AFML-TR-71-70

ix

MECHANICS OF BRITTLE MATERIALS UNDER LINEAR

TEMPERATURE INCREASES

H. NEUBER A. WI&IMER INSTITUT FUR TECHNISCHE MECHANIK TECHIVISCHE UNIVERSITAT MUNICH, W. GERMANY

TECHNICAL REPORT AFML-TR-71-70

D D'C
APRIL 1971

X) fl1hT)

B
This document has been approved for public release aiid sale; its distribution is unlimited.
Ropmducdd bv

NATIONAL TECHNICAL INFORMATION SERVICE

Sprlngflold, Va. 22151

AIR FORCE MATERIALS LABORATORY AIR FORCE SYSTEMS COMMAND WRIGHT-PATTERSON AR. FORCE BASE, OHIO

DISCLAIMER NOTICE
THIS DOCUMENT IS BEST QUALITY PRACTICABLE. THE COPY FURNISHED TO DTIC CONTAINED A SIGNIFICANT NUMBER OF PAGES WHICH DO NOT .REPRODUCE LEGIBLY,

NOTICE

When Government drawings, specifications, or other data are used for any purpose other than in connection with a definitely related Government procurement operr.tion, the United States Government thereby incurs no responsibility nor any obligration whatsoever; and the fact that the government may have formulated, furnished, or in any way supplied the said drawings, specifications, or other data, is not to be regarded by implication or otherwise as in any wanner licensing the holder or any other person or corporation, or conveying any rights or permission to manpfacture, use, or sell any patented invention that may in any way be related thereto.

for ACCESSID.1

M
ac

SECTIO WHITE
MF SEC7t

A L WICED

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

,. .... .. o .o.. . .. .. -.... o........

3 0 IMIflO/AA-i9iiIT

COeD

Copies of this repcrt should not be returned unless return is required by security considerations, contractual obligations, or notice on a specific document.
iO.3Jne 1971-CO)0S-40-7I-676

AFML-TR-71-70

MECHANICS OF BRITTLE MATERIALS UNDER LINEAR TEMPERATURE INCREASES

H. NEUBER
A. WIMMER

I
IThis
document has been approved for public release and sale; its distribution is unlimited.

FOREWORD

The research described in this report was performed by the Institut fur Technische Mechanik, Technische Universitat Munchen, Germany under USAF Contract AF 61 (052) -875. The contract was initiated under Project No. 7350 "Refractory Inorganic Nonmetallic Materials," Task No. 735005, "Theory and Mechanical Phenomena." The work was technically administered by the Air Force Materials Laboratory, Air Force Systems Command, WrightPatterson Air Force Base, Ohio, with Dr. A.J. Herzog and Mr. James W. Poynter acting as Project Engineers. This report summarizes the results of research conducted between 1 October 1966 and 31 October 1969. The results of a related aspect of this overall program are reported in AFl-TR-68-23 dated March 1968. The manuscript of this report was released by the authors in September 1970 for publication as a techaical report. The authors wish to express their thanks to Dr. A.J. Herzog (now retired) and Mr. W.J. Trapp of the Air Force Materials Laboratory as well as to the officers of the European Office of Aerospace Research in Brussels for their cooperation. The advice and suggestions of Dr. Hahn and Dr. Heckel from the Institut fur Technische Mechanik, Technische Universitat Munchen, are greatly appreciated. This report has been reviewed and is approved.>/

C.T. LYNC/ Chief, Ad anced Metallurgical Studies Br. Metals and Ceramics Division Air Force Materials Laboratory

iii

UNCLASSIFE ____ Security C'assifiction

(Security claslticationof title, body of abstract and Indexing annotation muet be entered when the overall report it cleassitied) r ORGINATIN G ACTIVII v (Corporate author) 2a RCPORT SECURITY CLASSIFICATION

DOCUMENT CONTROL DATA - R&D

Institut fur Technische Mechanik

Unclassified
2b GROUP

Technische Universitat Munich, W. Germany

3 'tEPORT TITLE

Mechanics of Brittle Materials Under Linear Temperature Increases.

DESCRIPTIVE

NOTES (Type of report and Inclulive date)

S. AUTHOR(S)

H. Neuber and A. Wimmer

6. REPORT DATE 7a. TOTAL NO. OF PAG S 7b. NO. OF REFS

Apr 1971
Ba CONTRACT OR C.RANT No.

115

14

9a. ORIGINATORS REPORT NUMBER(S)

b. PROJECT NO.

AF6(052)-875 7350 Task 735005

Task73505

b. OTHER
tis

RPORT

NO(S) (Any other numbers that maybe alined

rport)

d. 10. AV A IL ABILITY/LIMITATION NOTICES

AFML-TR-71- 70
This document has been approved for public release and sale; its distribution

is unlimited.
II. SUPPL EMENTARY NOTES 12. SPONSORING MILITARY ACTIVITY

_Wriehi-Patterson 13.ABSTRACT

Air Force Materials Laboratory (LLS) Air Force Systems Command AFB OH 45433

Th4 s results of work done to show the effects of surface roughness, residual stresses, strain rate, microstructure, alternating bending fatigue loading and a non-uniform stress field on the strength of small (2.-8.mm diameter) alumina rods are given. The rods used in the work were characterized by determinations of grain size, porosity, hardness, surface roughness and damping factor. Static bend tests at room temperature are used to determine the Young's modulus, the fracture stress, and the strain at fracture and to show the influence of specimen volume on the fracture stress. The effects of grain size and of porosity on these characteristics are shown. A new tensile test

specimen is designed which achieves an almost uniform stress field in the test area. Some test results are given to show the validity of this procedure for brittle specimens.

DD IAN4 1473
FOM

unclassified. Security Classification

INCLASSTFIED

.. ,
LINK A ROLE WT LINKB ROLE WT LINK C ROLE WT

Security Classification
14."-KEY WORDS

Mechanical Properties - Alumina New Design Tensile Test Specimen Microstructure - Alumina Porosity - Property Relationships Residual Stresses in Alumina Rods Damping Characteristics - Alumina

INSTRUCTIONS 1. ORIGINATING ACTIVITY. Enter the name and address of the contractor, subcontractor, grantee, Department of Detense activity or other organization (corporate author) issuing the report. 2a. REPORT SECURITY CLASSIFICATION: Enter the overall security classification of the report. Indicate whether "Restricted Data" is included. Marking is to be in accordance with approoriate security regulations. 2b. GROUP: Automatic downgrading is specified in DoD Ditective 5200. I0 and Armed Forces Industrial Manual. Enter the group number. Also, when applicable, show that optional markings have been used for Group 3 and Group 4 as authorized. 3. REPORT TITLE: Enter the complete report title in all capital letters, Titles in all cases should be unclassified. If a m,,aningful title cannot be selected without classification, snow title classification in all capitals in parenthesis immediately following the title.
4. DESCRIPTIVE NOTES. If appropriate, enter the type of

imposed by security .lassification, using standard statements such as: (1) "Qualified requesters may obtain copies of this report from DDC." (2) "Foreign itnnouncement and dissemination of this report by DDC is not authorized." (3) "U. S. Government agencies may obtain copies of this report directly from DDC. Othe qualified DDC users shall request through "U. S. military agencies may obtain copies of this report directly from DDC. Other qualified users shall request through "All distribution of this report is controlled. Qualified DDC users shall request through
__ _

(4

(5)

'

report, e.g., interim, progress, summary, annual, or final. Give the inclusive dates when a specific reporting period is covered, 5. AUTHOR(S): Enter the name(s) of author(s) as shown on or in the report. Enter last name, first name. middle initial, If rilitary, show rank and branch of service. The name of the principal . ithor is an absolute minimum reqwtrement. 6. REPORT DATE. Enter the date of the report as day, month, year, or month, year. If more than one date appears on the report, use date of publication.
a

If the report has been furnished to the Office of I echnical Services, Department of Commerce, for sale to the public, Indicate this fact and enter the price, if known.

I1L SUPPL .MENTARY NOTES: Use for additional explanatory notes. I 12. SPONSO. iNG MILITARY ACTIVITY: Enter the name of the departmental project office or laboratory sponsoring (paying for) the research and development. Include aidress. 13. ABSTRACT: Enter an abstract giving a brief and factual summary of the document indicative of the report, even though it may also appear elsewhere in the body of the technical report. If additional space is required, a continuation sheet shall be attached. It is highly desirable that the abstract of classified reports be unclassified. Each paragraph of the abstract shall end with an indication of the military security classification of the information in the paragraph, represented as (TS), (S), (C). or (U) There is no limitation on the length of the abstract. However. the suggested length is from 150 to 225 words. 14 KEY WORDS: Key words are technically meaningful terms or short phrases that characterize a report and may be used as index entrie3 for cataloging the report. Key words must be selected so that no se .urity classification is required. Identifiers, such as equipment model designation, trade name, military project code name, eeographic loc..tion, may be used as key words but will be followed by an indication of technical context. The assignment o links, rules, and weights is optional.

should follow normal pagination procedures. i.e., enter the number of pages rontauiing information. 7b NUMBER OF REFERENCES, Enter the total number of references c:ted in the report, 8a. CONTRACT OR GRANT NUMBER. If appropriate, enter the applicable number of the Lontract or grant tinder which the report was written. 8b, 8c, & 8d. PROJECT NUMBER- Enter the appropriate military department dentiftc tion, such as prvjet nuniuer, subproject number, system numbers, task number, etc. 9a. ORIGINATOR'S REPORT NUMBER(S) Enter the offit-tal repvrt n.mber by which the doc.ment will be identified and .ntrolled by the originating activity. This number must be unique tj this report. , 9b OTHER REPORT NUMBER(S): If the report has been assigned any other report numbers (either b) the originator or byl the sponsor), also enter this number(s). 10. AVAILABILITY/LIMITATION NOTICES Enter any timitations on further dissemination of the report. other than thse

TOTAL NUMBER OF PAGES: The total page c,)unt

AFLc-wPAFr-ocT 66 S.V

Security Classification

ABSTRACT

The results of work done to show the effects of surface roughness, residual stresses, strain rate, microstructure, alternating bending fatigue loading and a non-uniform stress field on the strength of small (2.-8.mm diemeter) alumina rods are given. The rods used in the work were characterized by determinations of grain size, porosity, hardness, surface roughness and damping factor. Static bend tests at room temperature are used to determine the Young's modulus, the fracture stress, and the strain at fracture and to show the influence of specimen volume on the fracture stress. The effects of grain size and of porosity on these characteristics are shown. A new tensile test specimen is designed which achieves an almost uniform stress field in the test area. Some test results are given to show the validity of this procedure for brittle specimens.

iii

TABLE

OF

CONTENTS Page No.

1. 2. 2.1. 2.1.1. 2.1.2. 2.1.3. 2.1.4. 2.1.5. 2.2. 2.2.1.

Introduction Experimental Investigation of Alumina Test Material Chemical Ana]ysis Grain Size and Porosity Surface Roughness Hardness Damping Performance of Strength Tests Influence of the Surface Roughness on the Fracture Stress, Young's Modulus and the Fracture Strain of Alumina Residual Stresses in Alumina Rods Influence of the Loading Speed on the Fracture Stress, Young's Modulus and the Fracture Strain of Alumina Alternating Bending Fatigue Tests Dependency of Fracture Stress, Young's Modulus and Fracture Strain f':om Grain Size and Porosity Influence of the Specimen Volume and the Stress Distribution on the Fracture Stress of Alumina at Various Grain Sizes and Porosities Tension Tests References Tables and Illustrations

1 2 2 2 5 8 9 11 12 12

2.2.2. 2.2.3.

13 15

2.2.4. 2.2.5.

16 20

2.2.6.

23

2.2.7. 3. 4.

30 33 35

iv

LIST OF TABLES

Table [Al 2 Chemical analysis of Degussit Al 23, 23 PT, Degussit Al 24 and Degussit test substance 2461 Distribution of the grainsize and the porosity over the length of Alumina rods Dependency of the "Rauhtiefe" Rt, "Gldttungstiefe" Rp and the CenterLine-Average CLA from the grainsize and the porosity of Degussit A1 2 0 3 round bar specimens Dependency of the Vickers-microhardness from the grainsize and the porosity of 4 mm 0 Degussit A1 2 0 3 round bar specimens Vickers-micro-hardness measurements on single grains of different sizes. Determined on a Wesgo Al 995 rod 6 mm 0, porosity p = 3.70% Variation of the damping factor of 4 mm 0 Degussit Al 0 specimens with the porosity and t e 3 grainsize Dependency of the fracture stress, Young's modulus and the fracture strain from the surface condition of 6 mm 0 Degussit Al 0 specimens determined from a eding test at room temperature Residual stresses of a 6 mm 0 Wesgo Al 995 specimen, grainsize 6 = 17.2m porosity p = 3.99% Residual stresses of a 6 mm 0 Wesgo Al 995 specimen, grainsize G = 16.5/.# porosity p = 4.08%

Page 35

36

39

AID

41

42

43

44

45

List of Tables (Continuation)

Table 10 Residual stresses of a 6 mm 0 Wesgo Al 995 specimen, grainsize G = 16.4/,u porosity p = bG.9% Dependency of the fracture stress, Young's modulus and the strain at fracture from the strain rate of 6 mm 0 Wesgo Al 995 specimens at room temperature determined from a bending test Dependency of the fracture stress of 5 mm 0 smooth Wesgo Al 995 specimens from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature Dependency of the fracture stress of notched round bar Wesgo Al 995 specimens (nominal diameter 5 mm) from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature Dependency of the fracture stress of 5 mm 0 smooth Degussit Al 23 specimens from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature Dependency of the fracture stress of notched round bar Degussit Al 23 specimens (nominal diameter 5 mm) from the numberp of cycles to fracture determined from alternating bend fatigue tests at room temperature Dependency of the fracture stress; Young's modulus and the fracture strain from the porosity and grainsize of 4 mm 0 Degussit Al OQ specimens determined from a bengitig test at room temperature vi

