Construction and Building Materials 158 (2018) 670–682

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Theoretical modelling and acoustic emission monitoring of RC beams

strengthened with UHPC
Prabhat Ranjan Prem ⇑, A. Ramachandra Murthy, Mohit Verma
Council of Scientific and Industrial Research – Structural Engineering Research Centre, Chennai 600113, India

h i g h l i g h t s

Non-linear fracture mechanics model to predict moment capacity.

Strength based model to predict load deflection response of strengthened beams.

Classification of acoustic emission parameters to study the progress of crack.

Effect of reinforcement ratio on AE parameters in RC beams strengthened with UHPC.

a r t i c l e i n f o a b s t r a c t

Article history: In this study, theoretical modelling and acoustic emission (AE) studies are carried out on damaged rein-
Received 7 July 2017 forced concrete beams strengthened with ultra high performance concrete (UHPC). The contributions from
Received in revised form 10 October 2017 the paper include (i) development of an integrated nonlinear fracture mechanics model to predict the
Accepted 12 October 2017
moment carrying capacity of composite beam considering (a) stress intensity factors due to cement paste,
Available online 17 October 2017
steel, UHPC and applied load (b) vectorial contributions of crack displacement from external moment, steel
and UHPC force (ii) prediction of load deflection response of strengthened beams by strength based model
Keywords:
which applies concepts of force equilibrium and strain compatibility method (iii) analysis of the effect of
UHPC
Fracture mechanics
reinforcement ratio on AE parameters from parametric analysis. The presented results can be effectively
Strength model utilized for designing UHPC thickness and health monitoring of structures strengthened with UHPC.
Acoustic emission Ó 2017 Elsevier Ltd. All rights reserved.
Strengthening

1. Introduction the composite members by UHPC. Recent, published research

The growing demand to meet stringent requirements of sus-
tainability, serviceability and durability characteristics, research
lead towards development ultra high performance concrete
(UHPC). From literature, it is found that UHPC have a very dense
cementitious matrix with inclusion of micro steel fibers leading
to, compressive strength generally greater than 150 MPa, ductility,
deflection hardening and deflection softening properties, tough-
ness, energy absorption capacity and blast resistant characteristics
[1,2]. The excellent material properties of UHPC can be utilized to
reduce section enlargement while doing strengthening using jack-
eting methods or external plate ponding. UHPC as a retrofitting
material have been found to overcome the drawbacks found in
fiber reinforced polymer such as delamination and brittle failure.
The studies presented by Tayeh et al. [3], have shown promising
results in rehabilitation of crash barrier wall, bridge piers,
beam-column joints, enhancing shear and flexural resistance of
⇑ Corresponding author.
E-mail address: prabhat@serc.res.in (P.R. Prem).
showed highlighted superior performance of strengthened beams
in terms of flexural capacity and torsional performance with UHPC
jacketing [4,5]. From literature review, it is observed that the crit-
ical parameters considered for strengthening flexural damaged
structures include (i) existing reinforcement ratio in the reinforced
concrete (RC) beams (ii) damage state of substrate member (iii)
strengthening thickness of UHPC for jacketing or overlaying (iv)
health monitoring of strengthened members with UHPC. The pre-
sent paper forms the continuation of earlier work carried out by
the authors on development of mix design [6,7], methodology of
curing [8], determination of fracture [9], impact properties
[10,11] and repairing flexural members [12–14]. In the present
paper, theoretical model is presented for predicting the moment
capacity of strengthened beams by using non linear fracture
mechanics approach and strength based theory. The second part
of the paper examines the effect on AE parameters during
strengthening when the reinforcement ratio of beams are different
(0.57% – A, 0.89% – B and 1.30% – C) and strengthened with an
overlay of 15 mm UHPC thickness.

https://doi.org/10.1016/j.conbuildmat.2017.10.063
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682
Table 1
Mix design of UHPC.
Cement Silica Quartz
fume powder
kg/m3 kg/m3 kg/m3
787 196 314
Fig. 1. Load versus crack mouth opening displacement.
2. Research significance
In the present paper, theoretical studies are conduct towards
development of non linear fracture mechanics model and strength
model. The fracture model accounts for bi-linear tension softening
parameters to predict the moment capacity of strengthened beams
while strength model evaluates the flexural behaviour by consider-
ing force equilibrium and strain compatibility. In other studies, para-
metric analysis is done to study the effect of reinforcement ratio in
substrate beams after strengthening with UHPC. The values of
acoustic emission parameters are classified in five different damages
zones to study failure mechanisms of composite beams. From over-
all studies an analytical and parametric study is presented to study
the behaviour of strengthened reinforced beams with UHPC.
3. Material properties
3.1. Ultra high performance concrete
The mix design of UHPC is presented in Table 1 [8]. The ductility
of UHPC is improved with straight micro steel fibers having aspect
ratio of 81 and 38. The fibers are of 6 (F1) and 13 mm (F2) length
Fig. 2. Cross section of control beams: (a) A (b) B (c) C.
671
Fine Water Superplasticizer
aggregates
kg/m3 l/m3 l/m3
867 173 14.63
(l), each with diameter (d) of 0.16 mm. Five batches of UHPC are
cast, of which details are - U1 with 0% fiber, U2 with 2% F1, U3 with
2.5% F1, U4 with 2% F2 and U5 with 2.5% volume fraction (V f ) F2.
The split tensile strength of U1–U5 mix is found to 11.8, 18, 20.2,
22.6 and 23.8 MPa. In order to study the fracture response of UHPC,
notched prisms (100  100  350 mm) is measured under three-
point loading as per RILEM recommendations [15]. The fracture
response of UHPC prisms is shown in Fig. 1. From the obtained
results Mix U4 is chosen to cast UHPC overlay.
3.2. Reinforced concrete beams
A set of virgin RC beams (A, B and C) of size (100  200  1500
mm) are tested under four point bending to determine average
flexural capacity. The average ultimate load ðPu Þ of control beams
is determined from Fig. 4a. The properties of beam, instrumenta-
tion and test set up is given in Fig. 2 and Fig. 3, respectively. After
testing of control beams, virgin RC beams of each type are cast
again and preloaded till 90% of ultimate load to induce damage.
The damaged beams of each category are now strengthened with
precast UHPC overlays of 10 mm (A1, B1 & C1), 15 mm (A2, B2 &
C2) and 20 mm (A3, B3 & C3) thickness in tension region by epoxy
and tested till failure. The flexural behaviour of strengthened
beams under bending is given in Fig. 4b–d.
4. Analytical modelling
4.1. Inverse analysis of UHPC mixes
The load versus crack mouth opening displacement of UHPC
prisms given in Fig. 1 are used to evaluate stress and crack dis-
placement ðr  wÞ relationship from inverse analysis through
crack hinge model. The obtained parameters are used for mod-
elling UHPC as a strengthening member for damaged RC beams.
The r  w relationship for UHPC is derived from the model devel-
oped by [16,17]) for fiber reinforced concretes. In the model the
tensile stress before crack initiation is modelled elastically while
the cracked state is done by crack hinge model given by Eq. (1)
and (2) and further simplifications.


