Use of Case Histories to Illustrate the Effect of Installation Activities on the Side Resistance of Pipe Piles in Plastic Soils

This paper presents observations of the influence of pile installation methods on the measured side resistance of driven pipe piles bearing in plastic soils. Pile installation practices that can reduce the capacity of completed piles in plastic soils from that calculated include re-driving of piles after partial or full dissipation of excess pore water pressure, long installation times, slow jacking, the influence of surface casing in soft soils, and the use of vibratory hammers. Selected case histories are used to illustrate how deviations from rapid and nearly continuous pile installation can result in poor performance. Designers and Contractors need to be more aware of the damaging effects of these practices and that the selected design approach may not effectively consider the selected installation procedures. Where project requirements dictate use of these installation approaches that damage pile side resistance, the proposed construction influence factors may be used to modify a SHANSEP-based side resistance method to estimate the potential reductions in side resistance associated with the selected approach. Although the number of case histories available are too few to characterize the reliability of the construction influence factor approach, the cases sufficiently demonstrate the damage that can occur when these practices are used and the importance of construction procedures for driven pipe pile foundations that minimize or avoid damage to side resistance and to highlight where special design and testing are merited. struction actions or interruptions can have on the pile side resistance. A generalized design approach is then presented to help designers and contractors address this issue. Saye et al. (2013) presented an empirical approach to assess the average side resistance of driven pipe piles in plastic soils using an adaptation of the Stress History and Normalized Soil Engineering Properties, SHANSEP, concept (Ladd and Foott, 1974). This approach includes assessments of the soil overconsolidation ratio, OCR, developed using both oedometer tests and an empirical correlation to undrained shear strength data. The empirical correlations in the SHANSEP-based approach incorporate the results of selected pile loading tests and laboratory OCR assessments presented by Almeida et al., (1996) and others to capture the effect of soil state on the strength of the soil-pile interface. Experience applying the SHANSEP-based approach shows that some pile installation practices can damage the pile side resistance. These damaging practices are identified. This paper then applies the SHANSEP-based approach to a small set of loading tests and soil property records to demonstrate how reductions in the side resistance of driven pipe piles in plastic soils may result from specific pile installation practices. When recognized, these damaging practices can either be avoided or considered in the planning and installation of the foundations. Specifically, this paper seeks to: 1) assemble a set of case histories where the pile installation method appears to have adversely affected the measured pile side resistance and clearly identify these damaging practices, 2) develop construction influence factors to estimate the detrimental influence of redriving, slow jacking, effect of 1 Senior Geotechnical Engineer, Kiewit Engineering Group, Inc., 3555 Farnam St Suite 1000, Omaha, Nebraska, 68131, USA 2 Geotechnical Engineer, HDR Inc., 1917 S 67th St, Omaha, Nebraska, 68106, USA 3 Professor, School of Civil and Construction Engineering, 101 Kearney Hall, Oregon State University, Corvallis, OR, 97331, USA * Corresponding author, email: steve.saye@kiewit.com © 2020 Deep Foundations Institute, Print ISSN: 1937-5247 Online ISSN: 1937-5255 Published by Deep Foundations Institute Received 21 January 2020; received in revised form 28 June 2020; accepted 3 November 2020 https://doi.org/10.37308/DFIJnl.20200121.213 Introduction Driven pipe pile foundations are widely-used in numerous industries, ranging from the transportation infrastructure sector to the energy sector. Large variations in pipe pile diameter and length now occur potentially extending well beyond the empirical record of selected design procedures. Common design procedures such as the Semple and Rigden (1984) alpha method do not directly consider the influence of how piles are installed. Any variations in construction activity are reflected in the selection of pile tests for the empirical correlation. Some empirical approaches proposed over the years indirectly capture these possible responses. This paper presents a set of case histories of pile loading test records to help designers and contractors realize that there exist certain practices that damage pile side resistance in plastic soils. An empirical