Page 46

11

47

12

49

13

50

14

51

15

52

16

List of Tables (Continuation)

Table 17 Dependency of the fracture stress from the specimen volume, the porosity and the grainsize, determined from a bending test at room temperature Dependency of the exponent from the porosity and grainsize, aetermined at room temperature

Page 63

18

75

:1
vii*

-I

List of Figures

Fig. 1 2 Polished and etched Alumina specimens Distribution of the grainsize and the porosity of the round bar Alumina specimens delivered by Degussa Distribution of the porosity and the grainsize over the length of a 4 mm diameter Degussit Al 23 specimen Distribution of the porosity and the grainsize over the length of a 4 mm diameter Degussit Al 23 PT specimen Distribution of the porosity and the grainsize over the length of a 4 mm diameter Degussit Al 24 specimen Distribution of the porosity and the grainsize over the length of a 6 mm diameter Wesgo Al 995 specimen Dependency of the poresize and poredistance from the porosity at constant 6 .1 tum grainsize G Dependency of the poresize and poredistance from the grainsize at constant porosity p--5.6% Dependency of the "Rauhtiefe" Rt, "Gldttungstiefe" Rp and the CenterLine-Average CLA from the grainsize and the porosity of 4 mm diameter Degussit A1.0 3 specimens

Page 77 78

79

80

81

82

83

84

85

10

Dependency of the Vickers-micro-hardness 86 on the grainsize and the porosity of 4 mm 0 Degussit A1 2 03 round bar specimens Dependency of the Vickers-hardness from tho size of single grains of a Wesgo Al 995 rod 6 mm 0, porosity 3.7%
viii

11

87

List of Figures (Continuation)

Fig. 12 Vickers-impression on an Alumina specimen. Pressure of the diamond pyramid 94 P. Magnification 1400 x Broken pieces of an Alumina rod photographed shortly after a bending test with a Hitachi high speed motion camera Fdrster elastomat for the. determination of the damping constant 6 Variation of the damping factor of 4 mm diameter Degussit A1 2 03 specimens with the porosity and the grainsize Bending test equipment for the determination of the fracture stress, Young's modulus and the fracture strain of Alumina.at room temperature Dependency of the fracture stress, Young's modulus and the fracture strain from the Rauhtiefe Rt of 6 mm diameter Degussit Al 0 specimens determined from a bending tgst Dependency of the fracture stress, Young's modulus and the fracture strain from the Gldttungstiefe R of a 6 mm diameter Degussit A1 03 specimens determined from a bending2 test

Page
87

13

88

14 15

88 89

16

90

17

91

18

92

19

93 Dependency of the fracture stress, Youngis modulus and the fracture strain from the Center-Line-Average CLA of 6 mm diameter Degussit A120 3 specimens determined from a bending test Measuring equipment for the determinati.on of the residual stresses on 6 mm aiameter Wesgo Al 995 specimens 94

20

ix

List of Figures (Continuation

Fig. 21 Strains 6 plotted versus the remaining area received from the grinding off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress over the cross-section determined from the above diagram Strains C plotted versus the remaining area received from the grinding off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress over the cross-section determined from the above diagram Strains 6 plotted versus the remaining area received from the grinding off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress over the cross-section determined from the above diagram Test equipment for the determination of the influence of the loading speed on the fracture stress, Young's modulus and the strain at fracture Bending test equipment for the determination of the influence of the loading speed on the fracture stress, Young's modulus and the strain of fracture Photographs of a bending test taken on Tektronix cathode ray oscillographs. Crosshead speed of the Zwick electronic tension test machine 600 mm/minute. Dependency of the fracture stress, Young's modulus and the strain at fracture from the strain rate of 6 mm diameter Wesgo Al 995 specimens determined from a bending test at room temperature
x

Page 95

22

96

23

97

24

98

25

99

26

100

27

101

List of Figures (Continuation)

Fig. , 28 29 30 31 32 33 pecimen for alterDegussit A120 nating bend fatigue tests Schenck-Webi fatigue test machine and Tektronix cathode ray odcilloscope Scheme of the Schenck-Webi test machine Clamping device for alternating bend fatigue tests Vibration of a specimen mounted in a Schenck-Webi fatigue test machine Dependency of the fracture stress of smooth and notched Wesgo Al 995 specimens from the numbers of cycles to fracture, determined from an alternating bend fatigue test at room temperature Dependency of the fracture stresz of smooth and notched Degussit Al 23 specimens from the numbers of cycles to fracture, determined from an alternating bend fatigue test at room temperature Fracture shapes of smooth Wesgo Al 995 specimens broken under static and cyclic loads Fracture shapes of notched Wesgo Al 995 specimens broken under static and cyclic loads Dependency of the fracture stress, Young's modulus and the fracture strain from the grainsize of 4 mm diameter Degussit Al 0 specimens determined from a beNdlng test at room temperature
xi

Page 102 102 103 104 104 105

34

106

35

107

36

107

37

108

----------------------------------------------

List of Figures (Continuation)

Fig. 38 Depe-ndency of the fracture stress, Young's modulus and the fracture strain from the porosity of 4 mm diameter Degussit Al 0 specimendetermined from a be~ddng test at room temperature Dependency of the fracture stress of round bar Alumina specimens from the specimen volume, determined from a bending test at room temperature Dependency of the fracture stress of 4 na 0 round bar Degussit A1 2 03 specimens from the porosity with the grainsize as parameter. The values were determined from bending tests at room temperature 1 Dependency of the Weibull exponent M from the porosity for various grainsizes Specimen shape for tension tests with Alumina Tension test specimen of Alumina Complete test setup for a tension test on Alumina

Page 109

39

110

40

111

41

112

42

113

43 44

114 115

xii

1. Introduction For an understanding of the fundamental properties of polycrystalline brittle materials the study of the mechanical behavior of such solids is very important. In former times there was little interest in brittle materials, mostly due to their irregular behavior. Only in recent years a gradual realization of the importance of brittle materials exists as they provide a very high temperature strength. The need of such materials in high temperature engines leads to an intensified attention focused on factores affecting the strength of brittle solids. This resulted in an increased research effort in order to understand the phenomenological properties and the behavior of these materials. Today the initiation of cracks is generally accepted as being due to the mobilization of dislocations, which are rather uniformly dispersed. Once the dislocations are activated, either by temperature or by applied stress, they continue to move until an obstacle is encountered. This is most often provided by a grain boundary or some other barrier, where the dislocations pile up. If this occurs for sufficient numbers of dislocations a microcrack will form, which then leads to crack initiation in the material. In many cases the materials contain numerous finely dispersed flaws, inclusions or vacancies in the form of pores, which becanne operative at crack initiation mechanisms long before dislocation activation can act. Under these circumstances the material may fail at

-- r

stresses far lower than those required to mobilize the dislocation movement. Brittle crack propagation can be described with the theory derived by Griffith /l/. He postulates, that a flaw can propagate, provided its extension is accompanied by a surplus of strain energy in the material. The Griffith criterion entails only one constant, the crack depth. The energy required to create new surfaces is based upon the theoretical surface energy of the substance. This concept was proved to hold for completely brittle materials like glass in numerous investigations. Subsequent work extended the utility of the Griffith criterion to the case of multiaxial stresses and nonuniform stress fields and it has been applied with some success to explain the influence of the grainsize on the fracture stress or for strain rate effects. At the same time, the Griffith theory cannot explain size effects and different strengths corresponding to various loading conditions. Another theory, proposed by Weibull /2/, considers, that a random distribution of flaws exists in the material, assigning a certain probability of failure which depends upon the size and the state of stress of the body. This statistical approach to fracture can account for the variation in strength with volume, state of stress or type of loading. Regarding, however, such parameters as grainsize, porosity, surface effects or stress gradients thetheory is reduced to a condition requiring arbitrary adjustments in the empirical constants contained in it. The Weibull theory, however, is incapable to consider strain rate effects or delayed

-2-

fracture for instance. In performing the research brittle materials had to be chosen which allow a well-planned approach to fundamental investigations of the parameters influencing the strength and failure characteristics. Oxide ceramics are very suitable for this purpose. Therefore studies have already been done for several years (see former report /3/ and /4/) on Alumina, Zirconia, Magnesia and Spinell. The first year's effort consisted essentially in the performance of compression, bending and some not very accurate tension tests. As the results from the bending tests turned out to be most reliable this method was maintained for the determination of the fracture stress, Young's modulus and the fracture strain (strain of fracture) of the above mentioned oxide ceramics in the temperature range from -195 0 C to 12000C or 14000C. Alumina showed the most promising material properties and was therefore selected for further research work. This report deals mainly with a study of such factors affecting the strength of Alumina as effect of surface roughness, resiuual stresses, influence of strain rate, fatigue, effect of microstructure, size effect and non-uniform stress field. The decision for these initial investigations was largely based on the importance of information for the designing engineer, who must be able to predict the failure of construction elements by knowing the constitution of the material and the loading conditions.

2. 2.1.

Experimental Investigation of Alumina Test material For the conduction of the planned tests round bar Alumina specimens with diameters of 2, 4, 6 and 8 mm and with different lengths of 60, 110 and 250 mm were obtained from the companies "Degussa", Frankfurt/Main, Western Gold and Platinum Co., Berkley, Calif., USA and Hrand Djevahirdjian, Monthey, Switzerland. The obtained Alumina rods were denoted as Degussit Al 23r Degussit Al 23 PT, Degussit Al 24, Degussit test substance 2461, Wesgo Al 995 and Djevax Korund.

2.1.1.

Chemical analysis All above mentioned Alumina specimens had a chemical purity of at least 99,5%. An analysis of the Degussit rods, given by Degussa is presented in Table 1. Al 23 and Al 23 PT are equal. Al 24 differs from both materials only by the content of MgO. Investigations performed by the manufacturer in order to determine the effect of the contents of NgO on the compression strength exhibit, that the influence up to a content of 3% of MgO is below 3,5% in a temperature range from 200 to 11000 C. From Wesgo Al 995 specimens it is only known, that the purity level io beyond 99,5%. With the specimens of extreme purity like Degussit test substance 2461 and synthetic corundum it is intended to compare the coincidence of their test results with the others.

-4-

2.1.2. Grainsize and porosity The microstructure of Alumina plays an important

part for the mechanical properties, therefore the grainsize and the porosity was determined of each specimen. For the determination of the grainsize the specimens had t,- be polished and etched. For this reason small parts were cut with a diamand saw close to the fracture surface form the tested specimens. Then 2 the parts were ground with silicon carbide on castiron disks and polished with diamond compound of a particle size of 3/k. After these preparations the specimens were etched for two minutes in melted Borax at 1100 C and then boiled for about half an hour in a 50% hydrochlorid acid solution and for neutralisation another half an hour in water with little Amonia. The measurements of the grainsiz; uf A1 2 0 3 which is an irregular and multifarious product can be carried out in several ways. In accordance to other works we used the Rosival method, which consists at drawing a straight line of known length across many grains and counting the number of grains, crossed by the line. For a more exact determination 2 lines in perpendicular directions were drawn on 7 arbitrary spots. The average "graindiameter p " can then be received from the equation

G
1

=-

(1)

3 -5-

where ZZ is the number of grains counted in one line of length t . Fig.1 exhibits photographs of a polished and etched Alumina specimen magaified 500 times. The porosity is a measure of the gaseous inclosures in the material and is usually expressed in percent as the ratio of the pore-volume to the total volume of the specimen. The pore-volume can easily be calculated from thu difference of the specific weight of the testmaterial to the theoretical density of a pore-free Alumina bpdy. The theoretical density 4
3 by X - ray examinations was determined to be 4,00 p/cm of pure o4-Aluminumoxide. The same value could also

be achieved asymtotically by the method of the buoyant force. Thus, the value of the porosity can be expressed in percent by

p = (1
where

100

[%]

(2)

P = porosity [%]
=

4,00 [p/cm) of A1 2 03

4.he

theoretical density

=-the specific weigth of the Alumina test specimen The grainsize and the porosity of the Alumina rods delivered from Degussa for the conduction of the tests varied from 1,5 1 um to 60 AM for the grainsize and from 1% to 20% for the porosity. Unfortunately the values were not distributed uniformly ovsr the whole range. Fig.2 shows, that especially for small values of the porosities large grainsizes are lacking

-6-

and that for porosities from 10% to 15% only grainsizes from 5 c1 to 7 are present. In order to determine the influence of the grainsize and the porosity on the fracture stress it would be valuable to know both quantities exactly at the fracture area. As the porosity is a mean value of the whole specimen and as the grainsize was determined only close to the fracture surface their distributions over the length and the diameter of each group of the Alumina specimens were measured. For this purpose the specimens were cut into 16 pieces and the grainsize and the porosity were determined from each piece. While the grainsize and the porosity were almost constant over the whole cross-section, the distribution over the length showedconsiderable variation. The results are exhibited in Table 2 abd in Fig.3, 4, 5 and 6. It can'be seen from the diagrams that especially'with the Al 23 PT and Al 24 specimen the grainsize and porosity change over the specimen length and that this fact requires to determine both values right on the fracture surface. However, it is not possible to measure the porosity on the fracture surface. The simplest method is to photograph a polished specimen and measure the area of the dark spots, then calculate the ratio of the area of the dark spots to the whole area. But this method is unreliable as always pieces of grains will be torn out during the process of cutting, grinding and polishing. The fact, that we only can take an average porosity of each specimen partially can explain the scatter of the fracture stress if the scatter of the porosity over the specimen length is taken into account.

