Bored Sockets in weathered Basalt L. Maertens Manager Engineering Department Besix, Belgium Associate Professor Catholic University Leuven Keywords: sockets, open-end piles, tensile piles, basalt ABSTRACT: Offshore structures are often supported by open-end piles installed from marine equipment such as Self Elevating Platforms (S.E.P.). Depending on the subsoil conditions, the penetration of the driven piles can be sufficient to resist uplifting forces or not. In the case of very hard-cemented soils or weathered rocks with underlying sound rock layers, it becomes impossible to install the piles to a depth ensuring sufficient friction to resist uplift forces. It is then needed to install pre-stressed rock anchors inside the open-end pile or bore sockets beneath the open pile tip and install a reinforced concrete pile in this bored hole. These sockets can be drilled by rotation or by percussion. The present paper deals with the installation of 61-mm diameter bored sockets through open-end piles (dia. 762-mm) in weathered and sound basalt. The working compression loads reach 4 kn and the tensile loads 2 kn. Design, testing and installation of the sockets will be discussed. 1 INTRODUCTION India s first L.N.G. Terminal was constructed on the western coast along the Arabian Sea, about 16-km south of Mumbai. Steel open-end piles support the 175-m long jetty, the jetty head, the berthing and mooring dolphins, the walkways and the navigation dolphins (see figure 1). All structures are designed to resist live loads, wave loads with a significant height of 9-m, currents of 1- m/sec and earthquake loads with a ground acceleration of.16-g. Figure 1: General view of the terminal The subsoil consists of three subsequent layers: Soft clay layer with a thickness between and 6-m. Weathered Basalt with a thickness between 1 and 5-m, and a RQD value varying between to 9%. Sound basalt with unconfined compression strength between 29 and 115 MPa. A significant problem is the definition of the installation procedure for the piles reconciling the requirement to guarantee an adequate bearing capacity and the requirement of limiting the deformation of the pile tip in such a way that the installation of the socket through the steel open-end pile remains possible without damaging the drilling equipment. This problem will be treated shortly in an addendum. The design values to be applied for the bond between the concrete sockets and the (weathered) basalt on one hand and the bond between the sockets and the steel pile on the other hand are a second problem. To better assess both problems, an onshore test campaign was organized.
2 BORING EQUIPMENT Figure 3: drilling hammer 3 SOIL CONDITIONS AT ONSHORE TEST LOCATION Boring at test location TCR RQD (%) 1 2 3 4 5 6 7 8 9 1 TCR 5 RQD Figure 2: RCDS-3 drilling hammer The drilling equipment used in Dabhol is specially designed and built for the site by Geotec International (Belgium) and consists of a Numa Reversh Circulation Hammer (Massachusetts, USA) combined by a RCD rotary head (NCB, Italy). It allows for boring 61-mm diameter sockets in the weathered and sound basalt trough the 762-mm diameter pile. The RCDS-3 drilling equipment consists of (see figure 2): 1. Casing clamp 2. Working platform 3. Raking cylinder 4. Mast inclination cylinder 5. Rotary head 6. Mast 7. Pull-down hydraulic gear motor 8. Suction pipe 9. Drill rod 1. Casing 11. Stabiliser 12. Down-the-hole hammer DEPTH (m) 1 15 Figure 4: Boring at test location To perform the onshore trial pile test, a series of onshore borings were carried out in order to find a location with a geological profile as similar as possible to the available offshore borings. The aim was to find a location with a sufficient thick layer of weathered basalt. A typical boring at the test location is given on figure 4. Nine unconfined rock core tests were performed, giving an UCS of respectively: 72,5 43,1 61,6 5,8 113,3 52,6 65,1 3. and 42,8 MPa
