FEATURE ARTICLE Embedded Instrumentation For Deep Foundations Over the past 70 years, instrumentation has matured from simple mechanical and hydraulic devices to complex electrical transducers, which when combined with advances in computing, have allowed geotechnical engineers to improve and optimize deep foundation design, verify construction techniques and ensure quality foundations. Applied Foundation Testing (AFT) has been utilizing embedded instrumentation since its inception in 1997. Beginning with installing and monitoring embedded strain gages as part of the SR 292 over Bayou Chico project in Pensacola, Fla., our firm has continued to use embedded sensors such as strain gages, accelerometers, Linear Variable Displacement Transducers (LVDTs), pressure transducers and other sensors in all types of deep foundations. More recently, AFT assisted the University of Florida and Smart Structures, Inc. in developing and deploying the wireless, dynamic testing, Smart Pile System™, generically referred to by the Florida Department of Transportation as the Embedded Data Collector. Embedded Strain Instrumentation Engineers routinely install embedded strain instrumentation to determine load distri- bution along deep foundation types ranging from large-diameter steel pipe piles and drilled shafts to small-diameter micro- piles. Embedded strain gages can also pro- vide indications of pile damage or defects. Examples of some of the more challenging installations include a 300 ft (91.4 m) long, 48 in (1.2 m) steel, pipe pile driven as part of the Tappan Zee Bridge Load Test Program in New York. The owner specified 12 levels of 3 arc weld- Figure 1: Instrumented pile segment at the Green Bayou Pipe Mill prior to shipment to the Tappan Zee site able strain gages for a total of 36 strain gages with the deepest located at approximately 300 ft (91 m) under water. The pile was constructed by field splicing one 160 ft (48.8 m) segment with one 140 ft (42.7 m) segment after the lower segment was partially installed. Because of the critical nature of the test program, 12 additional sister bar strain gages and 4 accelerometers were grouted, after the pile was installed, into a small-diameter pipe welded to the test pile. The rapid loading associated with the Statnamic test used on this project required us to use resistance based gages, which have the unique ability to be sampled at extremely high frequencies, normally only limited by the capabilities of the data acquisition system. AFT’s data acquisition team sampled all the instrumentation simultaneously: 48 strain gages, 7 acceler- ometers and one 10,000 kip (45 mN) load cell. Our sampling rate was 5,000 samples per second. Test scheduling required the strain gages remain operational at depth for over one month after the spliced pile was vibrated and driven to the final test elevation. During testing, all gages functioned, and all but three gages provided reliable strain distribution information. AUTHOR Donald T. Robertson, P.E., Applied Foundation Testing, Green Cove Springs, Fla. 80 • DEEP FOUNDATIONS • MAR/APR 2013 Augered cast-in-place (ACIP) piles also use strain gages. On the recent Brickell Citi Centre in Miami, Fla., a massive private development project, we used extensive strain instrumentation to measure performance of tangent pile groups consisting of four 24 in (0.6.m) ACIP piles to depths of 123 ft (37.5 m). Prior to construction, 3 full scale tangent pile test installations were constructed and load tested with a 10,000 kip (45 mN) Statnamic device. Each test footing consisted of a concrete cap over four strain inst rumented tangent pi les. We instrumented each test foundation unit with 144 strain gages embedded from top to bottom of the 4 piles in the tangent pile group. This type of installation challenges the robustness of the strain gage since they must be plunged to great depths into fresh sand/cement grout. AFT engineers monitored the strain gages simultaneously using a data acquisition system during the 10,000 kip (45 mN) Statnamic load tests. They followed through with compre- hensive data regression and analysis and presented the results to the design engineering team. These measurements provided valuable information for design; most notably, load sharing of each
