Вы находитесь на странице: 1из 8

Proceedings of the Institution of

Civil Engineers
Structures & Buildings 158
February 2005 Issue SB1
Pages 77–84

Paper 13993
Received 02/12/2003
Accepted 25/05/2004
Keywords: Deric J. Oehlers Irene Liu Rudolf Seracino
composite structures/concrete Associate Professor, Adelaide Postgraduate student, Adelaide Senior Lecturer, Adelaide
structures/design methods & aids University, Australia University, Australia University, Australia

Shear deformation debonding of adhesively bonded plates


D. J. Oehlers PhD, MSc, MIEAust, I. S. T. Liu BSc and R. Seracino PhD, MASc, MIEAust

Adhesive bonding of fibre reinforced polymer (FRP) and PIC IC debonding resistance; IC debonding force
steel plates to the surfaces of reinforced concrete Ppres prestress force
structures is an efficient form of retrofitting, as the Py yield capacity of the steel plate; Ap fy
plated structure is mechanically efficient, the plate is RC reinforced concrete
unobtrusive, and—for FRP plates—the plate is durable sfp side face plate
and the application is inexpensive owing to the lightness tfp tension face plate
and flexibility of the pultruded or wet lay-up plate. In this tp plate thickness
paper it is shown: from a comparison of the main FRP V shear
plating guidelines, that there is now a general agreement VAy interface shear stress approach
on the major plate debonding mechanisms; that current Vc shear capacity without stirrups; concrete component
rules in national standards for the shear capacity of of shear capacity
prestressed beams can be used to quantify a major plate Vcrack shear to cause a diagonal crack
debonding mechanism due to beam shear deformations, (Vincr )pp passive prestress shear capacity increase due to
that is critical diagonal crack (CDC) debonding; and that plating
these developments allow a simple design approach to (Vpp )ACI prestressed beam concrete shear capacity from ACI
be used for FRP and steel plated beams and slabs with approach
longitudinal plates. (Vpp )Euro passive prestress concrete shear capacity from
Eurocode approach
NOTATION Vpres shear capacity passive prestress approach
Ap cross-sectional area of plate Vslide shear to cause sliding across the diagonal crack
Ast area of tensile reinforcement Vtest shear capacity from test
a shear span Vunpl Zhang’s concrete shear capacity of the unplated
b width of beam beam
bp width of plate x position of the diagonal crack root from applied
CDC critical diagonal crack concentrated load
cfp compression face plate y distance to compressive face
d effective depth of beam Æ beam or pull-test coefficient
dps prestressing force position L partial anchorage parameter
Ep longitudinal Young’s modulus of plate p effective width parameter
F force  strain
Fps prestressing force c concrete crushing strain
FRP fibre reinforced polymer db IC debonding strain
fc cylinder compressive strength of concrete frac fracture strain of the FRP
f frac fracture stress rebar fracture strain of the rebar
ft tensile strength of concrete Ł angle of inclination of the weakest diagonal crack
f tef effective tensile strength of concrete in beam r area of reinforcing bars
fy yield stress  horizontal normal concrete stress; stress
h depth of the beam  IC IC debonding stress
IC intermediate crack p prestress
ku d depth of neutral axis  pp passive prestress
L anchorage length  shear stress
Le minimum anchorage length to achieve maximum PIC Rd basic design shear strength
mp modular ratio of plate material to concrete
Paxial maximum axial force in the plate 1. INTRODUCTION
PE plate end The mechanism that induces plate debonding due to rigid body
Pfrac fracture strength of the FRP plate; Ap ffrac shear deformations in a beam is illustrated in Fig. 1. The rigid

Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al. 77
of carbon FRP or steel plates, as the rigid body displacement
Critical
diagonal across the CDC causes the plates to debond before the stirrups
crack can resist any significant amount of shear.
V
V Rigid body
displacement The main aim of this paper is to show how existing design
5
(crack sliding) codes, and in particular the Eurocode model for the concrete
Direction of crack propagation component of the shear capacity of prestressed beams, can be
CDC debonding crack
easily adapted and used for determining the CDC debonding
resistance of plated beams. This use of prestress codes has
Fig. 1. Critical diagonal crack (CDC) debonding mechanism allowed the development of a simple design procedure for
adhesive bonding plates of any type to any surface of a beam.
Because it is suggested that designers will feel more
comfortable and more confident in being able to use both
body displacement or crack sliding across a critical diagonal established standards as well as simple design procedures for
1
crack (CDC) causes the plate to debond; the CDC is the crack longitudinal plating beams and slabs, it is felt this will help
associated with the shear capacity of a beam without stirrups— promote the use of this effective retrofitting technique.
that is, the concrete component of the shear capacity, Vc . An
example of CDC debonding of a tension face plate in a hogging
region is shown in Fig. 2, where the sudden appearance of a
2. GUIDELINES FOR PLATED BEAMS AND SLABS
CDC caused the plate to debond. The CDC crack should not be
confused with the flexural or flexural/shear cracks that are also
2.1. Debonding mechanisms
visible and which can also induce interface cracking that can
The major debonding mechanisms are illustrated in Fig. 3(a).
lead to intermediate crack (IC) debonding, which is another
2–4 These have been referred to as:
form of major debonding. Tests have shown that the
presence of stirrups does not prevent or inhibit CDC debonding
(a) intermediate crack (IC) debonding, which is associated with
the axial strains in a plate that intercepts an intermediate
Interface
crack such as a flexural or flexural/shear crack
cracking (b) critical diagonal crack (CDC) debonding, which is
Flexural
associated with the shear deformations across a CDC that
crack CDC debonding crack intercepts the plate
Flexural/ Critical (c) plate end (PE) debonding, which is associated with the
shear diagonal crack
curvature in the plate adjacent to its end
crack
(d) VAy debonding, which is associated with the interface
shear stress away from intermediate cracks, and which can
Fig. 2. CDC debonding amongst flexural and flexural/shear
cracks be derived from elementary structural mechanics using the
well-known VAy/Ib or VQ/Ib equations.

Point of contraflexure Point of contraflexure

Uncracked section Uncracked section


Cracked section

4
5
1 2 3
(a)

VAy CDC PE
IC

(Mechanism 1) (Mechanism 2) (Mechanism 3) (Mechanism 4) (Mechanism 5)


AUST:4 IC AUST: IC AUST: IC AUST: CDC AUST: PE
debonding debonding debonding debonding debonding
EUR8 mode 1: EUR mode 2: EUR mode 2: EUR mode 3: EUR (mode 6):
peeling off in peeling off at peeling off at peeling off caused concrete rip-off
uncracked flexural crack flexural crack by shear cracks BRIT:
anchorage BRIT: BRIT: EUR (mode 5): peeling failure
BRIT:7 debonding failure debonding failure, plate-end HK:
debonding failure HK: peeling failure shear failure concrete cover
HK:6 intermediate crack HK: BRIT: referenced separation
intermediate crack induced debonding intermediate crack HK: plate end
BRIT: referenced
induced debonding USA: referenced induced debonding interfacial
USA: referenced USA: referenced USA: not recognised
USA:9referenced.

(b)

Fig. 3. Comparison of debonding failure mechanisms: (a) guidelines; (b) debonding mechanisms

