You are on page 1of 17

A design method for drilled piers in soft rock

Geotechnicnl Research Centre, Fac~lltyof Engineering Science, The University of Western Ontario, London, Ont.,
Canada N6A 5B9

Golder Associates, 500 Nottinghill Road, London, Onr., Canacla N6K 3 P l
Received May 20, 1986
Accepted October 16, 1986
A new procedure for the design of drilled piers socketed into soft rock is presented and the selection of design parameters
discussed. The design method is based on (I) satisfying a specified design settlement criterion and (2) checking to ensure that
there is an adequate factor of safety against collapse. The design allows consideration of the possibility of some slip occurring
at the pier-rock interface under working load conditions.
The design procedure is illustrated by a series of detailed calculations relating to two piers in Queenston shale and the
observed behaviour of these piers is compared with design expectations. It is shown that these piers satisfied the design
settlement criterion while having a significant proven factor of safety against collapse.
Key words: rock, piers, piles, settlement, bearing capacity, design.

Une nouvelle procCdure pour le calcul de pieux forts dans la roche molle est presentee, et le choix de parambtres de calcul
est discute. La mCthode de calcul est basCe sur un critkre donne de calcul de tassement, et sur un coefficient de sCcurit6 adCquat
contre une rupture complkte. Le calcul permet de prendre en consideration la possibilitC d'un glissement i I'interface
pieu-rocher sous des conditions de charge de travail.
La procCdure de calcul est illustrCe par une sCrie de calculs dCtaillCs de deux pieux dans le schiste Queenston; leur
comportement observC est compare avec les prCvisions calculCes. L'on montre que ces pieux satisfont aux critbres de tassement
calcule tout en ayant un coefficient CprouvC de sCcuritC contre la rupture.
Mots clks : roche, pieux, tassement, capacitC portante, calcul
[Traduit par la revue]
Can. Geotech. J. 24, 126- 142 (1987)

Introduction Diameter of Recessed Pier

Historically, the design of drilled piers socketed into rock
(referred to herein as socketed piers) has been primarily based
on consideration of "bearing capacity" and construction con-
straints (Armitage and Rowe 1983). However, procedures for Surface
determining the "bearing capacity" of axially loaded drilled
piers in rock tend to be crude and conservative, as evidenced
by the fact that collapse is rare even in tests intended to deter-
mine bearing capacity. Because of the difficulty in causing
collapse, "bearing capacity" failure is often associated with the
displacement attaining an "excessive" value. Consequently,
piers designed on the basis of a de facto limited settlement Socket ,
criterion. (A similar situation exists when considering the I E b
bearing capacity of cohesionless sands (see e.g. Peck 1976).) Diameter i f Socketed Pier
Recently, a number of researchers have advocated design FIG. I. Notation for a recessed complete socketed pier in rock.
procedures involving relatively high allowable side shear
resistance values (compared with those of earlier empirical
design methods) in conjunction with the use of elastic theory uniform and when consideration is given to the potential vari-
for determination of settlement (e.g. Pells and Turner 1979; ability in side shear resistance itself, it can be expected that
Horvath 1980; Williams and Pells 1981). In the determina- under normal conditions (i.e. unless extremely low loads are
tion of settlement, these approaches implicitly assume that applied), some slip will occur along the socket irrespective of
the available side shear resistance is sufficient to avoid any whether it is considered in design (Rowe and Armitage 1984).
slip (i.e. failure) at the pier-rock interface under design Thus it would seem reasonable to adopt a design procedure that
conditions. does explicitly consider the effect of slip. This paper presents
As shown theoretically by Rowe and Armitage (1987), and one such approach.
as demonstrated by the field results of Williams (1980) and The paper firstly outlines some recent theoretical develop-
Horvath (l980), the majority of the load carried by a socketed ments that will allow for consideration of slip in the design of
pier will be carried in side shear until slip occurs at the inter- socketed piers. Secondly, selection of design parameters will
face. Once slip occurs, the majority of additional load is then be considered and a new design method proposed. This ap-
transferred to the base of the socket. In many instances, the proach will then be verified by comparison with the actual
shear stress distribution along the socket will be quite non- behaviour of two piers tested by Horvath (1980). Finally, con-

2. adjacent to this pier and a modulus E .. = I. 1. The pier. struction considerations such as cleaning and inspection of the rock below) side shear resistance. The pier has a Young's modulus E.5. = 0. The general problem being considered is shown in Fig. L . of diameter D . The pier may be recessed to a depth L. ( b )E b / E . Elastic settlement influence factors for a complete socketed pier: ( a ) E b / E . The rock has a mass modulus E . The dimensionless modulus E.e. assuming that the pier is not recessed (i. Initially. ( c ) E. It is assumed that the rock/soil to this depth has a socket using a finite element analysis. is bonded to the rock over a length L . below the Theoretical solutions pier. sockets are discussed. ROWE AND ARMITAGE: 2 FIG.0. below the ground/rock elastic solutions were developed assuming a fully bonded surface. = O). = 2./E. but does not contribute significant (compared with settlement I defined by .

The use of these solutions will be described in a later section. they developed solutions relating the settlement influence factor I at the head of the pier to the proportion of load trans- ferred to the base of the socket (Qb/Qt) for the situation where partial or complete slip has developed along the socket. D . The full set of these nonlinear solutions along with the elastic solutions given in Figs. ( c ) E. The situation where slip can occur at the interface has been examined in detail by Rowe and Armitage (1987). A limit state design philosophy is adopted. J . LC> 0) will result in a decrease influence of the loaded pier. 2-5 provide the theoretical basis for the design method presented in the following section. 1 for definition of socket head) and Q. and flexural failure may occur./E. the pier behaviour is perfectly elastic). A typical set of results is shown (see Fig.. Eh/E. The load interface. is the diam.).) (6) there are no voids or open cavities within the zone of Recessment of the pier (i. punching. 2-5 are for the limiting case where no slip occurs along the socket length L (i. in peak average side shear resistance at the concrete-rock etries (LID) and modulus ratios (E. hand method for the design of axially loaded drilled piers in soft rock. (The use of these charts will be described in ficially roughened. where p = average settlement of the head of the socket (see (4) the sides of the drilled shaft are clear of any contamination Fig.0./E. drilling mud) that could result in a decrease the head of the socket.e. pE.128 CAN. Also.g. = 2. Qb. carried in end bearing.) such side shear resistance at the concrete-rock interface may be that the settlement at the ground surface is given by estimated either from empirical correlation or from field load tests. Range of applicability-assumptions The proposed design method is considered appropriate for socketed piers in rock where FIG. (2) ensuring that there is an adequate "factor of safety" against the possibility of collapse. 1). a later section.D cant time-dependent (creep) behaviour at the design loads (e.0. The first portion of the expression given in [2] simply represents elastic compression of the pier over the length LC. (5) the sides of the shaft are in either a natural state or arti- tions from Fig. where I is the influence factor for L. Qt [2] propriate for rock that may be expected to exhibit signifi- . rock and no soft compressible material impairs this contact Qt (cleaning and inspection of sockets will be discussed later). (RF). (7) the expected mass modulus of the rock and average peak tion can be expressed in terms of a reduction factor (RF).D (3) the base of the pier is in direct contact with the underlying [l] I = . Eb/E.. yet rational.. = 0 (see Fig. is given in Fig. Elastic load distribution curves for a complete socketed ( l ) the load is axial and static in nature (cyclical loading has not pier: ( a ) homogeneous. rock smear. The solutions given in Figs.E. 4 and 5 for a range of situations.5. The probability of exceeding the design settlement can then be related to the choice of partial factors.The second portion of the expression gives the settle- ment due to rock deformation. E. been considered). in settlement at the head of the socket (see Fig. 4LcQt The approach described in this paper is not considered ap- p = -+ -1 (RF).e. Thus. = 0. crushing. 1987 eter of the pierlpile over the length LC. corals) in which a correction factor for the effect of recessment. The reduction factor (RF). = applied load at (e. design parameters are obtained by applying partial factors to the ex- pected values of the deformation and strength parameters.rrD.g. 3. 24. A wide range of solutions is given by Rowe and Armitage (1987). Design philosophy The design method described in this paper is intended to provide a simple. ( b ) E. 2)./E. (2) the load is carried in both end bearing and side shear. may be estimated for elastic condi. is given in Figs. gypsum) or for porous rock (e. = 1. VOL. Using an axisymmetric finite element anal- ysis. is rock salt.3. This reduc. 7). this . The approach is based on (1) satisfying a user-specified design settlement criterion (which will govern the design in most cases).g. 2 for a range of geom. CEOTECH.and all other terms are as previously defined.