4 STATIC TENSILE TESTS: Average uplift (mm) Pile load test T2 (15-16/9/99) - Interpretation of socket load Tensile Load (kn) 25 5 75 1 125 15 175 2 First Loading Slope,1,2,3,4,5 Unloading,6 Load supported by friction on steel pile Second Loading Slope,7,8 Figure 5: Tensile test Two static tensile pile tests were performed on pile T1 with a socket of 6-m and on pile T2 with a socket of 3-m. A tensile test up to 4 KN was performed on pile T1 and one up to 2 KN on pile T2. As the result of the tensile test on both piles T1 and T1 was totally satisfactory, it was decided to carry out a pull out test on pile T2 up to 5 KN (limit due to the strength of the testing frame). The pile characteristics are given in Figure 6. Average uplift (mm) Figure 7 : Test result T2 (2 KN) T2 - superposition of loading and unloading curves to 2 kn and 5 kn Tensile Load (kn) 5 1 15 2 25 3 35 4 45 5,5 1 1,5 2 2,5 Working load 3 Testpiles 762x16 mm Sockets 61 mm 9,59 9,6 T1 T2 8,54 8,59 3,5 Figure 8: Test result T2 (5 KN) Pile load test T1 (1-11/9/1999) - Loading and unloading curves Tensile Load (kn) 5 1 15 2 25 3 35 4, First Loading Cycle,5 5,95 4,35 Average Uplift (mm) 1, 1,5 2, Second Loading Cycle 2,5 3, 1,35 Figure 9 : Test result T1 (4 KN) -,5 Figure 6: Test piles The results of T2-test, T1-test and pullout test are given in figures 7, 8 and 9. Results of the tensile test: Pile Socket length (m) Wor king load (kn) Deflection at working load (mm) Test load (kn) Deflection at test load (mm) T1 6. 2 1.5 4 2.8 T2 3. 1.5 2.7 T2 3. 1 5 3.4
As pile tensile capacity is not only generated by friction on the socket, but also by friction on the steel pile, and since the uplift design is neglecting the latter, it was advisable to split up both. In Figure 7 one can distinguish two slopes in the loading curve. We assumed that the change in slope corresponds to the start of mobilisation of the friction on the socket One can see that the uplift resistance generated by the dead load together with the friction between the steel pile and the weathered basalt was found to be 125-kN. This corresponds to a bond stress between the steel and the basalt equal to 115 kpa or,115 MPa. According to Hobbs and Healy [1], Tchepak [8] and to Tomlinson [2], the ultimate skin friction for driven tubular steel piles were published as follows: Type of soil Ultimate skin Weathered chalk [1] Weathered to unweathered chalk [1] Weak coral [2] Moderately strong sandstone [2] Weak calcareous sandstone [2] Medium weathered siltstone [8] friction (MPa),26,1 to,1,45,28,45 3 to 5 This shows that the uplift resistance due to the friction between the steel pile and the weathered basalt is probably over-estimated, but this is on the safe side for the estimation of the ultimate skin friction between the socket and the basalt. The remaining uplift force (375 kn) was supported by the socket; the ultimate skin friction between the concrete of the socket and the basalt was at least,65 MPa. 5 CALCULATION OF THE UPLIFT FORCES. The calculation of the uplift forces for sockets bored in rock is complicated since the ultimate skin friction and the bond between concrete and rock depends on many factors: q uc : unconfined rock strength (MPa) Discontinuities and fractions in the rock Concrete strength (MPa) Socket roughness, expressed by the mean roughness height Δr (mm) or by the roughness factor RF (non dimensional) E