IOP Conference Series: Earth and Environmental Science

PAPER • OPEN ACCESS

Photoelastic stress analysis of mode I fracture toughness tests using

PMMA samples
To cite this article: A Muñoz-Ibáñez et al 2021 IOP Conf. Ser.: Earth Environ. Sci. 833 012031

View the article online for updates and enhancements.

This content was downloaded from IP address 152.171.31.180 on 26/09/2021 at 17:07

Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

Photoelastic stress analysis of mode I fracture toughness tests

using PMMA samples

A Muñoz-Ibáñez, M Herbón-Penabad, J Delgado-Martín

School of Civil Engineering, Universidade da Coruña, Campus de Elviña s/n, A
Coruña, 15071, Spain

[email protected]

Abstract. Rocks are usually inhomogeneous and anisotropic materials. The presence of foliation
planes, grain boundaries or even microcracks may alter the stress distribution. In order to identify
whether unusual behaviours in rocks are due to these imperfections or result from other factors
(e.g. experimental configuration), the analyses of homogenous and isotropic materials is an
useful approach. We have performed a series of mode I fracture toughness (KIC) tests using
polymethyl methacrylate (PMMA) samples, which has the advantage of allowing photoelastic
stress analysis based on its birefringent nature. Three different testing configurations were
considered in the study: S\ emi-circular bend (SCB) test, the pseudo-compact tension
(pCT) test, and a new alternative configuration based on the previous two that we have called
pseudo-SCB (pSCB) test. To perform the photoelastic analysis, all the experiments were
complemented with a specially-designed experimental setup consisting in two orthogonally
arranged circular polarizers placed on both sides of the tested specimens. Using a source of white
(polychromatic) light on one end it is possible to record the stress distribution using a digital
camera aligned with the samples on the other end. As the load increases, a distinct evolving
pattern of colour fringes can be visualized in the samples illustrating the spatially distributed
stress levels. Based on this analysis we observe in some of the tests performed non-symmetrical
stress fields. Although this behaviour could be related with the testing configuration, results
suggest that other features, such as the shape of the notch tip, imperfections in sample
preparation, or the misalignment of the samples in the testing device may also have an influence
in stress distribution.

1. Introduction
Rock materials are usually discontinuous at all scales [1]. At the microscale, the presence of pores, grain
boundaries, and pre-existing fractures can lead to high stress concentrations under increasing load. At
the macroscale, foliation or bedding planes are characterized by having lower resistance and represent
zones of weakness. These discontinuities might be determinant in fracture growth and, therefore, play
an important role in the success of engineering projects involving rock materials (e.g., hydraulic
fracturing, nuclear waste disposal or geothermal energy). Mode I fracture toughness (KIC) represents the
resistance of a material to the propagation of tensile cracks under the presence of pre-existing
discontinuities. In KIC assessment in rocks, properties such as grain size, particle arrangement, or the
degree of cementation can determine the propagation path of the cracks [2]. To identify additional
factors the use transparent models made of materials such as polymers, resins or glass is an interesting
approach specially to visualize fracture development. For instance, Wu et al [3] monitored the process

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

of crack growth on hydraulic fracture tests using polymeric samples. Their results suggest that far-field
stresses determine the orientation of the final fracture orientation but have little effect on fracture
initiation. Although the materials used in this type of studies are usually homogeneous and have,
therefore, a limited ability to improve the understanding about the behavior of heterogeneous samples,
they are valuable to identify additional factors (e.g. geometrical features, loading conditions) that can
have an effect on fracture propagation.
Photoelasticity is a non-destructive testing technique used to visualize stress distributions and
determine points of stress concentration in a material subjected to an external load. The method is based
on the opto-mechanical coupling of birefringent materials, which is characteristic of many common
transparent polymers [4]. A birefringent material, when properly aligned with other elements (i.e., light
source, polarizers, retarders), will exhibit chromatic fringe patterns as a function of the applied stress. If
polarized white (polychromatic) light travels across a birefringent stressed material, some characteristic
optical features develop: dark bands (or isoclines) and colored fringes (or isochromes). The isochromes
represent lines of constant principal stress difference in a plane normal to the direction of light
propagation, while the isoclines indicate locations with the same direction of the principal stresses [5].
In rock mechanics, photoelastic birefringent materials can be used to investigate stress patterns arising
from specific testing methods and for model matching.
In this study we have carried out KIC tests using polymethyl methacrylate (PMMA) samples. This
material was selected not only for being homogeneous and isotropic but also for its birefringent
properties, which allow performing photoelastic stress analysis (PSA) concurrent to test execution. The
study has been focused in three different testing methodologies: i) the semi-circular bend (SCB) test,
which is one of the suggested methods proposed by the International Society for Rock Mechanics [6];
ii) the pseudo-compact tension (pCT) test, recently proposed by to assess KIC in rocks under pure tensile
conditions [2]; and iii) the pseudo-semi-circular bend (pSCB) test, which can be regarded as a
combination of the two mentioned testing methods, and that is further described in [7]. In this work, the
KIC values of PMMA derived with the three methodologies are discussed, while the stress distribution
in the samples was also investigated.

