Sage Journals: Discover world-class research

Abstract

Increased lightweight design via composite sandwich structures is a promising approach due to their exceptionally high weight-specific mechanical properties. The involved high cost when using sandwich composite structures has hindered applications in cost-sensitive markets up to now. However, full thermoplastic composite sandwich structures enable a cost reduction using novel processing routes based on fusion bonding of core and facesheet. In order to optimize such full thermoplastic composite sandwich structures and to ensure proper bonding of facesheet and core after manufacture, valid test methods for the quantification of skin-core interfacial bonding are required. This paper reviews existing test methods for the determination of Mode I dominant interfacial fracture toughness of sandwich structures. The main focus is set on cantilever beam tests as well as on peel tests. Based on a definition of requirements for suitable test methods, their applicability for full thermoplastic composite sandwich structures is evaluated. The Mandrel Peel Test results as most promising test method for thermoplastic sandwich structures, especially with thin facesheets.

Keywords

Sandwich structures interfacial fracture toughness Mode I thermoplastic composites skin-core bonding

Introduction

Lightweight design plays an increasingly important role in modern transportation, due to growing energy efficiency requirements. The electrification of cars is a popular example. In order to meet more and more rigorous CO₂ emission limitations of car fleets, the share of battery electric vehicles will grow if automotive companies want to avoid drastic financial penalties.¹ The traction battery however introduces several 100 kg of weight to the car, and despite simplifications of the electric drivetrain compared to internal combustion engine (ICE) vehicles in general, it leads to a significantly increased car weight as observed in modern battery electric vehicles (BEVs). For this reason, different approaches have been developed to reduce the mass of BEVs, such as the extensive use of carbon fibre reinforced plastics in BMW i3 and Polestar 1, yet with high costs involved. In order to increase cost-efficiency, thermoplastic composites are a promising candidate and a variety of different thermoplastic composite components have been developed and introduced in automotive applications using novel processing methods.^2–5 In order to further increase lightweight design of cost-efficient thermoplastic composites, full thermoplastic sandwich structures and process methods have been of interest by different researchers due to their high weight-specific mechanical properties. For this reason, a number of different thermoplastic-based sandwich composite structures and processing techniques have been developed in recent years.^6–24

Sandwich structures in principle consist of a three-layer cross-section with a core and two covering facesheets.²⁵ The resulting composition is similar to an I-beam structure with a clear division of tasks between facesheets and core, following the stress distribution in the cross-section induced by bending loads.²⁶ The facesheets carry the majority of the normal stresses. Contrary, the core shall bear mostly the transverse shear stresses between the facesheets and support the factsheets at compressive loads in thickness direction.²⁷ In addition, the core leads to an increase of the second moment of area due to its thickness, resulting in an over-proportional growth of the flexural rigidity.^28,29 The relatively low stress levels in the core compared to the cover areas of the sandwich cross-section, enable using a core material or structure with low-density.³⁰ This significantly increases material and weight efficiency of the sandwich structure, making those structures interesting for many lightweight design applications.³¹ As discussed, the covering facesheets of sandwich components need to provide high mechanical properties. For this reason, metal and continuous fibre reinforced plastics (CFRP) are commonly applied.³² Regarding the latter, a shift in polymer materials used can be observed. For decades, CFRP facesheets for sandwich structures have been based on thermoset matrices such as epoxy.³² However, composite sandwich structures based on thermoplastic polymers are increasingly investigated, early examples are the works of Åström and Akermo^33–35 in the late 1990s among others. The reasons for this trend are the beneficial properties that thermoplastic polymers offer against thermosets. Brittle mechanical behaviour and limited debonding resistance of thermoset composite facesheets are major challenges.^13,36 Furthermore, the production of thermoset sandwich composites is time- and cost-intensive.^8,9,20 This is because even advanced manufacturing techniques for thermoset composites such as high pressure resin transfer moulding (HP-RTM) or wet compression moulding are directly depending on the curing time of the resin, defining the composite manufacturing cycle and the processing sequence. This makes thermoset sandwich composites less attractive for cost-sensitive high-volume applications. In contrast, sandwich structures with thermoplastic composite facesheets and thermoplastic core, referred to as full thermoplastic sandwich, can be considered superior compared to their thermoset-based counterparts with respect to the aforementioned issues. If facesheets and core are based on the same polymer, processing techniques based on fusion bonding can be applied, which offer the potential to produce sandwich composite structures much more efficiently than with thermoset systems. In addition, this single-polymer sandwich design provides largely improved recyclability properties.^9,10,13

Generally, it is essential for sandwich structures that the facesheets are properly bonded to the core, as bonding of facesheets and core (referred to as skin-core bonding) is fundamental to enable the full mechanical potential of sandwich structures.²⁸ Moreover, loss of skin-core bonding leads to the loss of structural integrity of the whole structure and ultimately to its total failure.^37,38 In order to evaluate and optimize the interfacial bonding behaviour, different testing methods have been developed within the past decades focussing on the interfacial fracture toughness as strain energy release rate (SERR) during crack initiation and propagation. Although there are very promising candidates, no specific test methods for the determination of interfacial fracture toughness of composite sandwich structures have been standardized yet contrary to monolithic composites.³⁹

This is even more true for full thermoplastic sandwich structures, which introduce specific requirements to interfacial bonding testing. As thermoplastic matrices are often more compliant than thermosets and as some exhibit unipolar characteristics, respectively, load application to the specimens may be challenging. To the knowledge of the authors, little to no comprehensive research was performed on the applicability of test methods for the evaluation of skin-core bonding of full thermoplastic sandwich structures. To fill this gap, this paper provides an overview of existing test methods for evaluating interfacial fracture toughness. Based on a definition of requirements for suitable and valid test methodologies, a comprehensive evaluation of the applicability as well as limitations of specific test methods for full thermoplastic sandwich structures is presented.

Requirements for test methods

For an evaluation of test methods for applicability at full thermoplastic sandwich structures, evaluating criteria and requirements need to be established first. In his work, Ratcliffe⁴⁰ defined requirements for a standardized test method enabling the evaluation of Mode I dominant skin-core crack resistance of sandwich structures in general, leading to the Mode I fracture toughness referred to as $G_{I}$ . These requirements are also applied in the present work for the evaluation of test methods and their applicability at full thermoplastic sandwich structures. These requirements are:

(1) Complexity of the test apparatus and procedure must be minimized

(2) The data reduction method used for computing $G_{I}$ should not require detailed stress analyses, such as a finite element calculation. Solutions for specimen compliance should be closed-form.

