Determination of Bearing Pressure

The superstructure loads calculated by a structural engineer will enable the gross foundation bearing pressures to be determined using the bearing capacity theory. However, it is the net foundation bearing pressure which should be used for both bearing capacity and settlement calculations. The net pressure is that part of the gross applied pressure that requires the shear strength of the ground to support it. The bearing capacity of a foundation is one of the classical problems in geotechnical engineering. There are several different methods that are used to calculate the bearing pressure of foundations, as noted in the following list (Day 2006):

  • 1. Engineering analyses or static formulae: Based on the results of subsurface exploration and laboratory testing, the bearing capacity of a foundation can be calculated using engineering analyses that are based on the principles of soil and rock mechanics (indirect, theoretical or semi-empirical) or correlations to in situ test results (direct and empirical). These methods will be summarized in Chapters 4-6 for shallow foundations, offshore spudcan penetration and deep foundations, respectively.
  • 2. Field load test: Prior to the construction of the foundation, a field load test could be performed to determine its carrying capacity. Because of the uncertainties in foundation design based on engineering analyses, a field load test is common and often results in a more economical foundation than one based solely on engineering analyses (Abu-Hejleh et al. 2015). For example, the Building Engineering Division/Institution of Engineers, Singapore/Association of Consulting Engineers, Singapore (BCA/IES/ACES) Advisory Note 1/03 that is based on other standards mandated (1) one number or 0.5% of the total piles, whichever is greater for an ultimate load test on preliminary pile (preferably instrumented); (2) two numbers or 1% of working piles installed or one for every 50 metres of proposed building, whichever is greater for working load test; or (3) two numbers or 2% of working piles installed, whichever is greater for non-destructive integrity test (for the purpose of quality control). This section will focus on the introduction of the field load test that is considered the most reliable method to determine foundation capacity.
  • 3. Application of pile-driving resistance: In the past, pile capacity could be estimated from the driving resistance during the installation of the pile. Pile-driving equations (e.g. Engineering News Formula) were developed that relate pile capacity to the energy of the pile-driving hammer and the average net penetration of the pile per blow. However, studies have shown that the accuracy of pile-driving equations is unsatisfactory as compared to the field load test. Terzaghi and Peck (1967) concluded that the use of pile-driving equations is no longer justified.
  • 4. Wave equation: This method is based on using the stress wave from the hammer impact in finite element analysis. It was first put into practical form by Smith (1962) and later by others. A more detailed discussion of the principles and a reasonably sophisticated computer program are available in Bowles (1997). This method has particular application for piling contractors in determining pile drivability.
  • 5. Specifications and experience: Other factors that should be considered in foundation design include governing building codes or agency requirements. In addition, local experience in terms of what has worked best in the past for local soil conditions may prove valuable in the design and construction of a foundation.

Types of Foundation Load Tests

A load test can be performed on either a non-instrumented or instrumented full-scale pile. With technological advancements, the methodologies and instrumentations for pile load tests have continuously remained under refinement. Generally, there are three types: (1) static load test (SLT) based on dead weight, a reaction system or Osterberg cell (O-cell); (2) Statnamic, now called force pulse (rapid) load test (RLT); and (3) dynamic load test (DLT). Table 3.7 summarizes the main characteristics of three types of pile load tests (Holscher and van Tol 2009). It should be noted that the data is based on current typical approaches for testing. Generally, SLT is expensive and time-consuming (become more so with increased load requirements) but has the advantage of simple analysis and interpretation. On the contrary, DLT and RLT are quick to carry out with more specialized equipment and analysis and hence are cheaper than SLT.

Static Load Test (SLT)

Head-Down Load Test

SLT applies load incrementally to a pile while measuring the pile movement. Types include axial compression and axial tension. The schematic diagram for head-down compression tests given in Figure 3.9 is easily understood.

Table 3.7 Characteristics of three types of pile load test (Source: data taken from Holscher and van Tol 2009)




Duration of loading

1 6 hours

1 00 to 200 ms

7 ms

Number of test per day




Reaction mass needed (percentage of capacity)

1 00% (exception of O-cell)

5%-1 0%


Time needed for result


10 mins

4 hours

Tension stress in pile




Prefab pile




Cast in place




Stress in soil




Poor water pressure in sand




Costs (Euro per tonne)








Note: ms = milliseconds: mins = minutes.

