# Reliability of Pile Testing in Hong Kong

Subject: Tech & Engineering 16 4495 16 min College

## Introduction

The Pile test is mainly used to test piles using the application of an axial load or force which covers vertical and raking piles which are often analyzed when they are compressed (GIF 1). The term compression is hereby used to define a pile that would resist an axial force that forces it to penetrate the ground. This happens at an angle that subjects the load to several forces which allow the pile to penetrate further into the ground. The vertical raking pile is however subjected to several pressures that force the pile to be extracted from the ground. This force is normally used in tension piles where an axial force is resisted, thereby forcing the pile to be extracted from the ground. Proof coring is also subjected to the above forces, with regards to pile shafts and integrity tests.

The pile test is designed to determine the bearing capacity and the settlement behavior of single piles for several projects, through several testing techniques which ensure the reliability of the test (Zhang 1051). In the determination of the settlement behavior, the pile test is normally used to determine if a pile test settles significantly in a load test. The same procedure is used to establish if a tension pile rises significantly in a pull-out test which is meant to establish that the pile’s load limit is within the safety standards. The loading device used in the pile test is usually a hydraulic cylinder that acts as a yoke and is anchored to the ground by tension bars or several anchors. There should at least be a 2.5-meter difference between the anchors supporting the yoke (Zhang 1051). Comparatively, the same distance can be four times the diameter of the pile to be tested. Sometimes, dead weights act as yokes. Nonetheless, regardless of the nature of the yokes, they should withstand loads of up to 1.1 the weight of the test load (Beim 3).

The test load is normally measured using a hydraulic load cell or an electric load cell. These load cells should accommodate remote reading and meet the standards of first-class accuracy. Pile testing should generally lead to the realization of unique results because displacement is also unique in all types of soils. Here, a special displacement time model is used to come up with a correct assessment of the settlement behavior in pile testing (Zhang 1051). The Cementation Piling and Foundations Limited explain that “Test equipment and an analysis system have been developed and they allow the unique definitive load-settlement characteristics to be established using any reasonable test schedule for any static load pile test” (Cementation Piling and Foundations Limited 1). However, it should be understood that all types of analysis techniques can be used on any size of a pile or any type of pile.

The pile test is also a special test designed to be implemented under strict conditions. For instance, the test requires specialized testing pieces of equipment which force the contractor to ensure the hydraulic jack and load measuring devices are properly mounted on their respective jacks. The test also requires the input of qualified personnel if it is to be effectively carried out and for all safety standards to be met.

The interpretation of the pile test results has been a controversial issue for a long time. For instance, several empirical rules have been developed to govern the relationship between the pile settlement and the load relationships. However, the greatest concerns about the pile test have been evident in the ascertainment of the reliability of the results. For instance, different parts of the world have different types of soils which ultimately affect the test reliability. In as much as there has been controversy regarding the reliability of the test, there has been a corresponding effort among many engineers to come up with sustainable solutions to the above concerns (Cementation Piling and Foundations Limited 2). For instance, one strategy suggested by Chan (cited in Cementation Piling and Foundations Limited 2) is to plot the settlement against the load. Here, the latter part of the information is usually assumed to be the correct assessment of the ultimate capacity to be ascertained by the pile test.

This paper seeks to analyze the reliability of the pile test in Hong Kong, based on three tests. These three tests are the Dynamic Pile Test, CAPWAP Analysis, and Static loading test. The static load test directly measures the pile head displacement and is regarded as the standard measurement for the pile testing method. It is also normally applied to physically applied test loads. The static load test is mainly applied using the compression, lateral, and tension configurations, and the dynamic pile test is mainly applied through a jack that acts against a reaction beam. The reaction beam is normally restrained by an anchorage system or it can be restrained by jacking up against a reaction mass (in the form of cable reactors or reaction piles) (Cementation Piling and Foundations Limited 2). These cable reactors are installed in the ground to provide tension resistance, while the nominated test load is applied according to the appropriate codes. Alternatively, the same procedure can be implemented according to the standards of the project. Under the static load test, the level of load increment is normally sustained in a predetermined period. Alternatively, the rate of pile movement can be sustained at levels that are lower than the nominated value.