(3) Specimen loading must be Mode I dominated, and this condition should be insensitive to debonding length

(4) The test must result in debonding growth either at, or in the vicinity of the facesheets/core interface, parallel to the plane of the interface.

(5) Debonding growth must be stable, although stick-slip growth may be permitted

(6) Specimen response must be linear elastic and must be appropriate for analysis using linear elastic fracture mechanics (LEFM).

(7) The test must result in the measurement of quasi-static values of Gc, including an initiation value and subsequent propagation values.

(8) It must be possible to test a practical range of sandwich systems including structures with compliant composite facesheets.

The first requirement attributes to an intended high working efficiency of the test methods. If the specimens and test methods can be prepared and conducted in a quick and reliable way, more tests can be performed and more data can be generated. Moreover, data generation and evaluation during testing such as video assisted crack tip assessment should be kept as simple as possible, as complex test performances are prone to inaccuracies and mistakes. The second requirement addresses the calculation of the desired test methods results. Different data reduction methods have been developed taking into account different material and test specific constraints. In order to reduce data reduction complexity and hence possible inaccuracies, simple analytical approaches are favourable. The next requirements address the process of debonding of sandwich structures. It is general consensus, that the most critical opening mode at the tip of a skin-core crack in sandwich structures is Mode I.^41–44 For this reason, a Mode I dominated test method shall be found (Requirement III). In addition, the crack is required to propagate at or in the very vicinity of the skin-core interface (Requirement IV). Crack propagation shall occur in a stable manner, though the frequently observed slip-stick behaviour of debonding interfaces is also accepted (Requirement V). The structural behaviour during testing should be linear elastic enabling the use of linear elastic fracture mechanics (LEFM) (Requirement VI), which builds the basis for the calculation of the SERR at crack initiation and propagation⁴⁵ (Requirement VII). The last requirement that Ratcliffe⁴⁰ presents is that the test method should enable using sandwich structures with different materials and geometries (Requirement VIII) in order to be able to optimize sandwich structures and to quantitatively compare their skin-core bonding characteristics. A further specification of this last requirement is made here for the specific application of very thin thermoplastic composite facesheets, which may suffer excessive bending during testing due to their low intrinsic flexural rigidity. This may result in fibre kinking and fibre damage of the facesheets, especially in the vicinity of the load introduction area.⁴⁶

After the definition of requirements for evaluation, the most promising candidates for Mode I dominant testing of skin-core interfacial bonding of thermoplastic sandwich structures will be presented and critically discussed in the following segments. The selected group of test methods comprises interfacial fracture toughness test methods as well as promising peel test methods.

Interfacial fracture toughness test methods

The theoretical analysis of the crack propagation along an interface between dissimilar materials has been of great interest for researchers for decades, beginning with the work of Williams.⁴⁷ In this work, he studied the distribution of stress around a crack at the interface between two dissimilar and homogeneous isotropic materials, such as layers of differing rock material or welded joints. This was followed by the analytical calculations of stress intensity factors at dissimilar material interface cracks by Erdogan,⁴⁸ Rice⁴⁹ and England,⁵⁰ all published in 1965. Experimental analysis of interfacial cracks of layered structures and of sandwich structures in particular followed only later. Experimental analysis of the interfacial skin-core bonding of sandwich structures mainly determine the SERR as measure of crack resistance during the debonding of facesheet from core.^40,51 Without claims of completeness, the most common test methods are presented according to the opening Mode:

1. Mode I:

(a) Double cantilever beam (DCB) (ASTM D 5528⁵²),

(b) Single cantilever beam (SCB),

2. Mode II and Mixed Mode:

(a) Tilted sandwich debond (TSD),

(b) Modified three-point bending (MTPB),

(d) End notch flexure (ENF).

The focus of this paper lays on Mode I fracture toughness test procedures, considering Requirement III (Mode I dominance). However, even those test methods providing a global Mode I loading of the interface between core and facesheet may include a Mode II component when applied at sandwich structures. According to Hutchinson and Suo,^53–55 this is due to the asymmetry of material properties of core and facesheets as well as of the loading condition during testing. This incorporation of a Mode II component hence leads to a combination of Modes, which can be expressed as Mode Mixity Phase Angle or Mixed Mode Ratio.⁵⁶

Most Mode I dominant test procedures can be mainly grouped into DCB and SCB test variants.⁴⁰ This paper will hence focus on these test methods. In addition, peel test methods will be evaluated, as they show promising characteristics for the SERR determination of layered material combinations such as sandwich structures. In the following sub-chapters, the DCB and the SCB as well as variants thereof are presented and critically discussed. After these sections, an overview of promising test methods based on peel testing methodologies is given. This main section ends with an overview of a test method specific evaluation regarding their applicability for ski-core bonding of full thermoplastic sandwich structures based on the defined requirements.

Double cantilever beam

One of the most popular examples of Mode I fracture toughness test methods for polymer composites is the double cantilever beam (DCB), as standardized in ASTM D 5528.⁵² The test method originates in the work of Obreimoff⁵⁸ for the evaluation of the splitting strength of mica. The test set-up was modified by Gilman⁵⁹ and is basically unchanged in ASTM D 5528 except for details regarding the load applications. Though originally developed for monolithic material architectures, the DCB set-up was adapted by Prasad and Carlsson⁶⁰ for sandwich structures. Contrary to the standard for monolithic composite structures, sandwich structures do not have the pre-crack in the middle of the cross-section height, but at the interface between core and upper facesheets, see Figure 1. Except for this difference regarding symmetry of specimen load application, the testing procedure is equivalent to the one for monolithic structures. The test method involves a load $P$ that debonds a facesheet from the core of a sandwich specimen with specific length $L$ , width $b$ and thickness $h$ . Before testing, the specimen is equipped with a pre-crack with defined length $a$ at the skin-core interface. The test load $P$ is applied via a piano hinge or connecting blocks, which can be bonded to the test specimen adhesively or mechanically. With increasing load $P$ and load point displacement $δ$ , this results in the initiation and subsequent propagation of the crack.

Figure 1.

