Design and Load Testing of Large Diameter Open-Ended Driven Piles (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

33 chapter four PRIVATE-SECTOR PERSPECTIVES ON LARgE DIAmETER OPEN-ENDED PILES This chapter outlines some of the perspectives of several private-sector individuals on the design, installation, and test- ing of LDOEPs. The information discussed is based primarily on interviews. Those requesting to be interviewed represent consultants and contractors. A brief introduction of the par- ticipants is included first, followed by summaries of the topics discussed. The summaries presented are based on all of the interviews unless specifically noted. Detailed notes of each interview are included in Appendix D. PARTICIPANTS The private-sector participants included: • Dr. D. Michael Holloway, P.E.—Consulting Engineer • Mr. Mike Muchard, P.E.—Applied Foundation Test- ing, Inc. • Mr. Steven Saye, P.E.—Kiewit • Dr. Robert Stevens, P.E.—Fugro–McClelland Marine Geosciences, Inc. • Mr. Scott Webster, P.E.—GRL Engineers, Inc. Holloway is in a private consulting practice based in the San Francisco Bay area. He has provided on-site consulta- tion and testing services addressing foundation engineering problems for more than 40 years. Holloway specializes in testing and analysis of deep foundation systems, with empha- sis on driven pile foundations, marine design and construc- tion, instrumentation, and static and seismic soil–structure inter action. He provides expertise for design, construction, and forensic investigations on assignments nationwide and overseas. Muchard is a founder of Applied Foundation Testing, Inc., a firm that specializes in deep foundation testing, including dynamic, Statnamic® (rapid), static testing, and instrumen- tation. Muchard has 25 years of geotechnical engineering, geo-structural engineering, and heavy civil construction expe- rience. His LDOEP experience includes steel piles, spuncast concrete, and bed-cast cylinder piles for deepwater offshore structures, waterfront, and bridge structures. His expertise is in instrumentation, Statnamic® (rapid), dynamic, and static load testing of LDOEPs. This expertise extends into the numerical analysis and interpretation of LDOEP load test measurements. In addition to the testing and instrumentation, he often pro- vides consulting related to the design, drivability, and installa- tion problems associated with LDOEPs. Saye is a Senior Geotechnical Engineer and Design-Build Geotechnical Technical Lead for Kiewit, supporting projects across North America. Saye has 38 years of experience and is a recognized expert in the implementation of geotechnical engi- neering for design-build projects, as well as the design of soft soil ground improvements. Key design-build projects involv- ing LDOEP foundations include the Permanent Canal Closure and Pumps project in New Orleans, Louisiana; the Port Mann Bridge Project in Vancouver, British Columbia; and the Pitt River Bridge project in Vancouver, British Columbia. Stevens has 40 years of experience as a geotechnical engi- neer, primarily in deep foundation design and testing for offshore projects. He joined the Special Projects Group of McClelland Engineers, Inc. in March 1978, and has since worked on offshore projects with LDOEPs throughout the world. He currently chairs the in-house advisory group for pile driving monitoring and analysis at Fugro–McClelland Marine Geosciences, Inc. He is a member of the Pile Founda- tions Standards Committee of ASCE and also the Geotech- nics Chair for the ASCE Coasts, Oceans, Ports, and Rivers Institute (COPRI) Marine Renewable Energy committee. Webster has 33 years of experience in construction and testing of driven piles for both offshore and on-shore struc- tures. Working for both STS Consultants and GRL Engineers has allowed him to develop a strong background in dynamic testing for driven piles. Webster has worked extensively since 1986 with dynamic testing and analysis techniques on a vari- ety of driven pile projects. Since about 1994 most of his work has focused on drivability analyses and dynamic pile testing for offshore projects within the United States and abroad. ISSUES AND EXPERIENCES Pile Plug Behavior A common issue or theme among the participants was the issue of pile plugging. As noted in chapter two, this is one of the most significant topics of LDOEP behavior that is not completely understood. The participants all mentioned that plugging or the absence of plugging dominates the driv- ing behavior of the pile, so that understanding if it is plug- ging or not is key to understanding how the pile will drive. Because plugging is not well understood, it is difficult to predict and thus is often treated as if the choice is one or the other: plugged or unplugged. The actual behavior of the pile