The static load test procedure is normally applied to all types of tests which are either undertaken in water or on land (Franki 1). The same test can also be undertaken on production piles or sacrificial piles (of a trial nature) (Cementation Piling and Foundations Limited 2). Trial piles are normally designed to be used in load tests and therefore, they are bound to be designed to the ‘failure’ status. When testing production piles, the situation is normally different because the test is aimed at proving that the piles perform to the desired standards. The test is also designed to affirm that the pile has an overload that meets the desired nominal value.

The dynamic pile test was specifically designed to provide a foundation in civil engineering. Dynamic pile tests have therefore been used on several construction sites in civil engineering and because they are economical and simple to use, they have been widely accepted by many engineers (Cementation Piling and Foundations Limited 2). However, the test demands a high strain testing that allows for a heavy mass to be dropped on a given pile where the velocity and force at the point of impact are measured to determine the bearing capacity, pile integrity, pile stress, and hammer performance. This type of testing is the most frequently used dynamic pile testing method around the globe because it has repeatedly been used to give an overall correct bearing assessment. However, hammer performance and pile stress measurements are determined during the point of pile installation.

In the same context, it should be understood that two types of tests occur in the dynamic pile test. As mentioned in the preceding paragraph, the dynamic pile test is normally done under a high strain. This is the first type of dynamic pile test. However, the dynamic pile test is also done under low strains to measure the velocity and force of the pile test. Here, the pile integrity is determined, and a small hammer is used (as opposed to a heavy hammer) (Cementation Piling and Foundations Limited 12).

The dynamic test is also done using the cross-hole sonic logging which requires that a hole is drilled on a shaft before pouring concrete on it (Morgano 490). Here, pulses are sent from one tube to another. The time the pulse arrives at the receiver tube is an indication of the integrity and quality of the concrete found within the tubes. In several countries, foundation engineers have come up with a list of civil engineering projects where the dynamic pile test is used. The test is therefore aimed at ensuring there are certain building codes and standards observed in the entire construction process. However, the foundation of engineering does not recommend a blanket application of the dynamic pile test because there are specific design aids for quality control that are predetermined by foundation engineers for quality control (Cementation Piling and Foundations Limited 22). As a result, the dynamic pile test has been used by several engineers around the world for quality assurance. Morgano explains that this has been going on for more than two and a half decades now.

The dynamic pile analysis is however synonymous to the Case Pile Wave Analysis Program (CAPWAP) technique because the CAPWAP program is used to estimate the static side and end-bearing components, through an analysis of the force and acceleration information which is recorded at the top of the pile. Alvarez explains,

“The CAPWAP Analysis makes use of field measurements obtained by PDA and wave-equation type analytical method to predict pile performance such as static load capacity, pile-soil load transfer characteristics, soil resistance distribution, soil damping, and quake values. CAPWAP Analysis is carried out on the PDA data after the test is complete” (Alvarez 1).

The results which are usually obtained in this manner are normally compared to the static pile load test.

CAPWAP Works as software that measures the shaft or pile capacities using the input from the pile analyzer (which contains data regarding the velocity and force along the shaft and the toe). Experts note that, whenever the pile driving analyzer test is used, the CAPWAP program should be used at least on one foundation per project because the program is known to complete the dynamic load testing procedure (Alvarez 1). Equally, it can simulate a dynamic load test because it contains improved features that analyze the shaft. Comprehensively, the CAPWAP technique possesses automated features which are adjustable and allow the users to make calculated improvements to the program’s results. Normally, the program is operated in English and it either uses metrics or standard units which are also based on the equation theory and signal matching procedures where the program derives most of its applications.

### Objective

From the understanding of the above three tests (CAPWAP, dynamic pile test, and the static load test), this paper seeks to justify the reliability of the pile test, based on the results of the three tests. In other words, this paper establishes the reliability of the pile test, based on the test results obtained from the three tests mentioned above. An inconsistency in the three tests would establish a lack of reliability of the pile test but consistency in the test results (of the three tests), implies that the pile test is reliable.

## Literature Review

CAPWAP has often been considered a signal matching analysis in the determination of capacity evaluation (from a high strain dynamic pile testing data) and although there are many applications for dynamic pile testing, bearing capacity is considered the main one (Alvarez 3). The process of predicting the correct static capacity in the dynamic pile test has often been shrouded in controversy and equally, it has initiated several discussions and studies on the same. Such studies have been performed since 1980 where several researchers have tried to establish the correlation among several piles tests. For instance, related studies establish that there is a correlation between the results of CAPWAP and static load tests (Alvarez 3). Similar findings also establish that there is a perfect correlation between the results of the two types of tests. However, it is a standard practice in most civil engineering projects to perform signal matching analyses to establish correctly the capacity to the dynamic tests, based on simple guidelines that guarantee a reliable examination of the correlation between the capacity to the dynamic and static load tests.

