1

Study on cyclic pile load test of pile resting and socketed in rock Sumisha A P & Arvee Sujil Johnson Geotechnical Engineering, College of Engg Trivandrum ABSTRACT In this study, different data of cyclic pile load test of pile resting and socketed in rock have collected and analyzed to compare it with pile sand for separating skin friction and end bearing. We all know that the transfer of load in different pile is different, friction takes the load in friction piles or end bears in end bearing piles but in rock socketed piles both friction resistance and end bearing are act together against the load. The load transfer is also depending on the type of soil in which the piles rest. Cyclic loading on pile is the best method to determine, how much of the total load takes up by the friction or by the point resistance. There already exists one graphical and analytical method to separate the skin friction and end bearing of pile resting on sand. In case of pile socketed in rock or resting on rock, the separation of skin friction and bearing is done with an assumption that rocks are dense sand. The data of cyclic loading of pile in rock are analyzed as per the IS method and found to be the results are erroneous. So a separate method is essential to analyze the pile in rock. Initially a common trend in load v/s settlement curve to be finalized for introducing a new method to separate skin friction and end bearing. The introduced new methods are worked out and that shows, graphical method gives almost acceptable results but the analytical method not. Further studies are required for clarifying graphical method and to modify the analytical method. 1 INTRODUCTION 1.1. Background of the present study

The increment in load to foundation and available hard strata at higher depth are lead to deep foundation. The piles are classified into friction pile and end bearing pile based on the load transfer. The rock socketed pile posses both skin friction and end bearing over the socketed depth. The skin friction and end bearing of pile can be separated by conducting cyclic loading on pile and the settlements are noted under loading and unloading. The cyclic loading test is safer because it tests the worst condition of pile under continuous loading and unloading.

Piles resting on sand have a particular IS code graphical and analytical method to separate skin friction and end bearing. These same methods have been used for the piles resting on the rock or socketed into the rock with an assumption that rock is dense sand. This assumption is lead to this project. Why will not introduce a new method for separating skin friction and end bearing of pile on rock. A single project data is not sufficient to introduce such a method since the lack of clarification of load deflection curve. So have to collect different journals and case studies for analysis to fix the common trend of load deflection curve of pile on rock. This paper is the partial result of study dealing with the behavior of pile in rock subjected to cyclic axial loads. As part of this project, an extensive search of existing data on the cyclic behavior of piles in rock was performed. A large number of publications were collected dealing with laboratory tests, model pile tests and pile load tests. The data were analyzed and the results of the analysis, which are represented in this paper served a guideline for introducing a new method for separating skin friction and end bearing of pile resting or socketed in rock. The comparison of load deflection curves of pile on sand and rock also included. The data of cyclic pile load test on rock were analyzed

as per the IS code method and it found to be erroneous. So a new method is essential to analyze the pile in rock. After finalizing the common trend of load v/s settlement curve of pile in rock, a new graphical and analytical method are introduced. The acceptance of results of these methods is discussed in this paper and further study required areas are also mentioned in it The main objective of this study are

• To compare the results of cyclic pile load test of pile on both sand and rock

• To determine how they are different

• Determination of common trend p-y graph of pile in rock

• Comparison of graphical and analytical method to separate skin friction and end

• bearing

• To introduce a modified graphical and end bearing method for separating skin friction and end bearing .

1.2. Data collected The cyclic pile load test data were taken from these journals. Radhakrishnan et al.(1989)”Load transfer behavior of rock socketed pile”, Omer et al. (2002)”Instrumented load tests in mudstone: pile

capacity and settlement prediction”

2. METHEDOLOGY

2.1. General Different journals and case studies of cyclic load test on pile resting on rock or socketed in rock were collected and the required data were taken. Mainly it requires the settlements on each loading and unloading. If the collected data in graphical form it should be extracted to the values. The collected data were tabulated to find

2

settlement of sub grade for each loading then plotted graph between settlement of sub grade and load. As per IS code method (2911 part 4) the graph was analysed to separate skin friction and end bearing. Actually this IS code is used for analysing cyclic pile load test on pile which resting on sand so here we are going to do the analysis on pile resting on rock with an assumption that rock is a dense sand.

