rJ

m TR 64-i8o HARD COPY $ .

I MICROFICHE $ .6 7,

\vq• CORRELATION OF THE DSR WIT,{ THE STRENGTH OF GLASS

OF DIFFERENT COMPOSITIONS AND CONFIGURATIONS

c-,

TECHIm CA )CUMETARY REPORT NO. ML TDR 64-180

August 1964

Air Force Materials LaboratoryResearch Technology DivisionAir Force Systems Command

Wright-Patterson Air Force Base, Ohio

Project No. 7381, Task No. 738102

(Prepared under Contract No. AF 33(657)-11219 by the Pittsburgh Plate GlassCompany, Glass Research Center, Harmar Township, Pennsylvania;

R. W. Ansevin, author)

ARCHIVE COPY

NOTICES

When Government drawings, specifications, or other data are used forany purpose other than in connection with a definitely related Governmentprocurement operation, the United States Government thereby incurs noresponsibility nor any obligation whatsoever; and the fact that the Govern-ment may have formulated, furnished, or in any way supplied the said draw-ings, specifications, or other data, is not to be regarded by implication orotherwise as in any manner licensing the holder or any other person orcorporation, or conveying any rights or permission to manufacture, use, orsell any patented invention that may in any way be related thereto.

Qualified requesters may obtain copies of this report from the DefenseDocumentation Center (DDC), (formerly ASTIA), Cameron Station, Bldg. 5,5010 Duke Street, Alexandria, Virginia, 22314.

This report has been released to the Office of Technical Services, U.S.Department of Ccmmerce, Washington 25, D.C., for sale to the general public.

Copies of this report should not be returned to the Research and Tech-nology Division, Wright-Patterson Air Force Base, Ohio, unless return isrequired by security considerations, contractual obligations, or notice ona specific document.

/

P I TT S B U R aR P . AT R r. T, A q C OM PA NMY

GLASS DIVISION RESEARCH LABORATORIES

CORRELATION OF THE DSR WITH 1HE STRIEqGTH OF GLASS

OF DIFFERENT COMPOSITIONS AND CONFIGURATIONS

FINAL TECHNICAL REPORT

GOVERNMENT CONTRACT NO. AF 33(657)-11219

PROJECT NO. 7381, TASK NO. 738102

Prepared By Approved By P. W. £4 4 •R. W. Ansevin P. W. French, HeadProject Leader Optics Department

Date August 7, 1964 Approved ByB. Y * nnison, ManagerResearch Planning and

Control Division

FOREWORD

This report was prepared by the Pittsburgh Plate Glass Cmpany, GlassResearch Center, under USAF Contract No. AF 33(657)-11219. The contract wasinitiated under Project No. 7381, Task No. 738102. The work was administratedunder the direction of the Air Force Materials Laboratory, Research TechnologyDivision, Mr. E. W. McKelvey, Project Officer.

This report covers work conducted I July 1963 to 30 June i #64.

The author would like to acknowledge the guidance and technical supportprovided by both Mr. P. W. French and Mr. L. E. Orr. The study of limitationswas performed by Mr. R. J. Obenreder. Credit goes to Dr. F. H. Gillery forthe method used in the estimation of membrane stresses. Mr. A. N. George andMr. J. Shelestak are to be thanked for their assistance in the preparationand testing of the many samples used in this work.

ABSTRACT

The Differential Surface Refractameter (DSR) is an instruaent recently

developed at the Glass Research Division of the Pittsburgh Plate Glass

Ccmpany to detect and e~timate the stress in the surface layers of glass by

nondestructive means. This work shows a correlation betveen the instrument

readings and experimental break strength data for glass of different ccnposi-

tions and configurations. In addition, the utility and some of the limitations

of the instrmuent are discussed.

This report has been reviewed and is approved.

W. P. CONRARY, ChiefMaterials Engineering BranchMaterials Applications DMvisionAF Materials Laboratory

iii

TA•TLE OF CONTE1•TS

'D A rP-V

INTRODUCTION -

CORRELATION REQUIR5ENTS ---------------------------------------------- 2

PROCEDURE ------------------------------------------------------------ 3

Adaptation of the DSR to Other Glass Types ------------------------ 3

Calibration ------------------------------------------------------ 3

Sample Preparation -------------------------------------------- 4

Surface Compression Via DSR Readings ------------------------------ 5

Center Tension Measurements -------------------------------------- 5

Destructive Strength Tests --------------------------------------- 5

DESTRUCTIVE STRENGTH TEST RESULTS AND CORRELATION ---------------------- 6

Results ---------------------------------------------------------- 6

Discussion of Results -------------------------------------------- 6

CONCLUSIONS ---------------------------------- 6STUDY OF LIMITATIONS -------------------------------------------------- 9

Curvature -------------------------------------------------------- 9

Other Limitations ----------------------------------------------- 10

Waviness --------------------------------------------------- 10

Edge Proximity --------------------------------------------- 10

Surface Finish --------------------------------------------- 10

REFERENCES ---------------------------------- ii

APPENDIX I DETAILED PROCEDURE -------------------------------------- 12

CALIBRNTION --------------------------------------------------------- 12

Calibration Methods ---------------------------------------------- 12

Type I MIL-G-25667 ------------------------------------------ 12

Chemcor ---------------------------------------------------- 12

iv

PAGE

Alimino Si l icate 14

Herculite ii -14

Index Measurements --------------------------------------------

DESTRUCTIVE STRENGTH TESTS -------------------------------------------- 19

Testing Methods ------------------------------------------------- 19

Testing Techniques ---------------------------------------------- 20

Calculation of Bending Stresses ---------------------------------- 21

Rectangular Specimens (Beam Bending) ------------------------ 21

Square Specimens (Concentric Rings of 2:1 Ratio) ------------- 21

Corrections for Fracture Origins Outside Maximum Stress Area .....- 22

Corrections for Membrane Stress in Concentric Ring Testing -------- 22

APPENDIX II INDIVIDUAL TESTING RESULTS ----------------------------- 26

V

ILLUSTRATIONS

FIGURE PAGE

1 Representative Calibration Results for Type I MIL-G-25667Glass (Polished Plate) ---------------------------------------- 13

2 Representative Calibration Results for Chemcor Glass ----------- 15

3 Representative Calibration Results for Alumino SilicateGlass -------------------------------------------------------- 16

4 Representative Calibration Results for Herculite II Glass ------ 17

5 Corrections for Fracture Origins Outside Maximum StressArea for 4 and 12 In. Span Beam Loading Method ----------------- 23

6 Corrections for Fracture Origins Outside Maximum StressArea for Concentric Ring Loading Method ----------------------- 23

7 Graphical Estimation of Correction Factor for MembraneStresses for Typical Full Tempered Herculite II Sample --------- 25

vi

TABLES

TABLE PAGE

1 Glass Types Investigated --------------------------------------- 2

2 Summarized Mean Calibration Results ----------------------------- 4

3 Birefringence Constants ---------------------------------------- 5

4 Summarized Destructive Strength Test Results and Correlation --- 8

5 Curvature Limitations ------------------------------------------ 9

6 Calibration, Surface Index and Stress Optical ConstantResults ------------------------------------------------------- 18

7 Individual Testing Results, Full Temper ChemcorSample Size 4 x 4 x 1/10 In., Testing Method l1l/2 and 3 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 15000 Psi/Min. ------------------------------------ 27

8 Individual Testing Results, Annealed Herculite IISample Size 4 x 4 x 1/10 In., Testing Method l1l/2 and 3 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 15000 Psi/Min. ------------------------------------ 28

9 Individual Testing Results, Full Temper Herculite IISample Size 4 x 4 x 1/10 In., Testing Method l-l/2 and 3 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 15000 Psi/Min. - -- ------ ------------- 30

10 Individual Testing Results, Annealed Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 7000 Psi/Min. ------------------------------------- 31

11 Individual Testing Results, 1/2 Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 7000 Psi/Min . ------------------------------------- 32

12 Individual Testing Results, 3/4 Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 7000 Psi/Min. ------------------------------------- 33

13 Individual Testing Results, Full Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and C In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 7000 Psi/Min. ------------------------------------- 34

vii

TABLE PAGE

E 14 Individual Testing Results, Annealed Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/2 In.. Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min . ------------------------------------- 35

15 Individual Testing Results, 1/2 Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min ------------------------------------- 36

16 Individual Testing Results, 3/14 Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min . ------------------------------------- 37

17 individual Testing Results, Full Temper Alumino SilicateSample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 38

18 Individual Testing Results, Annealed Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 3/A6 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min. ------------------------------------- 39

19 Individual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 3/16 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min. ------------------------------------- 40

20 Individual Testing Results, 3/4 Temper Type I MIL-G-_2>667Sample Size 6-1/8 x 6-1/8 x 3/16 in., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min. ------------------------------------- 41

21 Individual Testing Results, Full Temper Type i MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 3/16 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 5000 Psi/Min. ------------------------------------- 42

22 Individual Testing Results, Annealed Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 6000 Psi/Min ------------------------------------- 43

23 Individual Testing Results, 1/2 Temper Type i MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/4 in., Testing Method 3 and 6 In.Diameter Concentric Rirgs, Testing Machine - Instron,Loading Rate 6000 Psi/Min. ------------------------------------- 4

viii

TABLE PAGE

24 Individiual Testing Results, n/) Temper Type i MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/4 in., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 6000 Psi/Min. ------------------------------------- 45

25 Individual Testing Results, Full Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 6000 Psi/Min. ------------------------------------- 46

26 Individual Testing Results, Annealed Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 403c, Psi/Min . ------------------------------------- 47

27 1'.dividual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Instron,Loading Rate 4000 Psi/Mi. ------------------------------------- 48

28 Individual Testing Results, 3/4 Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 49

29 Individual Testing Results, Full Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 1/2 In., Testing Method 3 and 6 in.Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 50

30 Individual Testing Results, Annealed Type I V,[L-G-25667Sample Size 6-1/8 x 6-1/8 x 3/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 PsMin. ------------------------------------- 51

31 Individual Testing Results, Full Temper Type I MIL-G-25667Sample Size 6-1/8 x 6-1/8 x 3/4 In., Testing Method 3 and 6 In.Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/in ------------------------------------- 52

32 individual Testing Results, Aimefaled Type I MIL-G-25667Sample Size 12-l/h x 12-!/4 x 1/2 In., Testing Method 6 and12 In. Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 53

33 Individual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 12-1/4 x 12-1/4 x 1/2 In., Testing Method 6 and12 In. Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 54

ix

TABLE PAGE

-4 dTesting~ Results, 3/'4 Temper Type I MiL-G-25667Ssmple Size 12-1/4 X 12-1/14 x I/f r .! n., Testing Method 6 and12 In. Diameter Concent.i. Rir~ , Testir. Machine - Baldwin,

Loading Rate 5000 Psi/Min. 5535 IfTld.vi•da. Testing Results, Full Temper Type I MIL-G-25667

Sample Size 12-1_/4 x i2-1/4- x 1./2 in., Testing Method 6 and12 in. Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min. ------------------------------------- 56

36 Individual Testing Result-s, Annealed Type I MIL-G-25667Sample Size 12_1/4 x 121./4 x 3/4 in.,, Testing Method 6 and12 In. Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Min - ------------------------------------- 57

37 Individual Testing Results, Fu'Lll Temper Type I I,-_-G-.G25667Sample Size l2-1/4 x 12-1/4 x 3/4 In., Testing Method 6 and12 In. Diameter Concentric Rings, Testing Machine - Baldwin,Loading Rate 5000 Psi/Mn. -------------------------------------- 58

38 Individual Testing Results, Annealed Type I MIL-G-25667Sample Size 3 x 13 x 3/16 In., Testing Method - Beam Loading -

4sand 12 In. Span, Tes-ing Machine - Instron,Loading Rate 4000 Psi/Mn. -------------------------------------- 59

39 individual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 3 x 13 x 3./16 In., Testing Method - Beam Loading -

4 and 12 In. Span. Testing Machine - Instron,Loading Rate 4000 Psi/Min. -0

40 Individual Testing Results, 3/4 Temper Type I MIL-G-25667Sample Size 3 x 13 x 3/16 in., Testing Method - Beam Loading -

4 and 12 In. Span, Testing Machine - Instron,Loading Rate 4000 Psi/Min . ------------------------------------- 61

41 Individual Testing Results, Full Temper Type I Mi-IL-G-25667Sample Size 3 x 13 x 3/16 In., Testing Method - Beam Loading -

4 and 12 In. Span, Testing Machine - Instron,Loading Rate 4000 Psi/Min------------------------------------ 62

42 Individual Testing Results, Annealed Type I IL,•.G2•,.7Sample Size 3 x 13 x 1/i In., Testing Method - Bewa Loadir -

4 and 12 In. Span, Testing Machine - instron,Loading Rate 4000 Psi/Min .6

43 Individual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 3 x 13 x 1/4 In., Testing Method - Beam Loading -

4 and 12 In. Span, Testing Machine - Instron,Loading Rate 8000 Psi/Min . ------------------------------------- 64

x

.4 Individual TevUirl ResuI~e. q/4 TScnner 'P'n i ,Sample Size 3 x 13 x 1/4 inA, Testing Method - Beam Loading -

4 and 12 In. Span, Tezting Machine - In, ron,Loading Rate &0N P iMin- 65

45 Individual Testing Results. F-ull T1mper T.-•e i MIL-G-25667Sanmple Size 3 x 13 x 1/4 In., Testing Method - Beam Loading -4 anld 12 In. Spam, Testing Machine - instron,Loading Rate 8000 Psi/M--in -......... 66

46 individual Testing Results, nnealed Type I MIL-G-25667

Sample Size 3 x 13 x 1/2 in.. Testing Method - Beam Loading -

4 and 12 In. Span, Testing Mai'achine - instron,Loading Rate 5000 Psi/M.in ------------------------------------- 67

47 Individual Testing Results, 1/2 Temper Type I MIL-G-25667Sample Size 3 x 13 x 1/2 in., Testing Method - Beam Loading -4 and 12 In. Span, Testing Machine - Instron,Loading Rate 5000 Psi/Min. ---------------------- 68

h8 Individual Teeting Results, 3/4 Temper Type I MIL-G-25667Sample Size 3 x 13 x 1/2 In., Testing Method - Beam Loading -

4 and 1.2 in. Span, Testing Nachine - Instron,Loading Rate 5000 PsiMin. ------------------------------------- 69

49 indivridual Testing Results, Full Temper Type I MIL-G-25667Sample Size 3 x 13 x 1/2 In., Testing Method - Beam Loading -

4 and 12 in. Span, Testing Machine - Instron,Loading Rate 5000 Psi/Mn. ------------------------------------- 70

50 Individual Testing Results, Annealed Type I MIL-G-25667Sample Size 3 x 13 x 3/4 In., Testing Method - Beam Loading -

,and 12 In. Span, Testing Machine - Instron,Loading Rate 5000 Psi/Min - --------------------- 71

5i Individual Testing Results, Full Temper Type I MIL-G-25667Sample Size 3 x 13 x 3/4 In., Testing Method - Beam Loading -

4 and 12 In. Span, Testing Machine - Instron,Loading Rate 8000 Psi/Min. ------------------------------------- 72

xi

INTRODUCT ON

Numerous attempts have been made in the past to determine the strengthof tempered glass by nondestructive examination of the finished product.The results of these attempts, however, were not accurate enough to be ofpractical value and, in general, the strength of glass had to be found bythe destructive testing of a large number of samples. With the developmentof the Differential Surface Refractometer, we believe the nondestructiveevaluation of the strength of tempered glass is now possible.

The br~akage of glass takes place under tension, and in general, thedegree of tension required to cause failure is controlled by the presence ofsurface flaws. The surface condition has a considerable influence on thestrength Of glass; the stability of glass is dependent upon- the- magnitudeand distribution of these surface flaws so that under ordinary conditions, thestrength of a piece of glass is determined by the strength of its surface.The dependence of strength on the distribution of flaws explains the largevariations obtained in -any determination of the modulus of rupture. Due tothis variation, the strength of glass is a property which has a meaning onlyin a -statistical sense. Hence, in this sense, the strength of annealed glassof a given production process can be expressed in terms of a critical orbasic tensile stress. The tempering of glass causes the surface to beplaced in a state of ccmpression- which has to be overcome before the glasssurface can be subjected to a tensile stress of sufficient magnitude tocause failure. As -a result, the strength of glass is increased by thepresence of compressive stresses on the surface. This strength or modulusof rupture of tempered glass may be assessed as the -sum of the basic strength,i.e., -the -stress requiredc to cause failure in the absence of compressiveforces on the surface, plus the compressive stress induced by the temperingprocess. Therefore, with a knowledge of the -basic strength, the ability tomeasure the degree of surface compression present on glass by nondestructivemeans would provide a method of determining glass strength.

Among the optical methods investigated over the past decade for thenondestructive evaluation of the strength of tempered glass, surface refrac-tometry was found to be particularly well suited for the quantitative measure-ment of the surface compression of tempered glass. The Differential SurfaceRefractometer (RSR), an instrument based on the principles of surfacerefractometry(l , was recently developed at the Pittsburgh Plate GlassCmapany, Glass Research Center, expressly for this purpose. After suitablecalibration, the DSR is capable of measuring the degree of surface compressionpresent on tempered glass directly in pounds per square inch.

The primary intent of this report is to show a correlation between theDSR and the strength of glass of various compositions and configurationsthrough the use of experimental break strength data obtained in the laboratory.Specifically, the following table lists the types and thicknesses of glassthat were investigated:

Manuscript released by the author 5 June 19b4 for publication as an RTD"* Technical Documentary Report.

1

TABLE 1GLASS TYPES INVESTIGATED

Thickness Degree of Temperin Inches None (Annealed) 1/2 3/4 Full

1/1O to 3/16 E -. D and E3/16 A A A A1/4 A and B A and B A and B A and B1/2 A and B A and B A and B A and B3/4 A A

Legend: Type of Glass

A - Type I MIL-G-25667, polished plate glass obtained frcm PittsburghPlate Glass Company.

B - Altmino Silicate, high strain point glass No. 6695 obtained fromPittsburgh Plate Glass Company.

D - Chemcor, code 0311 glass obtained from Corning Glass Company.

E - Herculite II, No. 7265 glass obtained from Pittsburgh PlateGlass Company.

