REVIEW OF DEEP FOUNDATIONS INTEGRITY TESTING METHODS AND CASE

REVIEW OF DEEP FOUNDATIONS INTEGRITY TESTING METHODS AND CASE HISTORIES

Samuel G. Paikowsky1, and Les R. Chernauskas2

ABSTRACT Deep foundations integrity testing is employed for assessing the soundness of in-place constructed elements. The increased use of these members in the New England area over the past 20 years has resulted in an increased demand for quality control testing. A basic review of small and high strain integrity testing techniques used for deep foundations along with local case histories is presented. The issue of the required number of tests is addressed in light of the recent recommendations developed for the Deep Foundations section of the LRFD specifications. Various applications including pile length determination, integrity of pressure injected footings and piles during driving are included. Previous comparative studies are referenced and strengths and weaknesses are highlighted. The integrity tests are shown as a useful and important tool. This is especially true when a match exists between the implemented technique, foundation type, the user expertise and the owners’ expectations. Solid engineering judgment, analysis and decision making enhances the ability to utilize the test results and hence their usefulness and importance. 1.0 INTRODUCTION Integrity testing is the process in which the soundness of the inspected object can be determined. Integrity testing of deep foundations has become common over the past 20 years due to the combination of construction requirements and technological advances. Growth in the use of in-place constructed foundations (e.g. drilled shafts) along with higher design loads and increased legal activities are the main motivation behind the need for integrity testing. Advances in the areas of instrumentation, data acquisition, and signal processing accompanied the increased power of personal computers. These advances enhanced the capabilities and reduced the cost of developing methods for integrity evaluation of foundations. Deep foundations integrity testing mostly applies to foundations constructed from concrete/grout, such as drilled shafts, drilled mini piles, pressure-injected footings, and precast concrete piles. The testing is required for quality control during construction to detect flaws in the pile (e.g. necking, cracking, void, poor quality material, etc.). Such defects are applicable to cast (or injected) in place concrete piles, and to a lesser extent to precast concrete piles. In some cases, the determination of the foundation length is required. Integrity testing can then be performed on any deep foundation type, (including timber and steel piles) with some methods

1 Professor, Geotechnical Engineering Research Laboratory, Department of Civil and Environmental Engineering, University of Massachusetts, Lowell, 1 University Ave., Lowell, MA 01854, and a Principal at GTR. 2 Project Manager, Geosciences Testing and Research, Inc. - GTR, 55 Middlesex St. Suite 225, North- Chelmsford, MA 01863.

Paikowsky and Chernauskas, Review of Deep Foundations Integrity Testing Methods and Case Histories 2003 BSCES-GEO-INSTITUTE DEEP FOUNDATION SEMINAR P1

being capable of determining foundation length even when the foundation is not directly accessible e.g. structure/cap coverage of the pile’s top, (Finno et al., 1995). This paper combines the material presented by Chernauskas and Paikowsky (1999) and Paikowsky et al. (2000), Chernauskas and Paikowsky (2000), and Paikowsky (2003). Determining the integrity of a material can be accomplished by intrusive or non-intrusive methods. Intrusive methods are more conventional, and include drilling, coring, or penetration via preinstalled conduits. These methods can include destructive testing (e.g. on core samples) providing direct information about the condition of the structure under consideration, but may compromise the structural integrity once testing is completed. Non-intrusive testing can provide information about the condition of the structure without altering its structural integrity. Integrity testing by non-intrusive methods is often more cost effective, but requires sophisticated equipment and specialty training to yield meaningful results. This paper reviews the basic theory and application of the more common Non-Destructive Testing (NDT) methods applied to integrity testing of deep foundations. Both intrusive and non-intrusive techniques are described and relevant case histories from the New England area are presented. For additional information and analyses, the reader is directed to publications by Baker et al. (1992), Holeyman (1992), Rausche et al. (1992) and Vyncke and Van Nieuwenburg (1987). 2.0 BACKGROUND Two techniques broadly categorize pile testing: small and high strain testing. High strain testing is aimed at the pile capacity evaluation with the ability to determine its integrity. Small strain testing is aimed at investigating the pile integrity alone and is based on two principles related to the measurement of sound/stress waves by either direct transmission or reflection. Common direct transmission techniques include: (1) crosshole sonic logging, (2) single hole sonic logging, and (3) parallel seismic logging. In these methods a sonic pulse is produced with one transducer (transmitter) and the signal is picked up with another transducer (receiver). The transducers typically consist of a geophone or accelerometer. The methods differ in the location of the transducers and the pulse generation method. Common surface reflection techniques include (1) pulse echo (a.k.a. sonic echo), (2) transient dynamic response (a.k.a. impulse response), and (3) conventional high strain dynamic testing. In these methods reflections of waves generated at the top of the pile are measured. As both generated and reflected signals are measured at the same location, more sophisticated instrumentation (typically accelerometers and strain gages), data acquisition, and signal processing procedures must be employed. The major difference among these techniques relates to whether the generated impact pulse propagates under high strain or low strain conditions. Other common reflection techniques include the use of high frequency, electromagnetic pulses such as X-ray, and radar. These methods are more commonly used for subsurface soil evaluation (e.g. stratification, groundwater, and bedrock) and/or concrete slab mapping (e.g. rebar, voids, thickness and condition determination).


3.0 DIRECT TRANSMISSION TECHNIQUES - Crosshole Sonic Logging (CSL)

3.1 Technique The most common direct transmission integrity testing method is crosshole sonic logging (CSL or sonic coring in Europe). The method is used to evaluate the condition of the concrete within cast in place piles (caissons or drilled shafts) and slurry or diaphragm walls. A piezoelectric transducer is used to generate a signal that propagates as a sound (compression) wave within the concrete and another transducer is used to detect the signal. Each transducer is placed into a vertical PVC or steel tube that has been attached to the reinforcement cage and filled with water prior to the concrete placement. The water acts as a coupling medium between the transducer and the tube. A typical tube arrangement and testing principles are presented in Figure 1. The source and receiver transducers are lowered to the bottom of their respective tubes and placed such that they are in the same horizontal plane. The emitter transducer generates a sonic pulse (on the order of 10 pulses per second), which is detected by the receiver in the adjacent tube. The two transducers are simultaneously raised at a rate of around 1-foot per second until they reach the top of the drilled shaft. Typically this process is repeated for each possible tube pair combination (perimeter and diagonals). Figure 1b shows the six tube combinations that can be tested (logged) using a configuration of 4 tubes within a drilled shaft. Increased shaft diameter calls for a larger number of tubes, which increases, the number of combinations and thereby the resolution of the testing zone.

