PILE TESTING 1. INTRODUCTION

University of Sydney Analysis and Design of Pile Foundations, 2007 PILE TESTING 1. INTRODUCTION In relatively recent times, pile testing has been r...
2 downloads 2 Views 832KB Size
University of Sydney Analysis and Design of Pile Foundations, 2007

PILE TESTING 1.

INTRODUCTION

In relatively recent times, pile testing has been revolutionised largely as a consequence of high powered computers. Fifteen to twenty years ago, testing options were restricted to static loading tests, with some costly and slow forms of integrity testing available. Now, a variety of tests are available to estimate or measure pile resistance, together with numerous methods available for quickly and economically testing piles for structural integrity. This paper presents details on the pile testing regime used in Australia today. Attention will be focussed on:

2.



static loading tests



dynamic testing



Statnamic testing



integrity testing

STATIC LOAD TESTING

This test simply involves application of a static load to a pile. Tests are performed for compression, tension and lateral loadings. The load is most commonly applied via a jack acting against a reaction beam that is restrained by an anchorage system (comprising cable anchors or reaction piles), or by jacking up against a mass (kentledge). The load is usually measured by a calibrated hydraulic jack (now “illegal” under AS2159-1995) or a load cell. Pile movements are normally measured by dial gauges acting off simply supported reference beams. Typical static pile testing setups for axial loading are shown in Fig 1.

Compression

Tension Fig 1 Typical static pile test setup

Page 1 of 22

University of Sydney Analysis and Design of Pile Foundations, 2007

Static load testing remains the most reliable form of testing of a pile or pile group. However, a number of sources of error may occur if the test is not carried out in accordance with proper procedures. The sources of error are well documented (Pile Foundation Analysis and Design, Poulos and Davis, Wiley, 1980) and will not be elaborated on here. Suffice to say that the major sources of error are associated with interaction effects between the pile and/or the anchorage system; and also the interaction between the measuring system and the kentledge. It should be noted that these errors may be accounted for by reasonably rigorous analyses. It is most important to note however, that the above errors are usually not of significance when common sense and normal sound testing procedures, as outlined in AS2159-1995, are adhered to. Static loading tests in Australia are usually limited to about 4000kN (compression loading) for small diameter piles. Tests to significantly higher loads have been performed, but those are unusual because of the high costs involved. It is important to note that the overwhelming majority of tests performed in Australia are not performed to determine the “ultimate” capacity of a pile. Most tests are done to prove that a pile will satisfactorily support the design serviceability load plus some measure of overload to ensure the pile has a satisfactory reserve capacity (load factor) above the serviceability load. Acceptance criteria for pile performance is written into the Specification, or simply referred to as a requirement to comply with the criteria imposed of AS2159-1995. The major problems with Static loading tests are the time required to setup and do the test, and the high costs involved. For these reasons, static load testing has reduced dramatically in recent times, in favour of less expensive methods. 3.

DYNAMIC PILE TESTING

Dynamic pile testing was introduced into Australia in 1982, to test large diameter bored piles socketed into rock. The results of those tests, when compared with static loading tests performed on the same piles, were in good agreement (see Fig 2).

CAPWAP Prediction (MN)

WEST G A TE FR EEWA Y: C OR R ELA TION OF C A PWA P PR EDIC TION S WITH STA TIC LOA DI N G TEST R ESU LTS Correlation line 3 +30%

2.5 2

Fig 2 Dynamic & Static Loading Tests on Bored Piles

-30%

1.5 1 0.5 0 0

1

2

3

Static Loading Test Result (MN)

Page 2 of 22

University of Sydney Analysis and Design of Pile Foundations, 2007

By the early 1990's, dynamic pile testing had become the predominant means of pile testing in this country. 3.1

The Test

The purpose of dynamic pile load testing is to ESTIMATE the performance of a pile. In the process, an estimate of the suitability of the pile to perform its design task, is made. It is most important to appreciate that direct measurements of the pile load and associated movements are not made with the dynamic test. The resistance mobilised during the test is a prediction, as is the subsequent load-movement behaviour, of the pile performance. As with static load testing, dynamic tests are seldom done to determine the “ultimate” loads. They are normally done to “prove” pile performance. The equipment used for carrying out dynamic pile loading tests comprises the following: •

At least two sets of strain gauges and accelerometers (bolted at diametrically opposite faces of the pile).