Alvarez explains that dynamic tests should be performed on driven piles after a specific period which allows for the soil to stabilize. However, the time for the static load test and the dynamic pile test are supposed to be the same, but several researchers have established that the dynamic pile test should be undertaken immediately after the static test (Vulcan 11). Nonetheless, the unforeseen challenges in the project schedule force many project managers to delay the dynamic pile test which results in a lack of realization of the complete setup increase. It has often been affirmed that for the correct measurement of drilled shafts or auger cast files, there should be sufficient time allocated for the concrete to be completely strong and stable. This procedure is often deemed to be crucial because it allows the soil to gain stability after it has been drilled. The drilled soil is also deemed to experience a reasonable net set per blow for its full capacity to be effectively realized because the dynamic setting of the drilled shaft is synonymous with a small set per blow. Here, it is easy to realize a bias on the conservative side because of the nature of the capacity predicted (Vulcan 11).

In two major studies (cited in Vulcan 14), the CAPWAP model was equated to the classical smith model because of their similarities in predicting soil resistance, but the main shortcoming of such a comparison was that the Smith model failed to take into account the fact that, it did not consider extensions of unloading behavior. Most of the findings were derived from studies done on steel pipes, but similar studies which were collaborated by the Federal Highway Administration in Ohio concluded that there was a consistency in the accuracy of the initial tests (Vulcan 14). This conclusion was arrived at after analyzing the accuracy of studies using Timber and concrete piles.

A correlation database of the two tests was later affirmed from the same study after undertaking another study sponsored by the Federal Highway Administration of Ohio. The same findings were also supported by evidence that was amassed from independent dynamic testing organizations. However, the same findings were different from the 1980s findings which used steel as the main material of analysis because only 36 out of the 83 piles used in the initial study was steel. Also, 19 pipes were used in the second study (Vulcan 14). The two studies were later analyzed using the CAPWAP analysis which excluded any human input from the calculations. It was affirmed that the correlation results were extremely commendable and the faith in the capacity analysis was restored. In turn, there was an increased faith in dynamic pile testing as well.

Reports emanating from the six stress wave conferences report a strong correlation between the findings and reliability of the CAPWAP model and the static load test (Cementation Piling and Foundations Limited 12). These findings were established after a dynamic determination of 119 driven piles and 23 cast-in-situ piles (which include drilled pipes and auger cast after enough time was given for the concrete to retain its stability after drilling through the soil) (Cementation Piling and Foundations Limited 2). Many of the project engineers only included numerical data regarding the CAPWAP model and the static load test. Vulcan further explains,

“In all cases, the author’s determination of the static load test result was used. For example, there was a CW prediction of 21,200 kN for the Chin SLT projection of 31,700 kN, even though the maximum applied load for that test was only 20,000 kN (the plotted static test curve was flat at the max 20,000 kN)” (Vulcan 14).

In instances where a different parameter of the static load test was used, the Davisson method was used as a correlation method but almost all the researchers ensured they included the status of the spoil, though most of them failed to include the set per blow. Recent studies have shown an increased interest in the design load and dynamic load tests when high-capacity drilling shafts are used (Cementation Piling and Foundations Limited 12). However, due to the volatility in the pile capacity time changes, accurate capacity changes often include time changes, but unfortunately, many researchers who have undertaken studies on dynamic pile tests and static load tests have included this variant in their studies. The inclusion of time discrepancies in the finding is a crucial procedure because it is equally crucial to include time ratio analysis when establishing the reliability of either test.

Here, the time ratio describes the time of the dynamic test over the time of the static test. A ratio of one to zero is usually recommended here. However, extremely sensitive soils may demand a different ratio. Nonetheless, re-strike tests that include the static load test and the dynamic pile test conclude that it is more difficult to predict the reliability of the dynamic pile test as opposed to the static load test (Holeyman 725). This significant variation in results was often noted after a re-strike was done in extremely short periods of a day or so. However, if the re-strike was done in longer periods of six days (for example), there was a lower variation in the results of the static load test and the dynamic pile test. The same results were also obtained for periods of more than 0.25.