Ee Pre-CrackedState
r¼ ð1Þ
gðwÞf t Cracked
672 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682

Fig. 3. Experimental test set up and instrumentation.

(a) (b)

(c) (d)
Fig. 4. Flexural response of (a) A, B and C (b) A, A1–A3 (c) B, B1–B3 (d) C, C1–C3 [14].

8
< b1  a1 0 6 w 6 w1 w1 ¼ a1b 2 authors in [18] and given by Eqs. (3)–(5). The additional results
1 a2
gðwÞ ¼ ð2Þ in Fig. 5 and Table 2 shows the tensile behaviour of UHPC caused
: b2  a2 w1 6 w 6 w2 w2 ¼ ba22 by change in fiber volume.
The obtained load versus crack mouth opening displacement of
UHPC is taken as input and a least square fitting data is obtained 1 NX
max

E ¼ min ðP  Pexp Þ2 ð3Þ

from the crack hinge model to derive the bilinear parameters. E Nmax 0
The details of inverse analysis can be found in earlier work of
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682 673

Table 4
Material Properties of UHPC in Tension.

UHPC

ft E Gf w1 emax
(MPa) (GPa) (N/mm) (mm) (–)
18 43 14.323 0.42 7.6  104

sffiffiffiffiffiffiffiffi 
Ef 1 w w2
rðwÞ ¼ 2hLV f agK IC   2 ð6Þ
dEb 8 2l 2l

Z w¼2L
GF ¼ rðwÞdw ð7Þ
Fig. 5. Load-cmod curves generated by analytical model and experiment. w¼0

In the equation h; a; g; Ef ; Eb and K IC denote snubbing factor, snub-

1 NX
max

f t and a1 ¼ min ðP  Pexp Þ2
f t ;a1 Nmax 0 reinforced un-cracked matrix and fracture toughness of hardened
1 NX
max  2
a2 and b2 ¼ min P  Pexp
a2 ;b2 Nmax
0
4.2. Strength model
In the present model, compressive stresses are contributed by
the compressive forces developed in compressive region of
concrete. The tensile stresses are attributed by tensile forces devel-
oped from steel and UHPC overlay. The material properties of plain
concrete and steel considered for analytical modelling are given in
Table 3 and adopted as per Fib Model code 2010. In Table 3
f cm ; Ec ; Ec1 ; ec1 ; ecu ; f ctm ; GF ; wc , are compressive strength, tangent
modulus, secant modulus, peak strain, ultimate strain, direct
tensile strength, specific fracture energy and crack opening at
termination of tension softening, for concrete, respectively. f y
and Es are the properties of steel. The characteristics of UHPC are
given in Table 4.
The stress strain profile of plain concrete is assumed parabolic
and for steel perfectly elastic plastic. The properties of UHPC is
obtained from tests under compression, tension test and inverse
analysis. The related expression for finding fracture energy ðGF Þ
after formation of through crack is given by Eqs. (6) and (7)[19].
ð4Þ bing friction coefficient, elastic modulus of micro fiber & un-
cement paste, respectively. The load deflection response is evalu-
ated by strain compatibility method assuming Bernoulli hypothesis.
ð5Þ
During flexural testing there was no delamination between the
UHPC overlay and parent member, hence a perfect bond between
the components is assumed. The stress distribution for UHPC over-
lays is obtained by evaluating crack opening (w) from the product of
strain at bottom tension fiber of UHPC overlay ðeUHPC Þ and effective
length of overlay (Leff ). Leff for modelling is considered as double the
distance between tension reinforcement and UHPC [20].
To predict the load versus deflection behaviour of tested beams,
a program is written in MATLAB. Fig. 6 illustrates the flowchart of
the model. For developing the model, the section of the beam is
divided into several segments and strain distribution is assumed
linear. In each segment after finding average strain, corresponding
stresses and forces are evaluated. After evaluating the forces in
compression and tension, neutral axis is balanced to obtain hori-
zontal force equilibrium. After attaining this equilibrium load ver-
sus deflection are plot. Fig. 7, shows the results from developed
model. A local drop observed in the graphs after attainment of
maximum load is due to the yielding of steel. The model slightly
overestimates the experimental results which may be caused by
the overvaluation of yield stress of steel. The predicted flexural
response of the beams shows satisfactory coherence with the test
results.

