Journal of Adhesion Science and Technology

ISSN: 0169-4243 (Print) 1568-5616 (Online) Journal homepage: www.tandfonline.com/journals/tast20

Y-peel characterization of adhesively-bonded

carton board: an objective method

Christer Korin, Magnus Lestelius, Johan Tryding & Nils Hallbäck

To cite this article: Christer Korin, Magnus Lestelius, Johan Tryding & Nils Hallbäck (2007) Y-
peel characterization of adhesively-bonded carton board: an objective method, Journal of
Adhesion Science and Technology, 21:2, 197-210, DOI: 10.1163/156856107780437426

To link to this article: https://doi.org/10.1163/156856107780437426

Published online: 02 Apr 2012.

Submit your article to this journal

Article views: 71

View related articles

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tast20
J. Adhesion Sci. Technol., Vol. 21, No. 2, pp. 197– 210 (2007)
 VSP 2007.
Also available online - www.brill.nl/jast

Y-peel characterization of adhesively-bonded carton

board: an objective method

CHRISTER KORIN 1,2,∗ , MAGNUS LESTELIUS 2 , JOHAN TRYDING 3 and

NILS HALLBÄCK 4
1 Korsnäs Frövi, SE-71880 Frövi, Sweden
2 Karlstad University, Department of Chemical Engineering, SE-65188 Karlstad, Sweden
3 Tetra Pak Carton Ambient AB, SE-221 86 Lund, Sweden
4 Karlstad University, Department of Mechanical and Materials Engineering,
SE-65188 Karlstad, Sweden

Received in final form 28 December 2006

Abstract—Carton board packages are often closed with an adhesive. The adhesive joint thus formed
has to meet the demands during the entire product life from converting to end-use. The adhesive joint
has to be characterized if it is good or bad for the actual application. Today such characterization is
done by manually peeling the joint, immediately after the adhesive application in the gluing machine.
The manual peel test is a subjective test that is operator dependent. An operator needs long experience
to be able to perform a manual peel test. Therefore, the packaging industry is interested in a test
method that can objectively predict good or bad adhesive joints. The adhesive joints have been tested
in the so-called Y-peel test arrangement. An advantage of the Y-peel test is that it gives an objective
result from the force–elongation curve. Testing has been performed with carton boards of two different
thicknesses. Hotmelt adhesive was used and the open time was varied in the glue applicator. It was
found that the Y-peel test gives results in qualitative agreement with the manual peel test. Moreover,
by evaluating the energy consumption (dissipative energy) during the Y-peel test it was possible to
obtain not only a qualitative but also a quantitative assessment of the adhesive joint.

Keywords: Peel test; adhesive strength; adhesive joint; carton board; fibre tear; fracture energy.

1. INTRODUCTION

Carton board blanks are most often slid glued. During slid gluing the adhesive is
applied along one carton board side, and the opposite side is placed above the glued
side to form the joint. The process by which the glue joins two board sides together
involves a number of steps [1, 2].

∗ To whom correspondence should be addressed. Tel.: (46-581) 37456; Fax: (46-581) 31559;
e-mail: [email protected]
198 C. Korin et al.

