Response of buried pipelines to repeated shaking in liquefiable soils through model tests

https://doi.org/10.1016/j.soildyn.2021.106629Get rights and content

Highlights

  • We presented some shaking table tests on buried pipelines in loose saturated sand.

  • We applied a sequence of sinusoidal loading with growing and decreasing amplitude.

  • We measured the pipeline uplift onset and uplift rate for following shakings.

  • We discussed the impact of a shaking history on pipelines with different apparent unit weight.

  • We highlighted the importance of transient horizontal pipe displacements during liquefaction.

Abstract

Liquefaction hazard induced by seismic events might affect the stability of buried pipelines, with permanent displacements of the embedded structure as a consequence. In this context, the pipe uplift is of main concern, as already widely discussed in the literature. In this paper, further insights to assess the phenomenon and the related effects on pipelines are provided based on the results of shaking table tests on pipes of different apparent unit weight. A multiple shaking sequence is adopted on each model. Among others, data here reported confirms that, with liquefaction occurrence, upward displacement is expected for a pipe with an apparent unit weight smaller or equal to the unit weight of the sand deposit; the opposite for larger apparent unit weight. The pipe vertical displacement is initiated with the first earthquake that in sequence induces liquefaction extensively in the soil deposit. The uplift rate is higher for this shaking, and then it reduces for the next earthquakes even for bigger shaking amplitude. The onset of uplift varies for successive earthquakes depending on the magnitude of the input motion, the relative density and the weight of the structure. A transient horizontal movement of the transversal cross section of the pipe, due to the earthquake loading inducing liquefaction in the soil deposit, can be significant and induce pipe damages.

Introduction

In the field, evidence of uplifted embedded structures induced by soil liquefaction was observed with numerous earthquakes. The 1964 Niigata Earthquake and the 1983 Nipponkai-Chubu Earthquake [15] raised the interest in studying the phenomenon and marked the beginning of the research activities on liquefaction and related effects on buried structures. Further confirmations may be found in the 1993 Kushiro-Oki Earthquake, due to surrounding backfilled sand and alluvial deposits [34]; the 1994 Hokkaido-Toho-Oki Earthquake [19]; the 2004 Niigata ken-Chuetsu Earthquake, due to surrounding refilled loose sand near sewage structures [30,33]; the 2010 Chile Earthquake [10]; the 2011 Tohoku Earthquake [1,6]; and, the 2010–2011 Christchurch Earthquake series [8]. Starting from these manifestations, different studies have been done trying to understand the responsible mechanism, through model tests (e.g. Refs. [7,12,17,27], and numerical simulations [11,21,24] by accounting for the influencing parameters and their effects on the final uplift displacement. A large overview of these researches is summarized in Castiglia et al. [2].

By briefly framing the phenomenon, as it was figured out by Koseki et al. [17], the uplift process comes with the occurrence of liquefaction that reduces the soil shear strength allowing lateral deformations of the soil around the structure, that is a first component responsible for the uplift. Moreover, the effective stress below the structure is different from the value in the free field at the same depth, thus generating a horizontal hydraulic gradient in excess pore pressure with the movement of pore fluid from the free field towards the bottom of the pipe, that is a second component responsible for the uplift. Both these aspects consist of a 2-D phenomenon of lateral deformations and pore fluid migration. Chian et al. [7] observed a global soil deformation following a circular flow around the pipe, that develops due to the lifted soil filling the cavity below the pipe while the structure uplifts. The cavity is quickly filled by the low-strength liquefied soil, thus impeding the structure to go back to its previous position. In this condition, the soil cover would be reduced leading to lower uplift resistance for the next earthquakes, even if the excess pore water pressure would also be reduced due to the soil densification; in this way, effects are compensated. The authors highlighted an effect of dilation for the soil close to the pipe due to the cyclic response of the pipe to horizontal earthquake shakings, that induce stresses in the surrounding liquefied soil, thus allowing the liquefied soil around the pipe to be displaced towards the cavity.