-I-

The stress concentrations at the pores are mostly affected by the pore-shape and the pore-distance. Therefore the relation between both parameters and their dependence on the grainsize and the porosity was investigated. For this reason two groups of specimens were selected, one with constant porosity of 5,6% and varying grainsize, the other with constant grainsize of 6,1 pkm and varying porosity. In Figs.7 and 8 the photographs of the carefully polished specimen are presented and they show clearly, that with constant porosity and changing grainsize the size of the pores grows with increasing grainsize. Consequently the number of pores diminishes by the same amount. At constant grainsize the number of pores increase with increasing porosity whereas the poresize remains unchanged. These facts make it possible that with changing grainsize not only the dislocation mechanism, but also the varying pore-shape and -distance is responsible for altering the fracture stress. 2.1.3. Surface roughness The surface roughness may influence the fracture stress of Alumina, therefore the variation of the surface roughness with changing grainsize and porosity is of h'igh interest. For these Investigations Degussit Alumina specimens with changing grainsize at constant porosity or changing porosity at constant grainsize were examined in the as received condition of the material. The surface texture was measured on several spots by means of a Perth-O-Meter, a surface measuring instrument, equipped with surface tracers which explore the fine surface and transmit the oscillation caused by the peaks and valleys to an electromechanical converter.

Three characteristic values were measured by-the Perth-O-Meter, the "Rauhtiefe" Rt, which is the total height from the highest peak to the lowest valley, the "Gldttungstiefe" Rp, which is the maximum -eak height above the centerline of the modified profile and the Center-Line-Average CIA, which represents the arithmetic mean value received from the absolute amounts of the distances of the surface profile to the centerline. Generally one of the characteristic data is not sufficient as the surface might, for example, exist of many small peaks or of wide hills. In both cases the values of Rt could be the same, in spite of the different surface profiles which affect the fracture stress. It turned out from the measurements, that all characteristic surface data decreased with increasing grainsize, that means the surface becomes smoother with larger grains, whereas with growing porosity the roughness of the surface increases (see Fig.9 and Table 3). 2.1.4. Hardness According to the hardness scale of F. Mohs which extends from 1 to 10, where number 10 is assigned to diamond as the hardest material, Alumina has the hardness 9. However not the degree of hardness is of main interest, but the possibility to measure the plastic deformability of Alumina by micro-hardness tests with little effort. Therefore hardness tests according to the Vickers method were conducted on Alumina specimens with varying grainsize and porosity. Throughout the whole test series the Vickers pyramide was pressed upon the specimens with constant load P

-9-

A measure for the size of the impression is then the diagonal dH and the micro-hardness can be calculated according to HV = 1,854o
.kp/mm

2]

(3)

On each specimen 10 impressions were arbitrarily distributed and evaluated. In Table 4 the arithmetic mean values of the Vickers-hardness are given with respect to the grainsize and the porosity of each specimen. The results are also plotted in Fig.1O. They show at small grains an increasing hardness. There is a maximum at a grainsize of about 18 fUm and from there on the hardness decreases with growing grainsize. With increasing porosity however the Vickershardness decreases monotonously. The low values for specimens with high porosities are in good agreement with the observations made by E. Ryshkcritch /5/, who states that the brittleness of Alumina decreases with increasing porosity. The mentioned hardness tests were distributed arbitrarily over the cross-section. The grainsize however, as Fig.1O shows, exerts a considerable influence, therefore it was necessary to measure the Vickers-hardness individually on single grains of certain sizes. For these tests a Wesgo Al 995 rod, 6 mm diameter was taken which scattered in the grainsize from 12/4m to 50 /Am . 10 grains of almost equal sizes were selected in certain intervals for hardness tests. The results are assorted in Table 5 and plotted in Fig.11. The dependency of the hardness on the grainsize was similar to that shown in Fig.1O. One of the Vickers impressions in the polished and

-10-

etched Wesgo specimens is demonstrated in Fig.12. 2.1.5. Damping The fracture face of the tested Alumina rods like the fracture surface or the number of broken pieces of an Alumina rod after a bending test (see Fig.13) is partly determined by the interior damping of the material. Therefore damping measurements were performed at room temperature with Degussit A1 2 0 3 round bar specimens having a diameter of 4 mm and a length of 100 mm. The damping constant 5 , also called the logarithmic decrement, is a measure for the amount of damping which the specimen possesses. If the amplitude of a vibrating specimen decreases during N cycles from a value A2 to a value A1 = A 2 /e (where e is the basis of the natural logarithms) it follows (4)

Ito

For these tests we used "Fbrster's Elastomat" (see Fig.14). In this apparatus the specimens are excited a certain resonance vibration (torsional, longitudinal, transversal) and the N cycles of a given amplitude range during the natural diminution are counted. The excitation can be done by means of piezoelectric or magnetic systems. The magnetic systems are only applicable to ferromagnetic materials, whereas the piezoelectric ones are spplicable to all oscillatory matri-ls. In order to be able to apply the magnetic system, small metal sheets were Zlued to the specimens in a

-11-

way, that the specimens were excited to transve-sal vibrations. The test results are presented in Table 6 and plotted in Fig.15. The damping constant shows an enormous increase with increasing porosity and a slight decrease with increasing grainsize. This result coincides with the observations that specimens with a low porosity break into more pieces than specimens with a high porosity. 2.2. Performance of strength tests

2.2.1. Influence of the surface roughness on the fracture stress, Youngs modulus and the fracture strain The influence of the surface condition on the fracture stress, Young's modulus and the fracture strain was investigated at room temperature with Degussit A1 2 03 specimens fired in one kiln under equal conditions in order to achieve constant grainsize and porosity for all speciuens. For the test series specimens with a diameter of 6 mm and a length of 250 mm were chosen. The average grainsize was 9,0pm and the average porosity 5,37%. The surface condition of the specimens was varied by grinding with Silicon carbide of different grainsize or polishing with fine grained diamond compound. Three characteristic values, as described in chapter 2.1.3., were measured by a Perth-O-Meter, the "Rauhtiefe" Rt, the "Gldttungstiefe" R and the Center-Line-Average CIA. For determining te fracture stress, Young's modulus and fracture strain in dependence from the surface condition, the prepared specimens were tested in bending. The bending test equipment is shown in Fig.16. -12-

It is in principal the same as described in

13/

and

/4/. The evaluation of the fracture stress, Youngts modulus and the fracture strain remained unchanged. The test results are presented in Table 7 and in Figs.17 to 19. They show, that the decrease rate of the fracture stress and fracture strain with increasing roughness is almost negligible. Furthermore, Young's modulus remains constant. These data reveal that fracture originates primarily on the flaws. It is suspected that only at low porosities the groves of the surface texture influence the fracture stress evidently. However only a test series covering the whole range of grainsize and porosity can give a complete answer. 2.2.2. Residual stresses in Alumina rods As the residual stresses of a material affect the fracture stress and fracture strain their determination was attempted on our common Alumina specimens. For this purpose 20 Wesgo Al 995 specimens with a length of 250 mm, a diameter of 6 mm and little variation of grainsize and porosity were selected. By this specimen shape, which corresponds to our bending test specimens, only the grinding off method which was elaborated by G. Sachs /6/ could be applied. With this method the diameter of the specimen has to be ground off stepwise and the change of length must be measured after each step. As only longitudinal strains can be measured a simplificating assumption has to be made, that there exists only a uniaxial stress state. According to G. Sachs this leads to the following equation

-13-

OIGZf)e-

+e

(5)

where

is Young's modulus the remaining cross-section the axial strain of the specimen

from which the distribution of the residual stresses over the cross-section of the specimen can be determined. Much care was taken for the grinding process in order not to produce additional stresses. The surface was ground off stepwise from 6 to 3,5 mm, however, only over a part of the length in order to have the specimen always fixed in the same position of the measuring equipment. After each step the specimen was stored far about one hour in a room with almost constant temperature, then the change of the specimen length was measured by means of a Huggenberger Optimeter with an accuray of + 0,0003 mm. The maximum variation of the temperature was approximately 100. The elongations resulting from this change of temperature were eliminated by comparison with a gaging rod. Fig.20 shows a measuring equipment which was mounted on a heavy steel plate in order to exclude the deformations between the supports and the optimeter. The test results of three specimens are presented in Tables 8 to 10 and plotted in Figs.21 to 23. The tests show, that all kinds of distributions are possible, the stresses however are negligible small. It also follows, if the arithmetic mean value of the elongations of all 20 tests is taken that the corresponding residual stress is zero throughout the whole cross-section.

-14-

2.2.3. Influence of the loading speed on the fracture stress, Young's modulus and fracture strain of Alumina One can imagine that the strain rate, as long as dislocation movements take place, affects the fracture stress and fracture strain. Therefore a test series for the determination of the influence of the loading speed at our bending tests on the fracture stress, Young's modulus and the strain at fracture was conducted at room temperature. For the bending tests Wesgo Al 995 specimens with a diameter of 6 mm and only slightly differing porosity and grainsize were taken. ll specimens were fired in one kiln, they had an average porosity of 3.7% and an average grainsize of 12 am The bending test setup which was built in into our Zwick electronic tension test equipment is shown in Figs.24 and 25. One can see the arrangement of the electric strain gages which measure the load applied to the leverarms (1) and the strains of the specimen (2). The electric signals of the strain gages were given into two Tektronix cathode ray oscillographs wbich recorded a load-strain and a strain-time diagram. Figs.26a and 26b show examples of a test. Prior to the tests the load, strain and time signals were gaged for analysing the photographed diagram. The straight line from Fig.26a results from the euperimposition of the load and strain signals, the inclination of which is in connecticn with the length of the leverarm and the moment of resistance a measure of Young's modulus. The end of the line which shows the maximum load and
strain is a measure for the fracture stressad sta ai

The curved line of Fig.26b is a superimposition of time and strain. The inclination is a measure of the

-15-

strain rate. The tests were performed at four crosshead speeds of the Zwick tension test machine, namely 0.6, 6.0, 60 and 600 mm per minute. It can be seen from the photograph that the strain rate becomes constant at the end even at the maximum speed of 600 mm per minute, therefore inertia influences can bc neglected for the determination of the fracture stress, fracture strain and for Young's modulus if the measurements will be done in the region of the constant strain rate. The errors which are caused by the inclined and dis-

placed glueing of the strain gages on the tension and compression side of the specimens are kept small because of the special form of the metal film strain gages, which had A width of 0.75 mm and a length of 15 mm. The test results of this series are given in Table 11. This table contains also the porosity and the grainsize of each specimen in order to eliminate their influence. The values of the fracture stress, Young's modulus and the strain at fracture are reduced to an uniform porosity of 4% and a grainsize of 12 . The test results are also plotted in Fig.27. The diagrams exhibit, that the fracture stress and the strain at fracture increase slightly with increasing strain rate, whereas Young's modulus remains constant, 2.2.4. Alternating bending fatigue tests An important reason for investigating fatigue phenomena in Alumina is that very little experimental work under fatigue conditions has been done with materials which approach the ideal concept of brittleness. Thus a knowledge of the fatigue behavior of ceramics in general should benefit a far wider field.

-16-

Alternating bending fatigue tests were started with Wesgo Al 995 and Degussit Al 23 Alumina rods. For the tests cylindrical smooth and circumferentially notched specimens were selected which are shown in Fig.28. The manufacturing of the smooth and notched specimens was done with diamond tools of equal grainsize. The specimen was rotating and the diamond disks of 175 mm and 2 mm diameter were grinding parallel to the direction of the centeraxis in order to achieve similar surface conditions. At first it was intended to determine the familiar SN curve with the stress plotted on a linear or logarithmic scale versus the numbers of cycles to fracture shown logarithmically. The tests were carried out on two Schenck-Webi fatigue test machines (Fig.29) which were developed for bending and torsion fatigue tests on all kinds of materials. It is possible to superimpose the dynamic loads with a static preload. For the momentary tests, however, the static preload (bending moment) was kept zero. The "Webi" is a mechanically driven machine with a steady adjustable lift which is necessarily constant during the operation. The frequency is 1495 cpm. Fig.30 shows schematically the method of operation and the measuring equipment. The double eccentric (8) can be adjusted to each permissable lift. With the help of the connecting rod (7) the lift can be transmitted to the loading lever (6). The specimen (5) will be fastened with a special clamping device (see Fig.31) on the loading and measuring lever (2). mh c.uLpJi1g device had to be equipped with roller bearings in order to avoid torsional stresses which otherwise would originate by tightening the screws

-17-

of the clamps. In addition it is necessary that the specimen is clamped in such a way that the neutral axis is in the same level as the center of the axis of rotation (4). The measuring level is mounted on leaf-springs (3) which enable a motion in the longitudinal direction of the specimen so that tension and compression stresses which arise from the contraction of the specimen under bending load are excluded. The bending moment is generated by the lift of the loading lever and by the resistance of the spring (10) which can be installed with different strengths. The measuring lever enlarges the motions of the spring and with the help of the two dial gages the amplitude of oscillation can be determined. The mounted spring and the indication of the dial gages are a measure for the applied load which can be determined from the diagram of the characteristic curves. If a specimen is broken, an automatic switch turns the motor off. The numbers of cycles of the test can be read from a counter. Before starting the test series several possibilities of error of the fatigue machine were investigated. For instance, the upper and lower bending stress and the preload which can be adjusted statically with the two dial gages were checked by means of electrical strain gages and the diagrams of the characteristic curves of the springs were improved according to our measurements. The already mertioned torsional stresses were also measured and it turned out that they are negligible small because of the installed roller bearings in the clamping device (Fig.31). Further,

-18..