Portable field computer to condition and collect the data, and to store the signals.

The impact for the test is usually provided by a piling hammer. This causes a stress wave to be propagated down the pile, to reflect off the toe. The downwards travelling wave may be partially or completely reflected by irregularities or discontinuities in the pile shaft, and by interaction with the surrounding soil to produce “upward travelling waves”. The field computer receives the measured signals of strains and accelerations, and these are integrated to produce force and velocity results. A number of relationships are used to model the passage of upward and downward travelling waves and it is from these relationships that a prediction of pile performance is made. The predictions are made initially by PDA (Pile Driving Analysis/Analyser) methods and should be confirmed by “signal matching methods”. 3.2

PDA

The PDA produces an instantaneous prediction of the resistance mobilised during the blow, using methods such as the “Case” or “TNO” or “Impedance” methods, the differences between the methods being shown in Fig 3. The most commonly used PDA method is the Case method. The Case method basically assumes a model incorporating a spring and dashpot at the toe of the pile. The shaft performance is not modelled with this method, so it is best suited to piles deriving essentially all resistance from end bearing. The magnitude of the mobilised load can be heavily dependent upon the soil damping factor (the so-called “J value”) adopted by the operator. Mathematically, the expression for mobilised pile resistance is given as follows:

Page 3 of 22

University of Sydney Analysis and Design of Pile Foundations, 2007

Rtotal

=

Rstatic + Rdynamic

& Rdynamic = J. vtoe

where J vtoe

= damping constant = velocity of pile

So it can be seen that the J value may have an important influence on the magnitude of the pile resistance predicted by the Case method. The appropriate J value is often little better than an educated guess and should always be correlated to static loading tests to produce the most reliable results. Under no circumstances should pile testing for prediction of load resistance comprise PDA testing only. A minimum requirement should be to perform more detailed analyses using signal matching techniques (discussed later) and preferably, correlated with well executed static loading tests for greatest reliability. It should be noted that a prediction of the load distribution between the shaft and the pile toe can not be made with PDA methods only.

Fig 3 PDA methods

A typical PDA output, as most frequently viewed by the operator, is shown in Fig 4.

Page 4 of 22

University of Sydney Analysis and Design of Pile Foundations, 2007

Fig 4. Typical PDA Screen 3.3

Signal Matching

Signal matching provides the most reliable means of predicting the performance of a pile tested by dynamic methods. The pile and the soil data are modelled according to the best estimates made by the operator performing the analysis, and a calculation is made using wave equation methods. The calculated signals are displayed on the computer screen along with the measured signals. The operator then performs a number of iterations, varying the input data until a satisfactory match between the measured and calculated signals is obtained. Once a satisfactory match is obtained, a plausible model of the pile-soil system is deemed to be established, and from this, the mobilised static loading can be predicted. An example of a satisfactory signal match is shown in Fig 5.

Fig 5: Results of signal matching analysis Page 5 of 22

University of Sydney Analysis and Design of Pile Foundations, 2007

A further advantage of signal matching methods is that the distribution of the resistance of the pile down the pile shaft and the pile toe is predicted. A further sub-routine of the signal matching process permits a prediction of the static loadmovement performance to be made, and example of which is depicted in Fig 6.

Fig 6 Prediction of Load-Movement Performance 3.4

Monitored Results

The information that is collected and may be displayed during a dynamic test is most impressive, and includes the following: (i)

Force:

The impact forces imparted to the pile plotted against real time for each gauge. Average forces are also shown.

(ii)

Velocity:

The velocity of the pile at the measuring level against time determined from each gauge. Average velocity is also shown.

(iii)

Force and Velocity times Impedance: A graph of the force imparted to the pile, plotted against the product of the velocity times impedance (impedance being defined as the product of the pile modulus and area, divided by the wave speed), plotted against time. An example of a “Force - Velocity” curve is shown in Fig 4 above. The characteristics of the curves may provide valuable information to an experienced operator, including an idea of the relative distribution of the shaft and toe resistance of the pile. The shape of the curves also provides Page 6 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

information on the structural integrity of the pile, as shown in Fig 7. (iv)

Downward wave:

The computed downward travelling stress wave plotted against time.

