Vous êtes sur la page 1sur 8

5/12

Experience in Grouting Rock Bolts


Essais d ’injection de boulons d’ancrage
b y T . J. D . L eech , C. B. E., B .E ., B.Sc.,
and
E . B. P ender , M .E ., B.A. Snow y M o u n tain s H y d ro -electric A u th o rity , C o o m a, N .S .W ., A u stra lia

Summary Sommaire
The paper describes the developm ent of groutable rock bolts Le rapport décrit l’utilisation des boulons injectés pour le sou­
fo r perm anent support of underground excavations in the Snowy tènem ent perm anent des souterrains des ouvrages de Snowy
M ountains Scheme in south-east A ustralia. M ountains d ’Australie.
The advantages of grouting rock bolts are briefly described. Les avantages de ces boulons ont été exposés ici brièvement.
T he results are given of tests for flowability, settlem ent shrinkage Il est rendu com pte des résultats de recherches po rtan t sur la
and bleeding of grouts having various proportions of cement, maniabilité, la com pressibilité, le retrait, et le " bleeding ” des
flyash, sand and water. Tests were also made to find the percentage coulis contenant divers pourcentages de cim ent de cendres volantes
o f alum inium pow der which would enable any shrinkage to be et de sable. Il a aussi été fait des essais pour trouver le pourcentage
avoided. A grout of neat cement and w ater with 0 005 per cent d’alum inium en poudre susceptible d ’éviter tout tassem ent du
alum inium pow der as an expanding agent and a water-cement coulis au voisinage de l’ancrage du boulon. Ils o n t abouti a un
ratio o f ab o ut 0-40 was selected. An air operated hopper was coulis de cim ent avec une proportion d ’alum inium en poudre de
developed for injection of the grout into the space around the 5 pour mille et un rapport d ’eau à ciment d ’environ 0,40. Un
bolt. m élangeur à air com prim é a été mis au point pour injecter le
F o u r types of groutable rock bolts are shown — one has an coulis dans le vide subsistant au to u r du boulon.
attached de-aeration tube of metal or plastic. The three others Q uatre types de boulons sont examinés. Le prem ier com porte
have hollow cores — one of these has a deform ed shank to un tube en métal ou en plastique pour perm ettre l’expulsion de
enable m ore effective bond between the bolt shank, the grout and l’air ; les trois autres ont une tige tubulaire. L ’un com porte une
the rock. tige striée qui assure une meilleure adhérence du boulon, du
W ithin less than two years of the development of the pro­ coulis et du rocher.
cedure described, nearly 40,000 of these bolts have been installed En moins de deux ans, près de 40.000 de ces boulons o nt été
in the works of the Snowy M ountains Scheme. mis en place dans les chantiers de Snowy M ountains.

Introduction and Background

D u rin g th e c o n stru c tio n o f th e first p ro je c t o f th e S now y W ith a g ro u te d b o lt o n th e o th e r h a n d , if th e an c h o r