Table 2
Bilinear Parameters of UHPC.

Peak load Peak load Tensile strength Tensile strength w1

Exp (N) Model(N) Model (N=mm2 ) Model (mm)

U1 4034 4138 11.8 9.74 0.045

U2 10295 9707 18 16.21 0.29
U3 12496 12228 20.2 18.49 0.33
U4 15715 14813 22.6 19.32 0.42
U5 19104 18162 23.8 21.39 0.46

Table 3
Properties of Concrete and Steel.

Concrete Steel
Compression Tension

f cm Ec Ec1 ec1 ecu f ctm GF w1 wc fy Es

(MPa) (GPa) (GPa) (MPa) (N/mm) (mm) (mm) (MPa) (GPa)
35 32.5 19.8 0.0022 0.00315 3.8 0.0625 0.016 0.123 415 205
674 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682

Input dimensions of beam, and material Parameters of plain concrete, steel, UHPC and
thickness of overlay

Assume compressive strain in extreme fiber: εc (i)

Assume neutral axis depth: xu (i) Goto : εc (i+1)

Goto : xu (i+1)

From Bernoulli’s hypothesis evaluate corresponding strain in

Tension Steel UHPC overlay Concrete in Compression

Stresses and Force in Stresses and Forces in

Stresses and Forces in
Reinforcement (Fst) Compression Concrete (Fc)
UHPC (FUHPC)

N
Check force equilibrium
Fc=Fst+ FUHPC

Y
Evaluate Moment capacity of section (MI)

Y
N
Check Moment Capacity εc (i) ≤

If (MI >MI+1) εcu

Evaluate loads and Deflection till MI

Plot (Load, Deflection)

Fig. 6. Flowchart of Strength based model.

4.3. Nonlinear fracture mechanics model
The model presented in Fig. 8a accounts for the Mode I crack
propagation in RC beam, by assuming the behaviour of steel as
elastic-perfectly plastic and no growth of crack while steel
remained elastic [21]. The tension force produced from steel (F s )
is found to be remotely applied from the crack plane. It also consid-
ers no slippage between steel and concrete, however in reality it
does slip.
Later, this drawback is over come by Bosco and Carpinteri
model [22,23], in which crack is allowed to grow before yielding
of steel. The undetermined reinforcement reactions are evaluated
by applying closing forces on crack face instead of infinity. In this
research the mechanism for the proposed non linear fracture
mechanics (NLFM) based model is explained in Fig. 8b. The resist-
ing force’s against bending moment (M) include tension forces
from steel and UHPC.
4.3.1. Bridging forces due to steel
For evaluating F s , several experimental [24], theoretical [25]
and empirical [26] observations are available in literature. How-
ever due to similarity between the tested specimens, the relation
developed by [27] expressed by Eq. (8), is adopted for the present
study. It is an empirical relation to predict crack opening in tension
face of flexural members with high strength deformed bars. The
equation is used in the proposed analytical model to evaluate the
approximate relation between steel force and the crack opening
displacement.
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682 675

8
< AF s As ¼ Steel Area F UHPC ¼ rU ðwU ÞbtU ð18Þ
s
fs ¼ 
: 11876:5ws
A ¼ Area of Concrete surrounding each bar
1
A4 wU ¼ tU
wð1  2D aÞ ð19Þ
ð8Þ

rU ðwU ÞbtU tU
K IUHPC ðaÞ ¼ Y s a; ð20Þ
D1=2 2D
4.3.2. Present model
From the Fig. 8c, it is understood clearly that closure pressures From the crack opening compatibility conditions, crack width at
are exerted by steel and UHPC. The composite beam can be simi- the level of steel reinforcement (ws ) is obtained from Eqs. (9)–(12)
larly assumed here to be under n concentrated forces [28]. The vectorial contributions from crack opening is expressed
F i ; i ¼ 1; . . . n. Based on recent formulations developed by, Carpin- by Eq. (21).
teri et al. [28] generic point load translation’s di is given by Eq. ws ¼ ðws ÞM  ðws ÞS  ðws ÞU ð21Þ
(9), where kij are the local compliance’s.
X
n ðws ÞM and ðws ÞS are the crack opening due to external moment
di ¼ kij F j ð9Þ and crack closing values due to steel, respectively. This are com-
j¼1 puted in terms of there respective compliance coefficients using
Further with the growth of crack length (da), the increment in local Eq. (12), and given from Eqs. (22)–(26). kSM M and kSS F s represent
compliance’s (dkij ) with respect to the variation of total potential the compliance coefficients in which Y M ðxÞ and Y s ðx; cÞ are the
energy by Clapeyron theorem and (K Ii ) at the tip of the crack by geometry factors. The stresses obtained after minimum variation
Irwin relationship is obtained from Eqs. (10) and (11). kij due to is given in Eq. (25). Ae is the effective area of concrete and n is total
presence of crack length a is given by Eq. (12). number of tension bars.

