Scribd Logo
Upload

Welcome to Scribd!

  • Normalized Shear Modulus Reduction and Damping Ratio

    Uploaded by

    Horacio Ulloa
    17 views15 pages
    Modulo de corte

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    Modulo de corte

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    17 views15 pages

    Normalized Shear Modulus Reduction and Damping Ratio

    Uploaded by

    Horacio Ulloa
    Modulo de corte

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    of 15
    Soils and Foundations 2013:53(6)879-893 @ ‘The Japanese Geotechnical Society ® Soils and Foundations cones ‘warw.sciencedirect.com Journal homepage: www.elseviei.com/locate/sandt Normalized shear modulus reduction and damping ratio curves of quartz sand and rhyolitic crushed rock Kostas Senetakis"*, Anastasios Anastasiadis”, Kyriazis Pitilakis” "School of Civil Engineering and Technology, Sirindhorn International Insitute of Technology (SID) Thanmasat University, Rangst Campus, Pathuon Thani 12120, Dian Department of Civil Engineering, Avsuote University of Thessaloniki, Greece Received 27 June 2012; received in revised form 31 July 2015; accep 28 August 2013 ‘Available online 22 November 2013 Abstract “This paper presents a laboratory investigation of the strain-dependent dynamic properties of voleanic granular soils composed of a rhyolite crushed rock along with addtional experiments on quartz sand through a high-amplitude resonant column testing program. The sands were ested ina dry state in torsional mode of vibration and thus the degradation ofthe normalized shear modulus and the increase of damping rato in shear ‘sa function of the shear strain amplitude (7) were examined. It was revealed thal, for a given mean effective confining pressure (c,) and coefficient of uniformity (C,), the volcanic sands showed higher Tinearity in comparison to the quartz sands and that tis tend became more pronounced with decreasing 7), and increasing C,. In contrast to the general trend observed in the quartz soils, the confining pressure and the srain-size characteristics handly affected the rate of normalized modulus degradation and damping increase in the voleanic sands. These ifferences are possibly related to the micro-mechanisms that dominate at particle contacts i the range of small to medium shear strain ampliudes. For example, the possible more pronounced crushing of the asperites during the elevation of the confining pressure and during the dynamic loading along wih the lower intr-paticle friction angle and siffness of the voleanic sands of crashable particles in comparison to the {quartz sands of stonger particles might play an important role in the energy dissipation during the dynamic excitation and thus on the rate of damping inerease oF modulus degradation, © 2013 The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved. 1. Introduction practice, the straindependent dynamic properties of soils, commonly expressed in terms of the G/Gorlog y and D-log 7 ‘The small-strain shear modulus (Go) along with the normal- curves used direcly in one- or two-dimensional computer ized modulus degradation (G/Go) and damping ratio increase codes for seismic response analysis studies, as for example the (D) from small to large shear strains (7) ate key parameters for SHAKE (Schnabel et al., 1972) and the QUADAM (Idriss seismic response analysis studies. In geotechnical engineering et al., 1973, Hudson et al., 1994), are derived from available literature data on similar materials. For example, itis common practice to adopt in computer programs the proposed upper and “omesponng author. Te: +66 2 96 XH, +56 2986 9101 lower bounds of the GiGg-log y and D-log y curves by Seed “165.2 Sot SBGC912 Maile 6 9191 624 and Idriss (1970), Seed et al. (1986) or Rollins et al. (1998) Evmeil oddest koenctak @eit act (K. Sentai) ‘with respect to granular soils, or the comesponding curves by a Sun et al. (1988), Vucetic and Dobry (1991) or Stokoe et al. (2004) with respect to soils of variable plasticity. index. ‘Alternatively, analytical models or equations of empirical form peel Frodeetion and hosting by seevs are often utilized in computer programs, where the normalized (038 0806 © 2013 The Japunose Geotechnical Society. Pocton and hosting by Elvi B.V. AI ighs reserved, aap dong. 10164 san. 2013.10.07 80 1K. Senetals eta. / Sols and Foundations 53 (2013) 879-803 shear modulus and the damping ratio are commonly expressed a8 a function of the shear stain amplitude, , the mean effective confining pressure, o{,, a8 well as soil parameters such as the plasticity index, Pl, and the coefficient of uniformity, Cy. ‘Two of the most commonly used analytical models in soil and cs depend ‘on the curvature coeflicient, given in Eg. (3) for granular soils, while parameter b of Eq. (6) is used in order to account the 001 ¥ (6) Jog y curves of volcan sands ofthis sudy: Representative rst 01 effect of the number of loading cycles on damping ratio. Delails of these relationships have been presented elsewhere (Darendeli, 2001; Meng, 2003). In Eqs. (4)-(6), the damping ratio is given in percentile scale. 