C,

it was assumed that the weight of the measuring lever varies the statically applied load during cycling and disturbes the sinusoidal vibration more and more with decreasing flexural rigidity EJ, For this reason tests were performed with specimens of different materials and diameters in order to vary Young's modulus E and the ratio of inertia J. The electric signals of the applied strain gages were given into a Tektronix ray oscilloscope which recorded a strain-time diagram. No disturbance of the measuring lever on the bending stress or the vibration was noticed. This test was not interrupted for five days and we found that the stresses and the sinusoidal vibration remained unchanged. Fig.32 shows an example of one vibration cycle. The test program was started with 30 Wesgo Al 995

Kcethode
j

and 20 Degussit Al 23 smooth and notched bending specimens at room temperature. The test results are presented in Tables 12 to 15 and plotted in Figs.33 and 34. It can be seen that the fracture stress of the smooth and notched specimens decreases with increasing numbers of cycles. It seams however that the decrease rate of the notched specimens is less than that of the smooth specimens, so that for the Degussit Al 23 specimens the stress of the nonruptured specimens which ha e passed the fatigue test is the same for the notched and the smooth specimens and for the Wesgo Al 995 specimens the nominal fracture stress of the notched specimens is even higher than those of the smooth specimens at a number of cycles greater than 107. in order to draw conclusions from these introductory tests, these observations should be substantiated by more test results.

-19-

In Figs.35 and 36 some fractured fatigue specimens are shown. Viewed from bottom to top the first specimen is a typical static test, then the numbers of cycles increase, The fracture surfaces in the fatigued Alumina are characteristic in shape, being flat across most of the section but turning into a distinct breakaway lip. This shape is basically similar for the static failure only the breakaway lip tends to be a little greater. Williams found a simple explanation to interpret the general shape of the fracture surfaces. The crack originates at the tension surface and runs straight and fast at right angles to the imposed stress until it encounters the influence of a shock wave reflected from the opposite surface. The crack then moves to left or right of its original path. It is interesting that at all fatigued specimens the crack runs into the direction of the increasing cross-section, whereas in the case of static failure the crack turned into the opposite direction which is the direction of the increased stress. Similar observations can be made with the notched specimens, where the crack of the static bend test runs perpendicular to the applied stress through the smallest cross-section and the crack of the fatigued specimens deviated from the right angle into increasing cross-section. 2.2.5. Dependency of the fracture stress, Young's modulus and fracture .strain on the grainsize and the porosity The m ,,cture of a brittle material affects considerably its fracture stress, Yolngs modulus and

fracture strain.

-20-

Experience shows that the strength increases with decreasing grainsize and that, moreover, the brittle fracture of a polycrystalline material begins in one grain and then spreads to the others in the surroundings. The condition for the penetration of the crack across the boundary is probably the attainment of a critical stress at the end of a crack. Since the stress for the extension of a crack is inversely proportional to the square root of its length (Griffith /1/) one would expect the brittle strength of a polycrystalline material like Alumina to be inversely proportional to the square root of the mean grain diameter (Orowan E./8/). Another influence is exerted by the porosity. It is evident, however, that with increased porosity the strength of the resulting body will decrease. The relative loss will depend on the pore-size and pore-shape. E. Ryshkewitch /5/ has performed compression tests with porous sintered Alumina whose grainsize ranged from 1 to 3A and the experimental data can be represented by the following expression

"
=

(6)

strength of the porous body = strength of the nonporous body of the 0 same material porosity expressed as a fraction = material constant Statements of the coincident effects of grainsize G and porosity p on the strength of Alumina seem to be complicated as the pore-size and pore-distance is determined by the grainsize (see Fig.8). It can be

U
I,

-21-

concluded therefore, that the grainsize affects the influence of the porosity on the fracture stress. On the other hand the pores prevent the crack propagation in a grain as they act as obstacles which means that the influence of the grainsize on the fracture stress is affected by the porosity. For the determination of the influence of the grainsize and the porosity on the fracture stress, Young's modulus and the fracture str4in 160 bending tests were conducted at room temperature with Degussit A1 2 0 3 specimens having a diameter of 4 mm and a length of 250 mm. The specimens were manufactured in a way that the porosity and grainsize showed a variety according to Fig.2. As the bending test equipment displayed in Fig.16 has proved to .be appropriate this test setup was used and the method of evaluation of the fraoture stress, Young's modulus and fracture strain was carried out according to the equations given in /4/. The test results are given in Table 16 and the dependency of the fracture stress, Young's modulus and the fracture strain from the grainsize and the porosity is shown in Fig.37 and Fig.38. The diagrams were received in the following way: For instance, the fracture stress was plotted versus the grain diameter for groups of specimens with porosities scattering by 0.5% at most, then the fracture stress was plotted versus the porosity for different groups of grainsizes where the range was kept within I t of the other diagrams. The diagrams show that the grainsize affects the influence of the porosity on the fracture stress and fracture . The curves received of one diagram allowed to construct the lines

-22-

strain not only by a linear displacement 'ut also by the curvature. This is also true for the influence of the grainsize on the fracture stress and fracture strain at grains smaller than 20 prn , where the porosity affects the decrease rate. For grains larger than 20 1 sr" the influence of the grainsize is very small. Young's modulus decreases linearly with increasing porosity in the range 2.27% to 20.5%. The grainsize affects this influence only very little as the influence of the grainsize on Young's modulus is almost negligible small. For small grains the relationship of equation (6) has been confirmed, whereas for large grains the experimental curves deviate more and more from the exponential character. The influence of the grainsize on the strength of Alumina also shows an almost straightline relationship on a log - log scale for small porosities. For large porosities the curves deviate from a straight line, as the pores interfere with the dislocation movements. 2.2.6. Influence of the specimen volume and the stress distribution on the fracture stress of Alumina at various grainsizes and porosities The existence of a size effect with Alumina specimens is known with certainty. Many possible explanations have been suggested. An increase of stress with decreasing specimen volume follows from statistical considerations. It is characteristical for ceramic materials, that their strength values scatter over a comparatively wide range.

-23-

The statistical theory of Weibull /2/ allowes an adequate prediction of the failure probability of simple shapes of brittle materials and takes full account of the volume effect as well as the nature of the stress distribution to which the body is subjected. For an application of this theory it becomes necessary to determine the variability of the strength for a certain material and certain environmental parameters. The problem must therefore be attacked on a statistical basis, assigning a given probability of failure to a specified stress level. Theories based on this concept postulate that each material contains a random distribution of flaws. Fracture will occur by the most stress raising flaw which the specimen happens to contain. The specimen must be visualized as being composed of individual infinitesimal volumes, each of which carries a certain failure probability corresponding to the stress level to which it is subjected. When one of the volume elements is ruptured the whole specimen fails. Hence it follows that the strength will clearly be a function of size, as the larger the specimen is the greater will the probability be that a severe flaw will be contained in it. Weibull took the elementary laws of probability as starting point in order to develop a theory the formulae of which may be readily brought to agree with the test results. The assumptions necessary for deriving these equations are only that a distribution curve of strength values exists, that the material is considered as a conLinuum and LhaL the properLies of Lhe material are such that the probability of rupture starting at any

-24-

point is equal, that means that the materi4l is isotropic. Witn these assumptions and with the laws of probability the following general expression for the ultimate strength can be obtained

a,
0

= (T)(7)
1() 10,9 - 4

Here O11/is the volume element and

So

is the experimentally determined strength distri.

bution curve as a function of 6

If the material is such, that the risk of rupture is independent of the stress direction, then Y in the formula is a scalar quantity. For stresses distributed in an arbitrary way, G is to be replaced by another scalar quantity which is definitely determined by the three principal stresses. If 6 in the above equation denotes this comparison stress, the formula will be valid for any combination of stresses. The function #(8) has to be determined from experimen-

tal data and can most frequently be written as

(8)
By introducing (8) into (7) the ultimate strength is obtained

II

0 f 0

Instead of k we may introduce the constant 6(5 and we may put 7. VF/6 .Then the final

k-=

-25-

formula for tension tests is

-e
of radius r
,

(1)
6' can

For bending tests with circular round. bar specimen where f

= const the stress

be written as

S(11)

and the volume element

0V=

dV

as

' T

;r

(12)

Consequently if 0 (11)

in equation (8) is replaced by

we get from (7) with (8) and (12)

where

---

rI It is supposed that the power is a function of

the grainsize and the porosity and can be obtained from the strength distribution curve S. In order to avoid the enormous number of tests, which would be necessary to determine the strength distribution curve for each grainsize and porosity, bending tests were conducted with varying specimen volume for

-26-

'I

constant grainsize and porosity. In the first test series it was examined whether there is an influence on the test results by changing the specimen diameter or the specimen length. It turned out however, that only the volume of the specimen is decisive. It was altered within the range from 180 mm 3 to 3750 mm 3 . An example of a test diagram is given in Fig.39, where the fracture stress of specimens from a certain group of grainsize and porosity is plotted versus the specimen volume on a double logarithmic scale. According to I equation (7) it is possible to determine the power from the inclination angle of the f straight line given by

7IL= 7e

7( PI Y)/

(14)

As the values scattered very much a great number of tests were performed at the volume of approximately 780 mm 3 with 4 mm diameter Degussit A1 2 0 3 round bar specimens where the grainaize and porosity again showed a variety as in Fig.2. The usual bending test setup was maintained and the influence of grainsize and porosity on the fracture stress was determined in the same manner as described in chapter 2.2.5. The volume of the test specimens of this mentioned section was approximately 2390 mm5 and it was possible to calculate the value according to equation (14) by comparing the diagrams. The results of the bending tests are given in Table 17 and the dependency of the fracture stress from the porosity at a volume of 780 mm3 is given for several grainsizes in Fig.40.

-27-

I.

The diagram again shows that the fracture stress decreases monotonously with increasing porosity. However, the grainsize affects the influence of the porosity by a linear displacement and also by the curvature. The values are given in Table 16 and the inis

fluence of the porosity on the exporent plotted in Fig.41 for several grainsizes.

I The preliminary results show that the exponent W is very low for instance for grainsizes smaller than 6pn and for porosities below 2% and also for in connection / with - was large grainsizes beyond 20t4m

porosities above 15%. The highest exponent grainsizes of approximately 10 /A

determined in the range of 5 to 10% porosity and at

From Fig.41 an empirical equation can be established for 1if k with respect to the grainsize and the pcrosity. is replaced by the grainsize G in / which If this relation is inserted into equation (13) and was done in order to harmonize with the results of other investigations (/9/,/10/,/11/,/12/ and /13/) then this expression renders the size effect caused from the distribution curve. Equation (13), however, does not include the reduction of the effective crosssectional area by increasing porosity and the stressconcentration created by the pores. Both factors (I-p) and G(
,

where the stressconcentration

can be

determined empirically from Fig.38 or 40 with respect to the grainsize and porosity, must be incorporated in equation (13). This leads to the following expression for the influence of the grainsize, porosity

-28-

and the specimen volume on the fracture stress for a fourpoint loaded bending test of a round bar Alumina specimen:

, ,k
=48 kp/mm2

.(15)

where

IP
I
_

= = = =

grainsize in porosity in volumefraction 3 specimen volume in mm empirichl factor caused by the stressconcentration [046.A 2.9
+.0.OI

P()60uJM

Weibull exponent

= (I-op)(/bpC4o0.s) ' a. = .(* ( c7/4) 0 60

- o.322 ( G/(,, - i )o.,.8 e-

+0.4
/G.0 't) 4 0, 3i.

c= o.ozI (G/oa- ,o'-Oo.ss( 1)0-"

Equation (15) can also be written in terms of Ganmafunctions which are tabulated and facilitate the evaluation.

-Io"'"o -p= "

(

e-'/-'r(
V i,,

"

')
r'(-)

(1o
(IS-

r..M

As the bending tests were only conducted in the ranges covered with dots in Fig.2, equation (16) is valid in a porosity range from p = 2.27% to o = 19.5% and in a range of grainsize from

6.0

to

= 40.5 P.r".

-29-

2.2.7. Tension tests According to equations (10) and (13) the relation between the fracture stress from a tension test and from a bending test with specimens of )qual volumes and equal microstructure is: =
!0

,('17)
j

As was always positive in the tested region the fracture stress for tension must be smaller than that for bending. The results from bending tests therefore can give an estimate for the results of tension tests and this should give the starting point for further investigation. In testing of materials the tension test is a very important basis of research and engineering. However, it is well known that one of the main problems associated with tensile testing of brittle materials is the difficulty of applying a true axial load to obtain a uniform stress field in the test area. Considerable effort was applied in designing a tension test specimen which achieves an almost uniform stress field in the test area. For instance for a tension specimen with circular cross-section it can be shown from the simple beam theory that the percentage of increase in stress resulting from eccentricity is 8 times the r io of the eccentricity to the diameter of the rod. As only small specimens can be manufaotured an eccentricity of a few thousands of a millimetre can produoe an appreciable stress increase on one side of a tensile specimei which causes failure at lower loads

[F

-30-

than those corresponding to true axial loading.

4which

Therefore, a tension test device had to be developed excluded the eccentricity in the specimen and the bending moments caused by the clamps and the test machine. It is known, that the most accurate symmetry can be achieved with circular ground objects. Therefore, a specimen shape and a clamping device

Vwhich

considered.

consisted only of circular ground parts was

Fig.42 shows the proposed tension test specimen. Here it is of primary interest that the centeraxis of the ground rod coincides with a maximum discrepancy of 0,003 to 0,004 mm with the centeraxis of the drilled hole in the bushing on the front surface. This demand can easily be achieved by very careful grinding of the Alumina specimen and the bushing. Through the hole of the bushing exact fitting wires with sufficient strength will be put into the hollow spaces and fixed by filling it up with a low melting point alloy of lead, bismuth and thin. It is important that no play is allowed between the bushing and the specimen and that therefore the bushing has to be made of a material with very little temperature expansion. A normal steel would break the Alumina specimen during the preheating or the fillingup process. The thin wires with a length of approximaTely 50 cm on each side are clamped in the tension test machine. No remarkable bending moments in the specimen caused by the friction in the pins or the alignment of the tension test machine and so on are possible.