approach is presented to clarify the effect that specific convol14no2saye213.indd 1 27/01/21 5:53 PM 2 | DF I JOURNAL © Deep Foundations Institute Saye, Kumm, Stuedlein | Use of Case Histories to Illustrate the Effect of Installation Activities on Side Resistance surface casing in soft soils, and vibratory hammer installation on the calculated side adhesion capacity, and 3) present a generalized design approach that addresses these damaging activities. Consideration of the aforementioned practices should improve the rate of successful installation of driven pipe piles and interpretation of anticipated capacity. Furthermore, load test records where the installation damaged the side resistance should be identified and excluded from databases used to develop empirical assessments of pile capacity for nearly continuous installation with an impact hammer. Shansep-Based Approach to Assess Pile Side Resistance In the SHANSEP-based approach to assess the undrained side resistance of closed-ended and small diameter open-ended piles driven into plastic soil proposed by Saye et al. (2013) and placed within the reliability-based LRFD framework by Stuedlein et al. (2020), the average undrained side resistance, qs, is normalized by the effective overburden stress, σ’vo, and then related to the average soil OCR in the same format as the SHANSEP concept (Ladd and Foott, 1974): qs/σ’vo = 0.19 (OCR) 0.7 (1) where: qs/σ’vo = the normalized pile side adhesion, and OCR = the overconsolidation ratio. This approach is described on Figure 1. The method assumes that the pile response represents undrained conditions (rapid loading) with static loading tests only lasting a few hours or less. Most piles are installed continuously with an impact hammer and the SHANSEP-based approach provides a good assessment of the side resistance, as shown on Figure 1. Reported cases of closed-ended piles that were driven nearly continuously with a limited time for splicing to depths corresponding to slenderness ratios, defined as the ratio of length, L, to diameter, D, exceeding 50 were evaluated to confirm that qs/σ’vo does not decrease with increases in length. Endley et al. (1979) described a “long” test pile (i.e., L/D > 50), designated Pile 16 and located at Site E, bearing in the southeast Texas coastal plain deposits. These marine deposits exhibit high OCRs near the ground surface due to desiccation (Mahar and O’Neill, 1983); at Site E, normally-consolidated conditions were encountered first at the depth of 32 m as indicated by the su/σ’vo profile derived from good-quality UUC specimens shown in Figure 2a. The pile was closed-ended with a diameter of 0.274 m and a length of 32 m. The full soil profile is considered overconsolidated as a result of desiccation and overburden stress changes. Pile 16 was driven to a depth of 32 m without significant delay and allowed to set up for 9 days prior to static load testing. Although this setup time may appear short in duration, as a result of the small diameter and the high coefficient of consolidation associated with desiccated and overconsolidated plastic soil (e.g., Tavenas and Leroueil 1980; Jamiolkowski 1983) it does not appear to have impacted the measured capacity. The calculated average OCR and average qs/σ’vo values for this “long” continuously-driven pile presented on Figure 2 are shown to be consistent with Equation (1) suggesting that a reduction in the calculated pile capacity due to pile length is not necessary with the SHANSEP-based approach. The case histories presented in this paper show that some measured pile behavior is significantly different from Figure 1, with variations related to damage to the side resistance in the pile installation sequence. These “reduced capacity” pile loading tests are shown as unreliable data on Table 1 and Figure 3. Figure 1. qs /s’vo vs. OCR in the SHANSEP-based approach format for closed-end and small-diameter open-end pipe piles referencing reliable stress history data (adapted From Saye et al. 2013) vol14no2saye213.indd 2 27/01/21 5:53 PM © Deep Foundations Institute DF I JOURNAL | 3 Saye, Kumm, Stuedlein | Use of Case Histories to Illustrate the Effect of Installation Activities Table 1 is annotated with the activity associated with the reduced capacity and damage to the side resistance. Herein, an empirical construction influence factor, designated CI, is proposed to modify Equation (1) to account for the effect of the specific pile installation method and sequence using the following approach and estimate the reduced capacity: qs/σ’vo = CI 0.19 (OCR) 0.7 (2) Continuously-driven closed-ended pipe piles and small diameter open-ended pipe piles are assigned CI = 1, as would be the case for Pile 16 described above and reported by Endley et al. (1979). The CI factor is intended to help estimate the Table 1. OCR vs. qs/σ’vo and qs/su vs. su/σ’vo Data Considered “Unreliable” Load Test No. L (m) D (m) Open or