100 [ig Yeas INU + Yret)/ Fe | Detaing = —— |4 x @) tae [ let 760) ° Dass = €1Dstasing + 2D iting +P asing ® Gy" D-Do= (Z) x Deceret co) ‘Thereupon, cach theoretical curve of Figs. 2 and 3 concems the specific C,, value of each specimen (given in Table 1) as well as the specific 0%, magnitude where each specimen was tested (illustrated in Figs. 2 and 3). For the estimated damping ratio values in these figures, a spectrum of number of loading ceycles that represent the RC experiments were used. The typical range of the number of loading cycles during the RC tests for quartz and voleanie sands are presented briefly in Section 4, As shown in Fig. 2, the analytical relationships proposed by ‘Meng (2003) satisiaetorily predict the response of specimens N4-3 and Q2-1. However, as shown in Fig. 3, the voleanie ss 1K. Senetalis eta. / Sols and Foundations 53 (2013) 879-803 sands exhibit systematically more linear shape of the G/Go- log y and D-log 7 curves in comparison to the theoretical ones. This non-satisfactory convergence between the G/Gerlog y and D-log 7 curves of specimens V4-1 and V5-1 and the theoretical ‘curves proposed by Meng (2003) was observed in all the voleanie sands of this study. In Fig. 4a and b we plot the measured G/Go and D values against the shear strain amplitude of all the quartz sands and in Fig. 4e and d we do the same for the volcanic soils. In general, 1no significant effect of particles shape, in terms of roundness, ‘on the GiGo-log y and D-log y curves of the quartz sands was ‘observed, and thus, the experimental results of the sub-rounded to rounded and sub-angular to angular quartz sands are plotted in the same figure. In addition, no monotonic trend of the effect of void ratio (or relative density) on the GiGorlog y and D-log y curves was observed in either the quartz and the voleanic sands, which is in agreement with the available data presented in the literature (e.g., Kokusho, 1980; Meng, 2003). In Fig. 4 we also plot a spectrum of G/Go-log y and D-log y curves derived from the analytical relationships. proposed by Meng (2003) for granular soils as presented in Eqs. (1)-(6. According to these equations, the GiGorlog y and D-log 7 curves of granular soils exhibit more linear shape as 0}, increases and C, decreases. Thereafter, we decided to use an upper bound for the theoretical G/Go-log y curves plotted in Fig. 4 (as well as a respective lower bound for the D-log 7 ccuves) that corresponds tothe lower C, values of the materials under study, equal to 1.3 for the quartz and 1.5 for the volcanic sands, and to the higher o, where HARCT were performed, ‘equal {0 200 kPa. The lower bound for the theoretical GiG- Jog y curves of Fig. 4 (as well asthe respective higher bound for the D-log y curves) corresponds to the higher C, values of the materials under study, equal to 3.2 for the quartz and 4.2 for the volcanic sands, and the lower ¢, where HARCT were performed, equal to 25 kPa With respect to the quartz sands, it is observed in Fig. 4a and b that the measured G/Go and D values are plotted, in ‘general, within the lower-upper bounds of the analytically derived curves. The variability ofthe GiGy and D values of the ‘quartz sands at a given shear strain amplitude is atributed, Primarily, to the variability of the mean effective confining pressure (7). as also depicted in Fig. 2 with an additional effect of the coefficient of uniformity (C,) and the mean grain size (dsa). With respect tothe volcanic sands, itis interesting to note in Fig. 4 and d that all the measured G/Go values are located at the upper bound, whereas the D values are located at a b B uate sands ae al 06 |. Quate sands : o ® cok ° 02 ° ° 00001 0001001 1 a0 oot okt dl v4) v(%) c d 1 6 Voleanic sands 08 ae 7 Voleanie sands, ost 5 oe = © oat 0 oe 5 : L L L ° 0001001 oot ot 1 econ oot okt 1 ¥(%) v(%) Fig. 4. GiGi and D values agains the shear stain amplitude: a) (b) quar sands and (and () volcanic sands along with uppr-lower bounds propose by Meng (2003), K, Senetais eta. / Soils and Foundations 53 (2013) 879-893 85 the lower bound of the literature curves. As is also illustrated in Fig. 3, this is partially attributed t the higher linearity that the volcanic sands exhibit in the range of medium shear strain levels in comparison to the quartz sands. Although the volcanic sands of this study had variable grain size distribution and the specimens were tested at elevated confining pressures such as the quartz soils, the variability of the GlGo and D values of Fig. 4e and d at a given shear strain amplitude is less pronounced in