m and ν: rock mass modulus (MPa) and Poissons ration. R s: socket radius (m) The length to the diameter ratio of the socket The mass of mobilized soil Installation method of the socket. In the Dabhol case, following parameters are known (Test pile T2): q uc = 36 MPa : this is the characteristic calculated from the test values given in 3 after deleting the highest and lowest value Discontinuities : the boring results given in figure 4 show over the socket length RQD values between 4 and 95%. This corresponds to a number of fractures per meter between 1 and 1. Socket length: 3m R s =,35-m (socket radius) Mass of mobilized soil can be calculated from the socket depth (7,24-m), socket length (3-m) and the rock density (= 22,5 kn/m³) Installation method: percussion without use of drilling fluid. According to Seidel and Collingwood [3] a reduction factor η c on the ultimate skin friction is recommended as follows (table below). Construction method Control η c No drilling fluid High level 1. No drilling fluid Low level.3.9 Bentonite slurry High level.7.9 Bentonite slurry Low level.3.6 Polymer slurry High level.9 1. Polymer slurry Low level.8 Indicative values for reduction factor, η c 5.1 Design Method proposed by Tomlinson [2] This method is based on research by Williams and Pelles [4], Rosenbergh and Journeaux [5], and Hobbs [6] and is given in figure 1. α,9,8,7,6,5,4,3,2,1 Rock Socket skin friction according to Tomlinson α-values Rosenbergh and Journeaux fs=αβquc Williams and Pells β-values by Hobbs Fractures per meter 15-2 8-15 5-8 1-5 1,1 1 quc (Mpa) 1 1 Figure 1 For q uc = 36 MPa α value is between,5 and,1 β value is between,6 and,9 An average value for the ultimate skin friction is given by: ƒs =,75 *,75 * 36 = 2,2 MPa β.4.6.7.8.9
The ultimate uplift force (Fs) is: Fs = π *,61 * 3 * 2,2 = 11,61 MN In case of a low control level during construction, a reduction factor η c =,3 to,9 has to be applied. For an average value of,6 the ultimate friction becomes 1,21 MPa and the ultimate uplift force 6,97 MN. 5.2 Design Method proposed by Horvath [7] Horvath et all [7] developed a new factor characterizing the roughness of the socket wall: Δr h RF = * Rs Lt L s Δr h = mean roughness height L t = total travel distance along the socket wall profile L s = length of the socket = 3 m According to Seidel and Collingwood [3], the average value of Δ r h varies between 5 and 1-mm for quality boring without producing artificially grooves. For the ultimate skin friction, one proposes: α =,8 * (RF),45 and f s = α * q uc This means that for a boring without grooves (Δ r h = ), the skin friction becomes zero and thus no bond between concrete and rock is considered, which is very conservative. To better understand the influence of grooves, let s consider a triangular groove with a depth d each 1- mm (see figure 11): α,26,24,22,2,18,16,14,12,1,8,6,4,2, Groove Shape: d 1 mm 3 mm (dotted) Alfa Factor in function of Groove depth f s = α x q uc 2 4 6 8 Groove depth d (mm) Figure 11 3mm Pile 6mm Pile 12mm Pile This figure shows that the performance of a socket can be improved significantly by creating artificially grooves in the socket wall. As one can observe the skin friction factor α is also independent 1 from the strength of the rock. This is in contradiction with the approach given in 5.1. The graphs in full line are the values that correspond to the assumption of one groove each 1 mm, and the graphs in dotted line to one groove each 3 mm. This shows that increase the depth of the grooves is much more efficient than increase density. Considering for the Dabhol T 2 case a Δr h value of only 3-mm, we find α =.14 and f s = 5,4 MN/m² which gives F