34 is actually somewhere in between. Specifically, how plugging affects driving is not just a function of the soil, but is related to pile diameter and hammer selection. Although plugging behavior is not well understood and is difficult to predict with certainty the general consensus is that LDOEPs tend to not plug during driving. The gen- eral observation is that the piles are advancing as a “cookie cutter” into the soil. Stevens’ experiences on more than 250 offshore platforms indicate that plugging is rare. He has witnessed LDOEPs driven up to 300 ft into clay soils with no plugging. This is not to say that plugging during driving does not occur; however, most observations tend to be unplugged. Saye emphasized the need for additional research to add to the understanding of this behavior. A common question when assessing drivability and plug- ging is what is the proportion of skin friction development between the outside and the inside surfaces of the pile? There is no clear understanding of how the soil inside the pile behaves and how much of the friction resistance is derived inside the pile and how much outside. Designers often make an assump- tion based on local experience or rules of thumb. Holloway noted that the old rule of thumb was to assume two-thirds of the friction is outside the pile and one-third inside the pile. How much skin friction develops inside the pile is prob- ably related to the relative acceleration of the pile to the soil mass inside. Stevens’ early work included investigating analyses of plugging behavior by evaluating the accelera- tion of the soil mass in the pile with respect to the accel- eration of the pile. The inertia of the pile is almost always greater than the soil mass under the large forces from the hammer on the pile. Webster noted that the blow of a hydrau- lic hammer results in higher pile acceleration than that from a diesel or steam hammer, commenting that the hydraulic hammer acceleration is somewhat similar to the action of a vibratory hammer with respect to the acceleration helping to drive the pile with each blow (the action is similar, but the acceleration of vibratory hammers is much smaller). It is thus important that inertial forces be carefully evaluated in the wave equation analyses to attempt to adequately model the pile–soil interaction along the inside of the pile. Although there is general agreement that most LDOEPs usually do not plug during driving, a majority also agree that the behavior under static loading will usually involve plugged behavior to some degree. There is still enough uncertainty among designers, however, that considering the pile to behave as a friction pile (unplugged) or a displacement pile (plugged) can affect the approach the designer takes when calculating the static resistance of the pile. Plug behavior for long-term static capacity still requires more investigation to be fully under- stood for pile design. There have been some recent cases of LDOEPs driven with devices installed inside the pile to encourage partial or full plugging of the pile in an attempt to ensure plugged behavior for long-term static resistance (Muchard et al. 2009; Terracon 2014). Stevens has observed such behavior only a few times, but believed it was relatively successful for achieving a plugged condition. Dynamic and Static Testing Dynamic testing of LDOEPs is a common method for evalu- ating pile resistance. However, there is concern in the industry that dynamic testing does not adequately indicate the avail- able pile resistance (see also discussion in chapter two). With the high loads that LDOEPs are capable of, it can be difficult to have a hammer of the appropriate size to verify or test the pile resistance. The lower soil resistance during driving allows for the use of a smaller hammer for installation than may be needed to demonstrate the full available resistance of the pile. The effect of plug behavior on dynamic testing is not fully understood, implying that it can be very difficult to accurately assess plugging through dynamic testing. Improved instru- mentation means that reliable strain measurements on pipe piles can now be obtained to help answer some of the ques- tions about soil plugging. Holloway made special note on evaluating pile plugging through testing, explaining that static or pulse loading tests should be considered to help with plug evaluation either by doing an uplift/pullout test or drilling out the soil in the center of the pile and then doing dynamic testing to quantify friction distribution along the pile shaft. If an internal plate is used to fix the plug in place, soil mass needs to be added to the pile in the WEAP® model for the portion below where the plug is expected to form as well as increasing toe quake. All of the participants emphasized that proper use of PDA® equipment to monitor and test LDOEPs is essential to proper assessment of pile resistance and pile driving behavior. Vari- ables to be considered when testing LDOEPs include: • Instrumentation location and quantity. The larger diame- ters require the use of two sets of gauges to better average the stress–time history across the pile section. For small diameter piles, one set of gauges consisting of two strain transducers and two accelerometers set 180 degrees apart on the pile perimeter is sufficient. For LDOEPs, two sets consisting of four transducers and acceler- ometers, each placed 90 degrees apart around the perimeter of the pile are needed. Using only one set on an LDOEP as is common on small diameter piles usually yields poor quality data, especially with spiral- weld pipe, where the placement of the instruments can affect the interpretation of the results. Recent experi- ence by Saye on hurricane protection projects in New Orleans, Louisiana, has confirmed the necessity of this practice for LDOEPs.

