The indentation fracture toughness (KC) and its parameters: the case of silica-rich glasses

https://doi.org/10.1016/j.jnoncrysol.2004.07.021Get rights and content

Abstract

The fracture toughness of several glasses, including soda-lime-silica and aluminosilicate glasses, was measured by both the indentation fracture (IF) and the single-edge notched beam (SENB) techniques. The flow densification mechanism occurring during indentation for silica-rich compositions leads to discrepancies between the two methods. The influence of the indentation load and the post-indentation fatigue duration on the indentation on the crack length and on the fracture toughness measurement were investigated. Fatigue curves (VK) were obtained directly from indentation experiments and fatigue parameters were derived. The sub-critical crack growth resistance was found to be improved by increasing the silica content. In comparison with the studied oxide glasses, a Y-SiAlON oxynitride glass exhibits much greater fatigue resistance. The fatigue phenomenon has a major effect on the estimation of KC from indentation-cracking measurements.

Introduction

Structure–property relationships are of great interest in various fields of glass research. From the point of view of the contact-induced deformation, glasses may be divided into two classes: normal glasses (e.g. soda-lime-silicate glasses) in which a volume-conservative deformation process takes place and anomalous glasses (e.g. silica glass) in which flow densification occurs [1]. The indentation cracking features of these glasses are also different [2]. However, the effects of the vitreous network structure on the crack growth behavior is not yet fully established.

The experimental determination of the fracture toughness may give rise to different values depending on the way the measurement is achieved. The single-edge notched beam (SENB) test is a self-consistent method suitable to obtain a critical stress intensity factor, K1C, close to the intrinsic value [3]. The Vickers indentation fracture (IF) technique allows for an estimation of the ‘so-called’ indentation fracture toughness, KC. This paper discusses the differences observed between the IF and SENB fracture toughness values. The discrepancies observed between the KC values depending on which expression is used for the calculation were already well documented [4]. Different parameters have a direct and significant influence on the KC values, among which: (i) the indentation load, (ii) the indentation loading time and (iii) the time which elapses between the indentation and the crack-length measurements [5]. it is thus necessary to study the influence of such parameters in order to compare properly the indentation fracture toughness values of different.

The sub-critical growth of indentation cracks was already reported [6], [7] and investigated in the case of a standard soda-lime-silica glass [8], fused silica [9] and borosilicate glasses [10]. In this work, we attempted to characterize and compare the behavior of different glasses from the point of view of the median-radial crack propagation. Fatigue resistance results obtained for the silica-rich glasses were compared with those of different glasses such as soda-lime-silica, borosilicate and aluminosilicate glasses. The effect of both the indentation environment and the fatigue (post-indentation) duration is discussed.

Section snippets

Materials

The glass compositions are given in Table 1. WG1 is a standard window glass, WG2 is a silica-rich glass and WG3 is a borosilicate glass containing heavy-metal oxides. WG4 and WG5 belong to the soda-lime-silica system: WG4 contains 71 mol% of SiO2 while WG5 glass silica content is high as 80 mol%. Three aluminosilicate glasses were also considered in this study: CaSiAlO, YSiAlO and YSiAlON (nitrogen content ∼7 at.%). The details of the preparation and properties of these glasses were given in a

Influence of the applied load

Specimens of WG1, WG2 and WG3 glasses were indented for 15 s with loads ranging from 0.98 to 49.05 N. Fig. 1 shows how the radial crack lengths c change with the applied load P in a logarithmic plot. As can be seen, data could be nicely fitted well with linear intercepts (correlation factors were all higher than 0.998). From Eq. (1), it follows that lnc = k1 ln P + ln k2, where k1 and k2 are constants: k1 should be close to 2/3 and k2=(χKC)k1. However, in this study, k1 = 0.688 for the standard window

Conclusions

The indentation fracture toughness parameters were studied for three silicate glasses. The radial crack length increased as a function of the indentation load. For a 49 N load, circular cracks were observed, which is evidence for a Hertzian-type contact. At constant indentation loads, the post-indentation radial crack length increased as a function of the loading time. Fracture toughness values are different depending on considered experimental technique, i.e. IF or SENB. For the silica-rich

References (26)

  • A. Arora et al.

    J. Non-Cryst. Solids

    (1979)
  • V. Sglavo et al.

    J. Europ. Ceram. Soc.

    (2001)
  • S. Hampshire

    J. Non-Cryst. Solids

    (2003)
  • K.W. Peter

    J. Non-Cryst. Solids

    (1970)
  • H. Lemercier et al.

    J. Non-Cryst. Sol.

    (1996)
  • S.R. Choi et al.

    Mater. Sci. Eng. A

    (1992)
  • R.F. Cook et al.

    J. Am. Ceram. Soc.

    (1990)
  • J. Mencik
    (1992)
  • C.B. Ponton et al.

    Mater. Sci. Technol.

    (1989)
  • K.K. Ray et al.

    Brit. Ceram. Trans.

    (1999)
  • V.M. Sglavo et al.

    J. Am. Ceram. Soc.

    (1995)
  • J. Salomonson et al.

    J. Am. Ceram. Soc.

    (1995)
  • P.K. Gupta et al.

    J. Am. Ceram. Soc.

    (1981)
  • Cited by (0)

    View full text