The double cantilever beam used for interfacial testing of sandwich structures: asymmetric crack location in thickness direction and load application using piano hinges, based on Ratcliffe.⁴⁰

During testing, the SERR can be assessed applying different data reduction methods.^61,62 Without claims of completeness, they include the modified beam theory (MBT),⁶³ compliance calibration method (CC),^64,65 compliance-based beam method (CBM)⁶¹ and the elastic foundation analysis method (EFA),^66–68 which was originally developed by Kanninen⁶⁹ for isotropic materials and extended to orthotropic materials by Williams.⁷⁰ These data reduction methods however are only valid if the requirements of LEFM are satisfied, including small scale deflections and linear elastic material behaviour. According to Ural et al.⁷¹ who studied skin-core bonding of carbon fibre reinforced PMI sandwich structures with honeycomb cores, the DCB method meets these requirements.

For the calculation of SERR using these data reduction methods, the specimen width $b$ , the applied load $P$ , the load point deflection $δ$ and the crack length $a$ must be determined during testing.⁶² The modified beam theory (MBT), established by Williams,⁶³ yields valid results and due to its simplicity Shivacumar et al.⁶² recommend it for determining the SERR of a skin-core interface at Mode I crack opening $G_{I}$ as

G_{I} = \frac{3 P δ}{2 W a}

(1)

A correction factor $Δ$ should be added to the crack length in order to take account of the elasticity of the core.⁶² The integration of the core as elastic foundation in MBT can be achieved using the effective crack length $a_{e f f}$

a_{e f f} = (a + Δ)

(2)

Here, $Δ$ can be determined by using the cube root of compliance $C^{1 / 3}$ during testing and the corresponding crack length, see Figure 2.

Figure 2.

Determination of the correction factor $Δ$ , based on Avilés and Carlsson.⁶⁶

The correction factor $Δ$ denotes the intersection of the fitted data points with the crack length axis.

C^{\frac{1}{3}} = m (a + Δ)

(3)

Integrating the effective crack length in equation (1) results in

G_{I} = \frac{3 P δ}{2 W (a + Δ)}

(4)

During testing, the applied load and crack length plot shows two distinct stages of crack growth. $G_{I, o n s e t}$ describes the strain energy release rate at crack initiation and $G_{I, p r o p}$ during crack propagation.⁷² Mostly, the SERR value used for evaluation of interfacial fracture toughness is the propagation value.

Despite all efforts to adapt the DCB method from monolithic materials to sandwich structures, challenges still exist. Firstly, the propagating crack during testing leads to an increasing rotation of the specimen around the bottom load application^71,73,74 and bending moments develop in the core area resulting from the eccentric load application.⁴¹ In addition to the test set-up induced difficulties, the results can be influenced by specimen geometry. According to Ural et al.⁷¹ and their studies on sandwich structures with honeycomb cores and carbon fibre reinforced polyimide (PI) facesheets, the determined $G_{I}$ value depends on the thickness of the sandwich facesheet. Furthermore, the sandwich material has an effect on the test results of DCB testing. Shivacumar et al.⁶² studied different data reduction methods for DCB testing of sandwich structures with glass fibre reinforced plastic (GFRP) facesheets and polyvinyl chloride (PVC) foam cores with different density. They found that with increasing foam core density, $G_{I}$ increased until reaching a stable value at a specific density value. In addition to the global core and facesheet material properties, also the local material properties near the skin-core interface have an influence on the fracture behaviour. Bragagnolo et al.⁷⁵ studied the effect of different foam cores on fracture toughness of sandwich structures with carbon fibre reinforced thermoset facesheets. The penetration of facesheets matrix into the core significantly increased crack initiation resistance. However, the crack propagation resistance was not affected.

The most significant negative finding however is that the DCB test method used for sandwich structures involves a non-negligible Mode II fraction.^74,76 This can be attributed to the asymmetric load introduction and material properties.^47,50,54 It is generally agreed that a bi-material interface such as the sandwich skin-core contact leads to the coupling of normal and shear deformations.^{47,48,54,55,57} Similar findings were published by Burlayenko et al.,⁷⁷ who studied the debonding behaviour of sandwich structure with glass fibre reinforced epoxy composite facesheets and PVC foam cores using the DCB and SCB method. Experimental and finite element analysis showed that the SERR were significantly higher using the DCB method, which they attributed to considerable Mode II fractions involved. In addition, they found that shear stresses at the crack tip were likely to lead to crack kinking into the core. This finding is supporting previous publications, in which crack kinking at DCB specimens was observed, which was attributed to a positive shear stress at the crack tip.^60,74,78 He and Hutchinson⁷⁹ described the conditions leading to the kinking of a crack between two brittle materials out of their interface. According to their work, the tendency of a crack kinking into the core is a result of positive shear stresses at the crack tip and the satisfaction of a balance of involved SERR. Based on the studies of Prasad and Carlsson,^60,80 crack kinking into the core of a DCB sandwich specimen is increasingly probable with decreasing Young’s modulus and density of the latter. Hence, for a typical sandwich structure combination of a foam core with significantly lower modulus than the facesheet, this often leads to a crack propagation from the skin-core interface into the core.

Considering these findings and the comparison with the imposed requirements to suitable Mode I sandwich testing procedures based on the definitions by Ratcliffe,⁴⁰ the DCB is considered to be not suitable for Mode I dominated testing of the Mode I dominant fracture toughness resistance of the skin-core interface of full thermoplastic sandwich structures.

Single cantilever beam

A variety of different SCB variants has been developed and studied, including SCB with different specimen fixations as well as introducing a varying peel angle using the Tilted Sandwich Debond (TSD) variant,^81–84 which was first proposed by Grenestedt.⁸⁵ According to extensive studies by Adams and co-workers^44,74,86,87 as well as Ratcliffe and Reeder,^40,41 the SCB variant with the sandwich specimen fixed to a rigid base plate is most promising for determining a Mode I dominant fracture toughness of sandwich skin-core interfaces. For calculation of the SERR, data reduction methods can be applied equivalent to the DCB method.⁸³ In this specific variant, the lower part of the sandwich specimen is fixed to a rigid base plate avoiding bending and rotation, respectively, of the specimen, while the upper facesheet acts as cantilever beam that is peeled off the core via the introduction of a vertical load. However, since during the testing procedure the crack tip moves constantly, the load application is equivalently moving during testing which results in a varying test angle. This would lead to the development of shear stress in the core and a Mode II component at the crack tip depending on the angle at the crack tip.^82–84,88 Consequently, measures are taken to enable a constant location of the crack tip with respect to the load displacement axis of the testing machine during the test. Two approaches are used in literature: the first makes use of a moving base carrier,^37,83,84 whereas the other uses a long rotatable loading rod.^40,41,88

Cantwell and Davies^37,83 developed a single cantilever beam with a sliding carrier and steel base plate on which the specimen can be attached to, see Figure 3. A similar method was used by Glaessgen et al.⁸⁴ The horizontal movement of the base carrier makes sure that the load application only acts as vertical force, as any horizontal load components would lead to a movement of the carrier.