A - A'

A A'

(a)

(b)

defect

Transmitter ReceiverSignalPath

Figure 1. Typical CSL Testing Set-Up Showing (a) Transmitter and Receiver at

Different Depths, and (b) Plan View of the CSL Tubes with Possible Test Combinations.


In homogeneous, good quality concrete, the stress/sound wave speed, C, is typically around 12,000 to 13,000 feet per second and is related to the modulus, E, and unit weight, γ, as follows:

C =•

E g

(1)γ

If for any reason the condition of the concrete is compromised, the wave speed will be reduced relative to that of the “sound” concrete value. Figure 2 presents a typical sonic signal for which the propagation time between the transducers is measured. The vertical axis is the signal amplitude (microvolts) and the horizontal axis is the time (microseconds). The point where the amplitude begins to rapidly fluctuate indicates the arrival time of the signal to the receiver (a.k.a. threshold time). Since the distance between the two tubes is known, the wave speed of the concrete between the tubes can be evaluated. The signal arrival times can then be plotted with depth to generate a log for the particular tube combination as presented in Figure 3. In addition to the threshold times, the energy of each signal may also be plotted with depth. This information can be used to compare signals of one zone to another where lower energy and/or longer arrival times correspond to a compromised quality and/or defect.

Time (microsec)

Noise

Am

plitu

de (v

olts

)

Threshold

Figure 2. CSL Testing Typical Signal


De p

thDefect

Soft Bottom

Debonding

Energy

Defect

Energy (volt-microsec)

Threshold Time (microsec)

Threshold

Figure 3. Presentation of CSL Test Results in the Form of Threshold Time and Energy vs Depth.

Advantages to this method include the direct assessment of pile integrity and the ability to position the transducers in different elevations to create more signals, allowing the development of a tomographic presentation of the investigated zone. The limitations of the method include detection of defects only when they exist between the tubes. The testing can be performed only on drilled shafts for which access tubes were installed. Also, the method can only be used for drilled shafts, as other deep foundations are usually too small or constructed using different methods that do not lend themselves for accommodating the access tubes. Debonding between the tubes and concrete is common if testing occurs long after the concrete placement. Testing in fresh concrete is also difficult as certain zones may cure at a lower rate, creating difficulties in the interpretation of the threshold time and energy. These zones may therefore be interpreted as poor quality concrete.

3.2 CSL Case History – Equipment and Testing

3.2.1 The PISA CSL/SSL Testing System The PISA (Pile Integrity Sonic Analyzer) is a modular system allowing for adoption,

upgrade and incorporation of additional integrity testing technologies. The current integrity testing options available in the PISA include cross-hole sonic logging (CSL) and single-hole sonic logging (SSL) using CHUM (Cross-hole Ultra Sonic Module) and sonic echo (a.k.a. small strain propagation) using the PET (Pile Echo Tester) module. Additional modules are currently under development.

In addition to its modularity, two advantages of the PISA integrity testing system over other systems include its software and portability. The PISA is A Windows based system and is also compatible with Word 2000. The software is updated periodically to incorporate new developments


and algorithms that make data collection, interpretation, and report preparation easier and efficient. The PISA is lightweight (only 42.3 N (9-1/2 lb)) and self powered, hence can be easily carried around from shaft to shaft or site to site. This feature is also beneficial for air travel. The system can be also used as a standard laptop, saving the cost and space required for an additional personal computer (PC) when using a dedicated CSL testing system.

Figures 4 through 8 present photographs of the PISA system, including computer and sensors. Figure 5 presents the layout of the pile screen, where one can enter the pile information and select the tube orientation/locations. Selection of the desired tube combinations is accomplished by drawing a line between any two tubes. Figure 6 presents the data collection screen, where real-time graphical presentation of the concrete integrity is provided during testing. If a suspect zone is detected in this stage and the tomography option is enabled, the probes are lowered and raised relative to each other around the suspect zone, to further investigate and delineate the area. The signals can be examined and adjusted by manually picking the points or using preset algorithms to automatically determine the first arrival time (FAT) as shown in Figure 7. Figure 8 represents the typical graphical output for time and energy plots.

Figure 4. The Pile Integrity Sonic Analyzer (PISA) System with CHUM and PET Modules

1 The PISA Laptop PC 2 The PISA Module 3 Wheels for Cables and Depth Encoders Mounted on the CSL Access Tubes 4 Transducers and Cables 5 PET Transducer


Figure 5. Layout of the “Pile” Screen Figure 6. Data Collection Screen

Figure 7. First Arrival Signal Identification Screen Figure 8. Typical Graphical Output for Time and Energy Plots


3.2.2 Recent Advances in CSL Testing Following recent technological advances, a new concept in NDT equipment has emerged (Amir and Amir, 1998a,b). The use of generic laptop PC based systems and modular equipment components seem to be taking the place of the older dedicated systems. Naturally, the new concept allows small size, lighter, independent equipment with broader NDT applications. Such equipment has the advantage of employing common operating systems (e.g. MS Windows), conforming to other requirements (i.e. graphics presentations and word processing). In addition, such systems can easily utilize updated algorithms, for example, real time on screen tomographic presentation. The first experience in the USA of such equipment was the PISA (described in the previous section) and a relevant case history is discussed in the next section.

