Académique Documents
Professionnel Documents
Culture Documents
Friction and end bearing tests on bedrock for high capacity socket design
A N D NOELL. JOURNEAUX
PETERROSENBERG
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14
Geotechniccrl Depcrtment, Professionul Services Division, Warnock Her-sey Professional Services Ltd.,
128 Elmslie Street, LaSnlle, P.Q., Crrnrrda H8R I V8
Received October 22, 1975
Accepted May 4, 1976
Rock sockets are used to transfer axial loads to bedrock by a combination of periferal contact
bond and end bearing. The load distribution with embedded depth in relation to rock properties
actually measured is discussed. Case histories are presented.
The results of full scale and bearing and friction tests carried out on various rock types are
given. A relation between the unconfined compression strength of rock and the developed
concrete to rock bond value is reported.
Les emboitures dans la roche servent a transmettre les charges verticales a la roche par le
frottement lateral et la resistance a la pointe combines. Des cas types de la ripartition de la charge
selon les proprietes de la roche en place sont decrits.
On prtsente les resultats d'essais en vraie grandeur de la capacite portante a la pointe et en
For personal use only.
This method requires that the socket base end bearing value is assigned and it is assumed
be extended to a depth sufficient so that the that the rest of the stress is carried by bond.
stress does not exceed the allowable end bear- The socket length is adjusted so that the allow-
ing capacity of the rock. able end bearing and bond values are not ex-
The effect of bond between the concrete ceeded.
and the surrounding rock is ignored. In this The main difference between this and the
method the assumption made that all the axial other methods is that the relation between ap-
stress is transmitted to the socket base. Since plied axial stress and actual developed end
the load transferred to the rock by bond is bearing stress for various socket embedments
ignored, the result is that actual stresses on and rock properties is required.
the rock at the socket tip may be significantly
lower than assumed, since actual field tests Load Transfer
indicate that even in fractured rock the con- Finite element studies by Osterberg and Gill
crete to rock bond is significant. (1973) indicate that the percentage of load
For personal use only.
FIG. 1. b a d transfer curves for varying moduli ratio computed for embedment ratio of
3.0 (adapted from Osterberg and Gill 1973).
and does not reflect the zones of weakness in six times the unconfined compressive strength
the rock mass; such as clay filled joints, soft of the rock material. It was also determined
layers, or the pattern of jointing. Any relation that as the depth of embedment increased up
between the ultimate bearing capacity based to 1.0 the ultimate bearing capacity also in-
on measured laboratory strengths can be used creased from 25 to 70% depending on the rock
only after the laboratory strength test results tYPe-
have been reduced to represent the average Based on the bearing capacity theories re-
in situ strength of the rock mass within the viewed, it would appear that the ultimate
zone stressed by the socket. Such reductions bearing capacity of rock as obtained from the
may be based on experience with the local above is in excess of any allowable bearing
rock types and with the degree and nature of values presently recommended in building
jointing known to exist. codes. For example, the National Building
Teng (1962) reports that the ultimate bear- Code of Canada (1970) permits for footing
ing capacity of rock varies from five to eight design an allowable pressure of 10 ton/ft2
times the unconfined compressive strength of (9.7 kg/cm2) on the surface of limestone.
intact cores. Coates (1967) considers that the This value is intended to allow for the existence
ultimate bearing capacity is about three times of voids, open joints, and soil filled seams be-
the unconfined compressive strength; this theory low the selected bearing level. The Chicago
is based on the Griffith Fracture Theory which Building Code as quoted by Baker and Khan
concerns the initiation of cracking. (1971 ) allows end bearing pressures of 100
Research by Rehman and Broms (1971) ton/ft"97.6 kg/cm2) on 'sound' limestone
determined that the ultimate surface bearing with a 20% increase for each foot of con-
capacity of solid rock varies from about four to finement up to a maximum of 200 ton/ft2
NOTE
FIG. 2. Load transfer curves for varying moduli ratio computed for embedment ratio of
4.0 (adapted from Osterberg and Gill 1973).
(195.2 kg/cm" at a 5 ft (1.5 m) depth. The available which although not carried to failure
higher values may be used provided the rock are of interest. Table 1 presents the results of
is core drilled and inspected to insure its such plate load tests, included are tests carried
quality. out by the authors.
Freeman et al. (1972) based on in situ pres- It would appear from the results of limited
suremeter tests on shale with a lower bound data presented that the range in allowable end
unconfined compressive strength of about 1350 bearing values for design of sockets would be
psi (94.9 kg/cm2) allowed a bearing value of from about 15 ton/ft2 (14.6 kg/cm2) in poor
50 ton/ft"48.8 kg/cm2) for embedded cais- rock, 25 to 75 ton/ft2 (24.4 to 73.2 kg/cm2)
sons. in medium rock and 100 ton/ft2 (97.6 kg/cm2)
Plate bearing tests carried to failure on the in sound rock. These values are realistic pro-
surface of bedrock are costly to perform and vided adequate drilling below the fourdation is
therefore only a few are published and readily done to insure that no voids exist below the
available for review. There are, however, tests socket tip.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14
TABLE
1. Results of plate and pile load tests on bedrock
Maximum load or
Reference Rock type failure in tons/ftz (kg/cm2) Remarks
For personal use only.
(WHI Report 1967) AXT size 95% (225) load; settlement 0.02 in. (0.5 mm) Aw
.
