Beruflich Dokumente
Kultur Dokumente
concrete structures
M.D. Macdonald and A.J.J. Calder
Additional steel reinforcement can be bonded by epoxy the Quinton bridges. One of the beams was tested to failure
adhesives to hardened concrete to improve its structural without modification and mild steel plates were bonded to
performance. This requirement may be for a number of each of the remaining beams. In each case the steel was
reasons: blast-cleaned to Swedish standard Sa2½ finish 12 and the
• to correct for an error in design; concrete was blast-cleaned to remove surface laitance and
• to correct for an error in construction; cement paste. A structural grade epoxy resin adhesive was
• to maintain or increase the load carrying capacity of the prepared and applied to the surface of the steel plate. This
structure; or was weighted down onto the tensile flange of the concrete
• to constrain cracking. beam which had already been placed in the loading rig (Fig
1). After a minimum period of seven days for the adhesive
Although the technique could be applied to almost any to cure fully and for the beam to be instrumented, it was
concrete structure, the work at Transport and Road loaded to failure. The variables studied included two differ-
Research Laboratories, Crowthorne, UK (TRRL) has been ent adhesives (A and B) and a jointed plate, all using 10 mm
applied to highway bridges I. The first major application to thick plates, and a 6.5 mm thick plate. All the mild steel
bridges in the United Kindom was in 1975 when a group of plates were 140 mm wide and 4200 mm long. Adhesive A
four bridges, at an interchange on the M5 motorway at was used in all but one of the tests. With the jointed plate
Quinton, needed to he strengthened to cater for increased test, the steel bonded to the concrete had a transverse 3 mm
traffic loads 2' a. More recently, in 1977, another pair of gap near mid-span which was covered by a 1220 mm plate
motorway bridges at Swanley were strengthened a'4. The bonded symmetrically over the joint using the same adhesive
results of flexure tests at TRRL on beams with steel plates During the tests the performance of the beams was moni-
bonded to their tension flanges s' 6 were used to provide tored using a variety of instrumentation.
advice for these projects. Subsequent laboratory work has
provided more detailed information on the structural per-
Results
formance of this technique 7 and an extensive programme
of exposure testing is in progress s. Other UK research into The overall behaviour of the plated beams was similar. As
the bonding of steel to concrete using epoxy adhesives has the applied load was increased, tension cracking of the
been reported and includes flexure tests on beams and concrete occurred at mid-span and developed through the
slabs 9, lo, 11. section of the beam. The spacing of the cracks was influ-
enced by the presence of shear reinforcement stirrups with
Flexure tests on 4.9 m beams intermediate cracks developing at high loads. Eventually,
the concrete failed in horizontal shear at one end of a
The first phase of laboratory work carried out at TRRL s'6 reinforcement plate allowing separation to occur; this was
was to test some externally strengthened reinforced concrete followed by the plate springing away from the beam up to
beams to provide data for the proposed strengthening of about mid-span. Varying depths of concrete, including the
0143-7496/82/020119-09 $03.00 © 1982 Butterworth & Co (Publishers) Ltd
1, 2 Not strengthened w
As-cast
3, 5 Adhesive* A 85 x 3.60 24
4,6 B 85 x 3.60 24
7,8 C 85 x 3.60 24 Plated, as-cast
10 Plate geometry B 150 x 1.06 142
11 B 123 x 2.14 57
13 B 57 x 4.75 12
14 Reference B 85 x 3.60 24
15 B 150 x 1.06 142
16 Jointed B 85 x 3.60 24 Plated, precracked
17 Thick glue-line B 85 x 3.60 24
18 Partial bonding B 85 x 3.60 24
19 Multilayers B 3 (47 x 2.0) 24/12/8
/ Cracking
When plates were bonded to beams as-cast using a stiff
adhesive, loads to produce the onset of visible cracking were
increased on average by 100% compared with the as-cast
control beam, and loads to produce a minimum crack
width of 0.1 mm were increased by about 190%. Hated
beams with a stiff adhesive generated more cracks at a closer
c. Partial bonding d, Multi-layer=
spacing than either the as-cast beams or plated beams using
Fig. 2 Variations in plate details (dimensions in ram) a flexible adhesive. When a flexible adhesive was used,
crack spacing was the same as for as-cast beams (probably,
at points of high strain and incipient cracking the adhesive
in the constant moment span and then held in this condition. distorted locally and allowed cracks to form, whereas the
The prepared plate was then added to the beam, the instru- stiffer adhesive distributed the strains). When plates were
mentation completed, and loading continued. bonded to precracked beams, the subsequent cracking
behaviour was modified to give a greater number of cracks
Results at a closer spacing. The number of cracks developing in the
constant moment span and their average spacings are given
Deflections in Table 2 for all tests, together with the load required to
When assessed by their load vs deflection curves, the overall produce the first visible crack and a minimum crack width
performance of beams plated as-cast was virtually unaffected of 0.1 mm.