used and may be correlated with the categories Rl-R4. Reduction factors for a recessed socket: (a and b) E.g. 5. particularly discontinuities. Kulhawy and Goodman 1980. ROWE A N D ARMITAGE: 2 FIG. preferably. R3. Thus both the deformation and strength characteristics will be directional and will reflect the properties of both the intact rock and the discontinuities. chemical swelling. faults. Peck 1976). (c and d) E. The design parameters can be selected to allow for this potential variation by applying partial factors to the expected value of the mass parameter (be it the modulus of the rock or the average socket side shear resistance) as described in the following sections. = 10. Geological considerations Soft rock will typically be anisotropic and nonhomogeneous. and Rowe and Armitage (1984). statistical variation in parameters may be expected to occur owing to minor local variations in conditions. and the construc- tion method used to form the socket. = 100. To assist in the selection of FIG. (1980). the roughness classification R l . Even on an apparently uniform site. prototype field tests at the particular site of interest. . For these reasons it is important that the design parameters be related to actual field behaviour by means of either empirical correlations or. since the mass parameters are dependent on the discontinuities. seams. careful consideration of the geological factors is required when selecting design parameters. the roughness of a socket can have a significant influence on the expected side shear resistance. As indicated by Pells et al. Variation insettlement reduction factor with modulus ratio. Horvath (1980). (1980) and given paper does not consider factors such as solutioning in lime. and bedding planes (see e./B. in Table 1 will be adopted in this paper. etc. parameters for a given situation. The roughness of the socket walls will depend on both geo- logical factors. subsidence. and R4 proposed by Pells et al. R2. having variable weathering with depth beneath the rock surface and naturally occurring discontinuities such as joints. which may vary appreciably even over a given site. Alternative roughness stone.4. that may be critical classifications such as that of Horvath could equally well be to the design of foundations in certain situations. However./E.

higher values may be adopted have been proposed (for a review. reduction factor of at least 0. where 'i = expected side shear resistance. and rock mass "allowable bearing pressure" of 2 . i. shear resistance and mass modulus as follows: recognizing that the majority of the deformations for founda- tions on most rock will occur during construction. see Pells and Turner 1980. may provide an initial estimate of the expected side same philosophy should also apply to foundations on rock. Particular care should be adopted when ben.OO mm deep than 3% in both cases. when supported by appropriate data. shear resistance. (In considering the allow- .. the sockets of each pier should be care- roughened sockets. The maximum pier head settlement. vari- sistance should be verified for the ex~ectedfield conditions able weathering.g. a field test on a prototype pier.[5] are Expected side shear resistance and rock mass modulus appropriate for a specific site. from a field test. the predicted shear resistance and mass modulus is by performing a field load bearing capacity of a circular footing on rock with + = 350 test On a prototype pier. 7 would ~ ~ seem reasonable. width greater than 10 mm. .=215< mance of the structure. For example. In general. the probability of exceeding a Roughness specified design settlement is less than 30% if partial factors class Description f. even adopting the lowest theoretical pre- Design values for side shear resistance and mass modulus diction and applying a factor of safety of 3 would give an Design values of side shear resistance T.5 p . pd = 0.5. seams. u. or other substance that will inhibit good bonding be. R2.. There has been considerable debate concerning the bearing Equation I[' be quite conservative for capWty of circular foundations on rock.). Thus.7 are adopted and less than 11% forf. in MPa. (Exception [3a] C =0 . the compression of the length LC should be calculated using a A clean socket is classified as a socket where there is no concrete modulus and design load that have been factored in significant auger smear (i. This assumption may be appropriate for likely to be of greater concern than the actual magnitude of the smooth sockets but can be quite conservative for sockets with settlement. Horvath (1980)) may be close to fully inspected in the field and the designs adjusted as necessi- that of clean sockets.... grooves or indentations The probability of exceeding twice the design settlement is less less than l . = aver. 1987 TABLEI .e.. data. R2 Grooves of depth 1 -4 mm. accordance with normal limit state design practice.) R1. which were developed for rock where there reasonable allowable design settlement. GEOTECH. in MPa. is then calculated by subtracting the compression of the pier or age uniaxial compressive strength of rock core (ASTM column above the actual socket (e.i. coincidentally. Careful consideration should be given to the adoption of a These correlations. ben. and rock mass modulus.. smooth-sided socket.130 CAN.) The design tonite. but the actual available side shear re.. R3 as defined in Table 1) It is suggested that the maximum groundline settlement [3b] . = 0. and E.7 be applied even when the at spacing 50-200 mm socketed pier design is based on parameters back-figured from R4 Grooves or undulations of depth greater than 10 mm.10 mm. based on purely theoretical considerations. p. side shear resistance should be neglected for any sockets In the design of socketed piers. cause the onset of first failure based on incipient failure theory [5a] T~ = f.i = 0 . see e. viz. = f E = 0. Roughness classification (after Pells et ell. The scatter in experimental results is also quite high resistance or to justify the adoption of side shear resistance and the available data does not lend to any values larger than those recommended above. in MPa. theory. In the absence of direct field test imum settlement divided by 2. test may serve to the ranges from to 3zUc depending upon which theory is adequacy of initial empirical estimates of available side shear adopted. Numerous theories roughened sockets. etc. the side the settlement of each pier will be the same. it is recommended that a R3 Grooves of depth 4. It is well accepted that were no open discontinuities within the zone of influence of the foundations on soil settle and still perform adequately. J. piers designed on the basis of Rowe and Armitage (1984) have correlated the socket side these parameters should be proof-tested to verify the design. an allowable bearing pressure at the base of the [5b] Ed = f E E . tated by differences in the rock at different sockets (e. settlement for the socketed pier is then taken to be the max- tween the rock and concrete. at spacing 50-200 mm Unless previous testing and experience indicate that the parameters empirically selected on the basis of [3]. If differential shear resistance of very rough sockets (this includes artificially settlement is critical. Nevertheless..g. Couetdic and Barron 1975) and there is a significant variation The most means of determining the side in the predicted bearing capacity.'i (see Pells and Turner 1980). differential settlement is that are not clean. VOL. To make allowances for potential variability in rock proper- at spacing 50-200 mm ties with both position and time. width greater than 2 mm. I) from the maximum groundline settlement. ANowable etld-bearing pressures tonite is used. as discussed later. 7 ~ This. modulus Ed should be obtained by applying a reduction factor also approximately corresponds to the pressure expected to (or partial factor) f to the expected values of . piers should be designed to ensure that roughness R 3 or R4 (see Table I). 1980) Statistical studies by Rowe and Armitage (1984) have shown that. The socket. pier of 2 . Clearly. 24. due to the length LC in D2938). (Note that rock.E.e. RI Straight. deduced from a large number of field tests with the average unconfined compressive Allotvable design settlement strength for weak rock deposits in which the socket is founded. based on available data. = expected rock mass modulus of the Fig. zone of remoulded rock).g. = f . 6 6 for clean rough (R4) sockets should be selected to be the maximum settlement (at the surface of the soil/rock deposit) that is consistent with good perfor- [4] E. Even with smear..456 for regular clean sockets (roughness must be made for rocks particularly susceptible to creep. width greater than 5 mm. ~ .

