc. B u l l . , 54 286-2 9 0 ( 1 9 7 5 ) [40] H.Hawthrone, A . N e v i l l e , T .Troczynsk i , X . H u , M.Thammachar t , Y . X i e , J.Fu, Q.Yang, "Character izat ion o f Chemica l ly Bonded Composi te Sol -Gel Based A l u m i n a Coatings on Steel Substrates," Surf. Coat. Technol . , 176 [2] 243-252 (2004) [41] S.Wi lson, H .M.Hawthorne , Q.Yang and T .Troczynsk i , "Scale Effects i n Abrasive Wear o f Composi te Sol -Gel A l u m i n a Coated L i g h t A l l o y s , " Wear, 251 1042-1050 (2001) [42] A .S .Wagh and S.Y.Jeong, "Chemica l l y Bonded Phosphate Ceramics: I, a Dissolut ion M o d e l o f Fo rmat ion , " J. A m . Ceram. S o c , 86 [11] 1838-1844 (2003) [43] A .S .Wagh and S.Y.Jeong, "Chemica l l y Bonded Phosphate Ceramics: I I I , Reduct ion Mechanism and Its A p p l i c a t i o n to I ron Phosphate Ceramics, " J. A m . Ceram. S o c , 86[11] 1850-1853 (2003) [44] h t tp : / /www.psrc .usm.edu/macrog/Rlass .h tm [45] http:/ /mineral.gal leries.comyminerals/phospha<yberl ini t /berl ini t .htm [46] ht tp: / /mineral .gal ler ies.com/minerals/phosphat/var isci t /var isci t .htm [47] S.J.Bull and D.S.Rickerby, " N e w Developments i n the M o d e l i n g o f the Hardness and Scratch Adhes ion o f T h i n F i l m s , " Surf. Coat. T e c h n o l , 42 149-164 (1990) [48] Y . X i e and H .M.Hawtho rne , "E f fec t o f Contact Geometry on the Fai lure Modes o f Th in Coatings in the Scratch Adhes ion Test," Surf. Coat. T e c h n o l , 155 121-129 (2002) [49] Y . X i e and H .M.Hawthorne , " A Contro l led Scratch Test fo r Measur ing the Elastic Property, Y i e l d Stress and Contact Stress-Strain Relat ionship o f a Surface," Surf. Coat. T e c h n o l , 127 130-137 (2000) 99 [50] Y . X i e and H .M.Hawthorne , " A M o d e l for Compressive Coat ing Stresses in the Scratch Adhes ion Test ," Surf. Coat. Technol . , 1 4 1 15-25 (2001) [51] J.Malzbender, J .M.J. d e n - T o o n d e r , A.R.Balkenende, G.de W i t h , "Measur ing Mechanical Properties o f Coatings: a Methodo logy A p p l i e d to Nano-Par t ic le-F i l led Sol-Gel Coatings on Glass," Mater . Sci. Eng. , R 3 6 47-103 (2002) [52] D . R . M u m m , A.G.Evans and I.T.Spitsberg, "Character izat ion o f a Cyc l ic Displacement Instabi l i ty for a Thermal ly G r o w n Oxide in Thermal Barr ier System," Ac ta Mater., 4 9 2329-2340 (2001) [53] K. B. Bobb , D. A . L indquis t , S. S. Rooke, W . E. Y o u n g , and M . G. K leve , "Porous Sol id o f Bo ron Phosphate, A l u m i n u m Phosphate, and S i l i con Phosphate," Mater. Res. Soc. Symp. P r o c , 3 7 1 211-215 (1995) [54] H.Chai , B .R .Lawn, "C rack ing in Br i t t le Laminates F r o m Concentrated Loads, " Ac ta Mater., 50 2613-2625 (2002) [55] B .R.Lawn, " Indenta t ion o f Ceramics w i t h Spheres: a Century after Her tz , " J. A m . Ceram. S o c , 8 1 [8] 1977-1994 (1998) [56] V .Te ixe i ra , "Residual Stress and Crack ing in T h i n P V D Coat ings, " V a c u u m , 6 4 393-399 (2002) [57] C.H.Hsueh, "Some