78 Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al.
2.2. Guideline approaches to debonding mechanisms shear capacity, Vc , are proposed. However, there is slight
Numerous papers have been published on adhesive bonding confusion with the European approach, which gives this
longitudinal plates, and most manufacturers give their own or mechanism two labels in Fig. 3(b) and describes plate-end
refer to guidelines. However, this comparison is restricted to shear failure (identified as mode 5) using Fig. 4. This figure
the five approaches listed in Fig. 3(b), which consist of the shows a simply supported beam with plates terminating almost
Australian research brought together and documented in a at the supports and with a horizontal debonding crack on the
4
book as a design guide, the Hong Kong research and design right-hand side and a diagonal crack beyond the plate end on
6 7 8
guide, also issued as a book, and the British, European and the left-hand side. The authors are familiar with plate end (PE)
9
US guidelines. It is felt that these are fairly comprehensive debonding as in Fig. 3(b), which looks similar to the debonding
and reasonably independent. crack on the right side of Fig. 4, but this occurs only when the
plate is terminated in a region of high curvature, which does
The five debonding mechanisms encircled in Fig. 3(a) were not exist at the supports of the simply supported beam.
extracted from the five guidelines. Also included within a box Combining the two cracks in Fig. 4 resembles the CDC
is VAy debonding, which is rarely referred to in guidelines and debonding failure of a tension face plate that was terminated
from the author’s experience rarely if ever occurs in tests, as close to the supports in Fig. 5. Whether or not this is the plate-
intermediate and critical diagonal cracks induce stress end shear failure described in the European guideline, this
concentrations of a much larger magnitude. However, it is felt failed beam shows that it may be difficult to find an uncracked
that VAy debonding could become important in prestressed region of a beam within which to anchor the plate end as
beams or beams with thick plates where debonding might required in the European guideline, because shear cracks can
occur before the occurrence of flexural or shear cracks. The occur near points of contraflexure.
names used for each of the five debonding mechanisms in Fig.
8 8, 11
3(b) may differ between the guidelines, but there appears to be In the European guidelines, Blaschko in a seminar paper
an almost unanimous agreement on the mechanisms of suggested that peeling off at shear cracks may be prevented by
debonding, which is a very important first stage in the limiting the shear force to the concrete component of the shear
development of design rules. The US guidelines do not describe resistance of RC members, and suggested adopting code
5
the failure modes directly but do refer to three conference requirements by transforming the cross-sectional area of plate
papers that clearly describe failure Mechanisms 1–4; this has into an equivalent steel area of longitudinal reinforcement of the
been referred to as referenced in Fig. 3(b). Similarly, the British same axial stiffness but not axial strength. Hence this approach
guideline also refers to a conference paper for Mechanism 4. does not consider the maximum axial force that the plate can
Much of the agreement would appear to ensue from an resist, and does encourage the use of thick plates, which tend to
10
excellent conference paper by Blaschko et al. The European debond at lower stresses. Furthermore, the bond or anchorage
approach makes the distinction between the cracked region of behaviour of externally bonded plates is totally different from
4,8
a beam and the uncracked region in Fig. 3(a), and it is a that of the internal reinforcing bars encased in concrete. Hence
requirement that the plate is anchored in the uncracked region. converting externally bonded FRP plates, which have a brittle
This leads to the three IC debonding failure regions: at the bond characteristic and a brittle material characteristic, to
anchorage zone (Mechanism 1); at the position of maximum embedded steel reinforcing bars, which have ductile bond and
moment (Mechanism 3); and at flexural/shear cracks material characteristics, should be used with care.
(Mechanism 2).
8 12
Also in the European guidelines, Jansze in his PhD thesis
2.3. Intermediate crack and plate end debonding computes an effective concrete component of the shear
For Mechanism 1 in Fig. 3(b), there is a general agreement in
calculating the axial force in the plate allowing for the bond
length. There is also general agreement for Mechanisms 2 and Diagonal
3, although in the latter there is a slight confusion with the crack

British approach, which first refers to it as debonding failure


but then suggests using the equation for peeling failure in this
region.
Horizontal debonding crack
For Mechanism 5, there is also general agreement. The
8
European approach describes this failure, identified as mode 6 Fig. 4. FIB bulletin 14 diagram for ‘FRP plate-end shear failure’
in the box, but does not give design rules and, furthermore, the
US approach does not recognise this failure mechanism. This is
probably because plate end (PE) debonding is very unlikely to
control design when using FRP plates as they are usually very
thin. Also, the European approach requires termination of the
plate in an uncracked region, which by definition will be at or
close to a point of contraflexure, and hence the curvature, a
requirement for PE debonding, is minimal.

2.4. Critical diagonal crack debonding


There is general agreement for Mechanism 4 in Fig. 3(b) where Fig. 5. CDC debonding of tension face plated beam
design rules originating from the concrete component of the

Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al. 79
13
resistance of the beam that does not appear to depend on the Mohamed Ali et al. then adapted Zhang’s approach to allow
cross-section of the FRP plate but on the extent of plating, and for the effect of steel plates on the concrete shear capacity. This
which suggests that if the plate is terminated at a support then was accomplished by allowing for the increase in the shear
the shear capacity is infinite. However, terminating a plate at a capacity due to the maximum possible axial force in the plate,
support as shown in Fig. 5 does not prevent shear failure or Paxial in Fig. 6, which is the lesser of the yield capacity of the
debonding. Furthermore, as can be seen in Fig. 2, plating does steel plate, Py , the fracture strength of the FRP plate, Pfrac , or
not prevent the formation of a critical diagonal crack within the IC debonding resistance of the plate, PIC . The latter depends
the plated region, although it does tend to force the root of the on the position of the diagonal crack, as this controls the
4, 13
diagonal crack towards the plate end. anchorage length of the plate, such as LB for position B. It can
now be seen how the IC debonding resistance of the plate
6
In the Hong Kong approach, Teng et al. restrict the vertical affects the CDC debonding resistance, and also why Zhang’s
15
shear to 1.4 times the concrete shear capacity of the unplated approach was used, as it defines where the critical diagonal
beam, Vc , which is reasonable as the addition of longitudinal crack is and hence the anchorage length and consequentially
13
plates can easily increase the shear capacity by 40%, but it is the plate IC debonding resistance.
felt that this is more of a rule of thumb guidance.
13
Mohamed Ali et al. converted the plate axial force into an
equivalent area of reinforcing steel by dividing Paxial by the
2.5. Summary yield capacity of the steel plate. This approach was found to
Prior to the recognition of the fundamental debonding 13
give good predictions for the concrete shear capacity, which
mechanisms illustrated in Fig. 3, journal and conference papers is the CDC debonding resistance. However, this approach
would reference different generic approaches, such as the shear cannot be applied to FRP plates as they do not have a yield
approach or the fracture mechanics approach, as possible capacity. Instead it is proposed that the axial force in the plate,
solutions to the overall problem. However, it can now be seen Paxial , be treated as a prestressing force. This prestressing force
that specific approaches are required and are now being acts only after the diagonal crack is formed in Fig. 6, and it is
developed for each unique debonding mechanism. The IC induced through separation of the crack faces due to aggregate
debonding resistance can be determined directly from pull interlock: hence the term passive prestress. Both approaches
tests, which give a lower bound to the IC debonding resistance are correct for steel plates, as it can be shown through Zhang’s
6
in beams, or more advanced techniques are being developed mathematical models that there is an equivalence in
16

to determine the increased strains associated with IC debonding Mohamed Ali’s additional longitudinal reinforcement approach
8, 14 4
in beams. Design rules are also available for PE and the passive prestress approach. However, the passive
debonding, which can also be prevented by terminating the prestress approach can be used for brittle FRP materials, and
plate close to a point of contraflexure. therefore has a much wider application.

The following sections deal with critical diagonal crack


debonding, and conclude with a simple design procedure that 3.2. Steel plate passive prestress CDC debonding
17
covers all debonding mechanisms. In a companion paper, Zhang’s approach has been directly
adapted into the passive prestress approach, and the theoretical
shear capacities, Vpres , have been compared with 55 steel plated
3. CDC PASSIVE PRESTRESS APPROACH beam tests, Vtest , in which it was known that critical diagonal
cracks formed. The results are reproduced in Fig. 7, where Vunpl
3.1. Passive prestress CDC model is the concrete shear capacity of the unplated beam based on
The passive prestress model is illustrated in Fig. 6; it is based 15
Zhang’s original approach. It can be seen that there is good
15
on the premise by Zhang that the concrete component of the correlation between test results and the passive prestress model,
shear capacity depends on the shear to cause a diagonal crack, bearing in mind that Zhang’s approach is meant to be a lower
Vcrack , and the shear to cause sliding across the diagonal crack, bound, and that plating has increased the concrete shear
Vslide . The angle of inclination, Ł, of the weakest diagonal capacity by up to 150%. It is also worth noting that the test
crack occurs where Vcrack ¼ Vslide , and this is referred to as the results included tension face plates (tfp), side face plates (sfp),
critical diagonal crack.