M o u n ta in s Schem e — th e G u th e g a P ro jec t — ro c k bolts o r n u t ceases to be ab le to m a in ta in th e te n sio n in th e b o lt the
w ere u sed occasio n ally as a m eans o f tu n n e l su p p o rt. Since g ro u t w ill, o v e r b o n d distan ce, ta k e u p th e an c h o ra g e fu n ctio n
th e n , this use o f them has co n tin u ally increased. In th e m achine a n d as show n in F ig. I th e b o lt w ill rem ain c o m p eten t
h all o f T u m u t 1 P ow er S tatio n [ref. I a n d 2] w hich was ex c a­ to s u p p o rt th e ro ck . A g ro u te d ro ck b o lt h as th e a d d itio n a l
vated d u rin g 1956 a n d is 306 ft. long, 77 ft. w ide a n d 111 ft. m in o r ad v a n tag es o f b o n d in g a d ja c e n t b locks o f ro ck d irectly
deep, a to ta l o f 3,316 bolts w ere used to g e th er w ith som e su p ­ to o n e a n o th e r a n d o f ac tin g as a p in to resist sh earin g forces —
p le m en tary steel su p p o rt. T h e tra n sfo rm er hall, 127 ft. long b o th o f w hich c o n trib u te so m eth in g to o v erall stru ctu ral
a n d 73 ft. w ide, ex c av ate d in 1957, was s u p p o rte d by ro ck sufficiency.
b olts alone, 1,489 being used. In th e tailw ater tu n n e l an d In general, th e p ro b le m reso lv ed itself in to th re e m ain
tailw ater surge ch a m b er, w hich w ere e x c av ate d d u rin g 1957-58, p arts :
11,469 b o lts w ere used to s u p p o rt a n o v erall length o f 4,340 ft. (1) F in d in g a su ita b le g r o u t;
T h e use o f ro ck bo lts o n this scale fo r c o n stru c tio n s u p p o rt (2) C o m p letely filling th e h ole w ith o u t releasing the b olt
raised th e q u estio n as to w hether the ro c k bo lts could n o t te n s io n ;
fo rm p a r t o f th e p erm an en t s u p p o rt o f th e ro c k ex c av atio n s (3) S ealing th e g ro u t in th e h o le u n til it sets.
a n d en ab le less co n c re te lining to be used th a n w ould o th e r­
wise be the case. A s, how ever, th e w aters o f the S now y M o u n ­
tains a rea a re acidic it w ould be necessary, if th e bo lts w ere A S uitable G rout
used fo r p erm an en t su p p o rt, to p ro te c t them ag a in st c o rro ­ M o st ro c k b olts a re v ertical o r n e a r vertical. S u ch settl­
sio n o r to use bolts o f n o n -co rro d ib le m aterial. em en t sh rin k ag e as occu rs in th e g ro u t w ill ta k e p lace in the
G ro u te d bolts offered s tru c tu ra l design a d v a n tag es w hich n e ig h b o u rh o o d o f th e an c h o ra g e w hich is th erefo re th e p a rt
n o n -co rro d ib le bolts d o n o t have. F o r ro ck bolts to c a rry o f th e b o lt m o st likely to be ex p o sed b y sh rin k ag e. R ed u c tio n
o u t th e ir fu n ctio n o f su pplying tensile forces to p roduce o f settlem en t to a negligible a m o u n t is th erefo re essential.
eq u ilib riu m a b o v e a n d a ro u n d th e ex c av atio n , they m u st T h e use o f fine san d , flyash a n d a lu m in iu m p o w d er as m eans
rem ain in tension th ro u g h o u t the life o f th e w orks. If the o f c o n tro llin g sh rin k ag e w as in v estig ated . F in e san d , test
an c h o r, o r th e n u t o f th e u n g ro u te d bolt, is d am aged by c o rro ­ results o n w h ich a re sh o w n in F ig. 2, w as rejected ea rly in th e
sion to th e e x te n t th a t tension is released, th e b o lt ceases to in v e stig atio n b ecause o f its te n d en cy to b lo ck fine passages
fu n ctio n . o f th e g ro u tin g a p p a ra tu s . F ly ash , alth o u g h h av in g a high

445
PA T T E R N OF RO CK B O LT S ' T H E LOST A N C H O R A G E IS
AND C O M P R E S S IO N Z O N E I REPLACED B Y B O N D
IM M E D IA T E L Y AFTER ■A D H E S I O N O V E R A LENGTH
IN S T A L LA T IO N O F B O L T S . I O F T H E B A R . B O LT R E M A IN S
IN T E N S IO N B U T IT S
| E F F E C T I V E L E N G T H IS
\ RED UCED

Fig. 1 Effect o f loss of anchorage in a grouted rock bolt.


Effet de la perte de la force de l'ancrage dans un boulon
injecté.

Fig. 3 Effect o f various perceniages o f alum inium pow der in


expending the grout.
Effet de divers pourcentages d'alum inium en poudre sur
l’expansion du coulis.

“ pum pability” was measured by the time taken to pum p the


grout into a model rock bolt unit. This consisted of a fully
set up 20 ft. bolt encased in a clear plastics tube suspended
20 ft, above the level of the injection apparatus. Fig. 5 shows
the flowability of various sizes.