• Application of the glue on the board substrate

• Wetting of the board by the glue
• Penetration of the glue into the board
• Consolidation of the glue by drying or cooling
• Final joint formation under pressure.
The adhesive joint should withstand all loads the carton board package is exposed
to during its total lifetime, such as filling, transportation, storage and consumer
handling. The mechanical loads are, e.g., internal and/or external pressure that can
build up during part of or the whole lifetime of the package.
To obtain a good adhesive joint, the following process parameters are important:
• Open time, i.e., the time from application of the glue to the sealing
• Close time, i.e., the time the sealed blank is under pressure
• Pressure on the sealed zone
• Amount of glue
• Temperature of the glue.
To check the adhesive joint quality, generally peel tests are employed. Tradition-
ally, the tests are performed through manual peeling, i.e., a person peels the adhesive
joint until it fractures. Thereafter, the fractured surface is studied and graded. No
common nomenclature for the grading exists today, but an adhesive joint is gener-
ally considered as good if the failure is in the form of fibre tear in the carton board.
An example of a grading nomenclature is:
• Fracture without any fibre tear or with less than 50% fibres on the fracture surface
is not acceptable.
• Fracture with more than 50% fibres on the fracture surface is acceptable.
The manual peel test can tell if the adhesive joint is acceptable or not. The
main objection to this test is that it is dependent on the individual who performs
the test, i.e., the method is subjective. Long experience in manual peel testing
is needed to adjust the gluing machine in a specific production unit to produce a
better adhesive joint after the manual peel test has been performed. Manual peeling
gives no objective information about the load at which the tearing starts or how the
tearing depends on the peel speed, glue amount, moisture content, temperature and
the geometry of the glue string. These parameters are important when the gluing
machine is to be adjusted to produce a better adhesive joint.
To overcome the shortcomings of manual peeling, a variety of instrumental test
methods have been proposed. In such test methods the load–displacement response
is recorded during peeling. Peel tests suggested in the literature for carton board
based packaging materials are, e.g., the T-peel [3], angle-peel [4] and Y-peel tests.
The Y-peel test is a redesign of the constrained T-peel test. The constrained
T-peel configuration is discussed in Ref. [5]. These methods have advantages and
disadvantages. Figure 1 shows the experimental set-ups for these three methods.
 Y-peel characterization of adhesively-bonded carton board 199

Figure 1. Peel tests suggested in the literature for carton board based packaging materials are the
(a) T-peel (reproduced from Ref. [3]), (b) angle-peel (reproduced from Ref. [4]) and (c) Y-peel tests.

A comparison of the methods in Fig. 1 shows that the adhesive joint in the T-peel
test is not fixed and is free to move during the test. The way to overcome this
movement is to stabilize the tail by holding the tail still or by adding a stabilising
tab [3], cf., Fig. 1a. In the angle-peel and the Y-peel tests no arbitrary movement
is allowed. It is further observed that in the angle-peel test the sample needs to
be bonded to a support block, cf., Fig. 1b. This makes the preparation work for
the angle peel test more time consuming compared to the other two methods. For
testing adhesively-joined carton boards, the Y-peel test is preferred, since it is easy
to use and the set-up is mechanically well defined, ensuring that the results depend
mainly on the adhesive joint (see Appendix A).
The objective of the present study was to investigate if the Y-peel test could be
used for objective assessment of adhesive joint characteristics. By varying the
open time, adhesive joints with different strength characteristics were obtained.
The validity of the test was examined by testing these joints and by checking
the capability of the test to capture the different characteristics of the joints. An
adhesive joint is here defined as two sheets of carton board with hotmelt adhesive
in between.

2. TEST DESCRIPTION
The Y-peel test offers the possibility for an objective assessment of the mechanical
properties of an adhesive joint, and this test was, therefore, adopted in the present
investigation. Sample preparation for the Y-peel test and evaluation of the load–
elongation curve is reported in Ref. [6].
By separating the load–elongation curve from the Y-peel test into elastic and
inelastic parts (see Fig. 2) it is possible to derive the inelastic behaviour of the
200 C. Korin et al.

Figure 2. Typical load–elongation curves from Y-peel tests (a) and their division into elastic (b)
and inelastic (c) parts (for details see Appendix A). The total elongation is denoted u (see panel 2a),
whereas the elastic part of u is denoted ue . The inelastic part of u is derived as u − ue .

Figure 3. Line force (fn ) versus inelastic deformation (δnie ) of an adhesive joint with an ascending
and a descending part.

adhesive joint (see Appendix A). This is based on the fact that the proportional limit
(yield strength) in the in-plane direction is at least 10-times higher than the out-of-
plane failure stress [7, 8]. Inelastic deformation is hence confined to out-of-plane
deformation of the adhesive joint. Using equations (A14) and (A15) the force versus
elongation curve can be transformed into the line force versus inelastic deformation
curve for the adhesive joint (see Fig. 3). The line force is here derived as the force
(Fn ) acting on the adhesive joint normal to the carton board divided by the width
(B, see Fig. 1c) of the Y-peel sample, i.e., fn = Fn /B. The inelastic deformation
of the adhesive joint (δnie ) is defined as the inelastic opening of the adhesive joint in
the direction of Fn .
 Y-peel characterization of adhesively-bonded carton board 201

