SKIN FRICTION OF PILES COATED WITH BITUMINOUS COATS Makarand G. Khare 1 and Shailesh R. Gandhi 2

Transcription

1 SKIN FRICTION OF PILES COATED WITH BITUMINOUS COATS Makarand G. Khare 1 and Shailesh R. Gandhi 2 1 Ph.D Student, Dept. of Civil Engineering, Indian Institute of Tech. Madras, Chennai, India makarandkhare@yahoo.com 2 Professor, Dept. of Civil Engineering, Indian Institute of Tech. Madras, Chennai, India srgandhi@iitm.ac.in ABSTRACT: Piles are often coated with a slip layer such as bitumen to reduce the dragload. The friction between granular soil and coated as well as uncoated pile surfaces were studied using a modified direct shear test apparatus. The lower half of the direct shear box was replaced with a concrete or steel mild steel block to represent a pile surface. Two types of bituminous coat namely Shalikote and Bitumen of grade were evaluated in this study. Shalikote and Bitumen coating reduced the interface friction by more than 70% and 90% respectively. The shearing resistance of pile material-coat-soil is found to be influenced by normal stress, type of coat, and coat thickness. The effect of saturation and the rate of soil settlement on the interface friction are also discussed. Pull out tests were carried out on coated and uncoated model aluminum pile placed in a circular tank filled with granular soil. The results of pull out tests are compared with interface friction tests. INTRODUCTION The settling soil imposes dragload on piles and may cause excessive settlement of pile foundation. The large magnitude of dragload may necessitate higher pile cross section and/or deeper pile penetration which increase the cost. The pile design must ensure that the dragload is accommodated without causing any structural distress and excessive settlement of pile. In the past, various methods have been adopted to reduce the dragload depending upon the field condition. Coating the pile with bitumen is the most economical method for reducing the negative skin friction (Baligh et al., 1978). The effectiveness of slip layer in reducing dragload depends on the characteristics of the pile, the type of soil strata through which pile passes and the properties of coating material itself. In case of fine grained soils, the shearing behavior depends on the average rate of soil settlement. In case of coarse grained soils, soil particles may penetrate into the coat during pile driving. The particle penetration may adversely

2 affect the efficiency of coat in reducing the skin friction. The ideal coating material should have low viscosity to permit the slippage of soil surrounding the pile shaft and at the same time it should have adequate strength to adhere the pile shaft during storage and pile driving. The cost of coated pile can be much higher than that of uncoated pile (Briaud and Tucker, 1997). The granular soil penetrates into coat during pile driving and may result in scrapping off the coat and higher skin friction. Therefore it is important to study the effectiveness of coating material in reducing the interface friction between pile material and granular soil. The selection of type of coat and thickness forms an important aspect of pile design for dragload mitigation and overall economy of the project. This paper discusses laboratory tests to compare the performance of two types of bituminous coats in reducing the skin friction at soil-pile interface. The results of soil-pile interface tests on coated and uncoated specimens are compared with the pull out tests. LABORATORY TESTS Interface Friction Tests Interface friction tests were conducted by modifying the conventional direct shear apparatus as shown in Fig. 1. Pile material was represented by a mild steel block or a concrete block of mm. The properties of granular soil used in study are listed in Table 1. The granular soil used is classified as poorly graded sand (SP) and hereafter referred as sand. Normal load Top half of shear apparatus Fig.1. Interface friction test between coated pile material and sand. Table1. Properties of granular soil D 50 Mild steel / concrete block 28 D 10 C u C c G s γ max γ min (mm) (mm) (kn/m 3 ) (kn/m 3 ) Grid plate Sand at 70% relative density Coat