(v)

Upward wave:

The computed upward travelling stress wave developed by reflections at discontinuities (eg pile toe, mechanical joints, cracks) and the interaction with the surrounding soil.

(vi)

Total static and total dynamic resistance: These are computed from the “force-velocity” curves.

(vii)

Displacement:

The displacement of a pile is shown for the period prior to impact through the testing period, to residual displacement.

(viii) Energy transfer:

The energy transfer from the impact, as calculated at the measuring level, displayed against time.

(ix)

Compression and tension stresses can be continuously monitored for every blow during driving if desired, to provide a check that driving stresses do not become excessive so as to cause possible structural damage to a pile. This is especially important for concrete piles where tensile stresses are easily established during driving through soft soils in particular.

Driving stresses:

All of the above items are stored automatically during the test, and may be displayed during the test, with selected items being presented for reporting purposes.

Sound pile Damaged pile Fig 7: Typical Force and Velocity times Impedance Curves showing damaged pile. 3.5

Comments

The following selected items may assist in clarifying a number of misconceptions relating to the performance and subsequent analyses of dynamic pile loading tests. Page 7 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

3.5.1 Accuracy The accuracy of dynamic testing, when compared with static loading tests performed on the same pile have been reported at numerous venues to be unerringly accurate; often within a few percent, thus demonstrating that exceptional results can be achieved. However, a series of well conducted comparative tests performed during various contests both locally (eg 4th ANZ Conference) and internationally (Brussels, 5th Stress Wave Conference) provide not so glowing results. The results of such a contest, held at the 4th Stress Wave Conference are shown in Fig 8 below.

Fig 8 Results of Prediction Exercise - 4th Stress Wave Conference Fig 8 shows large discrepancies in the pile performance as predicted from dynamic tests and the actual static test results, but it is emphasised that the majority of participants managed to be in reasonable agreement with their predictions. Results such as the above tend to be less convincing than the glowing reports that have been issued by us practitioners in the past, but overall it can be concluded that dynamic pile loading tests can usually predict the static test result within an order of accuracy of around 10% to 25%, which should be regarded as being acceptable for geotechnical work. Also, there appears to be no doubt that the order of accuracy increases for tests performed on preformed piles in comparison to cast insitu piles. Notwithstanding, good results have been reported for all pile types. When the above-mentioned accuracy is put into context with the comparison of the costs of dynamic pile load testing to static pile load testing, the value of dynamic testing should become immediately apparent. 3.5.2 Operator “Error” Contrary to claims made by some practitioners with the introduction of dynamic pile loading tests, the solution/prediction obtained upon completion of the signal matching process is by no means unique. Different operators achieve different results. Fellenius (1988), in a study that Page 8 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

involved 18 participating “CAPWAP” practitioners evaluating the same data for preformed piles, showed standard deviations of between 5% and 14% for the four piles studied. Personal experience of the author suggests that a greater variability can be expected for both preformed and cast insitu piles. It is clear then, that different operators will produce differing answers. There is no one unique answer to any set of monitored data, but generally, experienced operators will produce reasonably similar predictions given the same set of data. 3.5.3 PDA Results vs Signal Matching There appears to be two schools of thought with respect to how many pile should be subjected to analyses involving detailed signal matching on any specific project. Probably the most common practice, nationally and internationally, is to test a proportion (typically 5 to 10%) of piles on a project and a “representative” number of these selected for the more rigorous CAPWAP/TNOWAVE analyses. It should be noted that the more piles tested and subjected to signal matching, the greater the confidence in the piles and hence higher reduction factors can be incorporated into the design or the required proving load. This is clearly highlighted in AS2159-1995. Upon completion of these analyses, the average soil characteristics via the “J” value, are then applied to all other piles tested by PDA methods only, to come up with a more reliable estimate of the mobilised pile capacity than that indicated from raw PDA methods only. The intention of this practice is simply to test a large number of piles; establish an average soil characteristic appropriate for the site/area and to use this value to predict the performance of all piles. The objective of restricting the number of CAPWAP/TNOWAVE analyses is quite simply to reduce costs, ostensibly without compromising technical standards. An alternative practice, particularly promoted by the TNO organisation of The Netherlands, is to subject all piles to the more rigorous signal matching procedure. The philosophy adopted in promoting this practice is to obtain the highest possible degree of confidence in the results, at a relatively low cost. In further support of this argument, it can be stated that once an accurate signal match has been obtained for one pile, relatively little work has to be done to obtain satisfactory matches for other piles tested. Many practitioner argue as to “the correct” practice that should be adopted. It is probably best left to the designer and the testing authority to arrive at an acceptable testing regime best suited for the project, prior to testing and subsequent analyses. Under no circumstances however, should piles be tested by PDA methods only, as significant errors in mobilised pile resistances may result. It is not sufficient to adopt “J” values from the literature as being sacrosanct, for considerable departures from the published values are common. A minimum requirement should be to perform a representative number of signal matches to obtain a representative “J” value. A preferred alternative is to correlate dynamic test results with a well executed static loading test.