X
n X
n
1 ðws ÞM ¼ kSM M ð22Þ
dW ¼  F i F j dkij ð10Þ
2 Z a
i¼1 j¼1 2
kSM ¼ Y M ðxÞY s ðx; cÞdx ð23Þ
DbE c
n X
X n
K Ii K Ij
dW ¼  tda ð11Þ
i¼1 j¼1
E ðws ÞS ¼ kSS F s ð24Þ

Z a 3:76  11876:5
2 K Ii K Ij fs ¼ ð25Þ
kij ¼ tda ð12Þ 1
E 0 FiFj Ae 4
n

From the studies reported by Zhang and Li [29], it is known that Z a

the net stress intensity factor (K I Þ is zero. However the stresses can 2
kSS ¼ ðY s Þ2 ðx; cÞdx ð26Þ
be only finite when the structural matrix is in elastic state with no bE c
cracking. The micro cracking in cement paste drives non linearity
The compliance coefficient and closing crack width due to unit
in pre- peak response and establishes large fracture process zone
force in the UHPC is obtained in Eqs. (27) and (28). The value of
[30]. Hence the K I , due to moment (K IM ), steel (K IS ) and UHPC
F UHPC is substituted from Eq. (18).
(K IUHPC ) are equal the fracture toughness of cement paste (K IC ).
The related expression is ðws ÞU ¼ kSU F UHPC ð27Þ
K IC ¼ K IM  K IS  K IUHPC ð13Þ Z a

2 tU
kSU ¼ Y S ðx; cÞY s x; dx ð28Þ
The expression for K IM and K IS is given by Eqs. (14)–(17), where a is bE c 2D
a
and c is Dc [31,32]
D Z a

2rU ðwU Þt U tU
M ðws ÞU ¼ Y S ðx; cÞY s x; dx ð29Þ
K IM ¼ Y M ðaÞ ð14Þ E c 2D
D3=2 b

6a1=2 ð1:99 þ 0:83a  0:31a2 þ 0:41a3 Þ 4.3.3. Results and discussions

Y M ðaÞ ¼ ð15Þ
ð1  a3=2 Þð1 þ 3aÞ The sequential steps to predict the moment carrying capacity are
presented in the form of flowchart and shown in Fig. 9. The model as
Fs described in the flowchart is programmed in MATLAB to get output.
K IS ¼ 1=2
Y s ða; cÞ ð16Þ
D b The crack length(a) in the strengthened beam is stated in terms cor-
"" # responding crack opening (w) at the mouth. For potential crack
 
2 3:52 1  ac 4:35  5:28 ac depth and crack width corresponding crack closure forces, crack
Y s ða; cÞ ¼ pffiffiffiffiffiffiffiffi  opening forces, crack closing width and crack opening width due
Pa ð1  aÞ3=2 ð1  aÞ1=2
2 33 to applied loads are obtained. On applying smooth closure condition
of crack faces from Eq. (13) and crack compatibility conditions from
61:30  0:30c3=2
 c  c 7 7
6 77 Eq. (21) intersection values of crack length, crack width and moment
þ6
 2 1=2 þ 0:83  1:76 a  1  1  a aÞ7
a
7
4 55 are identified. From the set of intersection values the value of max-
1  ac imum moment is selected. The experimental and theoretical model
results are given in Table 5 and Fig. 10. It is observed that the model
ð17Þ
according to Kaar formula does not give accurate values of moment
The contribution of the UHPC overlay can be similarly evaluated capacity. This values are observed to be underestimated than the
by determining the crack opening (wU ) at the position of F UHPC from test results. The reason could be due to empiricism in Kaar model
Eq. (18), where wU can be obtained from Eq. (19). K IUHPC given in Eq. and presumption that the tension stress over the effective area of
(20) is obtained by substituting Eq. (18) in Eq. (16). concrete is uniform. However this may not be true as the tensile
676 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682

(a) (b)

(c) (d)
⁄
Fig. 7. Flexure behaviour (a) A, B and C (b) A1 and A3 (c) B1 and B3 (d) C1 and C3; E: Experiment; M: Present Model

Fig. 8. (a) Fracture model by Carpinteri’s based on LEFM principles [21], (b) Present Model.

stress distribution should be non uniform. The present model satis-
factory predicts the experimental observations.
5. Acoustic monitoring
Acoustic emission (AE), testing is an non destructive method for
health monitoring of structures. This method helps to evaluate the
properties of the structure by capturing the dissipated energy from
the source due to crack formation, without intruding the structure.
In the present study, AE testing and data capturing is done by
Physical Acoustic Corporation (PAC) system and AE win software,
respectively. The location and mounting of sensors on the speci-
mens are given in Fig. 3. The accuracy of the reproducibility of
acoustic emission sensor in the acoustic system is done by
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682 677

Input dimensions of beam, and material Parameters of plain concrete, steel, UHPC and
thickness of overlay

Assume crack length ‘a’ after strengthening

Assume crack width ‘w’

Crack Closure Forces: (i) KIS- Tension Steel Crack closing width: (i) (WS)S – Tension Steel
(ii) KIUHPC- UHPC Overlay (ii) (WS)U - UHPC Overlay
Crack Opening Forces: (i) KIM- External Moment Crack opening width: (i) (WS)M - External Moment

Evaluate ax, wx, Mx Evaluate ay, wy , My

N Check IF
Wx=Wy, Mx=My

N
Check IF
ax = ay, Mx=My

Print ax = ay, Wx=Wy & Mx=My= Moment capacity of strengthened beam

Fig. 9. Fracture Mechanics Model.

Table 5
Crack Opening (w), Crack length ðaÞ and Moment Capacity ðM F Þ.