From the line force versus inelastic deformation of the adhesive joint the dissipa-
tive energy per unit width (hereafter referred to as dissipative energy) can be com-
puted. The dissipative energy is caused by the damage (formation of micro-cracks
and subsequent coalescence of these cracks) and by the inelastic deformation in the
through-the-thickness direction of the carton board and/or the inelastic deformation
of the adhesive. Hence, the dissipative energy is a measure of the amount of energy
that is consumed during the separation of the joint. As can be seen in Fig. 3 the
dissipative energy can be divided into an ascending and a descending part.
In some cases, however, a rapid failure occurs at the maximum line force, so that
the descending dissipative energy is essentially zero. This situation is shown in
Fig. 4. The ascending part is the first part of the curve which ends at the maximum
line force (fn max ), cf., Figs 3 and 4. The energy per unit width that is consumed
during the ascending part of the line force versus inelastic deformation curve is
denoted WA . The descending part of the curve starts at the maximum line force and
ends at the line force where final failure occurs. During the descending part of the
line force versus inelastic deformation curve, coalescence of micro-cracks and fibre
pull out in the carton board results in a softening behaviour (see Ref. [7]). The
softening behaviour may, however, also be due to deformation and fracture in the
adhesive.
The energy per unit width that is consumed during the descending part of the line
force versus inelastic deformation curve is denoted WD .

Figure 4. Line force (fn ) versus inelastic deformation (δnie ) curve of an adhesive joint with only an
ascending part.
202 C. Korin et al.

3. EXPERIMENTAL
In order to experimentally investigate how the maximum line force and the ascend-
ing and descending parts of the line force versus inelastic deformation curve de-
pend on the adhesive joint characteristics and how these relate to the appearance of
the fracture surface, a series of tests were carried out on two different grammages
(weight per unit area according to ISO 536) of a 4-layer sulphate carton board (Frövi
LIGHT 250 and Frövi CARRY 425 from Korsnäs Frövi, Frövi, Sweden) with vary-
ing open time during gluing. The carton board was clay coated on one side and
uncoated on the other side. Two hotmelt glues (Swift B569/38, Forbo Swift Swe-
den, Gothenburg, Sweden and Henkel TS100, Henkel Norden, Gothenburg, Swe-
den, which are two adhesives on the market for the same field of application) were
used in the investigation. The tests were carried out in a controlled climate at a
temperature of 23◦ C and at 50% relative humidity. The adhesive joint consisted of
two carton board specimens glued together with the coated side of one carton board
specimen placed against the uncoated side of the other carton board specimen. In
Table 1 the test series 1 to 7 were performed with the board Frövi LIGHT 250 g/m2
(LGT250), glued with Henkel TS100 and the test series 8 to 15 were performed with
the board Frövi CARRY 425 g/m2 (CRY425), glued with Swift B569/38. In the test
series 1 to 7 the open time was varied from 1.4 s to 2.6 s and for the test series 8 to
15 the open time was varied from 1.4 s to 2.8 s. The other process parameters, i.e.,
pressure time (1 s), nozzle size (diameter 0.46 mm, Nordson Saturn Precision Noz-
zle number 322418, Nordson, Malmö, Sweden), application temperature (160◦ C in
the tank, 165◦ C in the hose and 175◦ C in the nozzle), were kept constant.
Table 1.
Y-peel test results

Test Number Board Open WA ± std. dev. WD ± std. dev. Maximum Percent of
series of type time (mJ/m) (mJ/m) line force ± std. dev. fibre tear
samples (s) (N/m) samples
1 10 LGT250 1.4 319 ± 82 1844 ± 134 971 ± 131 100
2 10 LGT250 1.6 351 ± 129 1778 ± 392 989 ± 90 100
3 10 LGT250 1.8 316 ± 182 1653 ± 307 987 ± 160 100
4 10 LGT250 2 241 ± 80 1822 ± 276 893 ± 81 100
5 10 LGT250 2.2 427 ± 175 660 ± 408 1239 ± 204 90
6 10 LGT250 2.4 221 ± 106 232 ± 134 763 ± 52 20
7 10 LGT250 2.6 107 ± 71 44 ± 28 659 ± 57 0
8 9 CRY425 1.4 142 ± 64 1677 ± 356 1535 ± 135 100
9 10 CRY425 1.6 200 ± 88 2606 ± 265 1565 ± 84 100
10 10 CRY425 1.8 280 ± 229 2377 ± 571 1534 ± 156 100
11 10 CRY425 2 187 ± 55 2418 ± 555 1561 ± 169 100
12 10 CRY425 2.2 187 ± 124 1857 ± 851 1733 ± 141 100
13 10 CRY425 2.4 218 ± 182 1039 ± 962 1659 ± 142 90
14 10 CRY425 2.6 173 ± 232 713 ± 1087 1820 ± 236 50
15 10 CRY425 2.8 60 ± 160 109 ± 62 1451 ± 240 0
 Y-peel characterization of adhesively-bonded carton board 203