Schematic diagram for head-down compression tests

Figure 3.9 Schematic diagram for head-down compression tests

SLTs are typically performed to a maximum applied load equal to a multiple of the design load or to geotechnical failure. Compression tests utilize an overhead reaction beam and frame with resistance to the applied loads provided by dead weight or reaction piles. Tension tests may also utilize an overhead reaction beam and frame, or they may use only a reaction beam supported on mats. The benefits of SLT are that they (1) provide information that can be used in foundation design confirmation and design refinement, (2) allow the use of a lower FS (allowable stress design) or higher resistance factor (load and resistance factor design) to save construction cost, (3) optimize design from detailed load-transfer information and (4) calibrate static analysis methods (the focus of this book). Although SLTs are considered to be the most reliable way for evaluating pile capacity, they are expensive and time-consuming. Guidance on SLT can be found in ASTM D1143 (2013) for compression and ASTM D3689 (2013) for tension.

Applied loads are determined using a load cell and hydraulic jack pressure, while movement can be measured using digital or mechanical dial gages, multiple of types of displacement transducers, string potentiometers or a combination of these devices. The measured load versus movement is plotted in a way that can be used to define the foundation’s geotechnical capacity, as discussed in the following section. An example is presented in Figure 3.10. Additional embedded instrumentation consisting of strain gages or telltales can be used to measure foundation strain from which load in the foundation can be estimated. An example of load distribution with its instrumented elevation is given in Figure 3.11. Unit shaft and tip resistance values can be determined from load-transfer profiles.

ASTM D1143 (2013) suggested the following procedures for SLT:

1. Quick test: Load is applied in increments of 5% of the anticipated failure load. Each load increment is added in a continuous fashion

Load-movement curve from axial compression test (Sources

Figure 3.10 Load-movement curve from axial compression test (Sources: data from the database maintained by the FHWA)

I Load distribution with the depth (Sources

Figure 3.1 I Load distribution with the depth (Sources: data from the database maintained by the FHWA)

and immediately following the completion of movement readings for the previous load interval. Load increments are added until reaching a failure load but do not exceed the safe structural capacity of the pile.

2. Maintained load (ML) test: Pile is loaded to a maximum maintained load of 200% of the anticipated design load for tests on individual piles in increments of 25% of the design load. Each load increment is maintained until the rate of axial movement does not exceed 0.25 mm per hour.

  • 3. Loading in excess of a maintained test: After the load has been applied and moved in a maintained test, the test pile is reloaded to the maximum ML in an increment of 50% of the design load, allowing 20 minutes between load increments. Additional load in increments of 10% of the design is then applied until reaching the maximum required load or failure.
  • 4. Constant time interval test: Load is applied in increments of 20% of the design load in one hour.
  • 5. Constant rate of penetration (CRP) test: Load is applied to maintain a pile penetration rate of 0.25-1.25 mm per minute for cohesive soil or
  • 0.75-2.5 mm per minute for cohesionless soil. The maximum applied load is maintained until a total pile penetration of at least 15% of the pile diameter is obtained.
  • 6. Constant movement increment test: Load is applied in increments to produce pile head movement increments equal to approximately 1% of the pile diameter.
  • 7. Cyclic load test: Load is applied in increments in accordance with the maintained test. After the application of loads equal to 50%, 100% and 150% of the pile design load, maintain the test load for one hour and remove the load in decrements equal to the loading increments. After removing the maximum applied load, reapply the load to each preceding load level in increments equal to 50% of the design load.

Whether the pile is tested by means of CRP or ML, the rate of load applications is much higher than the loading rate of a building under construction.

Bi-directional SLT

Most pile design practices employ the head load-movement curve as the primary means for interpreting the SLT results, providing the least information on the soil response to the loading procedure (Fellenius 2015). To take full advantage of an SLT, a pile could be instrumented to measure the load distribution. For this reason, the SLT has advanced into the bi-directional method of testing over the past thirty years. As reviewed by Salem and Fellenius (2017), early bi-directional SLT was performed by Gibson and Devennv (1973), Amir (1983) and Horvath et al. (1983). Meanwhile, an independent development took place in Brazil by Elisio in 1983, leading to a method for the piling industry. In the mid-1980s, Dr. Jorj O. Osterberg, professor emeritus of civil engineering at Northwestern University, also saw the need for and use of a test employing a hydraulic jack arrangement placed at or near the pile toe to perform a bi-directional SLT (Osterberg 1995). Based on the existence and availability of the Brazilian device for bi-directional SLT, Dr. Osterberg invented and developed a hydraulically driven, high capacity, sacrificial loading device called the Osterberg cell (O-cell). It can satisfactorily meet the construction industry’s needs and provide an innovative and effective method to test high-capacity drilled shafts and piles (e.g. Osterberg 1995; Hasan et al. 2018; Kalmogo et al. 2019). Outside Brazil, the bi-directional load test is now called the “О-cell test.”