Figure 3.

Single cantilever beam with sliding carriage base plate avoids the variation of the peel angle, based on Cantwell.⁸³

Another option to achieve nearly vertical load application throughout the testing is to use a long road connecting the specimen with the testing machine. McGarva and Åström⁸⁸ included a long rotatable rod that connects the testing machine and the hinge mounted on the facesheet for interfacial testing of sandwich structures with glass fibre reinforced polyamide 12 (PA12) facesheets and PMI foam core. Though they call their testing procedure modified DCB, it is actually a SCB method with specimens fixed to a rigid base plate. According to their test set-up, a rotatable long rod avoids the Mode II component and shear development in the core during testing, which was similarly proposed by Ratcliffe and Reeder in later publications.^40,41

According to studies by Ratcliffe and Reeder⁴¹ as well as Adams and co-workers,^39,86,87 the use of a long rotatable loading rod leads to similar results compared to using a sliding base carrier while offering a much simpler test apparatus. They propose the SCB with long rod for application and standardization, respectively, in interfacial bonding testing and assessment of Mode I interfacial fracture toughness of sandwich structures, see Figure 4. There are several reasons for this. First of all, the SCB used for sandwich structures enables a significantly more Mode I dominant failure compared to DCB testing.^73,77 Weaver⁷⁶ as well as Adams and co-workers^{39,44,74,86,87} published studies in which different test methods for interfacial fracture toughness testing of sandwich structures are directly compared with respect to the Mode I dominance and the crack kinking phenomenon. In a first study, they compared the DCB, a modified DCB, the Three Point Flexure (TPF), the SCB with cantilever beam support and the SCB with fixed base plate. According to their studies, the SCB with fixed base plate is Mode I dominant with minor fraction of Mode II due to shear stress at the crack tip leading to a Mode I fraction of more than 95%. Also, the Mode I dominance was found independent of the crack length and almost independent of facesheet thickness.^74,87 Similar results indicating a clear Mode I dominance using the SCB method were found by Burlayenko et al.,⁷⁷ who studied the Mode I dominance of DCB and SCB test methods for sandwich structures with different material and geometry parameters. According to their results, the SCB is clearly more suitable to enable a Mode I dominated failure during testing compared to DCB. In addition to its Mode I dominance, the SCB method exhibits only a minimal tendency of crack kinking into the core, which was found by several authors.^39,78,83 Burlayenko et al.⁷⁷ even established a negative shear stress at the crack tip using the SCB method, which would result in a crack propagation towards the actual skin-core interface or even to the facesheet. In order to provide consistent and reproducible $G_{I}$ values using the SCB test method for sandwich composites, Ratcliffe and Reeder⁴¹ proposed specific constraints regarding the specimen and the testing procedure. As a result of these efforts and the positive findings, the SCB test method is in discussion to become a standard. The method was drafted as ASTM standard in 2014.⁸⁹ Following this, the SCB test method was conducted by several institutes around the globe in a large round-robin test leading to results contributing to a possible standardization of the method in the future.⁹⁰

Figure 4.

SCB test method variant with long rotatable rod, based on Ratcliffe and Reeder.⁴¹

Despite the discussed positive characteristics of the SCB with fixed base plate, some issues still exist. Variations of sandwich materials may influence the results of the calculated $G_{I}$ values using the SCB test method for sandwich structures. Maleki and Toygar⁷⁸ found that the density of the core in GFRP sandwich structures with PVC foam cores has an effect on the determined SERR results. According to their findings, sandwich structures with cores of higher density exhibit higher SERR using the SCB method. It must be noted that their SCB test set-up did use the fixed base plate design yet missing the long rod for vertical peel load application during crack growth. The thereby incorporated Mode II fraction during testing may make the results more vulnerable to the development of shear stress in the core and to a different material behaviour of the core. Findings of Viana and Carlsson⁹¹ may support this, as they also found increasing SERR with higher density foam core sandwich structures using the TSD set-up, which includes a Mode II component by design. Despite the low Mode II fraction as reported by Weaver⁷⁶ as well as Adams and co-workers^74,87 of less than 5%, Ratcliffe and Reeder⁴¹ as well as McGarva and Åström⁸⁸ consider the Mode II fraction non-negligible and use $G_{C}$ instead of $G_{I}$ . In addition, the structural layout of sandwich structures may affect the calculated SERR. Burlayenko et al.⁷⁷ found that the dominance of Mode I using this SCB variant depends on the thickness of the sandwich facesheets. Similar findings were reported by Adams and co-workers,^39,44 who studied the effect of using adhesive doublers to avoid excessive bending of thin and flexible composite facesheets during SCB testing of honeycomb sandwich structures. With increasing doubler and hence total facesheet thickness the SERR significantly increased, which was attributed to a change of crack growth location into the core. This makes it impossible to quantify the skin-core bonding of sandwich structures with very thin facesheets, which is the last requirement for suitable test methods for full thermoplastic sandwich structures.

Peel tests

Other test methods for the assessment of interfacial bonding of sandwich structures originate in the determination of the peel strength of metal-metal bonds, which was first proposed by Spies.⁹² In addition to metal-metal bonds, a large variety of other material combinations has been developed that require interfacial bonding evaluation. According to Moore,⁹³ bonded material combinations can be grouped according to their principal mechanical behaviour into:

(1) Flexible-flexible,

(2) Flexible-rigid,

(3) Rigid-rigid.

Rigid-rigid material combinations include structures of welded thermoplastic composites,^94–96 hybrid structures consisting of a composite reinforced metal substrate^97–101 as well as adhesively bonded metals.^102,103 Flexible-rigid combinations include the combination of an adhesive on a rigid, typically metal substrate¹⁰⁴ or a thin layer on a thicker substrate, such as coatings^105–107 or printed circuit boards.¹⁰⁸ Flexible-flexible material combinations typically include packaging and multi-layer foils, respectively.^109,110

For the testing of interfacial bonding of these material combinations, a large variety of peel tests and variants has been developed in the last decades. Standardized and widely adopted test methods include:

(1) Floating Roller Peel (FRP) (ISO 4578,¹¹¹ DIN EN 1464,¹¹² ASTM D 3167¹¹³),

(2) 90° Peel Test (ASTM D 6862,¹¹⁴ ASTM D 3330,¹¹⁵ DIN EN 28,510–1¹¹⁶),

(3) 180° Peel Test (ASTM D 3330,¹¹⁵ DIN EN ISO 8510–2¹¹⁷),

(4) T-Peel Test (ASTM D 1876,¹¹⁸ DIN 55,529¹¹⁹),

(5) Climbing Drum Peel Test (CDP) (ASTM D 1781¹²⁰),

(6) Mandrel Peel Test (MPT).

