Recoendations on Two Acceleration Measureents with Low Strain Integrity Test Liqun Liang, PhD., P.E., 1 Scott Webster, P.E., 2 and Marty Bixler, P.E. 3 1 Pile Dynaics Inc., 3725 Aurora Rd, Cleveland, Ohio 139; e-ail: lliang@pile.co 2 GRL Engineers Inc., 3725 Aurora Rd, Cleveland, Ohio 139; e-ail: SWebster@grlengineers.co 3 GRL Engineers Inc., 3725 Aurora Rd, Cleveland, Ohio 139; e-ail: MBixler@grlengineers.co ABSTRACT As ore existing structures need to be reevaluated, repaired or expanded, the deand of use of a Low Strain Integrity Testing (LST) to deterine/corroborate the existing pile/shaft length and integrity under existing structures has also increased. When applying LST to the piles under existing structures, ultiple downward travelling waves will superpose with the upward traveling waves (reflection fro defects and the toe). This akes the data interpretation very difficult when only one acceleration easureent is obtained. Two acceleration easureents were introduced about twenty years ago as an extension to LST and significantly iprove the reliability of application of LST to piles under existing structures. However, due to the coplexity of testing conditions, lack of clear guidelines and experience, this technique has not been frequently applied. After discussing the background of this ethod, this paper presents case studies and recoendations on the application of this technique. INTRODUCTION Various nondestructive testing (NDT) inspection ethods are available to evaluate the structural integrity of cast-in-place shafts. Widely used dynaic ethods include Cross Hole Sonic Logging Testing (CSL), High Strain (HST) and Low Strain Integrity Testing (LST) (Likins, and Rausche, 2, Rausche et al, 2, Webster et al, 211, Liang and Rausche, 211). Also a ethod based on teperature easureents during concrete curing enjoys increased acceptance due to its speed and ore conclusive results (Mullins, 2). However, due to its low cost, easy operation, quickness and flexibility (no pre-preparation such as the installation of inspection tubes), LST is often specified in any different countries for quality assurance. LST is also used to deterine the integrity of prefored piles such as concrete or tiber piles. As there is an increasing use of existing piles such as for cellular towers, reconfiguring, repairing or expanding existing structures, the deand for the evaluation of the integrity and length of piles under existing structures has significantly increased. Both LST and parallel seisic ethods are used
for these purposes. However, the parallel seisic ethod is liited to evaluate pile length only, and is not further discussed here. The LST ethod, in use since the 197s, is standardized by ASTM 52. Traditionally, it priarily relies on the easureent of the pile top acceleration as the result of a light haer ipact. The easured accelerations are usually converted to velocity for interpretation in the tie doain. Stress wave reflections fro increases or decreases in shape or aterial quality (i.e., ipedance change) along the pile are registered by the pile top easureents and then interpreted by the test engineer. If, for exaple, the cross section sharply decreases at a certain distance below the pile top, then at a tie which depends on the distance of that reduction fro the top, a positive velocity change (in the sae direction as the ipact pulse) will be registered. The LST can be applied to piles under existing structures where the pile top is incorporated in the superstructure and the pile length is unknown. However, two coplications need to be addressed for proper evaluation of the velocity records: 1) The wave speed of pile aterial such as concrete or tiber is not always known, which is an iportant variable affecting the accuracy of pile length deterination; 2) For piles tested below their head (when they are incorporated in a structure), stress waves not only travel downward but also upward where they are reflected by the structure, pile top or non-uniforities. These secondary reflections traveling down fro above the velocity easureent location have to be identified so as not to be confused with upward traveling reflections fro pile ipedance changes below the velocity easureent location or the pile toe. For this reason, Johnson and Rausche (1996), presented a new ethod by taking two acceleration signals siultaneously along the pile shaft that provide a eans of deterining the