Fig: 1 separation of skin friction and end bearing

This is the existing IS code graphical method for pile resting on sand and the analytical method is shown below. In graphical method a straight line is drawn which passes through the points in the straight line portions and another straight line which passes through zero and parallel to the previous line. For each load the end bearing and skin friction are shown in the above figure. In analytical method the value of m is found out by the equation and solved the two equations to find point resistance and skin friction. In other way the m is the slope of the load v/s elastic compression of sub grade curve and can determine directly from the curve.

m = T

AELTS

))/(*( .......... 1

(IS code 2911 part 4 2013) m = a constant

s =change in elastic settlement

T = change in load, L = length of the pile A = cross sectional area of pile, E = modulus of elasticity S = mT (S = corrected settlement , T = total load on pile)

T = P+F and S = m P +AE

LFT *))2/(( ....... 2

P = Point resistance, F = skin friction The graphs shown below are the collected data.

Fig: 2 Cyclic pile load data (Source: Load transfer behavior of rock-socketed piles Radhakrishnan et al

(1989))

Fig: 3 cyclic pile load data (Source: Instrumented load tests in mudstone: pile capacity and settlement prediction Omer et al. (2002)) For the study we require settlements on corresponding loading and unloading. The graphs are extracted to get settlements corresponds to loading manually which is shown below

Fig: 4 manual extraction of settlement (Source: Load transfer behavior of rock-socketed piles Radhakrishnan et al (1989))

3

Fig:5 extraction of settlement (Source: Instrumented load tests in mudstone: pile capacity and settlement prediction Omer et al. (2002)) The extracted data are shown below and tabulations to get elastic settlement of sub-grade and load for the 1st data Table : 1 basic data of rock socketed pile

Basic data of pile

Length of pile 12.4 m

Grade of concrete 35 N/mm2 Elastic modulus of pile

33721654.76 kN/m2

Dia of pile 0.705 m The graph of load v/s elastic settlement of sub grade was drawn as per IS code (2911 part 4) by assuming no frictional force and no elastic settlement on pile. Three trials are required to finalize the curve so in 2nd trial, the separated friction from 1st trial is taken as frictional resistance for calculating elastic settlement of pile. Then new curve was plotted and repeated this until the separated point resistance and skin friction in adjacent two trials are almost same. Table :2 IS code trial analysis

Load (T)

Elastic recovery mm

Frictional resistance

Elastic settlement of pile mm

Elastic settlement od sub grade mm

Point resistance kN

Frictional resistance kN

0 0 0 0 0 0 0 203.9 6 0 0 6 90 113.9 356.8 12 0 0 12 180 176.8

1019.4 16 0 0 16 240 779.4 According to IS code method, the skin friction and end bearing are separated by both graphically and analytically and compared the results obtained.

Fig :6 load v/s elastic settlement of sub grade (three trials) for data 1 Table:3 separated skin friction and end bearing

Lload (T) Graphical Analytical

Point resistance kN

Frictional resistance kN

Point resistance kN

Frictional

resistancekN

0 0.0 0.0 0.0 0

203.9 80.0 123.9 197.28 6.57

356.8 150.0 206.8 345.26 11.52

1019.4 180.0 839.4 986.54 32.85

In this curve, the settlement increases abruptly with initial loading and less settlement at the end of loading. This is clearer from the separated values of skin friction and end bearing. Both end bearing and skin friction are act together to resist the load at all loading. The initial inclined straight line portion represents the development of frictional resistance and point resistance and curve shows both increases. Comparing to point resistance the frictional resistance is more all the time. By analysing the graph as is per IS code method at a particular portion the frictional resistance values are same. The further portion analysed and that shows large increase in frictional resistance. The point resistance is increasing all the time.

The separation of skin friction and end bearing have done by analytical method that shows point resistance is greater all the time and only a small part of friction is there to resist the coming load.

