Pile Capacity Prediction Research Articles

PREDICTING GEOTECHNICAL AXIAL CAPACITY OF REINFORCED CONCRETE DRIVEN PILE USING MACHINE LEARNING TECHNIQUE

Modified Meyerhof method is a popular method to calculate pile geotechnical axial capacity in Malaysia currently. From past experience, pile design based on empirical and analytical method produce variability of predicted capacity, in which, there is a wide scatter of predicted capacities and tendency for the predictions to be conservative, i.e. to underestimate the load capacity. This study provides options of machine learning and statistical approach for prediction of pile capacity based on soil investigation and dynamic pile load test result. It serves as an additional checking for engineer during design of pile based on conventional empirical method. It also helps to provide deeper insights of non-linear variables related to pile capacity through machine learning and statistical approach. This study helps engineer to design pile foundation optimally, economically and safely. The prediction of pile geotechnical axial capacity with machine learning technique and statistical approach for local marine clay soil in Penang, Malaysia is proposed in this study. The information from soil investigation report and dynamic pile load test report are gathered from six projects at Batu Kawan and Nibong Tebal located in Penang state that contributed 439 numbers of data. The skin friction factor, end bearing factor and pile geotechnical axial capacity are computed and predicted using empirical method, machine learning model and statistical model. Support Vector Machine illustrate best fit model for predicting skin friction factor with R2 of 0.517 while Random Forest seems to be the best fit model for predicting end bearing factor with R2 of 0.264. Random Forest is found to be the best model in predicting the geotechnical pile axial capacity compare to other models as it explains 96.2% of the variability of pile capacity.

Deriving a p-y curve from lateral pile load tests – Case studies in alluvial OC clays

Estimating lateral pile capacity is one of the most typical works for geotechnical engineers. Although it has been a routine calculation, obtaining reliable results is still challenging, especially when the ground condition is unfamiliar. Fortunately, the lateral pile capacity prediction can be performed by deriving a p-y curve from the load-deformation curve measured during pile testing. In cases where a preliminary test pile exists, fortunately, the p-y curve can be generated easily using a subgrade reaction approach suggested by Davisson and Gill (1963). Two lateral pile load test results were used to show the prospect of this method for practical applications. The field tests were on driven and bored piles with different diameters and lengths embedded in alluvial OC clays at two different sites in Jakarta, Indonesia, where both piles were categorised as long piles. Derivation of p-y curves was performed with a simple spreadsheet. Then, the obtained p-y response was used to estimate the load-deformation curve. Results showed that the back-calculated curve agree well with the measured data. This outcome might be because the p-y curves were directly derived from the field test where the pile experienced in-situ stress, resulting in a more representative of the actual condition.

Forecasting the Capacity of Open-Ended Pipe Piles Using Machine Learning

Pile design is an essential component of geotechnical engineering practice, and pipe piles, in particular, are increasingly being used for the support of a variety of infrastructure projects. These piles are being used with dimensions that exceed those used in the development of the most widely used design approaches. At the same time, the growth in pile dimensions calls for the evolution of the state-of-the-art at a similar pace. The objective of this study is to provide an improved prediction of pile capacity. A database of 112 load tests on pipe piles ranging in diameter from 10 to 100 in. (0.25–2.5 m) and in length from 10 to 320 ft. (3–98 m) was employed in this study. First, design capacities were computed using four popular design methods and compared to capacities interpreted from a load test. For the employed dataset, the Revised Lambda method was found to best predict capacities of pipe piles obtained from a load test, among the four examined methods, and was thus employed as a reference standard for assessing the performance of ML methods. Next, eight ML regression models were trained to compute the capacity of pipe piles. Several trained ML models predicted capacities for the testing data set on par with the Revised Lambda method, and three were selected for further investigation. A variety of pile dimensions and soil properties were examined as input properties for ML and the trained models performed surprisingly well with only the pile dimensions used as input. In addition, ML models exhibited satisfactory diameter and length effects, which have been areas of concern for some traditional design approaches. The work thus demonstrates the feasibility of employing machine learning (ML) for determining the capacity of pipe piles. A web application was also developed as a tool for forecasting the capacity of pipe piles using ML.