1·6
Possible range of critical diagonal cracks
1·4

θ 1·2
Shear resisted by B 1
Vtest/Vpres

aggregate interlock A
h
V V 0·8
Reinforcing bar Critical 0·6
Plate diagonal
0·4 tfp
crack sfp
0·2 angle
tfp+sfp
Paxial  0 Paxial Paxial Paxial Paxial cfp
0
LB LA 0 0·5 1 1·5 2 2·5 3
x Vtest/Vunpl

17
Fig. 6. Passive prestress model Fig. 7. CDC debonding of steel plated beams

80 Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al.
angle sections, beams with both tension face plates and side The problem with both ACI approaches is that they are specific
plates (tfp + sfp), and compression face plates (cfp). to prestressed beams so that, as both p and pp tend to zero,
Furthermore, in these tests the plate thickness was varied from the shear capacity does not tend to the shear capacity of
1 mm to 20 mm, the depth of the beam from 180 mm to unprestressed beams as given by the ACI, so there will be a
370 mm, and the shear span from 500 mm to 1200 mm, and step change in Vc when the resulting normal stress  ¼ p +
the CDC debonding resistances varied from 26 kN to 169 kN. pp ¼ 0. It is felt that this will not be suitable for plated
structures because, as the area of the plate tends to zero and
Generic equations for the shear to cause cracking, Vcrack , and hence Paxial ! 0, the concrete shear capacity also tends to that
the shear resistance to sliding, Vslide , applicable to all of the unplated unprestressed section. Furthermore, the passive
configurations of plating have been produced in the prestress pp ¼ Paxial /bh occurs only after the formation of a
17
companion paper. In their simplest form for simply diagonal crack, which would also exclude the use of the
supported beams with tension face plates, the generic passive rigorous approach and possibly the direct approach. It needs to
prestress equations become be stressed that there is nothing wrong with the conceptual
approaches of the ACI. It would appear that the ACI is
   recognising the fact that, for prestressed beams, the shear load
bf tef f t mp Ap
1 V crack a ¼ ð x 2 þ h2 Þ þ þ ð Fps dps Þ to cause cracking, Vcrack , as represented by equation (1), is
2 h
generally greater than the shear load to cause crack sliding
after the crack has formed, Vslide , as represented by equation
2 3 (2): hence the importance of determining when cracks occur in
  sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2ffi
F þ Paxial 4 x x prestressed beams.
 5
ps
V slide ¼ 0:4 f c bh 1 þ 2 1þ
2 f c bh h h
18 5
In contrast to the ACI approaches, the Eurocode approach is
3 f 1 ð f c Þ f 2 ð hÞ f 3 ðrÞ applicable to both prestressed and unprestressed beams and is
given in the following form, with an additional term of 0.15pp
for the passive prestress:
where a is the shear span, x is the position of the diagonal
crack as in Fig. 6, h is the depth of the beam, b is the width of
ð V pp ÞEuro ¼
the beam, ftef is the effective tensile strength of the beam, ft is    
the tensile strength of the concrete, mp is the modular ratio of 4 40Ast
Rd 1 6  d 1 2 þ
ð : Þ : : :
þ 0 15 p þ 0 15 pp bd
the plate material to the concrete, Ap is the cross-sectional area bd
of plate, Fps is the tendon prestressing force, dps is the tendon
prestressing force position, fc is the cylinder compressive
strength of the concrete, and r is the area of the reinforcing where Rd is the basic design shear strength. As (Vpp )Euro ! Vc
bars. An iterative approach has to be used to determine the as p + pp ! 0, this passive prestress code approach will be
position x, which occurs where the equations are of equal used and compared with test data. It can be seen in equation
magnitude and which is the position of the critical diagonal (4) that the increase in the shear capacity due to plating is Vincr
crack. This approach is cumbersome for design, but may be ¼ 0.15pp bd, and as pp ¼ Paxial /bd then
necessary when complex plating systems are being used. Hence
there is a need for a simpler code-based approach, as follows. ð V incr Þpp ¼ 0:15Paxial
5