In addition, the utility and some of the limitations of the DSR weredetermined.

CORRELATION REQUIREMENTS

In order to correlate the readings of the DSR with the strength oftempered glass it is necessary to determine the following information foreach glass type listed in Table 1:

(1) The DSR calibration value for the glass.

(2) The basic strength of the glass, i.e., the strength of theglass in the annealed state.

(3) The degree of surface compression present on the temperedglass by means of the DSR readings.

(4) The breaking strength, or modulus of rupture, of the temperedglass.

From this data we would want to show the following equation to be validwithin the limits of experimental accuracy:

{ sic tR heading of Modulus ofStrengthl + Surface Compressionj ' f Rupture I

2

Of course, the three components of this equation are to be interpretedstatistically and each component would represent the mean result obtainedin the examination of a number of samples.

PROCEDURE

Adaptation of the DSR to Other Glass Types

Since the DSR was originally designed to operate specifically on plateand sheet glass of a given index range, some slight changes were required inthe basic design of the instrument to extend its use to the other glass typesinvestigated in this work. This was necessitated by the different indicesand/or the magnitudes of stress which are characteristic of the other glasstypes. As a result, three different instruments were used in this study.The basic design, the DSR, was used on Type I MIL-G-25667 glass (polishedplate); the Modified Differential Surface Refractometer-I, or MDSR-I, wasused for the measurement of surface compression on both Chemcor and HerculiteII; and, the MDSR-II was used for the examination of the Alumino Silicateglass.

Calibration

The DSR, suitably modified for each glass type when required, wascalibrated by subjecting fully tempered samples to mechanical beam loading.Loading was applied in suitable inc:rements; at each load increment, acorresponding reading of the DSR instrument was made. The DSR reading wascompared with the induced stress calculated from beam bending theory. Fromtlis data, the surface compression of the sample at zero load was estimatedusing the method of least squares. Dividing the value obtained for thesurface compression at zero load by the DSR reading for the same conditiongives the instrument's calibration value, K, in psi per micrometer eyepieceunit, for each sample.

In addition, the stress-optic constant for each glass type was determinedby measuring the difference between the surface refractive indices, n,, andn1, for light polarized parallel and perpendicular to the surface of thestressed sample, respectively. The stress-optic constant C, in psi per unitbirefringence, is given by

C =0 /n,, - n

where a is the surface laminar stress.

The difference n,, - n. used to calculate the stress-optic constant ismeasu•ed directly with the DSR. This direct measurement of n, - n. is moreaccurate, to at leastan order of magnitude, than the measurement of n, and

n, separately and then forming the difference. The separate surface indiceswere also measured, however, to give an indication of their relative values.These results are summarized below in Table 2.

3

Vi

TABLE 2SUMARIZED MEAN CALIBRATION RESULTS

[Calibration Stress-OpticSurface Value, K. Constant**., SurfaceStress psi/Eyepiece Standard C, psi/Unit Indices

Glass Type Instrument Unit Deviation Birefringence n_, 2&

Type I DSR 60.18- o.41 5.4 x 107 1,5246 1.5250MIL-G-25667

Chemcor MDSR-I 173.4 * 13. 4.4 x 107 1.5121 1.5130Alumino MDSR-II 167.2 * 2.5 .8xO 1.5576 1.5580Silicate

Herculite II MDSR.I 189.3 * 11. 5.0 x 107 1.5142 1.5153

*The relative differences in the magnitudes of the calibration values listed

here are, to a great extent, accounted for by a difference of telescope powerbetween the DSR and its modified forms.

**The stress-optic constants listed here represent values obtained by examiningthe surface layers of each type glass. These values may be different frcmthose obtained by examining the bulk material. This would be especially truefor Chemcor and Herculite II.

The methods used in determining the surface refractive indices and theirdifferences, and the details of the calibration procedure are contained inAppendix I. Detailed results are listed in Table 6 of Appendix I.

Sample Preparation

All specimens, with the exception of Chemcor, were made from regularproduction glass fram the Pittsburgh Plate Glass Ccmpany. The cutting washandled by normal plant methods with an added stipulation that all cutt~ingbe made fran the same surface. These samples were tempered at the GlassResearch Center in accordance with Table 1. No edge work was performed onthese samples. The Chemcor samples were obtained fully tempered from theCorning Glass Company. These samples were received with the edges groundto a rounded contour.

In general, 25 specimens were prepared for each category listed inTable 1. Approximately twice this amount, however, were prepared for thetesting of the Chemcor and Herculite II glass types.

The glass surfaces vere carefully protected against mechanical damageduring preparation of the specimens and during measurements. Paper wasplaced between each sample and its neighbor in order to avoid surface defectsfrcm glass to glass contact. The thickness was measured at the cornerswithout touching the central portion of the specimen. Width measurements,when required, were taken with equal care.

4

Surface Compression Via DSR Readings

DSR readings were made in the center of each tempered sample on thesurface with the cutter marks. The surface on which the DSR readings weremade was the side that would be under compression during the destructivetests. This was done in order to eliminate the possibility of affecting theresults of the destructive testing by contacting, and possibly damaging,the side to be placed under tension. It was assumed that the surface stressin the central portion of each sample would be essentially isotropic andequal from side to side. No less than four, and in general, five readingswere made on each sample and the average value was used, by means of thecalibration value, to determine the surface compression in psi.

Center Tension Measurements

Although not required in this investigation, center tension measurementswere made to provide additional information for those accustomed to gagingthe degree of temper by this means. -Measurements were made- using-a quartzwedge graduated to read relative retardation in-millimicrons. For thestrips, 3 x 13 in-., the measurements were made across the width in the centerof each sample. The square samples were measured across diagonally opposite*corners, along a 2 in. optical path for the 6-1/8 -in. squares, and alonga 4 in. optical -path for the 12-i/4 in. square samples. The 4 in. squares ofChemcor and Herculite II were measured through the whole width of the samples.S This data was converted to psi by using the following conversion factorsnrovided by the manufacturers.

TABLE 3BIREFRINGECE CONSTANTS

Manufacturer and Birefringence ConstantGlass Type Code Number psi/mmu/in.

Type I MIL-G-25667 PPG Polished Plate 2.13Chemcor Corning 0311 2.28Alumino Silicate PPG 6695 2.17Herculite II PPG 7265 2.26

Using this information of the center tension and the values obtainedfor the surface compression via the DSR, the ratio of the surface compressionto the center tension was calculated for each type glass in their variousstates of temper. These results are summarized in Table 4. More detailedinformation is contained in Appendix II, Individual Testing Results.

Destructive Strength Tests

The strength of the annealed and tempered specimens was measured byeither a modified beam loading method or a concentric ring method. Both ofthese testing methods have been in use at the Glass Research Center for thepast 15 to 20 years. The beam loading method involves loading to breakage,

5

rectangular specimens simply supported near each end and loaded at twopoints symmetrically placed about the center. In the concentric ringmethod, a square plate is loaded by a circular member which moves coaxiallyto a support ring. The details of these methods are contained in Appendix I.

DESTRUCTIVE STRENGTH TEST RESULTS AND CORRELATION

Results

The results of the destructive strength tests and the results of themeasurements of the surface compression. and center tension are summarized inTable 4. The results listed in this table indicate that the principle ofsurface refractcmetry, as used in the DSR, the MDSR-I, and the MDSR-II, tomeasure surface compression of tempered glass is capable of predicting thebreaking strength to within +10% for seventy per cent of the samples testedand to within +15% for the entire testing program. The average of thepercentage deviation from breaking stress is -2.6% with the sign of thedeviation taken into account. The mean of the absolute values of the percent deviation, however, is 8.1%.

Tables of individual testing results are contained in Appendix II.

Discussion of Results

The agreement of the equation

f~Basic +l DSR Reading of. _ Modulus ofjtrengthJ {Surface Ccmpressionj Rupture J

to within 10% certainly indicates the validity of surface refractcmetry, asexemplified in the stress instruments used in this study, as a means ofassessing the strength of tempered glass. With instrumentation of thistype, it is now possible to specify and control the strength of temperedglass of a given production process without the need of extensive destructivetesting. That is, once a DSR instrument has been cailbrated for a particularglass and the basic strength of this glass determined, this report shows thatthe strength of the tempered product can be determined nondestructively withstatistical certainty.

CONCLUSIONS

The Differential Surface Refractcmeter, an instrument based on the prin-ciples of surface refractcmetry, and modifications of this instrument asrequired to extend its usefulness to other glass types, is capable of deter-mining the magnitude of the surface compression, in psi, present on temperedglass by nondestructive means. Since the strength of glass is dependent uponthe strength of its surface, this ability to measure the surface compressionof tempered glass enables one to assess the strength of this glass withoutthe need for destructive tests.

6

This report has demonstrated the correlation of the DSR readings -with the

strength of tempered glass of four dilierent c .posi..iom n. ver.ety....

figurations. Specifically, the glass types and the stress instruments which

were correlated are:

Type I MIL-G-25667 glass, PPG Polished Plate, and the DSR;

Chemcor Glass, Corning Glass Company Code 0311, and the MDSR-I;

I{erculite II Glass, PPG No. (265, and the MDSR-I;

Alumino Silicate, PPG No. 6695, and the MDSR-II.

7.

T -

SU)OMARIZED DESTRUCTIVE STRENGTH TEST REST

FRAiES Number Numberof of Mean

kEdge Testing CenterBreaks insl Tension

Cheatcor 1 0 1/10 x + x 4 CR 1-1/2 x 3 6530Herculite II, Annealed 0 1/10 x 4 x 4 CR 1-1/2 x 3 30Herculite Il 51 0 1/10 x 4 x 4 CR 1-1/2 x 3 6950Polished Plate, Annealed 2j 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6Polished Kate, Annealed 2 0 1b x 6-1/8 x 6-1/8 CR 3 x 6Polished Plate, Annealed 29 16 1/2 x 6-1/8 x 6-1/8 CR 3 x 6Polished Plate, Annealed - - Same data as above but neglecting edge breaksPolished Plate, Annealed 22 13 3/A x 6-1/8 x 6-1/8 CR 3 x 6Polished Plate, Annealed - - Same data as above but neglecting edge breaksPolished Plate, Annealed 29 0 1/2 x 12-lb x 12-1/ CR 6 x 12 235Polished Plate, Annealed 31 4 3/4 x 12-1l x 12-1/ CR 6 x 12 311Polished Plate, 1/2 Temper 28 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6Polished Plate, 1/2 Temper 25 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 1060Polished Plate, 1/2 Temper 25 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 4310Polishe Plate, 1/2 Temper 26 0 1/2 x 12-1l x 12-1/ CR 6 x 12 U80Polished Plate, 3/4 Temper 25 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6Polished Plate, 3A Temper 26 0 lb x 6-1/8 x 6-1/8 CR 3 x 6 6430Polished Plate, 3A/ Temper 27 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 5090Polished Plate, 3/4 Temper 27 0 1/2 x 1-1/2 x b 2-1/ CR 6 x 12 5J20Polished Plate, l Temper 27 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6 6170Polished Plate, Full Temper 25 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 7160Polished Plate, Full Temper 26 3 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 7850Polished Plate, Full Temper - - Sam. data as above but neglecting edge breaksPolished Plate, Full Temper 25 7 3/4 x 6-1/8 x 6-1/8 CR 3 x 6 8190Polished Plate, Full Tonper - - Same data as above but neglecting edge breaksPolished Plate, Full Temper 25 0 1/2 x 12-1/4 x 12-1/4 CR 6 x 12 48oPolished Plate, Full Temper 26 0 3/4 x 12-1/4 x 12-1/4 CR 6 x 12 ROOAluslAo Silicate, Annealed 25 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 225Alilaino Silicate, Annealed 22 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 343Alimlno Silicate, 1/2 Temper 22 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 397oAlmino 31licate, 1/2 Temper 21 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 3480Alvnino Silicate, 3A Temper 20 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 5190Altmino Silicate, n/ Temper 21 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 5540AlumLno Silicate, Full Temper 21 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 7450Albnio Silicate Full Temper 21 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 7 00Polished Plate, Annealed 31 6 3/16 x 3 x 13 Beam 4 x 12 147Polished Plate, Annealed 28 13 1/4 x 3 x 13 Beam 4 x 12 11+2Polished Plate, Annealed 28 11 1/2 x 3 x 13 Beam 4 x 12 256Polished Plate, Annealed 2Z 25 3/4 x 3 x 13 Beam 4 x 12 171Polished Plate, 1/2 Temper 2 12 3/16 x 3 x 13 Beam 4 x 12 4030Polished Plate, 1/2 Temper 25 21 1/4 x 3 x 13 Beam 4 x 12 4030Polished Plate, 1/2 Temper 25 16 1/2 x 3 x 13 Beam 4 x 12 4360Polished Plate, 3/4 Temper 25 11 3/16 x 3 x 13 Beam 4 x 12 5320Polished Plate, 3/ Temper 25 24 1/4 x 3 x 13 Beam 4 x 12 5230Polised Plate, 3/4 Temper 27 17 1/2 x 3 x 13 Beam 4 x 12 5290Polished Plate, Pull Temper 23 20 3/16 x 3 x 13 Beam 4 x 12 7700Polished Plate, Full Temper 25 23 1/41 3 x 13 Beam 4 x 12 7370Polished Plate, Full Temper 25 13 1/2 x 3 x 13 Beam 4 x 12 86 0Polished Plate, Full Temper 26 15 3/4 x 3 x 13 Beam 4 x 12 9640

Legends CR - Concentric Ring Method, numbers folloving CR indicate the diameter of the rings In inct

Beam - Beam Loading Method, numbers following Beam indicate stress span and load span in inc

8

TABLE 4,RUCTIVE STRENGTH TEST RESULTS AND CORRELATION

MeanSurface

Mean Mean on Mean Basic StrengthTesting Center Surface Center Breaking Standard Plus SurfaceMethod Tension Comgreinn Ratio Stress Deviation Ranze :omuression

CR 1-1/2 x 3 6530 4,5770 7.07 51180 7230 28200CR 1-1/2 x 3 30 11330 2780 11300CR 1-1/2 x 3 6950 56970 8.21 67010 65+0 40000 71300CR 3 x 6 10020 2310 9600CR 3 x 6 9270 2080 8900CR 3 x 6 7930 3090 9800

ieglecting edge breaks 10760 2430 9000CR 3 x 6 7410 4020 i•400ieglecting edge breaks 9580 4150 1 700CR 6 x 12 235 10740 1200 4900CR 6 x 12 311 10570 2890 10200CR 3 x 6 8680 17980 3130 13000 18700CR 3 x 6 4060 9800 2.41 21980 2560 9800 19070CR 3 x 6 4310 9240 2.14 23420 4.310 16900 20000CR 6 x 12 4480 9920 2.22 23510 2150 7900 20660CR 3 x 6 11160 22040 2620 11400 21180CR 3 x 6 6430 15900 2.47 29360 310 11+100 25170CR 3 x 6 5090 12540 2.47 27090 3380 14600 23300CR6 x 12 5320 12910 2.42 27460 2880 11500 23650CR 3 x 6 6870 15270 2.22 26370 3280 13000 25290CR 3 x 6 7160 17300 2.42 30780 2480 10100 26570CR 3 x 6 7850 18980 2.32 29920 271+00 26910ieglecting edge breaks 310o 4544 151+00 2971+0

CR 3 x 6 8190 23770 2.90 32060 11190 3R800 31180ieglecting edge breaks 36930 7640 28400 33350CR 6 x 12 7410 18810 2 32610 3350 14600 29550CR 6 x 12 8100 23370 2:. 35990 0 12000 33940CR 3 x 6 225 10920 180 7200CR 3x6 343 10970 1280 o000CR 3 x 6 3970 10340 2.61 19350 1830 8500 21260CR 3 x 6 3480 10920 3.14 20790 2180 7300 21890CR 3 x 6 5190 13410 2.58 24340 1780 7 00 24330CR 3 x 6 5540 17500 3.15 25690 3500 12000 2870CR 3 x 6 7750 21150 2.73 30630 2220 10700 32070CR 3 x 6 7400 23080 3.12 30250 31+10 14200 31+050Beam 4 x 12 147 10940 2610 10400Beam 4 x 12 142 9050 1920 7200Beam 4 x 12 256 121+10 2260 11000Beam 4 x 12 171 6670 3410 161.00Beam 4 x 12 4030 9370 2.31 21570 2080 7800 20310Beam 4 x 12 40O30 9380- 2.33 20250 2780 10500 18430Beam 4 x 12 4360 10570 2. 2 21470 2510 1040 22980Beam 4 x 12 5320 11700 2.19 23750 2750 11100 22640Beam 4 x 12 5230 11860 2.27 20360 2630 10700 20910Beam 4 x 12 5290 13080 2.47 24590 3230 15900 25290Beam 4 x 12 7700 17160 2.22 27040 2190 9700 28100Beam 4 x 12 7350 17210 2 .2 23700 2800 9700 26260Beam 4 x 12 86 0 21430 2:48 30860 3240 13100 33840Beam 4 x 12 9610 26830 2.78 36500 3690 16300 33500

.ameter of the rings in inches.

:s span and load span in inches.

8

TABLE 4SUO4ARIZED DESTRUCTIVE STRENGTH TEST RESULTS AND CORRELAI

Number Number Suof of Mean Mean

" Edge Testing Center Surface CeiTeste Brks Sample Mt Tensig on2 in L

0 1/10 x 4 x 4 CR 1-1/2 x 3 6530 1+5770 7.inealed 0 1/10 x 1 x 4 CR 1-1/2 x 3 30

51 0 1/10 x 4 x 4 CR 1-1/2 x 3 6950 56970 8.Annealed 27 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6Annealed 2i 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6Annealed 29 16 1/2 x 6-1/8 x 6-1/8 CR 3 x 6Annealed - - Same data as above but neglecting edge breaksAnnealed 22 13 3/4 x 6-1/8 x 6-1/8 CR 3 x 6Annealed - - Same data as above but neglecting edge breaks%nnealed 29 0 1/2 x 12-1/. x 12-1/4 CR 6 x 12 235Annealed 31 1+ 3/4 x 12-1/4 x 12-1/4 CR 6 x 12 3111/2 Temper 28 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6 8680L/2 Temper 25 0 1/4 x 6-1/8 x 6-1/8 CR 3 x 6 1+060 9800 2.1/2 Temper 2 0 1/2 x 6-1/8 x 6-1/8 CR x6 4310 921+0 2.

