Interpretation of Compressibility Properties of Soil Surrounding Bored Piles from Pile Load Test Results

Unsatisfactory bored pile may result due to low skin frictional resistance or due to low end bearing resistance. A new methodology is proposed herein to estimate the shear stiffness of the material along the pile shaft using the static load settlement curve. The stiffness, estimated in terms of the Young's modulus, of the material along the pile shaft is an indication of the amount of ultimate skin friction. The new methodology, in combination with the earlier method proposed by Thilakasiri and Silva [6] to estimate the Young's modulus of the material below the pile bottom, can be effectively used in interpretation of static load settlement curves to determine the type of the material along the pile shaft and that of the material below the pile bottom. The developed methodology is implemented for static load-settlement curves obtained from testing of bored piles and the advantages of the new methodology are demonstrated.


Introduction
Cast in-situ bored piles are constructed by creating a borehole in the ground by drilling and subsequently filling it with concrete.During the drilling process bentonite slurry is used to stabilize the sides of the borehole with or without steel casings.The concreting is carried out from the ground surface using a tremie pipe and systematically bentonite is replaced by concrete during the concreting process.However, if the slurry full of hole is kept for a long period of time, a thicker and harder "cake" will be formed and replacement of the slurry with concrete will be difficult resulting in a layer of bentonite slurry between the wall of the hole and the fresh concrete.This layer of bentonite, invariably will reduce the development of skin friction between the soil and the concrete.
In the design of such piles, it is a common practice to neglect the skin friction resistance, mainly due to the presence of bentonite in the borehole during concreting, and design of such piles is entirely based on the possible end bearing from the bedrock.In the local design practice, the bored piles should be socketed to a minimum of 300 mm into tho bedrock.However, based on the results of static and dynamic load tests, Thilakasiri [4] showed that there is reasonable skin frictional resistance developed in bored and cast in-situ concrete piles constructed in Sri Lanka.
Carrying capacity of cast in-situ concrete piles may be reduced due to two main reasons: (i) False identification of the bedrock or improper cleaning of the borehole resulting in low end bearing capacity; and (ii) Presence of a thick layer of bentonite between the pile and the surrounding soil due to formation of a hard filter cake resulting in low skin frictional capacity.The weathered rock layer consists of different type of material depending on the degree of weathering and the type of the parent rock.At most locations, the transition from highly decomposed rock to the sound bedrock is gradual and there is no definite demarcation between the weathered rock and the bedrock.Due to this reason, the geotechnical engineers are faced with the problem of identification of the so called 'bedrock' and socketing the pile 300mm into the bedrock.Thilakasiri and Silva [6] proposed a new methodology to investigate the end bearing condition of cast in-situ concrete piles based on static load test results.However, there is no proper method to investigate the shear stiffness of the surrounding soil and hence the skin friction provided by the same.
The objectives of this research paper could be summarised as follows: • Introducing a new methodology to investigate the stiffness of the soil surrounding the pile shaft; • Hence, the separation of the load acting on the pile during initial stage of the loading into skin friction and end bearing components; and • Application of the new methodology to rock socketed bored piles in Sri Lanka to investigate the end bearing condition of such piles.

Determination of the
Compressibility Parameters of the Material Surrounding the pile from Static Load Test Results