Considerations o f Determinat ion o f Residual Stress on Y o u n g ' s M o d u l i i n Ceramic Coat ings," J. A m . Ceram. S o c , 7 4 [7 ] 1646-1649 (1991) [58] C.H.Hsueh and A.G.Evans, "Residual Stress in Meta l /Ceramic Bonded Str ips," J. A m . Ceram. S o c , 6 8 [5] M a y (1985) [59] A S T M Standard C1259-94, Amer ican Society for Test ing and Mater ia ls, Phi ladephia, 1994 100 [60] J.F.Shackelford and W.Alexander , "The C R C Mater ia ls Science and Engineer ing Handbook, " C R C Press, Boca Raton, F L , pp.436-438, 1992 [61] R. C. Rossi, J. R. Cost and K. R. Janowski , " In f luence o f the Shape o f Dispersed Particles on the Elastic Behav ior o f Composi te Mater ia ls , " J. A m . Ceram. S o c , 55 234-237 (1972) [62] J. M . Ch iou and D .D .L .Chung , " Improvement o f the Temperature Resistance o f A l u m i n i u m - M a t r i x Composi tes Us ing an A c i d Phosphate B inder , " Part I Binders, J. Mater. Sci., 28 1435-1446 (1993) [63] J. M . Ch iou and D .D .L .Chung , " Improvement o f the Temperature Resistance o f A l u m i n i u m - M a t r i x Composi tes Us ing an A c i d Phosphate B inder , " Part I I Preforms, J. Mater. Sci., 28 1435-1446 (1993) [64] J. M . Ch iou and D .D .L .Chung , " Improvement o f The Temperature Resistance o f A l u m i n i u m - M a t r i x Composites Us ing an A c i d Phosphate B inder , " Part I I I A l u m i n i u m -matr ix composites, J. Mater. Sci., 28 1435-1446 (1993) [65] Caro lyn M o o r l a g , "Chemica l l y Bonded Composi te Sol -Gel Ceramics: a Study o f A l u m i n a Phosphate React ion Products," U B C , Master Thesis, Augus t 2000 [66] W . D . K i n g e r y , "Fundamenta l Study o f Phosphate B o n d i n g in Refractor ies: I I , Co ld -Setting Propert ies," J. A m . Ceram. S o c , 33 [8] 242-247 (1950) [67] A.E.Scheidegger, " T h e Physics o f F l o w through Porous M e d i a , " T h i r d Ed i t ion , Univers i ty o f Toronto Press, Toronto , pp. 127-129, 1974 [68] J .H.Hampton and D.A.Thomas, " M o d e l l i n g Relat ionships between Permeabi l i ty and Cement Paste Pore Microst ruc tures, " Cem. Concr. Res., 23 1317-1330 (1993) 101 [69] J.P.Ol l iv ier and M.Massat, "Permeabi l i ty and Microst ructure o f Concrete: A Rev iew o f M o d e l l i n g , " Cem. Concr. Res., 22 503-514 (1992) [70] Y . Joshua D u , Ma t t Damron , Grace Tang, H a i x i n Zheng, C-J,Chu and Joseph H.Osborne, " Inorgan ic /Organ ic H y b r i d Coatings for A i r c ra f t A l u m i n u m A l l o y Substrates," Prog. Org . Coat., 41 226-232 (2001) [71] P.M.Koinkar , M.G.Wankhede, M . A . M o r e , P.P.Patil, S.A.Gangal , " In f luence o f Synthesis Temperature on Electrochemical Po lymer izat ion o f O-An is id ine on L o w Carbon Steel," Syn. Met . , 130 193-201 (2002) [72] P.K.Sinha, R.Feser, "Phosphate Coat ing on Steel Surfaces b y Electrochemical M e t h o d , " Surf. Coat. Technol . , 161 158-168 (2002) [73] E.Winter, "Ename l Paint ing Techniques," N