Fig. 2 Flow ability of m ortars o f various com positions.


Fluidité de mortiers de diverses compositions.

silica content, was obtained from cyclone precipitators and


had a high carbon content and water requirement. The third
alternative was the use o f sufficient commercial grade alumi­
nium powder to give an expansion to the grout sufficient to
counteract the am ount o f settlement shrinkage, but not
enough to make the grout flocculent and therefore weak.
Fig. 3 shows the effect o f various percentages o f aluminium
powder. The fineness o f the cem ent used also was found to
be significant. Figs. 4 to 9 inclusive record the results o f these
investigations. They indicate that both settlement shrinkage
and bleeding are a t a minimum when no flyash and no sand
are used and the percentage o f aluminium powder is 0-005 Fig. 4 Flow ability o f cement grout containing various per­
per cent by weight o f cement. centages o f flyash.
“ Flowability ” was measured by the time taken for the Fluidité d ’un coulis de ciment avec divers pourcentages
flow cone shown in Fig. 2 to em pty and, later in the work, des cendres volantes.
446
<£Y:

Fig. 5 Flowability of cement grout made with different brands


and types of cement.
Fluidité de coulis de ciments de divers types et de diverses
marques.

The grout which was adopted is a neat cem ent paste in


which :
(a) The flowability is between 25 and 30 seconds.
(b) W ater-cement ratios of grout satisfying (a) are, subject
to variations in the cement, usually between 0-38 and 0-44.
Fig. 7 Settlement shrinkage of neat cement grout made with
(c) The cement should have fineness (specific surface) o f various types and brands of cement.
approxim ately 4,500 sq. cm/gm as measured by the Blaine Retrait de coulis de pate pure faits avec divers types et
method. U sually this means th at the suitable cements are diverses marques de ciment.
those sold on the Australian ¿narket as high early strength
cement. Filling the hole
(d) Commercial aluminium powder, 0-005 per cent, by It was taken as a given condition that any grout injection
weight o f the cement, is added as an expanding agent. apparatus should be operated by the air pressure o f 80-100 lb./
It was found necessary to create a uniform suspension of sq. in. (5-6 to 7-0 kg/sq. cm) commonly available in under­
cement and aluminium particles in water. This means that ground excavations.
a high-speed mixer, o f an egg whisk type driven by an electric This set a limit to the pressure available to force the grount
drill, m ust be used. Three minutes mixing produces the best into the hole and, together with the restricted size o f the tubes
(£ in. internal diameter) which could be used, set limits to
results, for care is necessary to avoid emulsification with air.
H and mixing o r slow paddle mixing lead to blockages. viscosity o f the grount. The problem was to obtain a grout
o f no shrinkage and the greatest possible strength yet which
G routs meeting these conditions flowed well in the hole, still could be used with the apparatus. The problem of filling
gave complete coverage, and attained strengths of 4,000 lb./ the hole was therefore investigated a t the same time as the
sq. in. or m ore when tested in 6 in. X 3 in. (15*2 cm X 7-6 cm) grout.
cylinders a t 90 days. The bolts in use had a nominal diameter o f 59/64 in. (23*4
mm) and were set in holes of 1 § in. (34*9 mm) to 1 § in.
(41*3 mm) diameter, which set a practicable limit o f £ in.
(6-3 mm) to the outside diameter of the tubes which could be
used for grout or de-aeration.
As the annular space between the bolt and the walls o f the
hole offered a greater cross sectional area than any tube it
was decided to use this space for grout and the tube for
de-aeration a short tube being used to pass the grout
through the seal. The tube used initially was a type known
in Australia as “ Bundyweld” , being flat steel sheet rolled into
a double-walled tube and welded. Polythene tubing was used
on some works but it is considered use o f such tubes is struc­
turally inferior, as a relatively soft plastic is interposed bet­
ween the steel, grout and rock, and the capacity o f the bolt
and grout complex to resist any shearing forces is reduced.
However, bolts with long attached tubes are awkward to
handle in the confined spaces of underground work and the
tubes become damaged. The final development has been
hollow core bolts, in which the central passage acts as a de­
aeration tube. To ensure grout cover of the anchorage it is
necessary that the de-aeration tube or hollow cored bolt
shank extend beyond the anchorage. Fig. 10 shows the bolts :
Fig. 6 Relation of flowability and pumpability. (a) is the attached tube slot and wedge bolt — (b) and (c) are
Relation entre la fluidité et l’aptitude au pompage. hollow core bolts with expansion anchors — (d) is a hollow
447
R e t r a i t e t f l u i d it é e n f o n c t i o n d e s p o u r c e n t a g e s d e c im e n t , d e c e n d r e s v o la n t e s e t d ’e a u .