In the tests reported by Glos and Briggs (1983). no compressible/weak gouge-filled seams) having an average uniaxial compressive strength a ... there was no sign of signifi- cant yield at the base of the pier when the tests were terminated at base bearing pressures ranging from 0.960. it should be remembered that we are only concerned here with preventing significant failure of the rock below the base of the pier.5u. .e. and 2. q. and f.50.. to 1. the load carried by the pier was still increasing when the test was terminated. given by and a maximum bearing pressure at the base of the pier q. When these tests were terminated at loads ranging from 2. L = length of the socket. This requirement is intended to ensure that the rock beneath the base of the pier behaves "elastically" (i. and (iii) there are no solution cavities or voids beneath the pier. the base pressure q . = the design pressure at the top of the socket (which must not exceed allowable stresses in the concrete). To provide some measure of safety in terms of the ultimate capacity of the rock socket.290. In the tests reported by Horvath (1980).. defined as L L [gal q. In all cases.950. the load-displacement curve was still rising and collapse had not occurred. given by In the design procedure to be described in detail in the following section. ROWE AND ARMITAl able bearing pressure.. i = the expected side shear resistance along the socket.. The magnitude of the settlement is considered separately-in the follohing section and depends primarily on the mass modulus of the rock.) Table 2 summarizes the results of a number of well- documented field tests on recessed end-bearing piers with diameters of 0. .. it appears adequate to adopt an allowable bearing pressure at the base of the pier q..e.41~.350. first yield occurred at base pressures between 0..50.3 m or greater..250.-iD for q .4/. = a reduction .. < 30 MPa) where (i) the base of the socket is at least one diameter below the rock surface. to 10. i. r (4fsET) where q . determined from the design charts will be required to be less than the allowable bearing pressure q. and 1.. although the load-displacement curve was still climbing steeply when the tests were terminated at base pressures between 1.. the maximum base pressure. (ii) the rock to a depth of at least one diameter beneath the base of the pier is either intact or tightly jointed (i. For the purpose of designing piers socketed into soft rock (0. and 1. = q.) and in these tests some yielding of the rock (as inferred by the change in shape of the load-displacement curve) was observed at base pressures be- tween 1. without significant yielding) under design conditions.e... The tests performed by Williams (1980) were carried to higher load levels (relative to u..40. . D = diameter of the socket.

. it is recommended that f . Field lized shear resistance being less than that assumed in [8] is less tests may then be performed to obtain improved estimates of than 0. 5 b ) srEp 3 Modulus ratios E p / E d . 4 ) expected rock mass modulus beneath socket. this implies that the probability of the mobi.i. E p [Factored] Unconfined compressive strength of the rock.5%. Under these circumstances. 24./Er . 1987 factor for ultimate conditions.~+ q . cedure described below to develop a preliminary design. Thus. may excessive clay filled or weathered seams exist along the walls be very small. D applied load. Qt [Factored] Estimate : Modulus of pier. must be less than the maximum Proposed design procedure bearing pressure qhnl.132 CAN. see Eq. [8] is intended to ensure than even if only rock properties and the design can be revised as necessary. of the socket. socket diameter. Fig. Based on the statistical studies by Rowe and This information can be used together with the design pro- Armitage (1984). For a complete socket where the pier is not recessed and no Note that for lightly loaded piers with long sockets. Eb Select : Partial factors to be used in design fT . A preliminary geotechnical investigation should be under- taken to assess the in situ conditions of the rock mass and to determine a representative uniaxial compressive strength of the For the purposes of design.g. Ts / 7 ( L / D $ . 6): - STEP l Select : design settlement Pd . Eb/Er srEp 4 (L/D)mOx a Q t / ( % ~ 2 ~ d ~ I d ' PdEd D/Qt re there seams whose presence is not implicitly included in the rock mass modulus and side shear resistance? I NO Y S Estimate : Depth . the design method proceeds as follows (see also An acceptable design should satisfy both 171 and [9]. 3 ) expected rock mass modulus adjacent to socket ( ag. 30% of the expected side shear resistance was mobilized at the pier-rock interface. ~ ~ = ( L / D l m a X / ( l . the entire pier-socket sustem will still Design of an unrecessed complete socket have an adequate factor of safety against which seams extend L S Length of socket L i Proportion of seams S = C( seam thickness ) / L i Seam modulus E S Seam side shear resistance Ts Calculate: €. see Eq. Flow chart for the design procedure .e.. VOL. taken to be 0. GEOTECH. 5 0 ) Design mass modulus E d = f E . 7 FIG. [7] will govern. fE Calculate : Design s~de shear resistance 7' = f T .6.19) I. q.) ( ~ q 21) . E r (Eq. Uc STEP 2 Estimate : expected side shear resistance 7 (e.s) (~q. be rock as well as to determine any other appropriate parameters. J. = rd (I-s + SE~IE. ( Eq.3.

4 (E)d fs [ ~ q 1601 . 121 4ba = Uc [ ~ q 131 .L2) !/ Deslgn IS Complete FIG.e. -. . 7 = (I-s + Z 7 [ ~ q 1401 . (Concluded). ( Q b / Q t ) d from the ~ntersectionof the factored design line for (L/D). 2 . 4 or 5 Calculate 1: = Id / ( RF)p Determine a revised value of ( L / D ) d and ( Q b / Q t I d from Figs. ( L / D ) m o x ) L e t L 2 = L* (calculated using 12 ) L = min(LI. 3 4 Was a value of ( L / D I d found I YES Calculate L = ( L / D ) d .5 Uc [Eq.s)T . 151 L qbu = st . [ E ~ l. ROWE A N D ARMITAGE: 2 STEP 5 Determine: ( L / D ) d .e a ] qbm = 2.. . or ( L / D ) ~with ~ ~ the contour for Id( o r 1: ) using des~gn charts allowlng for slip (Rowe 8 Armitage 1987 ) YES I NO Determine the value of ( L / D ) d for given Id ( o r 1.) using Fig. using Id. 2 Deterrnine the corresponding ( Q b / Q t )d for ( L / D Id using Fig.D STEP 6 q t = Qt / ( ' X D ~ /) ~ [Eq..6 . Redesign required Go to step 1 YES +YES t NO L e t L I = L S + L o ( calculated above assuming S = 0 i. 3 Determine the recessment length L e YES Determine ( R F )p from Fig.