3 Two types of coating materials were used in this study. The Shalikote is a dispersion of selected grades of bitumen in water and is used as a protective coating over steel to prevent rusting. The Shalikote is applied cold on a surface. It can withstand temperature variations and vibrations. The bitumen coat used in present study had a penetration value between 30 and 40 and softening point between 55 C and 60 C. The first set of tests was conducted to measure the peak and residual shear stress at the interface of pile material and sand. The top half of the direct shear apparatus was placed on a mild steel block and secured in the position using locking pins. Sand was placed in the top half of the direct shear apparatus at 70 percent relative density by pluvial deposition technique. The normal stress was then applied on the sand surface and sample was sheared at 0.25 mm/min rate of shear. The interface friction between sand and concrete block was measured using identical procedure. The second set of tests was conducted to study the reduction in the interface friction by coating the blocks of pile material with bitumen or Shalikote. The bitumen was heated to 150 C and poured in the 60 mm 60 mm mould placed on the top of block. The coat was allowed to remain in the mould for 24 hours. The mould was then removed and the top half of the direct shear box was placed on the block without disturbing the coat. In case of Shalikote, the semi solid coat was thoroughly mixed and applied cold at a uniform thickness inside the 60 mm 60 mm mould placed on top of block. The Shalikote took more than 24 hours to cure. The end of curing is indicated by the change in the color of coat from brown to black. The initial coating thickness of 2 mm, 3 mm and 5 mm reduced to 1 mm, 1.36 mm and 2.16 mm after curing. Sand was then directly deposited at 70 percent relative density using the same technique on the top of bitumen or Shalikote coated block. Manual compaction was avoided to ensure the identical properties of coating material before applying the normal stress. After placing sand, normal stress was applied. The top half of direct shear apparatus was then lifted with the help of three lifting screws so that it remains just above the top of coat as shown in Fig. 1. Soil was then sheared against the coated block at an ambient temperature of 31 C. In all 42 interface friction tests were conducted as described in Table 2. Table 2. Test program of interface friction tests. Pile Coat Type Coat Thickness Normal Total Tests Material (mm) Stress (kpa) Uncoated - 3 Mild steel Bitumen 2, 3, 5 25, 50, =9 Shalikote 1, 1.36, =9 Uncoated - 3 Concrete Bitumen 2, 3, 5 25, 50, =9 Shalikote 1, 1.36, =9

4 The objective of above test program was to observe the effect of the pile material, the type and thickness of coat and the normal stress on the interface friction. In addition few tests were carried out to study the effect of the rate of settlement on interface friction. The tests discussed so far were conducted using dry sand and at an ambient temperature of 31 C. Dragload is most common in low lying areas where filling is required to raise the ground level. Soils in such areas are saturated. The effect of presence of water on the interface friction of coated and uncoated specimens was also studied. Model Pile Pull Out Tests Model pile pull out tests were performed to simulate the negative skin friction on coated and uncoated piles. The tests were carried out since the interface friction tests do not perfectly simulate the load transfer-displacement behavior of circular piles. In pull-out tests, the side friction generated would act downward on the pile as in the case of pile subjected to dragload. In addition, the relationship between the pull-out test and skin friction has been confirmed through field tests conducted by Keenan and Bozuzuk (1985) and Indraratna et al. (1992). Pull out tests have also been used by Chow and Wong (2004) to study reduction in the dragload using polyethylene sheets. The experimental set up used for the pull out test is shown in Fig. 2. Loading Frame Mechanical Jack Lifting Frame Load Cell Dial Gauge Model Pile Sand 500 mm 830 mm Fig.2. Model pile pull-out test set up.

5 The tests were conducted in a circular steel tank of 830 mm diameter and 500 mm height. The zone of influence of pile installation and loading depends on the density of soil and is reported to be in the range of 3 to 8 pile diameter (Meyerhof, 1959; Kishida, 1963; Robinsky and Morrison, 1964). In the present tests, the dimension of the tank provides a minimum lateral clearance of 10 diameters and satisfies the above criterion. An aluminum pipe of 38 mm diameter was used as a model pile. The poorly graded sand used in the interface friction tests was used in pull out tests. A homogeneous deposit of sand with 70 percent relative density was obtained by the pluvial deposition technique. The model pile was then driven into the soil up to a depth of 450 mm. A light weight hammer weighing 21.6 N and a 300 mm height of fall was used to drive the pile. The pull out test was conducted one hour after driving the pile. Bitumen used in interface friction tests was used in pull out tests on coated piles. The bitumen was heated to 150 C and applied on pile surface in uniform thickness of 2, 3 and 5 mm using a brush. The bitumen coat was allowed to cool before commencing the pile driving. RESULTS AND DISCUSSION The typical shear behavior of uncoated, shalikote coated and bitumen coated mild steel block with sand is shown in Fig.3. Tests with Shalikote showed initial increase in the interface friction followed by a substantial reduction as sample was sheared beyond 2 mm. In case of bitumen, the friction increased to a maximum value and then remained constant. The results of the interface friction tests carried out to study the effect of pile material, type and thickness of coat and normal stress are presented in Tables 3 and 4. Table 5 compares results of interface friction and pull out tests in terms of peak and residual shear stresses and relative movement required to mobilize these stresses. The interface friction of uncoated specimens of the concrete block and sand was higher than that with the mild steel block whereas after coating, the frictional resistance is almost same irrespective of the surface. In case of the bitumen, the peak and residual friction was of equal magnitude. Tests on coated specimens showed that the full friction is mobilized at a relative movement of 1 to 2 mm. In the present study the shear stresses at 6 mm relative deformation are compared for evaluating the reduction in the interface friction. The bitumen coated specimens showed 85% to 97% reduction in the shear stress when compared to the uncoated specimen. The Shalikote showed reduction in the interface friction by 20% to 70% to that of uncoated specimens. The analysis of the test results show that the bitumen coat achieved maximum reduction in the interface friction for all normal stresses and all thicknesses.