2. Materials and methods

2.1. Materials and sample preparation

PMMA, also known with the trade-names of Perspex®, Plexiglass®, or Lucite®, is a transparent brittle-
elastic polymer of high strength and rigidity [3,9]. Reference values for the compressive strength (83-
124 MPa), tensile strength (30-76 MPa) or elastic modulus (2950-3300 MPa) of this material at room
temperature are reported in the literature [9,10].

Table 1. PMMA specimen dimensions: D = diameter; B = thickness; a/b or a/R = notch length
ratio; R = radius; s/D = span length; Gd = U-shaped groove depth; Gw = U-shaped groove width.

Method D (mm) B (mm) a/R or a/b s/D Gw (mm) Gd (mm)

SCB 50 25 0.28-0.60 0.8 - -
pCT 50 25 0.50-0.75 - 10 5
pSCB 50 25 0.50-0.75 - 10 5

In this study, samples for fracture toughness tests were obtained from 50-mm diameter PPMA bars,
which were sliced into discs (thickness-to-diameter ratio of ~0.5) using a lathe. For the SCB specimens,
the discs were halved and a straight notch was cut in the center of the flat surface. In the pCT and pSCB
specimens, a U-shape groove was carved in the discs to allow sample loading, and a straight notch was
also cut to act as stress concentrator. Subsequently, the surfaces of the samples perpendicular to the ray
of light were polished to enhance fringe visualization. We considered a range of notch length ratios for

2
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

each testing method. Dimensions of the samples are given in table 1 and a schematic illustration of the
three geometries is shown in figure 1. It should be noted that the PMMA specimens were not annealed
after machining. Consequently, we would expect observing a residual stress field in the samples before
testing.

Figure 1. Specimen geometries and loading configuration. a) SCB test; b) pCT test; c) pSCB test.

2.2. Fracture toughness testing setups

Mode I fracture toughness tests were performed following three different experimental approaches:
SCB, pCT, and pSCB methods. The corresponding setups are illustrated in figure 2. SCB specimens
were tested following the recommendations of the ISRM [6] with a three-point bending fixture located
in a servo-electric frame equipped with a 4,5 kN load cell. In this configuration the sample is supported
by two steel rollers (separated a fixed distance, s). A third roller is used to deliver the load to the top of
the sample. In this work, the span length (s/D) was set to 0.8 as suggested for testing strong materials.
pCT and pSCB experiments were carried out following the procedure described in [2]. Tests were
performed with a specially-designed testing device equipped with a 50 kN load cell. All the experiments
were conducted at room conditions with a constant displacement rate of 0.1 mm/min. Load (P) and load
point displacement (LPD) were recorded continuously during testing. LPD corresponds to the
displacement of the loading roller in the SCB tests, and to the movement of the steel jaw in pCT and
pSCB tests.

Figure 2. Experimental setup for mode I fracture toughness tests. a) SCB; b) pCT; c) pSCB.

Level I testing was considered in this study and, therefore, only the value of maximum load (Pmax)
recorded was considered to compute mode I fracture toughness. KIC was calculated using the expressions
given in [6] for the SCB specimens, and in [2] for the pCT and pSCB specimens. For the latter method,
the expression of the dimensionless stress intensity factor (Y’) derived for 50 mm- diameter specimens
is given in [7].