Page 9 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

3.5.4 Mobilised vs Ultimate Load One of the major problems with dynamic pile load testing is that hammer blows of insufficient energy might be used to predict/prove the “load capacity” of the pile. Most dynamic tests are used as a means of providing a satisfactory proof load, not to predict the “ultimate load”. As a consequence, sufficient energy is seldom imparted to the pile to produce geotechnical pile failure, nor is it required. Many engineers appear to have difficulty in accepting the concept of “mobilised load”. Put simply, a light tap from a hammer will move the pile toe nil to a negligibly small distance, and will not then mobilise anywhere near the available geotechnical resistance. With a heavy blow, the pile toe can be made to move a greater distance and hence realising a higher load as a consequence of mobilising more end-bearing resistance. In short, if the pile toe can not be made to penetrate the end-bearing layer, then only a proportion of the maximum available pile resistance will be mobilised. It stands to reason then, that if a pile is founded on strong rock and a nil set is registered with an appropriate blow from an appropriate hammer, then the pile will fail structurally before reaching the maximum available geotechnical support. As a general guide, a 6 tonne and 8 tonne hydraulic hammer can be relied upon to mobilise around 3000kN and 4000kN respectively for precast piles commonly used in the Australian piling market. A 20 tonne drop hammer has been used to mobilise almost 30MN! 3.5.5 Time effects The resistance of a driven pile usually exhibits some form of “set up” or increase in capacity, with time. Usually these “set up” effects are the consequences of the dissipation of pore water pressures and can be most dramatic. In clay soils, increases in capacity of 2 to 6 greater than that registered upon completion of driving, have been obtained. Even in sands, increases of up to 70% have been reported. These aspects are most important for dynamic pile load testing. If the magnitude of proof loading is required to prove a maximum load factor, then clearly, dynamic testing should not be performed until an appropriate time delay and the test is referred to as a “restrike” test. Often a “restrike” test is performed 1 to 2 days after completion of driving, but obviously this time can be varied to suit site circumstances. The practice of firstly mobilising a pile by subjecting it to a number of blows prior to restrike testing should not be followed, as this directly conflicts with the objective of performing the restrike test. Testing performed at the completion of driving can usually provide better information relating to the potential end-bearing resistance. Restrike testing, which incorporates set up effects, will provide a better indication of maximum available shaft friction. It is acceptable to assume that the maximum mobilised pile resistance is the sum of the shaft resistance determined from restrike testing, and the end-bearing resistance as obtained from the “end of drive” conditions. Some specifications require that dynamic tests performed at the completion of driving and with a restrike test to be carried out at a nominated time (typically not less than hours) later. The value of such testing may be insufficient for set up effects of any significance to be established. Often an indication of the set up effects can be gauged simply by comparing the traditional “set” at the “end of drive” and at “restrike” testing. Page 10 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