35 • The durability of the pile during driving. High required nominal pile resistance can lead to aggressive driving and an increasing risk of pile damage. • Temperature of piles for instrumentation before and after installation, especially steel pipe piles. • Accounting for residual stresses from manufacturing in spiralweld pipe piles and concrete piles. • Sizing the hammer large enough to move the pile suf- ficiently to mobilize the target nominal resistance to be demonstrated. • How plugging is interpreted in the data. Owing to the large size of the pile, static tests are not as common for LDOEPs as for smaller piles, although several case histories are available in the literature, some of which are summarized in chapter five. In some cases, there is signif- icant disagreement between dynamic testing and static test- ing, with dynamic testing usually under predicting the static pile resistance. Although this is conservative with respect to design, if dynamic testing significantly under predicts that static resistance, designs become inefficient. Some experi- ences, however, such as those of Stevens generally observe very good agreement (within 5%) of pile resistance deter- mined from CAPWAP® analyses of dynamic test results and that of static load tests. Saye has observed that load test data for LDOEPs appear to be limited with respect to information on the distribution of resistance with depth. Instrumentation in the lower por- tions of piles can have difficulty surviving the pile installa- tion process. Until recently, Muchard experienced the same problems with instrument survivability. He reports that sig- nificant advances in strain gauge designs for LDOEPs have been made, allowing for greater survival of the instruments. For example, a 280-ft-long LDOEP was strain-instrumented along its length, successfully measuring load distribution during rapid load testing on the Tappan Zee Bridge Pile Demonstration Project (still to be published). Similar recent strain instrumentation successes on LDOEP rapid load tests have been reported by MnDOT. Embedded strain instru- mentation has also been installed and monitored in spun- cast concrete LDOEPs (Muchard 2005). In one case study, the strain gauges were monitored during the pile post- tensioning process, during driving, and then during subse- quent load testing. Muchard emphasizes that the actual static load distribu- tion along the pile has proven beneficial to the understand- ing of the performance of LDOEPs and is an extremely useful design and verification tool. More of this type of data is needed. Static Axial Resistance Calculations From the interviews and literature review there emerged a general opinion that the more widely used methods of axial analysis significantly underestimate the pile resistance of LDOEPs. These methods do not appear to adequately capture the influence on axial resistance that the construction prac- tices for these larger piles have. Most of the design methods used by transportation agencies are based on tests performed on small diameter piles. For piles above 36 in. in diameter, there are not as many well-documented case histories of the evaluation of static analysis methods, especially cases where the geotechnical limit state was evaluated. In many cases, significantly more resistance is available than con- sidered in design. Webster’s and Stevens’ offshore experiences with regard to why higher resistance values are not taken advantage of are typical of the comments of the others interviewed, as well as the experiences of the authors and others in the practice. Reasons often cited are related to schedule risks rather than a lack of knowledge, although it still occurs: • Owners and/or designers are unwilling to investigate higher resistance. Owners are willing to trade lower pile resistance for the reduced risk of construction problems or claims (e.g., “We have never had problems in the past doing it this way”). • The project schedule or other constraints do not allow the time or access needed to demonstrate and use higher resistance resulting from soil setup. In both the agency and private practice interviews, those that realize analysis methods are conservative for LDOEPs generally use static analysis methods for a rough estimate of the pile resistance. These estimates are tempered by experi- ence, with dynamic testing and wave equation analysis inter- pretations providing better estimates of pile resistance. In some instances, for example temporary works such as tres- tles, a contractor requires installation of a pile to a certain resistance to be available soon after driving. Being able to calculate the long-term static resistance of the pile is of less consequence than knowing that the pile has sufficient resis- tance at the end of driving to be placed into service almost immediately. Although the methods available to transportation structure designers are included in the various FHWA and AASHTO design documents, the offshore industry uses the API RP2 GEO design guide and its associated methods as noted in chapter two. Stevens believes that estimating static resistance with API RP2 GEO is probably the most effective approach and entirely applicable to transportation structures. He has consulted on bridge projects using the API methodology (Bay Bridge in San Francisco and Trans Tokyo Bay Bridge in Japan). API regularly updates the procedures based on data and experiences from the field and ongoing research. Recent updates include modifications to the design of piles in sands based on CPT test data. This procedure provides better esti- mates of pile resistance in very dense sands; however, he