A newer version of that system recently developed and distributed is presented in Figure 9. The USB (Universal Serial Bus) based CHUM is a generic box connected to any PC using the USB port (shown in Figure 9 connected to a tablet PC), hence allowing a flexible yet affordable system. This kind of device will most likely result in a new generation of NDT equipment that is more affordable and better suited for versatile testing demands, advanced analyses, and field application.

Dimensions 250mm x 210mm x 74mm

Figure 9. USB based CHUM Connected to a Tablet PC.

3.2.3 Testing CSL testing using the PISA were carried out by Geosciences Testing and Research, Inc. (GTR) of N. Chelmsford, MA, on over 100 drilled shafts installed for the support of a major roadway interchange in Boston, Massachusetts. The shafts ranged between 2.1m (7 ft) and 2.7 m (9 ft) in diameter and tapered to 1.2 m (4 ft) to 1.5 m (5 ft) in diameter over the lower portion (15 m (50 ft) to 30 m (100 feet)). The total lengths varied between 36 m (120 ft) and 67 m (220 ft) below ground surface. Various penetrations into rock were required depending on the loading conditions. The shafts were constructed using temporary steel casing to the top of the clay and slurry throughout the remainder of the drilling and concrete placement process. The concrete was placed using a tremie process.

Eight schedule 80 PVC CSL access tubes were attached to the reinforcement cage and placed within the shaft prior to the placement of the concrete. The tubes were filled with water prior to placement in the shaft. CSL testing was performed primarily along the four diagonal


tube combinations and four edge tube combinations as shown in Figure 10. Additional testing was performed as needed depending on the results.

CALLEDNORTH

1

2

3

4

5

6

7

8

Figure 10. CSL Tube Layout

The CSL testing indicated significant anomalies in four of the first group of shafts installed for phase 1. Typical time and energy plots for the shafts (designated as "1 through 4") showing the various anomalies, are presented in Figures 11 through 14, respectively. The presented data illustrate two soft bottoms (shafts 1 and 2) and two problems in the upper 12.2 m (40 ft) (shafts 3 and 4). Table 1 summarizes the results of the repeated CSL testing performed for all four shafts. Following the testing and outline of anticipated problematic zones, a coring program was undertaken to verify the identified anomalous zones. For shafts 1 and 2, one 10.16-cm (4-in) core was drilled down the center of each shaft into the underlying bedrock. For shafts 3 and 4, multiple cores were taken between 6 m (20 ft) and 9 m (30 ft) as shown in Figures 15 and 16, respectively. The lateral extent of the suspect zones as identified by the CSL testing is presented in the two figures. The concrete core samples retrieved from the suspect zones in shafts 1 (lower 6.1 m (20 ft)) and 2 (lower 2.4 m (8 ft)) were completely raveled, segregated, and disintegrated suggesting major discontinuity in the lower section of the shaft. The beginning of the defective concrete obtained from the coring coincided closely with the suspect zone identified during the CSL testing. These two shafts had gone through extensive repair procedures to transfer the loads through the defective zone to the underlying rock below.


Table 1 Cross-Hole Sonic Logging Test Summary

Concrete Tested 1Average 2

Shaft Age when Length Tube Stickup Condition 3

Tested (feet) (feet)

14 days 197 to 205 10 to 15

Edge Tube combinations: showed weak signal or loss of signal for a 2 to 3' zone in each profile within lower 20to 30' of shaft (soft bottom).Diagonal Tube combinations: showed frequent weak signal or loss of signal in each profile over lower 20 to 30'of shaft (soft bottom).

20 days 197 to 204 10 to 15 Diagonal Tube combinations: showed slight improvement over lower 20' of shaft, but still indicated soft bottomin each profile.

30 days 199 to 203 10 to 15Diagonal Tube combinations: showed further improvement over lower 20' of shaft, but still indicated softbottom in each profile.

9 days 109 to 116 15 to 20 All tested tube combinations indicate soft bottom conditions (lower 2 to 6 feet) except tube combination 1-3.

30 days 111 to 114 15 to 20 All tested tube combinations indicate soft bottom conditions (lower 6 to 8 feet).

4 days 177 to 178 15 to 20 Anomaly 25 to 35 feet below top of tubes in most profiles.

12 days 175 to 178 15 to 20 Anomaly 25 to 30 feet below top of tubes in northwestern quadrant of shaft.



44 days 133 to 136 15 to 20 Anomaly 35 to 45 feet below top of tubes in northwestern quadrant of shaft.

87 days 126 to 133 15 to 20 Anomaly 35 to 45 feet below top of tubes in northwestern quadrant of shaft.Notes.

Notes:

Shaft 1

Shaft 4

Shaft 3

Shaft 2

1. The tested length represents the average length of tube accessible during CSL testing, including the stickup above the top of the concrete. The tested length may vary for several reasons, including different tube stickups, slack in cables, blockage in tubes, and/or slippage between depth wheel and cable.2. The average stickup was not the same every time, but varied between the numbers indicated for these tests.3. The lengths and depths are referenced from the top of the tubes. Subtract the tube stickup to determine the depth in the drilled shaft.


0 0.2 0.4 0.6 0.8 1Time (ms) and Energy (relative scale)

200

175

150

125

100

75

50

25

0

Dep

th (f

t)

Arrival TimesRelative Energy

Figure 11. Shaft 1 - Time and Energy with Depth

0.0 0.2 0.4 0.6 0.8 1.0Time (ms) and Energy (relative scale)

120

100

80

60

40

20

0

Dep

th (f

t)


Figure 12. Shaft 2 – Time and Energy with Depth

Paikowsky and Chernauskas, Review of Deep Foundations Integrity Testing Techniques and Case Histories 2003 BSCES-GEO-INSTITUTE DEEP FOUNDATION SEMINAR P11


180

160

140

120

100

80

60

40

20

0

Dep

th (f

t)




120

100

80

60

40

20

0

Dep

th (f

t)




CALLEDNORTH

1

2

3

4

5

6

7

8

Legend

Weak to no signal from 30 to 45 feet below tubes

Estimated maximum possible area of suspect zone

Tube stickup 15 to 20'

Good signal

Realistic area of suspect zone(around 18 square feet in area)