It should be obvious, but nonetheless needs to be stated, that long term effects such as creep can not be assessed by dynamic testing methods. 3.5.6 Driving stresses One of the major advantages offered by dynamic testing is that driving stresses may be continually monitored, hence providing a control measure to ensure that piles are not damaged through over driving. This is true in the overwhelming number of tests performed. However, personal communication between the author and a number of practitioners indicates that most practitioners have reported that tension stresses as indicated by the PDA are not always correct. It has been reported that when driving reinforced concrete piles onto rock, unrealistically high (>10MPA) tensile stresses have been indicated by the PDA. That magnitude of tensile stress, if real, would result in pile damage which would be detected by the shape of the Force-Velocity curve. This phenomenon, whilst not being infrequent, is not what might be termed as being a common occurrence. 3.5.7 Young’s Modulus The PDA program requires the value of Young’s Modulus to be input, from which the computer estimates the Forces in the pile using conventional stress - strain relationships. The value of Young’s Modulus, if over-estimated, will result in higher forces and hence an over-prediction of the pile resistance. Correct values of Young’s Modulus can be deduced during signal matching, where the average values can be determined by conventional stress wave mechanics. 3.5.8 Impact equipment Dynamic pile load testing is a relatively complex task, but as with all technical data, the accuracy of the predictions is very much dependent upon the quality of the data obtained during the tests. Dynamic testing equipment, regardless of the “brand name” used, is of comparable high quality. It is most important then, that the energy delivered to the pile is delivered by a hammer that has a high degree of control. It is not satisfactory to believe that any mass, dropped from a nominated height, will produce good signals. A piling frame, because it can be manoeuvred to reduce force eccentricities as measured by the gauges, is preferred. This will result in higher quality signals, which in turn will provide operators a better chance of making higher quality predictions and help to reduce uncertainties that may otherwise result. 3.5.9 Concluding statements The dynamic pile loading test procedure has been outlined in simplistic terms which hopefully will provide those who have not had direct experience with the test, some insight of the value of the test. Dynamic testing has been a rapidly evolving field that has justifiably established itself as a cost effective test that can produce results of sufficient accuracy to be regarded as a valuable means of testing piles. However, it is the author’s opinion that the test enjoys a greater reputation for precision than can Page 11 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

be reliably produced for the all cases in the field. The test enables estimates of sufficient accuracy to be regarded as an invaluable asset to the construction industry, particularly when the low costs of the test are considered. 4.

PILE INTEGRITY TESTING

As mentioned above, dynamic testing provides an indication of the structural integrity of a pile. As this requires a large hammer to provide the impact, relatively large pile movements result. For this reason the test is referred to as a “high strain” test. The more common means of testing piles for integrity only is termed a “low strain” test and this is elaborated on in this section. The sonic pile integrity test is a non-destructive test that quickly and economically checks the structural integrity of a pile shaft. The test can be used on cast insitu piles and preformed driven piles (concrete, steel, timber). The test does not, and cannot, give any information on the load capacity of the pile. 4.1

The equipment

Testing equipment comprises a field computer, hand-held transducer and a plastic mallet. The equipment is robust and portable (fitting into one briefcase) and requires only one person to carry out the tests. 4.2

Basic principles

The hammer blow induces a stress wave which travels down the pile shaft as a packet of energy, reflects off the toe and is registered by the transducer at the surface as a “toe reflex” (Fig. 6). If the pile material is homogeneous, the wave will travel at a generally constant velocity. The time taken between the hammer blow and the wave to travel down, then up the pile shaft will be: Δt = 2 L / c

where Δt = time (ms) L = pile length (m) c = wave speed (m/s)

The wave speed “c” for say concrete is dependent upon concrete quality as follows: c = ( E/ρ)½

where c = wave speed (m/s) E = Young’s Modulus (N/m2) ρ = density (kg/m3)

It is well known that the density of poor concrete is about the same as that for high quality concrete. However, Young’s Modulus for poor quality concrete is much lower than for high quality concrete, and this is reflected by the wave speed. In many cases the pile length is known with a reasonably high degree of certainty. The operator performing the test then adjusts the wave speed until the pile length and the toe reflex correspond. In this way, the concrete consistency is indirectly indicated.

Page 12 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

Fig. 9. Stress Wave Propagation for Intact Pile The stress wave paths for a pile with a reduced cross section (“neck”) are shown in Fig. 7 below. Here, the stress paths are more complex, with waves being reflected off the necked section eventually meeting waves reflecting off the full cross section. The depth to the neck is calculated by the computer knowing the time for the wave to travel down to, then up from, the neck and also the wave speed.

Fig. 10. Stress Wave propagation for Necked Pile Reflectograms are displayed as velocity versus pile depth. provides a qualitative indication of major pile discontinuities.