e w Y o r k , Praeger pp.342-344, 1970 [74] U.Oppi , "Ename l ing on Meta ls , " N e w York , Greenberg pp.232-233, 1957 [75] A.G.Evans, "Fracture Mechanics Determinat ions," Fracture mechanics o f ceramics, vol .1 concept, f laws, and fractography, Sympos ium on the fracture mechanics o f ceramics, Pennsylvania State Univers i ty , pp.17, 1973 [76] M . A Meyers , R. W . A rms t rong , H. O. K i rchner , "Mechan ics and Mater ia ls , " N e w Y o r k , W i l e y , pp.25-28 1979 [77] D.Chicot , P.Demarecaux, J.Lesage, "Apparen t Interface Toughness o f Substrate and Coat ing Couples f r o m Indentat ion Tests," T h i n sol id f i lms , 283 151-157 (1996) [78] Toschi , C .Melandr i , P.Pinasco, E.Roncari , S.Guicciardi G. Por tu, " In f luence O f Residual Stresses on the Wear Behavior o f A l u m i n a / A l u m i n a - Z i r c o n i a Laminated Composi tes," J. A m . Ceram. S o c , 86 [9] 1547-1553 (2003) 102 [79] J. M . Y e h , C. L. Chen, Y . C. Chen, C. Y . M a , K. R. Lee, Y . W e i and S. L i , "Enhancement o f Cor ros ion Protect ion Ef fect o f Po ly (O-E thoxyan i l i ne ) V i a the Format ion O f Po ly (0 -E thoxyan i l i ne ) -C lay Nanocomposi te Mater ia ls , " Polymer, 43 2729-2736 (2002) [80] T .Lan, P.D.Kavi ratna, T.J.Pinnavaia, " O n the Nature o f Po ly im ide-C lay H y b r i d Composi te , " Chem. Mater. , 6 573 (1994) [81] H .L .Tyan, Y . C . L i u , K . H . W e i , "The rma l l y and Mechan ica l l y Enhanced Clay/Poly imide Nanocompos i te v ia Reactive Organoclay, " Chem. Mater. , 11 1942 (1999) [82] Z .Wang , T.J.Pinnavaia, "Nanolayer Reinforcement o f Elastomeric Polyurethane," Chem. Mater. , 10 3769 (1998) [83] J .W.Gi lman, C.L.Jackson, A . B . M o r g a n , R.Hayyis, E.Manias, E.P.Giannelis, M .Wuthenow, D .H i l t on , S.H.Phi l ips, "F lammab i l i t y Properties o f Po lymer-Layered-Silicate Nanocomposi tes. Polypropylene and Polystyrene Nanocompos i tes , " Chem. Mater., 12 1866 (2000) [84] D.C. Lee, L .W.Jang, "Character izat ion o f Epoxy-C lay H y b r i d Composi te Prepared by Emuls ion Po lymer iza t ion , " J. A p p l . Po lym. Sci., 68 1997 (1998) [85] I .M .Z in , R . I .Howard , S.J.Badger, J.D.Scantlebury and S .B .Lyon , " T h e M o d e o f A c t i o n o f Chromate Inh ib i to r i n Epoxy Pr imer o n Galvanized Steel , " Prog. Org . Coat., 33 203-210(1998) [86] J .W.Gi lman, C.L.Jackson, A . B . M o r g a n , R.Hayyis, E.Manias, E.P.Giannelis, M .Wuthenow, D .H i l t on , S.H.Phi l ips, "F lammab i l i t y Properties o f Po lymer-Layered-103 Silicate Nanocomposi tes. Polypropylene and Polystyrene Nanocompos i tes , " Chem. Mater. , 12 1866-1869(2000) [87] Z .Wang, T.J.Pinnavaia, "Nanolayer Reinforcement o f Elastomeric Polyurethane," Chem. Mater. , 10 3769-3774 (1998) [88] B. Wessl ing, "Passivat ion o f Metals by Coat ing w i t h Polyani l ine: Corros ion Potential Shi f t and Morpho log ica l Changes," A d v . Mater. , 6 [3] 226-227 (1994) [89] J .M.Yeh, S.J.Liou, C .Y .La i , P.C. W u , T .Y.Tsa i , "Enhancement o f Corrosion Protect ion Ef fect i n Polyani l ine v i a the Format ion o f Po lyan i l ine-Clay Nanocomposi te Mater ia ls , " Chem. Mater . , 13 1131-1138 (2001) [90] S.H.Messaddeq, S.H.Pulc inel l i , C.V.Sant i l l i , A .C.Guasta ld i , Y.Messaddeq, "Microst ruc ture and Corros ion Resistance o f Inorganic-Organic (Z r02 -PMMA) H y b r i d Coat ing o n Stainless Steel , " J. N o n Cryst. Sol ids, 247 164-170 (1999) [91] J.Masalski , J.Gluszek, J.Zabrzeski, K.Ni tsch, P.Gluszek, " I m p r o v e m e n t i n Corros ion Resistance o f the 3 1 6 L Stainless Steel by Means o f AI2O3 Coatings Deposi ted by the Sol-Gel M e t h o d , " T h i n So l id F i lms, 349 186-190 (1999) [92] M.Sexsmi th and T .Troczynsk i , "The Var iat ions i n Coatings Properties across a Spray Pattern, I n Thermal Spray Industr ia l App l i ca t ions , " Thermal Spray Proceedings, A S M internat ional, Oh io , pp.751-757, 1994 [93] C . K . L i n and C.C.Berndt, "Measurement and Analys is o f Adhes ion Strength for Thermal Spraying Coat ings, " J. Thermal Spray Technol . , 3 [1 ] 75-104 (1994) [94] Y . X i e and H .M.Hawtho rne , " A Contro l led Scratch Test for Measur ing the Elastic Property, Y i e l d Stress and Contact Stress-Strain Relat ionship o f A Surface," Surf. Coat. Technol. , 127 130-137 (2000) 104 [95] S.Dallaire, " In f luence o f Temperature on the B o n d i n g M e c h a n i s m o f Plasma-Sprayed Coat ings," T h i n Sol id F i lms, 95 237-244 (1982) [96] M . M e l l a i , P.Fauchais and A . Gr imaud, " In f luence o f Substrate Roughness and Temperature on the Adhes ion / Cohesion o f A l u m i n a Coat ing , " Surf. Coat. Technol . , 81 275-286 (1996) [97] M . M e l l a i , P.Fauchais and A . Gr imaud, " In f luence o f Substrate Roughness and Temperature on the Adhes ion / Cohesion o f A l u m i n a Coat ing , " Surf. Coat. Technol . , 81 275-286(1996) [98] P. Pantucek and I. Kvernes, " W e a r Resistant & Thermal Barr ier Coat ings in Diesel Engines, Methods for Op t im iza t ion o f Strength and Service L i f e , " i n C.Coutsouradis et al. (eds.), Mater ia ls for Advanced Power Engineer ing, Part I, K l u w e r Academic , Dordrecht, pp741-764, 1994 [99] A . I toh , M . H i ra ta and M . A y a g a k i , "Ef fects o f Substrate Temperature D u r i n g Spraying on the Properties o f Sprayed Coatings, I n Thermal Spray Coat ings: Research Design and A p p l i c a t i o n , " Thermal Spray Proceedings, A S M internat ional , Oh io , pp.593-600, 1993 [100] J.M.Zaat, " T h e r m a l Spray ing, " A n n . Rev. Mater. Sci., 13 9-42 (1983) [101] A . Hassui, S.Kitahara, T. Fukush im, " O n Relat ion between Properties o f Coat ing and Spraying A n g l e i n Plasma Jet Spray ing," Trans. N a t l . Res. Inst. Met . , 12 [1] 9-20 (1970) [102] J.Masalski , J.Gluszek, J.Zabrzeski, K .Ni tsch, P.Gluszek, " Improvement i n Corrosion Resistance o f the 316L Stainless Steel by Means o