3
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

2.3. Photoelastic setup

A polariscope was used for visualizing the stress field on the tested specimens (figure 3). The setup
consists of a source of led white light and two 82 mm-diameter circular polarizers (K&F Concept)
equipped with ¼- waveplates (or retarders). The configuration of the setup is cross-polar (i.e. the
polarizer, next to the light source, and the analyser, next to the camera, polarize light at 90º) with the
sample located between the two filters. With no sample, the cross-polar configuration produces light
beam extinction (i.e. a dark field) while the insertion of the sample illuminates the field with coloured
(isochromes) and black (isoclines) fringes whose development is related with the stress distribution in
the sample: A non-stressed sample will theoretically result in a single isochrome occupying the whole
surface of the sample. The experiments were recorded with a digital camera (KOPACE, KP-2100) with
a resolution of 1920x1080 pixels and a frame rate of 60 f/s.

Figure 3. PSA setup attached to the testing equipment used to perform pCT tests.

3. Results and discussion

3.1. Mode I fracture toughness assessment

Typical load-displacement (P-LPD) curves obtained from the experiments are plotted in figure 4.
We observe that load increases linearly (elastic phase) up to Pmax. Afterwards the sample fails abruptly,
proving the brittleness of the material and in agreement with the behaviour reported by Torabi et al [10]
for PMMA samples loaded under mixed mode I/III. As previously discussed in [2] for rock samples, the
pCT test configuration allows good control on fracture propagation even in the post-peak behaviour: the
material losses strength but still retains some cohesion. This is also evidenced by the loading curves
recorded for the pSCB sample geometry. Contrary, because the SCB configuration is more prone to
elastic energy storage, its release at the critical stress leads to a fast, uncontrolled fracture propagation.

4
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

Figure 4. Load vs load point displacement (P-LPD) curves (left); and mode I fracture toughness (KIC)
vs notch length ratio (right).

Mode I fracture toughness results obtained from the SCB, pCT, and pSCB tests have been plotted as
a function of notch length ratio in figure 4. SCB tests result in the higher KIC values while the lower ones
correspond to the pSCB tests. Apparently, fracture toughness has some dependence on the ligament
length for the ranges considered in this study, with decreasing values for larger notches (i.e., shorter
ligaments). This behaviour has been also observed by Visser et al [11] in PMMA samples during tensile
impact testing. In our experiments, this effect appears to be more prominent in the case of the pSCB
specimens.
Although a graphical analysis can be useful as a first approach, we opted for performing statistical
analysis with the aid of the free software Past 3.0 [12] to get quantitative information about test
repeatability and reproducibility. To this aim, first we split the values of KIC obtained from the
experiments into three groups according to the testing method. Then, we performed two statistical
checks considering a significance level of 95%: i) a within-group analysis of repeatability aimed at
assessing normality through a non-parametric Shapiro-Wilk test, and ii) a between-group analysis
focused on the comparison of means and based on a one-way analysis of variance (ANOVA) and the
non-parametric Mann-Whitney pairwise test. The three groups considered conform to normal
distributions, which suggests that the length of the notch might not be a determinant factor on the
assessment of KIC. Contrary, the between-group analysis to assess reproducibility was unsuccessful in
the three cases considered. This may be indicating a strong dependence of fracture toughness on the
testing conditions and sample geometry for this kind of material, in agreement with what we can
perceive from the plots. Mean values of KIC are given in table 2 for each testing method. We observe
that KIC has the lowest variability in the case of the pCT test, even considering a broader range of notch
length ratios than for the SCB and pSCB tests. This behaviour agrees with that observed previously by
Muñoz-Ibáñez et al [13] for pCT and SCB tests performed with four different rock types.

Table 2. Mode I fracture toughness (KIC) values (mean ± standard error of

the mean) for PMMA samples. Number of experiments is given into brackets.

Method KIC (MPa m1/2)

SCB 1.96 ± 0.06 (9)
pCT 1.80 ± 0.03 (13)
pSCB 1.54 ± 0.04 (5)

5
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

The values of KIC reported in this study are higher than those previously given by Sedighi et al [14]
for thin (B = 4.7 mm) compact tension (1.28 MPa m1/2) and SCB (1.30 MPa m1/2) specimens. However,
the fracture toughness values reported by Weerasooriya et al [15] for four-point bending experiments
(1.42-1.68 MPa m1/2) are in agreement with those obtained for the pSCB specimens. Interestingly, the
KIC results obtained by Ayatollahi and Saboori [16] for PMMA samples loaded under pure tensile
conditions (1.7-1.8 MPa m1/2) are in line with those of the pCT tests but not with those of pSCB tests.
These results suggest that not only the loading conditions but also the specimen geometry have an impact
in KIC determination.