4

Fig:7 separation of skin friction and end bearing The frictional part is more than point resistance in graphical method but in analytical, frictional part is less than point resistance. The sudden downward fall of curve shows some type of failure or complete mobilization of load. From the previous analysis the skin friction is getting mobilized completely because same value of friction has obtained. As per the IS code method further increase in skin friction is also shown here. If the continuous same value of friction shows frictional failure, then it is not possible to see such increase in friction further so this is the part when the IS code method fails to determine or separate the skin friction and end bearing for rock socketed pile. So modification is required to separate skin friction and end bearing for rock socketed piles. This is another data of rock socketed pile under cyclic pile load. Table: 4 basic data of rock socketed pile

Basic data of pile

Length of pile 28 m

Grade of concrete 50 N/mm2 Elastic modulus of pile 40305086.53 kN/m2 Dia of pile 0.9 m

The same steps were done for the 2nd data

Table :5 IS code trial analysis

Load (T)

Elastic recovery mm

Frictional resistance

Elastic settlement of pile mm

Elastic settlement od sub grade mm

Point resistance kN

Frictional resistance kN

0 0 0 0 0.00 0.0 0 611.62 4.78 0 0 4.78 220 391.6 920.5 7.7 0 0 7.70 400 520.5

1220.2 11.6 0 0 11.62 600 584.5 1651.4 16.2 0 0 16.24 840 811.4 1834.9 17.9 0 0 17.92 940 894.9

Fig: 8 load v/s elastic settlement of sub grade data 2 Table :6 separated skin friction and end bearing

load (T) Graphical Analytical

Point resistance kN

Frictional resistanc

e kN

Point resistanc

e kN

Frictional resistancekN

0 0.0 0.0 0.0 0.0

611.62 270.0 341.6 605.4 6.2

920.5 550.0 370.5 908.09 9.34

1220.2 560.0 660.2 1210.78 12.46 1651.4 530.0 1121.4 1634.55 16.82 1834.9 620.0 1214.9 1816.17 18.69

5

Fig: 9 separation of skin friction and end bearing

A large settlement variation in initial loading and increament of settlement deacreases after a particular loading.The frictional part obtained from graphical method is greater than the frictional part obtained from analytical part. An increase and decrease can be seen in graphical method and a slight continuous increase in analytical method. In graphical method the 4th and 5th loading gives almost same friction that means complete mobilization of friction. So further increase in friction is not possible but the method shows such increase in 6th loading so the method is not suitable for pile in rock. The graphically obtained frictional capacity is ten times the analytically obtained frictional capcity. In this it also shows, after a continuous same value of friction large increse is vissible which is not acceptable in the sense of taking higher load after a failure. A new method is essential for separating skin frictionand end bearing for a rock socketed pile 3.0 MODIFICATION The shape of the curve can be explained with the load transferred through the pile. First you should aware that the load coming to the pile always distributed to both skin friction and end bearing. Every load on pile has both frictional and end resistance but the proportions are different at different stages. The end bearing is always found to in an increasing manner and the skin friction increases initially then attained a constant value. The parabolic shape shows the initial increase of skin friction and constant value is obtained from the straight line portion of the curve. The next straight line with less settlement shows the increase in skin friction as per the existing IS code method for sand. For a rock socketed pile the skin friction mobilizes before the mobilization of end bearing so the further increment of skin friction is not possible. This will be and typical variation of curve for rock socketed pile under cyclic pile load test.

Fig: 10 assuming trend of graph

This is a typical trend of rock sockketed pile under cyclic pile load test. The third portion which is in straight line shows a small increase in settlement with load. We are all know that the skin friction values are constant for different loading means it is the maximum value that the skin friction can hold and in other way constant values means complete mobilization of skin friction has occurred. But the third portion increases the skin friction as per the existing IS code method which is not possible since the skin friction has attained its maximum capacity. The less settlement means the bottom rock support is very much high which is known as end bearing. The point resistance is stronger and cause less settlement than a frictional resistance, so the less settlement portion reveals the major part of the load is taken by point resistance. The capacity increase in curve at the end is the increase in end bearing not by the increase in skin friction by IS code method. So the method should be modified to show the actual process happening there. 3.1 Modified graphical method The maximum value of skin friction is taken as constant for every load after the mobilization of friction as frictional part and the rest of the load on pile is taken as end bearing. For executing this, the mobilized skin friction value is taken as offset from all the next loads and drawn a new curve by joining the left tip of the offsets. The horizontal distance in x axis from y axis to the new curve is corresponds to the point resistance and the new curve to previous curve (offset) corresponds to skin friction. The pictorial representation of this method is shown below. Comparing the modified and IS code graphical method, the mobilization of skin friction and increase in end bearing is only visible in modified graphical method. Is code method shows end bearing is less than skin friction at the end of loading. Table :7 comparison of IS and modified graphical

method

Load (T)