stress wave speed as well as the necessary inforation for separating downward fro upward traveling waves. Since the ethod and the related analyses have been ipleented in soe devices and software in early 2, its application has gained popularity. However, due to the coplexity of the application, difficulties have been experienced both in testing and analysis due to lack of guidance and experience. In this paper, the background and principals are presented and a real test pile is used to verify the ethod and gain better understanding. A case study is then presented to deonstrate the applicability. Finally a coprehensive recoendation is presented to help guide future application. BACKGROUND Figure 1 deonstrates the stress wave traveling paths for a siple case when acceleration signals are taken at two locations along a unifor pile with a length of L and wave speed of the pile aterial of c. The first acceleroeter labeled as A 1 is ounted along the pile at a distance of Z 1 with reference to pile top and the second acceleroeter labeled as A 2 is ounted along the pile at a distance of Z 2 with reference to pile top. The absolute reference (pile top for this case) is not of iportance, but both acceleroeters ust use the sae reference, so the distance between the
can be calculated. When a haer ipact is applied at the pile top, a downward travelling wave reaches A 1 first at tie t 1, then A 2 at tie t 2. When the stress wave reaches the pile botto at tie of L/c, the stress wave is reflected to becoe an upward travelling wave. The upward traveling wave will reach A 2 first at tie t 3, then A 1 at t. Finally, at the tie of 2L/c, the upward traveling wave will arrive at the pile top. Let s define: Distance between A 1 and A 2 : D = Z 2 Z 1 Tie for wave travel between A 1 and A 2 : T = D/c And we have following relationships: t 1 = Z 1 /c; t 2 = Z 2 /c; t 2 = t 1 + T t 3 = t 2 + 2*(L Z 2 )/c; t = t 1 + 2*(L Z 1 )/c; t = t 3 + T For interpretation purposes, the easured accelerations are integrated to velocity and the resulting V 1 and V 2 velocities are coputed fro the accelerations easured by A 1 and A 2 respectively. V 1 and V 2 are the pile particle velocities registered at the acceleroeter locations and include both coponents fro downward and upward travelling waves. For this siple illustration case, V 1 and V 2 only include coponents fro the downward traveling wave before the t 3 tie because of 1) a unifor pile; 2) no resistance forces acting on the pile, i.e., no wave reflection prior to reaching the pile toe; 3) ipact applied at pile top. At tie t 3, V 2 will only include coponents fro the upward traveling wave and at tie t, V 1 will only include the coponents fro the upward traveling wave. For a ore coplicated (and realistic) case, V 1 and V 2 ay include coponents fro both downward and upward travelling waves fro ipedance change (including toe) reflections. Upward travelling stress waves include the ost eaningful inforation since they are reflected fro pile ipedance changes including defects and pile toe and soil resistance, so it is iportant to separate velocities into downward traveling and upward traveling waves. With two acceleration easureents, the velocity at A 1 location fro the downward traveling stress wave can be coputed by (Johnson, M., and Rausche, F., 1996): V 1 (t) = V 1 (t) - V 2 (t-t) + V 1 (t-2t) (1) Where: V 1 (t) velocity coponent fro downward travelling stress wave at tie t, coputed for A 1 location V 1 (t) velocity including coponents fro both downward and upward travelling stress waves at tie t, easured by A 1 V 2 (t) velocity including coponents fro both downward and upward travelling stress waves at tie t, easured by A 2
Now the velocity at A 1 location fro the upward traveling stress wave can be coputed by: V 1 (t) = V 1 (t) - V 1 (t) (2) Where: V 1 (t) velocity coponent fro upward travelling stress wave at tie t, coputed for A 1 location. Figure 2a illustrates a ore coplicated case: a pile with a defect and also an effective ipedance change at the botto of the pile cap. One acceleroeter A 1 is attached at the side of the pile at a distance Z 1 below the top of pile cap and above the defect location. A haer ipact is applied near the pile axis at top of the pile cap to induce a stress wave traveling downward. When this wave reaches the botto of the cap, it separates into two waves due to the ipedance change: 1) Part of the wave is reflected, which