-31-

Two specimens have been manufactured and tested so far. Fig.43 shows the test specimen and Fig.44 the complete test equipment. An analysis of the stress field by means of electrical strain gages showed that the bending strain of the first specimen could be kept below 2% of the total strain during the whole test procedure. The manufacturing process can still be improved and it is expected to diminish the still existing errors further.

-32-

3.

References /l/ /2/ Griffith, A.A.: Phil.Trans.Roy.Soc. A 221 (1920) S.163-198 Weibull, W.: Ing. Vetenskaps Akad.Handl.No 151 (1939) No 153 (1939)Kungl.Tekniska Hogskolans Handl. no 27 (1949) Appl.Mech.Rev.(Nov.1952) 449 "Experimental Investigations of the Behavior of Brittle Materials at Various Ranges of Temperature", Contract AF 61/052 273 Final Report 30 April 1965 "Experimental Investigations of the Behavior of Brittle Materials at Various Ranges of Temperature", Technical Report AFML-TR-68-23 March 1968

/3/

Neuber, H.:

/4/

Neuber, H. a. Wimmer, A.:

/5/

Ryshkewitch,E.: Compression strength of porous sintered alumina and zirconia, 9th communication to ceramography J.Amer.Ceram.Soc.36 S.65-68 (1953) Sachs, G. a. Espey, G.: The measurement of residual stresses in metal, Iron Age (1941)

/6/

/7/

Williams in Kriegel, W.W. a. Palmour, H.: Mechanical Properties of engineering Ceramics, Intersc.Publ. New York, London 1961, Ch.18, 245-302 Orowan, E.: Petch, N.J.: Knudsen, F.P.: Fracture strength of solids Repts. Progr. in Phys.12 S.185-232 (1949) The fracture of metals, Progr. in Metal Physics 5 (1954) S.1-52 Dependence of the mechanical strength of brittle polycrys~.aline specimens on p'rosity and grainsize J.Amer.Ceram.Soc.42/376-387 (1959)

/8/ /9/ /10/

-33-

/11/

Petch, N.J.:

Cleavage strength of polycrystals J.Iron Steel Inst.(London) 174

Part I S.25-28 (May 1953) /12/ Spriggs, R.N.

a.Vasilos, T.: Effect of grainsize on transverse bend strength of Alumina and Magnesia, J.Amer.Ceram.Soc.46/ 224-228 (1963) /13/ Passmore, E.M., Spriggs, R.M. a.Vasilos, T.: Strength-grainsize-porosity relations in Alumina, J.Amer. Ceram.Soc.48/S.1-7 (1965) Neuber, H.: Stress and concentration design factors for grooved shafts. Unpublished diagrams.

/14/

-34-

j,

4.

Tbles and Illustrations Table 1 Chemical analysis of Degussit Al 23, Degussit Al 23 PT, Degussit Al 24 and Degussit test substance 2461 Al 23 Al 23 PT Al 24 test substance 2461 A1203 SO2 99.5% 0.05-0.1% 99.5% 0.05-0.1% 99.6% 0.05-0.1% 99.9% .0-0.02%

Fe20 3
MgO CaO Na 2 0

0.02-0.05% 0.02-0.05%
0.2% 0.05-0.1% 0.1-0.3% 0.2% 0.05-0.1% 0.1-0.3%

0.02-0.05%
0.01-0.02% 0.05-0.1% 0.1-0.3% 0.08-0.1% 0 0

-35-

Table 2 Distribution of the grainsize and the porosity over the length of Alumina rods. (Rods with a length of 240 mm were cut into 16 pieces. The numbers represent the grainsize and the porosity of each piece. The numerical order is arranged in the same way as the parts sticked together originally.) material and average porosity Porosity p Grainsize

Degussit Al 23

3.98 3.88 3.77

3.77

20.4 19.0 19.9

19.4

p
i

3.80%

3.85 3.85 3.80 3.93 3.95 3.99 3.87 3.74 3.93 3.99 4.00 4.24 Degussit Al 23 PT 7.37 7.71 7.55 8,25 7.79

16.7 17.3 19.0 19.0 18.0 21.7 19.2 20.9 20.2 17.1 19.4 23.7 28.1 29.9 25.9 25.3 27.5

7.58%

-36-

continuation of Table 2 material and average porosity Porosity Grainsize

7.94 8.07 7.84 7.56 7.72 7.06 7.31 7.86 7.61 7.65 7.12 Degussit Al 24 18.21 17.97 18.11 17.97 18.55 17.95 17.35 17.48 18.16 17.94 17.12 17.86 18.04 18.32 17.88 17.31 -37-

24.5 28.4 25.6 26.5 25.6 24.2 25.6 25.3 21.9 26.2 23.8 71.1 66.2 64.7 66.9 61.9 54.3 54.8 72.9 59.4 60.0 58.8 60.0 60.6 58.2 57.6 55.9

17.75%

cnntinuation of Table 2 material and average porosity Porosity p Grainsize

Wesgo Al 995

3.66

11.6

3.58
3.59 3.58 3.62 3.63 3.73 3.76 3.80 3.77

10.7
11.5 11.0 11.1 11.4 i0z5 11.0 10.7 10.9

3.80 3.84
3.82 3.86

10.9 10.2
10.1 11.1

3.94
3.82

10.7
11.6

-38-

Table 3 Dependency of the "Rauhtiefe" R , "Gldttungstiefe" RP and the Center-Line-Average CLA from the grainsize and the porosity of Degussit A1 2 03 round bar specimens. grainsize porosity Rauhtiefe Gl1ttungstiefe CenterLine-

RL

RP

Average
CLA

14.6 18.6 17.5 15.3 16.0 15.7

18.18 9.07 8.02 6.04 5.64 5.26

22.50 11.77 15.20 6.83 5.60 7.20

12.07 3.57 5.23 3.63 1.79 2.51

4.47 1.60 2.13 0.97 0.92 1.04

15.3
20.9 20.4 9.3

3.39
8.47 8.63 9.37

5.13
7.57 8.47 16.53

2.30
1.82 2.15 7.37

0.76
1.09 1.30 2.63

8.9 8.7 7.9

8.77 8.96 9.37

17.00 20.43 22.47

7.30 9.73 8.93

3.07 3.17 3.03

-39-

Table 4 Dependency of the Vickers-micro-hardness from the grainsize and the porosity of 4 mm 0 Degussit A1 2 03 round bar specimens. (The values of the Vickers-hardness are the arithmetic mean values of 10 hardness test on one cross-section.) grainsize porosity load diagonal Vickershardness kp/mm 1 8.39 7.85 7.b9 8.01 9.36 9.85 8.38 8.71 11.56 10.95 9.71 8.86 2687 3070 3199 2948 2159 1961 2690 2490 1418 1580 1042 1260

G
tIarnc

(%3
8.64 8.55 8.55 8.48 8.61 8.37 2.69 4.36 7.71 10.96 14.75 18.33

102 102 102 102 102 102 102 102 102 102 53* 53*

23.7 21.0 17.1 13.2 911 5.6 6.6 6.3 6.2 6.2 6.0 6.2

*) The low load was necessary, as the impression with the original load was larger than the free areas at specimens with high porosities.

-40-

Table 5 Vickers-micro-hardness measurements on single grains of different sizes. Determined on a Wesgo Al 995 rod 6 mm 0, porosity p = 3.70%.

(The values for the grains and the Vickers-hardness are the arithmetic mean values of 10 tests.) grainsize load diagonal Vickers-hardness HV

6 Lp)
12.2 20.4 30.2 41.5 50.0 94 94 94 94 94 11.48 -0.77 10.90 11.13 11.39

J[kp/mm i
1333

2-

1495
1473 1416 1355

-41-

__

Table 6 Variation of the damping factor of 4 mm 0 Degussit A1 2 0 3 specimens with the porosity and the grainsize. porosity grainsize damping factor

P
3.39 4.58 5.26 5.64 6.04 8.02 9.07 17.63 18.18

G
15.3 19.4 15.7 16.0 15.4 17.5 18.7 14.0 14.6

Sr
0.73"10 - 4 1.05.10 - 4 0.99.10 - 4 0.71.10 - 4 1.12.10 - 4 1.22"10 - 4 0.92-10 - 4 7.05-10 - 4 10.20-10 4

9.37 8.96 8.77

7.9 8.7 8.9

1.49.10

-4

1.15"10 1.08"10 - 4

9.37
8.02

9.3
17.5

1.1010 -4
1.22-10 - 4

9.07
8.63 8.47

18:"
20.4 20.9

0.92'10

1.05-10 - 4 1.25'10 - 4

-42-

Table 7 Dependency of the fracture stress, Young's modulus and the fracture strain from the surface condition of 6 mm 0 Degussit A1 2 03 specimens determined from a bending test at room temperature. Rauhtiefe Ri Gldttungstiefe RP CenterLineAverage CLA fraoture stress Young's modulus fracture strain

f~uL rn
17.17 14.70 14.03 9.77 8.80 8.60 4.77 4.90 4.43 3.10 2.23 2.18 1.62 2.02 2.05 1.96 2.10 1.26 1.55 0.84 1.69 1.77 1.45 0.91 0.92 0.46 i.44 0.74 0.61 0.80

4,v
2.53 3.07 2.30 1.43 1.44 1.22 1.00 0.90

kp/mm
26.7 23.9 27.5 26.5 23.6 23.8 25.9 23.0 28.7 27.0 25.8 27.6 29.8 24.7 29.0 26.2 27.9 29.3 27.6 25.3 29.0 25.1 27.2 27.2 -43-

kp/m

Pra2

%
0.700 0.565 0.725 0.770 0.650 0.655 0.685 0.640 0.765 0.725 0.690 0.730 0.810

3.68.104 3.73,104 3.70"104 3.64"104 3.65104 3.67104 3.68.104 3.68-10~ 3.70104 3.70104 3.64"104 3.71104 3.70104

7.66
7.13 7.03 7.03 6.77 6.43 5.90 5.70 4.93 4.87 4.57 4.43 4.30 4.27 4.20 2.77 2.70 2.07

112
1.14 1.04 0.85 0.88 0.70 0.80 0.79 0.71 0.58 0.57 0.42

3.70104 0.650 3.67-10~ 0.805 3.63"104 0.720 3.71-10 4 0.715 373 3.68-104 3.66-1o~ 3.70-104 3.68.104 3.72-104 0.775 0.760 0.690 0.770 0.685 0.690

0.64
0.29 0.32 0.20

3.73-10~ 0.740

-,

Table 8 Residual stresses of a 6 mm 0 Wesgo Al 995 specimen, grainsize * Fig.21) ground length Lmm] 199.9 remaining area [mm2 I#n 7 28.04 25.29 22.81 20.39 17.13 14.32 11.85 9.51 0.0 0.1 0.0 -0.2 -0.3 -0.4 -0.7 -0.7 change of length strain residual stress [kp/mm2

G = 17.2 pum

porosity p

= 3.99% (see

0.0000 0.0005 0.0000 -0.0010 -0.0015 -0.0020 -0.0035 -0.0035

0.33 -0.03 -0.22 -0.23 -0.13 -0.03 0.02 0.07

-44-

Table 9 Residual stresses of a 6 mm grainsize G = 16.5. (see Fig.22) ground

length

Wesgo Al 995 specimen, = 4.08% , porosity p

remaining
area

change of
length

strain

residual
stress

11mm)I
199.7

mjnm2.
28.04 25.42 22.98 20.43 17.31 14.55 11.76 9.56 0.0 -0.3 -0.4 -0.4 -1.0 -0.8 -1.2 -1.3

[%]IMMJ
-0.0000 -0.0015 -0.0020 -0.0020 -0.0050 -0.0040 -0.0060 -0.0065 -0.53 -0.38 -0.26 -0.15 -0.04 0.04 0.12 0.16

-45-

VN

Table 10 Residual stresses of a 6 mm 0 Wesgo Al 995 specimen, grainsize G = 16 .4"#n , porosity p = 4.09% (see Fig.23) ground length I'emaining area (m
2

change of length

strain

residual stress

[mm]
199.9

[~
0.0 1.2 1.8 1.9 2.0 2.6 2.6 2.3 0.0000 0.0060 0.0090 0.0095 0.0100 0.0130 0.0130 0.0115

te) kp/mm2]
2.44 1.18 0.51 0.08 -0.24 -0.40 -0.47 -0.48

28.04 25.34 22.90 20.47 17.31 14-.59 11.73 9.48

-46-

Table 11 Dependency of the fracture stress, Young's modulus and the strain at fracture from the strain rate of 6 mm 0 Wesgo Al 995 specimens at room temperature determined from a bending test. porosity grainsize strain fracture Young's fracture rate stress modulus strain

L%J
3.826 3.703 3.789 3.593 3.704 3.756 3.764 3.807 3.713 3 3.553 3.708 3.708 3.701 3.543 3.800 3.721 3.799 3.865 3.778 3.709 3.691 11.88 11 99 12.20 11.94 12.43 12.26 12.15 12.49 11.15 11.06 12.04 11.94 12.37 13.02 12.37 12.12 12.26 11.48 12.78 11.63 11.78

[ seec] -12k2 kj
mm 4.38 4.60 21.44 23.53 22.83 21.57 24.76 25.17 24.43 27.17 22.07 21.44 24.60 27.65 25.65 26.30 26.41 26.72 27.91 24.90 26.09 27.68 28.06

(7B

10-4.E

[2]
mm 3.640 3.759 3.669 3.729 3.713 3.577 3.661 3.690 3.716 3.613 3.713 3.680 3.703 3.688 3.722 3.622 3.680 3.730 3.748 3.771 3.713

o f %]
0.5689 0.6137 0.6156 0.5512 0.6370 0.6689 0.6470 0.7227 0.5854 0.5594 0.6434 0.7346 0.6891 0.6914 0.6927 0.7825 0.7203 0.6628 0.6657 0.7-;8 0.7362

cB

4.17
4.38 45.3 49.3 48.2 47.8 48.6 46.7 482 450 482 470 468 452 437 4399 4500 42401 r A 4520