6 The available literature shows that the rate of shear has no effect on interface friction (Heerema, 1979). However these studies are limited to the uncoated construction materials and soils. The effect of the rate of shear on the interface friction for the mild steel block coated with 2.16 mm Shalikote and sand is shown in Fig. 4. The tests show that the interface friction is directly proportional to the rate of shear. Shear Stress (kpa) Uncoated 2.16 mm Shalikote 2mm Bitumen Horizontal Movement (mm) Fig.3. Typical shear behavior of uncoated and coated specimens. Residual Shear Stress (kpa) Rate of Shear (mm/min) Fig.4. Interface friction of sand and Shalikote coated mild steel block. Table 3. The residual shear stresses (kpa) for bitumen coated specimens Normal Uncoated Stress M.S. Block Bitumen Coated Mild Steel Block Uncoated Concrete Bitumen Coated Concrete Block (kpa) 2mm 3mm 5mm Block 2mm 3mm 5mm Table 4. The residual shear stresses (kpa) for Shalikote coated specimens Normal Uncoated Shalikote Coated Mild Uncoated Shalikote Coated Stress M.S. Block Steel Block Concrete Concrete Block (kpa) 1mm 1.36mm 2.16mm Block 1mm 1.36mm 2.16mm The sand particles penetrate into coat under normal stress and shear stress which result in the visco-frictional behavior of coat. Therefore interface friction developed on the coated pile can be expected to vary with the rate of soil settlement. The interface friction at soil-pile interface is expected to be high immediately after placement of fill because of the faster rate of settlement.

7 Table 5. Comparison of interface friction and pull out tests Pile Coating type material Steel surface Concrete surface Peak shear stress (kpa)* Residual shear stress (kpa)* % Reduction in residual shear stress Displacement at peak shear stress (mm) Minimum displacement at residual shear stress (mm) Uncoated mm Bitumen 3 mm mm mm Shalikote 1.36 mm mm Uncoated mm Bitumen 3 mm mm mm Shalikote 1.36 mm mm Uncoated mm dia. 2 mm Model pile 3 mm (aluminum) Bitumen 5 mm *The peak and residual shear stresses correspond to the normal stress of 50 kpa in case of the interface friction tests. In case of pull out tests, the average radial stress is estimated as 8.5 kpa and the values of residual shear stresses reported are extrapolated for 50 kpa radial stress.

8 The results of the interface friction tests carried out to study the effect of saturation in case of the coated mild steel block and sand are presented in Table 6. From Table 6 it is observed that the saturation has no significant effect on the interface friction of the uncoated mild steel block and sand under identical normal stress. The interface friction of the Shalikote and bitumen coated specimens in saturated test condition was 42% and 380% more compared to the dry test condition. The substantial increase in the interface friction can be attributed to the fact that the ambient temperature in case of the tests carried out under saturation was 25 C as against 31 C in case of the uncoated specimens. As the ambient test temperature reduced from 31 C to 25 C, the viscosity of bituminous coats and the interface friction increased. The bitumen coat appears to be more sensitive to temperature than Shalikote. Table 6. Effect of saturation on coated and uncoated specimens Type of specimen Residual shear stress (kpa) Remarks Under normal stress of 50 kpa Dry Saturated Uncoated No significant change Shalikote coated % increase in shear stress Bitumen Coated % increase in shear stress The results of pull out tests on model uncoated and bitumen coated pile are shown in Fig. 5. The pull out load was normalized to the average effective vertical stress to calculate the shaft friction parameter β as defined by Burland (1973) and the relative movement is expressed as a percent of the pile diameter. The bitumen coating reduced the shear stresses by 90 to 98% to that of uncoated pile. The shear stresses mobilized on 2 mm and 3 mm bitumen coated pile are practically same. Sand was carefully removed without disturbing the pile after the test. A thin layer of bitumen coat was removed using a hot knife to check the extent of particle penetration into the bitumen coat. The particle penetration was observed near the surface of the coat and most of the coat was free from sand particles. The pile movement necessary to mobilize the maximum shear stress was 10% and 1% of pile diameter for uncoated and bitumen coated piles respectively. The failure in the bitumen coated piles takes place within the coat itself. The shear strength of coat is very low therefore the pile movement necessary to mobilize the full strength at the interface is less compared to that of the uncoated pile. The present investigation has a limitation of the scale effect. The non-dimensional shaft friction parameter (β) will be lower for the large diameter piles installed in the field (Turner and Kulhawy, 1994). The relative displacement required to develop the maximum shaft friction is however independent of the pile diameter (Vesic, 1964) and free from any scale effects. The aim of this study was to find the effectiveness of bituminous coat in reducing the dragload. As the percentage reduction is significantly high it is believed that even in the case of field piles the percentage reduction can also be significant.