These tests usually evaluate the force needed to peel the top adherent off the substrate and put it into relation with the interface contact area resulting in a peel strength. The peel strength however can be translated into a fracture toughness based on an energy balance analysis of the respective peel method, if the dissipative energy fractions can be validly considered in the energy balance.¹²¹ This makes them promising for possible use as Mode I dominant test method for thermoplastic sandwich structures. Nevertheless, some of the peel tests do not directly control the bending radius of the bending arm during peel testing, see Figure 5. This does not comply with the last requirement for suitable interfacial bonding test methods for composite sandwich structures with thin facesheets. Using these test methods, invalid failure due to the introduction of fibre damage may occur in the bending arm, that is, the composite facesheet of the sandwich structure.¹²²

Figure 5.

Free peel test (left) and roller-assisted peel test (right), based on Kawashita et al.¹²⁴

For this reason, only peel test with means to limit and control the bending radius will be considered in this paper for applicability at full thermoplastic sandwich structures, that is, the climbing drum, floating roller and the mandrel peel test.

Climbing drum test

The climbing drum peel test (CDP) was first standardized as ASTM D 1781¹²⁰ in the 1960s. The test method is often applied for testing the delamination resistance and skin-core bonding of sandwich structures, respectively,^124–128 also for thermoplastic sandwich structures.¹²⁹ The test procedure begins with the fixation of the upper sandwich facesheet to a drum, see Figure 6. This drum exhibits two sections with different radii. The facesheet is attached to the inner section with radius $R_{1}$ , while straps are attached to the other drum section with radius $R_{2} > R_{1}$ . The straps as well as the sandwich specimen are fixed in opposite sides on a uni-axial testing machine, which introduces the torque via the straps to the drum inducing its rotation. The rotation of the drum along the sandwich specimen surface leads to the debonding of the facesheet off the core in a stable manner at constant load.¹³⁰

Figure 6.

Climbing drum peel test principle, based on Nettles et al.¹²⁴

In its standard version, the climbing drum method yields a peel strength, that is, a force per width, and only enables a quantitative comparison of test results if the geometry is unchanged.¹³⁰ A study by Nettles et al.¹²⁴ however presents a method to determine the SERR using the CDP based on the work of Okada and Kortschot.¹³¹ Due to the kinematic constraints, the crack length during testing $D_{1}$ is directly dependent on the crosshead displacement $δ$ and the different drum radii $R_{1}$ and $R_{2}$ , leading to

D_{1} = \frac{R_{1} δ}{R_{2} - R_{1}}

(5)

During debonding, two load stages can be observed.^124,130 First, the facesheet is pulled around the centre drum with radius $R_{1}$ utilizing the pre-rack hence without skin-core debonding to take place; the associated load is $P_{1}$ . This is followed by the initiation of the debonding process leading to a stable load introduction $P_{2}$ .

The grey area in Figure 7 denotes the work for skin-core debonding during the climbing drum peel test. Nettles et al.¹²⁴ translated the work for peeling into a critical strain energy release rate $G_{I C}$ by relating it to the peeled surface with the specimen width $b$ and crack length $D_{1}$ , leading to

G_{I C} = \frac{(P_{2} - P_{1}) δ}{b D_{1}}

(6)

Figure 7.

Schematic of experimental force-displacement plot, based on Daghia and Cluzel.¹³⁰

Using the climbing drum peel test and the associated assessment method of a SERR, they evaluated experimentally the debonding process of carbon fibre reinforced epoxy facesheets on honeycomb cores. They compared the results with those obtained from DCB testing and found them to yield similar $G_{I}$ values. The CDP test outperformed the DCB test however with respect to the ease of conduct.

In addition, comparisons of the CDP test with the SCB method were conducted. Adams et al.⁴⁴ compared the CDP test with the SCB with fixed base plate using sandwich structures with honeycomb core and varying composite facesheets. Results showed that while the SERR varies with different facesheet thicknesses using the SCB, the climbing drum peel test yields more thickness-independent SERR values. However, using the climbing drum method and changing the roller diameter resulted in different SERR values, according to their study. This may be due to the conformity issue of roller peel tests without predefined alignment of the bending arm to the roller. The calculation method for the SERR using the CDP test prerequisites perfect conformity of the bending arm to the drum, which may not always be ensured.¹²⁴ Conformity of the facesheets to the drum can be improved, if the weight of the drum itself is increased, as Daghia and Cluzel¹³⁰ propose. In their study, they evaluate the use of the CDP for $G_{I}$ testing monolithic composite structures with varying layup and compare the results with DCB method. In accordance with the results of Nettles et al.¹²⁴ for sandwich structures, they find similar $G_{I}$ values for both CDP and DCB methods.

However, limited research exists specifically on the Mode I dominance of the CDP test. The discussed study of Adams et al.⁴⁴ indicates a Mode I dominant behaviour, as it yields similar SERR values as using the SCB variant, which they showed to be clear Mode I dominant. Similar findings were published by Daghia and Cluzel¹³⁰ who consider the CDP peel test to be Mode I dominant. According to Cañas et al.,¹³² who performed experimental and finite element studies on the CDP and a new proposed horizontal climbing drum variant, the resulting SERR is Mode I dominant but dependent on the material parameters.