2" Core 1 2" Core 2

2" Core 34" Core

Note: Core locations are approximate

CALLEDNORTH

1

2

3

4

5

6

7

8

Legend

Weak to no signal from 25 to 30 feet below tubes Good signal

Estimated maximum possible area of suspect zone

Tube stickup 15 to 20'

Realistic area of suspect zone (around 9 square feet in area)

2" Core 1

2" Core 2

2" Core 3

4" Core

4" Core

Note: Core locations are approximate

Figure 15. CSL Summary - Shaft 3 Figure 16. CSL Summary - Shaft 4


The concrete core samples retrieved from the suspect zones in shafts 3 (upper 1.5 to 3 m (5 to10 ft)) and 4 (upper 6.1 to 9.1 m (20 to 30 ft)) indicated mixed results. Moderate to severe segregation was observed in some of the cores for shaft 3. In one core, complete disintegration was observed between 1.5 to 3 m (5 to 10 ft). Some slight to moderate segregation was observed in the core samples retrieved from shaft 4. Compressive strength, Elastic Modulus, and unit weight testing were performed on selected samples from the cores obtained from these two shafts. The compressive strength testing indicated high variability in the values for both shafts. Between 1.5 to 3 m (5 to 10 ft) below the top of shaft 3, the compressive strength was considerably lower (some values were much less than the required 28 day strength) than the areas above and below. Lower compressive strength values were also observed for shaft 4 specimens results below 6.1 m (20 ft). The concrete specimens tested from this shaft exhibited a wide variability in the compressive strength. These depths correspond to the anomalous zones identified during the PISA CSL testing, suggesting compromised concrete condition. Shafts 3 and 4 were re-evaluated for their structural load carrying ability and subsequently were found to be adequate inspite of the compromised zone. 4.0 ADDITIONAL DIRECT TRANSMISSION TECHNIQUES

4.1 Singlehole Sonic Logging (SSL) Singlehole sonic logging (SSL) is a variation of the direct transmission CSL method in which the source and receiver are placed in the same tube and the signal travels in a vertical direction (refer to Figure 17). The method is limited to defects adjacent to the tube and is usually used only when a drilled shaft requires integrity assessment after construction. Due to high coring costs, typically, a single hole is advanced (often down the middle) to the bottom of the shaft or slightly below the depth where a defect is anticipated. It may also be desirable to perform SSL during CSL testing to isolate the location of a defect at a certain depth (i.e. distinguishing whether the defect identified using CSL is adjacent to the tube or in between the tubes). Brettman and Frank (1996) describe a comparison between CSL and SSL tests.

4.2 Parallel Seismic Logging Parallel seismic logging is another direct transmission integrity testing variation of the CSL test. The method is primarily performed for the assessment of the length of older foundations. Although large voids or bulges can be identified along the deep foundation edge, it is not typically used for identifying defects. Figure 18 presents the procedure in which a boring is drilled in the ground adjacent to the existing deep foundation (usually within 2 to 3 feet of the deep foundation edge). The drilled hole is advanced well beyond the estimated tip elevation to ensure that the entire deep foundation profile can be logged. A capped PVC tube is placed within the drilled hole and surrounded with bentonite slurry/grout that bonds the tube to the edge of the boring.


Figure 1Transmitte

Paikowsky and C2003 BSCES-GE

ReceiverSignal Path

(a) (b)

Depth

Threshold Time

Transmitter

Receiver

SignalPath

Defect

Figure 18. A typical Parallel Seismic Testing Arrangement Showing a) Instrumented Hammer and Receiver at Several

Depths, and b) Threshold Time versus Depth. 7. Typical SSL Test Set-up Showing the r and Receiver Placed at Different Depths.

hernauskas, Review of Deep Foundations Integrity Testing Techniques and Case Histories O-INSTITUTE DEEP FOUNDATION SEMINAR P15

A receiver transducer is placed at the bottom of the water filled tube and pulled upwards in intervals of around 1 or 2 feet. At each depth interval, the foundation top is struck with an instrumented hammer that sends a pulse down the pile and the soil, to be detected by the receiver. A typical profile of the signal arrival time with depth can be logged as shown in Figure 18b. A change in the rate of signal arrival time signifies either a large defect or the end of the pile.

The most attractive feature of this technique is that any deep foundation type can be tested as long as the drilled hole is close to the foundation. Due to the higher cost associated with drilling, the technique is used to identify foundation depth only when other methods (e.g. PEM) fail. 5.0 SURFACE REFLECTION TECHINQUES

5.1 Pulse Echo Method (PEM) The Pulse Echo Method (PEM, also known as the sonic echo method) is a surface reflection integrity testing technique. A high frequency accelerometer is attached to the pile top using a mild bonding agent such as petro wax or petroleum jelly. A lightweight hand held hammer (1 to 3 pounds) is used to strike the pile top and generate a small strain stress wave. The strains associated with the propagating stress wave are in the order of 1µε. Figure 19 presents a typical PEM integrity testing setup. The hammer is usually constructed from plastic to minimize the extraneous high frequencies generated by steel. The accelerometer attached to the pile top measures the acceleration at impact and reflections arriving to the surface from within the pile. This analog acceleration signal is recorded, digitized, and integrated (using a computer) to create a velocity record.

Defect

Figure 19. A Typical PEM Integrity Test Set-up.


The velocity record indicates the speed at which the pile material (at the point of measurement) moves due to the impact and reflected stress waves created by the hammer. A typical velocity record is shown in Figure 20a. This record can be further processed using algorithms that enhance the signal through filtering, shifting, pivoting and magnification. This manipulation allows enhancement of the velocity signal for weak toe reflections, reduce the effect of unwanted noise and drifts, thereby aiding in the interpretation of the pile response.

Vel

ocity

(fee

t/sec

)

Time (millisec)

Forc

e (lb

s)

Time (millisec)(a)

(b)

2L/C 2L/C

Figure 20. Typical PEM a) Velocity and b) Force Records.