Kaar model Present model

A2 w (mm) 0.36 0.14
a (mm) 132 118
MF (kNm) 11.96 13.57
w (mm) 0.46 0.19
B2 a (mm) 117 102
MF (kNm) 16.67 19.9
w (mm) 0.57 0.24
C2 a (mm) 97 83
MF (kNm) 19.45 26.23

Hsu-Nielsen source method [33]. Resonant type differential AE

sensors with low frequency ranges of 35–100 kHz are mounted
on the beams with viscous coupling agent. A 45 dB threshold level
is set during the testing to eliminate any mechanical and electrical
noise from surrounding area. The measured parameters from AE
testing are shown in Fig. 11a. The time taken for the AE signal to
cross the threshold value from the beginning and end of the signal Fig. 10. Comparison of Moment Resistance of A2, B2 and C2 beams.
678 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682

Fig. 11. (a) Acoustic signal parameters (b) Crack classification.

80
Ultimate Zone I ratio ðlÞ of the damaged beams are changing, but strengthened
load Zone II with same optimum thickness (here 15 mm) of UHPC overlay.
70 Zone III For the same, Type II beams that is A2 ðl1 Þ, B2 ðl2 ) and C2 ðl3 Þ
Crack
Zone IV
formation in
Zone V
beams are compared at different damage levels and discussed in
60 retrofit strip
the following section.
Concrete
Load (kN)

50 5.1. Results and discussions

crushing

40 Visible cracks
(i) From the present study, transitions in the RA and AF value
is studied, to identify the fracture mechanisms occurring
30 Hairline cracks in the damaged reinforce concrete beams externally
bonded with precast UHPC overlay. During experimental
20 testing, it is observed that the failure of the composite
beams progresses through five different stages. These
10 stages are, development of hair line and visible crack in
RC beam, followed by crack progress in UHPC and then
0 on subsequent loading compression failure due to steel
0 5 10 15 20
yielding. These stages are classified into five zones and
Deflection (mm)
observed crack patterns are shown in Figs. 12 and 13,
Fig. 12. Typical damage classification of strengthened beam. respectively. However, it is noted that there is no delami-
nation between the parent damaged beam and UHPC over-
lay till final failure. A typical failure pattern of
is called duration. The duration between the start and peak ampli- strengthened beam is given in Fig. 14. The main advantage
tude of the signal is known as rise time. AE Count is the total fre- of zone classification is, identification of the behaviour of
quency of AE signal amplitude crossing the threshold value while AE parameters with the progress of damage in structures.
AE energy is the integration of the amplitude envelope. Hit is During health monitoring of concrete members this
defined as any signal detection measured during testing. The changes will indicate the extent of damage and accord-
health of the structure can be evaluated by conducting parametric ingly necessary decisions can be taken for maintenance.
analysis between moving average of the rise angle value ðRAÞ and (ii) The values of RA and AF in terms of damage and cracking are
average frequency ðAFÞ and can be evaluated by Eqs. (30) and plot in different zones and presented in Figs. 15 and 16.
(31) [33]. Fig. 15 shows the parametric crack classification of l1 ; l2
and l3 beams for all zone. The changes in the AF and RA
Rise Time
RA ¼ ð30Þ value with the progress of damage is also shown for l1 beam
Maximum Amplitude
in Fig. 15. Fig. 16 shows the comparative plot of l1 ; l2 and
l3 beams superimposed in each zone separately. A typical
Counts comparison of AF and RA value for zone IV is also shown
AF ¼ ð31Þ
Duration for l1 ; l2 and l3 beams in Fig. 16. The measured parameters
After evaluating RA and AF, cracks are separated into tensile and are given in Table 6.
shear cracks as shown in Fig. 11b [34] as per the codal recommen- (iii) From Fig. 15, it can be observed that the AE activity is very
dations [35–37]. limited in the first, second and third zone. The cumulative
From the earlier reported results, 15 mm UHPC overlay is found energy till the formation of hair line cracks for l1 ; l2 and
as optimum thickness for strengthening the damaged beams l3 beams at load level of 22, 27 and 34 kN is found to be
[14,38]. Based on this observations authors have examined and 5895, 6828 and 13222 aJ. At this juncture a sharp change
described the behaviour of AE parameters, when reinforcement in stiffness is seen in the load deflection plot. Once crack
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682
Fig. 13. (a) Zone II (b) Zone III (c) Zone IV (d) Zone V.
Fig. 14. Typical failure of strengthened beams.
progress in the UHPC overlay at the load level of 47, 71 and
82 kN for l1 ; l2 and l3 beams, there is sudden jump in AE
activity. This can be observed by the comparison of cumula-
tive hit percentage in Zone I–Zone III versus Zone IV alone.
The total activity percentage from Zone I–Zone III for
l1 ; l2 and l3 beams is 6.59, 9.13, 12.24% which changes
abruptly to 60.74, 68.41 and 18.35%, respectively.
(iv) Maximum shear cracks can be observed in the fourth and
fifth zones. This is caused because once the crack penetrates
the UHPC overlay, fiber pull out contributes heavily as an
active fracture mechanism and trigger the shear mode. This
shear activity is also contributed from steel fiber debonding
in the UHPC matrix and rise of frictional forces during fiber
pull out. This phenomenon causes reduction in average fre-
quency in Zone IV–V compared to Zone I–III as shown in
Table 6. A similar observation, is also found by Reinhardt
et al. [39], where a decrease in AF is noticed during fiber pull
out tests. Multiple cracking observed in the UHPC overlay
increases the fracture energy of the composite beams as a
whole. This increased activity was found to be highest in
fourth zone of l1 & l2 beams. In the l3 beams there is sig-
nificant enhancement in AE events only when maximum
load is reached and macro crack is formed causing shear
mode activity. The same explanation is also associated with
the detection of hits which shows micro-cracking in first
three zones for l1  l2 . A substantial increase in the hits
observed in fourth zones for l1 & l2 beams and in fifth zone
for l3 shows macro-cracking [40,41].
679
(v) The flexural behaviour of the strengthened beams can also
be related in terms of AF and RA value, which shows that
once the load carrying capacity of the tested beams drops,
the value of AF drops considerably compared to initial stage
of loading. For all the beams, AF is found to be highest in the
first zone. This values are observed to be highest for lower
reinforcement ratio beams and lie in sequence of
l1 > l2 > l3 and obtained in domain of 117.56–54.62,
110.50–43.92 and 101.42–67.56 kHz respectively. This
change in AF indicate’s potential fracture in the beams
caused due to macro crack, steel yielding, and steel debond-
ing in concrete. In literature [42], it is mentioned that the
drops indicate delamination of external reinforcement. In
present study, however there is no delamination of UHPC
overlay with the damage beam.
(vi) On examining rise time, amplitude, count, duration and hit it
is observed that these values, increases linearly up to fourth
zone and then decreases in the final zone for l1 & l2 beams.
But this raised continuously for l3 beams. From the result of
mechanical testing, it was observed that the improvement in
flexural capacity is of sequence l1 > l2 > l3 , when com-
pared to control beams after strengthening. This indicates
that when cracking started in UHPC for l1 & l2 beams, it
was more effective compared to l3 , in retarding the damage
and causing delayed yielding and debonding of tension rein-
forcement. For l3 beams, the shear activity is restrained by
the increase in tensile reinforcement of the beam [43]. This
activity is further reduced due to the presence of UHPC
680 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682