-. 1/2 Temper 2 0 1/2 x 12-1/4 x 12-1/4 CRA x12 4480 9920 2.3/4 Temper 2 0 3/16 x 6-1/8 x 6-1/8 CR 3 x 6 111603/ Temper2 0 I/4 x 6-1/8 x 6-1/8 CR 3 x 6 6430 15900 2.3/ Temper 27 0 1/2 x6-1/8 x 6-1/8 CR3 x6 5090 12540 2.3/4 Temper 27 0 1/2 x 12-1/ x 12-1/A CR 6 x 12 5J20 12910 2.;'ull Temper 27 0 3/16 x 6-1/8-x 6-1/8 CA 3 x 6 6570 15270 2..ull Temper 2'0 1/4 x 6-1/8 x 6-1/8 CR 3 :6 7160 17300 2.?ull Temper 26 3 1/2 x 6-1/8 x 6-1/8 CR3 :6 7850 18980 2lull Temper - Same data as above but negleoting edge breaksPull Tosper 25 7 3/4 x 6-1/8 x 6-1/8 C1 3 x 6 8190 23770 2.ull Temper - - Same data as above but neglecting edge breaks"ull Temper 2 0 1/2 x 12-1/ 4 x 12-1/ 4 CR 6 x 12 440 18810 2.'till Temper 2 0 3/4 x 12-1/4 x 12-1/4 CR 6 x 12 :100 23370 2.

Annealed 25 0 !A x 6-1/8 x6-1/13 CR 3 x6 225Annealed 22 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 33,12 Temper 22 0 1/4 x 6-1/8 x 6-1/8 CA 3 x 6 3970 10340 2.1/2 Temper 21 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 3490 109203/4 Temper 20 0 1/4 2 6-1/8 x 6-1/8 CR 3 x 6 5190 13410Y4 Temper 21 0 1/2 x 6-1/8 x 6-1/8 CR 3 x 6 5540 17500 3.Ful• Temper 21 0 1/A 6-1/8 x 6-1/8 CR 3 x 6 7750 2150 2.Full Temper 21 0 1/2 x 6-1/8 : 6-1/8 CR 3 x 6 7400 2300 3.

inn•eled 31 6 3/16 : 3 : 13 Bea 4 • 12 11+7knneeled 28 13 1/4 x 3 x 13 Beam 4 x 12 11+2annealed 28 11 1/2 x 3 x 13 Beam 4 • 12 256nnealed 2 25 34 3 13 Beam 1+ 2 171/2 Temper 25 12 3/16 : 3 x 13 Beam 4 x 12 4030 9370/2 Temper 25 21 1/4 • 3 • 13 Boaz 4+ x 12 4030 93802"/2 Temper 25 16 1/2 x 3 x 13 Beam 4 x 12 4360 10570 2.11/4 Temper 25 11 3/16 2 3 x 13 Beam 4 x 12 5320 11700'/4 Temper 25 24 1/4 • 3 x 13 Beam 4 x 12 5230 11860/14 Temper 27 17 1/2 x 3 x 13 Beam 4 x 12 5290 13080

"*tll Temper 23 20 3/16 x 3 x 13 Beam 4 x 12 7700 17160•'ull Temper 25 23 1/4 x 3 x 13 Beam 4 x 12 7350 17210-'ul. Temper 25 13 1/2 x 3 x 13 Beam 4+ x 12 8640 21130"*ull Temper 26 15 3/4 x 3 x 13 Beam 4 x 12 964o 26830

zentric Ring Method, numbers folloving CR Indicate the diameter of the rings in inches.

sear Loading Method, nmbers following Beam indicate stress span and load span in inches.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _8

BLE 4

GTH TEST RESULTS AND CORRELATION

MeanSurface % Deviation

Mean Mean on Mean Basic Strength fromCenter Surface Center Breaking Standard Plus Surface BreakingTension rDresg n Ratio. . Stress Deviation Range Zomoression Stress

6530 45770 7.07 51180 7230 2820030 14330 2780 11300

6950 56970 8.21 67010 6540 40000 71300 + 6.410020 2310 9600

9270 2080 89007930 3090 9800

3 breaks 10760 2430 90007410 4020 15400Sbreaks 9580 4150 14700

235 10740 1200 4900311 10570 2890 10200

8680 17980 3130 13000 18700 + 4.04360 9800 2.41 21980 2560 9800 19070 -13.24310 9240 2.14 23420 4310 16900 20000 -14.6280 9920 2.22 23510 2150 7?00 20660 -12.1

11160 22040 2620 3.1 00 21180 : .96430 15900 2.47 29360 3110 14100 25170 -14.35090 12540 2.47 27090 3380 14600 23300 -11.o5320 12910 2.42 27460 2880 11500 23350 -14.96870 15270 2.22 26370 3280 13000 25290-.17160 17300 2.42 30780 2180 10100 26570 -13.77850 18980 2.32 29920 1 27400 26910 -10.1

breaks 31930 15400 29740 - 6.98190 23770 2.90 32060 1190 35800 31180 - 2.8

breaks 36930 7640 2.00 33350 -9M44o 18810 2 32610 3350 14600 29550 9z

23370 35990 39 12000 339 0 5.7225 10920 1?40 7200343 10970 1280 0003970 10340 2.61 19350 1830 9500 21260 + 9.9

3450 10920 3.14 20790 2150 73O0 21890 + 5.35190 13410 2.58 24340 1780 7o00 2+330 +. 0.15540 17500 3.15 25690 3500 12000 25470 .10.87750 2115o 2.73 30630 2220 10700 32070 * 4.77400 23080 3.12 302 0 3410 14200 34050 +12.6

147 10940 2610 10400142 9050 1920 7200256 12410 2260 11000171 6670 3410 16400

4030 9370 2.31 21570 2080 7800 20310 - 5.94030 9380 2.33 20250 2780 10500 18430 - 9.04360 10570 2.42 21470 2510 10400 22980 + 7.05320 11700 2.19 23750 2750 11100 22640 -. 75230 11860 2.27 20360 2630 10700 20910 + 2.75290 13080 2.47 24590 3230 15900 2490 + 3.77700 17160 2.22 270N0 2190 9700 28100 + 3.7350 17210 2.32 23700 2800 9700 26260 +10.88640 21430 2.48 30560 3240 13100 33840 + 9.79640 26830 2.78 36500 3690 16300 33500 - 8.2

rings in inches.

id span in inches.

8

STJDY CF LVIITATIONS

The limitations of the DSR, in its basic design, with respect to theeffects of curvature, waviness, edge proximity, and surface finish weredetermined by direct examination. Initially it was assumed that glasscomposition would not affect the limitations of the DSR and this work wascarried out using only Type I MIL-G-25667 glass. However, in the extensionof the basic design to the other glass types, especially Chemcor and HerculiteII, there is reason to believe that the limitations of the DSR may not, ingeneral, apply to the MDSR-I and MDSR-II instruments.

Curvature

The limitations of the DSR with respect to surface curvature weredetermined on four basic surface types: (1) concave cylindrical; (2) convexcylindrical; (3) concave spherical; and (4) convex spherical. In the caseof cylindrically bent glass the limitations must be established for twoorientations, one, when the optical axis of the instrument is orientedparallel to the cylinder axis, and the other, when the optical axis isorient' d perpendicular to the cylinder axis. Thus, a total of six caseshave *veen studied.

The limiting radius for each case was based on the quality of the stressimage observed in the DSR telescope, i.e., the point at which the imagequality deteriorates to a degree sufficient to affect the readability of theinstrument. Also, the radius of curvature at which the stress image disap-pears, or beccm-s impractical to read due to large variations, was determinedin each case. The results of this study are listed in the following table:

TABLE 5CURVATURE LIMITATIONS

Surface R1 R2

Concave Spherical 175 in. 100 in.Convex Spherical 250 in. 125 in.Concave Cylindrical parallel to Axis <5 in. -Concave Cylindrical perpendicular to Axis 200 in. 150 in.Convex Cylindrical parallel to Axis 100 in. 5 in.Convex Cylindrical perpendicular t+ Axis 150 in. 100 in.

Legend:

R, - Radius of curvature at which stress image quality begins todeteriorate.

R2 - Limiting radius of curvature

* - This limit has not been estaU--shed but would be less than 5.

9

At the limiting radius, R,, the stress image may be poor, but it isstill pT-^lt .... C -ne the -urCfae st.ress with a reasonab'_e amount ofaccuracy. Beyond this point, the stress readings become Lznpract..i.a. withthe DSI in its pf-•ent design. The term "reasonable amount of accuracy" isrelative and may be given more meanring when considered in light of thefollowing remarks.

When several readings are taken withl the DSR at one location 3n a flatsample of fully tempered glass, a variation in the readings of 2-3% isnormal. In the above work on curved surfaces, when the variation in readingsapproached 10-15%, the author considered this to be "reasonably accurate".Beyond this point, the usefulness of the DSR, in its present design, wouldhave to be determined by the application.

The variation in readings is not affected by a deterioration of thestress image as much as by a tendency to wobble or reck on the curved surfaces.This rocking can be correcCLed by proper changes in the design of' the D IT

instrument.

Other Limitations

The limitations of the DSR with respect to waviness, edge proximity,and surface finish were determined for flat samples only. These results aresummarized below.

Waviness: This will not be a problem in the use of the DSR on air,;-aftquality glass since the waviest sheet glass is readable.

Edge Proximity: The limitation on edge proximity is the prism contactlength, i.e., as lorg as the length of the prism is in contact, readingsmay be made at the edge of a sample. When making readings at an edge,however, it must be remembered that the stress is no longer isotropic in thisregion and that the stress instrument reads the surface stress in a direction900 to its optical axis.

Surface Finish: Short finish presents no problem - as the DSP iscapabTeof reading through surfaces ground to a matte finish with 800Xabrasive.

10

REFFER E1C nzS

(1) The theory of opera-ion of the DSR is discussed in Progress ReportNo. I. Ansevin, R. W. and French, P. W., C-rrelation of the DSR withthe Strength of" Glass f Different C..posi ns and Configurations,Progress Report No. 1. Ccvering the period 1 ju!Y 1963 to 30 September1963.

(2) Timoshenko, S., Strength .of Materials, Part I, Second Edition, D. VanNostrand, New York, 1940, p. 61.

(3) Roark, R. j., Formulas for Stress and Strain, Second Edition, McGraw-Hill, New York, 1943, p. 1b9.

(4) Vitman, F. F., Bartenev, G. M. Pukh, V. P., and Tsepkov, L. P.,A Method for Measuring the Strength of Sheet Glass, Translated fromSteklo i Keramika, Vol. 19, No. •, pp. 9-11, August, 1962.

(5) Vitman, F. F., and Pukh, V. P., A Method for Determining the Strengthof Sheet Glass, Translated from Zavodskaya Laborat"rya, Vol. 29,ITo. 7, pp. b63-867, July, 1963.

ApPW

DZTIAILED) PROuEDURE

CALIBRATION

Calibration Methods

Type I MIL-G-25667: The calibration of the DSR for Type I MIL-G-25667glass (polished plate) was determined by subjecting 10 fully tempe--d samples,of size 6 x 24 x 1/2 in., to beam bending using a Baldwin Testing Machine.Each sample was simply supported, with a load span of 22 in. a. a stressspan of 6 in., in such a manner that the surface which would experience thetension during the test was in a convenient position for examination withthe DSR. The DSR was placed on the sample in the area over the 6 in. stressspan and oriented to detect the resultant surface stress due to loading.The DSR reads the surface stress in a direction 900 to the optical axis ofthe instrument. At zero load the DSR reading corresponds to the magnitudeof the surface compression introduced during the tempering process. Loadingwas applied in suitable increments until the burface being examined was intension; the degree of this tension was held to a level below the expectedrupture point in order to avoid sample failure. At each load increment,the corresponding DSR reading was noted by taking several measurements.

The average DSR readings were then plotted against the surface stressescalculated fram beam bending theory, see Calculation of Bending Stresses,Appendix I. From this data, the surface compression at zero load wasdetermined by passing a line of least squares through the points recorded.This value obtained for the surface compression at zero load divided by theDSR reading for the same condition produced the calibration value, K, inpsi per micrcmeter eyepiece unit for each sample. The data for a typicalsample is illustrated in Figure 1. The linear relationship between thesurface compression and the DSR reading is quite evident in this illustration.

The results of examining these ten samples of Type I MIL-G-25667 glassdetermined the mean calibration value of the DSR for this glass to be 60.18psi per eyepiece unit with a standard deviation of 0.41. The stress-opticconstant, C, was found to be 5.4 x !07 psi per unit birefringence. This isan average result for the ten samples.

The pertinent data of the calibration study for this and the other glasstypes is contained in Table 6. Figures 1 through 4 illustrate the linearrelationship between the surface stress and the DSR readings for each typeglass.

Chemcor: The characteristics of this glass differ greatly from thoseof polished plate for which the basic DSR was designed. Therefore a differentversion of the DSR was required to measure the surface compression of thisglass. This version is termed the Modified Differential Surface Refractometer-Ior MDSR-I. Also, due to the sample size and thickness limitations at thetime this glass was procured, a considerably smaller sample in comparisonto the sample size of 6 x 24 x 1/2 in. used for polished plate, was used in

12

Applied strea s,psi 22,000. Mean DSR

Y Applied Stres s-PSI Reading-Drum Divisions

S0 303.01823 268. 3

20,00 3646 233.8

5469 204.57292 179.09115 148.8

18,000- 10940 116. 312760 90.514580 64.516410 26.0

16,000- 18230 T 1.020050 T 28.2

Calibration Value K = 18, 223To= 60.14

14,000 o303.0

12,000 -

10,000 - Line of Least Squaresy = -61. 25x + 18,223

8000 -

6000 -

4000 -4,

2000

x

0150 0 s0 100 IO 200 2501

Tension Compression DSR Readings, Micrometer Drum Divisions.

Figure I - Representative Calibration Results for Type I MIL-G-25667 Glass

(Polished Plate).

13

Appli

the calibration procedure. The size was nominally 2 x 7 x 1/8 in. Twenty

samples of fully tempered Chemcor were examined by subjecting them to beambend•.n•g with a ld span of 6 in. and a stress span of 2 in. Loading wasapplied using a Instron Testing Machine. Due to the small sazrple thickness,it was not possible to continue taking MDSR-I readings until the surfacewas placed in tension as was done in the calibration of Type I MIL-G-25667glass. The sample curvature became excessive after atout 2/3 of thesurface ccmpression was removed by loading with a subsequent deteriorationin the quality of the stress image. As a result of this deterioration,loading had to be discontinued before the surface was placed in tension.Consequently, the surface compression at zero load had to be determined byextrapolating the line of least squares for the points recorded. This needto extrapolate introduced a greater variation in the results as evidenced bya larger standard deviation in comparison with the results for Type I MIL-G-25667 glass. The results of examining these twenty samples of Chemcor glassdetermined the mean calibration value for the MDSR-I to be 173."A psi pereyepiece unit with a standard deviation of 13. The stress optic constant, C,for the surface layers of this glass was fcund to be 4.4 x 107 psi/Unitbilrefringence.

As in the case of the Type I MIL-G-25667 glass, there is a strong linearrelationship between the surface compression and the MDSR-I readings forChemcor glass, see Figure 2.

Alumino Silicate: The characteristics of this glass also differ frcmthose of polished plate to a degree sufficient to require a different versionof the DSR in order to measure the surface compression. This model is termedthe Modified Differential Surface Refractcmeter-II or MDSR-II. Otherwisethe calibration procedure was identical to that described for the calibrationof Type I PIIL-G-25667 glass. Ten fully tempered samples 6 x 24 x 1/2 in.were examined. The mean calibration value of the MDSR-.II for this glass wasdetermined to be 167.2 psi per eyepiece unit with a standard deviation of2.5. Figure 3 contains an illustration of the calibration results of atypical sample.

Herculite II: The same version of stress instrument as used on Chemcor,the NMSR'I, was used in determining the calibration value of this glass.The same procedure and technique as used with the Chemcor glass was appliedhere. Twenty samples, 2 x 7 x 3/16 in., of fully tempered Herculite II wereexamined. The mean calibration value of this data is 189.3 psi per eyepieceunit with a standard deviation of i1. The calibration result, for a typicalsample are illustrated in Figure 4. Further data is contained in Table 6.

Index Measurements

The surface refractive indices, ne, and n,, for light polarized Xileland perpendicular to the surface of the sample, respectively, were m( -edfor all calibration samples, see Table 6. The surface indices for tLe l'arge,6 x 24 in., samples were measured using the corresponding stress instrUmentfor that glass. Surface refractive indices can be measured with the stressinstrtuents by using a glass sample of known index as a reference. Thesurface refractive index, n, of a sample can be readily determined by notingthe angle between the critically reflected image of the reference sample and

14

pplied Stressib

Mean MDSR-I18. 000 y Applied :,tress-PSI Reading-Drum Divisions

0 255.6

1631 249.8

3261 242.216, 000 4892 234.0

6523 227.88154 216.49784 206. 0

11415 189.414, 000 13046 183..2

14677 176.,4

+16307 166.2

12,000 Calibration Value K - 44, 991

+ 255.6 - 176.02

10,000 -+.

Line of Least Squares

8000 + y =-172. 65x + 44, 991

6000 -

4000 -

2000

x

04

170 1O0 190 zoo lb., U Z3 Z0 Z5

MDSR-I Readings. Micrometer Drum Divisions

Figure 2 - Representative Calibration Results for Chemcor Glass

15

mo

Applied Stressp'si 22,000 Mean MDSR-II

Applied Stress-PSI Reading-Drum Division

0 137.620 001410 131.2

282o 120.84230 111.6

5640 104.87050 94.2

18,000- 8460 85.49870 75.2

11280 68.012690 63.0

16, 000- 14100 53.415510 44.416920 37.018330 26.8

14, 000- 19740 21.021150 13.2

+12, 000- Calibration Value K =23,087

137.6-

167.78

10,000 - + Line of Least Squares

y =-168. 46x + 23, 087

8000

6000

4000

2000-

01

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140

DSR Readings, Micrometer Drum Divisions

Figure 3 - Representative Calibration Results for Alumino Silicate Glass

16

i Applied Stress;psi Mean MDSR-I

35, 000- Applied Stress-r-I Reading-Drum Divisions

0 312.8

32, 504396 283.0

8792 266.2

13188 244.4

17583 223.430,000- 21979 203.2

26375 177.6

30771 148.027,500- 35167 122.4

Calibration Value K = 59, 315328= 189. 6325, 001- 312.8

22, 500 - Line of Least Squares+ y =-189.59x + 59, 315

20,000-

17,500- +

15,000

12,500-

10,u000

7500

5000

.