The overall mechanism is well caught but further studies are necessary to better clarify some open issues. The influence of the soil relative density and the earthquake magnitude on the uplift displacement of a buried pipe was already investigated by Yasuda et al. [35] and recalled in successive studies. Considering the difficulties of interpreting the test results, usually, only a single input motion is applied for each test to allow an easier understanding of the model response. This input motion set-up does not allow studying the pipe response for subsequent earthquakes, which is also an important point to be assessed. Few researchers used multiple subsequent loadings, such as Koseki et al. [17], Chian and Madabhushi [4] and Huang et al. [12]. Chian and Madabhushi [4] outlined a reduction in uplift for higher relative density achieved after the first earthquake, but if a longer duration is used for the second earthquake a bigger displacement can be experienced. This topic opens another unsolved issue related to pipeline performances. Worldwide case histories of repeated liquefaction manifestations are available, and among them, the 2010–2011 Christchurch earthquake series is a typical example. The effects of the seismic history on the liquefaction resistance of sand have been studied by different researchers and led to contrasting outcomes, due to the multiple factors that contribute to the process so that the effect of repeated liquefaction on the soil is still unsolved. Although it can be expected that the liquefaction resistance of soil may increase after previous earthquakes due to the pore water dissipation and the increase in relative density of sand, the soil can gain or lose resistance after the liquefaction history. Finn et al. [9] analyzed the effect of the strain history on the sand liquefaction through element testing and found that large strains would result in lower liquefaction resistance, while partial liquefaction occurrence or small strains show higher resistance of the soil. Seed et al. [26], by using large-scale model testing, observed an increase of liquefaction resistance after the application of low-level seismic shakings; five small earthquake shocks with magnitude 5 applied to a soil deposit of 54% relative density resulted in a resistance to liquefaction 45% higher than that available without previous seismic history. Ishihara and Okada [14] explained the change in soil liquefaction resistance for future occurrences as a consequence of the induced anisotropy of soil structure by the application of pre-shearing. This concept was again recalled by Oda et al. [23] who conducted a microstructural interpretation to explain the mechanism of a sharp decrease in liquefaction resistance after pre-shearing. Koseki et al. [20] have studied the multiple-liquefaction behavior of sand through both shaking table model tests and cyclic stacked-ring shear apparatus and observed that the previous liquefaction events can bring to both an increase and a decrease in the liquefaction resistance for future manifestations, depending not on the achieved relative density but on the past liquefaction history. In this context, the impact of liquefaction and repeated liquefaction of soil on the buried structure response is also a significant issue to be addressed together with the liquefaction resistance performances of soil. The re-liquefaction impact on buried pipelines has not been assessed yet. Moreover, the response in vertical displacement of a buried structure with different apparent unit weight was only partially addressed by previous researchers (e.g., Refs. [17,25,31]. Overall, it emerged that an underground structure having a unit weight smaller than the unit weight of the surrounding soil is subjected to an upward displacement, but the detailed displacement response of the buried structure has never been clearly assessed. Sumer et al. [28] studied the effects of the critical density of pipeline for flotation in soils liquefied by waves and found out that the pipeline floats if its specific gravity is smaller than that of the liquefied soil, otherwise it sinks.

This paper provides some additional insight into the mechanism of pipe displacements during seismic liquefaction. It presents the results of shaking table tests, in which the pipe response in uplift is investigated by applying a series of sinusoidal shaking steps with the same characteristics but different amplitudes. While changing the amplitude, the soil experiences a shaking/liquefaction history that changes its relative density and the stress-strain conditions. Due to the previous shakings, considerations on the uplift for each step are outlined. This aspect allows the understanding of how the liquefaction and re-liquefaction of soil deposits impact on the buried structures interacting with soils that have gone through multiple earthquakes. The same seismic sequence is applied to models with pipes having a different apparent unit weight to evaluate the differences in the behavior. In real applications, the different apparent unit weight is mainly related to the dimensions of the considered pipe and to the transported fluid. However, independently from the specific pipe application, it is important to understand whether it uplifts or sinks for proper risk assessment.