-47-

continuation of' Table 11 porosity grainsize strain fracture Young's fracture modulus strain stress rate

603i

4 .E 10-

6
[%0j 0.7300 0.7246 0.7670

E%
3.907 3.708 3.646

Y
'12.04

[P
--

fP-] h 28.62 27.94 29.25

[kp 3.889 3.667 3.731

12.04 12.15

4340
4270

-48-

Table 12 Dependency of the fracture stress of 5 mm 0 smooth Wesgo Al 995 specimens from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature. (Average grainsize G = 20.5 1 , average porosity

/0= 3.68%)
diameter bending stress [kp/mm 2 l 25.15 24.72 static bending test static bending test numbers of cycles to fracture

[mm 3 4.996 5.012

4.993
4.993

20.00
18.00

9.800
121.500

4.978
5.010
4.915

16.00
14.00
10.00

43.300
118.200 452.500 1.110.900

5.008

11.00

4.988 4.979
5.011 5-C3

8.00 9.00
7.00 6.50

2.922.300 5.478.400
5.450.300 no fatigue fracture

-49-

Table 13 Dependency of the fracture stress of notched round bar Wesgo Al 995 specimens (nominal diameter 5 mm) from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature. (Average grainsize 20.0&4m , average poi.'sity / = = 1.54, 3.68%, theoretical notch strength ratio according to Neuber /14/) minimum diameter nominal bending stress numbers of cycles to fracture

dbW
(mm

N Lkp/mm2 il
19.74 18.57 16.49 13.56 11.30 static bending test static bending test 2.000 8.800 29.400

4.978 4.990 4.950 4.935 4.955

4.948 5.008 5.045 4.970

5.013 5.095 5.010

12.00 11.00 9.73 10.18

8.92 8.03 9.44

154.900 95.300 711.000 no fatigue fracture

no fatigue fracture no fatigue fracture no fatigue fracture

-50-

Table 14 Dependency of the fracture stress of 5 mm 0 smooth Degussit Al 23 specimens from the numbers of cycles to fracture determined from alternating bend fatigue tests at room temperature. (Average grainsize p = 3.79%) diameter bending stress numbers of cycles to fracture G = 20.1m , average porosity

fIm7]
4.970 4.987 4.972

Lkp/mm2
21.24 21.02 17.62 900 1.400 18.400

4.958 5.007 4.988 5.002 5.022

13.03 11.85 9.36 6.65 5.47

50.000 147.600 117.300 1.953.800 no fatigue fracture

-51-

Table 15 Dependency of the fracture stress of notched round bar Degussit Al 23 specimens (nominal diameter 5 mm) from the numbers of cycles to fracture determined from alternating bend fatigue tests at room i;emperature. (Average grainsize = 21.6 pm , average porosity p = 3.68%, theoretical -notch strength ratio G = 1.54, according to Neuber /14/) minimum diameter d nominal bending stress 'bW numbers of cycles to fracture N

MI
5.017 4.995 5.135
5.022 4.850 5,103 5.090 4.970

kp/mm21 [
18.38 15.77 13.24
11.10 9.29 8.43 6.72 5.15

1.300 3.900 28.500

109.600 112.000 914.000 1.108.300 no fatigue fracture

-52-

Table 16 Dependency of the fracture stress, Young's modulus and the fracture strain from the porosity and grainsize of 4 mm 0 Degussit A12 03 specimens determined from a bending test at room temperature. grainsize porosity fracture stress [% kp/mrn 2] Young's modulus [kp/mm 2J fracture strain [%0]

[ 4,$J

8.2
17.9 6.6

2.33
2.21 2.69

27.0
26.5 27.0

40 210
40 125 39 730

0.672
0.661 0.679

8.2 8.8 18.0 18.4 18.8 19.4 23.1 6.3 6.8 7.4 7.9 20.7 24.3 25.0 6.9 8.4
12..7

2.89 3.31 2.87 2.93 2.87 3.29 2.67 4.36 4.92 4.29 4.02 4.35 4.38 4.33 6.02 5.10
5.45

27.8 25.9 22.5 23.0 25.2 22.7 19.2 24.7 27.3 24.1 22.5 23.7 19.6 18.4 26.6 26.2
18.6

39 37 39 40 39 44 39

745 732 800 032 718 300 948

0.699 0.686 0.564 0.574 0.635 0.525 0.481 0.656 0.727 0.617 0.574 0.604 0.509 0.468 0.725 0.699
0.517

37 641 37 580 39 038 39 39 38 39 230 200 455 100

36 578 37 427
36 055

-53-

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

p
f%)

6
30.4 27.6 23.8 15.3 16.5 23.5 22.5 19.7 26.2 26.9 26.3 25.3 26.8 29.3 27.9

C
35 981 35 708 36 633 36 400 35 900 35 034 34 914 36 600 34 420 34 240 34 314 34 694 34 470 35 000 34 477

6B
&o
0.844 0.773 0.648 0.420 0.457 0.669 0.642 0.535 0.758 0.784 0.764 0.728 0.776 0.834 0.807

[kp/Tpxan2 J jip/mm']

6.1 6.6 6.8 15.4 20.7 6.5 7.6 23.1 5.0 5.4 5.6 5.7 5.7 6.2 6.4

6.55 6.69 6.38 6.60 6.60 7.43 5.91 7.05 8,22 8.35 8.37 7.98 8.05 7.71 8.13

6.4 10.0
12.2

8.07 8.40
7.80

26.2 17.6
20.3

34 808 34 997
38 700

0.752 0.505
0.522

13.2 15.3 15.4 16.3 19.0

8.48 8.50 8.30 8.39 8.49

16.3 17.3 15.0 15.9 15.7

34 35 35 35 34

881 082 199 370 854

0.470 0.495 0.427 0.450 0.450

-54-

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

P
[/P1 %]

%
(kp/mm 2 J
29.0

E
[kp/mm2 )1
34 210

[%01
0.845

5.7

8.60

9.1
9.6 10.3 12.0 13.3

8.61
8.62 8.61 8.79 8.81

16.6
17.6 14.2 15.8 12.7

34 858
34 387 35 260 35 275 35 368

0.477
0.513 0.404 0.449 0.360

13.9 14.1 14.2 14.3 14.4

8.58 8.88 8.77 8.79 8.95

15.0 15.7 13.1 16.6 14.1

34 35 35 34 34 35 34 35 34

843 251 644 805 570 082 500 718 770

0.431 0.445 0.369 0.478 0.409

14.9
15.3 15.4 16.7 17.1
18.5

8.87
8.50 8.82 8.85 8.55
8.86

15.2
17.3 14.6 13.1 17.6
12.4

34 770

0.436
0.495 0.424 0.368 0.508
0.357

34 721

18.8 19.0 19.0 19.7 19.9 19.9 20.5 21.0 21.2 23.3

8.64 8.54 8.77 8.77 8.72 8.63 8.80 8.55 8.84 8.85

14.7 15.6 15.1 14.4 16.1 16.7 14.9 11.6 16.6 11.5 -55-

34 825 35 106 34 552 35 35 34 34 174 198 713 380

0.422 0.446 0.437 0.410 0.460 0.481 0.437 0.335 0.481 0.337

34 747 34 652 34 162

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

G
L ,23.7 28.6 5.4 14.4 15.5 15.9 16.3 16.5 16.5 16.8 18.8 19.3 19.9 20.0 20.1 23.1 28.6 4.5 4.7 5.1 5.4 5.9 6.3 44.5 47.1 47..8 51.6

p
(%7
8.64 9.00 9.51 9.05 9.40 9.07 9.32 9.15 9.32 9.08 9.16 9.25 9.35 9.02 9 21 9.37 9.00 17.70 17.24 17.52 17.71 17.30 17.46 17.68 17.52 17.25 17.69

LB
[kp/mm2j 15.7 14.1 22.8 14.9 12.3 13.2 15.6 17.9 17.4 13.0 12.3 12.2 13.4 13.1 14.0 17.3 14.1 20.5 20.2 21.3 21.4 17.5 20'.8 11.8 11.8 9.7 11.0 25 990 26 840 26 900

fkp/mm22
35 169 34 400 34 34 34 35 34 34 34 34 34 34 34 34 800 741 757 248 710 480 590 160 396 177 850 404

[% o 0.447 0.108 0.654 0.430 0.355 0.377 0.449 0.518 0.502 0.381 0.357 0.358 0.385 0.382 0.406 0.495 0.408 0.805 0.774 0.839 0.835 0.677 0.805 0.371 0.454 0.360 0.410

34 646 34 900 34 400 25 510 25 933 25 370 25 368 25 632 25 674

-56-

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

G
l/m3
4.2 4.3

6B

eB
2j

J
18.14 18.15

[kp/m
20.8 19.5

[kp/mm2J
24 836 24 685

[%o]
0.830 0.782

4.4 4.9 5.5

6.1

18.25 17.82 17.82

17.85

20.0 20.7 20.7

18.8

24 440 25 400 25 221

25 143

0.816 0.815 0.816

0.740

6.1 6.3 39.6 40.4

45.6 51.6 66.:

18.33 17.84 18.02 18.00

17.86 18.20 17.92

21.2 20.1 9.5 9.7

11.0 11.4 11.1

24 25 27 26

366 403 330 400

0.864 0.787 0.347 0.360

0.440 0.426 0.408

25 100 26 700 29 500

4.4 4.4 4.7 4.7

4.5

12.50 13.80 13.88 17.24

17.70

24.0 24.6 22.5 20.2

20.6

30 220 29 455 29 359 25 933

25 510

0.795 0.829 0.763 0.774

0.805

4.2 4.3
4.4 4.4

18.14 18.15
18.25 19.04

20.8 19.5
20.0 18.9

24 836 24 685
24 440 24 021

0.830 0.782
0.816 0.781

4.3 4.6 4.3 4.4 4.5

19.10 19.37 19-44 19.52 19.87

19.3 18.3 19-0 18.15 18.3

23 898 23 400 23 247 23 320 23 235

0.800 0.783 0.806 0.778 0.780

-57-

continuation of Table 16 grainsize porosity fracture stress 3e Youngts modulus fracture strain

Gp

CB

5.0 5.4 5.6 5.1 5.4 5.0 5.0 5.4 5.5 5.0 5.0 5.1 5.5 5.1 5.1 5.5 5.0 5.4 5.4 5.4 5.6 5.5 5.3 5.0 4.9 5.1 5.1

8.22 8.35 8.37 10.83 10.91 11.15 11.33 11.48 11.51 11.52 11.63 11.68 11.69 11.70 11.75 11.89 11.99 12.10 12.21 12.74 13.28 14.27 14.31 14.45 14.82 15.07 17.52

26.2 26.9 26.3 18.7 24.1 25.7 23.9 25.2 23.7 22.3 24.0 26.8 23.1 25.7 23.9 25.1 25.0 25.6 23.5 24.1 23.3 24.0 23.7 22.2 21.5 23.6 21.3

34 420 34. 240 34 32 31 31 31 31 31 29 31 31 31 31 31 31 32 30 31 30 314 026 619 700 539 513 661 141 493 315 449 100 167 288 170 770 039 213

0.758 0.784 0.764 0.581 0.758 0.810 0.755 0.794 0.745 0.758 0.759 0.852 0.731 0.825 0.762 0.778 0.773 0.831 0.754 0.793 0.782 0.838 0.828 0.772 0.765 0.852 0.839

29 650 28 486 28 492 28 645 28 130 27 660 25 370

-58-

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

G
(,uM/
5.4 4.9

p
[%
17.71 17.82

4C
k-P/mm2
21.4 20.7

5B
2

kp/m

[%o]
0.835 0.815

25 368 25 400

5.5 5.2 6.1 6.6 6.5 6.2 5.7 5.7

6.4 6.4

17.82 19.30 6.55 6.69 7.43 7.71 7.98 8.05

8.07 8.13

20.7 20.4 30.4 27.6 23.5 29.3 25.3 26.8

26.2 27.9

25 221 23 611 35 35 35 35 34 34 34 981 708 034 000 694 470

0.816 0.857 0.844 0.773 0.669 0.834 0.728 0.776

0.752 0.807

808 34 477

5.7 6.0
6.4 6.0

8.60 9.84
10.27 10.73

28.9 26.5
27.0 26.2

34 210 32 712
32 590 32 590

0.845 0.806
0.829 0.800

5.7 5.8 6.2 5.7 6.0 5.8 5.7

6.3

10.74 10.82 10.96 10.97 11.54 11.65 12.38

12.85

25.6 24.7 23.7 26.6 24.32 23.9 24.3

23.3

32 32 31 31 31 31 30

099 190 773 770 331 267 339

0.795 0.766 0.743 0.835 0.772 0.759 0.796

0.768

30 204

6.4

12.85

25.3

29 424

0.860

-59-

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

p
L%

(5
[kp/mm2]
24.2 23.7 24.1 21.3 17.5 20.8 20.1 18.7 21.2 20.3

Ir
[kp/nm 2
29 304 28 180 27 27 25 25 928 393 632 674

6B

6.3 6.0 5.7 5.7 5.9 6.3 6.3 6.1 6.1 5.7

13.61 14.75 15.22 15.49 17.30 17.46 17.84 17.85 18.33 19.25

0.820 0.836 0.857 0.772 0.677 0.805 0.787 0.740 0.864 0.852

25 403 25 143 24 366 23 594

5.7 6.8
6.9 6.8 7.4 8.2

20.49 4.92
6.02 6.38 10.70 2.33

18.8 27.3
26.6 23.8 20.2 27.0

22 688 37 580
36 578 36 633 32 170 40 210

0.820 0.727
0.725 0.648 0.629 0.672

8.2
8.8 8.1 8.1 7.9

2.89
3.31 3.67 3.79 4.02

27.8
25.9 26.6 23.4 22.5

39 745
37 732 38 575 38 526 39 230

0.699
0.686 0.688 0.607 0.574

7.4 7.6 8.7 7s5

4,29 6.91 9.65 9.67

24,1 22.5 20.3 21.3

39 038 34 914 34 360 33 800

0.617 0.642 0,92 0.634

-60-

continuation of Table 16 grainsize porosity fracture stress Youngts modulus fracture strain