2500x

0120 140 160 180 200 220 240 260 280 300

DSR Readings, Micrometer Drum Divisions

Figure 4 - Representative Calibration Results for Herculite II Glass

17

TABL9 6CALIBL.TIOX, SUIACE MrUo AND STraS OPTICAL CONSTANT RMSULT$

i Stroes OpticI ~Costant

Surface Stressmeter Directlg Measured Calibration Value C x 100Stress, de*&ps' psi/unitstres Sa oprassion Ki!l1alae: K, Apsl/B'kepiee.

q 3 60.69 1.5251 1. 242Tpe I 0 S6'7 25 3P- ?1 1Ty'pe 1 m"2667 051 p 18021 101.8 3:~ 1O6 1.R5 .Tipe I NIL-C-? 67 PP-ITrype I MIL-G-2%667 DS0t 3165 .3 1. 250

.60 60.06 1.5250 1.1246Type I MIL-1- 2667 8 PP- , 1, 2 '2.0 :61 60.' 1 .52.52•6Type I MIL-G-25667 OSPt pP 219 1.59 1.2. ,

316: ANN' Z2: 1:521*752Meen Meon Meamn Mean60.18 1.5250 1.5216 537

SD0.1*1

•l1uaino Sliaete 1•SR-I! £3-1 2 851 1•*7.1, 1*.22 161.83 1.5580 1.5576 565tLuaino Silicate I•R-II A5q-2 2421 19..9 1,.12 173.82 1.5m80 1.5575 5A t u m in o S il i ctee ~ ) 0 1- , ! £ 5 2 38 7 0 1 1 * 4. 6 4. 1 * 1 6 5. 3 6 1 . 5 5• 8 0 1 .5 7 65 7AluT eno Sil icate ?6DSR ! - 0 12928 1165.2 1,.12 169..79 1. 570 1. 5576 576Alual-no Silicalte P•SR- I! s-AS< c71• 11*9.6 1*.2,8 166.25 1.511.5576Alualno Silioate :'•Sa-r AS 21* 261.O11.4.16

.llaio ilcae I3-1I £39 1*01*'.o2172.12 1.5580 1.5576 6023 an itee ••-I! A-9lO1 .•170.27 1.558 1.5 76Ai laino Silicate .•S•*R-I £5.0- 1*O 115,.0 1*. 15 165.7`3 1.5580 1.9•7"Mean Keen Mean Mean167.2 1.5250 1.5576 587

SD2.5

Cheaoor 24358~q-? ! -1 1*01*76 269.2 10.55 15;0.36 1.* 127 1.5119 B""no M4 : 0-2 4121 272.8 10.69 161.73 1. •1 2 1.5122=•.aoo t S.-1 AS3 1,5 269.6 0o.57 165IN 1.5150 1.,120 52uai:or 11- t DSR -1- V*231 261.6. 10.33 1,79 1.514 1.512 5g9bAamcor 3038-I CC-S5 41751 266.0 10.13 16.56 1.551 01.5125760 5Alhumior 21DSt-I ;C--6 126 27. 10.71* 168.11 1.5101 1.5121

2717. 10665 1 .1-2 1.5121 1*8checmor :osItI 12- 6 274 162.12 1.5180 1.5126

lo 7185:.? .zoz.1o 17Chaueor 3D51-I tC1 :-9I3 268.6 10.18 182.12 1.5130 1.5119 76 5Cheacor 301SR-I 1 AS2 15.0

126.22 1.5127 1.5120.he*:or N058-I cc-i0 1#0499 259.6 10.02 17.Chmeaor MDS3-I CC-12 1*8164 270.0 10.58 17801. 1 1.5110.he ;or XDS8-1 C-1 480 270.6 10.62 17. 1.5130 1.5122 128

Chea:or 4D3S-I CC- 11241 262.6 10 1

07.0 10. 1 1791'.16"2 7 1'012. ,315120 4"•Chemor MD)SR.-1 CC-5 I I5 600t 10- .5 1 0 1.512010•.h morMDSR* I CC- -26 27 .0 10.6 1.12ChM4or :4S• Cl •,2226.79 1Al.l 1. 1 11.5•122 462

Cheor :R-I CC-19 4"61 261.6 10.275 190.25 1.51 2 1.5120 1386chea:or MS3-I CC-20 1*3018 272.8 10.69 157.69 1.51301.5120 1*02

'seHar Hmean Mean165.47 1.0 1.5121 162

13.

Ch:!rcle I? )V5-I 1C-2 n 5. 10.95 8 .21. 1. 51 215123

-o=Lt I 13• * 02 1. 11.97 178.• 1.,1` (1:1, 172Her- ulite I. :rMSR-I 1 513 21.2 10.60 160.22 1.512 1.511 2her I I! :4sD - i C •.609'1 66 11. 1 87 : 1: 2 1 411 92hor II ,038-I 11-5 4B7P 16.6 11.97 180 32 1.5151 1,58her eul te I! : 05*-I .- 6 58626 065 2.7 1 5 19 20 .2 51 1. 5 1208

C,:.:r ýOS. ! I"' CC;7 14 6921..b.F 189 1 152

h ::!:r.6 IIS5-I 11 11.91 25.176 08 1.5 1 1.51 21 419ieorouli7te I1 ;458- 3 1, 11. 1.5123.He.or, !! 3-I CC-9 97 2.6 11.2 11. 201o.77 1." 1 0:1.52

Coresi. I 3 R-I ,-li 960 21.2 180.2 5 1. 51: 1.51231*Me*rl I-I. I 11-10 318 07.8 5.6 15 1.51* 1.02

Heraul,. iI PO•-I 1-12 6 10.2 11.72 1 21,11_ * 11Itorc"ulie I!I 34D358-i -•iW.8 i92.6 1.•15 .112 5162er2lie 16o~I ;g.4 120. 1. 152 3121 518llt,14)SR-I H-15 1656260 122 7 15 47H.r.,o 1,. , I 30,-1 -2-16 9105 1:6 2. 11.812 151 1:51143 50

lHorer lte II .MDSa- I H-1 51332 11.8 11.79 11.151 2.516Her ulite 1 MD318-I H-8 50177 120.4 11.1 -' 1.5 11*0 16Herculite II MD5*-I 51-20 160.2 1.1.152 .5140Hercul.te 11 3,,,-I 113-9 59117 : 2 11.1" 1.11*2

a 1:771 .:51 1*

ean Moean Mean Mean189.,3 1.51 3 1.5112 501

11.

Legends nlU. 3w•faca index of refraction for light polarized parallel to the plane of the surface.n- S urface index of retraction for light polarized perpendicoular to the plane of tho surfaco.

Sd - Standard Deviat12on.

gulie MDH 1 2 611.72JB.9 18

the critically reflected image of the ,ar..Te ýr-g exPined. Fcr a stressedsample, two images will. be fcrmed, one .r-re >AJ.g t--- and the other

oepondiz.g --ne sur!ace refrac"2ive index is given by

n - iN2 _ Ein2

where N is the prism index of the stress i.nstrment and i e

•.e intrnen, ad is the emergent

angle of the prism for a given sample, i.e., the -a&grp at which the lightleaves the prism as measured fromr the nrormal.

The surface indices for the smaller. 2 x - in., samples were measuredby reflected light of the proper polarization using a Bausch and Lomb, Abbe- ýRefractometer. The proper polarization was a-ttaind by fitting the eyepieceof this instrutment with a polarizing filter. By rotating the eyepiece 90*,the two indices, n,, and n,, were measured direoti,.

The difference between- the two surface .ref.-actCive indices, n, - nL, can.be measured directly with the DSR in terms of eyepiece units. The ftllowingfactors apply to each glass type and their respective stress instrument:

Type I MIL-G-25667 g2ass using the DSR, 1.12 x '0-6;Chemcor glass using the MDSR-I, 3.92 x 10- 6

Herculite !I glass using the M.DSR-I, -.z8 x 10-6.Alumino Silicate glass using the IDSR-II, 2.86 x 106.

By multiplying the stressmeter reading, expressed in everiece urnits, by theproper factor, as determined above, we get the absolute salue of the indexchange directly. These converi..n fa,,, are good only in the index rangeslistea in Table 6.

DESTRIXTIVE STRENGTH TESTrS

Testing Methods

The beam loading method involved loading to breakage of 3 x 13 in.specimens supported 1/2 in. from either end and loaded at the two points4 in. from the supports. In the concentric ring riethod, a square plate wasloaded by a circular member which moved coaxially to the support ring.Three different pairs of ring sizes were u•ed or. three different samplesizes: 3 and 6 in. diameter rings with 6-l/8 in. square samples; 6 and12 in. diameter rings with 12-111. in. square samples; and 1-1/2 and 3 in.diameter rings with 4 in. square samples.

Both the beam loading and the ..on.entric ring loading methods were usedon the Type I MIL-G-2566T glass; then, on the basis of the results, theconcentric ring method was selected for the other glass types. The concentricring method was preferred over the beam loading method because of the largenumber of edge breaks encountered in the latter method. The evaluation ofthe strength of glass is a relatively comp'cated problem; It is know thatthe strength depends on a numiber of various edge and surface effects. There-fore, if the strength of glass with a specified surface treatment is to be

i rnv u m n

determined, it is necessary that the measured strength be characteristic ofthe Properties cf the sur+fc'e ver a&• -o-- c..nta. • rrors.

a possible effect of a weakened edge. The ._orn'ent-ic ring method, of proper....e-, .. cludes.to effctco oed defe:lý from t.e results or strengthmeasurement s.

In reference to concentric ring tes':s of proper design, it was initiallyintended that only the • and 6 in. diameter rings be used to test the Type I"MIL-G-25667 glass in a-.- thicknesses, however, a large nirmber of edge breakswere encountered in the testing of the 1/2 in. and 3/4 in. thicknesses. As-his testing method is intended to eliminate the edge effects during strengthmeasurements, larger samples were prepared to determine if this may hrave beenthe result of the =a!' sample size relative to thickness for these groups.Samples 12-1/4 in. square were tested on a 12 in. diameter support ring witha 6 in. diameter load ring. The results show that this corrected the prohlem.Another way to eliminate this type of break would be to have the sample extendbeyord the support rng y•. a greater amount. When the latter method is used,however, correctimons for the overhang have to be included in the calculationof the breaking stress.

Testing Techniques

Specimens having cut edges werc broken with the cutter marks on the sideunder compression. This surface was taped in order to preserve the originof failure. Loading rates were in the range '0f O00O to 15000 psi per minutedepending on the available loading rates of the testing machines used. Theloading rate used for a given test, however, was held constant to within+1000 psi per minute. 'When Possible, a loading rate of 5000 psi per minutewas used. These are the loading rates which were maintained during the latterhalf of the destructive tests. When necessary, in order to keep testing timeswithin a reasonable length, specimens were fast loaded to approximately 50*of the rupture strength. The temperature and humidity were held withincertain limits during the tests and at least 48 hours previous to the tests.The temperature was maintained between 70_750 F and the relative humiditydid not exceed 5,0%. A minimum of 20, and in general, 25 samples were testedfor each category listed in Table i. Approximately 50 samples, however,were used in the testing of glass types Chemcor and Herculite II.

Each sample was examined for defects which could possibly affect itsstrength. Samples whici' contained defects that are not normally representativeof that type sample werc rejected.

The location of the fracture origin and the direction of the maximumstress were noted for all samples. If a fracture origin occurred outside ofthe area of maxiunm stress, the breaking s-ss was given the value of thestress at the position of the origin rather than the maximum stress thesample experienced. The methods used in determining this value of stressare listed in the following section.

The amount of center deflection of each group of samples tested by theconcentric ring method was monitored in order to detect the presence ofmembrane sl es. When the deflection of the center exceeded half thesample -LhicKness, the degree of membrane stress was est-mated and the breaking

20

stress corrected tc eli•inate this effect. Groups of samples for which thisoccurred will have both the breaking stress including membrane effects andthe corrected breakiih slvtf'tbz ilzlted in the talvcs -f .. d...... ...... un

results. These tables are contained in Appendix II. This effect was foundto occur only in the thin glass samples which were subjected to high luadb;q/4 and full tempered 3/16 in.-thick Type I MIL-G-25667 glass; full tempered1/10 in.-thick Chemcor; and full tempered 1/10 in.-thick Herculite II.

Calculation of Bending Stresses

The data resulting from the destructive tests and the calibration studywas converted into maximum surface stress by use of the following formulassolved from bending theory(2)(3)(4)(5):

Rectangular Specimens (Beam Bending)

FL (2)amax. = 3/2'

where amax. = maximum surface stress within stress span, psi

P = load in pounds

L = effective span, i.e., the load span minus the stress span,in inches

w = width of spezýimen in inches

t = thickness of specimen in inches

Square Specimens (Concentric Rings of 2:1 Ratio)(3)(4)(5)

PUmax. 0 0.545 -

This fcrmula, however. is not corrected for overhang while the following

three are corrected()(5):

1. Formula for 6-1/8 in. square samples on 3 and 6 in. diameter rings.

azmax. = 0511 7

21

2. Formula for 12-1/4 in. square samples on 6 and 12 in. diameter rings.

p

10max. = 0.510 T-o

3. Formula for 4 in. square samples on 1-1/2 and 3 in. diameter rings.

PU3max. = 0.463 -t-

where amax. = maximum surface stress withI-i the load ring, psi

P = load in pounds

t = thickness of the specimen in inches

Corrections for Fracture Origins Outside Maximum Stress Area

The value calculated for the maximum stress at which a sample failedwas reduced by a "fracture origin factor" whenever the fracture originatedoutside of the maximum stress area. The graph in Figure 5 was used for the3 x 13 in. beam loading specimens(i while the graph in Figure 6 was usedfor the concentric ring loading specimens(3). In the concentric ring loadingspecimens, whenever the direction of the fracture stress was at some angleto the tangential and radial stresses, the fracture origin factor was estimatedby interpolating between the two curves of Figure 6.

Corrections for Membrane Stress in Concentric Ring Testing

Whenever large deflections are required to cause failure of a sample,i.e., whenever the deflection at the center of the glass exceeds half itsthickness, a method of estimating the breaking stress which excludes membraneeffects is more desirable than the misleading large strengths given by theformulas 1, 2, and 3. The theory for the concentric ring loading methodfrom which these formulas were derived assumes small deflections and purebending of the center portion of the plate. Fcr large deflections, theglass may stretch like a membrane as well as bend, giving rise to membranestresses. Since the mathematical methods for treating membrane stresses arecumbersome and lengthy, an empirical approach was used to estimate thisAffect.

It is known from the existing theory that for pure bending the load isproportional to the deflection, while for pure membrane stressing the loadis proportional to the cube of the deflection. Thus when a sample requiresa deflection of greater than half its thickness to cause failure, an examina-tion of the graph of the load versus deflection in the region beyond the halfthickness will show a nonlinearity due to the presence of membrane stresses.A correction factor for this membrane stress was estimated by forming aratio of the linear portion of the load at failure, determined by an extension

22

1.00

.90

. 9o

0

u .70-

"" .600 50-

U .40-

.30-

.20-

.10-

0I0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

Distance of Fracture Origin From Center

Figure 5 Corrections for Fracture Origins Outside Maximum Stress Area for4 and 12 In. Span Beam Loading Method

1.00

.80 - Tangential Stress

04)

0 .70-1..

0 .0-R ofdia Fau OintR

Fiue6-Crecin o rctr rgn utieMxmm Stress Ae o

S . 40--

, . 30-

.203

.10-

0 0.25 0.50 0.75 -1.00 1.25 1.500 .50 1. 00 1. 50 2.00 2. 50 3.00

0 1. 00 2.00 3.00 4.00 5.00 6.00Radius of Fracture Origin From Center

Figure 6 -Corrections for Fracture Origins Outside Maximum Stress Area for

Concentric Ring Loading Method

23

of the linear load-deflection ciirve, divided by the actual. load at failure.Fur a given group of samples, several estimates o'f the correetion factorwere me~de and the average value used to determine the corrected hrenk1ngstresses for the entire group. By this method, correction factors ýWereletennined for the following groups of ceniples which shoved membrane effects:3/4 and full tempered 3/16 in.-thick Type I MIL-G-25667 glass - a co:-rectionfactor or 0.9; full temp.:!red 1/10 in.-thick Themcor and Hercuiite II -acorrection factor of 0.8. Figure 7 illustrates the method as applied to asample of full tempcered 1Herculite "II. These samples were tested on "heInstron Testing Machine which gives a measure of the losd ring deflectiondirectly.

24

22, 500

2, 150 lb. JLoaad at Wai!are J

20, 00CEstimated Correction Factor 1,690

2, 150 - 07'

17, 500 1, 690 1b,

Linear Portion

of Load atFailure

15, 000"

Load versus Deflection

12, 500This lire is an ex-tension of the linear

0 portion of the load0-1 versus the defle '_on

10.000 curve.

7 5 t-

The existance of slack inthe loading systerm Ce

500- causes this slight no Center Deflection Equal

linearity when the To 1/2 Thickness

load is first

applied.,

250-

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0 08

Deflection at Load Ring

Figure 7 - Graphical Estimation of Correction Factor for Membrane Stressesfor Typical Fuli Tempered Herculite II Sample

25

INDIVIDUAL TESTING RESULTS

The individual results jf the destructive testing for each type glassare presented here.

The following symbols are used in this appendix to indicate the locationof the fracture origin and the direction of the maximum stress at failure.

Concentric Ring Testing

- Fracture origin located inside the diameter of the loadingring.

Ir - Fracture origin located at the diameter of the loading ring.

T 0 - Fracture origin outside of maximum stress area with maximumstress at failure acting in a tangential direction. Numbersin the blank before the T indicate the distance in inchesfrcm the load ring.

R 0 - Same meaning as above except maximum stress at failure actingin a radial direction.