VOL. Id may be adjusted as described in the (ii) socket diameter. I. Select the appropriate chart from those given as Fig. Calculate the dimensionless modulus ratios E. the percentage of compressible/weak weath- Determine the following design parameters: ered seams expected along the shaft of the socket.. I. then values Fig./E. (i) allowable design settlement (see previous section).0 FIG. (m). FIG. it is necessary to check whether the pier can be designed for the given conditions. and then proceeding as (i) Using a representative value of u. 7). point: bility of exceeding the design settlement will depend on the (i) Draw a vertical line to the L I D axis and read off the partial factors adopted as indicated in the preceding section. Note .: I = I . axis and read off the E ~ I E . (EllEd)implies that the reduction factorf./E. Step I (c) Estimate S ./Ed and (ii) Draw a horizontal line to the Qb/Q. Proceed to step 6. (MPa)... length L . This is achieved by Step 2 selecting an appropriate design chart for the given values of (a) Select expected values of side shear resistance and rock E. Then from the intersection side shear resistance and rock mass modulus ([5]).. a.. from those produced by Rowe and mass modulus by one of the following methods: Armitage (1987) (e. GEOTECH. 2 (e. (a) Calculate the dimensionless socket length that would be (c) If no intersection point can be established on this design required if all load was to be carried in side shear: chart. determine ?. = E. is applied to If an intersection can be found..g.ID. Eb/E." the socket and E b (the modulus of the rock at the base of the which is a straight line between the coordinates (LID = 0. = 1. that the use of the ratio E. 24..8. f.e./E. Design of a complete socketed pier allowing for no slip conditions. For S # 0.) Q. for the rock adjacent to (a) On the design chart. then a pier of length Ld will all the rock. . satisfy the design settlement criterion and it simply remains to check that there is an adequate "factor of safety" against over- Step 4 stressing the rock beneath the pier. (b) Locate the intersection between the "factored design (b) Apply partial factors (i. (iv) modulus of pier (factored according to limit state rules). Qh/Ql = 100%) and (LID = Ltll.134 C A N .. J. Find the intersection of this line with the curve for the . pier) from [3] and [4]. E. p.g./E. "design" length-to-diameter ratio (LID). and f. The proba. . 1987 loo h _IFactored Design Lsne €b/[ = 1./E. = 1. (see Fig.accurate field information is available. while allowing for possible slip at the pier-rock interface.) to the expected line" and the curve corresponding to the "design" settlement values found in step 2(a) to obtain the design parameters for influence factor. the values of (LID). 8) and draw a horizontal line for (b) Calculate the "design" settlement influence factor. D (m). give the "design" settlement influence factor I.. 7... and E . should be used. E. of E . (E.. following section.. draw the "factored design line.. Step 5 (MPa). Design of a complete socketed pier allowing for slip./Ed and E. (MN). E.*D. see Fig./E. determined from tests on follows: intact core samples. This step involves selecting the length of pier required to (v) unconfined compressive strength of the rock. 7). (S = C (seam thickness)/expected socket (iii) applied axial load (factored according to limit state rules). see Fig. = (L/D). Calculate the corre- Step 3 sponding length L. 7).) rather than "design" ratio of load carried to the base (Qh/QI). Qh/Qt = 0) (see (ii) If more .

= 4. = 5 1 % are read using the following approximate methods. 7 shows that there is and an alternative estimate of the required pier length should be no intersection point between the factored design line and the determined as described below. then move horizontally to read off the value of (Qb/Ql)d(e. eter must be increased to reduce the base pressure.. the diameter D . a lower modulus and a lower shear strength than the adjacent mined in previous steps). Calculate: FIG. In step 5(a). interpolation will be necessary between the bounding design charts. 9). Going to Fig. This pier can be and simply extend the pier to a depth Li below this zone (i./E.: (i) If there is an intersection point on this curve then the pier can be designed elastically (i. no slip occurs at the interface). The corresponding proportion of load transferred to the base (Qb/Q. 9. = 0). 7 shows that there is been obtained (e.. must be followed. consider a pier with (LID). soft weathered seams may be encountered this step. A pier with this diameter D cannot be designed to satisfy the specified design conditions with I. design has been controlled by settlement considerations and a ment (if the selected value was unrealistically low) or increase suitable pier with length Ld and diameter D has been selected. or adopt a larger pier diameter (or both). Proceed to step 6 to check bearing pressures./Er = 50 and then horizontally to get (Qb/Ql). Based on this rigorous line and the curve for I = 0.25. > D . = 1.25 and the procedure described in step 5(c) exist along the proposed socket. Going to Fig. = 2. = 0. = 0.. ROWE AND ARMITAGE: 2 135 for I = 0. it appears that the effect of seams can be considered the values of (LID). Now consider three possible values of I./E. ./Er = 50 is found at (LID). lie between two bearing controls the design and either the pier length or diam- design charts (or curves) given by Rowe and Armitage (1987). 3 by drawing a vertical line from the selected value of (LID).. the designer forl.. = 0. to meet the curve for the appropriate value of E.. Go back to step 1 and either increase the design settle. 1 (where these values would have been deter. tion of the seams along the socket. from a design neglecting seams by fol- no intersection point between the factored design line and the lowing steps 1-6) and it is found that weathered seams will curve for I = 0.7. designed elastically (i. The required (LID). These seams will generally have and E ~ I E=. The the revised length of the pier L . = 4. have developed theoretical solutions that allow the designer to -If 1.e. (b) If either condition [14b] and [16b] is violated then end NOTE:In situations where values of E.. = 4. can be found from Fig..5 is rock. an intersection with the appropriate elastic curve may choose to neglect the portion of the rock with these seams for E.7 to the curve for E. (a) If both conditions [14b] and [16b] are satisfied then the tions./Er = 50. 7) and analysis.. curve for I = 0. will then be given by corresponding value of Qb/Q. In this case it is necessary to go back to step 1 and either reassess the choice of allowable settlement p. 8 and drawing a horizontal line (a) If these seams are localized to a depth L. Suppose that an initial estimate of the socket length Li has -If 1. Qb/Q.. can be obtained From earlier recommendations.) may be determined from curves given as Fig. it is required that by dropping a vertical line from the point of intersection to the LID axis.e.15.45 then the intersection point of the factored design examine the effect of seams explicitly. it is required that iI criterion. = 100%) and ((LID). Calculate the design length L. see I Fig. .e. The effect of these seams will depend on these two marked on the LID axis and the "factored design line" is drawn properties together with the proportion of seams and the posi- between the points (LID = 0.. directly from the graph. Rowe and Armitage (1984) Qb/Q.. from earlier recommendations.. 9 by moving vertically upwards from (LID). 8 and drawing a horizontal line pier or are localized at some depth L.15 then an inspection of Fig.25 then an inspection of Fig.g. (LID). if the length L. = 13. This pier of length LID satisfies the design settlement Again. = 50 along the shaft of the socket. Design of a complete socketed pier for no slip conditions. negligible slip should occur under design conditions)../E. required length Li then this approach may be far too expensive -If 1.15. = 0. = 0. I (ii) If there is no intersection point then no pier of diameter D will satisfy the design requirements for the specified condi....45 can be found (see Fig..2 and (Qb/Ql).5%./E.15 and the procedure described in step 5(c) (b) If the seams are evenly distributed down the length of the must be followed. = 4.g. is a significant proportion of the Proceed to step 6. appropriate value of E. = (LID). Consideration of weathered seams To illustrate the three possible situations that may occur in In soft rocks. Proceed to step 6 in this case. Step 6 This step involves checking that the end-bearing pressure does not exceed allowable values that were selected to ensure an adequate factor of safety against overstressing the rock below the pier./Er = E.-D.5 with E. Eh/E. However. we find that there is no intersection of this line with the curve for E.