Findings obtained from the 2000 Stresswave conference showed that, if one static load test was correlated with another static load test method, there was a strong correlation of the results between the CAPWAP method and the static load test if a re-strike was done. Holeyman (725) further explains,

“Duzceer (cited in Holeyman 725) compared 12 failure criteria on 24 piles (14 driven and ten drilled). Because there is no universal consensus as to definitely preferred criteria, the average failure load from all Duzceer tests (but ignoring Chin result) was taken as the correct answer” (Holeyman 725).

To further compare the ratio of individual criteria to this “average”, Holeyman elaborates that “there is a wide difference in SLT failure loads for the different criteria and considerable scatter (COV), especially for methods with very low or high average ratios” (Holeyman 725).

However, the failure which is seen from the above analysis has been perceived differently by many researchers. Many of their opinions have bordered on the conservative or liberal fronts. For example, researchers such as Debeer and Housel (cited in Holeyman 726) have been conservative on the above failures but researchers such as Brinch-Hansen and Chin have been liberal on the same issue. The failures defined above have however been categorized into seven methods, falling within a range of 18% off the average range of failure, which is made up of a difference of 9% (plus) or 9% (minus) the average value (Holeyman 727). Research studies done by Davisson (cited in Holeyman 725) show a difference of 5% failure rate below the average value. However, the range of results of the CAPWAP model to standard load test ratio is seen to be less than the 18% failure window explained in the above analysis. This analysis means that CAPWAP is highly reliable and can be equated to the standard load test definition of failure. However, the results projected from the CAPWAP model are deemed to be conservative because it is less than a difference of 5% below the average range of failure, and this figure is lower than the average failure load.

Here, there is a direct correlation between the static load test and the CAPWAP model results because studies done to investigate the correlation between the two tests on a 20-millimeter displacement affirms the same fact. These results show that there is a perfect correlation between the two, but more specifically, it shows that the CAPWAP model is accurate and precise (Holeyman 725). However, the same studies have shown that the CAPWAP model sometimes over-predicts its findings. When we assess the results of the static load test viz-a-viz the results of the CAPWAP model, we can affirm that there is a normal distribution of results between the two tests and few cases exceed the 125% threshold where the static test failure load can be perceived to be arbitrary.

However, the shortcoming of the CAPWAP model is deemed to be within the range of 1%, and therefore, it is not statistically significant (Holeyman 725). Here, this over-prediction range is deemed to be within the safe limits and it cannot cause any problems if its general use is to be analyzed. It is also affirmed that in less than 9% of the instances when the CAPWAP model is used, there is a significant cause for concern over its over-predictability. If the static load test were to be undertaken to test with a larger displacement value, the maximum applied load is also bound to increase considerably. Here, the ratio between the CAPWAP and the static load test is also bound to decrease significantly and many of such cases are perceived to have a substantial setup where there was an early re-strike (Holeyman 725). Such cases are also not strange to an inactivation of the full capacity. Considering the ground fortunes that surround the use of the CAPWAP technique, there has been increased adoption of the CAPWAP model in several countries including Sweden which uses the CAPWAP model in 95% of all its projects (Vulcan 14). This is true because, in the country, most of the pile tests are prefabricated with steel.

In a study done by Svinkin (4) to determine the integrity of the pile test using the dynamic pile testing technique, a pile-driving analyzer using the CAPWAP software was used. The pile driving analyzer was used to store, display and process the output of the experiment. The system and driving criteria were justified using stress wave procedure and the required pieces of equipment were certified by the appropriate engineer. In this study, it was affirmed that the dynamic pile test was reliable because it was standardized across various regions of the world. However, the mode of standardization was varied across various states. For instance, in Australia, the high strain dynamic pile test is standardized by AS 2159-1995; in Germany, it is standardized by the German Society of Geotechniques and in the UK, the Institute of Civil Engineers (Svinkin 4) standardize the same test. There are also different standards set for the low strain dynamic pile testing in most of these countries, but many researchers note that the dynamic pile test has been paralleled in several countries because of the allowable stress and capacity factors which vary across several states as well.

However, it is crucial to note that, the dynamic pile test has received a good commendation from most civil engineers in these countries and it is equally crucial to note that, most countries have used the test for decades. For instance, Brazil has used the dynamic pile test method for more than two decades now and though not much codification has been done in the Latin American country, the test is standardized by the Brazilian Association of technical codes. In addition, it is also crucial to note that, the test has been accepted by most specialists operating in the country. The situation is not any different in Canada where the Canadian Geotechnical Society has endorsed the test as a reliable procedure for ensuring high-quality standards are maintained in static test measurement (Svinkin 4).