s = 28,98 MN. 5.3 Design method proposed by Seidel and Collingwood [3] The authors consider a coefficient called SRC (Shaft Resistant Coefficient): n Δ SRC = η * (1 + ) * c ν d rh s with n = Em/q uc (rock mass modulus to the unconfined compression strength). n 1 ν = Poisson Ratio ν =,25 By calculations using the software ROCKET, the authors conclude as follows: f s = α * q uc with α according to figure 12. α,4,35,3,25,2,15,1,5, Adhesion Factor from SRC f s = α x q uc quc =.5 MPa quc = 1. MPa,,2,4,6,8 1, 1,2 1,4 1,6 1,8 2, SRC Figure 12 In Dabhol T 2 case: ηc = 1 ; n = 1 ; ν =,25 ; Δ r h = 3 mm ; d s = 61 mm SRC =,38 α =,12 f s =,12 * 36 = 4,32 Mpa F s = 24,83 MN quc > 3. MPa
5.4 The mass of mobilized rock The mass of the mobilized rock depends on the shape of the considered rock conus. The angle of failure can be considered ϕ 1 = 3 for weathered basalt and ϕ 2 = 45 for sound basalt. By limiting the uplift force to the weight of this rock conus (see formula in figure 13), one neglects the cohesion or bond stresses at the failure surface, which is conservative. Uplift Conus Test Pile 2 8,59 5.6 Bond stress socket-steel pile In many cases, stud-bolds are provided to guarantee the load transfer from the concrete socket to the steel pile. According to B.S. 54, Part 5 shear connections can be avoided as far as the ultimate bond stress does not exceed,4 MPa in the case of concrete poured in a cylindrical steel pipe. In the T 2 pile, the length of the socket plug inside the pile was 2,3-m, giving a bonded surface of 5,3 -m² and an ultimate bond capacity of 5,3 *,4 = 2,12 MN. The applied force was at least 3,75 MN, giving a bond stress of,71 MPa, or 1,75 times the ultimate bond stress according to B.S. 54. L1 Weathered Basalt ϕ 1 = 3 6 ADDENDUM: INSTALLATION 3 PROCEDURE FOR PILES. 4,35 2,59 45 L2 Basalt ϕ 2 = 45 1,35 W = π/3*γ rock [ tg²ϕ 1 * ( (L1+L2) * tgϕ 2 /tgϕ 1 +L1*(1-tgϕ 2 /tgϕ 1 ) )³ + tg²ϕ 2 * (L2)³ * (1-tgϕ 2 /tgϕ 1 ) ] Figure 13 5.5 Evaluation of the discussed design methods Uplift capacity against socket length for testpile T2 L1= 6m - Pile embedded over 4.24m Uplift force (kn) 3. 28. 26. 24. 22. 2. 18. 16. 14. 12. 1. 8. 6. 4. 2. Socket Friction (Horvath), 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, Socket length L (m) Figure 14 Build in Safety Weight of the rock conus: s = 1.2 Friction socket-rock: s = 3 Weight of rock conus Socket Friction (Seidel) Socket Friction (Tomlinson) In figure 14 the allowed working load of the pile T2 is plotted against increasing socket length. A safety factor of 3 is applied on the skin friction and 1.2 on the weight of the corresponding rock conus. One can see that in this case the governing criterion is the weight of the rock conus, as far as the length of the socket is smaller than 3,25-m (Tomlinson), 7,5-m (Horvath) and 6,75-m (Seidel). Figure 15: Damaged pile tip Damage of pile tip as shown in figure 15 cannot be accepted since excessive damage of pile tip prevents the installation of the sockets through the piles. Driving analysis by TNO-WAVE (Pdp Wave) can predict for the considered soil profile the SRD (Static Resistance during Driving) as well as the stress in the pile during driving, for different Hammer Energy levels and different penetrations per blow. In figure 16, the results of this analysis are shown for a compression pile (76 * 16-mm): It shows that the stress during driving decrease significantly when the hammer energy is reduced. For a S9 hammer (hydraulic hammer from IHC), one can see that for an SRD value of 525 KN, the driving stress is 35 MPa for a full energy setting of 9 KJ and is reduced to 26 MPa when the setting of the energy is reduced to 45 KJ.