G
i~rn
7.9

p
1%3 E7
10.36

fE
(kp/m
20.1
2J

[kp/mm2 J
33 800

[o
0.594

7.4 7.5
9.6 10.0

10.70 10.89
8.62 8.40

20.2 17.4
17.6 17.6

32 170 31 997
34 387 34 997

0.629 0.540
0.513 0.505

10.3
16.3 17.1

8.61
8.39 8.55

14.2
15.9 17.6

35 260 35 370
34 770

0.404
0.450 0.508

16.7
16.8 16.5 16.5 16.3

8.85
9.08 9.15 9.32 9.32

13.1
13.0 17.9 17.4 15.6

35 718
34 160 34 480 34 590

0.368
0.381 0.518 0.502 0.449

34 710 39 520
44 34 34 35 300 854 713 198

19.5
19.4 19.0 19.9 19.9

3.25
3.29 8.49 8.63 8.72

23.7
22.7 15.7 16.7 16.1

0.601
0.525 0.150 0.481 0.460

19.7 19.0
20.0 19.3 19. 23.1 23.1

8.77 8.77
9.02 9.25 9 2.67 7.05

14.4 15.1
13.1 12.2 15.4 19.2 19.7 -61-

35 174 34 552
34 404 34 177 5 8 39 948 36 600

0.410 0.437
0.382 0.358 0.385 0.481 0.535

continuation of Table 16 grainsize porosity fracture stress Young's modulus fracture strain

G
4IIUm
23.7
23.3

B
[%] ip/mm2J 15.7
11.5

F
[kp/rx.
35 169
34 162 ~

C-B
o
0.447
0.337

IA
8.64
8.85

23.1 28.6 27.4

9.37 9.00 9.62

17.3 14.1 13.1

34 900 34 400 34 650

0.495 0.408 0.379

-62-

Table 17 Deperdency of the fracture stress from the specimen volume, the poro3ity and the grainsize, determined from a bending test at room temperature. volume porosity grainsize fracture
stress

3) (mm

(M
3.53
3.76

(
14.3
16.6

(kp/mm2)
27.53 25.44 34.39
30.95 35.50

180
180

180
178 183

1.92
2.26 2.20

13.4
12.2 12.5

174 375 369 378 374 372 375 387 374

386

2.70 3.35 3.60 3.30 1.70 3.18 3.43 3.50 3.28

3.43

9.9 14.4 13.9 12.1 11.6 11.6 14.1 12.1 11.8

12.9

34.48 22.89 23.77 27.37 30.31 26.62 22.10 26.28 27.81

28.60

373 364
380

1.63 1.40
1.63

12.1 12.0
10.8

27.95 29.64
29.56

382 381 382

378

370
378

1.41 1.78 1.55 1.53 1.73

1.13

12.3 12.5 14.5 13.1 10.4

12.1

29.28 28.67 26.30

28.89

30.84
30.23

-63-

!,'I

'S

continuation of Table 17 volume rorosity grainsize fracture stress

!(rmM3)
377 409 415 407 413 419 408 409 419 416 412 408 416 422 410 413 413 410 407 729 721 710 719 714 718 724

M%
3.33 2.25 2.04 2.70 1.88 1.78 2.98 2.64 3.45 2.80 3.10 4.20 2.26 2.35 3.74 3.58 3.58 3.30 3.30 3.00 8.65 8.99 9.07 9.07 8.59 8.61

(/M
14.1 18.2 16.6 22.2 16.6 16.6 9.5 12.5 20.4 26.4 16.0 12.4 22.8 15.2 18,8 11.2 11.2 9.5 9.5 1.8 5.9 6.2 6.2 6.2 6.4 9.1

(kp/mm2 )
21.35 26.22 25.35 25.42 23.81 23.39 34.60 36.41 25.40 24.35 28.15 28.05 25,98 22.25 24.53 30.15 29.07 31.11 35.47 31.93 26.86 33.00 29.89 30,15 26.36 18.37

-64-

continuation of Table 17 volume porosity grainsize fracture stress

(mm
695

W
7.60

(G M
11.0

(kp/mm2 )
23.20

721 707 724 721

720

8.77 8 79 8.30 8.85

8.64

14.2 1A.3 15.4 16.7

18.8

15.49 17.11 16.99 16.97

17.73

716
723 720 697

8.77
8.72 8.84 8.42

19.7
19.9 21.2 27.3

16.73
16.20 15.84 20.45

746

4.03

1.4

42.04

749
755

4.35
3.93

1.4
1.5

40.28
35.43

764
768

5.77
6.13

1.5
1.5

34.73
39.60

752 753 736

741

3.60 3.95 2.53

2.20

1.7 1.9 2.0

2.0

42.17 40.77 45.06

40.85

742
751

2.30
13.08

2.1
5.2

39.93
26.81

730

12.85

6.4

28.34

746
730
731 729 742

4,29
8.62
8.40 8.61 8.78

7.4
9.6
10.0 10.3 12.0

26.48
19.12
18.10 21.05

-65-

continuat.on of Table 17
volume porosity grainsize fracture

-tress

V
(MM3)
748
743

p
M%
4.15
7.80

G
,m
12.0
12.0

(kp/im 2) 25.11
21.80

748
748 758

5.45
4.68 6.50

12.7
12.8 12.8

21.51
27.61 22.13

765
730 736

6.40
8.48 8.81

13.0
13.2 13.3

19.31
19.32 15.28

732
743

8.58
8.88

1M,9
14.1

17.53
20.13

746 740
762

4.88 9.05
6.20

14.3 14.4
14.6

23.01 16.30
22.39

731 743

8.87 9.40

14.9 15.5

15.64 15.10

734 743
750

8.50 8.82
6.40

15.3 15.4
15.4

16.62 14.86
22.17

735 740 734 736

762

9.07 9.32 9.32 9.33

2.07

15.9 16.3 16.3 16.5

16.6

14.09 17.71 18.84 15.89

25.70

743 747
768 732

9.0 3.20
6.48 8.55

16.8 17.1
17.1 17.1

16.18 22.84
21.76 19.65

757
757

6.00
2.88

17.6
18.0

21.14
26.38

-66-

continuation of Table 17 volume porosity grainsize fracture stress

p
(3mm
730 751 754 736

(I)
8.86 9.68 6.38 9.15 18.5 18.5 18.7 18.8

(kp/mm2 )
15.58 14.99 20.94 14.25

739
736

8.49
8.77

19.1
19.1

17.01
19.10

747 733
731

9.25 8.63
9.35

19.3 19.9
19.9

14.91 16.37
12.88

743
739

9.02
9.21

20.0
20.1

13.89
15.76

740
, 751 736 763 743 761

8.80
6.60 8.55 7.65 7.75 6.30

20.5
21.0 22.0 22.2 22.2 23.1

17.34
15.94 16.41 18.42 18.99 21.60

755
761 752 741

6.90
6.97 2.67 8.85

23.1
23.1 23.1 23.3

19.22
19.59 25.27 15.13

736 J
746

8.64
4.25

23.7
24.0

19.68
21.99

742 761 753

757

4.38 2.57 6.75

7.70 -67-

24.8 25.0 25.0

25.0

20.49 24.93 19.73

15.87

continuation of Table 17 volume porosity grainsize fracture stress

__(mm

3)

M%
8.15 4.00 6.63 5.75 7.00 9.00 5.15 7.25 13.80 13.88 10.89 10.83 11.33 11.74 14.31 12.30 12.74 11.48 11.69 13.28 10.74 7.98 8.05 12.38 11.10 10.56 11.54

(M #V) j
26.1 27.3 27.3 28.6 28.6 28.6 33.3 33.3 4.4 4.7 4.9 5.0 5.0 5.1 5.3 5.3 5.4 5.4 5.5 5.6 5.7 5.7 5.7 5.7 5.8 6.0 6.0

(kpimm2)
19.18 22.98 18.37 18.93 19.44 13.91 19.03 17.98 25.84 25.34 23.95 24.97 24.48 25.56 26.01 22.92 25.59 26.03 24.82 24.24 27.93 29.01 29.20 25.i3 26.57 25.63 27.29

761 749 751 760 752 737 756 758 829 826 811! 815 812 806 816 801 822 816 813 826 809 797 797 820 810 804 617

-68-

continuation of Table 17 volume porosity grainsize fracture stress

3) (mM 811 818 823 791 805 790 794 790 812 815 814 810 817 799 805 792 801 799 793 804 799 803 824

p
(M
10.96 13.61 12.85 7.43 10.40 9.35 9.66 9.65 9.88 18.95 10.30 18.70 19.35 18.20 18.03 18.00 17.60 17.70 18.10 17.80 17.80 17.90 18.25

G
(
6.2 6.3 6.4 6.5 7.5 7.8 8.5 8.7 9.4 9.5 10.0 16.3 19.3 38.8 39.6 40.4 40.8 40.8 42.1 42.6 43.1 43.5 44.0

(kplmm2 )
25.92 26.01 24.75 30.05 22.36 20.37 19.19 19.80 18.14 13.11 20.11 14.36 13.94 10.12 10.57 10.90 11.03 11.94 10.89 10.62 10.96 10.04 8.76

793
791 795 806

17.80
17.60 17.80 18.40 -69-

44.5
46.1 46.1 46.1

12.53
12.28 11.37 9.27

Ti

continuation of Table 17 volume porosity grainsize fracture stress

(mm3)
799 789 808 792 802 805 790 793 791 793 802 789 794 793 795 828 829 820 844 837 870 855 871 854 857 871 862

W%
17.70 17.52 18.30 17.25 17.80 17.97 18.00 17.80 18.20 18.20 17.80 18.20 17o70 17.70 18.30 17.40 17.92 17.87 18.14 18.15 19.10 19.04 19.52 18.25 17.70 19.87 19.37

(/UM)
46.7 47.1 47.2 47.8 48.4 48.4 48.4 49.0 49.0 49.6 50.2 51.6 52.3 53.0 54.4 60.9 66.0 69.9 4.2 4.3 4.4 4.4 4.4 4.4 4.5 4.5 4.6

(kpl/21
10.70

11.13 10.03 11.33 9.90 10.06 10.60 '12.27 11.19 10.63 9.73 10.77 10.35 10.72 10.72 10.80 11.11 11.44 21.43 19.56 18.95 22.46 19.08 20.26 19.31 19.41 19.29

-70-

cQntinuation of Table 17 volume porosity grainsize fracture stress

31 (mm
851 854 864 851 865 865 871 850 855 859 852 1205 1199 1204 1196 1193 1195 1186 1199 1204 1188 1204 1202 1187 1201 1199 1194

%'%)
17.82 17.52 19.30 17.71 19.39 19.25 20.49 17.30 17.85 18.33 17.46 3.19 3.19 3.67 3.99 3.60 3.60 3.26 3.16 3.65 2.98 2,91 3.05 2.66 3.42 3.19 3.27 -71
-

66
5.0 5.1 5.2 5.4 5.7 5.7 5.7 5.9 6.1 6.2 6.3 11.8 11.8 12.0 12.8 15.0 15.0 12.4 12.7 16.0 12.4 14.3 13.3 11.1 i5.4 16.5 12.6

(kp/mm2 1
21.48 22.28 22.33 20.87 20.91 21.17 19.54 23.27 21.66 20.82 23.14 26.93 26.66 24.27 22.54 26.30 26.51 31.04 26.60 27.32 31.95 27.49 27.41 29.04 10.13 25.52 31.15

continuation of Table 17 volume porosity grainsize fracture stress

3) (mm
1199 1203 1191 1599 1591 1601 1613 1605 1582 1677 1677 1672 1667 1658 1658 1670 1662 1667 1691 1682 1667 1654 1665 1657 1779 1801

M%
3.26 3.26 3.23 3.22 3.40 3.11 2.81 2.48 2.39 4.30 5.59 5.59 5.59 4.20 4.20 3.89 3.74 3.74 4.22 7.79 3.82 3.81 3.49 3.94 3.04 2.97

gO

(/A W
14.9 14.9 20.1 14.6 18.3 14.2 22.5 18.2 15.4 13.8 18.6 17.0 17.0 10.8 10.8 7.1 12.1 12.1 13.5 23.3 14.3 18.2 15.4 8.0 19.0 23.3

Gb

(kplmm2 )
28.49 25.55 28.52 22.66 22.15 22.23 21.62 23.05 27.60 20.83 21.87 23.49 25.97 27.81 23.12 26.56 21.28 25.60 21.34 23.66 24.03 25.41 25.52 23.78 13.79 18.94

-72-

continuation of Table 17 volume porosity grainsize frac'ure stress

V
3) (mM

P
2.48 2.61 2.62 2.68 2.79 2.70 3.04 3.00 3.46 3.48 3.12 3.56 3.56 3.49 3.06 3.65 3.65

9
(um)
15.2 18.7 18.3 19.6 17.8 24.6 18.5 27.2 26.4 21.3 25.4 23.0 23.0 22.2 20.0 26.9 24.8

(kp/mm)
20.14 20.81 18.15 20.17 20.89 21.49 19.55 21.84 20.56 20.65 16.62 21.07 21.71 18.50 21.53 20.00 18.25

1793 1793 1781 1781 1808 1781 1786 3000 3022 2997 2999 3015 3010 3029 3016 3029 3036 3064 3059 3049 3047 3092 3076 3091 3778 3748

I3036

3.06
3.06 3.06 3.73 3.34 3.17 3.93 3.57 3.19 3.27 -73-

20.0
22.0 22.0 20.0 22.2 22.2 18.2 22.9 16.5 12.6

20.11
23.17 22.34 18.07 17.97 21.77 18.07 20.17 24.42 28.00

continuation of Table 17 volume porosity grainsize fracture stress

(amM3 )
3777 3754

W%
3.26 3.23

&
'
14.9 20.1

(kp/mm 2)
22.25 26.60

-74-

Table 18 Dependency of the exponent from the porosity and grainsize, determined at room temperature grainsize porosity m

G
6.0

P
6.0 9.0 12.0 15.0 18.0 0.064.5 0.0612 0.0671 0.0736 0.0764 0.0839 0.0830 0.0786 0.0790 0.0935 0.0964 0.0897 0.1080 0.0982 0.0982 0.1039 0.1188 0.1241 0.1384 0.1384 0.0691 0.1072 0-1155

7.0

4.0 6.0 9.0 12.0

7.96

2.0 4.0 6.o 9.0 12.0

9.86

16.6

4.0 6.0 9.0 12.0 15.0 18.0 2.0 4.0 6:0

-75-

continuation of Table 18 grainsize porosity m

9.0 12.0 15.0 18.0 19.5 4.0 6.0 9.0 12.0 15.0 18.0 23.3 2.0 4.0 6.o 9.0 4.0 6.0 9.0 17.5 4.5 8.8 9.3 5.6 7.3 18.5

0.1108 0.1141 0.0975 0.1150 0.1146 0.1139 0.1195 0.1110 O.i060 0.1108 0.0652 0.1080 0.1070 0.1045 0.1130 0.11-30 0.1120 0.0125 0.0937 0.0952 0.0791 0.1125 0.10424 0.0085

28.0

40.5

-76-

*1k

117
Fig.1. Polished and etchlled Alumina speoimens ; porosity p 6.2 (grainsize magnification 500-.times)
=2.33%;

-77

4-)

4i-1
0

(a)
0 0

A D)

P A

D) a)

____

___

____ __

___

___

'Ia)
** .