The shape of the reflectogram

Reflections of stress waves occur not only at the locations of pile discontinuities, but also at the boundary of soil layers. A soft soil may, for example, produce a reflection similar to a pile neck. It is essential therefore, that the operator be provided with all available information, which includes the following items: •

piling records showing nominal pile depths and geometry



construction details (eg lengths of temporary casing, concrete consumption, pile driving records.



soils details Page 13 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

4.2



structural details (eg concrete strengths, reinforcement details)



any unusual occurrences during construction (eg delays during concreting)

Soil Effects

Shaft friction usually plays a major role in pile performance. It also plays an important role in pile integrity testing. Together with internal damping characteristics of the pile material, shaft friction dampens or decreases the magnitude of the signal. This reduction in amplitude depends on the pile length, pile type, soil type and consistency and the length to diameter ratio of the pile. A practical on length for sonic pile integrity testing is about 50 diameters, although in stronger soils a limit of 30 diameters may be appropriate. 4.3

4.4

4.5

Phenomena Detected •

reflections from the toe (in most instances)



reflections from significant inclusions ( 5 to 10% of pile diameter)



reflections from horizontal cracks



reflections from joints (as for precast concrete piles)



reflections from increases and decreases in cross sections



reflections from changes in soil layers



reflections from changes in material properties

Phenomena NOT Detectable •

gradual increases or decreases in cross section



curved forms



small inclusions of foreign materials



local loss of cover



debris at the toe of the pile



cracks parallel to the pile axis

The Test

The test is performed by pressing the transducer on the pile head and hitting the pile head with a sharp blow from the plastic mallet. The induced stress wave travel down, then up the pile shaft, to be registered on the screen of the field computer. The signals (“reflectograms”) are reflected off discontinuities in the pile, such as necks, cracks, enlargements, etc. Once satisfactory signals have been obtained, they are stored in the internal memory of the computer, from which they can be down loaded onto a PC for further enhancement and reporting at a later stage. The signals are recorded either in the time domain (eg as used by TNO, Pile dynamics, IFCO). Testing authorities using this method are Ground Engineering, Franki and Wagstaff Piling). The alternative method records signals in the frequency domain (eg using equipment developed initially by CEBTP, France), as used by Pile Test International, Vibropile and Integrity Testing. Recording in the time or frequency domain produces similar results, with one method having no Page 14 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

major advantage over the other, although it could be argued that measuring in the frequency domain provides and indication of the pile head stiffness, which is basically the limit of elastic stiffness. Pile head stiffness is not to be confused with static load capacity. The value of Pile Head stiffness is to compare numerical values of different piles. Obviously, piles having significantly lower values require further investigation. The following presentation is oriented towards results produced in the time domain, for the sake of brevity. An experienced operator can, in most cases, provide an immediate on-site interpretation of the test result. The test usually takes between 1 and 5 minutes per pile, so production rates of 100 to 300 piles tested in one day are theoretically achievable in ideal conditions. It is a wise policy to test a number of piles on a project, so that reflectograms which differ from the norm, may be targeted for further investigation. Requirements for testing are simple and oriented towards achievement of high quality signals necessary for interpretation. Firstly, access to the heads of the piles tested should be such as to allow a hammer blow to be delivered, preferably without impediments such as spiral wire or pile cap reinforcing cages. Cracks or voids in concrete under the transducer or hammer impact locations can produce false signals, so the surface of the pile should be trimmed back to sound material and be free of water or other debris. No surface grinding or other special treatment is required. Best results are achieved when tests are carried out as soon as possible after installation. For driven piles this means that piles should preferably tested immediately after driving, when shaft friction will be minimal. Cast insitu piles can not be tested until the concrete or grout is cured to a sufficient degree (about 80% of its ultimate strength), which is normally reached 5 to 14 days after installation. At this stage, results of tests on piles which have been cast into a pile cap or floor slab have usually not been successful. So where possible, access to the head of the pile is desirable. 4.6

Typical Results

The results of a reflectogram for an intact pile is shown in Fig. 11. This cast insitu pile was constructed in soils comprising 6m of soft clays overlying very stiff clays and medium sands. The reflectogram clearly shows the influence of the soil strata. The pile toe is visible with a wave speed of 3900m/s, which is typical of that expected for high quality concrete. No shaft defects are indicated.