Prediction of Pile Capacity using Pile Loading Tests and C PT Tests

Prediction of Pile Capacity using Pile Loading Tests and C PT Tests

A framework for the design of vertically loaded piles in spatially variable soil

Each pile in an offshore wind farm may be designed solely on the results of a single cone penetration test (CPT) positioned at its center or based on sample statistics calculated from multiple investigated locations in the same geotechnical unit. This study investigates quantitatively the effects of spatial variability of soil on the efficacy of the two approaches to achieve a reliable design. Random field theory is used to generate sets of cone tip resistance profiles with different characteristics to represent different seabed conditions and piles are ‘designed’ using the above two approaches and the partial factor method to withstand a given vertical drained design load distribution. The effects of the inaccuracies in prediction of pile capacity from the CPT data are quantified in terms of the scatter in calculated reliability (or probability of failure). Outputs of the parametric statistical analysis demonstrate that the ability of the two CPT-based approaches to produce a reliable foundation design is dependent on the ratio between the horizontal scale of fluctuation and the pile diameter, and the overall variability (coefficient of variation) of the seabed properties. Modifications to deterministic, partial factor design approaches are suggested to account for these uncertainties to ensure target reliabilities are achieved.

Pile Shaft Capacity from Cone Penetration Test Records Considering Scale Effects

Abstract The cone penetration test (CPT) is used as a complementary approach for determining the bearing capacity of piles. Via direct CPT methods for pile capacity prediction, correlation coeffici...

A TLBO-optimized artificial neural network for modeling axial capacity of pile foundations

Due to a considerable level of uncertainty describing the pile–soil behavior, many pile capacity prediction methods have focused on correlation with in situ tests. In recent years, artificial neural networks (ANNs) have been applied successfully in many problems in geotechnical engineering, especially, axial pile capacity estimation for driven and drilled shaft piles. Training neural networks is a crucial task that needs effective optimization algorithms. The most popular algorithm is a back-propagation method (BP), which is based on a gradient descent that can trap in local minima. The paper proposes a new artificial neural network (ANN) in which the learning is performed using a recent teaching–learning-based optimization algorithm (TLBO), improving axial capacity predictions. The model is trained and validated on 479 data sets for a wide range of uncemented soils and pile configurations, obtained from the literature. Results show that the considered TLBO-ANN model outperforms other state-of-the-art models in the prediction accuracy and the generalization capability. For instance, we obtained a coefficient of determination $$R^2=0.941$$ and a variance accounted for $${\text{VAF}} = 94.09\%$$ for TLBO-ANN while $$R^2=0.871$$ and $${\text{VAF}} = 87.31\%$$ for the classical BP-ANN. In addition, error investigation with log-normal approaches demonstrates that the probability that predictions fall within a $$\pm \,25\%$$ accuracy level for TLBO-ANN model is 0.93 and that for BP-ANN model is 0.75. The proposed TLBO-ANN model predicts pile capacity with more accuracy, less scatter, and higher reliability.

Prediction of Pile Capacity of Socketed Piles Using Different Approaches

This paper presents the analysis of formerly available pile load test data of different places in Mumbai region to estimate the ultimate load carrying capacity using different empirical methods. Conducting a load test is the best way to analyse the load settlement behaviour of rock-socketed pile. But because of support reaction and time limitation, the test can’t be conducted till failure of pile. The load settlement database of six pile load test are extrapolated using different empirical methods like De Beer’s method (1967), Chin-Kondner Extrapolation method (1970), Vander Veen’s method (1953), Decourt’s method (1999) and Shen method (1980). These methods determine the yield limit load and Decourt’s method is showing satisfactory result as compared to previous case studies. The relationship between ultimate load carrying capacity and different pile diameter can be used for estimation of ultimate capacity of large diameter piles which can’t be tested using field pile load test due to economical constraints. Safe load capacity of piles is calculated according to IS 14593. For further advancement of study, relationship between socket friction and pile diameter can be recognized and comparisons between static and dynamic safe load can be illustrated.