0\
Tavenas (1971) Quebec City sandstone 250 Practically linear to 250 ton/ft2
RQD < 10% (244) (244 kg/cm2) then failure defined
Labrador Friable iron formations similar to 50 Linear to 45 ton/ft2 (44 kg/cm2)
(WHI Report 1968b) weathered and friable sandstone (49) then failed; settlement 0.5 in.
(12.7 mm)
*Wornock Hersey International Ltd.
NOTE
LOAD IN P S I (kg/crn2)
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14
'I I I I I
T F F ! EARS '
For personal use only.
surfaces;
loaded to a maximum stress of 170 psi (1 1.9 psi (17.22 kg/cm" and further movement
kg/cm". As shown on Fig. 2 the socket move- indicated no drop in strength.
ment at 160 psi (1 1.2 kg/cm" defined as The tcst results show that even in poor
failure, was 0.25 in. (6.35 mm); no drop in rock the bond value is significant, that it is
bond with further movement was observed. mobilized at small movements and that it re-
The test shown on Fig. 4 was performed on mains fairly constant after failure.
an 8 in. (0.20 m ) diameter, 36 in. (0.91 m) Examination of these figures as well as
long concrete socket formed at the bottom of other published data show that, the ultimate
a test pile. The socket was grouted onto a bond strength is developed with very little
steel anchor block through which steel cables movement usually less than 0.25 in. (6.35 mm)
were strung to the surface. Two tell-tales for after the peak strength has been obtained the
deflection measurements were installed in the strength does not drop off but tends to remain
plug. Loading was by means of a hollow stem constant even with further movement. This
jack, reacting against the pile casing. The test was also observed in an instrumented caisson
was carried out at a depth of 55 ft (16.7 m) by .Jackson et al. (1974).
below ground level. Failure was taken at 245 A list of observed values of bond between
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14
Bond resistance
Reference Rock type in psi (kg/cm2) Remarks
Moore (1964) Weathered fractured 140 Concrete 5000 psi (351.5 kg/cm2)
interbedded sandstones and shales (9.8) 20 in. (508 mm) diameter by 8.5 ft
(2.59 m) long plug; n o failure
For personal use only.
established
Matich and Weathered shales, 45 Concrete 5000 psi (351.5 kg/cm2).
Kozicki (1967) q, = 70 psi (49 kg/cm2) (3.2) Test on 24 in. (609.6 mm) diameter
by 13.5 ft (4.1 m) long plug; no
failure
Thorburn (1966) Weathered and fragmented shales; 35 Test on 3 in. (76.2 mm) diameter
joints spaced 0.5 to 1 in. (12.7 to (2.5) by 12 ft (3.65 m) long plug; 2
25.4 mm) ultimate bond obtained 0
1
Test on 19 in. (482.6 mm) diameter
rn
Seychuck (1970) Shale, joints spaced 3 in. to 10 in. 440
(76.2 to 254 mm), q, = 5000 psi (351 kg/cm2) (31) by 36 in. (914.4 mm) long plug;
no failure
Gibson and Deveny Sandstone 120 Failure defined
(1973) (8.4)
Jackson et al. (1974) Cherty limestone 235 Failure interpreted by authors
RQD 50 (16.5)
This paper Fig. 3 Highly fractured hard, Andesite core rec. Test on 18 in. (457.2 mm) diameter
in BXL core 33 to 75%, q. = 1500 psi by 22 in. (558.8 mm) long plug.
(105 kg/cm2) Concrete 3000 psi (21 1 kg/cm2)
This paper Fig. 4 Shale, joints spaced 2 in. (50.8 mm), 245 Test on 8 in. (203.2 mm) diameter
q, = 3000 psi (21 1 kg/cm2) (17.2) by 36 in. (914.4 mm) long concrete
plug. Concrete 5000 psi (351.5
kg/cm2)
CAN. GEOTECH. J. VOL. 13, 1976
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14
For personal use only.
L E G E N D
REF. REMARKS
0 THORBURN(1966) qu -
eat.
0 MATICH A N D KOZICKl(1967)No t a i l m .
A T H l S PAPER Fig.4
V THlS PAPER F i g . 3
+-
X SEYCHUCK(1970)
GIBSON AND DEVENNY
( 1973)
No faihrb
qu -
eat.
10
FIG.5. Tentative relation between bond and unconfined compression strength of rock.
concrete and rock is given in Table 2. Although depend on such factors as, socket wall rough-
these observations reveal a wide range of ness, concrete strength, cleanliness of sides,
values, and the precise data with respect to and groundwater conditions. The effect of water
the rock properties are not available for each in sockets was particularly pointed out by
case, the data have nevertheless been plotted Baker and Khan (1971).
on Fig. 5 using an estimate of rock compressive
strengths and give a tentative relation between Rock Socket Design
compressive strength and the ultimate bond Based on the information presented a pro-
which may be used to estimate a preliminary posal for the design of rock sockets is given
bond value. for consideration. This is as follows:
The bond value actually developed will also ( a ) On the basis of field and laboratory
NOTE 333
investigation estimate an allowable end bearing factor of safety against overall failure against
capacity for the rock. the end bearing value chosen.
( b ) Estimate a bond value to be used. This BAKER,C. N. and KAHN,F. 1971. Caisson construction
selection may be based on the tentative rela- problems and correction in Chicago. ASCE, J. Soil
tion given in Fig. 5 or on the results of field Mech. Found. Div. 97 (SM2), pp. 417-440.
tests. COATES,D. F. 1967. Rock mechanics principles. Mines
Branch Monogr. 874, Dep. Energy, Mines, and Res.
(c) Estimate the ratio of the rock to con- Ottawa, Can.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Texas A&M University on 11/14/14