I

~ ·'-<...,'-' • "'' ,,,•'l'•,.~,',,;"• ·,>••,.-. ~~~t;.~~,~~~~~~~· ~1<.·:'•::~~ ~l~>j:,\. '·,;__.,j- Jo,. ',~' ~,~·~.' ,1 - ,,.,.,_.·~,,·~ ,;'z't.·• ...... ':._.,o•,J.' • f _ '::_ ,.

Bond Strength of
Shotcrete Repair byoenisB•aupre,Ph.o
A good concrete or shotcrete repair must possess three prime char- is generally very good, due to the shotcrete com-
paction process and the normally low water/cement
acteristics: 1) the repair material must be durable in an aggressive
ratio of this material , particularly for dry-mix
environment, 2) the repair must be well bonded to the substrate and shotcrete (2). It is most probable that the phenom-
enon of rebound plays a more important role than
3) the repair must be as crack free as possible to efficiently protect
compaction or mixture composition on bond
any embedded steel reinforcement from corroding. strength than is recognized . When shotcrete starts
impacting on the receiving surface only the cement
hi s article deal s only with the second iss ue:

T
paste sticks to the surface. The other components
the bond strength of the repair. Before dis- rebound until a sufficient thickness of paste is built
cussi ng bond strength of shotcrete, it may be up. A well-compacted layer of low water/cement
helpful to look at bond strength of concrete repairs. ratio Portland cement paste is thus formed at the
In 1956 Felt wrote: interface between the old concrete and new shotcrete
" .. .it became apparent that factors influenc- layer.
ing bond of new and old concrete were not eas- This paper summarizes the results of a study on
ily isolated and controlled. The most the influence of surface preparation and mixture
important factor was the condition of the old composition on long term bond strength of shotcrete
swface-its cleanness, roughness and strength repairs. This paper also presents the results of a
or soundness. If the sUJface was clean, slightly new study, carried out in 1998, on the influence
rough andfree of weak outer skin, good bond of multi-layer applications on shotcrete interlayer
was generally obtained, otherwise relatively bond strength.
poor bond was obtained." (L)
Bond Strength of Shotcrete Repairs
In 1988 , our understanding of shotcrete bond In a study carried out by Laval University in 1988
strength was much the same as it was for concrete (3), many pull-out tests (over 700) were performed
in 1956. Very little information was available con- to evaluate the capacity of different shotcrete mix-
cerning the parameters that influence the long term tures to produce an acceptable and durable bond to
bond strength of shotcrete, particularly the influ- concrete. A secondary objective of this study was
ence of mixture composition and surface prepara- to evaluate the influence of surface preparation on
tion. In 1987, Schrader and Kaden reported that the shotcrete bond strength. Twenty-one different con-
bond between shotcrete and an old concrete surface crete slabs were cast, cured and allowed to dry for
one year. Then several different methods
of surface preparation were used to prepare
the slabs: sandblasting, jackhammering,
jackhammering followed by sandblasting,
Figure 1. Pull out test set-up. grinding, and hydro-milling . Following
this, slabs were covered with a thin layer
of shotcrete. Six dry-mix and four wet-mix
shotcrete mixtures of different composi-
tions were used (some mixtures contained
silica fume, latex , steel fibers, high early
strength cement, or a combinations of some
of these variables). All these mixtures were
comprised of good quality shotcrete.
Repaired concrete
Pull -out tests were performed to evalu-
ate the repair bond strength. A 3 3/ 4 in.
core (95 mm) diameter core was drilled with the