Material Present at the Pile Toe
Thilakasiri and Silva [6] investigated the loadsettlement behavior of bored piles and proposed a methodology to obtain the Young's modulus of the material present at the pile toe.According to Van Weele [8], when a pile is loaded, initially the load is carried mostly by skin friction until the shaft deformation is sufficient to mobilize the limiting value, as shown in the region oa in Figure 1.When the limiting skin friction is mobilized, the point load increases nearly linearly until the ultimate end bearing capacity is reached, as shown by region ab of Figure 1.At the point of the ultimate end bearing, the pile undergoes large settlement under a relatively small load increment.However, it should be noted that the load settlement curve from the static load test could deviate from the shape shown in Figure 2 due to unusual subsurface conditions and defective pile shaft.Moreover, depending on the magnitude of the applied load and the ultimate skin frictional resistance, for some piles ultimate skin frictional resistance may not be mobilized.For such piles, the region ab, shown in Figure 1, may not be visible.
Once the skin friction has reached the ultimate value, P 5k,n' further increment of skin frictional resistance is not possible.Therefore, any increment in the load on the pile head (L\p), after reaching ultimate skin friction, will be equal to the change in the end bearing load (L\P,nJ Therefore, it could be proved that the Young's modulus (EJ of the material below the pile toe can be expressed as equation [1] (Tomlinson [7]): If the skin friction has reached the ultimate condition equation [1] could be used to estimate the Young's modulus of the material beneath the toe of a pile.It should be noted here that the Young's modulus (Eb) depends on the strain level due to the highly non-linear nature of soil and rock.However, the settlement required at the pile toe to develop ultimate skin frictional resistance is generally in the order of 2 to 5mm (Thilakasiri et al. [5]).As a result, the strain level of the material at the pile toe, when ultimate skin friction is mobilized, is relatively low.Therefore, if the L\p and ap used in Equation [1] are obtained closer to the point "a" in Figure 1, estimated deformation modulus is corresponding to low strain levels.

Material Present along the Pile Shaft
Various theoretical methods were introduced by different researchers to estimate the settlement I ENGINEER 11( nd bearing bored piles.Most of the th oretical approaches in settlement estimation u • s idealized linear elastic medium urrounding the pile.As it is commonly known, t ii exhibits a highly nonlinear behavior, ,. p cially when subjected to large strain levels.l'h refore, applicability of such theoretical 111 thods is limited only to initial loading stage of the pile.One such widely used method is pr posed by Poulos [3] and in this method the l11.1d (P) applied on the pile top is related to pile top settlement (p) through Equations [2] and [3].
diameter and length of the pile, Young's modulus of the material present along the pile shaft, Young's modulus of the material present below the pile toe and the Poisson ratio of the material present along the pile shaft.Out of the above parameters, Young's modulus of the material present below the pile toe was obtained from the method proposed by Thilakasiri and Silva [6].Young's modulus of the concrete is taken as 35 GPa.The Poisson ratio of the soil varies between 0. Nh -Modification factor that depends on the ratio E/E, and k.
The relative stiffness factor k, which is defined tor the solid piles as the ratio between the Young's modulus of the pile material (E) and that of the soil along the pile shaft (E), is the main parameter that determines the modification factors.In addition, the ratio h tween the Young's modulus of the end h aring layer (Eb) and the average Young's modulus of the soil along the pile shaft (E) is u cd in estimation of the modification factor Rb. Load(kN)

Development of Skin Friction and End Bearing During Initial Stage of the Loading
Poulos and Davis [2] presented a methodology to separate the magnitude of the SF and end bearing under the low strain condition of the surrounding soils.Poulos and Davis [2] showed that the parameter �' defined as the ratio between the end bearing resistance to the total applied force during early stage of loading, can be mathematically represented by equation [4] 71 Where Where {3 0 -Parameter that depends on the length and diameter of the pile.
Cv -Parameter that depends on the Poisson ratio of the soil along the pile shaft.
Ck -Parameter that depends on the relative stiffness k; of the pile.
Cb -Parameter that depends on the ratio E/E, and k.
Knowing the Eb and E. of the material surrounding the pile, the ratio between the end bearing resistance to the total applied force, {3, during early stage of loading can be estimated.

Method of Study
A survey was conducted in 2005 to collect information of the static load tested bored piles socketed into the bedrock.The authors contacted the major piling contractors, prominent pile testing contractors and the relevant clients to obtain the pile loadsettlement data of piles from different parts of Sri Lanka.More than 75% of the organizations contacted by the authors were generous in giving the requested information Authors would like to express their gratitude to the organizations and individuals for making their data available to this study for further improvement of piling practices in Sri Lanka.D1;1ring the survey, load-settlement curve, details of the pile and, subsurface information were also collected of the tested piles.
Data gathered for 40 such piles were analyzed using the methodology proposed in this paper and the methodology proposed in Thilakasiri and Silva [6).