The first production-sized prototype О-cell was placed into experimental service by 1988. As the О-cell is pressurized, it loads the pile in two directions. Figure 3.12 shows one major technical achievement of О-cell test that allows the separation and direct measurement of shaft and toe resistance components of total capacity. The portion of the pile above the О-cell location is pushed upward to mobilize its shaft resistance. Simultaneously, the pile below the О-cell location is pushed downward to mobilize its shaft resistance and its toe resistance. At any pressure, the upward load is always equal to the downward load, eliminating the need for a surface reaction. The pressure is increased until the pile reaches its shaft or toe resistance or both. During a bi-directional SLT, hydraulic fluid pressure supplied to the О-cell is measured using an electronic pressure transducer. From these pressure readings, loads provided by the О-cell are determined. Cell expansion is measured using multiple electronic displacement transducers spanning the cell. Combined with telltale and pile head movement readings, the

Schematic diagram for О-cell test displacement transducer readings indicate the upward movement of the top of the cell and the downward movement of the bottom of the cell

Figure 3.12 Schematic diagram for О-cell test displacement transducer readings indicate the upward movement of the top of the cell and the downward movement of the bottom of the cell. These results are then presented as upward and downward movement versus load. In comparison to the head-down pile load test, bi-directional SLT (1) provides an efficient way of high-capacity test for drilled shafts (or bored piles), (2) separates resistance and movement data for pile head and toe and thereby determine the magnitudes of mobilized shaft and toe resistance, (3) gives the load-transfer mechanism along the shaft and (4) does not need a massive reaction system (e.g. reaction beams, additional piles, anchors or dead load platforms), saving cost and time. More details on the О-cell load test for drilled shafts and driven piles can be found in Osterberg (1995). Guidance on bi-directional SLT can be found in ASTM D8169 (2018).

Figure 3.13 (a beautiful “butterfly” curve) shows an example of the upward and downward load-movement curves measured by Elisio in 1983. The data was taken from Fellenius (2015). The pile was a 13 m long, 520 mm diameter bored pile constructed through 7 m of sandy silty clay and 6 m of sandy clay silt. As a bottom-up loading mechanism instead of the head-down loading mechanism is used in the О-cell test, the load- movement curve at the pile head is not directly measured, which must be constructed using measured shaft and toe load-movement curves. The constructed head load-movement curve is often referred to as an equivalent head load-movement curve. Three methods are available for this construction: (1) original (Osterberg (1995), (2) modified (Schmertmann and Hayes

1997) and (3) load-transfer analysis (e.g. Coyle and Reese 1966; Kwon et al. 2005). For simplicity, only Osterberg’s (1995) original method is introduced, which relies on three main assumptions: (1) pile is rigid, (2) load- movement behaviour above the О-cell is independent of the direction of the

Measured results from О-cell upward and downward movement versus load (the "butterfly” curve) (Sources

Figure 3.13 Measured results from О-cell upward and downward movement versus load (the "butterfly” curve) (Sources: data taken from Fellenius 2015) relative movement between the pile and surrounding soil and (3) load- movement behaviour below the О-cell is the same as when the pile is head loaded. In the Osterberg (1995) method, construction begins by determining two resistance values from the top and bottom load-movement curves at an arbitrary movement. Summation of the two resistance values and the chosen movement is a single point on the equivalent head load-movement curve. Following this procedure, the equivalent head load-movement curve is plotted for different values of movement. In practice, three different shaft response scenarios typically observed in the О-cell test are categorized as Case A (only shaft shear failure), В (only end-bearing failure) and C (failure is not achieved in either shaft shear or end bearing) (Kalmogo et al. 2019). In general, load tests are terminated when movement is insufficient to mobilize shaft and/or toe resistances. Extrapolation is required in most O-cell tests to define or interpret pile capacity. Hasan et al. (2018) evaluated the influence of extrapolation error on LRFD calibration.

Table 3.8 compares three construction procedures with their assumptions, advantages and disadvantages. Case studies in Seo et al. (2016) suggested that the differences among three methods to construct equivalent head load-movement curves were not significant in terms of pile capacity. However, in terms of head movement, the Osterberg (1995) method results in a significantly stiffer load-movement response than that measured in a conventional head-down load test. Both modified Osterberg and the load- transfer methods were practically accurate enough to estimate the head movement under the service load.

< Prev   CONTENTS   Source   Next >