= Li volves selecting a pier geometry such that and (1) the probability of exceeding the design settlement pd is acceptably small.S + ST. (MPa). 4 modified parameters (LID).e. from Fig.e./E. Suppose that 20% of this compared with that initially adopted. and the corresponding ratios L.e.5 pier if only 30% of the expected side shear resistance were mobilized will not exceed one-third of the estimated base and the new estimate of 1: is bearing capacity (i. Step 5(c) -modified for recessment and a new estimate of I:.1. where LC# 0).51 most probable rock and pier properties). < qb. if it was designed using step 5(c) and curves given in Fig.. L. = 4. This value may be (LID). and 1.e./T) = 5.2 m (i. = (LID). (iv) Repeat steps 5 and 6 as previously described using the determine a settlement reduction factor (RF)..136 CAN.). = 37%. However. take LZ = m). Condition ( 1 ) ensures that the serviceability limit state crite- rion is satisfied./Er = 0. 7\17. = 0.40) = 2.60./Er. (LID). = 3. In this case..4 and (Qb/Q. 24. To illustrate this procedure. (Qb/Q.. = 5. and E. then the percentage of seams S could be reduced and -the pier . from Fig. the design settlement is not the settlement that would be predicted for piers with the selected geometry assuming the (Qb/Qt)d = 0. represent 20% of the final length of the socket. 2.40. this will give a proba- gives (LID). 1) E.. 2 and 3 and this revised for E.. This design will be conservative for a recessed socketed pier (i..5. J. E. and if necessary./Ed = 50. Design of n recessecl complete socketed pier If the pier is recessed.40) = 3 . (ii) Deduce a modified design side shear resistance.2) 5(c) may be repeated again for a revised value of 1: calculated with E.. S = 0. viz. Determine the following additional design parameters: (i) recessment length LC(m) and (ii) modulus of the rock adja- cent to the recessed length of the socket (see Fig. I987 (i) Estimate-The proportion S of seams to total socket length The design length 3. GEOTECH./Er = E..e. this pier could be designed as shown in Fig. . in fact. However. socket length LZ that will satisfy the design conditions (if no Adjust the design value of I d as follows: such pier exists.5 then design settlement to be less than 30%.2 and Q. the expected observed settlement of a pier designed to satisfy (iii) If the seams are localized at the top of the pier. Having determined an initial estimate of (LID). (2) the bearing pressure at the base of the pier under design (ii) Deduce the modified maximum pier length for full slip conditions is less than a specified allowable value qhl. I:. = 0. 7 for be used to compute a new value of (RF).1.40 was calculated assuming that the seams (i.. then L.. with the most recent estimate of (RF). (i.e. Condition (2) is intended to limit yielding of the rock beneath the base of the pier under design conditions. the calculation repeated to give (after another iteration) Ld = (often taken to be zero).40.. viz./D. the initial (20). and 7: to determine a for the appropriate values of L I D . 2. (c) Take the total pier length (as measured from the rock surface) to be the minimum of L I and LZ as calculated above. Thus bility of exceeding the maximum allowable settlement (which LZ = 3. take conditions (1)-(3) above will be less than the design settlement L = min ( L . it should be initially designed ignoring recessment by following steps 1-6 above. the design obtained ignoring recessment may be a little too conservative if the pier (iii) Deduce a modified modulus design does not involve slip (i. S = 2 (seam thicknesses)/L). L. If. length)./E.. qb.)d= 0.2 L i and Additional comments The design procedure described in the previous section in- If the seams extend approximately to a depth L i .seam material interface strength T. Considerable research has been conducted into available side shear resistance of concrete sockets under short-term condi- L = min (4.. which becomes (3) the maximum pressure that would develop at the base of the (LID).. design. = 2. E. = 0. Normally. very little . the designer may adjust the design to take account of recessment as follows.. The adjusted value of (LID): should seams. (i) If the seams are at the top of the socket..37 tions (see Rowe and Armitage 1984). = 0.. and a modified maximum pier geometry The effect of recessment can be ignored if the pier design involves slip (i. seams only represent 20% of the top 2./Qb = 51%. Consequently.40 is twice the design settlement) of less than 3%. consider a pier to be designed and then repeat step 5(c) using Figs.. 3..e. 3 0 .). l(1 .45. if it was designed using figures given by Rowe and Armitage (1987))....60. 4 0 .. Neglecting value of 1: instead of I.. one would want the probability of exceeding the Repeating step 5 using 1: = 0. the -the modulus of the weathered seam material E.e. step socket length is composed of weathered seams (i.. VOL. If the seams are distributed along the pier./D .. (Qb1Qt).. L2) = min (2.37 and (LID). 3.

. Because of this the same settlement p should not be acceptable for the same uncertainty as to exactly how much load will eventually be structure constructed on soft rock. :. = 100) Fig.. design settlement less f . Any attempt to than even if only 30% of the expected side shear resistance is achieve negligible settlement will necessarily involve large- mobilized.'.. 141 Back-figured f T For probability of Probability of exceeding exceeding design settle. :. the rock and concrete.e./E. [ 141 < qh.e. research has been conducted into the long-term load-transfer than the concrete pier (i. 1131 4h (MPa) Eq. OK Less than p..IE_'i From lines 4 and I I Eh/Er 'Wssume weaker base rock Effect of weaker base because of fracturing rock included in back- (see text) figured parameters (L /D )r. from Fig.. . 1I 61 < y h. * . Based on the very limited avail. When designing piers in rock that is considerably less stiff If the concrete pier is only slightly stiffer than the rock (i. 15bI Eq.. neglect figured parametcrs effect of seams (LID 'For Eb/E. Nevertheless.5. (LID 1. 1151 qhu (MPa) Eq. OK Eq. Horvath 1980). Ladanyi 1977. OK Observed Less than p.g. Design calculations for pier P2 6) (ii) Design based on empirically Design based on back-figured determined parameters parameters Line Step Quantity Value Notes Value Notes P'l (m) "Specified "Specified D (m) Specified Specified Ql (MN) Specified Specified E. OK settlement (m) NOTE:Items marked * are specifically commented on in the text. ROWE A N D ARMITAGE: 2 TABLE3. 15bl E. *By linear interpolation 'Wy interpolation lines lines 170 and 17b for 17n and 176 for Eb/E. (MPa) Horvath ( 1980) Horvath ( 1980) a. Although there is insufficient data for Soft rock should be regarded in a sense as a very strong soil. to satisfy the settle- behaviour of socketed piers. Fig.e. 12b (Qh/Ql)d (%) 4 . .i (MPa) "Eq. expensive.. One cannot expect zero carried in side shear. 10b (E. then there is no reason why cal for smooth than for rough (e. = 0. 1141 < qh. :./E. from "For E. = 0. = I.I (LID )<I *For E b / E . any firm conclusions to be reached. ment criteria is quite straightforward and invariably increasing able data (e. E./Ed = 250. R4) sockets. > lo). OK Eq.g. [I 61 < qh. (MPa) Eq.. 15al Eq.75 E./E. condition (3) above is intended to ensure settlement for piers constructed on soft rock. it may be reasonable to i. it may not that some additional load transfer to the base of the socket will be possible to design a pier of given diameter and reasonable occur with time owing to the time-dependent characteristics of length ( L I D 5 10) if the settlement criterion is too restrictive.../E. 1101 I* Eq. 1301 Back-figured E . settlement p would be acceptable.. ment less than 30% than 30% T~ (MPa) Eq. OK qbn. presuming that for a given structure constructed on soil a speculate that this time-dependent behaviour will be more criti. -F . 1121 Eq. (MPa) Eq. IOa (E. 15al Ed (MPa) Eq. this may be too against bearing capacity failure of the pier. 1121 qhil (MPa) Eq.x Eq. 1151 Eq. 1111 S Minor occurrence of seams Effect of any seams based on interpretation included in the back- of the bore log.t = 187 (Qb/QO* (%I *Interpolation from lines '"Interpolation from lines 18n and 18b 18a and 18b Ld (m) 4 1 (MPa) Eq. it would appear the pier length will reduce settlement.. 1131 Eq.. l l O 1 Eq.. there will still be a "factor of safety" of at least 3 diameter piers and low applied pressures. 11 11 Eq. (MPa) Horvath (1 980) Horvath ( 1980) . = 100) ~ i g 120 : (Qh/Qt)d " ..