T 0 0 - Addit'cnal information contained in the blanik following tneT indicates the angle to the ta-gential direct-ln at whichthe maximum stress at failure occurred. Other symbols havethe same meaning as above.

Or - Fracture origin located over diameter of support ring.

OE - Fracture origin located at edge of sample.

Beam Testing

IC or Ic - Fracture origin located inside the area of maximum stress inthe center portion of the smple, i.e., not at an edge.

IE - Fracture origin located inside the a4ea of maxiP-.un stress,

but a- an edge.

OC - Fracture origin located outside the area of maximum stressaway from the edges. Numbers in the blank indicate thedistance in inches frcm the maximum stress area.

OE - Same as above except fracture origin located at en edge.

Statistical Information

SD - Standard Deviation.

26

TABLE 7INDIVIDUAL TESTINO RESULTS, FULL TEMPER CHENMCOR

sAxPLL sizE 4 x 4 x 1/10 IN., TESTING METHOD 1-1/2 AND 3 IN. DTAKMLTR CONCE?•TRIC RINGS,TkESTIXG l(AcHinL - rNsTRox, LoADiNO RATE; 15000 Psi/mIN.

Breaking StressAveraeo Center Avetage Surface Surface Breaking Fracture Breaking Stress Corrected for

w i ' Thikness Tension MR -I Compression To Center Load Macture Origin With Membrane Membranewoo ly. Pa Redn Pat Ratio lb. Orgi Factor Stress. vel Stress, psi

1 .10110 6509 268.02 671 7.14 1 1 70400 56300*TIo3 6 s2 5 96 69900 55300103 2 2go.6 3J.54.0980 68 2 .6 158 6.61 1 10 I 1 69300 55500.1049 6297 262.6 15535 7.23 1516 I 1 63800 510009 .0987 6715 2741.1 1T7 7.09 I 1 1 46200 000

10 .1035 6361 270.2 1568, 737 170000 600011 .101 6138 269. 11671 7.61 16 1 1 70800 560012 .101 7 11g0 9 62187 I 1 71700 7 -

73 21 Z1 00 1702 1 1 62800 5800. o101163 3 2;4.6 6.6 1;66 62800 50200

15 .09(5 6635 259.0 11911 6.77 1210 1 1 58000 610oo16 .00 301.6 522 U7 1318 I 1 54500 11600

1 .1060 MOT I 8 20h.19 1900 Zr 1 30120I .1011 7260 258.6 N1841 6.17 11111Z 1 7500 5960020 .1011 6791 26.0 4390 6.53 1538 I 1 65800 5260021 .1038 7034 266 457e8 6.5o 1701 I 1 73200 5860022 .1001 7239 212.6 43107 5.95 1366 1 1 62800 502002 .1055 5992 308.2 4112 8.92 1;26 I 1 71900 572 .1003 7223 265.8 U6090 6.38 1 18 1 1 66700 Z34002 1052 6635 268.8 16610 7.02 1232 1 1 5160026 .1052 5(13( 303.6 2644 8.86 1831 I 1 76800 61400

2 .1050 6119 271.2 417026 6.9 1570 1 1 418oC51028 oo6 7200 26;.0 46298 6.13 1616 1 12n 1026 7161 26 .6 16228 6.1,s 1200 I 1 002 003n .1016 7051 26T. 46367 6. 138 1 16 5160031 ini8 7070 264.6 45882 6. 18 1550 I 1 700uo 5600032 .104,0 687 249.1 43216 6.29 1092 1 1 76800 11o003 •oQ 7136 266.1 1 4194 6.19 1728 I 1 80800 64700

:(nQZ 62 192.6 339 5.20 11811 1 1 5550035 .10412 6796 259.2 r95 .63 1166 I 1 49500 3980036 67 262.2 56 6.7 10 4 I 63600 509006O 17 257.8 6.7 160, I 6860 ,037 6 L 1 60 54900.1021 6817 250.8 4 111903 .121 685 250.8 439 6.5 1486 1 1 66000W0030 .105o 587- 251. 4•59••. 60820 8600

.1041 53 311.1 53997 9.25 1696 I 1 71700 57100111 .1006 6069 182.4 3162 5.21 6 1 1 45600 3650042 .1U4 g812 304., 781 16'22 1 6900

115 .009 726o 26N46 2 6.31 1386M 1 61000 52506 .•106 6965 254.2 78 1676 1 71000 001 .1035 59 j 1317.1 5 92 12561 Ir 1 67700 51103. 9O3 9.202 Z.43 1810 1 1 75100 60100.103 7 260. 513 1 1490 I 1 63100 5070010 .lO021 6826 250.2 135J 6.36 1120 I 1 19700

50 .1003 5861 223.0 6 6.59 1256 Ir 1 57800 620051 .0Q03 7109 251.6 ?118 6.21 1420 Ir 1 66700 53400

Mean Mean Mean Mean Mean6530 115770 7O07 63970 51180

SD SD9030 7230

High Higho00 6470U

Low Low45600 36500Range Range35200 28200

27

ICCA

10W m

1-3

* 0 0 0 0 0 0 0 S 0 S 0 0 0 0 0 * 0 0 0 0 00 5 0 w 0 1-6 I- 1-D.*

oA C8,0z 00

1-.30 H t-4

m / 0 zWW4WW~~~~~~~~ WWWfuWt )WJWWN OfU 01 -

co

O0 OXA 0 0 ) t

~O I W-j

1-

I'l

0

0 0 0

Pwwww* 0 00 cogh

~cp

001

W -40 P yoo v a

r.)%O %.n LA MA %

oq 0

0)0

10I

I-d paIaI' f'

"I.2LA10 v0 '10 c

0 r too 808088008 0 1;'00 0 0 0 00 000

INDIVIDUAL TESTING RESULTS FULL TEMPER HLRCULITE IISAMPLE SIZE 4 x 4 x 1/10 IN. TESTING MUVTHOD I-1/2 AND 3 IN. DIAMETER CONCENTRIC RINGS,

TESTING MACHINE - INSTRON, LOADING RATE 15000 PSI/MIN.

Breaking Breaking

- Average Center .'verage Surface Surface Breaking Fracture Stress pat Stress psiSamole Thickness Tension MDSR-I Compression To Center Load Fracture Origin. With Membrane Corrected for

No. in. psi Reading psi Ratio lb. Origin Factor Stress Hembrane Stress

1 .1000 6075 290.0 5k9 7.87 1880 Zr 2 8 4Too 6T7002 100 708 21.2 5J124 7 5 59 Ir 1 82200

0 607 209.2 5r6 7.8 1o Zr 1 82200.1012 6975 281.8 35 ;:65 1746 I 1 o8900 631oo

1018 6919 294 8 OT 1798 I 1 ;9300 6 200.1011 6975 V6.2 57964 8.31 1618 Zr 1 73300 58600

000oo8 6919 296. 56 1 8.11 1H90 1 1 00000

:1608 6975 303.4 5743Z 82 11r 1.1006 6975 22.6 553 1 1920 Ir 1 87500 70200

10 .1016 6919 29.6 5578 .06 2160 1 1 9680011 .1000 7031 293.8 55616 7.91 2000 1 1 760

12 .09o6 7031 295.6 55957 7.96 1756 x 9000 65500

13 1005 69T5 308a4 8380 8.037 178 1 1 81600 65300.1010 6975 296.2 56071 1852 Ir 1 84OOO 67200.008 7031 2904 54973 7.82 I5 I 1 80100 6r 100

.100 7031 290.6 17 .6 Z 1 090OO

1N 1019 760975" 309.2 R321876 1 1 8360069081 2 Ir 1 66o69200

17 :1019 6919 285.8 7. 9102rS.1005 0 305. 57o1 8168 O2o Z 1 83700 66900

2m 1037 7594 321.2 6803 8.01 2040 1 87900 7030021 301.8 57131 8.13 85 Z 1 8,700 67700

22 .100 708558 30 5754T 8.12 22 1 2 102900lO. 6 6581 287.2 5436T 8.26 1988 Zr I 8o10o 673002 1056 6581 306.8 58 8.82 1920 Ir 1 79600 63700

25 .1052 6581 315.6 3 9.08 2150 Ir 1 90000 7200026 103 6525 321.2 72500

2; in8 6 303. 50Z5 12 1 90500T202ý 6 ~ 355.8 5 10.12175 I I 90600 72

72 70.1 1 T300520 6638O 300.2 N6 1r 62

3n .0091 7031 340.8 64513 9.18 1708 1 1 80500 640031 .0008 675 288.8 5OO8 1602 1 1 1

32 .1005 6919 297.20 56222 .13 1802 1 1 6 66100

1007 6qlci 289.6 54821 7.92 2020 Z 1 9 73W

.101k 66' 2882 54556 7.957 1482 1 1 6800 o00Oq 6M• 285.8 54102 .79 1780 I 1 82700

:1000 6919 293. 55541 1912 1 1 87000 696003; .1009 6919 302.0 57169 8.2 1932 Z 1 87900 70300

.1012 6919 307.2 558153 8.40 1776 Z 1 80300 6t3000 .1012 6919 295.2 55881 8.08 38 Ir 1 83100 500

.1018 6863 282.8 5353 1536 I 1 6870041 .O101 6863 296.4 561 .1:8 21 1 1 850 80

.1013 1Oi 2.6 557 8.06 190 Z 1 02 63004? .01 6863 301.2 0o17 83190 r170690:1011 6863 318.2 017 I.8 1704 1 1 76 o 61300

45 .1016 6863 29:.8 o5 8.13 2065 Z 1 92700 74200%F .0o96 714 2 7.85T2500 Zr 1 116800 93400

781 313.6 5932? 7. 9 1 1 81600 65300k 10~ T42 312.0 59062 7.5 119N 1 1 990 900

n .100 708M 3-7.2 60046 ;.#? 2814 1 1 66250 .1008 7088 289.4 54783 J.73 1720 I I51 .100I 6919 311.2 58910 .51 165k 1 1 75000

Mean Mean Mean mean Mean6950 56970 8.21 83790 67010

SD SD8170 6540

I~o & '100

Low ow66800 53o00Range Range50000 30000

30

CJO

w w \r\HHH H H HH N-S-FtJ Pi3 H- 00 00 --a.Z (7'%".R -FL'W N) H O'.o 00-13 C\"In -F- N) H 0 E3

HrCDI

HD

M M

~F C\O ý-

N)CD~~~~C Hcor3I 3 H H w ~

OCDW C N N)H J ) N WM N N HF-JFJN) ) -IN) ) ) N N r N)N)N) : C

t-4 C COt~

r\)C\LAJ,3---3F- cr 0 .o

0 -r~a\ 0 \ ON0 N o C\ -row oo o :t.

HH4 HH H I-HH H H H $-4H H HH HH HH -i ~ 1-4 wqH" 1ý F

*C)

H F-I~' HIH H H H HHHH" -JPH HHJ FHH H H Hf' H "~~W~~~0-~~~-~~~HOc mC rct~r3 EDN~ rI ~O~0\ ICN~r~s -~O~~.OW ~ O-) ~a ~0AJo

0o~0 ~ 8888 888888ooooo ~m0(D 0 OTcl

Wn tjt- I- m WHX HH F-J -J a F- t- F- H iJ FI H J FJ F- H n $

-3 W-Q g -h -J :) CD 1ý H J \ W H -f FJ N -- NOK) 000 N N)I'Dw 0%D tC)

N) : N C a \0 D U UjOD\, -Q _Q \ -- 3 0LA)N) FJ9N808 --ýJ~j1W-4On 0 r 8: 0808 08 804

(30 o 0 000

O( 9 000000000F1 AF

cl)

N) 1- 0 .. O 0:)-j 'cy"In -P'W . N) H 0'1C) 0011 G\Vn r ) H 0 ýd

H l

~~ CDO

;j; ; *(CD "0)

cnCD En

CD C) ID\0OCD \ O -ý CF"ODH \0 )W N ) 00 0:)~o~ ',D F- 0 , )W- AEn ::130) H \0 'J.0 CD'O wL.J Q\J F' O f o0. O-r N-r Hý I- CO0:: - -': H o \%O\,n C'ODN) ONN), 0 LJ \-n OD .''J'HlHf N)-0 0- 0 CD H-

M tI; -4

r C\YJ 0 C\ a\ C7'j a\ 0C-Q GvJ \J1 O\ 0 \ 0 0" \\-R (D < 0) ýN,-4W N)-3aJ 3ýJ-7 A )\TLJO- A NDF ý" W CIW CDo0

l-A CD ~

0CFD: c o'..O\O\\\L N0H00H0 N)0.OOO HHA))LLJ,%0O\\) n( F-3- W0 N) H .PO \ C7NL.. \0 H %.R Z LA -- - -:1F-J F -J \0 CONO F- En w H tX3 -t

(-'D 0 H ýP-

o cro r\ -~ \ \ \ r. , \ \. \)r)mWQN ) 0 "ý

r\)~~( (D~rN wO I- =

0"D, w "J~J VC\ 1 4ýN -3\J "'j 4\jp -,a H3 .4F 0 D 0- D0)HHc \) Q) C=H

bn" ~A0"\,\A0 0 0o~ nl 0 "A 0 0C00o" 0 -tO I4 ,

01~' -j Iqc

CD) 0

.-. C-4 l

CD C

0dW~t-J M MF-Z H HH F F- - r F- J r H Q H Q," H r H "

COib%,n0 ý UO(D -HI-4)\D\. H -HH 3HH0HHHHaHO10-0J0H \0 0HO t I-(D H%jl::3 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ c-' Wi t o : j0 00C 9 ' " A P,+wwwM - o4,-ýN

8aq 0 :)'AJ ,n.: Q0080080080880080 0m A

CD 0 0 0'HeOC 0000 CoN3%)H o0 j 0. \0 C 0 t "Enl)

A::~~~~~~~LAJ~~~~~( ::3 oLJ o0.JH ~ O)J~-~~JN - l

o %,0 00)--. O"Ji- -P-W~ N) F- 0 ',D CO-.3) 0\',-n -FL.J W H 010c

CD Cl)

P- O\ 0"'I ON 0N 0' 0\\i 0'l- 0'CN 0' C-'3 ON 0'% GN 0\ 07"n-Q '1r

nor\ (70 00)G\-P-WCo.J 0)L.j 0 \-.n CN ý- F~-. (1) M(n CD

I~l (DO 03 cI-'CD\, N'.OHl\-'r 0)rLJý3L.J.J'r (H~ (D 00. 0' - O% 00r. r. \)P)W0*D-FWH0"FJE n

~~U IN.

1-40 00

Li. CDL ýD 10Ii-ý6 0

0 -

-P-w \ 0 J O 0 ,A ,n 3 H\ L ) -0\n H. 0)01,0N)L 0'j U) CD HZ ,'

0 0 t- H

0 V) 0 0I.'

*~~~ ~~~~~~~~~ 0 *z .- .* * * * * * * * * tD 1) I-L'410.m) ml~T. N 0i N) m m )'.wxN4rN.Nj N) K) w -P Hi

co:: M4 0L'J Co--. 0)'DH P- H'w \)'JV..)H 0 ct C) 08MtI

(CD (D

0o'-k"R" j , " o " o'" jl C)\ .0V 0 ClF-

o org ýv

0 H

H 0-1 C0F-J ~ ~ ~ ~ ~ ~ ~ ~ -F"HP*P - - - - - - FJH H ctc -

10 p

to,

O(DO~~~~OH 00 00 0 0 00 000

cI)

H0 '.0 C0-.3 O\~-PW. rl H 0\0 00-.3 GN\Ar -P r~j N 0 E3

H(D

I

. . In.<C,.

CA C

(D 0

\,n0 W \.F-I VI) r%) \0 G\H C\--3~ WN G\ .DAJ(.AJ \0 \0 \OLJ O\\O r- j t C/) ONIý-4

rQ LAjLJ W\) -P-W LA) N) H N) N) N) rQN) LUJ Hý r\) LJ N) H H D, Cl) CD C) HHjý*rL.'Jn\O H WM- 4N3N HN- N) H- C\ \\JILAJ H- H- C7\\Jl\,R r w '1 :1

w C))

ElCV) ZHC')r') r of)r)r or N\)NH ) ~N)) NN)HM"N) N) N) HHF 0 )tjH 0) -a L..i a\ -,3 a, L. OD~ -r N) L.-J H H - -3 \) OL'ý0 H N) LA. Eil (D :3j ý -3 t-4\.n;3LA O000-J A -ý \, LA.) \D O\LAO\O0\0 LAJ 0 LJ\0O C\hC 0- c 0H"H-0 w -:, LAJOC\ 0) O \ ONH -,0 C ; ON) fUJ -F'JN) (n 0 Z t ~Cn F-

HCD ->0 t

H

OC-) HO t.j

-13 0 0" o 0)0) 6ýn 3\:Q 0 ýJ ON-)-.ý3s)'o O N3O~ ct (D wLAJi :3 VkOHnHC\-3LHP -3N 'n -r Fa-,3C\N) M -~ 0 w ý

(D CD HA w~

ol M~

wd 0 C-1-P"-'JWA~J t4.LJ ~ .ULLJ~ CD 0 H

H-' \ 0 \A LAJN O -3O -QN) 0 A) P- \DOO0O\ V~ OL ýl 1 00 0 0 00 J% 00 0 0 0OO0 0 \n0 0 \. *

c-s C+

z

HHHH HHH HHH HHH HHH OH.O C+

HN) t- t~ CD LAj P A \ U A oN N) N)L ow A iN) N) r\)00 W 7\ -'a~~o ý- tV0 0 N)\0H 0N)HDOH\0HF- 0 H -- \00-)0 0-3 C\ 1 c+s(D

0on 0lrN) LA) 00000)0 00 00000000000 FýMP;OCDO 0 0 0 0O0000000000000000000 CnF-ý

co)

CD N

N~fJU1 fh U a\'1U -r N) C\ -4 -.- t _ NN )0O' \at~4V C7'U -o->51U 4oIDiU co R)A. A o K) co No o noK oonr Da '

I H i-i ,-~~~~~~~~~~~~~~~~~~ V4LtJtL~JCLLLL(J~JaLJLLA DC -it~

-~CD 'J1C-F0. fl~O'U~Jfh1~J OL~ JU~ ~ C trj:E U

F-1 h -HC

-F,~~~~ tH otz2Dtx %0) 0 D ) w-- \ wm DI~dm r1-40 \ \ wON ww

"I-- C-t-I

tol0\ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~h (-), o4\nýn- - \ 4c\nuo\q \cpuq

-C-PJ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1 0\n0OHPO0N O ' DL)O -J - - r0000000000000000000 at-

4I Loizi 'i b

tFl-4U

0 1-

000- Uo 0 0 U0 o- ot-n \00 0 00 0 0 00 \ . - 0 o0 000- 1 "PC .4aq ) \om - o 4 \ow o-r---4 p- m-n -- \. %.pN) N)\,n --4'-4004 0 0 00-40 00000000000000000 --000 0 0 000 0000 000 0000 000 ca: -:

'-4

0T

HO \ýo 0 --iCN, O'U ý N) V-3 O\O 00-4 CAP -P- Pi' 04'

ýn 6n C) CDý pn0 Cf)

0\0 QC'OGNO.0fQNl) -9.f XI-4jLw N3 LJL.J'.O .P A'-P, .d\J 0' 0\ N) W~ %.0 W~ -4 .5P.J 4-'. I--.rN 400! CX) -. mn

k'JCD o oj- r) w .'LA.9- 9- U$c7LQJ\J CN'-n \i-- H K)o b :3 (D

ci-

*l) HP-

0 Ct2.

0:3 p- zu\n 0- UIoH %..JI\0 \0\0 N) 01. 7L-j CN HC"nA) z. cf'FO VJI -4 uJ_ 0~ 0\ XND L'J roOU 0I' v OPi0N) 00%-n H- OH P oo (A 0 -Cl

F- (D-3 C

0

HP)0 ccn-W:4 W%3H NO I-WAWw wL~~ wLw L. o 0'

I-j~~~~~~~~~~~ CD-C H 00% )N 0FJWLAN LJ; l * ý m"tUýj~

%. o-j P \\nK)p-0\, N AJ- -4 n ). : 1

0'C Cl- 0 H 3

CC)\ 00.a cl-r\ \. 0\--) N)N)N)NN tr~r N) N)lpwkn)Nz 0\L ro3~ w"' \.\n v )p 0= --C,\D -4 M-4 -PooL. Hr O'J 07~J 004.f-J2)NL.J -CO oH-N3 " .