Reliability of design approaches for axially loaded offshore piles and its consequences with respect to the North Sea

In the near future, several offshore wind farms are planned to be built in the North Sea. Therefore, jacket and tripod constructions with mainly axially loaded piles are suitable as support structures. The current design of axial bearing resistance of these piles leads to deviant results regarding the pile resistance when different design methods are adopted. Hence, a strong deviation regarding the required pile length must be addressed. The reliability of a design method can be evaluated based on a model error which describes the quality of the considered design method by comparing measured and predicted pile bearing resistances. However, only few pile load tests are reported with regard to the boundary conditions in the North Sea. This paper presents 6 large-scale axial pile load tests which were incorporated within a new model error approach for the current design methods used for the axial bearing resistance, namely API Main Text method and cone penetration test (CPT)-based design methods, such as simplified ICP-05, offshore UWA-05, Fugro-05 and NGI-05 methods. Based on these new model errors, a reliability-based study towards the safety was conducted by performing a Monte-Carlo simulation. In addition, consequences regarding the deterministic pile design in terms of quality factors were evaluated. It is shown that the current global safety factor (GSF) prescribed and the partial safety factors are only valid for the API Main Text and the offshore UWA-05 design methods; whereas for the simplified ICP-05, Fugro-05 and NGI-05 design methods, an increase in the required embedded pile length and thus in the GSF up to 2.69, 2.95 and 3.27, respectively, should be considered to satisfy the desired safety level according to DIN EN 1990 of β = 3.8. Further, quality factors for each design method on the basis of all reliability-based design results were derived. Hence, evaluation of each design method regarding the reliability of the pile capacity prediction is possible.

DYNAMIC MEASUREMENT TECHNIQUES FOR PILE CAPACITY PREDICTION

Pile Capacity prediction by the Stress-wave matching technique uses measured pile top velocity-or force-time signal as an input boundary conditjon. Reliable estimation of the pile capacity largely depends on availability of accurate and quality dynamic measurements of pile. . This paper discusses some dynamic measurements obtained from two instrumented precast concrete piles. Various aspects of measurement techniques and their shortcomings are also highlighted.

Comparative analysis of methods of pile-bearing capacity evaluation using CPT logs from tropical soils

This study presents the results of a comparative analysis of the performance of eight direct cone penetration test (CPT) methods in predicting the ultimate load-carrying capacity of a 300 mm diameter end-bearing pile, evaluated at defined soil depths using CPT logs obtained for various locations in the south-east and south-south regions of Nigeria. The methods used are the Schmertmann (1978), LCPC (Bustamante & Gianeselli 1982 - Laboratoire Central des Ponts et Chausees), De Ruiter and Beringen (1979), Tumay and Fakhroo (1982), Price and Wardle (1982), Philipponnat (1980), Aoki and De Alencar (1975), and the Penpile methods (Clisby et al 1978). The results of univariate analysis of variance indicated that the interactions between location and treatment (methods of pile capacity prediction), depth and treatment, and location and depth were statistically significant for cohesionless soils, but that the interaction between depth and treatment was not statistically significant for cohesive soils. Also, post-hoc tests (Least square difference and Bonferroni methods) showed that the LCPC and Philipponnat methods are best suited for cohesionless soils, while the LCPC, Tumay and Fakhroo (1982) and De Ruiter and Beringen (1979) methods are recommended for cohesive soils.

Serviceability limit state reliability-based design of augered cast-in-place piles in granular soils

This study proposes a reliability-based design procedure to evaluate the allowable load for augered cast-in-place (ACIP) piles installed in predominately granular soils based on a prescribed level of reliability at the serviceability limit state. The ultimate limit state (ULS) ACIP pile–specific design model proposed in the companion paper is incorporated into a bivariate hyperbolic load–displacement model capable of describing the variability in the load–displacement relationship for a wide range of pile displacements. Following the approach outlined in the companion paper, distributions with truncated lower-bound capacities are incorporated into the reliability analyses. A lumped load-and-resistance factor is calibrated using a suitable performance function and Monte Carlo simulations. The average and conservative 95% lower-bound prediction intervals for the calibrated load-and-resistance factor resulting from the simulations are provided. Although unaccounted for in past studies, the slenderness ratio is shown to have significant influence on foundation reliability. Because of the low uncertainty in the proposed ULS pile capacity prediction model, the use of a truncated distribution has moderate influence on foundation reliability.

Prediction of vertical pile capacity of driven pile in cohesionless soil using artificial intelligence techniques

Piles are used in substructures of different infrastructural constructions. Due to the complex nature of soil, there are different empirical models to predict the bearing capacity of piles. The objective of the present study is to develop prediction models for vertical loaded driven piles in cohesionless soil using a novel artificial intelligence (AI) technique multi-objective genetic programming (MOGP). Two other recent AI techniques, multivariate adaptive regression spline (MARS) and functional network (FN), are also used to compare the efficacy of different AI techniques. The results MOGP, MARS and FN models are compared in terms of different statistical parameters such as correlation coefficient (R), absolute average error, root-mean–square-error, overfitting ratio and P50. A ranking criteria approach has been implemented to assess the performance of the prediction models developed in this study along with other AI and empirical models available in the literature. The predictive model equations based on the MOGP, MARS and FN are also presented.