3.2. Photoelastic stress analysis (PSA)

Figure 5 shows some examples of the polariscope images resulting from the SCB and pCT tests.
Isochromes are shown for the different loading levels. In the SCB test, the lack of control of crack
propagation prevents stress assessment after Pmax. It must be noted that in this study we focused on
performing a qualitative analysis of the stress distribution from the images recorded, with the aim of
identifying features having an influence on the results. However, a complete study of the stress field
(i.e., the assessment of the direction and distribution of the principal stresses) is not reported.

Figure 5. Time evolution of the stress field (from top to bottom) in SCB (left) and pCT (right) samples.

6
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

The fringe patterns observed in the pictures provide information about the distribution and
concentrations of stresses. The colours of the isochromes appear in a sequence that is function of the
stress level [17]. In the pCT test, we first observe fringes appearing mainly at the notch tip (the area
where the highest stress should develop) which then extends towards the loading ends. The bottom part
of the sample remains mostly unstressed. As the load increases, the first fringes move towards zones of
lower stress and simultaneously, additional new fringes appear around the notch tip, indicating a
growing stress concentration in this region. At Pmax, the stress is located on the right side of the ligament
plane, and we observe that the crack develops towards this area after peak load crossing. As the crack
propagates, the stress is released at the notch tip and then mainly concentrates in the vicinity of the
growing crack tip. In the case of the SCB test, high stress concentrations are observed not only at the
notch tip, but also at the contact points with the rollers. At peak load, the high density of fringes observed
at the top part of the sample would suggest a significant energy storage in this area [18], which then led
to fast crack propagation with no post-peak control. The shape of the stress field at the notch tip (i.e.,
ellipses that become pointed at the tip) agrees with the prediction of the Linear Elastic Fracture
Mechanics (LEFM) theory, less clearly for the pCT sample and better defined for the SCB specimen.
Similar results were reported by Pitarresi et al [19] for epoxy samples loaded under three-point bending.
As previously mentioned, crack deviation from the ligament plane was observed in some
experiments. In fact, the images resulting from the polariscope reveal that the stress field in the sample
was not symmetrical, even for the SCB tests. The issue of fracture deviation has been previously
observed in pCT and SCB tests performed with rocks [2,13] and was attributed to their likely
heterogeneous and non-isotropic nature. However, crack deviation is also observed in this study for the
homogeneous PMMA samples, which may be indicating that there might be additional features causing
this behaviour. In this regard the asymmetric loading conditions, inherent to the pCT configuration and
common in the SCB tests if the specimens are misaligned with respect to the top roller [12], could have
a significant contribution. However, if that were the only cause for crack deviation, we would expect
that crack development would always occur towards the same side of the specimens, at least in the pCT
tests. Contrary, the deviation did not follow any discernible pattern, with specimens presenting cracks
deviating towards either side. This was also previously observed by Muñoz-Ibáñez et al [2] for rock
specimens. Therefore, there might be additional factors influencing fracture deviation. We conjecture
that this could be related with imperfections in sample preparation (e.g., misalignment of the notch tip
from the vertical plane or stress concentrations at the contact with loading ends) or the shape of the
notch tip.

4. Conclusions
In this paper, a photoleastic stress analysis was carried to determine the stress field in mode I fracture
toughness tests of PMMA samples. Tests were performed following three different testing
methodologies: SCB, pCT and pSCB tests. The loading curves of the PMMA samples failed abruptly
after the peak load, in agreement with the brittle nature of the material. However, despite this behaviour,
the pCT and pSCB tests were able to provide control even in the post-peak region. The results of KIC
obtained in this study suggest a dependence on the loading conditions and on the sample geometry. The
images from the polariscope revealed a high stress concentration at the notch tip (i.e., the point for crack
initiation) and at the loading ends, especially in the case of the SCB tests. The stress field was
asymmetric, and this behaviour would be caused not only for the asymmetric loading configuration but
also for geometrical features such as imperfections in sample preparation. Our results revealed that these
features can significantly influence the stress distribution in the sample and consequently play an
important role in fracture propagation.