f AI2O3 Coatings Deposi ted by the Sol-Gel M e t h o d , " T h i n Sol id F i lms, 349 186-190 (1999) 105 [103] D.Yan, J. He, X . L i , Y . L i u , J. Zhang, H. D i n g , " A n Invest igat ion o f the Corrosion Behavior o f A l 2 0 3 - B a s e d Ceramic Composi te Coatings in D i lu te HC1 So lu t ion , " Surf. Coat. T e c h n o l , 141 1-6 (2001) [104] C.P.Bergmann, " In f luence o f the Substrate Roughness on the Adherence o f Plasma Sprayed Ceramic Coat ings, i n Thermal Spray Industr ia l App l i ca t ions , " A S M Internat ional, Oh io , pp.683-686, 1994 [105] C.P.Bergmann, " In f luence o f Thermal Conduct iv i ty o f The Spray Powder on the Adherence o f Plasma-Sprayed Ceramic Coatings, I n Thermal Spray Coat ings: Research, Design A n d A p p l i c a t i o n , " A S M Internat ional, Oh io , pp.537-542, 1994 [106] M.Varde l le , A .Varde l le , P.Fauchais and C.Moreau, "Pyrometer System, for M o n i t o r i n g the Part icle Impact on a Substrate dur ing a Plasma Spray Process," Meas. Sci. T e c h n o l , 5 205-212 (1994) [107] A . R . N i c o l l , " T h e Ef fect o f V a r y i n g Plasma G u n Nozz le Diameters on the Surface Roughness, Hardness and B o n d Strength o f AI2O3 A n d C r 2 0 3 Coat ings, I n Thermal Spray Technology: N e w Ideas and Processes," Thermal Spray Proceedings, A S M Internat ional, Oh io , pp. 19-25, 1989 [108] L .B ianch i , F.Ble in, P.Lucchese, A . G r i m a u d and P.Fauchais, "Compar i son o f Plasma Sprayed A l u m i n a and Z i rcon ia Coatings by R F and D C Plasma Spraying, I n Thermal Spray Industr ia l A p p l i c a t i o n , " A S M internat ional, Oh io , pp.575-579, 1994 [109] C.Richard, J .Lu and J.F.Flevenot, G.Beranger and F.Decomps, "S tudy o f C2O3 Coat ing Mater ia ls , i n Thermal Spray: Internat ional Advances i n Coatings Techno logy , " Thermal Spray Proceedings, A S M internat ional, Oh io , pp.11-16, 1992 106 [110] J .A.Marceau, R.H.F i rminhac, Y . M o j i , " M e t h o d for P rov id ing Env i ronmenta l l y Stable A l u m i n u m Surfaces for Adhesive Bond ing and Product Produced, " U S Patents 4,085,012, A p r i l 1978 [111] H .W.Ho lmqu is t , L.E.Tarr, " M e t h o d o f Cleaning A l u m i n u m Surfaces," US Patents 4,793,903, December 1988 [112] M . D . D r o r y and J .W.Hutch inson, "Measurement o f the Adhes ion o f a Br i t t le F i l m on a Duct i le Substrate by Indentat ion," Proc. R. Soc. Lond . , A 452, 2319-2341 (1996) [113] "Meta ls Handbook" , n in th edi t ion, V o l u m e 13, " C o r r o s i o n " , 1978 [114] N . N . V o e v o d i n , N.T. Grebasch, W.S . Soto, F.E. A r n o l d , M.S. Don ley , "Potent iodynamic Eva luat ion o f Sol-Gel Coatings w i t h Inorganic Inh ib i to rs , " Surf. Coat. T e c h n o l , 140 [1] 24-28 (2001) [115] J.R.Taylor and A . C . B u l l , "Ceramics Glaze Techno logy , " Pergamon Press, f i rst edi t ion, N e w Y o r k , pp.134-138, 1986 [116] V .B .M iskov ic -S tankov ic , D .M.Draz ic and M.J .Teodorov ic , "E lect ro ly te Penetration through E p o x y Coatings Electrodeposited on Steel ," Cor ros ion Science, 37 [2] 241 -252 (1995) [117] A S T M C 747-93, "Standard Test M e t h o d for M o d u l i o f E last ic i ty and Fundamental Frequencies o f Carbon and Graphite Mater ials by Sonic Resonance," M a r c h (1993) [118] A S T M C 848-88, "Standard Test Me thod for Y o u n g ' s M o d u l u s , Shear Modu lus and Poisson's Rat io for Ceramic Whi tewares by Resonance," N o v e m b e r (1988) [119] A S T M D 4092-90, "Standard Termino logy Relat ing to D y n a m i c Mechanica l Measurements on Plast ics," June (1990) 107 [120] L. Chandra and T. W . Clyne, "Use o f an Ul t rasonic Resonance Technique to Measure the In-Plane Y o u n g ' s Modu lus o f T h i n D i a m o n d F i lms Deposi ted by a D C Plasma Jet," J. Mat . Sci. Letts., 12 [3] 191-195, (1993) [121] W . L . Headr ick, R. E. Moore , A . V a n Leuven, "Measur ing Refractory M O E at H i g h Temperatures," Issue o f Ceramic Industry, November (2000) [122] J. Schrooten, G. Roebben and J. A . Helsen, " Y o u n g ' s M o d u l u s o f B ioact ive Coated Oral Implants: Porosi ty Corrected B u l k Modu lus Versus Resonance Frequency Analys is , " Scripta Mater. , 41 [10] 1047-1053 (1999) [123] I. A . Ashcrof t , C. W . Lawrence, T. P. Weihs and B. Derby , " I n Si tu Measurement o f the Elastic M o d u l i o f Glass-Ceramic Th i ck F i l m s , " J. A m . Ceram. S o c , 75 [5] 1284-1286(1992) [124] Ch in-Chen C h i u and E ldon D. Case, "Elast ic Modu lus Determina t ion o f Coat ing Layers as A p p l i e d to Layered Ceramic Composi tes," Mat . Sc. Eng. , A 132 39-47 (1991) [125] R.C. Rossi, "Pred ic t ion o f the Elastic M o d u l i o f Composi tes, " J. A m . Ceram. S o c , 5 1 , August (1968) [126] T.P.Chou, C.Chandrasekaran, S.J.Limmer, S.Seraji, Y . W u , M.J.Forbess, C.Nguyen, G.Z.Cao., "Organic- Inorganic H y b r i d Coatings for Corros ion Protect ion," J. Non-Cryst . Sol ids, 290 153-162 (2001) [127] Y .Mass ian i , "Progress i n the Understanding and Prevent ion o f Cor ros ion , " Insti tute o f Mater ials, London , pp.179, 1993 [128] I. Skeist, " H a n d b o o k o f Adhesives," Second Ed i t ion , Skeist Laborator ies, Inc. L iv ings ton , N e w Jersey, pp628, 1977 108 [129] M . M e t i k o s - H u k o v i c , E.Tkalcec, A . K w o k a l , J.Piljac, " A n I n V i t r o Study o f T i and T i - A l l o y s Coated w i t h Sol -Gel Der ived Hydroxyapat i te Coat ings," Surf. Coat. Technol . , 165 40-50 (2003) [130] C.K.Tan and D.J .B lackwood, "Cor ros ion Protect ion by Mu l t i l aye red Conduct ing Polymer Coat ings," Corros ion Science, 45 545-557 (2003) [131] C .L iu , A . L e y l a n d , Q.B i , A .Ma t tews , "Cor ros ion Resistance o f M u l t i - L a y e r e d Plasmas-Assisted Physical V a p o r Deposi t ion T i N and C r N Coat ings, " Surf. Coat. T e c h n o l , 141 164-173 (2001) [132] V .B .M iskov i c -S tankov ic , M.R.Stanic, D .M.Draz ic , "Cor ros ion Protect ion o f A l u m i n u m by a Cataphoret ic Epoxy Coat ing , " Prog. Org . Coat., 36 53-63 (1999) [133] K. Ho lmberg and A . Mathews, "Coat ings Tr ibo logy : A Concept, Cr i t ica l Aspects and Future D i rec t ions , " T h i n Sol id F i lms, 253 173-178 (1994) [134] M . Fallet, H . M a h d j o u b , B. Gautier, J. P. Bauer, "E lec t rochemica l Behavior o f Ceramic Sol -Gel Coat ings on M i l d Steel ," J. Non-Cryst . Sol ids, 293-295 527-533 (2001) [135] T. L .Met roke, R. L .Parkh i l l and E. T. Knobbe, "Passivat ion o f Me ta l A l l o y s Us ing Sol-Gel Der ived Mate r ia l s -A Rev iew, " Prog. Org . Coat., 41 233-238 (2001) [136] C .L i , H.Zhao, M .Matsumura , T.Takahashi , M.Asahara, H .Yamaguch i , " T h e Ef fect o f N i C r Intermediate Layer on Corrosion Behavior o f Cr203 Ceramic Coated Mater ia ls , " Surf. Coat. T e c h n o l , 124 53-60 (2000) [137] J. S. Chen, J. G. D u h , F. B. W u , "Microhardness and Cor ros ion Behavior i n CrN/Electroless N i / M i l d Steel Complex Coat ing , " Surf. Coat. T e c h n o l , 150 239-245 (2002) 109 Appendix I Determination of Young's modulus by Sonic Waves A n ultrasonic wave tester (Gr indoSonic M K 5 " I n d u s t r i a l " Instrument, J .W.Lemmens, Inc., M O , U S A ) is used to measure the elastic modulus and Poisson's ratio according to A S T M C-1259-04. The dynamic Y o u n g ' s modulus can be obtained f r o m [117-119] : 10~ 7 , , Ed= — *AL2N2d (1-1) " 981 where £^=modulus o f sample (GPa) Z=d imens ion o f sample (cm) N= resonant f requency (Hz) ^ d e n s i t y o f sample (g/cm3) For rectangular shape substrate, shear modulus can be calculated f r o m [120, 121]: ALmN2 G= [B 1(1 +Ay] (1-2) wt where I = d i m e n s i o n o f sample, /=thickness, w = w i d t h , w=mass, 7V=resonant f requency, A, B =empir ica l correct ion factors dependent on w and /. Poisson's rat io can be calculated by: ^ = T ^ - 1 (1-3) 2GC where: 110 vc=Poisson's ratio o f coat ing .Ec^Young's modulus o f coat ing Gc=shear modulus o f coat ing Consider ing that there are some pores in the coat ing, the corroborated equation can be used as [122 ] : E = E0(\-aPc) (1-4) where E is the Y o u n g ' s modulus o f porous coat ing, E0 represents the modulus o f fu l l y dense coat ing, Pc is the vo lume f ract ion porosi ty. For per fect ly spherical holes the constant (a) is the f o l l o w i n g func t ion o f coat ing Poisson's rat io vc [122] : 3(9 + 5vc)(l-vc) a = — — — (1-5) 2(7-5vc) 1 V The f o l l o w i n g method includes testing o f uncoated substrates as a reference f r o m w h i c h the effect o f the coat ing may be obtained and the coat ing 's modulus can be calculated [123, 124]: 1. Measure the bare substrate's density Ds and thickness Ts 2. Measure the coated substrate's density Ds.c and thickness Ts.c. Coat ing 's density (D) can be calculated by: 111 T T S—C rri S rr\ 1 s-c *• S-C 3. Input substrate's Poisson's ratio . 