Characteristic soil strength for axial pile capacity and its estimation with confidence for offshore applications

Many offshore design codes require the characteristic soil strength for axial pile capacity to be estimated with caution or conservatism, i.e. – in statistical terms – it has to be estimated with confidence. This study is concerned with a reliability-based calibration of the necessary minimum confidence for such estimation. The study demonstrates how to estimate characteristic soil strength for axial pile capacity with confidence and provides an approach to such estimation that will render an incentive to obtain more soil strength data and thereby give credit to the geotechnical designer who opts to test more. The study capitalizes on existing load and resistance models for an offshore example pile from the literature and refers to a number of relevant offshore foundation design codes. In order to accomplish the purpose, a number of prerequisites for a successful study are dealt with before the confidence calibration itself is presented.A suitable approach to represent model uncertainty associated with axial pile capacity predictions is presented and implemented in the reliability analyses that are used for the calibrations. The approach implies that the standard deviation of the ratio between true and predicted capacity is represented as a function of the pile length rather than as a constant. This approach is supported both by theory and by data.Based on stochastic models for load and resistance in conjunction with a method for prediction of axial pile capacity, a second-order reliability analysis of the example pile is carried out. With an assumption of perfect knowledge of the soil strength, with characteristic soil strength defined as the mean value and with prescribed load factors for permanent and environmental loads, the necessary requirement for the material factor on the characteristic soil strength is calibrated by tuning the reliability analysis to meet a prescribed target failure probability.With the calibrated material factor kept unchanged, the reliability analysis is repeated, now with the stochastic model for soil strength altered to include statistical uncertainty owing to limited soil data. The value of the characteristic soil strength is adjusted downward until the result of the reliability analysis again meets the prescribed target failure probability. The resulting value of the characteristic soil strength is interpreted as “the conservatively assessed value” to be used in design when soil data are limited and this value is subsequently used to find the corresponding minimum confidence for characteristic value estimation by capitalizing on the properties of the Student’s t distribution. Results are presented for target annual failure probabilities in the range 10−5–10−4.

Relevance vector machine and multivariate adaptive regression spline for modelling ultimate capacity of pile foundation

This study examines the capability of the Relevance Vector Machine (RVM) and Multivariate Adaptive Regression Spline (MARS) for prediction of ultimate capacity of driven piles and drilled shafts. RVM is a sparse method for training generalized linear models, while MARS technique is basically an adaptive piece-wise regression approach. In this paper, pile capacity prediction models are developed based on data obtained from the literature and comprise in-situ pile loading tests and Cone Penetration Test (CPT) results. Equations are derived from the developed RVM and MARS models, and the prediction results are compared with those obtained from available CPT-based methods. Sensitivity has been carried out to determine the effect of each input parameter. This study confirms that the developed RVM and MARS models predict ultimate capacity of driven piles and drilled shafts reasonably well, and outperform the available methods.

Axial Pile Capacity Prediction Obtained from Environmental Friendly Jack-In Piling Test on Clayey Soil

Jack-in piling is environmental friendly system enabling placement of the pile into soil layer with minimum disturbance. It works with very low noise, low vibration, allows piling in confined area, relatively very fast in term of installation rate, better quality control and very fast in mobility. The main issue regarding the pile is bearing capacity; the ability of the pile to withstand axial load without failure. This study attempts to find the correlation between jack-in force and ultimate pile bearing capacity. The result of 5 piling record on clayey layer soil indicates that there is a good correlation between jack-in force and empirical ultimate pile bearing capacity.

Centrifuge modelling of cone penetration tests in layered soils

Penetration problems are important in many areas of geotechnical engineering, such as the prediction of pile capacity and interpretation of in situ test data. The cone penetration test is a proven method for evaluating soil properties, yet relatively little research has been conducted to understand the effect of soil layering on penetrometer readings. This paper focuses on the penetration of a probe within layered soils and investigates the layered soil effects on both penetration resistance and soil deformation. A series of centrifuge tests was performed in layered configurations of silica sand with varying relative density in a 180° axisymmetric model container. The tests allowed for the use of a half-probe for observation of the induced soil deformation through a poly(methyl methacrylate) window as well as a full-probe for measurement of penetration resistance within the central area of the container. The variations of penetration resistance and soil deformation characteristics as they relate to penetration depth, soil density and soil layering are examined. The results of deformation are also compared with previous experimental data to examine the effect of the axisymmetric condition. The effects of soil layering on both resistance and soil deformation are shown to be dependent on the relative properties between soil layers.