Changes in the pile cross-section, concrete density, and/or soil resistance affect the

impedance in the direction of the traveling wave and create reflections of the stress wave that propagate back towards the pile top. These reflected stress waves can return in compression or tension, depending on the type of impedance change. The pile properties that define impedance, Z, are the speed at which the stress wave propagates, C, the elastic modulus, E, and the cross sectional area, A expressed as:

Z = E A

C (2)

•

Figure 21 illustrates the relationship between the variations in the pile impedance, the

traveling wave and the reflections recorded at the surface. A reflected tension wave indicates a decrease in impedance. Conversely, a reflected compression wave indicates an increase in impedance. Combinations of these impedance changes can create complex reflections at the pile top. By inspecting the velocity record for these changes, the approximate location of the impedance change can be determined. At a time of 2L/C, where L is the pile length, the pile toe response can be identified by observing a reflected tension wave due to softer soil at the tip


(signal is in opposite direction of impact pulse analogous to free end conditions) or a reflected compression wave due to denser soil at the pile tip (signal is in same direction as impact pulse analogous to a fixed end condition). These reflections are subsequently discussed for the high strain testing. One of the most difficult tasks in the interpretation of the velocity record is distinguishing between the velocity reflectors due to pile defects (e.g. crack, neck, void, poor and velocity reflectors due to soil resistance. Detailed quantification of defects is difficult, if not impossible, since the interpretation is based on reflected waves and also relies on an assumed wave speed. The most reliable way to use the method is by comparing the response from a large number of piles at the same site. Piles that indicate a response that is different from the majority are further investigated.

Z2

Z1

Z1

Vel

ocit y

(ft/s

ec) Z2Z1 Z1

Time (msec)

Dep

th (f

eet)

Figure 21. Wave Propagation and Reflections versus Time and Depth.

The PEM testing is simple and quick and hence can often be performed on all the piles at

a site. PEM testing can be carried out on various deep foundation types and materials. Under certain conditions, the PEM can be performed on piles that have been covered by a cap or grade beam structure. The small strain PEM technique is generally effective to a depth of 20 to 30 pile diameters, depending on the magnitude and distribution of the frictional soil resistance.

5.2 Transient Dynamic Response (TDR) Method

The Transient Dynamic Response (TDR) method (a.k.a. the impulse response method) is based on the PEM except that an instrumented hammer is used to generate the impact pulse. An accelerometer mounted in the hammer, or a force transducer built in an impulse hammer, allows one to determine the impact force (using the hammer’s mass) in addition to the velocity records obtained in the PEM (see Figure 20). Since a force transducer is not attached to the pile, only the impact force is recorded. The force and velocity records can be converted from the time


domain to the frequency domain using a Fast Fourier Transform (FFT). The ratio of the velocity spectrum over the force spectrum yields the mobility spectrum (V/F in the frequency domain, presented in Figure 22), providing an indication of the pile’s velocity response due to the induced excitation force.

The TDR method allows additional insight compared to the PEM interpretation technique. Certain dominant frequencies can be identified and correlated to pile length and distance to variations either in the pile impedance or in the soil. In addition, the low frequency components (less than 100 Hz) can provide an indication of the dynamic stiffness of the pile. Although low strain methods allow an estimate of the static pile behavior, they cannot accurately determine the pile bearing capacity. In contrast to dynamic measurements during driving or static load test to failure, these methods do not fully mobilize the pile’s resistance.

Frequency (Hz)

Mob

ility

- V

/ F (f

t-se

c /lb

s)

df dfdf

Figure 22. A Mobility Spectrum (V/F versus Frequency) Using Records Obtained by the TDR

Method. 6.0 PEM/TDR Case History I - Pressure Injected Footings Approximately 600 Pressure Injected Footings (PIFs) were installed as part of the foundation system for a large entertainment complex in Worcester, MA. Two PIFs were visually observed to contain poor quality, low strength concrete reduced to a “putty like” consistency near the pile tops. The upper few feet of these PIFs were cut off to remove the material and assess the extent of the defective zone. Ten PIFs, including the visually observed defective piles, were selected for PEM/TDR integrity testing in order to assess the concrete quality in the shafts. The tests were carried out by Geosciences Testing and Research, Inc. (GTR) of N. Chelmsford, MA. The shafts consisted of corrugated metal shells filled with cast in place concrete. Reinforcement steel was installed within the upper 5.5 feet to allow for connection to the piles cap. The subsurface profile in the vicinity of the test area includes 5 to 20 feet of granular fill over dense sand and gravel. The PIF bulbs were formed in this denser stratum.


Figure 23a presents the velocity record with pile length for a ‘sound’ PIF. The signal indicates a decrease in the velocity around 24 feet, signifying a compression wave reflection due to the transformation from the shaft to the bulb, corresponding to an increase in the impedance. The velocity increases sharply at around 26 feet, due to a tension reflection from the bottom of the bulb (where the impedance decreases when transforming from concrete bulb to the surrounding sand). The mobility spectrum for this PIF is presented in Figure 24a where peaks about 256 Hz apart appear between approximately 400 and 1600 Hz. This change in frequency, ∆f, corresponds to a length of around 25 feet, based on the following relationship between time and frequency:

f

t∆

=1 (3)

and time and distance (considering reflection): L = C ⋅ t / 2 (4) The PIF shaft length was reported to be 23.7 feet, which closely agrees with the above-determined length considering the accuracy of the construction method and the testing procedure. Figure 23b presents the velocity record with pile length for a PIF that was found to have a major defect. The velocity increases sharply at around 7 feet due to a discontinuity associated with a large reduction in the impedance. In fact, the low magnitude stress wave could not pass through this defect and the reflections are repeated every 7 feet with the signal dampened with time. Even though the PIF shaft was reported to be 23.7 feet, the length indicated by the test, was around 7 feet, since the defect occupied almost the entire cross section. The mobility spectrum in Figure 24b looks significantly different from that of a “sound” pile presented in Figure 24a. In this case, the change in frequency is around 928 Hz, which corresponds to a length of 7 feet. The soil around the “compromised” PIF was excavated to a depth of 10 feet. The corrugated shell was torched off the shaft around 8 feet below the top of the pile. When the shell was removed, the PIF fell over, due to a complete break in cross section around 7.5 feet. Another PIF evaluated by the PEM/TDR to have a defect around 5 feet was also excavated. After the corrugated shell was removed, a large volume of water and “putty like” concrete fell out of the shell. An approximate 20% reduction in cross section was observed at a depth of 4 to 5 feet below the top of the PIF. As a result of the integrity testing and subsequent verification in the field, a reduced cross sectional area was used to reassess the load carrying ability of the foundations.