P e r t e d e l ’ e a u d e s c o u l i s e n f o n c t i o n d e s p o u r c e n t a g e s d e c im e n t , d e c e n d r e s v o la n t e s e t d ’ e a u .

448
Fig. 11 Grouting a rock bolt.
Injection de coulis autour d’un boulon.

(2-54 cm) thick and 2 in. (5-08 cm) thick. When the bolt
Fig. 10 Groutable rock bolts. is tightened the bearing plate forces the seal into the collar
Boulons injectables. o f the bolt hole and the containing ring is crushed into a
collapsed condition. This seal has proved satisfactory where
the collars of the drill holes are neat. F or holes with irregular
collars it was found preferable to use a quick-setting m ortar
core bolt with a deformed shank. The short grouting tubes composed o f 2 parts by volume o f Portland cem ent, 1 part
are not shown. sand, and 1 part of a proprietary quick-setting agent — a
The pattern o f deform ation on the bolt (d) is, because o f saturated solution of sodium carbonate/sodium silicate.
the test results of Brown and H ajnal-K onyi »[ref. 3 and 4], A fter 4 minutes a t 60° F (15-5° C) this m ortar will set rock
required to conform to ASTM 305-50 T. This bolt is regarded hard, even in running water.
as an end point in the development of the groutable rock bolt. Seals o f oakum dipped in grout have been tried but are
As, however, hollow core steel stock with a satisfactory useless in wet holes and have the disadvantage that appreciable
deform ation pattern is not always available in Australia, some quantities o f oakum may get caught between the bearing
attention has also been given to the use of bonding agents, plate and the rock surface and later contribute by plastic
such as epoxy resins. yielding to loss of tension in the bolt. “ D rypack” — a m ortar
o f 1 part cem ent to 2 \ parts o f sand passing a No. 16 screen
(openings 1-020 mm) — has also been used as a seal.
S ealing the hole

The greatest single difficulty in developing methods o f grou­ A pparatus


ting lay in sealing the collar of the drill hole around the bolt,
de-aeration, and injection tubes. The general aim was to The initial trial apparatus, which incidentally was after­
develop a seal which would be factory-m ade and capable wards used on the works, was built to have a capacity suffi­
o f meeting all drill hole collar conditions, w ithout any adjust­ cient for two 20 ft. (6-1 m) rock bolts and is shown in Fig. 12.
m ent or fitting on the jo b , and could be placed easily in pos­ A length o f 8 in. (20 cm) dia. pipe was used, a flange welded on,
ition when the original rock bolting was done, so that no and a concrete filler used to make the cone section. The
slackening o f the bolt tension at a later date would be necess­ apparatus, including hoses, was carefully designed so that
ary. The seal which was developed consists o f a pad o f expan­ there would be no sudden change o f section.
sion join t filler material in a containing ring of brass shim D uring the tests is was found necessary to add an air dis­
material 0-005 in. (0-127 mm) thick. It is pierced by holes for charge at the bottom o f the hopper to keep the grout agitated.
the bolt, the grout tube, and de-aeration tube and is used with W ith this, grout can be kept in usable condition for about
a bearing plate which has a second hole through which the one hour for satisfactory flowability, but not over 15 minutes
tubes project, Fig. 11. The seals are m ade in two sizes — 1 in. for effective expansion.