Results
Two methods were used to estimate the Young's modulus of the material present along the pile shaft and at the pile toe.The details of the piles and the Young's modulus of the material at the toe of the piles and along the pile shafts are given in Table 1.
Rock types and the status of the upper part of the rock, obtained from borehole investigation, for different projects are given in Table 2.
The results are shown in graphical format in Figures 3 and 4. Figure 3 shows the variation between the Young's modulus value at the pile toe (EJ and the residual settlement of the pile top after complete unloading.Figure 4 shows the variation of the Young's modulus along the pile shaft (E) and the residual settlement after complete unloading.
It is evident from the information given in Table 1 and Figure 3  When the material along the pile shaft is weak more load will be transferred to the pile toe.However, since the residual settlement will be governed by the stiffness of the material at the pile toe, stiffer material at the pile toe will reduce the residual settlement.3. The Young's modulus values given in the Table 3 are obtained from testing the soil or rock in-situ.Therefore, the stiffness is less than that is normally obtained from testing a solid rock specimen in the laboratory.
It is evident from the comparison of the Young's moduli values obtained for the material beneath bored piles and typical values obtained from literature, that the estimated Young's moduli values are not corresponding to• those expected for solid bedrock.In residual formations found in Sri Lanka, the bedrock consists of strong rock types such as biotite gneiss, charnockitic gneiss, granite etc. but the upper layers of the bedrock i. Possibility of the presence of loose debris at the toe of the pile; ii. Weathered and fractured nature of the bedrock; and iii.The small strain level of the ultrasonic method used in the testing to obtain the Young's modulus.
is fractured and partially weathered.Therefore, the Young's modulus of in-situ rock mass could be much less than that obtained from testing fresh solid intact rocks.Comparison of the Estimated Young's modulus of the material below the pile toe given in Table 1 and those obtained from laboratory testing of some rocks found in Sri Lanka given in Table 3 (Jayawardane [1]) clearly shows that the material present below the pile toe is soft.The main reason for such difference could be due to: ICTAD/DEV /16 on Specifications for Bored and Cast in-situ Reinforced Concrete Piles specifies that if the pile undergoes more than 12mm residual settlement, when loaded upto 1.5 times the working load, the pile should be considered as unsatisfactory.According to that guideline, there are four piles, which undergo more than 12 mm residual settlement.Out of these four piles, two piles have very low Eh values indicative of very loose end bearing conditions or a void.Furthermore, the ratio between Eh and E, for these two piles is also very low.The other two piles hav Eh values 74 ENGINEER The Young's modulus of rock depends on the degree of weathering the material is subjected to, the fracture pattern and spacing.Therefore, a considerable variation in the stiffness of the material at the pile toe could be expected.Moreover, depending on the quality of the construction methodology adopted, even loose debris could be expected at the pile toe of some piles as well.This condition is quite evident from the estimated Young's modulus values, as there is a large variation in the estimated values. •

Conclusions
The new methodology introduced makes use of the initial stage of the static load-settlement curve to determine the elastic compressibility properties of the material present along the pile shaft.The new method introduced in Figure 5 shows that the variation of f3is less than about 10% for the E/E, ratio less than about 3.However, it is not possible to establish a definite trend for f3 with E/E, ratio.The estimated f3 shows that the skin friction is predominant during the initial stage of loading.It also shows that contrary to the common belief among some geotechnical engineers, there is a significant contribution from skin friction to the total load carrying capacity of cast in-situ end bearing bored piles as previously shown by Thilakasiri [4]., orrcsponding to medium dense sand and that , ould be due to collapsing of the sides into the bored hole before concreting or the false id mtification of the competent bedrock. .