Fig. 15. AF vs. RA value for (a) A2, B2 and C2 beams superimposed in Zone I–V (b) A2 beam in Zone I–Zone V.

Fig. 16. AF vs. RA value plot of (a) beams A2, B2 and C2 in separate Zones I–Zone V (b) A2, B2 C2 beam in Zone IV.
 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682 681

Table 6
AE parameters of strengthened beams.

Rise Duration AMP AF CE Count % Hits

(s) (s) (V) (kHz) (aJ)
A2 Zone I 135.90 479.73 50.96 117.56 5895 5601 1.48
Zone II 256.09 711.76 50.01 103.87 9415 8038 2.88
Zone III 281.93 737.45 50.39 93.74 7124 15784 2.23
Zone IV 622.97 2974.96 56.50 54.62 1354446 160439 60.74
Zone V 251.35 1232.78 50.87 87.97 136621 24293 32.67
B2 Zone I 114.12 558.12 58.29 110.50 6828 4156 1.60
Zone II 228.76 989.29 60.69 105.95 14544 2175 4.29
Zone III 238.86 1074.27 60.21 77.39 21433 4040 3.24
Zone IV 741.08 3038.84 64.09 43.92 1625723 222049 68.41
Zone V 433.23 1614.12 59.97 72.74 161925 29660 22.46
C2 Zone I 132.55 894.53 51.12 101.42 13222 8460 2.36
Zone II 256.93 967.11 50.43 95.63 28538 496 0.21
Zone III 288.31 1062.64 50.90 87.32 65400 34751 9.67
Zone IV 370.20 1550.10 51.91 79.32 90902 35961 18.35
Zone V 532.87 2850.26 53.88 67.56 2029041 536576 69.40

Rise, Duration, Amplitude (AMP), AF and Count represent average data Acoustic Energy (CE) is cumulative data.