* * *~~~.*
S

02

*S

a)

*r

-78-*

7~

~ %_______

~0

_____

75

5 .o - ..- . -i

o-- --

p--

! mm
0 '

s0

160

240

100

f GPM

75

50

25
0

-00_-0n 0

>11
0

80

160

240

Fig.3

D'stribution of the porosity and the grainsize over the length of a 4 mm diameter Degussit Al 23 specimen

-79-

10

__

__-_

__

__._

10
0 0 0

2,5
1 mM 0 0
100 G /""M
-

80

160

240

75

50

25

0
_
0

0
O.
-_ _

J--"---

0r-,-.m0-- -0,

Smm

0'
0 Fig.4 80 160

240

Distribution of the porosity and the grainsize over the length of a 4 mm diameter Degussit Al 23 PT specimen

-80-

20

P_

0,00

0
0

00

175

15

12,5
I 100 0 so 160 240

(10
75
0

41m

80

160

240

Fig.5

Distribution of the porosity and the grainsize oyer the length of a 4 mm diameter Degussit Al 24 specimen -81 -

75

5,0

2,5

1 mm

80

160

240

30

20

10

o'=

"

--

"

""-

-o --'

'o --

- -

,.--.

___--o-.-0

0mm
I MM

80

160

240

Fig.6

Distribution of the porosity and the grainsize over the length of a 6 mm diameter Wesgo Al 995 specimen
-82-

~W1

, 2.69

9iA
436%

6. .55%

p=7.71%

p=8.37%

p=10.7%

P= 14.75%178
For
comparison

p=17.4%
G 60. 9 r

Fig.7

Dependency of the poresize and poredistance from the porosity at constant grainsize G''6.1,pm

-83-

". '
Vo
9
,

* ..*4
*1e
.w 0

#.vi

-, o. 3

6=

1.54

6.031u
cV =

.~

u~I
1

,~,D A,~ A

'

.7'

4*4

17.

A:*
20.0

3.3
12.3ju

comparison1
~
*~A.*27%

For

Fig. Deedec

oth oei -84- adprista from teganieacosatprsyp,56

i~'~L~ *-84-

-j-A

IsI
70

20_______

Ps'02 " 3 %

R/ MnJG,1',36-86S5

.,'

to

5 (,tut) n 0 0 10 20

5 P1%) 0 0 10 20

72

h~""~'12
P .02 - 9,37 %

IRA7I

I
G ,36, F865 Pm"

10

0~~

G","P(

-"

t0

20

t0

20

Ct.,Atum/

CLAIpm,)

P 8,02 - 9,37% 5

0 '14,36-18,65 ,mrn

o 0
!

r_ I

I,
00 10 20

. PM

0 0 to 20

Fig-9

Dependency of the "Rauhtiefe" Rt, "Gldttungstiefe" R and the Center-Line-Average CLA from p the grainsize and the porosity of 4 mm diameter Degussit A1 2 0 3 specimens -85-

HV10'

(kp/mn? )

px837-864~ M%

G (urn)
0 10 10-3M (kP/lrn2) 20 30

.3

p(01 0)

10

20

Fig.10

Dependency of the Vickers-micro-hardness on A. I -o-^s4+ir 0 o Deguss8t A1203 round bar specimens. (The points are the arithmetic maean values of 10 hardness tests on one cross-section) -86-

I
I
47750

HV(,)

7500

1120
~~~1000 0 Fig.11 ..

.. . 20 40 60

Dependency of the Vickers.-hardness from the size of single grains of a Wesgo Al 995 rod 6 mm 0, porosity 7.7%. (The points are the arithmetic mean values of 10 tests)

Fig.12

Vickers-impression on an Alumina specimen. Pressure of the diamond pyramid 94 P. Magnification 1400 x

-87-

Fig.13

Droken pieces of an Alumina rod photographed shortly after a bending test with a Hitachi high speed motion camera (Picture taking rate: 10 000 pictures per second)

~ser e'-as--t*o-a

for the determination

of the damping constant 6

-88-

70
G 14,0 19,4 /im

75 5.0-

50

10

20

p 3.0

8,02-9,37 %

2.0

00

'

1.0
" 0

0 Fig.15

10

20

30

Variation of the damping factor of 4 mm diameter Degussit A1 2 0 3 specimens with the porosity and the grainsize

-89-

D)

-P 0 00

4co

0 -P

-P D)

0
4;

rtq
)

4-DO

rr

40

____*_

20

10

*.

Io
10

Pt (#,n)
i

Fig.17

Dependency of the fracture stress, Young's modulus

and the fracture strain from the Raubhtiefe Rt of 6 mm

diameter Degussit A12 03 specimes determined from a

bending test (grainsize 6 -91-

9.Ofuw.

porosity

5.37%)

40

30

20

;0

0
Yd Eq1

2,
__

)qO

LL
Fig.18

40

Dependency of the fracture stress, Young's modulus and the fracture strain from the Gldttungstiefe Rp of a 6 mm diameter Degussit A1 2 0 3 specimens determined from a bending test (grainsize porosity

S=

9.0A m

5.37%)

-92-

40

20

50

, 0 2

tCA(n CLA (sn)

E mined .fo 0
22

) edn

2t et(rainiz

G=

2,

'I

porosity

= 5.37%)

-93-

Fig.20

Measuring equipw9nt for the determination of the residual stresses on 6 mm diameter Wesgo Al 995 specimens

-94-

0,002

-o~oo0z..2..
-0,004

-OP06

1012 [ive

0,41

-0,2
-014
-0,6
__ _ _ _ _ _ __ _ _ _ _ _ __ _ _ _ _ _

]?ig.21

Strains

plotted

received from the grinding off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress ov-er the cross-section f determined from the above diagram (grainsize 6=
17.2,1 wv porosity

versus the remaining area

3.99%)

-95-

0,002

[M M-208LZ~a
10

20
> 00

"

30

-0,002
-0,004

-0,006

lG-fklkPm n'
0,,4
0,2 "" ' rr,,.,..7 M.

-0,4

-0,6

Fig.22

Strains 6 plotted versus the remaining area received from the grinding off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress over the cross-section f determined from the above diagram (grainsize 6 = 16.5 j4m , porosity p = 4.03%)

-96-

0100

10

20

30

2 G-[flmrr

SFig.23 ] t J,

6 p-lotted versus the remaining area Strains received from the grind-ing off process of a 6 mm diameter Wesgo Al 995 specimen and distribution of the residual stress over the cross-section f determined from the above diagram (grainsize 6

-97-

Fig.24

Test equipment for the determination of the influence of the loading speed on the fracture stress, Young's modulus and the strain at fracture

-98-

..

. . .. . . .

III

ATi

Fig.25,~~~~~~ eteupetfrtedtria ~ ~ ~ ~
inlec ofteladnpedoh

Bedn
fatr

iono

-99

iIIIEIIOE Em,--,,,M
-U,--..MM

Em.--...MMOMM

Fig.26a Lo

d-sri

dIga

.E.-.E. .E...-EEEM E.E-EEEE.M E.EEEEE. EEEE-,,E,

.Fiz26b Time-strain diagrami a h d

e t ta e ra ocllgapsM n eo b M a CoshM M teniot g a f macie sed tst Ph 0nminute fth wikelcroi Frctr la P= 192k, ratresran 3= .86o time t .34 e.pe nt

n T k ro

MMMMMM"i7;4-100-MM

I20

to
11

A2

T
1
_

__1

50

100

500

1000

Wk

.I

[ I

"__
-_

.2 _ _ _ _ _-__

_---_ _---

__ _ "

50100

oo

1000

5000

i!

Figo27

Dependency of the fracture stress, Young's modulus adthe strain at fracture from the strain rate of 6mm diameter Wesgo A1. 995 specimens determined from a bending test at room temperature

-101 -

1!0

11

12

13

14

15

16

h7

Fi&.28

Degussit A120 3 specimen for alternating bend fatigue root radius 1 mm) tests (residual diameter 5 mm, notch

InI
Fig.29 Schenck-Webi fatigue test machine and Tektronix cat'hode ray oscilloscope -102-

t,

1co

Ic I4r-44-D 4-i

.P

P,0 a) 0

0
p
Cd 4-1 (-4-rq

4-"

0
A

C)

W
zP

;Sl

(1

a)C

A m

0 4

C\J *- tr

'-i

L* l- 00

-103-

Fig.31

Clamping device for alternating bend fatigue tests

Fig.32

Vibration of a specimen mounted in a SchenckWebi fatigue test machine (strain amplitude 0.41%o, frequency 1495 cpm)

-104-

I!l

CCD4-1

'~

'~~~C

CD4)~
,

4-D

4- r- U 11 i
0 Pg-a) 10 p -H
0

4Z rO'-C

o0

-P

C)

V) 34C 0

-P C-i Q 0 I 4- F4H P-1q 0 C f4 P 0P CD El 0D-P cd a) (0 0I~ 4-)OFA4 ul 0 0 C

I~qk
0 0 CDQ o

-PI
DC

~-P

SC-P

-P CD5

V~L

IL
- 05P

Ii
0 0 Wr

P1
t

-PC 0 1 *
.,I

ro-P.
(L) N c

o0

Q I0 54

)'0 r4f rTis-

-P 02 -0 02 0 0 pF-I

N -iH

->

r,: -D 0 Pi (-P Cd -P 0 F-I 0 (1 -P F-I9 (q 0~- 0H r02+' c

CH 02 0 0) 0 0)20 F4I HF44 02 0 02E

0)0

/
I
IH

j..

*di

p> 9-Pf-m a) 02H

-P

:j 0

02 c

00020

(1)
p'Lc

pi -P

0r

04

P-106-0

Numbers of cycles to fracture

5,478,000 2,922,000 452,000 43,000 static failure

Fig.35

Fracture shapes of smooth Wesgo Al 995 specimens

broken under static and cyclic loads

Numbers of cycles to fracture 1,162,000

155,000 29,000
2,000 static failure

110

ll

112 113

114

15

16

17

118

119

Fig.36

Fracture shapes of notched Wesgo Al 995 specimens broken under static and cyclic loads

-107-

30

20
,..u9 24./. 10

p-1 5(.)

t~
0 10

(( m) (t

20

30

+~ 3

Ph) '33 'I :_ . 4. 6,51 (-/./ P - .18(%.)

p =9,2 (.()

p. 17,5 ('/.)

2 G 0 10 20 30
(em)

0,5

,-65,('/,) 7 0 4,3J/.) 6.51 M%) p.,241/.)

1!7,5(

G(.m)
01 0
Fig.37 10 20 30 Dependency of the fracture stress, Young's modulus and the fracture strain from the grainsize of 4 nn diameter Degussit A1 2 0 3 specimens determined from a bending test at room temperature

-108-

-'0

20-

_____ ____

____

---.

Gs6,60-~t)

10

_3

_r

8__

p()
01

L,

EB (1090

04

4
II0
CV)__
-3

0)-P (1) d A~U

4-

Go

CX

0)

ri)l
$-4 F1 40~ p-0)

0U 4-1rO 0

U), HCw

CH

0
(1

:s(
-

-110--

PLQ

0\
D Q Q(D4zE-

0)2 4 a)Qcc ra

(0 0 *H - P4

pN
-4

m dC
Ci) b.

4-J ~

m CH 0

4-)+

Cd0 0)

c)0
P0) 0) A 0

to~

/1
Iq

0
+1 0

X)

0b0

Ia

0)
P0

4z0)

*~f-~TT
U,
C~4

Ii,

*rI

4,

I
I

I I

02 4,2 0) 4, 0 02

0)
4,

0 cl-I
C

0)
p4

02

ci~
C) 02
p4

tJ)
U,

-113-

&

fiL.3

Tension test specimen of Alumina

.,.I
-114-

Fig.44

Complete test setup for a tension test on Alumina

-115-