0~~~~~~1 Crw 0)qJ '.'ninCOo'0o;~- '.o ~ Jcn.

OCDO0 1-0 0 000000000000

FlFl

P-r,'- 0 H H r-l)a 0 %0.--4rN) (ON o 10 r%)N o ui N) r' 0' N

OCý - 9' ,n- QLJ )r )FJ1-- LAJJ -1 - --- 0 0 C

0O HO\

k)* ,r ro coo c,. ý

~ Ic~t''It3 -if T3-

F-I H-. (n104

10 z

00( 00 0 0000 0 F-ý-hO0 -j 0 HI-J 0 '1ý H

oý co 0n u C0O-1A 01 c jýr o --

LaJ~~~~~ ýJU~a tj O 'ZN

ý3% o~ 0 -

0 0- 0-6

0 ~ ~ U iC+

H F J - -- -H f - HH ý HH H -J %In0 H0 F HJlN)N)-C"W-9W -'OH 0 ý- -

"0 "0 NP' F-9

ýZL1t-4

C)

ýO 0 w

*Oqo0 O:T~OD10 000000000000000000000 F-*En0010 0 0 0 000000000000000000000 (1):

C/)

W~ -PWL.JU -MA,. Uj(A-.J L.jLjL~ L.jWW~t LAJLJJ .P- -C- -P-VThJn Z:)I6% q oD (Yo o f) N) 4-,%. x--4 w N) 00r) LA.)LIJ %.0 ?iN) A) NP,-r- CN C\ a'i N.0AJ~ LAJ -4 PILAJ r.o -- C1 00 .- z-r- (Dn

N)'U -1J -,X CDw00-,N OD\DU .- rLJ'~~ '"-4 ij-4N OL CDel

0.00 0 00 0 0 f-N) ~ Z\.n ~ ~ ~ ~ ~ ~ cC Pim0CQjWC% N 0-'P jm o

0 t'

I--t- P- P- F' P'o- H -- H tO- 3 H D~ t-- OOHCN 'a I= I03 .(D i-Puu\nu- 4q 7 DJ -4 c~o - coo\o \p\ a' %'.pO I-C " '

0:3 I-'JJ -OD 0' U #-a U 03%J' U \n \ 0\) ON) 0)J1"0 COO A) ) Hý O

0) C

0- C/)

W :iJZ W)(J INOW6 N LJL.LJJW IA) W WLLAJP.LA' P

ý. N) P- -" -D0 4 N)OýPN)P- LoOnR) W-4 -4U).-3

(D (D 0 I

000 -j\ %4 -PW JU A %"A~n 0 "N) W'J1~0 o q 0 Z

-V'\D --j 00-4 -P'O -rO0 00-4 -IONO N'%. 0 0 Mt 0 .)

H *n

t)

~ 0~10 H

:3 -4

C+0 C )

K) 0r CrAJ~ U' \-n m-r 0L~ ~0330 'no\-)-A-u -4N) p 03lo - 1f -- L0 'IC CD009 0 0 ýr0 '0 : 000000000000000000000 ji- 0 C-.0CD0 0 0 0 OO.00000000o 0 n

N)Nr11t t-j O tD-1iu 4:tt4 E

Pl oll~ \cl~ Ulcl --4~ll~ Ulr\\l 4Ll KH o) \o cCl)av ýN) 0

-r' .JClAi _-~ Lo) LA.) N) U'NLiJ f l% JLA-I WA. W. -C1A.LA rý ..-U- N "-r- 0\. 0 0 .CO-[ O--. -Pl--PUl 4 u'Oj . -rUkJ' * n 0)

i-~ -40a\ C0 o~o -r-A).o i-An H -4 cz)N)-l r\) .rl'Okfl (D n~

--4 --j -4 a\ -4. -- --1 -4 -4 --j --4 --3 O'O'C-4 ( D 0 com~C 00 -cwru. I-u - r c'-j OMU' u -u 0 \0\0o0 :3 COo~~~'~l-t.N0 -r '.- , D i oooo vl-4 LAj--j r) -r En :n F Xi-

ý- Fat t-'--.r~, "*JFc-'- H Fý 4

V4-W .i~j~ -91 rwLA wA ZUJ LA) w S17 -PtALw 4-%A) -91 m) wJi w mF-jPO-C'a\\I I- ,' o 00Il 0. L--4U10 Q10 \0 0N) 03C/) (D

D -lCO IN) t'j

N) 3 N) N)NMN) N)jN) N)r\3N) N)N)mN) N)N)IQN) JN)N) N)N) 1W (D LAJ WN)N)N)r3 ~r3-0--i LJL.',N) Lir.)LAi r) W H H r\) IV~ " " ý

0 -4 Oo\- OL0)-0 01-1P-0-4o-'4 -P\ L Ul -- Coo 0 En CD " - t~i mOD \0 O0 LN)N)LAJLAJO- 0-4- -4 9'0- W) 0 4co Hl

LA) : A A -. A ~ N)'. 0'W0 L O L4 1 N) W W W W -Jl -A-1 P 0 H;

(D (D

z- t1zH0j

F- -N) O"A).J A)'OOhJJ VI r,04o oOOO OO OCc\oo oo o cZ

0 N0

4 :z: m

0.- t4 txl -

'10'

r*OJ 0

ý-j a H FJH F" ýJ FJ -4 H F H -J H H 0 F 0 -3rt. on hI 0

C+ D

:cf Lo tx: COW UWLjWLj~LJP N) N) ~N)W oWN) N) N3N)W z " A1)N ULo On P- ro a) IýJL.JQ 4" -C-4 LA-'Ul N) N) F- \O \.00 U)0 N) C7\ o 1-1UJ F-uV, 1-0 oq0 0~~ Ul7 w \OOO000 00 0 0000 000 00 () :

oCD0 0 0 0 00O000000000000000000 I

NN) N) W .tA N) N) O\ 2. Nr%4-') - \ >- ~1.L~'N

0\H1

10-;4N I- . \ .p NJ N) t- N) -9 r* S- * *9ý LA. *r f\ I- La P* r% *. *n * S

~ *ox

tot-

0a'cx-4 cN-.J4-,-O-4 U-~ 'zr-4a ~ 1 t-.a-4. o'\Co-j . 0 l H3\D)--40\"0-"¼J OD OX~ WO \ %,0-. Co e-i Lo .C-N) 00 X- \0 U1 COU1 t--~ o 0) PV G

- 1-9

2 2 -M " 0

o~~~ W-H

LA) LA) u O 00 0 1

HH~~~~- HHH t-,HH H-' H

0tI5:

N~~~~t~~~0(I~~~~~I fON4 "'0 ' N 0O ))xLOO (l~ -t

CO )to~

trco P- :t t.A tý t" I--t-- clM X -!H

C '

ClU)

CD -

I-- Ir- 0-C (D'OD\ \C'. OD Mr OG CO a' OD O M%,0\0 \0 M'.0 M M M\'O '0 \0 M, Wi 0\0 Mv ::S # '-1 c)

\0~~~~~~~~~~~~~~C O-- N X' PZ-,.-r Vr - -% )N )N )ý -N A -N )

:r - En OD

K~~J1I) k~F.&)U -PFN) tU kp -r-L~-J -4 -LJ L.J tPW \_nU 4-J1 4- J -P 0 (~ D 0-*~~~~~~~~~~ * " W 13 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 0 t: 0 LJi"'

0 tli

0

0

0r N)0- cvcK) v-J'OD0 0, Q-) 0 r%)0 aJ0 N ' CL U,. \,.n'. oo LA) N i' N :, OD R) 1- C) rr% v V\ t-\ -ý 00 ) D \X.~O 0C 9-\OZ'0 \ 0 N .OO 4:- M(O-j'-4i U 03 PIV 0 -

:2 t-i

%,n N-1 t-4~

0 0 . . . . . .V1 0 .-~ 00.i -.~ . -. . 0F..qJ

00~~~o 00 zj H'

*~l 0i 0 0 o\ 0,-

J1t~)

t- N) LZC t 0.

DOJIL&04- 00 EUWý0%r)W 00W0b4 b414 C SCU 0 4 ýL)U D I-r Zu

DO A 0 00000 000 o)-C 00000- '.5"

7) (Do 0'. 0 0 00000000000000030000000 CA ::T)CD 0 0 Oo~oo~ooocoooO~o~ooo~o Cfl

(n~

(A Itto-

N

ChwAi\O R.) 'YnJ\ J ký -lý 0 rO PI. t, 4 J -- ,k- 0%k' * ** *n~ *j *n *3 0 *~ *n *n ; Y *" c*i * a n~~fONJU ~ ~ ~ ~ $- WO'aO~m~~0j lb~n

:3 b 04 c

:, 3"~9- 0%.~~ a'~ ~-rO-t.. 0%N) Ccm-nJ nfltv C-rt'jO%0%ONOr to~oO NO ~-4 -P'44 PC (,o %f~ -n~ j6 ? i%,-.r~j,' a pi - CAMt 14

()%Now - 0' cxCr-JuL.uJ 0 ak., M'tS-4ON --." .0- C"

0 000

3~ W3

;0 1ý iD0%knw-

FA 9-

N..~

*UP

to;4%

eQ0%

CA

al t

"X-t- m 0.V O nW O u -4 00 f) V ý i 04 0 f' " ( -4 -4 " W I.-row V, so M e

P~~ m l-Mstota~~ 8888

liJ Ocfl% 4-i..' NJ 0 ID CD'IJ Ol"J r-li N) ' D o "- a-~ '.fln~ r-%j low

aC a 0% w"jt~ 0%.n CD4-n")(0'-?) C) tDCID UOD 0b.l. 0C) l- UW"J M~- cn140

(A

cý~~~~~~~ ~ ~ ~ ~ -6. (7 ,0 y (aio IjN 3L y ,a ,0

0B 0 a, -.. 01 CA) C) l-1'O wJIJ.OO M-I0g

S-e

N) ) %J(l ri %j.n~ CND '..Jn X- o (1\ZK D'O 0 -- CýN N J w0%.o'JCTJ OL.) 10. m) t

'J :) " O JN 1 )0 -1 m4-c.) "ý N.) I'. CDDCO'J C1)~ OLj . 0'VJ)J WfC'JJ 0> r"% N0-j a

o Q C) O C) %- "NOD NWJD- 4- -'J K)I U- - J-J\ IA 0

o0 ) U

101"0U0

00 '-I

4-. Mw 0 , C 61 j .n-r :

-~~~~~~~ O'b :vr'N1L"vQ J 'l-t J gn~0000 0O0O'J'~000%fl'JO00.

:.- *4or

cl I

KiH zH

C, -

0

8.8 8 8~ 888 ~ o~~88~801

I- m00

CDt-.' Ain i Lo 'P %0 "WftNl-L.D . U P a

so.a 12 p

1J-J-JAtCD tv P w

t1i

co co)

02-.2 o7\%- -puj r\. "- 0ýo oo--4 a'ui- .4:Aw N) H ~ 0 o:0--4 CY"JI -PL-W N)~ 0 'o

CD a

00

:3-~ C)'C cz

-4jf r 0,m0w OOL.J-K0- m 0H-40Hm0m0mu-NJO'02LM C~D U0CCo ~ ~ ~ ~ ~ ~ ~ ~~~~E (Do 0 0~c0 -R~'o' ~ lH

r'-3

4- F-F 4(

'x~j 0 '-4

a' -AJ (. H0 63 C

t2l O<V210

0

F- F- -- F- - a) 4

Clt)n

~ 010

CA)

L. . . . .1A J. L. .E . ~ . I . . . . .r . -. -. . .

co W " 0%J"W.'"0 W% 0L. XJI C 0C..AJ"000 ESoq

N

000% 9-" %CL %ýW C-). (31.rf 4"A W4^ .FOU -rar -A 4-L-- 4fl0.VOLa 'JW WiC7 te

0c '$1AO toJýJ." OL.' 0L.j CJI 0'O'%D OJIV C.-"J CON ES cc044

0~

wp" - w PW 0--0 P- $. PI-- w ýw .V .w"P

KC)O-. MI-4I-

0~~ ~ %.A -4 % O--1~ ~

0~~*

.4Clow,

NX 0 V) -g

r"W 0.- r~ N I AM~ riJ ~ fIJ ~ N

4 0 -4 0, -r -4 -4 -~4Ad.0*, Piu PtW a7 ESý 0

00 0

0

-4-9"1H :zzoz7: f

C/)0)

O*),\U $1.W K) 1-- 0 \D c- ONJ TU1 .- L.J N) t- 0 -, :- a"s' ~-PIL N) F-- 010

H0

0 U' 0 0:) 00 U-) 0 I-j O \~,, JU~ Ou OYJI:%_ 0U \CL.-' CDo 0 : DcEn(nD Y

-IDC a\ r NUN r \' Ol OLA rPC ON)0as%-P O' ON CNUN) -r- XI n :3ýL~j o -, o),, -P o \. -r ro\.o o q)\ O'-' o co:)) N)--4 U'U1 -r00 oH o n ct

o) -r--4~ - q a-4 u' co-4 u' --4 -C 'NU, CDuU' ,Dc P as a% a--4 \' -P-4 a) u~ m-Jc3 a'i-,,.n 0 \1 C~r)N $~j ;"Lo ~OD C) 0,.o H- X- -N) -p) " Co-H N) N) a~ Cn '1 H-H

o~~~~~~~~~~~~~ W~AJJ 03ouu IN0 m~0. L- oJO.'. n'.'