overlay. UHPC overlay increased the flexural depth of the
section and decreased the tension forces acting on the ten-
sion reinforcement due to flexural loading. This led to
improved ductile failure with extensive yielding [44]. The
cumulative energy released till failure of beams increased
with higher reinforcement ratio and is 1513501, 1830453
and 2029041 for l1 , l2 and l3 beams respectively.
(vii) From all the plots opposite relationship between AF and
RA is observed. The measure of AF in the tensile mode is
greater than shear mode which is due to the shorter
waveform shape, lower rise time and vice versa in shear
mode [45]. The tensile mode of cracking is prevalent in
plain concrete which changes to shear due to presence
of steel fiber in concrete [46]. In the present case, this is
evident due to presence of UHPC overlay, that is there is
delay in the overall cracking. From the Table 6, this can
be also observed by increase in duration of Zone IV com-
pared to Zone I–III. Visually this noticed by crack splitting
and interface cracks at the level of steel, as shown in
Fig. 12. From the categorization of the cracks, it is evident
that in the first three zones, the formed cracks are tensile
crack and in the fourth zone and fifth zone maximum
shear cracks are formed. A shear crack or mode II fracture
represents a in-plane shear dislocation and tensile crack or
mode I fracture is similar to opening dislocation.
6. Conclusions
1. The moment capacity and load versus deflection response of
strengthened beams is predicted by strength based model and
non linear fracture mechanics approach. The results of the the-
oretical model are in good coherence with the experimental
observations.
2. In strength based model the flexural behaviour is evaluated by
strain compatibility method assuming Bernoulli hypothesis and
no slipping of bond between the damaged beam and overlay.
The compressive and tensile forces are balanced by taking into
account the contributions of concrete, tension reinforcement
and UHPC.
3. In fracture mechanics model stress versus crack width relation
of UHPC, stress intensity factors due to moment, steel, UHPC
and cement paste are considered. On applying smooth closure
condition of crack faces and crack compatibility conditions
intersection values of crack length, crack width and moment
are identified to find out the moment resistance of the compos-
ite beams.
4. From acoustic emission studies, the behaviour of UHPC is
evaluated with three reinforcement ratios of 0.57%, 0.89% and
1.30%. Parametric analysis is conducted between the average
frequency and rise angle at five different damage zones.
5. The analytical models presented from the paper can be utilized
in designing the thickness of UHPC overlay for strengthening
damaged beams and the results of acoustic emission can be
effectively used for health monitoring of retrofitted concrete
members.
Acknowledgements
The current study is funded from the award of UGC-UKIERI
between CSIR-SERC and Cardiff University, UK. Authors would like
to acknowledge discussions with Prof B.L.Karihaloo Cardiff Univer-
sity, UK and their colleagues Shri Nawal Kishor Banjara and Prof.B.
H.Bharatkumar for suggestions regarding the present work.
References
[1] K.V. Harish, J.K. Dattatreya, M. Neelamegam, Experimental investigation and
analytical modeling of the r–e characteristics in compression of heat-treated
ultra-high strength mortars produced from conventional materials, Constr.
Build. Mater. 49 (2013) 781–796.
[2] A.Q. Sobia, M.S. Hamidah, I. Azmi, S.F. Rafeeqi, Elevated temperature resistance
of ultra-high-performance fibre-reinforced cementitious composites, Mag.
Concr. Res. 67 (17) (2015) 923–937.
[3] B.A. Tayeh, B.A. Bakar, M.M. Johari, Y.L. Voo, Utilization of ultra-high
performance fibre concrete (uhpfc) for rehabilitation – a review, Proc. Eng.
54 (2013) 525–538.
[4] M. Al-Osta, M. Isa, M. Baluch, M. Rahman, Flexural behavior of reinforced
concrete beams strengthened with ultra-high performance fiber reinforced
concrete, Constr. Build. Mater. 134 (2017) 279–296.
[5] T.J. Mohammed, B.A. Bakar, N.M. Bunnori, Torsional improvement of reinforced
concrete beams using ultra high-performance fiber reinforced concrete (uhpfc)
jackets – experimental study, Constr. Build. Mater. 106 (2016) 533–542.
[6] P.R. Prem, B. Bharatkumar, N.R. Iyer, Influence of curing regimes on
compressive strength of ultra high performance concrete, Sadhana 38 (6)
(2013) 1421–1431.
[7] P. Ambily, C. Umarani, K. Ravisankar, P.R. Prem, B. Bharatkumar, N.R. Iyer,
Studies on ultra high performance concrete incorporating copper slag as fine
aggregate, Constr. Build. Mater. 77 (2015) 233–240.
[8] P.R. Prem, A.R. Murthy, B.H. Bharatkumar, Influence of curing regime and steel
fibres on the mechanical properties of uhpc, Mag. Concr. Res. 67 (18) (2015)
988–1002.
[9] A.R. Murthy, B.L. Karihaloo, N.R. Iyer, B.R. Prasad, Determination of size-
independent specific fracture energy of concrete mixes by two methods, Cem.
Concr. Res. 50 (2013) 19–25.
[10] M. Verma, P.R. Prem, J. Rajasankar, B. Bharatkumar, On low-energy impact
response of ultra-high performance concrete (uhpc) panels, Mater. Des. 92
(2016) 853–865.
682 P.R. Prem et al. / Construction and Building Materials 158 (2018) 670–682
[11] P.R. Prem, M. Verma, A.R. Murthy, J. Rajasankar, B. Bharatkumar, Numerical
and theoretical modelling of low velocity impact on uhpc panels, Struct. Eng.
Mech. 63 (2) (2017) 207–215.
[12] A.R. Murthy, V.R. Kumar, S. Gopinath, P.R. Prem, N.R. Iyer, R. Balakrishnan,
Structural performance of precast and cast-in-situ ultra high strength concrete