travels upward, defined as Stress Wave 1 (SW1); 2) The rest of the initially traveling downward wave continues to travel downward defined as Stress Wave 2 (SW2) and reaches A 1 as shown as pulse #1 recorded by A 1. Now the stress pulses observed at A 1 fro SW1 and SW2 (ignoring tertiary reflections) are: 1. Part of initial ipact stress wave (SW2): downward traveling; 2. Reflection fro top of pile cap of SW1: now downward traveling; 3. Reflection fro the defect of SW2: upward travelling;. Probably very sall secondary reflection fro top of pile cap of SW1: downward traveling; 5. Reflection fro pile toe of SW2: upward traveling The pulses observed at A 1 (1, 2,..5) include both downward and upward traveling waves. However, only #3 and #5 are of priary interest since #3 includes the inforation of the location and extent of the defect and #5 indicates the location of the toe. If there is only one acceleroeter used (the wave recorded at A 1 location), the data will show the result of all five waves superiposed, and it is ipossible to tell which is fro the iportant upward traveling waves. If another easureent is taken at A 2 location soewhere between A 1 and the defect location, equations (1) and (2) can be used to reove the downward traveling wave coponents fro the A 1 data as shown in Figure 2b. TEST PILE STUDY To further verify and deonstrate the ethod, a concrete test pile was created as follows: Total Length:.2 ( ft) Cross section area: Square 25x25 (1x1 inch)
An indent with depth of 75 (3 inch) between 3.5 and 3.96 (1 and 13 ft) fro pile top Concrete specific weight: 2 kg/ 3 (15 pcf) Wave speed: 3 /s (5 ft/s) t t L/c 1 2 t 3 t 2L/c z 1 z 2 A 1 A 2 Downward Upward Figure 1. Illustration of the two acceleroeter easureent: gage locations and wave propagation Figure 2. Illustration of a pile with a defect and haer ipact applied above the pile cap: a) all waves observed at A 1 including both downward and upward traveling waves; b) Only the wave coponents fro upward traveling waves The pile was laid horizontally on the ground to iniize soil resistance effects. As shown in Figure 3, two acceleroeters, A 1 and A 2, were side ounted at.57 (15 ft) and 5.1 (17 ft) respectively. A haer ipact was applied at the botto of the indent (3.96 or 13 ft) to siulate testing under an existing structure. The ipact generates stress waves travelling upward and downward and the paths of the stress waves are illustrated in Figure 3. Please note that only reflections fro the pile top and toe are included in Figure 3, while the reflections that are relatively sall fro ipedance changes along the pile are ignored for siplicity. In Figure 3, the dotted lines represent paths of the stress wave initially traveling downward fro the ipact location called SW2 while the solid lines represent paths of the stress wave initially traveling upward fro ipact location called SW1. Here is the list of the velocity variations observed at A 1 (the waves are nubered as labeled in the Figure 3): 1. Ipact pulse of SW2 2. Reflection fro pile top of SW1 3. Reflection fro pile toe of SW2. Reflection fro pile toe of SW1 5. Reflection fro pile top of SW2
6. Reflection fro pile top of SW1 7. Reflection fro pile toe of SW2 The waves #3, # and #7 recorded at A 1 are reflections fro the pile toe, which are upward traveling and of interest. Now let s look at the velocity easured at A 1 as shown in Figure. All of waves both fro upward and downward traveling waves appear on the recorded A 1 data. Please note that # and #5 are superposed together due to the relatively wide input pulse. Except for the first pulse which ay be easily judged as the initial ipact pulse and a downward travelling wave, the other waves are ipossible to tell if they is fro downward traveling waves (reflected fro pile top) or fro upward traveling waves (reflected fro pile toe).. c/s 6: # 1 A 1.6 13 ft 3.96 1 2 3 5 6 7.1 T1 Toe -.6 1 2 3 5 6 7 Figure. Test pile results: Velocity easured at A6: 1 # 1+1% 15 ft.57 A 1 A 2.5 17 ft 5.1 -.5 L/D=27 (D=2.66 c) 7.62 (31 /s) 1 2 3 5 6 7 9 1 11 13 1 15 17 1 19 2 21 22 (a) ft.2 Figure 3. Lay out of test pile, gage location and wave traveling (The reflections fro ipedance changes at indent edges are ignored) (b) Figure 5. Test pile: a) Velocity records at A 1 and A 2 ; b) Upward traveling velocity curve after two velocity analysis to reove downward traveling waves. Both velocity records collected at A 1 and A 2 are shown in Figure 5a and the two velocity analysis was perfored using equations 1) and 2) to reove downward traveling waves as shown in Figure 5b, which shows the reflections fro the toe under A 2 clearly and akes the data interpretation easier. If the data quality is not good enough to perfor two velocity analysis to obtain the result shown in Figure 5b, anual exaination of two velocity curves ay be needed (to look at phase shifts and deterine the upward waves). If the wave travels down, the dark