The Y-peel test device was mounted in a tension-testing machine (Zwick 005
Machine, ZR Sverige, Kullavik, Sweden). During the test, the force and elongation
were measured (the precision of the machine is 0.5% for the force and 0.2% for
the elongation, according to the manufacturer). The compliance, i.e., the inverse
stiffness, of the machine affects only the elastic part of the force–elongation curve.
Hence the accuracy of the inelastic part depends only on the precision of the force–
elongation measurement of the machine, provided that no sliding occurs in the
clamps.

4. RESULTS

The ascending and descending dissipative energy together with the maximum line
force for the tested samples are shown in Table 1. The percent fibre tear samples
are also given in Table 1. If a tested sample exhibited fibres on the glue string by a
visual judgement after fracture of the adhesive joint, the failure was considered as
fibre tear. If all tested samples in a test series exhibited fibre tear, the percent fibre
tear samples for that test series was set to 100%. If no fibre tear was observed on any
sample the percent fibre tear samples was set to 0%. As an example, if five samples
out of ten tested samples exhibited fibre tear, then the percent fibre tear samples for
that test series was set to 50%.
The maximum line force, fn max and WA are plotted versus the open time in
Figs 5 and 6, respectively. The figures show that fn max and WA were not affected
by the open time.
In Fig. 6 it is also seen that there is a large spread in the results for WA . This is
probably due to the fact that WA is small.
Plots of WD and the percent fibre tear samples versus the open time are shown
in Figs 7 and 8, respectively. In Fig. 7 it is evident that the descending dissipative
energy in the open time interval 1.4 s to 2.0 s was almost independent of the open
time. For open times longer than 2.0 s the descending dissipative energy decreases
with increasing open time. Figure 8 shows that the percent fibre tear samples were
not affected by the open time for values below 2.0 s. For open times longer than
2.0 s the percent fibre tear samples decreased with increasing open time.

5. DISCUSSION

Adhesive joints with different open times were made and evaluated with the Y-peel
setup. The open time was varied while keeping all other process parameters
constant. Two different grammages of carton board were tested.
Increased open time results in increased cooling of the heated hotmelt adhesive.
A likely effect is that, since a cooler adhesive has a higher viscosity, it would not be
able to wet and conform to the carton board surface during final joint formation to
the same extent as a hotter and less viscous adhesive.
204 C. Korin et al.

Figure 5. Maximum line force vs. open time with standard deviation.

Figure 6. Ascending dissipative energy vs. open time with standard deviation.

The distinct decrease in percent fibre tear samples was accompanied by a decrease
in WD at approximately the same open time (2.0 s, see Fig. 7). This is not surprising,
since the descending part dominates the total dissipative energy in cases of extensive
fibre tear (cf. Ref. [9] for further details on this topic). This indicates that WD may
be used for quality assessment of adhesive joints. This conclusion seems to be valid,
irrespective of the board grammage, at least for grammages in the range used in the
present investigation.
 Y-peel characterization of adhesively-bonded carton board 205

Figure 7. Descending dissipative energy vs. open time with standard deviation.

Figure 8. Percent fibre tear samples vs. open time.