ýA ODw ý, ODWýn\: : O

o m -4u N)'a a\ -',ý 7% -- Cr -P" k~n .0U %nq ")- ao ý,- c ~ x- HN oo z -3

o N)-- UN) aN )- -0La 00"C.9 N P \'~J.o i-J'. i00 '' 0 t- \ 0Ht E-.:U)~C 0 ZU)

ocn S~N)* . . . . . . .. . . . . . . . . 0) 0 "u0 U 4-UJ -PCY\ XLa' U .L.. X'W-P N) U H m-' a NLJA .Xr'Nji c-* (D

W HX--9,DN) CU 40 F-0 C\-4 F-A-'-4 G\CON)a~oaa CD0 rCOC - 0 =w-0Oct* 0

m DC

N 0 0'e-.N) c 0'U'A ',.-L .L. -LJ. Aa U 'J 1L J '4J

0

F-~

ro 0 0

00 'F 1-3100 000 0 00 0 0 0 0 0N)o 0 0 00 0 0 0 0 00 0 0 0 0 0 nZ

'n a

riN K ) )N) HHHH I H1- -jý-aH HHHH CliVI NFW ) H- OlO 00-3 a\". -P' N) H 0 \0 00-i a'\.Vu -F` N) 1H 0

* ~~ ~ ~ k k* *) *r *o ON * .. ***co .o * 0 (,D4 -

W Au LAbLJ WWL. PIW LAJ La j LmJLA)J -F- -FLAJ W P10LAJtALo WLX) W =IP " MNoN~\AM"\Awom \'nw"00000Nc CDO 01 'O(1 OCD I

I - CJ )

0'0 H - 10 \-F'\O %n s0:) 0 'N)O a\N) 00LH 003w & a0\O:: NO \O l. COO \n F. r) co-. 0) H (7\ W L JO N)N) )0'--W- FH-- -CD t'I I

\0 ~ ~ ~ ~ ~ 'O CO C 00%0~ 0O CO3-0 0- 00--c D- 3 coN ) 10 C- (D 9.H

*~~~~ * 9 6 *ý 69nz 0 6 0 0 r%) AM, =C n

0 1

-3 D - -J OD cO,-4 -4 -4-3- 0 C%-3 CO-Z. 0%-Q.~ 0\ -3C CON-ý N -Q '13 I' t

8: N \0 H 0 00 0 0N CDa 0\ NW " HH\n QC -'K " aFNH - 0 \nH r)-F, vo-,-F' H F-W 0N N)\O H H H 0 F-

0Ou ~0 t4N)

Wr .N Wi&~I F--\4J.n . .`W -FýILJ L~ O n W- a' ct CD :4L~

0 t0 z 0.CD MC zz D

-F-W'ý'LO HD OD CD' %n 00 COa Ft \FJJ0 \0 PWW 0A0~ +~'raH 0''C L0\OJ 8 . 0 0' 0 H _

o0 0 HC)

00

F-4 H I- H H I- 14 1- H H H e- - -l t30

t7'~

ft0HHHH ~ ~ ~ ~ ~ ~ I HHHHH HHHHH H

F*F jHH IjF -jFiF jF - - 01-0

( D

H Nt7'L*J M MjLAJ N) NLJ N)) WLJ N)N)JALAW N)LAJ NWI AJLAJW N)JJJAJ L00 2)-pO %J- H w) 0 CD -~H) -3co H H Co'.O Hj % H 0 %D H -- j W\O -IýQ 0\ -FHLj L c1 C+ D

CD 8 g'0' COW 88 8 8 88 8 880 C8888888 8888 0 0 88 HCD

C/3)OD

\o Qo-,j C\\n .'iJr\3 0 \0 o -, 0\\,n i-i ) 0 "D co-Q a\,-ni -F'-.- ro H- 0 ýo

HH

*ý CD

(ACD

XCO

4\J -- V. .'J~-n0' Tl~T1J' 0%J1.T'~J F J'JVl'i 0"J -0\ H t4 CD MNHF- 0 K 0 ~0 0coJn 0 - P'0 -1 3 CY\- 0 \ HH -"ID-F' a\3 3 oN o 0OCD-

00

00

7Qct '>

" -0 -- 4-4 0000I-000I--0-40 0 -0C)0 0 0 0 0 0 t4 !-4 CID H

H HH H- HHi H- H HHH O-j 0

0 0000 00 00 0 -Q000CDr,0"Z0a0ýw t-I : C/) x D -P' tx

co 0 F,~ CD Q

CDD

ý-j t-j H 1- Cl 2 0- ) tHo H .D N) H H) --a H Ho H1% H) -3OCC ONH'C _

K0.l-~ 80 Al0 CN OD F)H -P' N)J '0 (n -

080881 88088808 800 00 o 88800000 0 0 0080 8000 CDo

0

Ic

CD 10

Uf)

\0 00\10 LC)O % %. '0 NO %D0 CO OD O 0\0 \o'. \0'\0 'O 0 COD a 0ý CD'. 0,.. CO Z 4 tK) --a N) N.0 1O r%)-3 LA.) -.3 OD~- a-4.'0 N) H- N) N) \0 CD'%O H'.. , 0 r -cb

En CD I

co

-F,:K CD (1) xL.j () K FW 0 p N) w w w N) ujN)HO PH.)LA L) PoH Lai LAJN) Pi:3

F HHHI-'FJH H H FIH FHHHHHHH ýj J -N)HHH - ýJ ýHHHHH P D tA

CDH

o.

0

F* t 0 L4 * A * 0 * % * . 0 0 0a %D *A 0 I C\ -0 **% *F 0- 1\ ) 0A Ed ) t-

Ho \,O\A \0 ODF HOOLJ - a)LJC N \ HN)N)N)-D OH)\Wý 00--Q %AC H0 ý-F ~N) 7%'0~. 8

.)' N).0LJ'0ON a"n'.H DO"oON 8 r -ý CD 0 0 U)-N)E~~~~~~~~~~ z))))))))NH)NNNNNNNH N N -Lt.

Lo:tz

0 ~ ~ ~ ~ ~ ~ ~ ~ 3HiC CD j;66 0 * * 0 )0

o 0 OH NH* ~ ~ ~ ~ ~ ( M S -42N

'.31~~~~i (IN.J'1 Z() * CN)H-J- F-4~a 0- F-FJJH3'JHt-

0 N 0 DNIAD -' m0 w *D -jwH F- I- W\ %ýo-rw0 \ I-jt-4m0'

0~~ ~ ~ 1 0 H- i -

;o 6 )# 0 !Z .10 H$-A 0. a\~

'.0~ N)N)'J

P-1) 0ý'O ý- N)~%1-4 0N)CDOO'.N)C4DN) ý- Ho -4 0. Co H H c-tO'O~Q~0 !~.AJN) !48~83 ~ 888888 O 888 888 888

-~~ N)H 0 %'.003 C)N%) ; O -P^ PIJ ) I-j 0 %.0 oo~ C\". 4-Pi. W) H 0 1d(Da

-F~- o I-O D N'.0 a)'.0 00,10 00%.0 a'00'10 \.0 \0 \D co Co ~00% \D 0 00 m) OD 0) a) % 0)' \0 a)'o Z. K 1

U.A J~ " C" \.ALJ O.W,0L'J'.n 00a'jALJ.V.fa) .lL- COo\Ji 0 C) CD oq G\

co~'~Efl~1~J-~-A\ ~at'n'.t'J'~r~n D (D X-

0 CD H R F Frw'.FJ '-OH-OHHH000 0JFIHHH 0 0 00 000 10 c- Z OU0 t3 - -ra\ HL'.W'0 -,3 ") K) a% 'r.".7 N) 0 -AJ - -3*. N) C01.0 0 LiAJ \, (7 ' N) H-- I-m tr, Iý

NN) P) " )HN) N) r) P) N) N) )N"N) fHýN) K) K) HN) NN) "N P) r%)NN) CD ~ )NO'N) 0 OHHrQ 0 H H )oH O\.0'".ý-3HH'.D0 jH H 0 N) H tJ' ~CD ~H

a)(L& .COW \0.-rl 9'H 0O FA H Has0r a'~JN) CLO-4Y\0\Lfi N) I j :L- .N6o ****, * *0 *A 6* *~ 6*: 6 6ý * 6 6 iz u.o, o r Lr,

ODW).J w.J. o)'JLQ M'.a)J J~r'J0a'.COa ~u)

P)C MOL&)J)0W % \. QN N ) A )" N JU H L*JJ\.) HNHJN) ý % OL&J N) 'I Q t "' (Ii Cf\J1ro F- l) .~01- -F Flan ODaLLOW .. f0 "flLJJ 00'-- SA ) O 0 N CD~- F- Fý n( H C-;-r 3oa%. -woK ,p-~w" i-ýý % 0a' N)0 0F.0 nJ a 3 0 N) -F-N)0 H n o. ej P-4 t-o a)'-flO\N) 0%,01.0-3 CNvQ0 c7 a)T 4-F.'.W ~ a' IQcn Po-,~x0C n 0 Ir:

I-h (D "- L) ~ -30 Z

0 En 11 r1003 41) N))ONNN 6 0 )0 -

4rc4, " F- *F'O 0'LiJ0 0\' C\ .4.j N) r%) (7"i FA rw Jt'.oAJA" 0*Ji r CD "- tZLAJ%DN 3'Naa%- 3 O \ 3-- o NO a~)a 00 a0 '-4 N)' CD Cr) ý3A

CDCD H H

F-HH-JI-j -,H)H-Hýj -H HHHHjI-j HHHHHHHHPHHHHJ -JF- FJ PH '1J 'Jx-'.lN) )HH' P' HN L) -P J- H ) Lo OOA J 0N)H ý- -N) "\J H Fr!%3sJ LAJ H t.-4D OZ=

0000000080 00000 0 0 Pý tl?* *0 10 co?, tT1N

0N0 0 -o *a

00'

CD . (

0".'-L. 0 J 00 J Co'-.3 F HLA -jý-J0 F- 0' J 0--j 'A. F)N H )'~V. H i-j F-H H 1 o 0"

888~~~~~C O88888888888888 ct

O;a w 0%..n%-LJ "%A". ( 0- "~ *LJ MAI a %A'. co Q%~ 0 a m J M %A- m

0-4 C)

0

0%U 004 Vrc WW * % AO.% K)a-J % J..-.. K).D . Mi KE4 a- VA 004O~~ 0000 C "Wa Dw0%C ~0 1-w UrMr' tvIu--

C aD

wJ WJ w ww wAD w -- L. ww wF wfi~' P-0 IJ.ALA wJ.LL wJ 1-j NI-jJA PiHw 0

co %a-40% 1DJtA -- o%0l 00 Lai w %JN'NNc .f %AOJNNN go0Jtf (' -'00:3 ~ W~ W.i& ~ JLI. D 00COA I.~'

0*

0 W :

= -4 (fle

C% 8 -r-v'.%0 -PO K) %CCO 0 fCDJ%0 JVJIDiJ IAD C-"..NJ- -lCO0N0I-O t0 P

o 00000C00000000000000c00000 a . 0%4

Oa-4

00 o 0L

t* .r * _

04 0%

%0'0 %

-ja 28 -4tl0wUtOA0%.% .9'0%0 0% 4.a Di'J.3NJ V UND 'Ja Oj%ý20- %J taO I.Or~~~~ a-ORwurnL - C

-Cr

:r LAP ri

N O ý - ' c,-3 (7%%n N,- Fj 0*~ %0N-30%n-l~ I-I

G\

(.L&J IJaiLA.UJ -'-..'WLAJ-F' LAJ .PLJLAJ t~ LAJ -P--FIA ' -

HO 01- J1-W? c- o0 - A.~ okn~j'l C'¼l" 7 7% 0'-~ H'. \ * A) 0- \

U)( ný

~~~~0 1 -4l-CI

0 00 00 H 0H 000000 0

c~~~: co -

oo H H H 0c

00zL

%_n 00O F4 0 r0 L&Jc)L0 0 0000 00 " ooa' (D 'i:'m trm m t~l tel =:D

O-4

OM 0- 0 'x

0z

_r:3 ODI ma %-Fl') 00 8 " %.n -%3 I..J Pi\j 10LAJ COP '-I4D i - \ (0\8 H~J c3 8080822 0888??08080000

00~ 00 000 4

%A 4- Wv W- 00 %D OD- f a% 4-W PI-' ONO ONJ 0%'%A -rW.. K) fr.. . P-4 E/)N3~ JJJ - 3- - ý j j"D

WO ~m m- W W WO D"O O0%J-3C -Ij 'J W m %Q ~ m c c We'90 ~ ~ ~ ~ ~ ~ ~ w W.L.D%. DC N f%%f4-I-('A00Jl p

%00 W94%.n~ (7Cy%% 0--3 04-J N 1 A p

oc

4-W~ rU) 4" _"Wr

0 wý

NiX 41 "J LAi4U rfi4%~j4W-I- IN f% -VK P

4n w NO 51 Sw DOw DN001-r 5-3--oW 0 ( p 2

0~ ~.0~'~%~i%%O0C 'JA.IINi%4ACD..f~j~JO%%~.O' En

LAJ." rO~JO 47L.%J O0r'o'.3L'cEMo88888~o8')

W~~~~~~~~~~ WNW"fW...W"W" NNWzV -

OOOOI-eOO44O.t.e4ii4 Cn.:

*1 '1 ~ P1

P1

r.ZC0 C'.Jl'J~~~~~~ pwlJ *lp~lf

L*IJ~c 0 0 L4

-t xU)E~

ne--Jl~O: ~iv 0%OLA 0%J coI ~ ~ Q% Jj I'W~ %OJ0%4O Q- %J '.3"A NO~PQ

%0o cx,-3 %.A~ 4-FIA)~ iP3 I.-4 0 co-13 c\'oln -ML'N H\ 0 O0 cx0-3 01- -F.'W N' H- 13 1

CD)

1- HgF-Ch rp HO i r-4Or\8 ' on n~~ M r'3rr- 3 ~ ~ Q~ 0CDCDro-P 6'O''O O Z -- , Z I-

L %-0C %,D T%.0 40'1 % ".0%0JL )'J10 JL'0 %-0 r%D 0 0 0N 'D "O'D~i l\0 \CJ -F .rO A \D C') K *

,0 ZlJ

HD(D ! H ýt1N "N Q NN WN WP% r%)N %)N Nr% NrQM ~k) WN : :)rto QE -4H 3

00 000 0 0000000080 a-td:

LAJ 0 Z U= D

H~~~~~~~~~ H HHHI'HHH H HHI' /

0 -4

8zm

C"Jl -PL'. N'. H 0', C -Q aJ '.n PU J -,DH 0O c-3 cNJ ý-n r'i H

H

~~o C'OD40 ID%01.\0% 0 00 80\0 ID\ \0 1100 \1%D*0\0\ I' 0 nffHý,-3D" 0ýJPH .nCON aco-. N -'C\JIOC()tl cl- 2O~n ooN)-~lN OK K ) ) 00ilNcooo.nCOK N\A CD(

(n CD t~

oo F0 , H~LA--3F LoA 'N3a'\Q\.O % \on fLA)LA.) -43 .A N)~c~ Q N) (n ct00 co0 H00Ha0 0ýj %0 ' oH4 F,0 3 .0 F- C -. 0AJw Cr'% 1H -~( m ~j

o 1-4

\A-r Ov%.nr 0\ C- a\-.3 CA' a\ 0\--Z o~ C'- a\ cy\ý a\ C\ CA' a'N C)N" .n'i 0) %VCD

* . * . . . . * . & . * * . . . a 0 *- w Co II& U OH ti

(D 1-4N)

\0A %0L)0 0-*W-J-H %O H H\4 0HH H -HA H HO0C H N)" nCD -) lC

o -a 0 0ui\'n" 0%0%. )0 I%'D 0 'O a'L'AJF 4a'\ 0 co 0 -P\JL0 0 ool n 0 HHHCD H1-'CDH

Ocn HA Wi

\,.n 0 Aa00 " r) rJ0 K )%, P ODLJ CD-%3 \fl\0LL)-30 0 ODLJ ý0, C) 0)

(D (D

~ Htt(D 0,

0 * )0 D0HN % FJ IlH 00N H N) Fj H N)c0 0 0 0,J 0- CD 1

000000000800000000 0000 01" O

0 pi 1-4 N)'

ý- -- ý4ý4H0 - -4 1-41-qý ... 1-4 ý4 ý4 CtzC)

oilHHHHHHHHI'l 0HHHHHHHHH

C21 U)'0\ O -.3 1.- tvw IV C+ (D~'O ~ ~ ~ -- Ow0 \ 00 F-\n pJt0 0-Qf3 a\~(J~~ HH\r'.. P0LA a~r\.o C10

On8 '-: 00 88~8 8888OL 8)088888888 m I0CD 0:888 0 00 00 0

I En

%D \D-r \ J 0 ) \D\ %0\DO %V O\' *10 & No % H \o0\0 \D) \D S'\l-0~A \0 0 H

Ji'Ji'J~0 N)) P D\iC

Id NI

N)C loH HC- )N PF\n8o%04HN)NN Wi )H \0C-WtSE C. 14-30:3 0D\1 HH 0 HH0HYAJ0HoON)oo0. (ID

-J~~~~O~ OýHH 1 " PO LJ f~

"o H

PID ). N) r..J H N)LP.KJ N N)L.JA N)L.KL.rA N) N) N) (a N) LA N)A N) N) N) N) Fsj ro H) N)N l) ( 4H 0 H 00Ha' oH FJiH 0 " N HW1.OO) IjHW FNH~-r%.O " 1 0' ID0 4 0m

H 13 1ý0 coH o0 oCHA A)%n-p FW - 'DwF HAJ 000 ) \.nJl)H w~ C-(L En$1ý

\.w OHM)

N)N) Co 0H"03 i 00HP-.nN ) )r.) 00 N o Hc H0 0 ri\1 u V=%,U

(DI r)

H HH H HH HH HH FH HHHF- HH H -If-JH I- - H H H ý-H HH H 6LJJN) (DHL W ~ )~ N) H -FN) LAN) N) LAJLAJ N)Jti HA HA H H, H;-r K)FHN) LAJ -U NWN H 0 IH) CDN\ J\O -3 )- WFH 0C\ %,0 0 Q CoD ) HJ P'\o "AF"J. H C\T% (A (D M

0~ Z-444

o'to

O-41-I'-- t-

0 F5n\In. r%) N) N) Pi ,.'N))N) N.) NN) N) s) N)NN) IQN) N) . AN)K)N) N)K) N)K)N) AN) N) w J1 U

N).N) -.. )z 0 )0 o 3-pO 0'i-'L.- %ý 0'oH'N) .FP)'.3 0\ N) -1 A ) 0 0 _ F;r .OD ID -ý A NI8888~~ oo 88 O88 888jK w

Clo

NI -IF' K)- HO %Z CO-3 a'o~ -r'l N) HOj I'D 00-3 ON\.n -FL.J N) Hj 'd

l- 0CD

0 \0O\O\DOD .0 , 0 \0.D \ 0\0 1%0\0 \'O\D 0 .0O\0 0\DIDvi O N) N) 0 ") Ovl0 woO COO WnN\.M mno0\.'n 0 0 0 ) CD Uq N

f CD n D t,< t4

S0) -P- N aON Ck8 L 00 H -P-N O \jol-.J 0\O'. F= CO\,O -r\-L.WULJ H to) W ct H- H- t~

o: OCn ODI\0 O\-30a'%LAi0 C'ODH\D-\O-F\n\J~)CAJCOH"J ý- (D !1r..2I-4

In :11 on,2 \)Jr~3)LA (D 1-4

H~ H'jN H H H~~ Co.aj~- N ) H N H H Fj H' PiO-j 0- ý CD) W -

(D H- Hf -

O zC1 01CnC,

O) :K NO)J rCONr )K N )N )K N uNr)r) ro tjro Piw N N) W0 t~ 0 \4n 0 AG', Z &CD HZ~

L^): 1ý -3) ~ N -4 O\0w*"%0c P mol m -: 0 '

(D CD : x

I-,3

1-2 <

La%"Q4%D"3LAjl%3 a' -\ -FC\"Pj\ ON'J TVJ\AA'Jla t4~ CD 4-4\0 -FH-0-i &J.3~ - 0 H'A \A r) HJ a'Lo ý'0N \,nOLCOLA) \.n 0 W o'0 a) 8 z00...8C P,; \- F 8 CODH 8 W8 - O -' -;- C\ 0-F' a) K 0 * -

000000 000 00 00000000

00

* 0'i0

0. * w'J

(D

N) ~ ~ I t-4* -F' 0nL )L ~ P oN - jr)wwu \ ~ jwwL vK ow U

HHHxH \HHHHH-'HHHHHH HHI "0 H-C8 8 0000 808 8 0000 088 0 ctanc-t

'0LO n.

cf)WL.JJ N) N)N)N)N)N) N NNW N)H H H H HF ýH FH H 130- \0 'Oz-~3 CN'ý -FtJ N%) Hý 0 \0 C00-1 O"'A -PUJ W) 0 \0 cxr-i C7', %il N) Hý ý

H

ON)~(1 (Di-~f

(D1 (JDINL- 0iL LiLA FjWW A N )WNoW )00wMW 0 W H HL O N )LJ A LJLA X Wcl t3

(D 0ý W"r)r\)r P F p -- K -- 'H\. p-3N\ 0 \.o -pO Pj -1 0 N) cAH( tn '-

a)D tvC -

yq CD P- Ul)

o- I-C0

HCD H-ZHH0n - 14 -

I 3 ot7'~

I H 0HZ""11 OP-3

\, 0 c t~

CD CD

ný4 j .- F'Oj ~jIN O'W -IN ,0 0" 000 r"F ) M\o 0 FýN \I- t4CD OC-H-p-m ~ ~ ~ ~ ~ ~ HN 0-\)0M-% rwO)PWC---Hv0 w' Q

H O~C 0I0

F- 0 0 0 0 -4' crN) \j3 \jnN

0 0 '-0 0f4 - 0 0(Df

F, H-- HHH H-f F-HHH-HFj FjF- - - - -*~~~l n ctUY C )

a; 0 -lCD \0 H 0 00 0 N ) M'

CD

H- -H H- 1HH HHHH HHH H- H- HHH HHHHj - P0 -p- -' ) 0 W) -P' H c H N) 0 LAJ \4tH N\)t 0L' m %, P-~\0 -pC- o-, 0 coH. "J~C \N) - r\) -P ct CDN) HL.J co'..n CON uj .3 HOj -P0-F~ 01%L~ -F O aL\.J\ m H \Cý O''H C\0N \'~L 0 N\) 0)~- w (n~

88800 08888888888888808888888888888888 FH-

Mý,n ~ ~ \ cD- \)ýj0 , :--4 Owkn -ww' W~ P- 0 \, co-4 mkY1 -;=-w NO 0- El

1 W0~ Cf)-J:~ tO(D

O .JJ-~o1t .OcaOOw1O~\ .

En (D

mm wOD- coco o c OD 0 c coco o O--4 -4 4 -4 - -- co o c coco 0 ( (1

ý-(D o -1-4-w .)u. L)Chrvw \)P o k 0 O N- tn ý -nr00 0 10: zf F-4~

HP0 -n -w 0,%DLA)-4HHHI-0-4 Mo w W -f--I OH ; -;t-1-4 0 %D- 1-41-1 W -lC/ .S:

01-'H- A)1-nC(7C\-WW -gJ1-hI-J1nr\ WJ ý-H - JJ0 -;=IA.;= \) -J 0 o

C- D H g

0:0

ODD 0) ýDU - DQC10 0 -q OO OOOO 0000000-M--4QM0 -4C)N00 Wt( "Hýr

" 0.A)JApj wA o r )N )N lCl)

c),\-PL&)Pjba) co @C-pru -sr -'LO. m~i1U t co R C'3 o 03o\nK CD0-J o~J Ovnj, crotj, 0 (D 09CO (D~

'ko D c D%.o 00-4 CD\O 00 coD CO)D OD OCDO ko '.o0 CD CD. CC) Co CMD Ok 4 D C O-4 cx'.o c t-~ A) et ýAo0 \,n 0 0 .Jflu. O',Jn 0 t-flV1 0 \- 0 0Ot-n %- 0 0%.n 0'J tn 0 k 0 0 kflVn 0 0 _, 0 -4 H H

(f D 0- rZ

-~(D t= -;rLA -trw) -pr i-- t=,-t- -t .r zu -;rkn \..n v -Pý .ý\vi 4-p\n -n- v n -p -st -;r. -wt-'. %-n -or cn Ca ct H C--- )LJLk 0N AL)F )LA 4L.HL 4\OW LJXH. A - 1 A 7 )FiCU WC - -[2HD c !t

z~~~~ 0&)qa W-

0A t - ý )N )\ 0 \HO 00LA) - 0 00)ý t 00 NHý - O - 0

0H 0 Hi-4-4 3 00 HO -4- 0000 I-4HH -IO

09c

IC

co -J '4 Wu M-4 LA)--4 Lo) 0 r_ C

(DHj C I)-j~ *

\-" " N) 0 CDNolam CT) 100H :T &)-1 Ht:0 0 0 HHHaHH Fa 1HH 1- - $aF HjI-.H H- HH H D

LA) .JI N) 0 OLA) LA w.a HC03 Oki ODW. H\tO .tO N o %) N O'. I\\,N'.0-, 1\) F M'-4 m ýH .0 '1U M )C)0\0WCjUC P-f)N Dýj0O ;L. 44 )N DM0%nR CO CCD D

ý-: OOO0000 O000000O00000000000000000 c00 -nq00000000000000000000000000o0o00 col

-p r\): 0009 0OCD 0

LAI % :1JWAJ* D- -0J 1 -a -J %.A1 WL.)'),

*- . *)0-7" r c

0 o ýo 1 D o 'D~ -! 'oO O O1O3O' -: * 0ý0ý 0 D 3 13 o *0 *0 *'.0 WO'0%0 COCM'O -40 (DC0" CDMCD.0 C" OCD& Oa

LJ0 0-01N W I-t

03 6z; CDW ~J1 -4 Wga C" .JL.AJ C- 4 "C Abr-4 W IA 04,

'g~~~~~~IU~~~~~~~2 * ***.****

10

'oc %4i PA,-a') CT ~ 0~0 J 0~' 0 - ,0 oI 0 t9D "% loc o0 co i P1

0O~

I-', OctO

0

0 0

n%'-4 0 W

0 00

02)

0%

4 0

so00 co0 t .0Sot9 1r )c

'A

. . . .. .. . . ..6d I *

glln ~ ~ ~ ~ ~ ~ ~ ~ rr Ar ODE WJ rC "k n%"u in t 1.* . 9 9 9 * 9 9 9 9 . . 9 9 ~ .a

N00 .~~~ U ct 0-4'CO 0 %C OOiOa 0-. 0"0 WMC-4 0 ,009.0 a3 .J

ror~~t Laii lr-ru4

0 n

)ODJ40J,-u0ie-'7O 'OD0) C-4aJ1J- I4 4j .

IQ a

0 Vt

JCMW* 00 W0 W-- mc@' Htj7%D% % 09~%)o0r'l k

02'S