9 Uncoated Pile 2mm Bitumen Coated Pile 3mm Bitumen Coated Pile 5mm Bitumen Coated Pile Pile Movement as percent of pile diameter CONCLUSIONS Fig.5. Results of pull out tests The paper presents laboratory tests to study the effect of the pile material, normal stress, the type and thickness of coat, rate of shear and saturation on the interface friction. The following conclusions were deduced from this study. 1.Based on the interface tests, the bitumen and Shalikote coated specimen showed 85% to 97% and 20% to 70% reduction in the shear stress respectively. 2.The interface friction with Shalikote is directly proportional to the rate of shear. This behavior may be attributed to the visco-frictional nature of coat when sand particles penetrate into the coat. 3.The interface friction of Shalikote and bitumen coated specimens in the saturated test condition was 42% and 380% more compared to the dry test condition under the equal normal stress. The substantial increase in the friction of coated surface is due to the increased viscosity of the coat with the reduction in the ambient temperature. 4.The interface friction of coated pile and soil is independent of pile. The magnitude of interface friction at a given normal stress primarily depends on the type and the thickness of coat, and the rate of shear. 5.The displacement required for the mobilization of the peak shear resistance is much smaller (about 1 to 1.6 mm) in the case of the interface tests on coated as well as uncoated pile surfaces. 6.In case of pull out tests on the model piles, the displacement required for the mobilization of the peak shear resistance of an uncoated pile is much higher (about 6 mm) compared to that of coated model pile (< 1mm). The displacement in case of coated pile was always small as it depends on the peak deformation required to mobilize the strength of coating material. 7.In the case of uncoated pile surface the frictional resistance is found to increase with the normal stress (or radial stress) whereas for the coated surface the frictional resistance is practically independent of the normal stress.

10 References Baligh, M.M., Vivatrat V., and Figi, H. (1978). Downdrag on bitumen-coated piles. Journal of Geotechnical Engineering, ASCE, 104 (11), Briaud, J.L. and Tucker, L.M. (1997). Design and construction guidelines for downdrag on uncoated and bitumen coated piles. NCH Rep. 393, Transportation Research Board, Washington, D.C., pp Burland, J.B. (1973). Shaft friction of piles in clay- a simple fundamental approach. Ground Engineering, vol. 7, Chow, S.H. and Wong, K.S. (2004). Model Pile Pull-Out Tests Using Polyethylene Sheets to Reduce Downdrag on Cast In Situ Piles. ASTM Geotechnical Testing Journal, 27 (3), 1-9. Heerema, E.P. (1979). Relationship between wall friction, displacement velocity, and horizontal stress in clay and sand, for pile drivability analysis. Ground Engineering, 12 (9), Indraratna, B., Balasubramaniam, A.S., Phamvan, P., and Wong, Y.K. (1992). Development of negative skin friction on driven piles in soft Bangkok clay. Canadian Geotechnical Journal, vol. 29, Keenan, G. H. and Bozuzuk, M. (1985). Downdrag on a Three-Pile Group of Pipe Piles. Proceedings of the 11th International Conference on Soil Mechanics and Foundation Engineering, San Francisco, 2, Kishida, H. (1963). Stress distribution by model piles in sand. Soils and Foundations, 4 (1), Meyerhof, G. G. (1959). Compaction of sands and bearing capacity of cohesionless soils. J. Soil Mech. Found. Div., ASCE, 85 (6), Robinsky, E. I., and Morrison, C. F. (1964). Sand displacement and compaction around model friction piles. Can. Geotech. J., 1(2), Turner, J.P. and Kulhawy, F.H. (1994). Physical Modeling of Drilled Shaft Side Resistance in Sand. ASTM Geotechnical Testing Journal, 17 (3), Vesic, A. S. (1964). Model Testing of Deep Foundations and Scaling Laws. Proceedings of the North American Conference on Deep Foundations, Mexico City, 2,