solid line fro A 1 is ahead of the blue dotted line fro A 2. Otherwise the wave is fro the upward traveling waves. CHALLENGES IN APPLICATION The signals easured fro the test pile are clean since it is under ideal conditions, and this helps to ake the two velocity analysis easier. In reality, the signals taken fro a pile under a ore coplicated situation ay ake interpretation ore difficult. The following factors affect the quality of easured data: Non unifor pile; Coplicated existing structures attached to pile; Effect of soil resistance; Effect of three diensions due to: o Gages can only be ounted on the side; o Choices of ipact location and direction are liited. CASE STUDY In this study, the testing data fro the LST easureent for the FDOT Bridge No. 726- SR 15 over Clapboard Creek, Duval County, Florida were used. The ain testing objective was the assessent of the unknown in-place lengths of the 1-inch square concrete piles in-service supporting the bridge as shown in Figure 6. Data acquisition was done with a Pile Integrity Tester. Testing was perfored by applying hand-held haer ipacts to the top of the pile cap concentrically over the top of each tested pile. Pile otion easureents were obtained with two acceleroeters affixed at two locations along the pile lengths (Figure 6); locations are between to 2.1 ( to 7 ft) as the axial pile distance below the botto of the pile cap. The distances between two acceleroeters were between.3 and 1.5 (1 and 5 ft). Four sizes of haers were used with weights approxiately between.5 to 62.3 N (1 to 1 lb). For this project the ain purpose to use two acceleration easureents was to distinguish the wave traveling direction rather than wave speed deterination since there is a reasonable knowledge of wave speed for this type of pile, so the distance between two acceleroeters was kept less than or equal to 1.5 (5 ft). To deterine wave speed accurately, it is recoended to place two acceleroeters as far apart as possible and apply the highest sapling frequency. Several easureents taken with two acceleroeters placed 1.5 (5 ft) apart were used to check the wave speed used in the analyses. The collected data, in for of velocity fro both acceleroeters, coputed upward traveling velocity and wave traveling path for selected records fro different piles are displayed in Figures 7 to 13. The layout of each figure is: top left - velocity curves fro both acceleroeters; botto left - coputed upward traveling velocity; right - wave traveling paths. For the selected records,
2 9. 7 the pile lengths deterined varied between 1.3 and 2.1 (6 and 66 ft) with wave speeds varying between 39 and 2 /s ( and 13 ft/s). A typical wave speed for concrete pile is /s (13 ft/s). For prestressed concrete piles, it ight be higher. A variation of wave speed in one job site within ±5% ay be expected, so the estiated pile length ay have an error of ±5%. In this study, the wave speed was varied to atch the observed toe reflection tie to a reasonable pile length. It can be seen that clear toe reflections were observed fro the coputed upward traveling velocity plots (botto left). Two Acceleroeters Figure 6. Photos of the Bridge, Bents, Pile and Locations of Acceleroeters.6.3 Pile: BENT 5 PILE 2-3: # 596+1% -.3 -.6 x 6 1.29 (1 /s) 2 2 Figure 7. Bent 5 Pile 2: A 1 at.3 (1 ft) and A 2 at 1.37 (.5 ft) 1.3
. Pile: BENT 9 PILE 2-3: # 251 Pile: BENT 9 PILE 2.2 -.2 -. x 5 1.29 (39 /s) 2 2 Figure. Bent 9, Pile 2: A 1 at.6 (2 ft) and A 2 at 2.1(7 ft) 1.3 Pile: BENT 1 PILE 2.25.13 Pile: BENT 1 PILE 2-3: # 25-1% -.13 -.25 2 2 2 x 5 19.2 (2 /s) Figure 9. Bent 1, Pile 2: A 1 at.6 (2 ft) and A 2 at 2.1 (7 ft) 19.2.3 Pile: BENT 11 PILE - 3: # 357.19 -.19 -.3 2 2 x 5 1.9 ( /s) Figure 1. Bent 11, Pile : A 1 at.6 (2 ft) and A 2 at 2.1 (7 ft) 1.9 Pile: BENT 13 PILE 2.3 Pile: BENT 13 PILE 2-2: # 5-3%.15 -.15 -.3 2 2 2 x 1 19.1 (2 /s) Figure 11. Bent 13, Pile 2: A 1 at ( ft) and A 2 at 1.5 (5 ft) 19.