and hence to give a specified I ) may not be unique.45/(n x 0.0.e. VOL. (iii) design load Q.. 5 lo).35 m is of this paper for a relatively smooth-sided socket. and 7. tween the results for E. 5 ~= ~8. The design diameter may be preferable to increasing the length of the pier. (vi) ? = 0 . it can be seen that in this case no interpolation was necessary The design procedure is illustrated under column (i) in to get results for Ep/Ed = 95.. f.. and Queenston shale using rock and concrete data obtained by 2.0. = 6./Er = 50 are 1.hin step 5 (lines 17a and and hence a design settlement p../E./E. I.75 was (assumed).. = 0.4. then obtained (lines 17c and 18c) by linear interpolation be- (ii) pier diameter D = 0. Thus. . = 1. For this design. however.IE. = 17 mm deduce the values of (Qh/Q. and Id can then be deduced from [lo] and [I I ] . increasing the pier the modulus beneath the pier Eb = 0. 1 5 E.45 MN (assumed).) and the maximum (vii) E = 2152/675 = 560 MPa. = 1.45 = 0.0085 X 390 x 0. = 0. If rock (this was in fact the case of Horvath's pier P2) such that the end-bearing resistance is not adequate. obtained for E.9 1 and 44% and a. I. and 250 and E. deduced (line 19). 100.82) = 3.) are then checked against the design The partial factors are chosen to meet a serviceability limit recommendations and are found to be adequate (lines 20-24.. GEOTECH. bearing pressure (q.5 and 1.17 = 0. the pier length required to satisfy a given design condition (i. The value for E. provided there is adequate end-bearing resistance. the benefit of the higher concrete modulus can be partially The design parameters can then be deduced from [5]: offset by the weakening of the rock at the concrete-rock inter. = f. The design parameters are as follows: and so only the curves for Ep/Ed = 100 (Fig. viz. state such that the probability of exceeding the design settle. 1987 I ~ ' ~ ' ' " ' " FIG. 10. Example design calculation (xi) (LID). obtained above for E./E.71)/4. Determination of design L I D and Q b / Q .). = (0. To illustrate the use of the proposed design method..= 3. = 100. These competing effects can give rise to a situation where (ix) Ed = 0. (xii) I. Table 3)). it is (x) E.5 (b). parameters (LID). and (Q.7 (lines 8 and 9 of may be derived from socketing the pier into the rock.7 X 1. value of Ep/Ed = 95 in Table 3 (line 12) is very close to 100 given in this paper../E. Table 3). 1.e. = 4.. The allowable end-bearing pressure (q. = 0 .5 ~ mm 176. Since 50 < E./E. using empirical correlations. 50.7 X 560 = 390 MPa.0 ( a ) ..53. one could also deduce values of (LID).5. Table 3). E. 24. = 0.17 MPa. empirical parameters: (LID). generally best to adopt the shorter of the possible design It is assumed here that the pier will be founded in fractured lengths. some benefit (in terms of reduced settlement) ment is less than 30% (i..71' X 0. 4 5 m = 1. = 0. values for pier P2. 25. 10) were used to (i) maximum permissible pier head settlement of p. the design length of socket L.4. This completes the design. (v) representative unconfined compressive strength of the rock and (Qb/Q. In fact.7 1 m (assumed)..53... face.). E.75Er (line 13). Table 3 using the empirical correlations given in [3a] and [4] From line 17c. the values of (LID). = 4. = 0.75 MPa (from Horvath 1980). = 37 000 MPa (from Horvath for Ep/Er = 50 and interpolate between this result and that 1980). the actual Horvath (1980) and the empirical correlations for E. = 37 0001390 = 95.O./E.. < 100./E. In these cases. (viii) T~ = 0. (iv) modulus of the piers. 18a and 18b. = 0. consider Rowe and Armitage (1987) have provided design charts for the design of a concrete pier (designated as P2) constructed in Ep/Ed= 10. EhIE. CAN./Q..82 MPa.