3.3. Prestress code rules adapted for CDC debonding


18
The American Concrete Institute (ACI) has two code models, That is, the increase in the concrete shear capacity is directly
both of which are specific to prestressed beams. The ACI has a proportional to the maximum axial force in the plate.
rigorous approach in which the theoretical concrete shear
capacity is achieved when the principal tensile stress at the 4. COMPARISON OF EUROCODE RULE WITH TEST
centroid of the uncracked composite section is equal to the DATA
17
cracking strength of 0.33ˇfc . Mohr’s circle can then be used to For the steel plated beams in the companion paper, a large
determine the load to cause cracking of an element that is population of tests were available in which it was known that
subjected to a shear stress  ¼ VAy/Ib (or VQ/Ib) and a critical diagonal cracks occurred, and these were used to
horizontal normal concrete stress  due to the prestress force validate the crack sliding model as in Fig. 7. These tests will be
Ppres of p ¼ Ppres /bh. The ACI also has a direct approach in used to validate the passive prestress code approach. A similar
which the equation for the concrete shear capacity in regions population for FRP plated beams in which CDC debonding
of low applied moment is given in the following form, with the occurred is not available.
inclusion of an additional term, 0.3pp , for the passive prestress
where pp ¼ Paxial /bh: The maximum axial force in a plate of cross-sectional area
Ap is Paxial , which is the lesser of the yield capacity, Py ¼
 pffiffiffiffiffi  Ap fy , the fracture capacity, Pfrac ¼ Ap ffrac , and the IC
3 ð V pp ÞACI ¼ 0:29 f c þ 0:3 p þ 0:3 pp bd debonding resistance, PIC ¼ Ap IC . The IC debonding
resistance of a plate affects not only the flexural capacity,
such as at Mechanism 3 in Fig. 3, but it also affects the
where both p and pp act over an area bd, where d is the passive prestress across the diagonal crack, such as at
effective depth to the outermost tensile steel reinforcement. Mechanism 4 in Fig. 3. From an extensive study of the IC

Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al. 81
effective width of the
bp (b/2)1 (b/2)2 concrete section to which the
(b/2)1 (b/2)2 (b/2)1 (b/2)2
plate is attached (as shown in
b tp bp/2 bp/2 (b/2)2 Fig. 8) and which for non-
tp
bp/2 bp/2 symmetrical plated sections is
bp
bp the lesser of the two values
4
(b/2)1 (b/2)1 shown. The IC debonding
b
bp
resistance depends on the
b/2 (b/2)1 (b/2)2
anchorage length, L, between
(a) (b) (c)
the intermediate crack and
the plate end, such as LB in
Fig. 8. Effective widths: (a) symmetrically plated; (b) non-symmetrically plated; (c) adjacent plates Fig. 6. The following
minimum anchorage length
or effective length, Le , is
required to achieve the maximum IC debonding stress:
1·4

1·2
sffiffiffiffiffiffiffiffiffiffi
Ep t p
Le ¼ pffiffiffiffiffi N and mm
Vtest/(Vunpl  Vincr)pp

1.0 8
fc
0·8

0·6
and hence the L factor in equation (6) allows for plates that
0·4 are partially anchored, as follows
sfp (0·15Paxial)
0·2
tfp (0·15Paxial)

0 1 if L . Le
0 0·5 1·0 1·5 2·0 2·5 3·0 9 L ¼
Vtest/Vunpl
sin ðL=2Le Þ if L , Le

Fig. 9. Comparison of steel plated tests with 0.15Paxial factor


The coefficient Æ in equation (6) depended on whether IC
debonding occurred in pull tests or in beam tests. For pull tests,
which are equivalent to a beam with a single intermediate
Point of contraflexure Point of contraflexure crack, Æ has a mean value 0.427 with a 5% characteristic value
Shear of 0.315, whereas from beam and slab tests, where the
interaction between adjacent intermediate cracks is
tp Le Moment 8, 14, 17
Le known to increase the IC debonding resistance, Æ has a
Le
bp
Le mean value 0.887 with a characteristic value of 0.379.
Le Le
17
The test data in the companion paper for side plated (sfp) and
Hogging region Sagging region Hogging region tension face plated (tfp) beams are compared with the passive
prestress CDC debonding resistance from the adapted Eurocode
Fig. 10. Simple design procedure model in equations (4) and (5). In the test data population the
plates were not always fully anchored: that is, the anchorage
length, L, was not always greater than the effective length, Le ,
debonding resistance in both pull-tests and beams, Chen and in equation (9). Hence the maximum force in the plate, Paxial ,
6, 19
Teng developed the following equation for the IC was determined from the crack sliding analyses in the
debonding stress: companion paper, and this was used to predict the increase in
the shear capacity due to plating in equation (5). Furthermore
sffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffi in deriving PIC , the value of Æ ¼ 0.427 from pull tests was
Ep f c
6  IC ¼ ÆL p used, as a pull test represents a beam with one intermediate
tp crack, which is the case for a single diagonal crack intercepting
a plate. The results are shown in Fig. 9, and it can be seen that
there is very good correlation for both the side face plates and
where Ep is the longitudinal Young’s modulus of the plate, fc is the tension face plates, and that plating can substantially
the cylinder compressive strength, and tp is the plate thickness. increase the concrete shear capacity, and hence the resistance
The parameter p in equation (6) is given by to CDC debonding. Because of this good correlation with test
data, it is suggested that the European model of equations (4)
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi and (5) can be used to predict CDC debonding.
2  bp =b
7 p ¼
1 þ bp =b
5. SIMPLE DESIGN APPROACH FOR STRENGTH
A simple design procedure is illustrated in Fig. 10. As the
position of the critical diagonal crack is not known, the region
which depends on bp , the width of the plate, and b, the (hogging or sagging) has to be fully plated and—importantly—