Test number* Number of cracks Average space between cracks Load to produce:
(mm)
before plating after plating first visible minimum crack
before plating after plating crack width, 0.1 mm
(kN) (kN)
1 8 - 122 - 6 12
2 9 - 116 - 7 12
3 - 11 - 94 18 42
4 - 10 - 96 10 N/A
5 - 10 - 98 14 32
6 - 11 - 95 14 42
7 - 9 - 121 8 32
8 - 9 - 114 6 26
10 - 10 - 106 12 24
11 - 10 - 104 14 N/A
13 - 9 - 116 10 34
14 8 12 133 78 7 12
15 5 9 197 116 6 9
16 7 12 156 85 7 10
17 8 10 131 102 7 13
18 6 7 167 154 6 7
19 8 8 129 129 6 13
e . . .
Conditions as m Table 1
N/A Specified condition not achieved
'°i
3O
_o
y i
ii •* As-cart
A$-cist
85 x 360
123x214 57
24
t0
i ~ As.cast 150x106 142
10 20 30 40 50 60 70
Deflection (ram)
40i
z 30
30
z
v
~o
.=
~-Onset of cracking
I I I I I I I I I
2O0 400 60O 800 1000
Concrete turfaee strain at mid-sglm (u~)
124 I N T . J . A D H E S I O N A N D A D H E S I V E S A P R I L 1982
Micro-structural analysis
I D ,oa,~ [] o°,o.,., I
A microstructura] investigation was carried out at the
Year 2
50 Year 1 Year 2 Year 1 Year 2 Year 1 Year 2 Year 1 National Physical Laboratory to determine the mechanism
which has caused corrosion of the epoxy-bonded steel
plates 14. Visual inspection of the concrete revealed discrete
40
voids and air bubbles, and staining techniques were used to
extend the investigation down to the micro-structure. Small
30 triangular sections of epoxy resin were immersed in various
marker solutions for periods of up to six months. Methylene
blue, a water soluble organic strain, and silver nitrate rapidly
3 penetrated around interfaces between filler particles and the
c 20
Variate Source of variation Degreesof Sum of squares Mean square Variance Significance
freedom ratio level
50
r-
!
40 l
F
z
J¢ 30
I
dl
~0
!
:t
10 :/i i Fig. 17 Electron micrograph of shadowed replica from epoxy
resin containing a silica particle (× 3000)
Conclusions
0
2 4 I 2 2 3 A study of reinforced concrete beams tested in 4-point
S~'im: Teated soon I Tested I Tastedafter 1 year bending showed that a fully composite action could be
after plating of e x p m u r e at ~;ilverdale of exposure at Pilley
achieved when steel plates were bonded externally to the
Fig. 14 Mean failure loads of plated beams tension faces of the beams. This led to a significant change
in the cracking response of the beam and the ultimate loads
achieved. The following points emerged:
r//////x/// A progressive, soft failure can be achieved by using a
' ' I ~ Particle j ~ / wide plate.
To obtain the maximum increase in the serviceability
Broed fron! penetretlon limit conditions of stiffness and crack control, a fully bonded
plate using a stiff adhesive is necessary.
through golished face To combine the maximum increase in ultimate load with
a progressive soft failure, the plate width-to-thickness ratio
Fig. 15 Penetration of methylene blue through resin had an optimum value of about 60 for the type of beam
tested.
Broad-front )enatration R s-polilhed l u r f k - e
A significant amount of corrosion at the steel/resin
interface has taken place on all specimens which have been
exposed to natural environments. This has led to some
a reduction in strength of the exposed specimens compared
with specimens kept in the controlled laboratory environ-
ment where no corrosion has taken place.
The corrosion was fairly uniformly distributed over the
steel plate implying that moisture had penetrated through
the concrete and the resin rather than by seeping between
Penetration along crack
the steel and the resin at the edge of the plate.
Fig. 16 Penetration of methylene blue through crack The microstructure of the epoxy layer consists of a
labyrinth of very time cracks which are likely to be inherent
in the resin rather than caused by natural exposure or
stressing of the systems. These cracks are likely to have
cracks and pores has provided routes for rapid diffusion of provided routes for corrosive moisture to reach the steel
• the staining materials through the epoxy resin. plates.
A sample of the end of one of the steel plates was Application of a chromate primer to the surface of the
removed from a strengthened bridge after three years in steel plate appeared on limited evidence to have been gen-
service. Corrosion of the steel which contained traces of erally effective in preventing corrosion of the steel without
chloride and sulphate, was detected under the stin-sdherent greatly affecting the structural strength of the member.
epoxy resin. Microstructural analysis of a portion of the
epoxy resin revealed a system of similar cracks and discon- Ackno~ts
tinuities as to those observed on the small test beams. The
concrete and the epoxy resin were generally more defective The microstructural analysis of the composite system was
than in the experimental exposed beams, but there was less carried out by Mr G.O. Lloyd of the National Physical
corrosion on the surface of the steel. Laboratory.
I N T . J . A D H E S I O N A N D A D H E S I V E S A P R I L 1982 127