4. Measure the bare substrate's Young ' s Modu lus Ed.s by 90° f lexura l mode 5. Input in i t ia l coated substrate's Poisson's ratio v/ 6. Measure the coated substrate's Y o u n g ' s Modu lus Ed-C by 90° f lexura l mode 7. Measure the coated substrate's shear Modu lus Gd.c 8. Calculate the coated substrate's Poisson's rat io v. E i v = 1 2G 9. I f > 0 .02% then v,=v go to step 7 v 10. Calculate the coat ing's modulus Ec by: 77 T — T 11. Ec =(-*-) E S ) Ed 12. Calculate the coat ing's shear modulus Gc by: T T -T 13. Gc=-i-Gtl_,+-zzr-LGd s-c *• s-c 14. Calculate the coat ing's Poisson's ratio vc'. y c = - ^ ~ i 2 G C 15. Checked by the equat ion: 112 N' 11 , where N: resonant f requency o f coated substrate, l + a c « ( 1) T„-T, No: resonant f requency o f bare substrate, 1 s~c , Ec'. coat ing's elastic modulus, Es: substrate's elastic modulus, Dc'. coat ing's density, Ds: substrate's density. 16. F inal vc, Ec and Gc obtained. Note: this method shows the satisfactory results on ly when the thickness o f substrate is be low 4 0 0 u m and coat ing thickness is above 4 0 u m [125] . 11 Appendix II Original Data of the Radius of Curvature vs. Thickness Table II.l CB-CSG thickness vs. curvature radius for Group C Group C X ( m ) Y ( m ) ( |X|+ |Y|)/2 (m) Average(m) Bare substrate, 18.62 15.86 17.24 16.5 (0.63) 18.32 15.14 16.73 16.23 15.29 15.76 15.82 16.72 16.27 10±2um C B - 9.82 11.44 14.63 13.6(0 .95) CSG 10.89 16.25 13.57 14.11 10.57 12.34 14.93 12.79 13.86 2 0 ± 2 u m C B - 12.16 9.82 8.99 8.5 (0.55) CSG 3.99 -19.48 7.74 2.69 18.20 8.45 12.62 11.04 8.82 30±2 u m C B - 5.69 0.66 3.18 3.35 (0.31) CSG 3.26 3.02 3.14 2.29 4.84 3.81 -3.26 3.30 3.28 40±2 urn C B - -2.07 3.95 3.01 2.21 (0.58) CSG 2.09 4.03 2.06 -0.75 3.53 2.14 1.48 1.79 1.64 6 0 ± 2 u m C B - 2.15 1.61 1.88 1.51 (0.39) CSG 3.01 1.22 1.22 1.74 0.50 1.12 2.43 1.17 1.80 8 0 ± 2 u m C B - 1.76 2.12 1.94 1.49 (0.51) CSG 0.57 1.50 1.04 2.12 1.69 1.91 1.10 1.02 1.06 100+5 u m C B - 0.73 2.64 1.69 1.47 (0.35) CSG 1.48 1.45 1.47 0.67 1.27 0.97 1.50 2.09 1.73 *Values in parentheses show the standard deviation 114 Table II.2 Siloxane Thickness vs. Curvature Radius for Group A Group A X ( m ) Y ( m ) (|X|+ Y| ) /2 (m) Average(m) 0 p m siloxane 6.51 -0.45 3.48 3.5 (0.92) -8.26 -0.63 4.45 -3.82 -0.73 2.28 -3.29 4.48 3.89 5±2 p m siloxane 9.46 11.63 12.38 12.7 (0.99) 21.25 6.21 13.73 13.06 .. 13.40 • -13.23 8.72 -14.20 11.46 8±2 p m siloxane 13.80 11.60 12.70 13.5 (0.63) 20.30 7.10 13.70 15.40 11.40 13.40 -14.13 14.27 14.20 10+2 p m 11.76 13.20 13.48 14.3 (1.29) siloxane 14.00 15.22 14.61 21.20 0.75 15.98 17.06 -13.20 13.13 15±2pm -11.11 -18.81 14.96 14.4(1 .49) siloxane -11.61 -19.99 15.80 -12.87 16.21 14.54 -10.18 14.42 12.30 20±2 p m 21.98 7.78 14.88 14.9(1 .76) siloxane 16.25 9.57 12.91 20.81 8.41 14.61 17.26 17.14 17.20 *Values in parentheses show the standard deviation 115 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0078714/manifest

Comment

Related Items