sandwich panel, Int. J. Comput. Mater. Continua 44 (1) (2014) 59–72.
[13] P.R. Prem, A.R. Murthy, G. Ramesh, B. Bharatkumar, N.R. Iyer, Flexural
behaviour of damaged rc beams strengthened with ultra high performance
concrete, Advances in Structural Engineering, Springer, 2015, pp. 2057–2069.
[14] P.R. Prem, A.R. Murthy, Acoustic emission and flexural behaviour of rc beams
strengthened with uhpc overlay, Constr. Build. Mater. 123 (2016) 481–492.
[15] R. TC162-TDF, Test and design methods for steel fibre reinforced concrete:
bending test, Mater. Struct. 33 (2000) 3–5.
[16] J. Zhang, H. Stang, Applications of stress crack width relationship in predicting
the flexural behavior of fibre-reinforced concrete, Cem. Concr. Res. 28 (3)
(1998) 439–452.
[17] J.F. Olesen, Fictitious crack propagation in fiber-reinforced concrete beams, J.
Eng. Mech. 127 (3) (2001) 272–280.
[18] A.R. Murthy, B.L. Karihaloo, N.R. Iyer, B.R. Prasad, Bilinear tension softening
diagrams of concrete mixes corresponding to their size-independent specific
fracture energy, Constr. Build. Mater. 47 (2013) 1160–1166.
[19] D. Lange-Kornbak, B. Karihaloo, Fracture mechanical prediction of transitional
failure and strength of singly-reinforced beams, Eur. Struct. Integrity Soc. 24
(1999) 31–66.
[20] F.J. Alaee, B.L. Karihaloo, Retrofitting of reinforced concrete beams with
cardifrc, J. Compos. Constr. 7 (3) (2003) 174–186.
[21] A. Carpinteri, Stability of fracturing process in rc beams, J. Struct. Eng. 110 (3)
(1984) 544–558.
[22] C. Bosco, A. Carpinteri, P.G. Debernardi, Minimum reinforcement in high-
strength concrete, J. Struct. Eng. 116 (2) (1990) 427–437.
[23] C. Bosco, A. Carpinteri, Fracture behavior of beam cracked across
reinforcement, Theor. Appl. Fract. Mech. 17 (1) (1992) 61–68.
[24] M. Rashid, M. Mansur, Reinforced high-strength concrete beams in flexure, ACI
Struct. J. 102 (3) (2005) 462–471.
[25] K.H. Tan, M.K. Saha, Cracking characteristics of rc beams strengthened with frp
system, J. Compos. Constr. 12 (5) (2008) 513–521.
[26] D. Zhang, T. Ueda, H. Furuuchi, Average crack spacing of overlay-strengthened
rc beams, J. Mater. Civ. Eng. 23 (10) (2011) 1460–1472.
[27] P.H. Kaar, E. Hognestad, High strength bars as concrete reinforcement, part 7:
control of cracking in T-beam flanges, Portland Cement Association, Research
and Development Laboratories, 1965.
[28] A. Carpinteri, A. Spagnoli, S. Vantadori, A fracture mechanics model for a
composite beam with multiple reinforcements under cyclic bending, Int. J.
Solids Struct. 41 (20) (2004) 5499–5515.
[29] J. Zhang, V.C. Li, Simulation of crack propagation in fiber-reinforced concrete
by fracture mechanics, Cem. Concr. Res. 34 (2) (2004) 333–339.
[30] F.J. Alaee, B.L. Karihaloo, Fracture model for flexural failure of beams
retrofitted with cardifrc, J. Eng. Mech. 129 (9) (2003) 1028–1038.
[31] G. Guinea, J. Pastor, J. Planas, M. Elices, Stress intensity factor, compliance and
cmod for a general three-point-bend beam, Int. J. Fract. 89 (2) (1998) 103–116.
[32] H. Tada, P. Paris, G. Irwin, The Analysis of Cracks Handbook, ASME Press, New
York, 2000. p. 1.
[33] ASTM-E976-10, Standard guide for determining the reproducibility of acoustic
emission sensor response.
[34] K. Ohno, M. Ohtsu, Crack classification in concrete based on acoustic emission,
Constr. Build. Mater. 24 (12) (2010) 2339–2346.
[35] M. Ohtsu, T. Isoda, Y. Tomoda, Acoustic emission techniques standardized for
concrete structures, J. Acoust. Emission 25 (2007) 21–32.
[36] JCMSIII-B5706, Monitoring method for active cracks in concrete by acoustic
emission, Federation of Construction Materials Industries, Japan.
[37] R.T. Committee et al., Recommendation of rilem tc 212-acd: acoustic emission
and related nde techniques for crack detection and damage evaluation in
concrete, Mater. Struct. 43 (9) (2010) 1183–1186.
[38] P.R. Prem, A.R. Murthy, Acoustic emission monitoring of reinforced concrete
beams subjected to four-point-bending, Appl. Acoust. 117 (2017) 28–38.
[39] H. Reinhardt, C. Grosse, B. Weiler, Material characterization of steel fiber
reinforced concrete using neutron ct, ultrasound and quantitative acoustic
emission techniques, NDT. net 6 (5) (2001) 1–12.
[40] D.G. Aggelis, Classification of cracking mode in concrete by acoustic emission
parameters, Mech. Res. Commun. 38 (3) (2011) 153–157.
[41] D. Aggelis, S. Verbruggen, E. Tsangouri, T. Tysmans, D. Van Hemelrijck,
Characterization of mechanical performance of concrete beams with external
reinforcement by acoustic emission and digital image correlation, Constr.
Build. Mater. 47 (2013) 1037–1045.
[42] S. Verbruggen, D.G. Aggelis, T. Tysmans, J. Wastiels, Bending of beams
externally reinforced with trc and cfrp monitored by dic and ae, Compos.
Struct. 112 (2014) 113–121.
[43] S. Shahidan, R. Pulin, N.M. Bunnori, K.M. Holford, Damage classification in
reinforced concrete beam by acoustic emission signal analysis, Constr. Build.
Mater. 45 (2013) 78–86.
[44] Y. Hor, W. Teo, S. Kazutaka, Experimental investigation on the behaviour of
reinforced concrete slabs strengthened with ultra-high performance concrete,
Constr. Build. Mater. 155 (2017) 463–474.
[45] J. Feiteira, E. Tsangouri, E. Gruyaert, C. Lors, G. Louis, N. De Belie, Monitoring
crack movement in polymer-based self-healing concrete through digital image
correlation, acoustic emission analysis and sem in-situ loading, Mater. Des.
115 (2017) 238–246.
[46] D. Soulioti, N. Barkoula, A. Paipetis, T. Matikas, T. Shiotani, D. Aggelis, Acoustic
emission behavior of steel fibre reinforced concrete under bending, Constr.
Build. Mater. 23 (12) (2009) 3532–3536.