449
was using a grouting hopper o f the type described and grou­
ting o f rock bolts was a regular activity.
Some contractors made rock bolt grouting a Sunday main­
tenance o p e ra tio n ; others preferred to use a separate full­
time crew.
The ideal procedure would be to grout the rock bolt imme­
diately after installation, but because bolts have to be installed
close to the face under cramped conditions and poor visibility
it is not practicable to do so. Also, the practice is adopted of
not allowing bolts to be grouted closer than 50 ft. (15-25 m)
to any place where there would later be blasting, other than
blasting for cleaning up.
In fact, contractors often prefer to have the face team
place a plain rock bolt without attachm ents and, later, when
the face has advanced, have another team prepare the bolts
for grouting by inserting a de-aeration tube and seal. This is
not a desirable practice because, from the structural view
point, once a rock bolt is installed the tension in it should
not be relaxed and this is especially so when the face is no
longer nearby to give additional support. Cases have occurred
when unthinking workmen had released the tension in all of
the bolts over an appreciable area of wall and roof—fortunately
without any serious consequences. An interim practice was
adopted of allowing alternate bolts to be relaxed for up to
4 hours. This has not been wholly satisfactory and in current
contract work it is required that at the time o f installation
the bolts be prepared for any later grouting and that once the
bolts have been put in tension, the tension shall not be released.
The discharge of grout from the de-areration tube has
always been found to be an adequate indication that the
hole has been completely filled by grout and, if the tubes are
Fig. 12 G rout injection hopper. closed immediately, the holes should remain filled. However,
M alaxeur de coulis pour injection. as a final check o f normal inspection m ethods, some cores
containing rock bolts and surrounding rock have been drilled
out with a 6 in. (15-2 cm) diamond drill. N o defects in the
grouting have been found, but in addition to checking the
Field E xperience grouting these cores provided useful information on the
joints, natural o r artificial, in the rock surrounding the
excavation.
The method o f grouting rock bolts described above was
developed in the m id-half o f 1958 and introduced to the field In excavation which will remain unlined, exposed surfaces
from September 1958 onwards. o f rock bolts and accessories are coated with sand-cement
m ortar to a minimum thickness of £ in. (6-3 mm) o r a cement
A t this time work had just commended on the 9-mile Toom a- g rout to a minimum thickness of £ in. (1*26 cm). However, in
Tum ut Tunnel, the 10 £-mile M urrumbidgee-Eucumbene T un­ order to obtain uniform ity in the tensions put into the bolls
nel, and Tum ut 2 Power Station with its 3-mile headwater and by the fixed torques used to tighten the nuts, all bolt threads
4-mile tailwater tunnels. The contracts required the grouting are required to be coated at the factory with a plastic grease
of rock bolts but, in accordance with the A uthority’s contract o f the type used for bilge protection and containing rust-
policy of allowing a contractor fairly complete freedom to inhibiting compounds. Owing to the presence o f this grease,
choose his methods provided the required end result is achie­ there was difficulty in getting m ortar or grout to adhere.
ved, left each contractor free to develop his own methods. This was overcome by applying a priming coat o f bitumen-
Much of the difficulty in the first six m onths o f rock bolt aluminium paint. The two together — paint and m ortar or
grouting on these jobs arose from attem pts to use conven­ g rout — give excellent chemical and mechanical protection
tional grouting equipment, which is unsuited to rock bolt to the exposed portions o f the bolt and accessories. There
grouting. In rock bolt grouting the passageway sizes are so are obvious advantages in using the same grout for this as for
much smaller than in conventional grouting that shortcom­ filling the hole, and the selection of suitable equipm ent enables
ings which would not stop an ordinary grouting jo b cause this to be done. The use o f sprayed grout also enables relati­
continual hold-ups. For an 8 ft. (2-44 m) rock bolt the actual vely inaccessible boll ends and washers to receive good coa­
grouting operation can be completed in one and one-half tings o f grout.
minutes.
The development of rock bolt grouting methods took place
It has been found that only sufficient grout should be mixed during the concluding stages o f construction o f T um ut 1
for immediate needs and any left over a t the end o f an hour Power Station and tailwater tunnel, and the first grouted
m ust n o t be used. On the job, the workmen will rarely pay rock bolts were installed in these works in 1958, some 1,600
sufficient regard to correct mixing o r to dirty holes, ineffec­ bolts in the tailwater works being grouted. In the M urrum ­
tively cleaned tubes, and ineffective seals. Instead there is bidgee-Eucumbene Tunnel, Toom a Tum ut Tunnel and
always an attem pt to cure troubles by adding more w ater to Tum ut 2 Power Station, illustrated in Fig. 13 and 14, grouting
the grout until the inspector stops the work. of rock bolts is now routine procedure and approxim ately
Each contractor tried out various machines and methods 40,000 rock bolts have been grouted within less than 2 years
and had similar difficulties. However, within 6 m onths each o f the completion o f the investigation.
450
Fig. 13 M achine Hall T um ut 2 Power Station during construction.
Salle des machines de la centrale de Tum ut 2 pendant la construction.