Figure 23. PEM Velocity Records versus Time for a) a Sound Pile and b) a Defective Pile (for

PIFs).

Figure 24. TDR Mobility versus Frequency Responses for a) a Sound Pile and b) a Defective

Pile (for PIFs).


7.0 PEM/TDR Case History II – Precast Concrete Driven Piles Damage in driven piles can be detected while monitoring the pile capacity using high strain dynamic measurements (subsequently presented). These tests are traditionally carried out on a small number of piles even when compared to typical concrete pile breakage during installation (say 5 to 7%). Damage during driving or site work following the installation can result in piles with questionable integrity. Figures 25a & b present PEM testing results on 16 inch square (about 90 feet long) concrete piles driven for the support of multistory buildings in Cambridge, Massachusetts. The tests were carried out by Geosciences Testing and Research, Inc. (GTR) of N. Chelmsford, MA. A repetitive increased velocity reflection at times corresponding to a distance of about 15 to 20 feet is presented in Figure 25a. The repetitive reflection indicates damage extensive enough to prevent the signal from propagating any deeper than the indicated depth. No evidence is however provided by the test as to the compression load bearing capability of the pile. Figure 25b presents the results obtained from a nearby sound pile for which the propagating signal responded to the variation in the soil type (sand layer at about 25 to 30 feet and a till layer at about 60 to 70 feet). The tip response was magnified due to the small energy used in the PET testing. As a result, the techniques’ effectiveness at such depths is questionable. The usefulness of the method in time and cost savings was certainly a big advantage, allowing the identification of a large number of defective piles in a short period of time. The limitations regarding the nature of the damage and the structural ramifications need to be recognized as well.

Figure 25a. PEM Velocity Records versus Time for a) a

Defective Pile


Figure 25b. PEM Velocity Records versus Time for b) a Sound Pile (Precast Concrete

Driven Piles).

8.0 HIGH STRAIN INTEGRITY TESTING DURING PILE DRIVING OR DROP

WEIGHT TESTING

8.1 Technique Dynamic pile testing is commonly employed for evaluating the drivability and capacity of driven piles. The same method is also used to assess the capacity of cast in place shafts. When a ram strikes the pile head, it initiates a large strain wave that propagates down the pile as illustrated in Figure 26. External soil resistance or changes in the pile’s impedance (due to variations in the pile’s material or geometry) causes reflection waves that are recorded at the surface in what is in principle a surface reflection technique similarly to PEM/TDR methods using low strain waves. Typical dynamic pile testing instrumentation consists of two accelerometers and two strain transducers attached on opposite sides of the pile close to the pile top. Knowing the material properties and pile geometry at the point of measurement, the strain is converted to force, while the acceleration is integrated with time to produce a velocity record. These force and velocity records can be used to evaluate the pile’s integrity. As long as there is no change in the pile impedance or external forces (friction) are not activated, the force and velocity remain proportional. Reflections from the tip can be reviewed in light of two classical boundary conditions (see for example Timosheko and Goodyear, 1934). Free end (analogous to easy driving through soft clay) calls for zero stress and no velocity restrictions at the tip, resulting in a compression wave returning as a tension wave and velocity increase (theoretically doubling). Figure 27 presents such reflection from a 48-inch diameter pipe pile driven offshore during initial penetration of about 3 feet. The downward velocity and compression stress returned from the tip as a tension wave and an increased downward velocity. Fixed end conditions (analogous to hard driving against bedrock) calls for zero velocity and no stress restrictions at the tip, resulting in a compression wave being reflected with a greater magnitude than the incident wave and the tip velocity at about zero.


Defect

AccelerometerStrain Gage

RAM

Figure 26. A Typical Dynamic Test Set-up.

returning reflection wavestraveling impact (incident) waves

Figure 27. Measured Force and Velocity (Times the Pile Impedance) at the Pile Top versus

Time during Initial Penetration (Paikowsky, 1982).

If a pile contains a defect or is damaged during driving, the wave reflecting from the zone of decreased impedance is comparable to “free end conditions”. These reflections would arrive to the measuring transducers before the reflections associated with the pile’s tip as the damaged zone is at a point along the pile between the top and the tip. The detection of damage during driving is routine and usually is associated with tension cracking of concrete piles. Other structural damage (e.g. splice breakage) can also be identified as presented in the case history


below. The advantage of the high stress wave propagation testing over the small strain integrity testing is its ability to quantify the structural significance of the discontinuity. While a small strain wave would indicate a complete discontinuity for any size crack across the pile, the high strain stress wave would pass through these discontinuities enabling the transformation of compression forces, therefore indicating the adequacy of the structural member.

8.2 Case History – Driven Pile Several hundred H piles were installed for the support of an elevated walkway in the Boston area. Dynamic pile testing was specified for capacity monitoring and the driving operation progressed routinely. One of the inspected piles exhibited clear damage progression during driving. Figures 28a and 28b present the force and velocity (multiplied by the pile’s impedance) signals at the pile top shortly before and after damage detection. Since the early damage identification was dismissed, driving continued and the dynamic records for the subsequent blows are presented in figures 28c, d, and e. A clear velocity increase accompanied by a force decrease attests to the development of the damage. The records of Figure 28e suggest that the pile essentially “ends” at mid-point, indicating a complete detachment between the upper and lower pile sections. The identified damage is associated with a full penetration weld splice that apparently disintegrated during driving. When the pile was pulled out of the ground only the upper section was extracted with severe deformations at the weld connection.