SPT 에너지효율 측정 롯드를 이용한 매입말뚝의 선단지지력 예측

국내외에서 널리 사용 중인 표준관입시험(Standard Penetration Test, SPT)은 작업의 용이성, 다양한 지반정수와 강도정수를 결정방법 등의 장점을 지니고 있다. 이러한 SPT 수행시에 발생하는 해머 에너지는 N치의 산정 및 보정에 매우 중요한 영향인자가 된다. SPT시험시 해머 타격에 의하여 롯드로 응력파가 전달되는 현상이 말뚝 항타 시에 발생하는 현상과 동일하다하여 SPT에 PDA장비를 적용하여 관입에너지를 측정하는 방법이 개발되어 사용 중에 있다. 본 연구에서는 SPT 롯드를 말뚝으로 가정하여 SPT 에너지 효율 측정롯드를 SPT시험과 동적콘관입시험에 적용하여 얻어진 힘과 속도의 파형을 매칭기법(CAPWAP, CAse Pile Wave Analysis Program) 분석을 수행하여 매입말뚝에 대한 극한선단지지력을 예측하는 기법을 제안하였다. 또한, SPT 에너지 효율 측정롯드에서 예측된 선단지지력과 실물 말뚝정재하시험을 통하여 산정된 선단지지력을 비교분석하였다. 연구결과를 토대로 SPT나 동적콘관입시험을 통하여 매입말뚝의 지지력을 간편하게 예측할 수 있는 경험식을 제시하였다. The standard penetration test (SPT) has been widely used because of its usability, economy, and many correlations with soil properties among other factors. In SPT, hammer energy is an important factor to evaluate and calibrate N values. To measure hammer energy, an instrumented SPT rod was developed considering that stress waves transferring on rods during SPT driving are the same as stress waves transferring on piles due to pile driving. Using this idea, an instrumented SPT rod with a pile driving analyzer was applied as a pile capacity prediction tool in this study. In order to evaluate this method, SPT and dynamic cone tests with the instrumented SPT rod were conducted and also 2 pile load tests were performed on pre-bored steel pipe piles at the same test site. End bearings were predicted by CAPWAP analysis on force and velocity waves from dynamic cone penetration tests and SPT. Comparing these predicted end bearings with static pile load tests, a new prediction method of the end bearing capacity using the instrumented SPT rod was proposed.

New ultimate pile capacity prediction method based on cone penetration test (CPT)

Our research evaluated current cone penetration test (CPT) pile design methodologies (Schmertmann, Laboratoire Central des Ponts et Chaussées (LCPC), etc.), modified one of these methods (Philipponnat method), and proposed a new one to improve future driven pile designs. This was accomplished by evaluating 14 pile-capacity-design methods based on CPT data. Furthermore, load and resistance factor design (LRFD) resistance factors for each method were calculated using 21 cases from Florida and 28 from Louisiana. The resulting resistance factors were not satisfactory for any of these methods. A new design method — the University of Florida (UF) method — was proposed taking into account cementation, soil type, and historical data. The LRFD resistance factor was also assessed for this new method. The proposed UF method provides better LRFD resistance factors for both Florida and Louisiana soils. It could be a promising alternative for improving pile design.

Reliability assessment of CPTU-based pile capacity predictions in soft clay deposits

This paper focuses on the piezocone (CPTU) test results and the related interpretation procedures to predict ultimate pile load capacities of driven and jacked piles for the Jiangsu soft soil deposits in China. The pile foundation data including pile load tests on twenty six piles from several field sites are compiled and presented. A comparison was made to explain the advantages and fallacies of various CPT based interpretation methods. A modified CPTU-based method has been proposed, which was used to predict the ultimate capacity from effective cone tip resistance and excess pore water pressure. Comparison of the calculated pile capacities to the measured pile capacities showed that the modified CPTU method's predictions showed a better agreement with the ultimate load capacities determined from static loading tests. Also, there was less scatter in this method's predictions than other CPTU methods for deep foundations of soft clay deposits, indicating a higher reliability in the proposed methodology.