Graphical Modified graphical

Point resistance kN

Frictional resistanc

e kN

Point resistance kN

Frictional

resistancekN

0 0.0 0.0 0.0 0.0

203.9 80.0 123.9 83.87 120.0

356.8 150.0 206.8 233.78 123.0

1019.4 180.0 839.4 896.4 123.0

6

Fig: 11 modified separation of skin friction and end bearing data 1 3.2 Modified analytical method

Fig: 12 Determination of slope of data 1 Determination of m from cyclic loading data Load transfer behavior of rock-socketed piles radhakrishnan et al (1989)

In IS code analytical method the value of m is taken as slope of load v/s elastic settlement of sub grade. In new method the value of m is taken from cyclic pile load test curve as the slope of red line shown above. Then the results obtained are shown below. Table : 8 comparison of Is code and modified analytical method of data 1

Load (T)

Modified analytical Analytical

Point resistan

ce kN

Frictional resistanc

e kN

Point resistanc

e kN

Frictional resistancekN

0 0.0 0.0 0.0 0

203.9 152.46 51.41 197.28 6.57

356.8 304.9 51.88 345.26 11.52

1019.4 355.13 664.27 986.54 32.85

For the second data

Fig : 13 determination of slope of data 2 Table:9 comparison of IS code and modified analytical method

4. RESULTS AND DISCUSSION

Different data of cyclic pile load test of pile resting or socketed in rock were collected. And load v/s elastic compression of sub grade curves have drawn. The drawn curves are analyzed and a common shape is finalized. The common trend is explained by the load transferred through the pile which means the point resistance and skin friction corresponds to every load have determined and their magnitude variation gives such shape to the curve. The point resistance is increasing all the time but the skin friction increases initially and attains a maximum value then become constant. The further increment in load resistance capacity is by increment in point resistance. The IS code graphical method is found to be erroneous since it gives such result that skin friction is also increasing beyond the maximum mobilized value. An increase after maximum capacity is not a possible one. The mobilization is identified by the similarity in skin friction for different adjacent loading. This error in IS code method led to the introduction of new graphical and analytical method. The introduced graphical method gives almost favorable results and further studies are required to modify the introduced analytical method since it is also erroneous.

load (T) Modified analytical Analytical

Point resistance kN

Frictional resistance kN

Point resistan

ce kN

Frictional

resistancekN

0 0.0 0.0 0.0 0.0 611.62 389.55 222.07 605.4 6.2 920.5 633.69 283.74 908.09 9.34

1220.2 964.14 259.1 1210.78 12.46 1651.4 1350.43 300.94 1634.55 16.82 1834.9 1489.5 345.36 1816.17 18.69

7

5 CONCLUSIONS

A common trend in load v/s elastic compression of sub grade of pile under cyclic loading in rock has determined which is different from the trend of sand The IS code methods gives erroneous results for analysis of pile in rock under cyclic pile load test. For the same data, IS code graphical and analytical methods gives different value for point resistances and skin friction. Modified graphical and analytical methods are introduced and verified it’s accuracy by considering number of trials of different cyclic pile load test data collected. Further study is essential to introduce most favorable analytical analysis 6 REFERENCE Brierley, G S.,Thomson, D E., and Eller, C.W.(1979).

“Interpreting end bearing pile load test results”ASTM STP 670,pp. 181-198.

IS code part 4 (2013) “code of practice for the design and construction of the pile foundation load test on pile.” Bureau of Indian standards, New Delhi, India

Omer et al. (2002)”Instrumented load tests in

mudstone: pile capacity and settlement prediction”

Radhakrishnan et al.(1989)”Load transfer behavior of rock socketed pile”

Tham et al. (2013) “A study on socketed steel H piles under vertical load”

Vanwheele,A F.(1957) “Method fir separating skin friction and end bearing of pile” Proc. Of 4th international conf. soilmechanics foundation

engineering, London, pp.76-80.