8.3 Case History – CFA Shafts Paikowsky et al. (2004) report on drop weight testing of Continuous Flight Auger (CFA)

and Bentonite slurry construction shafts. The shafts were tested as pre-construction evaluation of possible foundation solutions for a large multi-story expansion of a pharmaceutical manufacturing complex in Haifa Bay, Israel. The three CFA shafts were 70cm in diameter and 25.0m long. The tests were conducted using a modular drop weight system shown in Figure 29. When analyzing the dynamic test results using the signal matching technique (CAPWAP) and the reported shaft dimensions, it became apparent that the lower parts of all three shafts exhibit substantially reduced impedance as depicted by the impedance distribution (solid line) in Figure 30. Often in drop weight testing, the produced stress wave is either not sharp enough or the energy is not high enough to mobilize the shaft’s tip, and hence its clear detection (as in the presented tests). As such, the reported constructed length of 25.0m was used in the analysis. The results depicted in Figure 30 can be interpreted as either the quality of the shaft in the lower 5m was compromised, (compared with an expected impedance of 3,765kN/m/s) or the shaft was not constructed to the planned depth of 25m. Further analysis was carried out for possibly shorter piles resulting with the impedance distributions depicted by dashed lines in Figure 30. These distributions reflect shaft lengths varying between 22.0 to 23.0m and suggest reasonably well-constructed shafts, but shorter than anticipated. This result was later confirmed to be the correct answer.


Figure 28. Force and Velocity Records Obtained during the Driving of a Steel H-Pile a) Shortly

before Detecting Damage, b) Showing Initial Damage, c) as the Damage Develops, d) as the Damage Progressed and e) Upon Complete Discontinuity.


STRIKER PLATE

MODULARMASSES

TRIPPINGMECHANISM

ADJUSTABLEDROP HEIGHT

LIFTINGDEVICE

INSTRUMENTATION

PILE ENCLOSER

GROUND SURFACE

CONNECTINGRODS

PILE

PILECUSHION

GUIDES

PDA

TO CRANE

(a) schematic

Figure 29. Israeli Drop Weight System (after GTR, 1997).

25

20

15

10

5

0

Dep

th B

elow

Gra

de (m

)

25

20

15

10

5

0

25

20

15

10

5

0

0 2 4 6 8 10 0 2 4 6 8 10

Impedance (MN/m/s)0 2 4 6 8 10

C1 C2

C3

Figure 30. Variation of Shaft Impedance with Depth for the CFA (C1 – C3) Constructed Shafts


9.0 NUMBER OF TESTS A study aimed at the development of new AASHTO Load and Resistance Factor Design (LRFD) specifications for deep foundations was recently completed (Paikowsky, 2002). One of the issues addressed in the study was the number of tests recommended/required either for dynamic tests on driven piles, or small strain tests for drilled shafts. The two have a different logic that supports the tests and, hence, lead to a different probabilistic approach in their solution. While the first looks for the number of tested piles required to verify capacity evaluation, the second looks for the number of tests to guarantee a limited tolerable number of defected shafts. The statistical approach for both is presented by Paikowsky (2002), and was carried out in collaboration with Professor Gregory Baecher (see Paikowsky, 2002). For the drilled shafts the analysis requires to define the maximum fraction of shafts with a major defect that the owner is willing to tolerate in a large set of shafts. A major defect was defined as any defect that significantly compromises the ability of the shaft to carry the assigned loads. A sample calculation was carried out assigning this value for 5% along with: (1) the “owner’s” risk of incorrectly accepting a defected shaft as a solid shaft to be 10%, and (2) the “contractor’s” risk of rejecting a set of solid shafts as defected being 10%. This calculation suggested that for 100 shafts in a sample, one needs to test 80 shafts in order to comply by the above requirements, which are not seen as stringent requirements. It was concluded, therefore, that in order to statistically assure very low rates of major defects within a set of drilled shafts, a very high proportion of the shafts must be tested. Thus, it became reasonable to require 100% of drilled shafts to be post construction tested for major defects. Reasonably such tests can be a combination of direct transmission (e.g. CSL) for a limited number of shafts, e.g. during initial stage of construction and at several locations based on site and/or construction variations, complemented by the more cost effective surface reflection tests (e.g. PEM) on all other shafts. 10.0 DISCUSSION A large variety of non-destructive, intrusive and non-intrusive deep foundation integrity testing methods and case histories are presented. The methods’ strengths and limitations are related to their effectiveness, time (in preparation, testing and interpretation) and associated cost. In general, the direct transmission methods necessitate considerable preparation and can provide higher accuracy in the zone bounded by the penetrating sleeves. Surface reflection techniques require only minimal preparation but are limited in their zone of meaningful operation and accuracy. The selected testing method needs to reflect the anticipated result and the associated line of action. It is with this approach in mind that one requires reviewing comparative studies of known embedded defects (e.g. Baker et al., 1992 and Smits, 1996).

The ability of a method to detect a certain defect should be examined in light of the defect’s influence on the foundation serviceability. This course leads to the selection of an integrity testing method based on the expected outcome. For example, the possible detailed data provided by the direct transmission methods should not result in a caisson rejection just because certain zones suggest a lower quality of concrete. Such decisions need to be associated with the design loads and the load bearing assessment of the tested caisson. The surface reflection methods on the other hand allow extensive testing with the expectations that detailed