Fig 11. Reflectogram of a sound pile

Page 15 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

By contrast, Fig. 12 shows a reflectogram of a 600mm diameter bored pile that did not exhibit the characteristics expected for the pile in the given soil conditions.

Fig 12. Reflectogram of an unsound pile Soil details for the pile depicted in Fig 12 were understood to comprise 2 to 3m of loose or soft soils overlying stiff clays, in turn overlying weathered rock at 7 to 8m depth. Temporary casing was used during construction. During the concreting process, some piles were reported to be making water and concrete was discharged directly into the excavation. The reflectogram indicates a change in impedance at 2.7m, consistent with a change in soil conditions or reduced pile diameter. A further change in impedance occurred below 4m, consistent with a reduced diameter or loss of concrete consistency. The pile toe is visible at 7.5m depth with a wave speed of 2200m/s only, which is not typical of that expected for good quality concrete. Subsequent coring of the pile proved the concrete to be porous. Below 4m depth, clay inclusions of up to 100mm were found down the length of the pile shaft. The total pile length was found to be 7.4m. The pile was rejected and replaced. 4.7

4.8

Advantages of Sonic Pile Integrity Testing •

tests are performed quickly and economically.



an immediate indication of pile integrity may be provided, permitting immediate rectification work to be carried whilst piling equipment is still on site, thus eliminating costly re-mobilisation costs and delays to the project.



no special treatment is required to prepare pile surfaces prior to testing.



Software is now available to assess the influence of defective piles on the pile group, thus providing an indication of the extent of desirable remedial works, if any.

Future Directions

Quantification of pile defects, using computer simulation techniques, are available. A new development is that Coffey Geosciences have undertaken some Research and Development to assess the influence that defective piles may have on the performance of single piles or pile groups. Page 16 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

Present developments in the Netherlands will in future provide for testing of piles which have been cast into a pile cap using measurements from the time domain, which is often possible with existing equipment that measure results in the frequency domain.

5.

STATNAMIC LOAD TESTING

Statnamic is the most recent form of testing introduced into Australia. Statnamic is a relatively new technology, having being jointly developed by the Berminghammer Foundation Equipment (Canada) and the TNO organisation of the Netherlands in the late 80's. 5.1

The test

The principle of the Statnamic test is depicted in Fig 13. The Statnamic test is one in which the downward directed force on the pile is obtained by burning fast-expanding solid fuel in a combustion chamber, resulting in a large pressure acting upward on a reaction mass (Fig. 14).

Statnamic set up Gravel container Gravel Reaction masses Silencer Cylinder Platform Laser sensor Laser Laser beam Piston Load cell Pile to be tested

Fig 13: Fig 13 Principle of Statnamic test

Fig 14 Fig 14 Statnamic set up

The mass is accelerated to 20g, in turn producing an equal and opposite force acting downward on the pile. The load is applied in a linearly increasing manner, followed by a gradual unloading which is achieved by controlled venting of the pressure. The reaction mass, usually rings of concrete or steel, provide the resistance, and needs to be only 5% of the total load to be applied to the pile. During the test, a state-of-the-art load cell and laser sensor, built into the Statnamic device, act in concert with a high speed laptop computer to measure load and pile movement directly, taking up to 4000 readings per second. Although the test appears more exciting and Page 17 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

dramatic than conventional static load testing, the procedure is safe and quiet. Comparative tests on piles subjected to Statnamic testing and conventional static loading tests have provided very good agreement in load-settlement performance. Statnamic devices are now available for routine testing of piles to loads in excess of 3000 tonnes and the technology exists to increase loadings well beyond 3000 tonnes. It is understood that a device that would enable test loading to 6000 tonnes is under consideration for manufacture at present. Tests have been conducted in the UK, USA, Canada, Malaysia, China, Japan, Korea, Indonesia, Germany, Israel, and now, Australia. 5.2

Typical results

Results from the Statnamic tests are immediately visible on computer screen during the test, and stored for subsequent reporting purposes. Results from three tests are shown, as reproduced from the field data, are shown in Fig 15. For the tests shown, the pile “working loads were around 5.2MN, thus all piles were effectively proof loaded to a load factor of about 3.

