Fracture tran rna lity we colu nche is g and S slab is a prerequisite for dry-snow slab avalanche release. If a macroscopic Got et al., 2010). Girard et al. (2012) set up an acoustic sensor network ase transition exists. e—microscopic cracks crack that will propahighly possible.
Cold Regions Science and Technology 110 (2015) 160–169
Contents lists available at ScienceDirect
Cold Regions Scienc evscopic instability where larger cracks form (Schweizer et al., 2003).
Since cracking within a material is accompanied by the release of elastic energy generating elastic waves (acoustic signals), recording
The application of the AE technique to snow within the context of avalanche research started in the 1970s, when several field studies for detecting acoustic activity (at frequencies ranging from 3 to 100 Hz)scopic cracks within the weak layer (Schweizer et al., 2003). Cracking on a microscopic scale is expected to always happen during snow deformation, but only if not compensated by re-bonding or sintering (e.g.
Reiweger et al., 2009b; Schweizer, 1999), it is expected to lead to amacroor disordered materials such as snow and a ph
The material can be considered in a stable stat form but do not coalesce to form a macroscopic gate—and an unstable state—catastrophic failurecrack reaches a size of 10 cmormore itmayquickly propagate in theweak layer across an entire snowslope. This process is called ‘crack propagation’, and, if the slope is steep enough, leads to the release of a slab avalanche.
The formation of the first macroscopic ‘initial’ crack is called ‘failure initiation’ and assumed to be due to the accumulation of damage, i.e. microin order to predict failure within rocks and permafrost. For analyzing emissions of a glacier to anticipate the break-off of a frontal lamella,
Faillettaz et al. (2011) applied a method from the framework of critical phenomena. According to the theory of critical phenomena (Johansen and Sornette, 2000) an analogy between the failure of inhomogeneousacoustic emissions (AE) can be used for det ⁎ Corresponding author.
E-mail address: firstname.lastname@example.org (I. Reiweger). http://dx.doi.org/10.1016/j.coldregions.2014.12.002 0165-232X/© 2014 Elsevier B.V. All rights reserved.ried surface hoar, faceted 2001), beneath a cohesive been used to investigate earthquake occurrence (Niccolini et al., 2011) and to predict the collapse of a limestone cliff (Amitrano et al., 2005;et al., 2003). A weak layer, often consisting of bu or depthhoar crystals (Schweizer and Jamieson,1. Introduction
The release of a dry-snow slab avala tion within the snowpack (e.g. McClunium plates which were then frozen to the snow. Localizing AE events during fracture of layered snow samples showed that the AE originated within the weakest layer, i.e. the relevant layer for snow failure. For finding an indication of imminent failure, we analysed the exponent β of the cumulative size-frequency distribution (‘survival curve’) of event energy. At the occurrence of instabilities, the β-curve deviated from steady behaviour and exhibited distinct ‘drops’, indicating that the power law behaviour of the distribution was not fulfilled anymore. Studying the temporal evolution of the exponent β might therefore provide useful information about snowpack stability also in the field—provided that the AE signals are not too strongly attenuated and can be detected in time before catastrophic failure occurs. © 2014 Elsevier B.V. All rights reserved. preceded by crack formachaerer, 2006; Schweizer growth (e.g. Lockner, 1993). In other quasi-brittle materials such as concrete (Ohtsu, 1996) or wood (Bucur, 2006) the AE technique is commonly used to study and predict failure processes (Johansen and
Sornette, 2000). In the field of natural hazards, acoustic signals haveKeywords:
Snow at 31 kHz. Best coupling to snowwas achieved by attaching the AE sensorswith silicone adhesive to thin alumin-Available online 10 December 2014 cold laboratory. Using snowMeasuring and localizing acoustic emission to fracture
Ingrid Reiweger a,⁎, Klemens Mayer a,b, Kevin Steiner a a WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland b University of Technology, Graz, Austria c Institute of Mechanical Systems, ETH Zürich, Switzerland a b s t r a c ta r t i c l e i n f o
Received 14 April 2014
Received in revised form 28 November 2014
Accepted 2 December 2014
Acoustic emissions (AE) are caused by changes in the inte to be a valuable tool for stabi wave propagation in snow, j ourna l homepage: www.e lsecting cracks and crackvents in snow prior
Jürg Dual c, Jürg Schweizer a sient elastic waves produced by a sudden redistribution of stress in a material l structure. In other natural, heterogeneous materials monitoring AE has proven estimation and failure prediction. After studying the characteristics of ultrasonic measured the acoustic emission signals during snow loading experiments in a mns we found that most energy of an artificial acoustic signal was transmitted e and Technology i e r .com/ locate /co ld reg ionswithin the natural snow cover were performed (Gubler, 1979;
Sommerfeld, 1977; St. Lawrence and Bradley, 1977). The authors concluded that avalanche formation was related to an increased acoustic activity. Bradley and St. Lawrence (1975) showed the Kaiser effect in snow based on laboratory measurements.
Sommerfeld (1982) and Sommerfeld and Gubler (1983), postulated that all avalanches should be preceded by acoustic activity. St. Lawrence (1980) developed an analytical model of AE response of snow. He expressed the acoustic activity as a function of strain and stress, and succeeded to reproduce the cumulative acoustic emission curve during a slow tensile experiment at a strain rate of about 10−6 s−1 with a snow sample in the laboratory. Buser (1986), Ishida (1965), and Oura (1952) performed laboratory experiments where they measured acoustic impedance, velocity, and attenuation of snow. However, within their measurement setups they could only measure the acoustic waves travelling within the pore space and not in the ice skeleton. The frequency range studied was about 100–4000 Hz.