12 Sh.otcrete Magazine • May 1999

cut extending beyond the bonded Table 1. Summary of repair bond strength in MPa.
inte1face into the original substrate
concrete. A circular steel plate was Type of Jackhammer +
attached to the top of the unbro- shotcrete Hydromilling Sandblasting Grinding Jackhammer Sandblasting
ken core by means of a fast set- Dry-Mix 1.6* 2.0 0.2 1.3 1.7
ting epoxy. The test sample was
then placed on the testing mecha- Dry-Mix+
nism and a tensile force applied silica fume + 2.0 2.3 0.8 1.1 1.9
until failure occured. A minimum fibers
of six pull-out tests were per- Wet-Mix 1.6 n/a n/a n/a n/a
formed at the ages of two and six
months for each panel tested .
*average from tests at 2 and 6 months
The results of these tests are
grouped and summarized in Table 1. This data is implications for the shotcrete contractor because the
now included in the ACI 506 Guide to Shotcrete. quality of surface preparation, which is a costly
Reference 3 of this paper lists further details of in- operation , can make the difference between good
dividual test results. Each result presented in Table and bad shotcrete bond. Sometimes, contractors
l is the average of at least 12 pull-out tests (6 tests need to place shotcrete in more than one layer. En-
at two months and 6 tests at six months) . gineers , seeing bond between layers as a potential
Apart from the improvement in bond strength source of trouble, and not knowing the best way to
evident from the combined use of fibers and si lica accomplish this bond, are sometimes reluctant to
fume , statistical analysis revealed no significant in- allow the placing of shotcrete in more than one layer.
fluence of the mixture compositions on the bond- The one- layer operation can sometimes cause lo-
ing strength obtained for a given swface prepara- gi stics problems for overhead application: when a
tion. However, the type of surface preparation had thick layer of shotcrete is placed overhead, there is
a significant influence on the bond strength of the a risk of "fresh decohesion" (delamination) (Figure
shotcrete repair. 2) during the finishing operation even if the shotcrete
The highest bond strength was obtained with does not actuall y fall from the surface. When
sandblasted surfaces . While this is not a very prac- decohesion is detected after the shotcrete has set,
tical method for removal of much concrete, it was a the contractor must then remove the un sound
very effective method to improve bond strength of shotcrete and reapply new shotcrete.
shotcrete to concrete . The next highest bond In order to obtain data on this issue, the Indus-
strengths were obtained with surfaces prepared trial Chair on Shotcrete and Concrete Repair at Laval
by hydro-milling, or by jackhammering fo llowed University has undertaken a series of tests on the
by sandblasting. Jackhammering alone did not multi-layer bond strength of shotcrete. The results
seem to produce sufficient bond strength because presented in this section are the res ults of the first
it left a great deal of unsound cracked particles phase test program .
that weakened the interface. Ground surfaces During the study, twelve concrete base slabs were
produced very poor results compared to other prepa- coated with two layers of shotcrete. The first layer
ration techniques. was applied to a sand blasted concrete surface and
Hydro-milling seems to have the advantage of produced excel lent bonding characteristics. That
removing the damaged concrete, leaving the sur- layer was finished in different ways: 1) no finish, 2)
face clean without weakening the surface layer of scratched with a steel trowel (but not finished) , 3)
the old concrete. It is fast, efficient, and requires scratched and finished with a wood trowel, and 4)
less labor than other methods. Chipping with jack- roughened with a broom. Half of each panel was
hammers can potentially weaken the surface, but in coated with a curing compound, either by spray or
this case, this phenomenon was not significant. Low- by using a brush. The second half of the panel of
mass hammers (15 kg/33 lb.) were carefully used, shotcrete was either water-cured or left ot dry. Af-
and sandblasting further helped clean the surfaces
by removing some of the residual fractured concrete
particles. These two methods of concrete surface Figure 2. Pull-off test for
multi-layer shotcrete.
preparation are presently the only ones accepted by
the Quebec Department of Transport for concrete
or shotcrete repairs. 2nd layer of shotcrete
1st layer of shotcrete
Study on Multi-Layer Shotcrete Bond
The bond characteristic of shotcrete is an important
issue for the engineer because it has important im-
plications for repair durability. It can have practical
core