investigations are carried out on the suspected caissons only. Choices, therefore, should be made regarding quantity and quality. Testing of many piles with the ability to detect major defects (where possibly undetected defects are not expected to compromise the pile’s load carrying ability) vs. detailed studies of a smaller number of piles or a combination of the two methods can be performed. Statistical evaluation of risk associated with major defects supports the logic of testing each and every cast-in-place deep foundation. 11.0 CONCLUSIONS The reviewed methods and the presented case histories demonstrate that deep foundations integrity testing is useful and has significant importance. The PEM/TDR and CSL techniques are routinely used to assess the quality and condition of cast in place and driven piles. Conventional dynamic testing is effective in evaluating pile integrity during driving or striking if in place constructed deep foundations. In some cases the results of the integrity testing were used to reject the piles where in other cases they were used to re-evaluate or redesign the piles. Frequently, the integrity testing is used to confirm anticipated defects in the piles. When using an adequate testing method along with engineering judgment, integrity testing of deep foundations can be employed as an important tool with sound economical justification. ACKNOWLEDGMENT Geosciences Testing and Research, Inc. (GTR) of N. Chelmsford, MA carried out and interpreted all the described tests. The cooperation of the contractors, consultants and owners (in particular MHD) associated with the described projects is appreciated. The case history describing the CSL testing method was carried out using the PISA, Pile Integrity Sonic Analyzer manufactured by Pile Test Com Ltd. Israel. The case histories described in the section related to the high strain integrity testing was carried out using the PAK 586 pile driving analyzer manufactured by Pile Dynamics Inc. of Cleveland, Ohio. The material presented in this paper was compiled mostly from the following publications: Paikowsky and Chernauskas (1999, 2000), Paikowsky (2002), and Paikowsky et al. (2004). The authors acknowledge the assistance of Mary Canniff in putting together the manuscript. REFERENCES Amir, E.I and Amir, J.M., (1998a), “Testing Piles with Virtual Instruments”, Proceedings of the

DFI conference, Vienna, Austria. Amir, E.I. and Amir, J.M., (1998b) “Recent Advances in Ultrasonic Pile Testing” 3rd

International Geotechnical Seminar on Bored and Augured piles, Ghent, Belgium. Baker, C.N., Drumright, E.E., Briaud, J., Mensah-Dwumah, F. and Parikh, G., (1992), “Drilled

Shafts for Bridge Foundations”, Final Report Under Contract No. DTFH61-88-Z-00040, Federal Highway Administration, February.

Brettman, T. and Frank, M., (1996), “Comparison of Crosshole and Single Hole Sonic Integrity Logging Methods”, Proceedings of the 5th International Conference on the Application of Stress Wave Theory to Piles, Orlando, Florida, September, pp. 698-707.


Chernauskas, L.R. and Paikowsky, S.G., "Defect Detection and Examination of Large Drilled Shafts Using a New Cross-Hole Sonic Logging System", Proceedings of ASCE Specialty Conference, "Performance Verification of Constructed Geotechnical Facilities", Luteneger A.J. and DeGroot D.J. edt, April 9 - 12, 2000, University of Massachusetts, Amherst, ASCE Geotechnical Special Publication No. 94, pp.66-83.

Chernauskas, L.R. and Paikowsky, S.G., “Deep Foundation Integrity Testing Techniques and Case Histories”, Civil Engineering Practice, Journal of the BSCE section/ASCE, Spring/Summer1999, pp. 39-56.

Finno, R., Prommer, P. and Gassman, S., (1995), “Non-Destructive Evaluation of Existing Deep Foundations”, Proceedings of US/Taiwan Geotechnical Engineering Collaboration Workshop, Taipei, January, pp. 125-135.

Holeyman, A.E. (1992), “Technology of Pile Dynamic Testing”, Proceedings of the 4th International Conference on the Application of Stress Wave Theory to Piles, Balkema, Rotterdam, September, pp. 195-215.

Paikowsky, S.G., (1982), “Use of Dynamic Measurements to Predict Pile Capacity Under Local Conditions”, Master’s Thesis, Dept. of Civil Engineering, Technion-Israel Institute of Technology, July.

Paikowsky, S.G. with contributions by Birgission G., McVay M., Nguyen T., Kuo C., Baecher G., Ayyub B., Stenerson K., O'Mally K., Chernauskas L., and O'Neill M., “Load and Resistance Factor Design (LRFD) for Deep Foundations", Final Research Report submitted to the National Cooperative Highway Research Program on Project NCHRP 24-17, December 2002, pp. 134 (not including Appendices).

Paikowsky, S.G., Chernauskas, L.R., Hart, L.J., Ealy, C.D., and DiMillio, A.F., "Evaluation of a New Cross-Hole Sonic Logging System for Integrity Examination of Drilled Shafts", Proceeding of the Sixth International Conference on the Application of Stress-Wave Theory to Piles, Niyama S. and Beim J. edt., September 11- 13, 2000, São Paulo, BRAZIL, pp.223-230.

Paikowsky, S.G., Klar, I., and Chernauskas, L.R., (2004), “Performance Evaluation of CFA Vs. Bentonite Slurry Drilled Shafts Utilizing Drop Weight Testing”, to be published at the Geo-Support 2004 Conference in Orlando, FL, February 4-7.

Piletest.com website for USB-based CHUM system information and photo. Rausche, F., Likins, G. and Ren King, S., (1992), “Pile Integrity Testing and Analysis”

Application of Stresswave Theory to Piles, Balkema, Rotterdam, September, pp. 613-617. Smits, M.Th.J.H., (1996), “Pile Integrity Tests”, Proceedings of the 4th International Conference

on the Application of Stress-Wave Theory to Piles the Hague/Netherlands, September 21-24, 1992, Chapter 3 of the Test Results volume, A.A. Balkema/RotterDam/Brookfield 1996.

Timoshenko and Goodyear, J.W. (1934) “Theory of Elasticity”, McGraw Hill Test Consult, (1998) “Parallel Seismic Test Fact Sheet”, Rosky, Warrington, UK, March 3, pp. 1-2. Vyncke, J and VanNieuwenburg, D, (1987) “Theory of the Dynamic Tests”. Proceeding of the

Conference on “Pile Dynamic Testing: Integrity and Bearing Capacity”. The Int. Society of Soil Mechanics, Foundation Engineering, Brussels, Belgium, 1987, pp. II-1 to II-103.


REVIEW OF DEEP FOUNDATIONS INTEGRITY TESTING METHODS AND CASE

Documents

Transcript of REVIEW OF DEEP FOUNDATIONS INTEGRITY TESTING METHODS AND CASE