In addition, the crack kinking phenomenon typical for DCB testing of sandwich structures is less distinct using the CDP test method. According to studies of Adams et al.,⁴⁴ the crack location is at or near the interface between Nomex honeycomb core and carbon fibre reinforced epoxy facesheet, independently of the facesheet thickness. Nettles et al.¹²⁴ finds similar results with phenolic honeycomb cores and carbon fibre reinforced epoxy facesheets. However, they also found that the SERR during debonding depends on the material parameters, if the skin-core bonding is good. This is because the crack tends to propagate into locations with less crack resistance than the interface, which may be the core or the facesheet. In case of sandwich structures with tough skin-core interface and low-density honeycomb cores, they observed a tendency of the crack to kink into the less resistive core material. On the other hand, using higher density honeycomb cores, the crack was found to kink into the facesheets. Consequently, they conclude that the skin-core interface must be the weak link in order to make sure, that the crack propagates at this location. This is a general issue when it comes to interfacial bonding testing and is addressed by several authors.^101,123 Freitas et al.¹⁰¹ even state that ‘The most important result in order to evaluate the peel performances of adhesive joints using peel tests is the failure mode’. While they conduct an evaluation of the crack location qualitatively, also a quantitative evaluation can be conducted. Kawashita et al.,¹²³ for example, propose a subjective approach by introducing the fracture index based on optical micrographs of the fracture surfaces.

Floating roller peel test

The floating roller peel (FRP) test method is another popular peel strength method. The test is standardized in ASTM D 3167¹¹³ and typically used for measuring the debonding resistance of flexible-rigid interfaces.¹⁰⁶ A specimen with a rigid substrate bonded with a flexible adherent is pulled through a set-up of two rollers with equal radius $R_{1}$ , see Figure 8. One of these rollers is the peel roller, over which the flexile adherent is peeled off the substrate by applying a load $P$ to the flexible adherent using a uni-axial testing machine. A certain but uncontrolled conformity of the flexible adherent is achieved by the fixation of the floating roller set-up to the testing machine leading to an alignment force $D$ . The test result is a peel strength, which relates the applied peel load to the width of the adherent, similar to the CDP method.

Figure 8.

Floating roller test set-up, based on Kawashita et al.¹²³

This test method was applied for the assessment of interfacial bonding of a carbon fibre reinforced polyamide 6 (PA6) UD tape on a steel substrate by Reincke et al.¹⁰⁰ However, they did not calculate a SERR but a peel strength during debonding. Similarly, Freitas and Sinke¹⁰¹ used the floating roller method to evaluate the interfacial bonding of different combinations of Aluminium and carbon fibre reinforced epoxy laminates with different layups. They indicate, that this test method only allows a comparison of bonds with the exact same flexible adherent as this has a major effect on the determined peel strengths. In addition, the peel strength should only be used together with an evaluation of the crack mode, that is, cohesive or adhesive crack.

As indicated in Requirement II, it is necessary that suitable test methods yield a Mode I dominant SERR value that can be compared quantitatively with other sandwich designs and research data, respectively. In order to assess a $G_{I}$ value using peel tests such as the floating roller, different approaches exist. Generally, the total energy during peel testing $G$ includes the actual energy for debonding $G_{a}$ and additional dissipative fractions $G_{p}$ depending on the material of the flexible adherent.^{102,106,121,133}

G = G_{a} + G_{p}

(7)

Consequently, it is necessary to assess the plastic work fraction $G_{p}$ during peel tests in order to receive the true debonding work $G_{a}$ . Assuming a perfect conformity of the strip to the mandrel, Sexsmith et al.¹³⁴ calculate the work of plastic deformation during roller-assisted peel testing of metallic foils in order to determine the actual work to separate two structures during the mandrel peel test. They also showed that the plastic energy for bending $G_{p}$ is significant and must be considered for calculation of $G_{a}$ . However, this approach is only valid if the bending arm conforms to the roller. Also, conformity of the bending arm to the roller cannot be assured before testing, which makes it even more challenging.¹²³ In addition to the work of Sexsmith et al.,¹³⁴ a large variety of methods to calculate the energy fraction of the flexible adherent during peel tests have been developed for various kinds of materials enabling the assessment of the true interfacial fracture toughness. Comprehensive literature on this topic can be found in an introduction to peel test methods by Moore.⁹³ However, the valid calculation of the structural material behaviour of the bending arm increases complexity of the test procedure, since these methodologies need to be validated based on experimental campaigns. Adding to the issue of the valid assessment of plastic work of the bending arm, the Mode I dominance of FRP is unclear, although Arouche et al.¹³⁵ found fracture characteristics typical for Mode I based on SEM analysis of fracture surfaces of carbon epoxy laminates after floating roller peel testing. Consequently, it can be concluded that the floating roller peel test is considered complex, prone to inaccuracies and not well suited for a quantitative comparison of Mode I dominant fracture toughness of full thermoplastic sandwich interfaces.

Mandrel peel test

As discussed, the conformity of the bending arm during peel tests is a critical issue. Based on the 90° peel test, Sexsmith and Troczynski¹⁰⁶ added a mandrel in order to improve control of the peel angle. They later modified the novel mandrel peel test with an alignment force, pulling the peeled-off strip at the end of the specimen, see Figure 9. This enables a direct control of the conformity of the adherent to the mandrel,¹⁰⁷ which distinctly separates it from the floating roller and the climbing drum test.

Figure 9.

Mandrel peel test set-up according to Sexsmith and Troczynski.¹⁰⁷

Another significant advantage is the direct experimental assessment of the dissipative energy component $G_{p}$ during testing, first proposed by Breslauer and Troczynski,¹⁰² avoiding complex calculations as usually required using other peel tests. Based on an energy balance of the roller-assisted peel test, the total energy during mandrel peel testing is the sum of the true debonding energy $G_{a}$ , the energy losses due to the friction coefficient and the alignment force $G_{p}$ as well as the energy for the bending of the flexible adherent $G_{b}$

G = G_{a} + G_{p} + G_{b}

(8)

Translating the energy into a force balance analysis and neglecting the elastic energy stored in the flexible adherent, Breslauer and Troczynski¹⁰² calculate the relationship between the interfacial fracture toughness $G_{a}$ , the plastic energy dissipation $G_{p}$ , the alignment force $D$ , the applied peel force $P$ and the friction coefficient $μ$ as

P (1 - μ) = D + b (G_{a} + G_{p})

(9)

Based on equation (9), the true fracture toughness $G_{a}$ of the interface can be calculated, if the plastic energy dissipation of the peel arm $G_{p}$ and the friction of the system is known. In contrast to other peel tests, these values can be experimentally assessed by plotting the width-specific peel load $P / b$ against the width-specific alignment load $D / b$ . For further details regarding this procedure and the corresponding $P / b$ – $D / b$ plots, it is referred to the work of Breslauer and Troczynski.¹⁰² The friction as well as the plastic energy during the bending of the flexible adherent can be assessed by a calibration test using an unbonded interface specimen. In a subsequent test using a bonded specimen, the true energy for debonding its interface can be determined using the $P / b$ – $D / b$ plot.