Shotcrete Magazine • May 1999 13

Table 2. Multi-layer bond strength in MPa. probably be similar to the results
from the first study, i.e. be as good
Type of finish between layers (results with no curing compound)
as a single layer shotcrete appli-
Time None Scratch Scratch + Wood Roughen with Broom cation . For technical and practical
reasons the author does not recom-
4 hours 2.1 1.8 2.1 1.9 mend using curing compounds be-
tween .layers, even if the waiting
1 day n/a 2.1* 2.1 n/a
period before applying the second
layer is long.
28 days n/a 1.8 n/a 2.0
Conclusions and
*average of 8 tests instead of 4 tests Recommendations
The test results described in this
paper indicate that the bonding
ter different periods of waiting (4 hours, 1 Table 3 compares the average test re- achieved between good quality shotcrete
day and 28 days) the second layer of sults from Table 2 with the corresponding mixtures and concrete surfaces prepared by
shotcrete was applied and finished with a average bond strength results for the cur- hydro-milling or chipping with light jack-
wood trowel. Pull-out tests were performed ing compound condition and for the single hammers, fo llowed by sandblasting, is gen-
28 days after placing of the second layer shotcrete layer condition. One can see that erally strong and durable . The other types
of shotcrete. there is little reduction in bond strength of concrete removal (grindi ng , chipping
Figure 2 shows the geometry of the when placing shotcrete in more than one with jackhammers without sandblasting)
sample used for this study. Table 2 sum- layer if curing compound is not used; how- resulted in either lower bonding strengths
marizes the results of the bond pull-out ever, there can be significant reduction, in or a reduction in bonding strength over
tests for the half panel made without cur- the order of 50%, if curing compound is time (see reference 3 for more information
ing compound. From Table 2 it can be seen used. The results presented in Table 3 were on bond durability). No significant differ-
that, for the waiting period and the types obtained without removi ng the curing com- ences were observed between the bond
of finish studied, there is no significant in- pound before the application of the sec- strength of dry or wet process shotcrete
f lu ence of these parameters on bond ond layer of shotcrete. However, if the cur- applied on hydro-milled surfaces.
strength. The average bond strength results ing compound was removed with efficient With respect to the multi-layer bond
from Table 2 is (190 psi) 2.0 MPa. sand blasting, the bond strength would strength of shotcrete, the presence of

14 Shotcrete Magazine • May 1999

shotcre te/shotc re te inte rfaces does no t
seem to create a large reduction in shotcrete Conditions Average Bond Strength
quality in terms of mechanical bond if no
c uring co mpo und is used . However, in No curing (from Table 2) 2.0 MPa
saturated conditi ons and in the presence of
freezing and thawing, engineering practice Curing Compound 1.2 MPa
typi cally requires that the bond interface
should be contained between the base con-
One layer (no joint) 2.4 MPa
crete and the outermost reinforcing layer.
Table 3. Average bond strength.
Acknowledgements
The author would like to ex press thanks to References: me nt of Tra ns po rta ti o n of Qu e bec ,
M arc Jolin , Pi e rre Lacomb e, a nd Eri c I. Felt, E.J ., " Res urfac ing and Patching 1992, 86 p.
Cl outier for the ir work in obtaining infor- Concrete Pavement with Bonded Con-
mation included in thi s articl e. The fin an- c rete ," Proceedin gs of Hi ghway Re-
Conversion factors: I MPa = 145 psi
c ial support of the Natural Sciences and search Board , 1956, pp. 444-4 79. Denis Beaupre is an associate professor
Engineering Research Council of Canada in the department of civil engineering at
fo r s uppo rtin g th e Indu strial Ch a ir o n 2. Schrader, E. and Kaden, R., " Durabil -
Laval University in Quebec City. His main
Shotcrete and Concrete Repair is greatl y ity of Sh otcrete,"(SP-1 00-57), Detro it,
research topics are shotcrete, both wet and
appreciated. The author also thanks all in- ACI SP- 100: Concrete Durability, 1987 ,
(by processes, as well as pumping and rhe-
dustri al partners of the Industri al Chair: St- pp . 107 1- 1102.
ology. He also works as a consulting engi-
Laurent Cement, Lafarge Canada, Quebec 3. Talbot, C. , Beaupre, D., Gendreau, M., neer with the "Se rvice d 'Expertise an
Department of Tran sport, C ity of Quebec, L a mo ntag ne A ., a nd Pi geo n, M . Materiaux" (S.E.M.) in Quebec City where
Ambex Technology, King Packaged Ma- " Durabilite des repara ti o ns e n beto n his wo rk p ri ma rily invo l ves shotc rete
te rial , Mas ter Builders Technol ogy, and proj e te (Du ra bi I i ty of sho tcre te re - proj ects.
Sika Canada. 1illoeo pairs) ," Report submitted to the Depart-

Shotcrete Magazine • May 1999 15