The fact that the maximum line force does not depend significantly on the open
time implies that fracture initiation occurs in a similar manner, irrespective of the
percent fibre tear samples. Fracture initiation is defined here as when a major crack
is developed due to micro-crack coalescence. Once a major crack has developed
the subsequent event will either be a fast running crack, leading to catastrophic
failure of the structure, or crack arrest. Crack arrest requires that the material
ahead of the crack tip is capable of absorbing the energy released due to fracture
206 C. Korin et al.

initiation (cf., Ref. [9]). In order to do so, the material must be able to sustain
continued deformation beyond the maximum force. Since fracture initiates at the
maximum force (cf., Ref. [7]) this means that the material must possess a softening
behaviour after reaching the maximum force. The descending dissipative energy
is hence a measure of the resistance against crack propagation, which makes the
joint less sensitive to local stress concentrations e.g., at corners and folds in a carton
board package. In this context it should, however, be pointed out that a softening
behaviour may be due to causes other than fibre tear.
If fracture takes place entirely within the adhesive, and the adhesive possesses a
softening behaviour after reaching the maximum force, the descending dissipative
energy could be high even though the percent fibre tear samples is zero. Evidently
this represents a situation which could not be reliably assessed using the percent
fibre tear samples. However, performing a Y-peel test allows for assessment
of adhesive joints independent of the fracture mechanism, via the descending
dissipative energy and the maximum force. In this work, the maximum force turned
out to be constant with respect to the open time. In other cases, or in cases where
other process parameters are varied, this may no longer be the case. This may,
for instance, occur if the location for the fracture initiation, due to varying process
conditions, shifts between the carton board, the adhesive, or the interface between
the carton board and the adhesive. How to judge the combined effect from a varying
descending dissipative energy and maximum force on the integrity of adhesive joints
is, however, not trivial. Further research is needed in this respect.

6. CONCLUSION

This investigation proposes the Y-peel test as a method for objective, and person
independent, classification of adhesive joints with respect to crack propagation
resistance (via the descending dissipative energy) and the propensity for fracture
initiation (via the maximum force).

Acknowledgements
The authors would like to thank Ms. Therese Björklind, former MSc Student at
Örebro University, and Ms. Anna Jansson, Korsnäs Frövi, for their help in analysing
Y-peel curves and Mrs. Inga-Lill Jansson, Korsnäs Frövi, for laboratory assistance.
The authors also offer their thanks to Dr. Ola Karlsson, Korsnäs Frövi, Prof. Lars
Ödberg, Royal Institute of Technology (KTH), and Ass. Prof. Gunnar Engström,
Karlstad University, for careful scrutiny of the results and the manuscript. The
authors are much obliged to Mr. Roogher Johansson, Forbo Swift Sweden AB,
and Mr. Jonas Jarhäll, Henkel Norden AB, for their encouragement as well as
for supplying the adhesives used in this investigation. The financial support and
laboratory resources from AssiDomän Cartonboard are gratefully acknowledged.
 Y-peel characterization of adhesively-bonded carton board 207

REFERENCES
1. L.-H. Lee (Ed.), Fundamentals of Adhesion. Plenum Press, New York, NY (1991).
2. L.-H. Lee (Ed.), Adhesive Bonding. Plenum Press, New York, NY (1991).
3. M. Edin, L. Ödberg and J. Sterte, Nordic Pulp Paper Res. J. 17 395–400 (2002).
4. A. J. Kinloch, C. C. Lau and J. G. Williams, Int. J. Fracture 66, 45–70 (1994).
5. R. P. Wool, Polymer Interface Structure and Strength. Hanser/Gardner Publication, Cincinnati,
OH (1995).
6. J. Tryding, M. Lestelius, C. Korin and J. Lewandowski, in: Proceedings of VVD 2003, Dresden,
pp. 277–294 (2003).
7. Q. S. Xia, Mechanics of Deformation in Paperboard, PhD thesis. Department of Mechanical
Engineeering, Massachusetts Institute of Technology, Cambridge, MA (2002).
8. Q. S. Xia, M. C. Boyce and D. M. Parks, Int. J. Solids Struct. 39, 4053–4071 (2002).
9. J. Tryding, In-Plane Fracture of Paper, PhD thesis. Division for Structual Mechanics, Lund
Institute of Technology, Lund (1996).
10. S. Timoshenko and D. H. Young, Theory of Structures, 1st edn. McGraw-Hill, New York, NY
(1945).

APPENDIX A

An analysis of the Y-peel set-up, under the assumption of small deformations, is

presented. The theory of deformation of truss and joints is discussed in detail by
Timoshenko and Young [10]. Due to symmetry of the structure only one half of
the set-up needs to be analysed. Consider the forces acting on one half of the set-
up as shown in Fig. A1. The total force applied to the upper end of the structure
(F ) is equally divided between the two symmetric parts, while the reaction force
in each leg is denoted R. Balance of moment with respect to the adhesive joint

Figure A1. Forces acting on one half of the Y-peel set-up.