~~~,j~~~~ C02. .aoa*a ~5 CC

ER r'.K t0-

o0 0* to

0 0

's0

n 02

0 0Os4 e0%Q ' tjtaR WQ-b~4-4 0 '010O-ga'-4C It uW ' ()N oh)1\ a h I QWW-4WO O% 4r a*0W

0 os 19 '3 9110S,0W 00oo oooooo oa"0R00 0 00000 00 00808 0000000

ArwjD- to M- ( rA -

* * .* * * . .* . . . . . . . . 0 * 0 . I.... OD )

3* '1'09 1-1009 % 63 0-4

o"D'- 0.D~-l 7 7 %q -4O uJN JOQ -4c w a-ow"%0 %30 a)4~

N I-

10

-A %CO M-4 -4-4 -4 -J4--4 -4 00aD a "-4 0%-4 -4i 0-. ' O

o 0g .- 4

C4 I

0*

U90%O0-4 ir k&9()%A a4 *-4 vf

o o oiC,

4w 0 -4 L" a

Zc 004~Nft ct z

x caIN ~ ~ 83 La w WWr )8 % 1)h oh )1)1)

-e3cD-J ~ ~ ~ ~ ~ ~ ; *wiru '0O~-3 w SL) *r' *-' N)Rc~3 J 4L) N)~

L41 LA.~U) W AJ -tr-C="WLA)-trA.) -=J UJJLJ A LA.) WL 4 A LA.) LA. -t 4r C) CDr%cn 0%c.n ino t-n O 00 ,n0 0 L)U k A 0D- k-n 0 LD LA.) \O'. I OLA)LkLAA 00 :5I ~

'U) OLAJ rOW A) LA)WWL)L L)WU 4 LAiL) Lf )jL~j N) r~~ OLA) N) W ~ W W-. N)-

CD Z. r.-

rI~-J3. tOLUJ Fj cD O O) awoLA -irQ~ 0 A)0 rc'O M -~A 0 00- CO4 ca ca &1-3 m

0 0

t4 cc) o c

t!1jHOW-I-41-40 ~ 1.48 -414II

0 0~~~1 :r - -

0\0 ct CD '-

000000000000O00o oo 00t10 m 000000000000000000000000 '-:4-

CJ0 l

'0

Z) 0 bo bi-,ý>Dbb; bo bý> bb O;D D 2- *

* ~ P - 1" "A . . ** * *SSSSa . . S** 5 A 4

-ww aw 44 *D* 0- "a P% tW-4%0 %p va, a" -J'%f40 r" w4 0-40- O'' .v

toA4 '~ J totot F o o IV o, t7,:*~~~~o

0 1 ,5 55Sa..3

0

~~ 0

"00C so"" NOW' 0 0 "W"NO~I 0WII J "" N t' -v

Fri 11"I"I H ii

N ~ N .~0 -ýD M-4 ON" .- 0 %4) caNJ 0'*AA r.t. r-,)

L.JLJkh XZ' Zt -WWLAJ LAJ W Li- £LAJL&* £ aD.IL AJ W L& Lmt& LI K siO)No 6lbO"60 CDC ~~ 6 CO 0 D-- IP CDO ODC-i'o0

0 8,0.0 0 00 0 08(30 880 %

rO

IIS4-3

CZ)1

0':3 WC N N%--D#- lb-CC D OO~ r -

0 0

6* * . e 6 6 e * . e * * & * * e p * .) 6,-C't,

C) -

to CWNW WW W WWWWWW

4r0- M.-N-- iW D %AN J% -, - m oo-jo P

0~~6 0 00 t -

I wINNJNLNPDLJOwfIADNQrLNNNNJL.J..DN NI* .- 4 g a

wo ~ ~ ~ ~ ~ ~ ~ -0%- \O9Hw Mf rWlA

109

t-04444 440044h(

N~rU JN NJJNNNNNNJN~r~.r%3rr N ju'fJi NJ N flCL.DL. .~A L . Lýa N ) L W W L .W L -'

(I)

rjW ~jWrj r)Ljj JjW~jWW W*JW-PW CU)- (D V

L.JW~~~~~~~~z NtaL.i N.Ib .0.JA .0 C6LJ. I'11j~ jLJ~D

%0 w ow 0 a*)

O0 % l%AW c - cmo00% F- J %OrNW. co. WJfr4

w

"%5%oow~o~" cn om

I ~42X

L'* LsI 'J-AJ t.4 L'F.'W' Coc tp., C..-I,

0 000l~ 0'

1 n140, -4 - OD4-i 44-4 40QV0"% . 4 OI.4OO P-4 CY%41~4 ýj(ý154E

.30ft0

LZ

(flSC~L D0- Vi C)I

W W C', w tu O:rDW & WW-r- N V-w -%OAw w - I. . Ix.wwrr) 3A

8~888888 8-

C/)

co--3 C'v, -PA~ -.HO 3 O N- \\Jn -Pt.v ro 0 \D 00 -3 C.n -rLF i 0 E3(

CD

0\ý COO DO OnknAJCrC> c300 0\,ALA' 0 00%-n 0\-nr 0 OW\j~.j" CO (D M~cn (D X-

(I L3

HLAý- C

N)PiN)L~ ) AJN)r N NN)LON wN ) )w AJN)N)LJ AJN)L) LJ A(-~I~D H- Z

C)

I HI(D C-) I H3CnH

N) ir'jr.)N NwN>r' o p, ,cD L)C1ý\j C I ýln _-F0 C\ 1 vJnn I C\0\0' -F0 C\fJ JlJ N C\ \ cn En: - nt

w CNý-9' CYCOCO -F-3 F-3 t0 D ý.4rf)t

to 0=t

0I 010I) 10

z -4 0'

a0IV N4.)

et'ct OQC+z Z

(D H-

U)

H000 Ni LiC NOH WN C\iH0PiHL N) (ID P 1~0' H ý 0000000O000 000000000~00 '-w0 -

OCDO 0 0 0 00000000000000000ooooo0co00000

~A tU %Q OD--- 01% 4W~ N~ W OV (JDNI 0%%A 4W~ W w

4.4I .- q

* r 9 or Or 9 or or or Or Or Or Or or or Or or Or or or 0 r as

ILm

4. 0-4

I:r 0 ty

t-~V V '90 A o ,1

o0 W 0

:a LA 2t..4j

0 t

0 9100 ** "99 0 * O S 9 0 9 9 O.v*~ 1,--..C-1 ;a% C)~ ~jt ~~Ai '

'SxciV

tv WN A)IV I IV u IVIV I WJ0b~I. 0d ~ 0 000

ot

9.~ 4I 49 - 090.K

0'

'I' :Iu0 C,.

~Mf~~e I 9iI~'i~~i et

"-4 ch"J ~ ' CMN) Oriji 4-w". 11 -1 ONO (ODN) 01%.nr~irj

Co--j 4 0 co OD cDNj --)% -j --j C% so co 0000OD DCD'. cz- o cC D'.q (,D i *cAA COCE&7 3W" OD Q Z&DG D 05?w '' 0

.a "I V1

iL..iu~~~~~~~~i~~~it D a~.I. v~LtDr~ 3r t.LJNLJrLD ~ &

*Q (D *, ai La L- Il *1% *o 1- w *Q Il I A *4 w 1- W N W w - D W

0:3 a~o -OOQQ.O~O~ 8888~2~8~ '-

rv w v N w N N N N w 0- N w ri N N ri .) p

v. 4 4("I.

0 -

a

LSO NJ 4A-'N)&DAD (U t7- r o V'W&~-N)JW W \ r 4'\AfjL. 4w 6ý (D" !

WQP\D D0\AJf "W\c%0o W -4 .' -91 OD0iv0cD3~~~~c n ~.~w0Z~to '.~---'-'-' ccoaA"J~J~L~o~*isa

W -4

.- 0

rru

0 -

014C

~ ~ all.

'09 a a I o:r 008'88800 80 L s8 8

04

----------F -----%A w

I ~ Qj~ ~.J N

'AC~fa%0AJOOO .AN~JO"Ou '0CCD- N%As 9

%0 %Q O'C'O %0 %00 .O 8, ' % % Q D 0O0%D. No ga

%rI-~NJ%%A' ~ O~.O%-' 'Arrr'c %Ja% . :01~ ZLn':3 P. w 0% .- ý- %A P* *OD 0 N"P

N~~ w o -wt%A~~~~~~ ~ ~ ~ ~ w i ~ LW0I4%%LjW 0%~ 2g ;0V

0 - -%3 W0 N W jNJNJDt WJ NJ O-NJNDJI'NN3 a~ UlI-'La NJ .OI-L.J

0 En t" r0t,

%A4cl%%jLO ~NNJ N_ 3o% N 70ý "00: r-WW0- 01 4% K 4-W *4 -,4 CO0Ws 3N W-

8 '4L. CID %0 %-2%0N 1-' N WD% J~ 1 0 0-1 431A &:SO

r%A N%-A C%%)Oý 110w % % O 'O 4 ("Q V*160A 000 0 0 U 0%0 0

"-I'0 -0 '0~ Z "

3-0i * r,

N 0J

%Q7s xI

"S ~ ~ ~ ~ ~~~ht . -148O"R...01- -

N ~ ~~~~~ ~~ n r 0WD CjWrl .

%A o- LJ0J

~' ~~~ 88 888888

Ul) U)\I H~L 0 \D O' o Q-13 O\R1 -Pi'LA N) H O'\0 OD-a3 C\\jn 4'LAJ N)H 0 13

I-0(D

C/)

1-4

En

'-4N) r% \. U* Q N) N* *A *A r\* N) N* N) LO W N) LO N) N) * N* . N) * N) CD -- P

(D 0 Z U) C

tr-

tz HNNNC&- F-NNL A f AJ LAJ N)NNN I-'H-jtýCD O-rt'>

L.n CFO ONn `-'\ -r' i-F'L 'Q I-J2 \, ro 0A -P~ o"F-x 0*O C\ \ \Zj(%0C rwP :4CD to~*

10 11D 0 Hr_

P-40 0 0 0 0 -4 ' 0 0Lm~ .& .- t- . O'L.J% 0 OmCDt tC)4-. 2 01 Cl

H H H HH HFH HN-Hý H ý4 CD

N) LA fl -jHw r O) N) CO 0 F- o

f:3N 0 0 aqý Qa 0 0001-' O'00008t'00e-3OU 0 0:, 0* 00 00000 000 000000 "

(D m t:Jr~3HLJ I- N N ~ DNn

Cl)

c'-n -F-L. ro H' 0 \o co:)-. O\%n F-P- r,, 9-' 0 %\o 0)-s3 cN".r -P-L.j N~) H 0

(D

3~%3~ I'(A CD

0 CD

Cl)

Hu<-F-" N3VJ o~j -p N) CC-) r pJ't JC C )NC a\ coo " - P'" a'\.n-% L) QoJ 'J jt; CD

rO0%4 3rvD % \ -r I-'F-OO V'o 0~o 'J0N)0 -- ON H o"JO N ,nFJ0 -. ' ON CZ) I'l t:

79 CD zwt

H0 q

ý-- CD Z /) LA4

* I -3~ tz Cfjt

tD 0 -t- F

h)~l 0 Pa'~ ~ jo~-- C7\ N-J-.~ o'--3 w"~ %,jtO"C a'o0 w 0 -o w o'. o' W- m

00 Hn 0 CfaWjOC- -ý t

co 0ivi r' 0 coNNNNNLJ))~f3)))))NNNN ft) C) ) x1~

(CD -

OD~ PF-I %.n oo

*C IV~J

a\'ic 0) co UCNa W lw~~~~~~~~~~~~~~~~~~~~~ -4"399-4\n ))40o 3\0I 0C% \0 JC fF 7\00 O \A - , , nC

9-42-o o i- 8 0 0009-49-41-41-4-49-49-49-4o '-w "--4I4- -00- ~ -4 C(mO C)lo0lci 0 OCl Clil0tif Cli 00000