Fig. 14 Tum ut 2 Transform er Hall.


Salle à transform a leurs de la centrale Tum ut 2.

451
Acknowledgements [2] M oye, D. G. (N ov. 1958). Rock Mechanics in the Investiga­
tion and C onstruction of T1 U nderground Power Station,
T h e au th o rs reco rd a p p rec iatio n o f the perm ission o f the Snowy M ountains, A ustralia. Symposium on R ock
S n o w y M o u n tain s A u th o rity to pu b lish this p a p e r a n d th a t Mechanics Geol. Soc. A m er., Published Geol. Soc. Amer.
a n y views expressed are n o t necessarily those o f th e A u th o rity . Case Histories , No. 3.
M an y officers co n trib u te d to the w ork. G ro u t investigations [3] B ro w n , K. J. (M ay, 1958). The G routing and Bond Strengths
w ere m ad e by M r. D .C . K e n n ard , B .E ., a n d M r. W .F . N av in , of Plain and Deformed H.T. Prestressing Bars. C iv il
B .E ., d eveloped the g ro u t in jectio n a p p a ra tu s. M r. A . D . H os- Engineering and P u b lic Works Review, vol. 53, No. 623,
k in g , B .E ., D .I.C ., D r. C.C . W o o d , B.Sc., B .E ., D .I.C ., p. 539.
P h. D ., an d M r. N av in , developed field practices a n d assisted [4] H a jn a l-K o n y i, K. (M ay, 1951). C om parative Tests on
th e c o n tra c to r’s sup erv iso ry staff in establishing th e m ethods V arious Types o f Bars as Reinforcement o f Concrete
o n th e w orks. T hese officers are m em bers o f th e A u th o rity ’s Beams. The Structural Engineer, p. 133.
Scientific Services D iv isio n , o f w hich M r. Leech is E ngineer- [5] F eald m an , H. (Dec., 1954). The Use o f Deformed Bars for
in-C harge. M r. P ender, w ho is the S enior E x ecu tiv e E ngineer, Concrete Reinforcement. C iv il Engineers and P u b lic
C iv il E ng in eering, w as responsible fo r th e o v erall direction Works Review, vol. 49, N o. 582, p. 1298.
o f th e d ev elopm ent. [6 ] L ang, T. A. (M arch, 1959). U nderground Experience in the
Snowy M ountains, A ustralia. Proceedings o f the Second
Protective Construction Symposium R and C orporation,
References published U.S.A.F. Project R and Report R-34I, Protective
[1] L ang , T. A. (Feb., 1958). Rock Behaviour and Rock Support Constructions.
in Large Excavations. Symposium on U nderground [7] G ilm o u r, L. W. (July, 1959). Some Aspects of U nderground
Power Stations, A.S.C.E. Power Division O ctober 1957. W ork-Snowy M ountains Scheme, A ustralia. Symposium
C onvention New York U.S.A. A bstract published C iv il on Shaft S in k in g and Tunnelling , The Institution of
Engineering , vol. 28, No. 2, p. 40. M ining Engineers, London.

452