China has also not been any different from Canada or Brazil, since it has been active in the use of the dynamic pile test in the last ten years. Currently, there are numerous efforts directed towards popularizing the test in Asian countries. However, though the dynamic pile test has been widely accepted as a reliable test across most countries, some experts note that the test should not be used to evaluate the design capacity in preliminary tests. However, recommendations have been made for the use of the test in measuring first-grade building foundations, second case-building foundations, third-grade building foundations, and supplementary tests. Due to these strict standards and specifications, several engineers have affirmed that the dynamic pile test can be deemed reliable (Svinkin 4). Though many researchers agree that dynamic pile testing increases the overall reliability of the Pile test, there is little consensus among the researchers regarding the ability of the pile test to reduce the safety factor concerns.

In Sweden, recommendations about the dynamic pile test can be traced to a document written by the Royal Swedish Academy of Engineering services, Swedish National Road Administration, Swedish National Rail Administration, and the Swedish Geotechnical Institute, which recommended the use of dynamic pile test, based on the findings of numerous studies done in Sweden and around Europe. Nonetheless, there has been a publication titled Guidance on Dynamic Pile Testing which exposes the pitfalls, limitations, and benefits of the Dynamic pile test by noting that the constituent material of any dynamic pile test should have a large differential module of elasticity (Svinkin 4). The same document also emphasizes a lot of caution to the fact that the results of the dynamic pile test should be carefully interpreted by qualified and specialized engineers for a correct assessment of its reliability. In the same context, they also caution that all methods used to assess the reliability of results of the dynamic pile test have their limitations.

In close reference to the pessimism expressed by these researchers, Svinkin also explains that the dynamic pile test is reliant on hardware and software tools that cannot be used to replace crucial engineering understanding of pile capacity. Svinkin explains,

“HSDPT has been used in practice for years. During the past decade, about a few thousand dynamic testings were made each year around the world. However, the actual accuracy of pile capacity determined by this method and areas of application of HSDPT are unknown” (Svinkin 4).

To back up the above claims, several engineers sharing this school of thought have pointed out that, though the high strain dynamic pile test is synonymous with excellent hardware and software support, there is still a lot of concern, since, the two tools do not have a strong engineering foundation. In this context, Fellenius, a European researcher (cited in Svinkin 5) noted several discrepancies with the high strain dynamic pile test because he explains that “routine pile static loading test with measurements of only the load-movement of the pile head is meaningless” (Svinkin 5). The same researcher also notes that the dynamic pile test is reliant on the pile head load movement, and due to this reason; it shares the same problems noted in static load testing. Such problems include the fact that the long-term behavior of the dynamic pile test can only be understood in the short-term and the fact that, the dynamic pile test provides a direct measurement of data, based on the strain force, but the same information is used for the indirect determination of pile force. The skepticism expressed by these researchers shows that the reliability of the dynamic pile test result is still controversial although the same can also be said of the static load test and the CAPWAP model.

## Works Cited

Alvarez, Camilo. Dynamic Pile Analysis Using CAPWAP and Multiple Sensors. Nd.  Web.

Beim, George. Standardization and Codification of Dynamic Pile Testing. Nd. Web.

Cementation Piling and Foundations Limited. Innovations in Static Load Pile Testing and Analysis. 1992. Web.

Franki, James. Static Load Testing. Sydney: Baulkam Hills, 2011. Print.

GIF. Pile Load Tests. Nd. Web.

Holeyman, Maertens. Results of an International Pile Dynamic Testing Prediction Event. Sao Paulo, Brazil: Balkema, 2000. Print.

Morgano, James. Pile Capacity As A Function Of Time In Clayey And Sandy Soils. Rotterdam: Balkema, 1994. Print.

Svinkin, Mark. Some Uncertainties In High-Strain Dynamic Pile Testing. 1997. Web.

Vulcan, Michael. “Discussion of A possible physical meaning of Case damping in pile dynamics.” J. Canadian Geotech 39.2 (2002): 490-492. Print.

Zhang, Tang. “Reliability of Axially Loaded Driven Pile Groups.” J. Geotech. and Geoenvironmental Engrg 127.12 (2001): 1051-1060. Print.