Section snippets

Model tests description

The research was conducted by means of the 1-g shaking table machine available at the University of Tokyo. Soil models with dimensions of length 2.8 m, width 0.4 m and height 0.55 m were prepared in a rigid soil box composed of a steel frame structure and transparent acrylic walls, by using a silica sand # 7 [29]. A simplified soil box scheme is sketched up in Fig. 1. Sand physical properties and the grain size distribution curve, for the employed material, are available respectively in Table 1

Response in acceleration and frequency

While studying the pipeline behavior, it is important to understand the response of the soil deposit interacting with the buried structure too. The seismic response of the soil deposit is influenced by the stiffness of the soil layers that is affected by the effective stress changing during the earthquake loading. Indeed, the stiffness of a liquefiable layer changes while the pore water pressure increases, thus modifying the amplification characteristics of the seismic waves through a soil

Discussion

The dependency among the different variables that have been accounted for in this experimental work is summarized in Fig. 16. Analyzing the effects of these variables on the pipe displacements, it can be highlighted that:

  • Three ranges of excess pore water pressure can be identified by looking at the pipe vertical displacements (Fig. 16-A); excess pore water pressures below 1.0 kPa does not induce significant vertical movements (i.e., below 2.0 mm); excess pore water pressures in the range

Conclusions

The susceptibility of underground structures to liquefaction, including buried pipelines, was proved during historical earthquakes. Given the importance of such structures for the development of urbanization and considering the damage effects that pose serious threats to safety, functionality and economy, the research on the topic should be of consistent support. In this paper, further insights into the understanding of the uplift-related phenomenon are provided with the support of model tests

Author statement

Castiglia Massimina: Term, Conceptualization, Methodology, Investigation, Validation, Formal Analysis, Data Curation, Visualization, Writing - Original Draft, Writing - Review & Editing.

Santucci de Magistris Filippo: Term, Conceptualization, Methodology, Supervision, Funding acquisition, Writing - Review & Editing.

Onori Filippo: Term, Conceptualization, Validation, Writing - Review & Editing, Project Administration.

Koseki Junichi: Term, Conceptualization, Methodology, Supervision, Resources,

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was partially financed by the Saipem corporation (Italy). Fruitful discussions with their technical staff are warmly acknowledged. Shaking table tests were executed with the assistance of Ms. M. Nucciarone and Mr. E. Di Biase, University of Molise; the staff members and students of the Geotechnical Engineering laboratory of the University of Tokyo are also warmly acknowledged for their precious continuous care and help.

References (34)

  • M. Castiglia

    The experimental study of buried onshore pipelines seismic-liquefaction induced vertical displacement in shaking table tests and its remedial measures

    (2019)
  • S.C. Chian et al.

    Effect of soil conditions on uplift of underground structures in liquefied soil

    Journal of Earthquake and Tsunami

    (2012)
  • S.C. Chian et al.

    Floatation of underground structures during the Mw 9.0 Tōhoku earthquake of 11th March 2011

    (2012)
  • S.C. Chian et al.

    Soil deformation during monotonic and seismic pipe uplift in liquefiable soil

    Journal of Pipeline Engineering

    (2015)
  • M. Cubrinovski et al.

    Performance of horizontal infrastructure in Christchurch city through the 2010-2011 Canterbury earthquake sequence

    (2014)
  • W.D. Finn et al.

    Effect of strain history on liquefaction of sand

    J Soil Mech Found Div

    (1970)
  • GEER - Geoengineering Extreme Events Reconnaissance Association

    Geoengineering reconnaissance of the 2010 Maule, Chile earthquake

    (2010)
  • Cited by (13)

    • Seismic response of buried pipes in sloping medium dense sand

      2023, Soil Dynamics and Earthquake Engineering
    • Mitigation systems for the uplift of buried pipelines in liquefiable soils under repeated shaking through model tests

      2021, Soil Dynamics and Earthquake Engineering
      Citation Excerpt :

      Despite the reduced amount of uplift with the ongoing loading sequence, the upward displacement once initiated continues also with the shaking steps characterized by smaller amplitude. This highlights an uplift-prone behavior of the structure for multiple earthquake occurrences [12]. For the tests with the application of a remedial measure, TP_DP records a cumulative maximum uplift of 32.5 mm, TP_GB of 8.2 mm and TP_G of 14.7 mm.

    View all citing articles on Scopus
    View full text