82 Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al.
Possible
pivotal strains ε (strain) σ (stress) F (force)

F1 (y1)
εc
kud F2 (y2)
d

εrebar
F3 (y3)
F4 (y4)
Tension εdb εfrac
εdb Pivotal
face plate
point
(a) (b) (c) (d) (e)

Fig. 11. Flexural analysis of tension face plated beam

the plate has to be fully anchored, so that Paxial is constant However, it is recommended that only prestress code models
throughout the region and is the least of Py , Pfrac and PIC fully for the concrete shear capacity that can be applied directly to
anchored. Hence it is recommended that the plates are both prestressed and unprestressed beams, such as the
5
extended beyond the points of contraflexure (or at least Eurocode model, should be used.
beyond a region of low moment) by an effective length Le
given by equation (8).
REFERENCES
As the plates are terminated near to the points of 1. OEHLERS D. J., PARK S. M. and MOHAMED ALI M. S. A
contraflexure, plate end (PE) debonding as illustrated in Fig. structural engineering approach to adhesive bonding
3 will not occur. Critical diagonal crack (CDC) debonding can longitudinal plates to RC beams and slabs. Composites Part
be prevented by ensuring that the shear force within a region A, 2003, 34, 887–897.
is less than that given by equation (4), in which pp ¼ Paxial / 2. OEHLERS D. J. Development of design rules for retrofitting
bd and where Paxial is the least of Py , Pfrac and PIC fully by adhesive bonding or bolting either FRP or steel plates to
anchored. The latter can be obtained from equation (6) with RC beams or slabs in bridges and buildings. Composites
L ¼ 1 and with Æ ¼ 0.315: that is, the characteristic value Part A, 2003, 32, No. 9, 1345–1355.
in pull tests. Alternatively, the concrete shear capacity, the 3. OEHLERS D. J. and MOHAMED ALI M. S. Critical diagonal
resistance to CDC debonding, can be assumed to be that in crack debonding of adhesively bonded plates: the shear
the national code plus 0.15Paxial . peeling mechanism. Proceedings of the International
Conference on FRP Composites in Civil Engineering, CICE
Having now ensured that PE and CDC debonding do not occur, 2001, Hong Kong, 563–570.
the flexural capacity can be determined where the maximum 4. OEHLERS D. J. and SERACINO R. Design of FRP and steel
axial force in the plate is either its yield capacity, its fracture plated RC structures: retrofitting beams and slabs for
capacity or its IC debonding resistance from equation (6) with strength, stiffness and ductility. Elsevier, Oxford, 2004.
Æ ¼ 0.379: that is, the characteristic strength in beams. A 5. BRITISH STANDARDS INSTITUTION. Eurocode 2. Design of
4
standard routine analysis for the flexural capacity is Concrete Structures. Part 1: General Rules and Rules for
illustrated in Fig. 11, where the strength can be controlled by Buildings. BSI, Milton Keynes, 2002, prEN 1992-1-1.
any of the strains in Fig. 11(b): that is, the concrete crushing 6. TENG J. G., CHEN J. F., SMITH S. T. and LAM L. FRP
strain, c ; the fracture strain of the rebar, rebar ; the fracture Strengthened RC Structures. John Wiley & Sons,
strain of the FRP, frac ; and the IC debonding strain, db , which Chichester, 2002.
can be obtained from equation (6) with Æ ¼ 0.379 for FRP 7. CONCRETE SOCIETY. Design Guidance for Strengthening
plates and elastic steel plates. However, if the steel plates are Concrete Structures Using Fibre Composite Materials.
designed to yield prior to IC debonding then the debonding The Concrete Society, Crowthorne, 2000, Technical Report
4
strain, db , can be much greater. In most cases for FRP plates No. 55.
the debonding strain governs the analysis, as in Fig. 11(c). It is 8. TASK GROUP 9.3. Externally Bonded FRP Reinforcement for
also worth bearing in mind that plate debonding often occurs RC Structures: Design and Use of Externally Bonded Fibre
at low concrete strains, so that allowance must be made for Reinforced Polymer Reinforcement (FRP EBR) for
part of the concrete being pseudo elastic and other parts Reinforced Concrete Structures. Federation Internationale
plastic, as shown in Fig. 11(d). du Beton, Lausanne, 2001, fib bulletin 14.
9. ACI COMMITTEE 440. Guide for the Design and Construction
6. SUMMARY of Externally Bonded FRP Systems for Strengthening
The research has shown that the axial force in an FRP or steel Concrete Structures. American Concrete Institute,
plate can be considered as an additional prestress, so that the Farmington Hills, MI, 2002, Emerging Technology Series,
concrete component of the shear capacity of prestressed beams ACI 440.2R-02.
can be used to design against CDC debonding. The implication 10. BLASCHKO M., NIERDERMEIER R. and ZILCH K. Bond failure
of this research is that design rules in national standards for modes of flexural members strengthened with FRP.
predicting the concrete component of the shear capacity of Proceedings of 2nd International Conference on Composites
prestressed beams can be used to predict CDC debonding. in Infrastructures, Tucson, 1998, 315–327.

Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al. 83
11. BLASCHKO M. Strengthening with CFRP. Munchner 16. BAY B., NGUYEN T., WEBSTER A. and WILKINS S. Vertical
Massivbau Seminar, TU Munchen, 1997. Shear Strength of FRP Plated Reinforced Concrete
12. JANSZE W. Strengthening of Reinforced Concrete Members Beams. Department of Civil and Environmental
in Bending by Externally Bonded Steel Plates. PhD Engineering, Adelaide University, 2003, 4th year
dissertation, TU Delft, The Netherlands, 1997. research report.
13. MOHAMED ALI M. S., OEHLERS D. J. and BRADFORD M. A. 17. OEHLERS D. J., LIU I. and SERACINO R. Passive prestress
Shear peeling of steel plates adhesively bonded to the sides approach for CDC debonding of adhesively bonded steel
of reinforced concrete beams. Proceedings of the Institution and FRP plates. Magazine of Concrete Research, 2004, 56,
of Civil Engineers, Structures and Buildings, 2000, 140, No. 8, 475–486.
No. 3, 249–259. 18. ACI COMMITTEE 318. Building Code Requirements for
14. NIERDERMEIER R. Zugkraftdeckung bei klebearmierten Reinforced Concrete/Commentary. American Concrete
bauteilen (Envelope line of tensile forces while using Institute, Farmington Hills, MI, 1989, ACI 318–89/ACI
externally bonded reinforcement). Doctoral dissertation, TU 318R-89.
Munchen, 2000. 19. CHEN J. F. and TENG J. G. Anchorage strength models for
15. ZHANG J. P. Diagonal cracking and shear strength of FRP and steel plates bonded to concrete. Journal of
reinforced concrete beams. Magazine of Concrete Research, Structural Engineering, ASCE, 2001, 127, No. 7,
1997, 49, No. 178, 55–65. 784–791.

Please email, fax or post your discussion contributions to the secretary by 1 August 2005: email: journals@ice.org.uk;
fax: þ44 (0)20 665 2294; or post to Journals Department, Institution of Civil Engineers, 1–7 Great George Street, London SW1P
3AA.

84 Structures & Buildings 158 Issue SB1 Debonding of adhesively bonded plates Oehlers et al.

Вам также может понравиться