Pile: BENT 13 PILE.1 Pile: BENT 13 PILE - 3: # 133.5 -.5 5 -.1 x 1 2. (2 /s) 5 1 15 2 25 1 15 Figure. Bent 13, Pile : A 1 at ( ft) and A 2 at 1.5 (5 ft) 2.1.3.15 Pile: BENT 1 PILE 2-2: # 52 Pile: BENT 1 PILE 2 -.15 -.3 x 5 1.59 (1 /s) 2 2 1.6 Figure 13. Bent 1, Pile 2: A 1 at.6 (2 ft) and A 2 at 2.13 (7 ft); Middle Size Haer ( lb) For Pile #2 in Bent #1 (Figure 13), the results fro different sizes of haers (1 lb and 1 lb) are shown in Figure 1 and Figure 15. Sall haers (or hard ipact tip) generate higher frequency input which helps identification of defects near pile top. However, the higher frequency content soeties is ore susceptible to show existing structure interference aking the data analysis ore difficult as shown in Figure 1. If this data has to be used to interpret results, then anual exaination has to be perfored. Let s look at the pulse P fro A 1 starting around (fifth pulse) and the corresponding pulse arriving at A 2 as P since P is the fifth pulse on A 2 (the blue dash curve). An additional check to ake sure that P and P are the sae pulse is that the distance between the is nearly the sae as the distance between the first pulses on A 1 and A 2. Since P is behind P, this pulse is part of the traveling downward wave. The right plot of Figure 1 shows the reflection paths as red dash lines. If P is part of the reflection (upward traveling wave), P and P would be on sae path, but they are clearly not. Thus the wave pulse P is not a reflection fro a defect or the pile toe. A larger haer or softer tip generates lower frequency content (Figure 15) which often results in a clearer interpretation of the toe reflection, and the wide input pulse helps reduce or filter out noise. The drawback is that the lower resolution ay result in the loss of detection of sall defects or defects very near the pile top. Therefore it is recoended to use ultiple sizes of
haers. For this project, the lb haer used gave the ost reasonable data quality, adequate resolution and enough energy to see the toe reflection..5.2 P P Pile: BENT 1 PILE 2-3: # 515-5% P Pile: BENT 1 PILE 2-3: # 515-5% -.2 -.5 x 5 1.59 (179 /s) P 2 2 1.6 Figure 1. Bent 1, Pile 2: A 1 at.6 (2 ft) and A 2 at 2.13 (7 ft); Saller Haer (1 lb).3 Pile: BENT 1 PILE 2-3: # 523-1% Pile: BENT 1 PILE 2.15 -.15 -.3 x 5 1.59 (1 /s) 2 2 1.6 Figure 15. Bent 1, Pile 2: A 1 at.6 (2 ft) and A 2 at 2.13 (7 ft); Larger Haer (1 lb) RECOMMENDATIONS Due to the difficulties of the application as discussed above, it is iportant to take a correct approach to use this ethod to obtain a reasonable result. The following are the recoendations based on our experience of applications and studies. Acceleroeter Mounting Two acceleroeters, A 1 and A 2, are installed along the pile at depths Z 1 and Z 2 which should be easured fro sae reference. It is iportant to accurately easure the distance between A 1 and A 2. Accelerations resulting fro an ipact applied soewhere above A 1 are recorded by both A 1 and A 2. Depending on the equipent, the signals fro both sensors should be captured and the trigger channel should be the A 1 sensor attached closer to the ipact location. Side ount acceleroeters are recoended and the sensors can be glued, bolted, or held by wax or putty to the side of the pile. Under wet condition or unclean surface, bolts are preferred to
attach the acceleroeter to anchors installed in the concrete piles. For tiber piles, the side ount sensors can be directly attached to the side of piles using lag bolts. If side ount sensors are not available, top ount sensors can be used (if they can be bonded parallel to the pile axis). Possible sensor ounting setups are shown in Figure. Ipact Technique Figure 17 shows three possible ipact spots. Directly ipacting above the pile generally gives the better signal if the connection between the pile top and structure is solid as shown in Figure 17a. If the top ipact is not possible or the connection between the pile and existing structure is not solid, the side ipact ay be tried, with the attept for the ipact to still be as parallel to the pile axis as possible. There are two ethods of side ipact as follows: 1. If allowed, a notch is created to apply ipact as shown in Figure 17b; 2. Otherwise, a block for ipact is attached to allow ipact axially. The aterial and attachent ethod of the ipact block will affect the signal quality (Figure 17c). When using a side ipact, a notch is preferred since the ipact can be applied without additional inference fro an ipact block attachent. Due to the potential coplexity of existing structures, it is recoended that testers try to apply ipacts both at the top and side. For saller size piles, ipacts ay be applied at a location 9 degrees away fro the gage. It is suggested to apply ipacts at least one diaeter above the top acceleroeter location. The selection of right haers is another iportant factor to obtain reasonable data. The frequency content in the input pulse induced by the haer ipact are affected by the contact between ipact surface and haer tip, the haer tip aterial and the haer weight. If the ipact spot is not sooth, higher frequency noise will be induced. A harder haer tip and/or lighter weight will generate higher frequency content. The higher frequency input helps identify the sall defects and defects near the ipact spot, but their energy decays faster and ay reduce the chance to detect the toe or defects in the lower portion of the pile. A softer haer tip (such as using Lexan instead of steel) or a heavier weight will generate a lower frequency input pulse. For piles under existing structures, due to ultiple reflections fro the interface between the pile and structure and/or ipact input energy splitting to different wave travelling paths, ore input energy with a lower frequency content is preferred when coparing to traditional testing at the pile top. Since there are several factors affecting the input frequency content and no haer design is standard, it is recoended that different sizes of haers should be tried to find the best haer for any particular situation.