36 cautions that care be exercised when estimating the mobilized end bearing for drivability analysis. One major issue in practice noted by Holloway is the use of different methods for estimating the base and side resis- tance. For example, using Meyerhoff for side resistance and Nordlund for base resistance. The approach to the pile–soil behavior for each method is different; therefore, mixing methods can lead to poor predictions of total pile resistance. An issue noted by both the industry and agency interviews is how to account for setup or long-term increase in side resis- tance over time. Testing LDOEPs after significant setup can be a challenge, as noted in some of the case histories in chap- ter five. Earlier in this chapter, it was noted that load tests are often not taken to failure; therefore, measuring the full gain with time is not accomplished. In some project environ- ments, such as temporary works or design-build delivery, allowing for setup to occur may create significant negative schedule (and thus cost) impacts. The piles are thus intention- ally designed to be less efficient to accommodate the lower resistance without setup. Another design issue that Holloway notes is that many designers do not account for residual stress in the pile analysis. This leads to overestimating side resistance and underestimat- ing base resistance. In many cases load tests are essentially “proof” tests (without reaching failure) confirming that the structure load can be supported. Muchard adds that this prac- tice can lead to non-conservative designs if the pile lengths are optimized (shortened) on the basis of overestimated side shear resistance. However, this does not help provide a clear understanding of the true resistance and how the soil–pile inter- action actually behaves, which can lead to significant inaccu- racy using unit resistance values in making adjustments to the pile toe elevations. Driving Behavior Holloway discussed how it is important that designers recog- nize that how an open pile drives is highly dependent on how the stresses are reaching the toe of the pile—the failure mech- anism at the toe. With relatively thin-wall steel piles, wave equation analysis will sometimes indicate that toe stresses are not very large relative to the yield stress, but that piles still have problems with collapsing during driving, usually because of poor driving alignment resulting in ovaling and the collapse of the piles owing to transverse and eccentric stresses. He rec- ommends that stresses at the toe need to be less than half of the yield stress of the steel to accommodate the eccentric forces encountered at the toe when significant end-bearing or poten- tial obstructions are anticipated. Saye discussed some experiences of Kiewit in clay soils with significant setup that indicated long delays in driving for splicing have the potential for damaging the long-term pile resistance. A common practice is for a contractor to set the first section of many piles, weld the next section on all of the piles started, and then return to drive the piles after all have been spliced. This practice can sometimes require weeks or even months between the driving of the first section and the re-start of the driving of the first piles that were set. It is possible that re-driving after such a significant time of setup could have a negative effect on the pile shaft resistance as the remolding of the soils along the pile could result in lower pile resistance than expected or would be available had splicing and the re-start of driving occurred within a short time period rather than weeks. In the practice of “sticking” a lower section of pile before splicing, it is common for a contractor to use a vibratory ham- mer to install the first section. It is also common for a vibra- tory hammer to be used to set a pile that is installed in one section. The design of piles for axial loading does not always take into account any vibratory installation, rather it assumes that the piles are impact-driven the entire length. The effect of vibratory installation on pile resistance is not well-understood, with very little comparison of the actual effect of the vibratory hammer on pile installation in clay available. Significant dif- ferences in opinion exist as to whether or not vibratory ham- mer installation has a negative influence on pile resistance, especially in clays. Saye noted that the U.S. Army Corps of Engineers specifications for some of the hurricane protection projects in New Orleans allowed vibrating up to 50% of the pile length in clay before impact driving. Most standard speci- fications do not address the use of vibratory hammers. Driving Piles to Bear on Rock Holloway and Stevens provided some insights when driv- ing steel LDOEPs to bear on rock. Holloway has observed instances where hard driving into sedimentary rocks can result in the breakdown of the composition of the rock. Relaxation of base resistance tends to occur as a result of the breakdown of the rock structure, resulting in less long-term base resistance than estimated (substantial toe relaxation). Restrikes frequently show a decrease in base resistance, with piles being driven 1 to 1.5 m into the rock in order to observe base resistance increase again in some cases. Holloway believes that careful evaluation of the impact of the pile at the rock interface is necessary for these piles. Stevens’ experiences offshore have included piles driven through layers of soft rock (very hard clay, shale, siltstone, gypsum, etc.) to achieve resistance. In such conditions, the toe stresses and toe displacements must be carefully moni- tored. Driving is to be halted if toe stress reaches 80% of the yield stress of pile. If the toe displacement turns negative, it indicates the toe is being damaged or crushed. It is also impor- tant to check drivability using 90% end bearing (modeling a sudden fixed-end condition) and evaluating the resulting stresses at the pile toe.

37 RESEARCH NEEDS IDENTIFIED BY PRIVATE SECTOR Areas that were identified by the practitioners as needing addi- tional research and investigation included: • Comparing dynamic and static load tests to better cor- relate dynamic testing and wave equation analyses with static resistance. • Defining the pile movement corresponding to the selected pile load test capacity for the LDOEP static load tests. • Evaluating the influence of vibratory pile installation on the capacity of piles, including LDOEPs. • Evaluating the effect of delays in pile installation for splicing on the side resistance capacity of LDOEPs. • Assembling good quality instructive case histories or databases of LDOEP behavior with cone penetration tests to characterize the soil conditions. • Investigating the specific differences between steel and concrete LDOEPs comparing the two types on the same site to investigate the differences in driving behavior, plugging, and pile resistance in the same conditions. • Further investigating pile plugging during driving. • More load tests investigating distribution of resistance with depth (e.g., strain instrumentation installed along the length of LDOEPs).