Following this work, other researchers used this mandrel peel method for the evaluation of interfacial fracture toughness of different material combinations. Kawashita and co-workers^{103,123,133,136} adapted the set-up of this method using a moving base carrier and applied it on the determination of interfacial fracture toughness of structural metal-polymer hybrid laminates.

A comparison of different peel tests suggests that the adapted mandrel peel test is very well suited for the valid experimental assessment of interfacial fracture toughness of metal-polymer hybrid structures and benefits from its ‘elegant simplicity’.^123,136 Following the work of Kawashita and co-workers, Grouve and co-workers^122,137 evaluated the use of the mandrel peel test for thermoplastic composite-composite interfaces manufactured using automated thermoplastic tape placement, see Figure 10. Thermoplastic tape placement of unidirectional fibre reinforced thermoplastic tapes (UD-tapes) involves thin UD-tapes being continuously placed and welded in one process step according to the desired layup. Grouve et al.¹²² calculated the SERR during debonding of these interfaces using an energy balance taking into account residual stresses in the bending arm, which may be induced due to the tape placement process and the involved thermal history. It was found that for the UD-tapes used the plastic work during bending and also residual stresses can be neglected.^122,137 This leads to the calculation the interfacial fracture toughness $G$ of a thermoplastic composite interface¹²²

G = \frac{1}{b} [P (1 - μ) - D]

(10)

Figure 10.

Mandrel peel test set-up according to Grouve et al.¹²²

This approach to determine the true SERR during debonding is equivalent to the Breslauer and Troczynski¹⁰² approach (equation (9)), considering the neglect of plastic work $G_{p}$ . Using this set-up and equation, it is only necessary to use a calibration test in order to determine the friction coefficient $μ$ during the test. The true fracture toughness for debonding can be calculated based on the forces involved during testing, that is, the applied peeling load and the alignment load.

In addition to the testing of the interfacial bonding of UD carbon laminate with polyphenylene sulphide (PPS) matrix and UD tape reinforced carbon-PPS woven laminate,¹²² the mandrel peel test was used for the assessment of the SERR of the debonding interface of various other combinations of thermoplastic composite interfaces^94,95 as well as thermoplastic composite metal interfaces.^97,98 In addition, interfacial bonding of full thermoplastic composite sandwich structures was evaluated using the mandrel peel test set-up.¹⁸ Based on the results of these studies, the test procedure provides a high reproducibility^94,136 and mostly avoids kinking of the crack into one of the thermoplastic composite adherents.^18,122 Kok et al.⁹⁵ however find that with high interfacial toughness between carbon fibre reinforced PEEK tape adherents, the crack may propagate into in the flexible adherent as the fibre matrix adhesion may exhibit lower crack resistances than the interface. In addition, they find that this is depending on the fibre orientation. In accordance with the findings of Kawashita and co-workers,^123,133 they indicate the necessity not only to focus on fracture toughness test results but also on the failure mode.

Regarding the Mode I dominance of the mandrel peel test, limited research is available, yet it indicates Mode I dominant crack propagation. Sacchetti et al.⁹⁴ find Mode I dominant crack propagation during debonding of woven carbon fibre reinforced PEEK composites, based on a comparison of SERR using DCB and the mandrel peel test. Both test methods yield similar results, although the mandrel peel test leads to a little lower SERR values. Also, Grouve et al.¹²² found that the SERR during debonding of UD-UD as well as UD-woven carbon fibre reinforced PPS interfaces was about 20 % lower than using the DCB method, which is partly attributed to decreased fibre bridging using the mandrel peel test compared to DCB. Another reason may be due to a more Mode I dominant behaviour at the crack tip using the mandrel peel method, which was shown to lead to lower SERR compared to Mixed Mode behaviour in case of DCB and SCB test methods.⁷⁷ In addition, SEM analysis of the fracture surfaces of DCB specimens and mandrel peel specimens showed similar fracture characteristics exhibiting signs of Mode I crack behaviour.^94,122

Direct comparison of mode I dominant interfacial fracture toughness test methods

The specific characteristics and the evaluation results of these test methods with respect to the requirements are summarized in Table 1. With respect to the defined requirements for suitable interfacial bonding test methods, the mandrel peel test method yields best results compared with all other methods included in this review. Consequently, the mandrel peel test is considered a very promising test method for the assessment of Mode I dominant interfacial fracture toughness of the skin-core interface of thermoplastic sandwich structures, especially with compliant facesheets.

Table 1.

Evaluation of Mode I dominant test methods for the assessment of skin-core bonding of thermoplastic sandwich structures: ++ ‘very good’, + ‘good’, o ‘neutral’, - ‘fair’ and -- ‘poor’.

	Test method
Requirements	DCB	SCB	CDP	FRP	MPT
I – Low test complexity	−⁵²	−^40,41	o^120,124	o^113,123,138	++^{103,123,124,137,138}
II – Simple data reduction method	−⁵²	−^40,41	+^124,131	−¹²³	++^{102,122,123,136,137}
III – Mode I dominant	--^75,77,78	+^{40,41,44,74,86–88}	+^44,130,132	o⁹⁹	+^94,122
IV – Crack position at the interface	--^74,76,77	++^39,77,78,83	++^44,124	−^99,101,137	+^{18,95,122,123,133}
V – Stable crack growth	+	+^37,83	+^124,130	+^123,138	+^94,95
VI – Application of LEFM valid	+^52,71	++^40,41	+	+	+
VII – SERR as test result for initiation and propagation	++⁵²	++^40,41	o^124,131	+^{101,102,106,121,123,133,138}	++^{46,94,95,122,139,140}
VIII – Enabling variations of specimen geometry and material	−^62,71	+^41,74,78,87	o^44,132	−^{101,102,106,121,133}	+^94,136

However, it is generally recommended that the determined $G_{I}$ value should be complemented by an analysis of the crack location. This is because minor deviations of the crack location away from the exact interface may be possible. If this is the case, the quantitative test result, that is, $G_{I}$ does not directly reflect the crack resistance of the interface itself. As the crack will follow a path with lowest resistance, it only indicates that it is higher than the crack resistance of the core and the facesheet material, respectively. Hence, the test results should include both the $G_{I}$ value but also the exact crack location and type of fracture, even when using the mandrel peel test or the SCB test method.