208 C. Korin et al.

results in that the moment carried by the adhesive joint must be zero. Moreover, the
symmetry constraint implies that the transverse force must be zero as well. Hence
the adhesive joint carries only a normal force (Fn ). Equilibrium in the vertical and
horizontal directions yields

F
− R cos(α/2) = 0 (A1)
2

and

Fn − R sin(α/2) = 0. (A2)

Combining these equations gives

F
Fn = tan(α/2). (A3)
2

Hence the force acting on the adhesive joint can be easily derived from the applied
force F . A sketch of the deformation of the Y-peel set-up is shown in Fig. A2,
where the region adjacent to the adhesive joint is shown in greater detail under the

(a) (b)

Figure A2. (a) Deformation pattern for one half of the Y-peel set-up. (b) Close-up of the joint region
assuming small deformations.
 Y-peel characterization of adhesively-bonded carton board 209

assumption of small deformations (see Ref. [10]). From Fig. A2 it is readily derived
that
δ1 = u − δt (A4)
δn
δ2 = δt cos(α/2) − sin(α/2), (A5)
2
where δ1 and δ2 denote the elongations of the board segments, and δn and δt are the
normal opening and the transverse sliding displacement of the joint, respectively.
The deformation of the board is assumed to be elastic, while the deformation
of the adhesive joint is assumed to be composed of elastic and inelastic parts
(the outline presented in Ref. [6] was somewhat incomplete regarding the elastic
part of the deformation). The elastic deformation is defined as that part of the
deformation which is recovered when the loading is released. The remaining part
of the deformation is defined as the inelastic deformation. The constitutive relations
for the carton board could, thus, be expressed as
F
δ1 = C1 , (A6)
2
F
δ2 = C2 R = C2 , (A7)
2 cos(α/2)
where equation (A1) has been utilized and C1 and C2 are the compliances, i.e., the
inverse of stiffness, of the carton board segments with lengths l1 and l2 , respectively.
For the adhesive joint we have
δn = δne + δnie , (A8)
where δne and δnie denote the elastic and inelastic parts of the deformation, respec-
tively. The elastic part of the deformation can be computed as
F
δne = Cn Fn = Cn
tan(α/2), (A9)
2
where equation (A3) has been utilized and Cn denotes the elastic compliance
of the adhesive joint. Combining equations (A4)–(A9), with some algebraic
manipulations, gives

  
2u C2 1 Cn
δn =
ie
− F C1 + + tan(α/2) . (A10)
tan(α/2) cos2 (α/2) tan(α/2) 2
Assuming that the inelastic deformation of the adhesive joint is zero, i.e., δnie = 0,
the response is then purely elastic so that
u = ue = Ce F, (A11)
where ue is the elastic elongation and

  
tan(α/2) C2 1 Cn
Ce = C1 + + tan(α/2) (A12)
2 cos2 (α/2) tan(α/2) 2
210 C. Korin et al.

is the elastic compliance of the Y-peel setup. Hence, the inelastic deformation of
the adhesive joint can be computed as
2u 2ue
δnie = − . (A13)
tan(α/2) tan(α/2)
From equation (A3) and equations (A11) and (A13) the force and inelastic defor-
mation in the adhesive joint can easily be computed from the recorded load and
elongation in the Y-peel test. Since the elastic compliance Cn of the adhesive joint
is unknown, and difficult to determine independently, equation (A12) cannot be
used to compute Ce . Instead, the elastic compliance (Ce ) is determined from the
initial slope of the force–elongation curve. When α = 90◦ , equations (A3) and
(A13) become particularly simple, i.e.,
F
Fn = , (A14)
2
δnie = 2(u − ue ). (A15)
It is convenient to normalise the force (Fn ) to the line force (fn ) according to
Fn
fn =, (A16)
B
where B is the width of the Y-peel sample. In this way results from different Y-peel
tests can be directly compared irrespective of the specimen width. The dissipative
energy per unit width at fracture of the adhesive joint can be computed as the area
under the fn –δnie curve, i.e.,
 (δnie )f
W = fn dδnie , (A17)
0

where (δnie )f is the inelastic deformation at final fracture of the adhesive joint.