Distribution of the Total Load between End Bearing and Skin Friction
/\ nother important factor to note here is the v iriability of the Young's modulus values of the Ill terial present below the piles from the same pr ject.Variation of the EJor P4 is 1194.5 MPa to 18.1 MPa whereas that for PU is 287.6 MPa to 3.9 MPa.The main reasons for such large variations I nuld be due to: variability of the material across th site; and inconsistency of the construction m thodology adopted by the piling contractor.
l'he ratios between the developed end bearing und skin friction (fJ) during the initial stages of I > ding were estimated using method proposed m Poulos and David [2].The estimated values of fl for the case studies are given in Table 1.There ,Ir different factors affecting the value of f3 luring initial stages of the loading, such as: tiffness of soil surrounding the pile (£,); tiffness of material at the pile toe (EJ; length of th pile (L); and diameter (D) of the pile; tiffness of the pile material etc. Due to the effect of large number of parameters on the value of {3, it i very difficult to establish a definite variation of� with any single parameter.The variation of fJ with the ratio E/£ 5 is shown in Figure 5. Ultrasonic methods on intact rock samples 11 rnblende Biotite Gneiss 37170 -48770 Jayawardane [1]  Ultrasonic methods on intact rock samples Biotite Gneiss 31460 -58630 Jayawardane [1] Ultrasonic methods on intact rock samples uartzite 35960 -54170 Jayawardane [1] Ultrasonic methods on intact rock samples ,. met Sillimani te 49950 -73630 Jayawardane [1] Ultrasonic methods on intact rock samples Marble 48130 -65070 Jayawardane [1] Ultrasonic methods on intact rock samples ,ranulite 40700 -52410 Jayawardane [1] Ultrasonic methods on intact rock samples 75 ENGINEER combination with the method introduced by• Thilakasiri and Silva [6] is used to obtain the elastic compressibility properties present along the pile shaft and at the pile toe.Load-settlement curves from forty load tested piles were analyzed using the proposed methodologies and the elastic compressibility properties present along the pile shaft and at the pile toe were estimated.The analysis of the estimated Young's modulus of the material present along the pile shaft and the pile toe clearly showed that the developed methodologies could be used to identify the reason for failure of piles.Furthermore, likely properties of different soil and rock types are also presented to guide the users to identify the likely soil types corresponding to the estimated compressibility properties.Using the elastic material properties estimated, the ratio between the• developed end bearing and skin friction (() for the initial stage of the loading was estimated.Estimated ( values show that contrary to the common belief among Sri Lankan geotechnical engineers, there is a significant contribution from skin friction to the total load carrying capacity of cast in-situ end bearing bored piles

( 1 ) 1 IFigure 1 :
Figure 1: Load settlement curve and the range ab, after mobilization of the ultimate skin friction.

5 L
2 for loose sand to about 0.45 for saturated clay (Poulos and Davis [21).Young's modulus of soil along the pile shaft /) -Diameter of the pile shaft I -Settlement influence factor for . .incompressible pile in a semi-infinite elastic half-space with a poisson's ratio of 0.ad -settlement curves obtained from static load settlement curves often show a linear p rtion at the initial stage of loading as shown 111 Figure 2. Two regions shown in Figure 1: oa .mdab are clearly visible in most of the load- •ttlement curves obtained from static load I • ting of bored piles.It is clear from the Equations [2] and [3] that the ttlement of the pile head depends on the ung's modulus of pile material (concrete), NA -Modification factorfor compressibility of the pile N,. -Modification factor for the poisson's ratio

Figure 5 :
Figure 5: Variation of f3 with E/E, that piles having high residual settlement after complete unloading have a lower Young's modulus value of the pile toe material.Similarly, Figure4shows that the piles having lower E. values undergo large residual settlement but there is considerable number of piles with lower E. values with lower settlement as well (Pile ID 8, 10, 14, 15, 16, 23, 31 and 33 in Table1).But in almost all the above mentioned piles considerably high Eb values are observed.

Table 1 :
Details of the piles,.observedsettlement at 1.5 x WL, Net settlement after unloading and the estimated Young's modulus of the material at the pile toe and along the pile shaft lh relative stiffness, k, of the material along the pll shaft falls within 140 to 1675 with an 1v rage of 944, neglecting the pile no 23, which I in talled in very soft overburden material.ENGINEER I

Table 2 :
General rock types and status of the rock present at the project sites given in Table1.

lable 3 :
Typical values of Young's Modulus of soils and rocks from published researches