comparison to Fig. 4a and b. It seems therefore ‘that parameters such as the mean effective confining pressure @j, along with the grain-size characteristics, that have been used in previous studies efficiently in order to quantify the rate ‘of modulus decrease and damping increase in quartz soils, ‘might not be such strong parameters in crushable soils, as shown by the voleanie sands used in this study. This assertion is supported by Anastasiadis et al. (2010), Senetakis (2011) and Senetakis et al. (2013a) who studied the dynamic proper- ties of the coarse fractions of a pumice soil (artificially crushed rock), which has different physical properties and compaction characteristics in comparison to the volcanic sands. of this paper. 3.2, Analysis of experimental HARCT results 5.21, GGorlog 7 curves ‘According 10 the RC est results and after analyzing the normalized modulus degradation curves through the modified hyperbotie model, no strong effect of 0, particle distribution, shape and mineralogy was observed on the curvature coe cient (@). The “experimentally” derived curvature coefficient Values of all specimens are summarized in able 2. Parameter (a) ranged from 0.85 to 1.14 with an average value equal 10 1.00 and a value of the standard deviation equal 1 0.06. ‘Thereupon, the modified hyperbolic model of Eq. (1) transformed to Fa. (7) G 1 Go” THT a In Fig. 5 we plot the “experimentally” derived A, values separately for the quartz and the voleanic sands, against the coefficient of uniformity (C,) and the mean grain size of particles (dso). These values are also summarized in Table 2. With reference to the quartz sands (Fig, Sa and b) itis shown that there is a general trend of decreasing 4, with increasing C,, and dso, but the data are scattered as also implied by the relatively low values of the coefficients of correlation R°. This slight trend of decreasing A, implies that 7a decreases. and thus, the sands exhibit higher non-linearity in the range of ‘medium to high strain levels with increasing C, and dso However, itis noted in Fig, Sa and b that a stightly beter correlation may be found between A, and C, than A, and dso with respect to the R? values. Therefore, the reference strain of the quartz sands was expressed analytically as a function of C, and not ds, In Fig. Se and d, it can be seen thatthe parameters C, and dso do not affect ina monotonic way the reference strain of the voleanic sands, and thus, for these soils. we Table 2 Reference stain parameters (A, sandy specimens ofthis study 1) and curvature coefiient values (a) of Code name A, my) Values at variable, (kPa) of specimen 2501 2 Oo # © © OM & 0661 O08 O48 — 095 095095 0964 ons 046 = = 10) LOL sks 0070 042 097 097 95 O6R2 0086 0.46 Los 099 = O60 00s 021 = = 0m 0.8 093% 0089 O31 10s 1.00 OTL 0086 026 092 ast 388 008s O18 — ORS 093. 1.00 0683 0062 067 — 092 099 Lon 09s) om7s 056 — 099 un 0515 002 046 095° — 102 102 0395 008 058 100 — 1102 0920 0032 008 — 009106, 0553 0035 049 Oh 097 Ls 0710 004 948 O88 099 O98 0823 0086 004 102 1041.0 961 0.105 OL LOL = v22 1250 0132 020 096 Le v32 898 0485 005 Lor ta vel m3 0097 oot = 097 in VS 0976 01S 006 = 087098 °Not included inthe regression analysis decided to use an average value of the reference strain, also depicted in the aforementioned figures The n, values of all specimens are summarized in Table 2. ‘The quartz sands had n, values ranging from 0.18 to 0.67 with fan average value equal 10 0.42, while the volcanic soils showed remarkably lower n, values ranging from 0.04 10 0.20, with an average value equal t0 0.08. Consequently, the overall effect of a, on the normalized modulus degradation ‘and damping ratio’ increase in the range of medium shear strains is very small for the voleanie sands, According to the experimental results. presented in Fig. 5 and Table 2, the reference strain (7a) of the materials under study may be expressed as a function of mean effective confining pressure through Egs. (Sa) and (8b), separately for quartz and volcanic sands. We note that fxr is given in percentile scale (%) and € corresponds 10 the Napierian logarithm, Quartz sands : ae = 0.159 xe MEK ( a) (8a) Py «2. 0. 100 5) 3.2.2. D-log y curves In Fig. 6a we plot the damping ratio values, expressed as D—Dp, against the corresponding GiGo values of all spe mens, where Do is the small-stain damping. In particular, there no clear tend observed of the effect of panicles Voleani sands © Yer (8b) 86 1K. Senetals eta. / Sols and Foundations 53 (2013) 879-803 7 b ox 018 Asaismeenns: Jen ie r 4 ak Aamwsome | = ReO47 O12 ae e + > + © oe a: Ale t . a - # < a + oo | +¢ 4 + + ¢ 4 ° 4 1 ° 1 1 1 1 1 2 2 4 o os 1 18 2 28 a (mm) © d C4 026 4 owt al Mean vue = 0.100% S 4 Eon + 8 . < te p+ os | al : L L L (ea es iaeecas eee ees esiztenas c dag(mm) Fig. 5, Etfostofeoxiicen of uniformity

    You might also like