which was de. For design (ii) the actual value of Ep/Ed = 187 (line 12. P2 and P4..3 mm for the designs based on (i) empirical sions as the piers tested by Horvath. explain the difference between the back-figured rock mass This implies that this design is based on partial slip along the modulus exhibited by sockets P2 and P4 (see Table 4). E. Thus both designs may be considered to have performed adequately. the empirical correlation given by [3n] was adopted in an average "R4" socket and as a consequence the value of ? selecting the expected side shear resistance ?.. based on back-figured parameters. (Qh/Q.9.. Comments regarding pier design at location P2 (see Table 3) To allow for this roughening.4 mm. these two piers have been "redesigned" using (i) empirical correlations given in this paper and (ii) parameters back-figured from Horvath's load tests Comrnents regarcling pier design at locatioil P4 (see Table 5) (Table 4). = 0. (MPa) ~71o"&i+ Measured peak average side shear resistance.e.. was taken to Horvath's load test (see Table 4) are given in column (ii) of be unity. T~~~~~(MPa) Roughness category Modulus of pier concrete (GPa) Length of socketed pier (m) Diameter of socketed pier (m) Recessed pier length (m) 11.7 have there was no fractured rock beneath pier P4. 12a and 12b shows that there is no inter- curves have been given by Rowe and Armitage (1987). 11.35. the difference between the design settlement and Horvath's tests (1980) the observed settlement is greatest for the design based on back-figured parameters.)d = 27%. methods but was then artificially roughened (i.) been applied to both the side shear resistance and modulus. Notice that in sumed to be homogeneous) since the effect of the base is this particular case the intersection between the factored design already included in the back-figured modulus. For design (i). 18c). previous section and column (i) of Table 3. the more conservative shown in Fig.)*. Referring to Fig. accuracy the design values (LID).. ROWE A N D ARMITAGE: 2 139 A TABLE 4. In each case. which had geometries as would have given the longer socket (i. were 6 and 8. is less than the design settlement in both cases. design settlements of 8. In fact. and so Fig. based on empirical correlations. it Ep/Ed = 100 and 250 (lines 17a and 17b. the actual socket was considerably rougher than reason. 18a and 18b) and the is seen that for I. I I. For this 2(a). (This may partly dashed curves for elastic behaviour and full slip (T.37 occurs between the duced assuming a homogeneous deposit. values of (Qb/Q. 11 /Load Cell (These design values correspond to maximum allowable settle- N ment p.6 m) under the design load was 8. method proposed in this paper.Ll. = 1..)dwere deduced both for 0. expected.e. polation (lines 17c. section between the factored design line and the curve for I = sequently. since socket for design conditions where partial factors of 0. elastically (i.. The actual performance correlations and (ii) back-figured parameters respectively.) SECTION A A The observed settlement at the head of the socket (i. = 2pd. reference to Fig.. the modulus ratio Eh/E. E. the pier can be designed value for E p / E d = 187 was then obtained by linear inter.e. Table 3).. i.. To obtain the proportion of load carried to the base of To obtain pier geometries comparable with those of the pier the socket. which. In this case.e. 17 and 23 mm respectively. as and P4 (from Horvath 1980). 8.e. Validation of the procedure with reference to In this case. no slip under design conditions) with (LID). In this case. a design based on back-figured parameters two end-bearing piers. the effect of fracturing in the rock below = 95.35. ( 1983) performed tests on has been adopted. a roughness R4) using a special pneumatic roughening tool (Horvath 1980).5 mm were adopted for the redesign based on (i) empirical correlations and (ii) back-figured parameters respectively. The redesign of back-figured from a load test and used in design (ii) is 15% this socket using empirical correlations was described in the higher than the value given by the empirical correlation. 12a was used to estimate with sufficient the pier was not explicitly considered (i.e. 9 gives (Qb/Q. The corresponding Since there was no highly fractured zone beneath the bottom calculations using modulus and side shear values deduced from of the pier at socket P4. step 5(c) is required.. Thus if the same design settlement Horvath ( 1980) and Horvath et al. a.~/T~ = 1). The of the piers can then be directly compared with design actual socket at P4 was formed using conventional drilling expectations. (MPa) TOP VIEW Back-figured rock modulus. line and the interpolated curve for Id = 0. the empirical equation [3b] was The socket at location P2 was formed using conventional used to estimate the expected side shear resistance ? in step drilling techniques and was relatively smooth-sided. For design (ii). Con. Ep/Ed = 95. the rock was as. Relevant design information for socketed piers P2 and P4 Horizontal 7' Strain Gouges - Telltales in Rock + Item Socket P2 Socket P4 Telltales in Pier O Average uniaxial compressive stength (at site) for shale. the design settlement pd has been The design settlements pd adopted for the design of socket P4 selected such that the "redesigned" pier has the same dimen. To provide some validation of the design design) for the given conditions. P2 actually tested by Horvath.5 and A check of the allowable pressure q . at a FIG. Plan and cross section of geometry for socketed piers P2 depth of 0.. and maximum bearing . an in- Table 3) is not particularly close to the values for which design spection of Figs../E.

... Design calculations for pier P4 (i) (ii) Design based on empirically Design based on back-figured determined parameters parameters Line Step Quantity Value Notes Value Notes I I Pd (m) 0. I151 16. [I61 < clh. [I41 < (ih. [I01 2. [13] 22 ~b ( M k ) 3.. [ I l l 16 4(c) s -0 Minor occurrence of seams .. including quately (i. 12b will not give an intersec- 18 (QI.35 Eq.9.O Eq.75 Horvath (1980) 6.6 Eq. [ ] I ] 0.IE.e. tests by Williams shear resistance. referring to Fig. Victoria has recently completed extensive field research con- cates that in both cases the design would have performed ade. GEOTECH. 'This side shear resistance may be significantly (1980) have shown that fresh concrete can force base debris a reduced if the socket is formed under bentonite.9 Eq. particularly if . (MPa) 11.. Clearly..6 Eq.56 "'Eq..2 Eel. roughness in developing the maximum peak and residual side pacity of a socketed pier.140 CAN.2 ~ q 1121 .. 2.[5a] 1. OK Observed 0. (MPa) 560 Eq. and Pells (1981)..9 Eq.J..35 1. 1984a. 9 20 6 9 . cerning the cleaning and inspection of sockets. [5a] II Ed (MPa) 390 Eq. = 1.37 Eq. 1301 for rough socket 1. 1501 264 Eq./Er I Sound base-no highly 1 Sound base fractured zone below pier 14 4(a) (LID). [I31 6../Q!)~(%) 29 27 tion between the fac- tored design line and 19 Ld (m) I ..)from Fig. This work has involved settlement and well below the maximum permissible set. (MPa) 6. (MPa) 6.OK 23 4htn (MPa) 16. Horvath (1980).75 Horvath (1980) 6 2(a) 'i (MPa) 1.7 1 Specified 0. OK 0..7 1 Specified 3 Qt (MN) 4.9 "For E. Comparison of the design settlements of 6 and 8.26 Eq.. 1141 < q..7 Probability of exceeding 0.45 Specified 4 E. = 100. broken rock.6 mm) with the placing of the concrete.75 Eq.3 mm it is important to ensure that base debris is removed prior to (maximum permissible settlement 12 and 16. The Research by Pells et al.0083 "'Specified 2 D (m) 0. side shear resistance in the lower regions of the socket. 95 From Lines 4 and 1 1 140 13 E./E. 8 2(b) fT 0.7 than 30% 0. (MPa) 37 000 Horvath (1980) 37 000 Howath (1980) 5 o. load can be carried in both side shear and end bearing.0059 Less than p. 1151 24 qhu(MPa) 7.) will obviously influence the settlement response and ca. [101 15 4(b) I d 0. OK 3 .1 Eq. (Qh/Q. OK settkment (m) NOTE:Items marked * are specifically commented on in the text pressure qbmagainst design recommendations indicates that the considerable distance up socket walls thereby also affecting socket design is controlled by settlement considerations. 1121 21 qh.. 21.7 Probability of exceeding design settlement less design settlement less 9 fc 0. 120 1. Adoption of cleaning and inspection techniques such as those proposed by Holden Construction considerations should provide sufficient confidence in socket construction to The design method presented in this paper presumes that allow for end bearing in the design of socketed piers. 8 gives an elastic solution for (LID).0059 Less than pd.9 mm at the head of the socket indi.2 Eq. OK 7.. VOL.2 Eq. = 0. The Road Construction Authority of observed settlement of 5.35 I. b). the observed settlement is less than the design sockets constructed under bentonite.. 1161 < ql. neglect figured parameters effect of seams 17 (LID 1. 12n.006 *:Specified 0.7 than 30% 10 T~(MPa) 1. Furthermore.09 Eq. Effect of any seams based on interpretation included in the baek- of the bore log.75 Eq. 1987 TABLE 5.9 Note: Fig. Williams compression of a layer of base debris (e. 15b] 12 3 E. reported by Holden (1980.36.8 Back-figured from field test data 7 E .. Fig.. clay. the development of important new techniques and has been tlement).g. and others has shown the importance of wall etc. L41 377 . 11..45 Specified 4. (1980).