Fig 15 Statnamic Test results, Quay West Project, Melbourne. 5.3

Advantages of Statnamic testing

The advantages of Statnamic testing include the following: •

the test is quick and easily mobilised.



pile performance is measured cost-effectively



high loading capacity is available



the system is flexible and adaptable eg single piles or pile groups can be tested for Page 18 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

compression loading and also lateral loading characteristics. • $

the test is quasi-static, and does not produce harmful compression and tension stresses that have the potential of damaging a pile. Statnamic test results are in good agreement with static loading test results, particularly for piles founded in stiff soils or rock.



the test can be used not only for testing pile foundations, but also to confirm the bearing capacity of soils or rock suited to pad footings, thus enabling optimisation of footing design.

RELATIVE MERITS OF VARIOUS METHODS OF LOAD TESTING Static load testing remains the definitive testing method, with the development of all of the more “modern” methods resulting in comprehensive comparative testing with static methods required to gain acceptance in the engineering community. However, static load testing is slow and expensive and generally being suitable for relatively low loads. It is for this reason that other methods such as dynamic and Statnamic have been devised and accepted. It may be of use to compare the three main methods of testing, which is discussed below. Fig. 16 summarises the three main methods schematically.

Fig. 16 Major load testing methods

Page 19 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

The resultant phenomena associated with the tests are shown in Fig. 17.

Comparison Load Tests STN

DLT

LOAD

TIME

DEPTH H



u

v

DLT

z

z 

v

z 

u

STN

z

z

z 

v

u

SLT

z

z

z

Fig 17. Phenomena resulting from various test methods. The following observations become apparent: $ the load for Statnamic testing is significantly longer than for dynamic testing, typically being 10 to 15 times longer. For this reason, it is closer to modelling static load performance. $ the stresses resulting from the tests are similar for static and Statnamic testing, whereas stress wave phenomena result from dynamic testing. $ velocity effects in dynamic testing vary down the length of the pile; are relatively constant during Statnamic testing; and non-existent during static testing. $ resultant displacements during dynamic testing vary down the length of the pile, but are relatively constant during static and Statnamic testing In short, the static test is closely simulated by Statnamic testing, especially for piles founded on an end-bearing stratum compared with dynamic testing. Poulos (Pile testing - from the designer’s viewpoint, 2nd Statnamic Seminar, Tokyo, 1998) Page 20 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

tabulated a summary of the capabilities of various forms of load testing, which are reproduced below. Table 1. Summary of capabilities of various pile load tests with respect to the results obtained. Test Type

Ult. Axial Geot. Capacity

Ult. Lateral Geot. Capacity

Loadsettlm’t

Lateral defl’n

Group Effects

Struct. Capacity & Integrity

Special Loadings

Ground Movs.

Static Uninstrumented

3

0

3

0

1

1

1

0

Static Instrumented

3

0

3

0

2

2

2

2

Static Lateral

0

3

0

3

1

2

2

0

Dynamic (PDA)

3

0

2

0

0

3

1

0

Osterberg Cell

3

0

2

0

0

1

1

0

Statnamic Uninstrumented

3

2

2

2

2

1-2

1-2

0

Statnamic Instrumented

2

2

2

2

2

2-3

2

1

Legend: 3 = very suitable; 2 = may be suitable under some circumstances; 1= possible but unlikely to be suitable; 0 = not suitable

Table 2. Summary of Various Pile Load Tests with Respect to the Accuracy and Relevance of the Results Test Type

Pile Loaded in Same Way?

Additional Stress Changes (Side effects)

Accuracy of Movement Measurement

Accuracy of Load Measurement

Similar Duration of Loading to Prototype?

Static Uninstrumented

3

2

2

3

3

Static Instrumented

3

2

2

3

3

Static Lateral

3

2

2

3

3

Dynamic (PDA)

3

2

1

1

1

Osterberg Cell

2

2

2

3

3

Statnamic

3

3

3

3

2

Legend: 3 = good; 2 = may be adequate; 1 = generally not good

This paper has not presented information on Osterberg Cell testing as this is not a test that has Page 21 of 22

University of Sydney Analysis and Design of Pile Foundations, 1999

been used to any great extent in Australia.

Page 22 of 22