Location and Distance between Two Sensors To Deterine Wave Speed The accuracy of wave speed deterination depends on the accuracy of the easureents of the distance between the two sensors and the tie (T) for the wave to travel fro A 1 to A 2. Increasing the distance between the two acceleroeters and/or increasing the sapling frequency generally iproves the accuracy of the wave speed deterination. If the saple frequency is 15, hz and thesensors are 1.5 (5 ft) apart, the accuracy could be within 2%. Thus the distance should be the largest that it can practically be. Please note that the wave speed deterined only represents the averaged value between two sensors, which ay not represent the overall wave speed for whole pile. Pile cap or existing structure Screw to concrete anchor or adhesively bond to piles Create a notch to adhesively bond gages. Notch < 2 o > b Attach ipact block Attach a block to the side and glue gage on the block. (a) (b) Figure. Sensor Mounting: a) using side ount sensors; b) using top ount sensors (a) (b) (c) b Haer tip size Figure 17. Ipact Locations and Methods: a) Ipact on top the structure above pile; b) Ipact on notch created on the side of pile; c) Ipact on a block attached to the side of pile To Separate Downward and Upward Traveling Velocities Since the atheatical anipulation is perfored (Equations 1 and 2) between two velocity records acquired fro two acceleroeters at different locations, it is very iportant to eet the following requireents:
1. Sae type of sensors; 2. Correct calibration factors; 3. The cross section of pile between A 1 and A 2 should be unifor. The distance between the sensor connected to the triggering channel and the ipact location or the nearest cross section change location should be at least larger or equal to.3 (1 ft); one diaeter of pile is preferred. The distance between two sensors is suggested as between.75 and 1.5 (2.5 and 5 ft). It could be larger for a wide input pulse with lower frequency content. CONCLUSION As a NDT ethod, LST has been widely used as a quality assurance tool for deep foundations. However, to test piles under an existing structure, the traditional LST with only one acceleroeter, or one acceleration plus ipact force, is often difficult to interpret due to the ultiple downward travelling waves induced by the initial haer ipact and reflections fro the existing structure. The extended LST ethod with two acceleration easureent helps to separate upward traveling waves fro downward traveling waves, and therefore to deterine the reflections fro potential defects and the pile toe. The case study presented in the paper is a good exaple of its successful application. To help assist in future applications, recoendations have been presented. REFERENCE Johnson, M., and Rausche, F. (1996). Low Strain Testing of Piles Utilizing Two Acceleration Signals, Stress Wave 1996, Orlando, FL, p.59-69 Liang, L., and Rausche, F. (211). Quality Assessent Procedure and Classifications of Cast-In- Place Shaft using Low Strain Dynaic Test, Proceedings fro Deep Foundations Institute 36th Annual Conference on Deep Foundations. Likins, G., and Rausche, F. (2). Recent Advances and Proper Use of PDI Low Strain Pile Integrity Testing, Proceedings of the 6th Int. Conf. on the Application of Stress Wave Theory to Piles. Sao Paulo, Brazil, pp 211-21. Mullins, G. (2). Theral Integrity of Drilled Shafts, Presentation to the Florida Suncoast Chapter, ACI. Rausche, F., Likins, G., Liang, L. (2). Challenges and Recoendation for Quality Assurance of Deep Foundations, Proceedings of the 33rd Annual and 11th International Conference on Deep Foundations. Deep Foundations Institute: New York, NY (CD-ROM) Webster, K., Rausche, F. and Webster, S. (211). Pile and shaft integrity test results, classification, acceptance and/or rejection, TRB 9th Annual Meeting, Washington, D.C.