Download paper, available by permission of the publisher.
Reference: Zekkos, D.P., Bray, J.D., Der Kiureghian, A. (2004), “Reliability of shallow foundation design using the standard penetration test”, 2nd International Conference on Site Characterization Proceedings, 19-22/9/2004, Porto, Portugal, Vol.2, pp. 1575-1582.
The sources of uncertainty, variability and bias in the performance of the Standard Penetration Test were identified using an extensive literature review and were classified in a manner that should assist the practice in recognizing and reducing them. These uncertainties and additional uncertainties introduced in the design process are then incorporated in a reliability-based evaluation of the design of shallow foundations. The results of the reliability analyses show that the factor of safety approach can provide an impression of degree of conservatism that is often unrealistic. At times,foundations with smaller factors of safety have smaller probabilities of failure than foundations with higher factors of safety.The reliability-based approach provides rational design criteria,accounting for all key sources of uncertainty in the foundation engineering process, and thus should be the basis of design.
An extensive literature review was made to evaluate sources of uncertainty in the performance of the SPT, based on the previous work of many researchers (Kulhawy and Mayne (1990), Schmertmann (1975), Barton (1990), ASTM D6066-96, Youd et al. (2001), Kulhawy and Trautmann (1996)).
Table 1 summarizes a total of 27 sources of uncertainty and bias, classified in five categories. This classification aims to assist the practicing engineer in taking the necessary measures to minimize the uncertainties in the performance of the test:
- Category A includes sources that depend on the type of soil, with emphasis in sources related to granular materials since the use of the SPT is generally not recommended for cohesive soils.
- Category B lists sources of uncertainty due to the presence of water.
- Category C includes reducible sources related to the equipment and its maintenance.
- Category D includes sources that are reducible if the test procedure is performed care-fully.
- Category E lists irreducible sources in the investigation procedure.
Figure: Note that for the same factor of safety you get different Probability of Failures, depending on the type of soil.
- The SPT can provide useful and reliable data with good maintenance of the equipment and quality control in the performance of the test. It can also provide very unreliable data. Efforts should focus on making the test more reliable.
- The classification of uncertainties shown in the paper aims to provide a framework for the practicing engineer to recognize the uncertainties of the SPT and take measures to minimize them. There are not sufficient data to adequately quantify most of the sources of uncertainty.
- Design philosophy for the bearing capacity criterion shows inconsistencies that affect the system reliability for small foundation widths. The design philosophy should not only take into consideration the theoretical basis of the mathematical model but also consider the variability and the uncertainties involved in the data that are used as input to this model i.e. the quality of the data. Ignoring the uncertainty in the input can result in unconservative design or at best at irrational design procedure.
- The current practice in shallow foundation design using factors of safety provides the engineer with a false impression of a certain degree of safety in the design. Foundations designed with a factor of safety of 2 may be safer than foundations designed with a factor of safety of 3 in different ground conditions. The reliability-based approach provides a more accurate assessment of the degree of conservatism and rational design criteria. The profession should specify the socially and economically acceptable probabilities of failure in the design.