Acknowledgements
This work was funded by the MINECO/AEI/FEDER, UE project BIA2017- 87066-R.

7
Mechanics and Rock Engineering, from Theory to Practice IOP Publishing
IOP Conf. Series: Earth and Environmental Science 833 (2021) 012031 doi:10.1088/1755-1315/833/1/012031

References
[1] Zhang JJ 2019 Basic rock fracture mechanics Applied Petroleum Geomechanics (Gulf
Professional Publishing) chapter 4 pp 133-61
[2] Muñoz-Ibáñez A, Delgado-Martín J, Costas M, Rabuñal-Dopico J, Alvarellos-Iglesias J and
Canal-Vila J 2020 Mode I fracture toughness determination in rocks using a pseudo-compact
tension (pCT) test approach. Rock Mech Rock Eng 53(7) 3267-85
[3] Wu H., Golovin E., Shulkin Y, Chudnovsky A., Dudley JW and Wong GK 2008 Observations of
hydraulic fracture initiation and propagation in a brittle polymer The 42nd U.S. Rock
Mechanics Symposium (USRMS), San Francisco, California
[4] Phillips JW 1998 Experimental stress analysis University of Illinois
[5] Dally JW, Rilley WF 1991 Experimental stress analysis (New York: McGraw-Hill Book Co)
[6] Kuruppu MD, Obara Y, Ayatollahi MR, Chong KP and Funatsu T 2014 ISRM-suggested method
for determining the mode I static fracture toughness using semi-circular bend specimen Rock
Mech Rock Eng 47 267–74
[7] Herbón-Penabad M, Muñoz-Ibáñez A and Delgado-Martín J (under review) Numerical and
experimental analysis of pseudo-compact tension (pCT) testing configuration using two
alternative sample geometries EUROCK 2021, Torino, Italy
[8] Laganà A, David M, Doutre M, de Groot S, van Keulen H, Madden O, Schilling M and van
Bommel M. 2021 Reproducing reality. Recreating bonding defects observed in transparent
poly(methyl methacrylate) museum objects and assessing defect formation J Cult Herit
(Preprint)
[9] Samavedi S, Poindexter LK, Van Dyke M and Goldstein AS 2014 Synthetic biomaterials for
regenerative Medicine Applications Regenerative Medicine Applications in Organ
Transplantation (Academic Press) chapter 7 81-99
[10] Torabi AR, Saboori B, Keshavarz Mohammadian S, Ayatollahi MR 2018 Brittle failure of
PMMA in the presence of blunt V-notches under combined tension-tear loading: Experiments
and stress-based theories Polym Test 72 94-109
[11] Visser HA, Caimmi F and Pavan A 2013 Characterising the fracture toughness of polymers at
moderately high rates of loading with the use of instrumented tensile impact testing
Engineering Fracture Mechanics 101 67-79
[12] Hammer Ø 2011 PAST PAleontological STatics Reference Manual (Natural History Museum.
University of Oslo)
[13] Muñoz-Ibáñez A, Delgado-Martín J and Juncosa-Rivera R 2021 Size effect and other effects on
mode I fracture toughness using two testing methods Int J Rock Mech Min Sci 143C 104785
[14] Sedighi I, Ayatollahi MR and Bahrami B 2020 A statistical approach on the support type effect
on mode I fracture toughness determined using semi-circular bend (SCB) specimen Eng Fract
Mech 226 106891
[15] Weerasooriya T, Moy P, Casem D, Cheng M, Chen W 2006 Fracture toughness for PMMA as a
function of loading rate.
[16] Ayatollahi MR and Saboori B 2015 A new fixture for fracture tests under mixed mode I/III
loading Eur J Mech A/Solids 51 67-76
[17] Vishay Precision Group 2011 PhotoStress Instruments: Introduction to stress analysis by the
PhotoStress method (Technical Note TN-702-2)
[18] Ju Y, Ren Z, Mao L and Chiang FP 2018 Quantitative visualisation of the continuous whole-field
stress evolution in complex pore structures using photoelastic testing and 3D printing methods.
Optics Express 26 (5) 6182-201
[19] Pitarresi G, Toscano A, Scafidi M, Di Filippo M, Alessi S and Spadaro G 2014 Photoelastic stress
analysis assisted evaluation of fracture toughness in hydrothermally aged epoxies Frattura ed
Integrità Strutturale 30 127-37