On another hand, the number of blows is increased from 28 to 18 blows per 1-mm penetration. This means that the driving time is almost 4 times longer as the number of blows remains 5 blows per minute. Field test shows the following damage: Pile Thickness (mm) Max blows for 1-mm penetration (at 9 kj) Maximum Driving Stress (from output of TNOwave model) (MPa) Damage at toe (m) 1 16 51 38.1 2 16 5 38.5 3 19 57 36.5 1 9 8 Definition of refusal for compression pile 16 mm Maximum SRD = 2 Maximum Working Load = 2 * 262 kn = 525 kn 9 kj Finally the installation procedure was as follows: Pile (% of full energy) 5 75 5 Compression Tension Refusal criteria for permanent works Hammer Energy 16-mm 19-mm 16-mm (*) 19-mm (*) (kj) 45 67.5 45 45 Blows per 1-mm penetration 1 1 2 (*) This criterion was checked by installing two additional raking piles on the test location onshore. After inspection, no damage at pile tip was observed as shown below in figure 17. 5 4 7 67,5 kj SRD (kn) 6 5 4 3 45 kj 2 1 1 2 3 4 5 6 7 8 9 1 11 12 13 14 15 Blows per 1 mm penetration 45 Maximum Driving Stress (MPa) 4 35 3 25 2 15 1 2 3 4 5 6 7 8 9 1 11 12 13 14 15 Blows per 1 mm penetration Figure 16 Yield stress = 415 MPa Allowable stress = 332 MPa As one can see, the maximum stresses during driving were close to the yield stress (415 MPa). In fact these maximum driving stresses are computed by the IHC model with the assumption that the stresses are uniformly distributed over the entire cross section. This is of course never true in reality, and an appropriate safety factor has to be used in the definition of the refusal criteria. Final installation criteria to guarantee the required SRD are governed by in-depth stress and damage analyses. It was concluded to allow 8% of the yield stress (= 332 MPa) for compression piles and 55% (= 225 MPa) for tension piles, since tension piles need a socket. 7 CONCLUSIONS Figure 17: Pile tip after driving The most appropriate foundation method for offshore structures in cemented soils or weathered rocks is a foundation on tubular steel pile. Due to the problem of penetrating the piles to sufficient depth, bored sockets are often needed to resist uplifting forces. This leads to following problems: 1. Pile driving criteria for piling to guarantee a very low damage level on the pile tips. This can be managed by driving analyses using appropriate software and in-situ testing. 2. Design methods for sockets are not yet standardized and existing methods are giving a large dispersion of results. The weight of mobilised rock is often governing the design. In situ tests are needed to confirm the calculations.
REFERENCES [1] HOBBS, N. B. and HEALY, P. R. Piling in chalk, Construction Industry Research and Information Association (CIRIA), Report PG6, 1979. [2] TOMLINSON, M. J., Pile design and construction practice, E & FN Spon, London, 1995. [3] SEIDEL, J. and COLLINGWOOD, B., The SRC method for estimating side resistance of drilled shaft, The Magazine of the Deep Foundations Institute, Fall 22. [4] WILLIAMS, A. F. and PELLS, P. J. N. Side resistance rock sockets in sandstone, mudstone and shale, Canadian Geotechnical Journal, Vol. 18, 1981, pp. 52-513. [5] ROSENBERG, P. and JOURNEAUX, N. L. Friction and end bearing tests on bedrock for high capacity socket design, Canadian Geotechnical Journal, Vol. 13, 1976, pp. 324-333. [6] HOBBS, N. B. Review paper Rocks, Proceedings of the Conference on Settlement of Structures, British Geotechnical Society, Pentech Press, 1975, pp. 579-61. [7] HORVATH, R. G., KENNEY, T. C. and KOZICKI, P. Methods for improving the performance of drilled piers in weak rock. Canadian Geotechnical Journal, Vol. 2, 1983, pp. 758-772. [8] TCHEPAK, S., CHIN, M.C. Statnamic testing of bored piles socketed into siltstone, BAPIII, Ghent, 1998. [9] MAERTENS, L. Design and installation of steel open end piles in weathered basalt, International Deep Foundations Congress, Orlando, 22