1980. Prolonged exposure of the rock socket (particu. Research reported by 1171 L . These calculations show that piers to casting of the pier. = to place the concrete as soon as practicable after construction [14al qh = (Qb/Qt)d(lt of the socket. Experiments by Equation Holden (1984a) have shown that under certain conditions. values for pier P4: (0)E. Holden ( 1984a). as always. care- is based on ful consideration should be given to the geological conditions (1) satisfying a specified design settlement criterion. however.s + SE. The ex. / E . [5bl E. - 1 2 3 4 5 6 7 8 9 1 0 O . ben.-E. Bamford and Washusen (1981) concerning the softening of 7 :I: = ( I .for satisfactory design larly to water or bentonite) may result in softening of the rock 1151 yh..?:" example.. 5 ~ ~ 116al qh.4 f . 0.) Similar comparisons conducted cedure are summarized in Table 6./Q./T)? Silurian mudstone in contact with water or bentonite has..? in designing socketed piers cast under bentonite./(1 . of course. However. The key equations used in this pro. Clearly. (17) E. Frequently.. ROWE A N D ARMITAGE: 2 -__:35 .45 (MPa) 100 mm on both mudstones and sandstones.. ever unusual conditions or uncertainties exist.. ) tent of exposure required to cause softening will. = f. tonite without additional reaming (Holden 1984a. The design method has been illustrated by a series of detailed the potential effects of softening should be considered in any calculations relating to the sockets P2 and P4 tested by Horvath situation where prolonged exposure of the socket occurs prior (1980) in Queenston shale. "factor of safety" is unknown.S + S E . Some recom. for L18al 118b1 .. The design method have shown similar encouragement. . = 100.6 6 (MPa) (roughness R4) [3bl the side shear resistance even in relatively rough (or artificially 141 E. Pells et a/. = 0..*D)/Ql= I d ( l . = 4 1 ./E.. The procedure is relatively by Rowe and Armitage (1984) for a number of other field cases simple and a socket can be designed in less than 15 min once documented by Glos and Briggs (1983) and Williams (1980) the designer is familiar with the technique. = 2 .s + ST. for satisfactory design 116bI the available side shear resistance.i: = fT .IDIQ.. + L. designed to have the same geometry as those tested by Horvath (1980) would have satisfied the design settlement criteria while Concluding discussion having a proven "factor of safety" against collapse of at least A procedure for the design of drilled piers socketed into soft 2. ( L / D ) d ? (f. 5 qh.. 1201 E:.. = ( I . number Equation tonite filter-cake can build up to thicknesses in excess of 130 I i= . a filter-cake of bentonite will form along the sides of a socket during construction. = 2 1 5 6 (MPa) roughened) sockets. depend on the type of rock and exposure conditions. it is generally advisable L131 qhi. resulted in a limit of 6 days being placed 03 the time 1191 (LID)?. the socket is relatively smooth (see e. 11461 q b 5 qh.. Determination of design L I D and Q. mendations for construction procedures to minimize the effect I101 (LID = Q I/ ( n D 2 ~ d ) of bentonite filter-cake in these situations have been given by 1111 Id = P&. 12 1 I I: = (pdE. b). 12. = 250.3) at the concrete-rock interface thereby substantially lowering y..g.)E./E. 6 ... = (L/D). 1 4 k b L - L - D D FIG. Summary of important equations TABLE Holden 1984a). considerable caution is required L5al 74 =fTfr.S + ST./E. This will reduce ii = 0.. 1121 91 = Q l l ( n D v 4 ) In the construction of socketed piers./?) sockets formed in this rock may be exposed to water or ben.. = L... (Since the test piles were not brought to collapse the actual rock has been proposed. at any specific site when designing piers in rock and field load (2) checking to ensure there is an adequate factor of safety tests should be conducted to verify design assumptions when- against collapse.

and KOZICKI.. H. 209-222. CEOTECH. A.0053 1 . F. discussions over many years and to Dr. JR. sockets in sandstone.P. 303-310. and ROWE. Proceedings of International Journal.K. Proceedings of the International Conference on Structural method of assessing the in situ strength properties of Western Cana. into weak rock. J. Ont. F. Methods of Department of Supply and Services Canada and the National improving the performance of drilled piers in weak rock. University of Western Ontario. J. 16(3). Endbearing on rock with particular reference to sand- COUETDIC. J. The construction of bored piles in weathered sedimen. pp. Rock sockets in Proceedings of the International Conference on Structural Founda- soft rock. pp. for Foundations and Slopes. Geotechnical Journal. dian coal. Research Report GEOT-4-83. and TURNER. N. and shale. ISU83-00082. 20(4). 2.. 1983. who acted as Scientific Advisor on this on discontinuous rock. thesis.. 1. Plate-load testing as a stone.. pp. Canadian Geotechnical Journal. ASCE Journal of Geotechnical Engineering. J. H. . and BRICCS. ASCE.. B. Research Report tions on Rock. Canadian Geotechnical Journal. 525-535. R. ence on Geomechanics. 1976. 91 -95. E. pp. F. P. I. high capacity socket design: Discussion. Design of foundations Research Council. Vol. R. pp.. A. 502-5 13. GLOS. Pells for many useful on Structural Foundations on Rock. 1975. Friction and end bearing tests on bedrock for useful comments on the manuscript. G. 1981. Development of a socket inspection device. J. M. 1984. Vol. and TURNER. Elastic solutions for the BAMFORD. The Proceedings of the International Conference on Structural Founda. tary rock. Canadian Research Council of Canada under contract no. Canadian Geotechnical tem for rock-socketed drilled pier foundations. Vol. Theoretical solutions for axial deformation of drilled Construction Authority of Victoria.. 24.0 . 1-7. Road 1987. and ARMITAGE. For the National Journal. H. R. 24. RowE. and WASHUSEN. H. Construction of bored piles in weathered rocks.P. KENNEY. CO.G. N. 758-772. socketed into weak rock. 69. WILLIAMS. Australia. Monash University. pp. pp. tions on Rock.R.. 1980. mental investigation into sideshear for socketed piles in sandstone. 1983. Perth. Bozozuk. 14(1). pp. F. neering Science. this issue.. H. 1987 Acknowledgements " Research Council of Canada. 1 -2 1 . M. 109(4).. M. G. 48 1 -487.R. Ont. and PELLS. 1979. mudstone.D.. Symposium on Weak Rock. The design and performance of piles socketed 1984b. K. Vol.J. Sydney. 1980. Boulder.P. Research project report on load transfer sys. shafts in rock. R. An experi- Sciences. H. and BARRON. 111. nical Report no. 1980. J. T.1 1-84. Sydney. 181.142 CAN. ROWE. Ph. J. P. N. 1980. Faculty of Engi. Proceedings of a Specialty Conference. 12(10). pp. A. H. I. Thanks are also due Mr. of the National KULHAWY. 18(4). contract serial no. Australia. Side resistance rock HORVATH. pp.R. Rock foundations for structures.. Sydney. and GOODMAN. Faculty of Engineering Science. 10X5. Many thanks are due Dr. Current practice for the PECK. PELLS. G. Kulhawy for his very LADANYI. Proceedings of the International Conference project. C.190. 1980. Proceedings of 4th Australia and New Zealand Confer. London. Rock Engineering design of socketed piles: Response to a survey. The research described in this paper was supported by the HORVATH.. 1981. WILLIAMS. E. Sydney. The design of piles HOLDEN. 1983. 11. GEOT. DSS file no. International Journal of Rock Mechanics and Mining PELLS. Weakening caused by design and analysis of rock-socketed piles. K. C. N. 1980. Melbourne. 1977. Canadian Geotechnical fluids used for drilling mudstones. Foundations on Rock. Tech. Voi. 291 -302. ARMITACE. 153-155. 1984a. pp. N. K.31155-9-4420. The University of Western Ontario. Tokyo. B. VOL. pp.. R. lSX79.. Melbourne. M. London.