Conclusion

The present work shall contribute to select and further on to establish a suitable test method for the assessment of Mode I interfacial fracture toughness of thermoplastic sandwich structures. For this reason, promising test procedures are reviewed and evaluated in this paper. The considered test methods are the DCB, the SCB and different variants of peel tests. To evaluate the applicability of these test methods, eight requirements for suitable skin-core bonding test methods for thermoplastic sandwich structures are defined.

In this review it is found that the DCB test method used for sandwich structures exhibits a certain Mode II fraction, which is mainly attributed to the asymmetric load introduction in addition to the bi-material effect of sandwich structures resulting from different properties of skin and core. In addition, the DCB method shows tendencies of crack kinking. This characteristic for the DCB method used for sandwich structures is attributed to a positive shear stress at the crack tip. Consequently, the DCB method does not comply with the requirements for suitable test methods for skin-core bonding of sandwich structures.

In order to avoid the crack kinking phenomenon, several researchers propose the SCB as test method for interfacial fracture toughness assessment of sandwich structures. The SCB method exhibits only a minimal tendency of crack kinking into the core and distinctly less compared to the DCB method. Also, the major issue of specimen rotation is avoided as the specimen is fixed on a rigid substrate. According to literature, the SCB method leads to a clear Mode I dominant failure at the skin-core interface of sandwich structures. Also, the Mode I dominance is almost unaffected by crack length and facesheet thickness. However, the SCB test method does not allow valid testing of composite sandwich structures with very thin and compliant facesheets, respectively, since excessive bending of the facesheets near the crack tip cannot be avoided leading to facesheet failure and invalid results.

For this reason, peel tests appear promising since some of them control the bending radius of the facesheets by design. Peel tests considered in this paper are the climbing drum, the floating roller and the mandrel peel test. Recently published data reduction methods enable the determination of interfacial fracture toughness using the CDP. Authors showed that the SERR are similar to those using the SCB method and that they are independent of the facesheets thickness. However, the SERR values may differ depending on the radius of the drum. The use the latter avoids excessive bending of composite facesheets, yet perfect conformity cannot be guaranteed a priori. This is also found for the floating roller test method, which is typically used for flexible-rigid material interfaces. Due to the insufficient conformity, testing results are prone to inaccuracies, as calculations of the dissipative energy fractions during testing involved by the adherent material behaviour are required. In order to avoid these complex calculations for the assessment of interfacial fracture toughness using peel tests, the mandrel peel test was proposed by different authors. According to their studies, the method enables the direct experimental assessment of the SERR of a debonding interface. Available literature suggests that the method is reproducible and Mode I dominant. Also, it avoids crack kinking into the core. Considering the defined requirements for suitable Mode I dominant test methods for thermoplastic sandwich structures, the mandrel peel test yields most promising overall characteristics of all methods considered in this work.

In future studies, experimental campaigns analysing the skin-core bonding of full thermoplastic sandwich structures should be conducted using the SCB and the mandrel peel test in order to provide a direct comparison of the SERR results as well as of the specific skin-core fracture surfaces. Also, the applicability of these methods for thermoplastic sandwich structures with different core structures and facesheets should be assessed.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the European Union’s Horizon 2020 research and innovation programme (grant number 770019).

ORCID iD

Felix Weidmann

Note

References

European Parliament . New Passenger Cars and for New Light Commercial Vehicles, and Repealing Regulations (EC) No 443/2009 and (EU) No 510/2011 (Recast) (Text with EEA Relevance), 2019. Regulation (EU) 2019/631 of the European Parliament and of the Council of 17 April 2019 setting CO2 emission performance standards for.

Cetin

Thienel

. Large-series production of thermoplastic door module carriers. Lightweight Des Worldwide 2019; 12: 12–17.

Häffelin

Wagener

Beutler

, et al. Lightweight Design on the Move: All-Plastic Brake Pedal with Multiaxial Fiber Reinforcement in Series Production. Kunststoffe, 2017.

Siebel

. All-Plastic Brake Pedal Suitable for Large-Scale Production. Online, 2020.

Nuyan

Constantinou

. Lightweight seat with plastic backrest. ATZ Worldwide 2014; 116: 16–19.

Weidmann

. Insulating sandwich housing structures for the thermal management of battery packs. Twelfth International Conference on Ecological Vehicles and Renewable Energies. EVER, 2017.

Grünewald

Orth

Parlevliet

, et al. Modified foam cores for full thermoplastic composite sandwich structures. Jnl Sandwich Struc Mater 2017; 6: 109963621770874.

Grünewald

Parlevliet

Altstädt

. Manufacturing of thermoplastic composite sandwich structures. J Thermoplastic Compos Mater 2016; 30: 437–464.

Menrath

Henning

Huber

, et al. Foam-injected sandwich panels with continuous-reinforced facings. In: proceedings of PPS-29: The 29th International Conference of the Polymer Processing Society-Conference Papers, Nuremberg, Germany. American Institute of Physics; 2013, pp. 477–481.

10.

Corvaglia

Passaro

Manni

, et al. Recycling of PP-based sandwich panels with continuous fiber composite skins. J Thermoplastic Compos Mater 2006; 19: 731–745.

11.

Neumeyer

Knoechel

Mühlbacher

, et al. Thermoplastic Sandwich Structures-Processing Approaches towards Automotive Serial Production, 2017.

12.

Åkermo

Åström

. Modelling component cost in compression moulding of thermoplastic composite and sandwich components. Composites A: Appl Sci Manufac 2000; 31: 319–333.

13.

Skawinski

Binetruy

Krawczak

, et al. All-thermoplastic composite sandwich panels - Part I: Manufacturing and improvement of surface quality. J Sandwich Structures Mater 2016; 6: 399–421.

14.

Åkermo

Åström

. Experimental investigation of compression molding of glass/PP-PP foam core sandwich components. J Thermoplastic Compos Mater 2016; 12: 297–316.

15.

Chen

Xue

Zhou

, et al. Preparation and performance of continuous glass fiber reinforced polypropylene composite honeycomb sandwich panels. MATEC Web Conf 2016; 67: 6082.

16.

Mcgarva

Åström

. Insert integration in thermoplastic-based foam core sandwich components. J Sandwich Structures Mater 2016; 1: 214–229.

17.

Loypetch

Tröltzsch

Nestler

, et al. Influence of talc in polypropylene foam cores of sandwich structures with skins made of thermoplastic prepregs. IOP Conf Ser Mater Sci Eng 2019; 480: 12012.

18.

Weidmann

. On the in-Situ Manufacture of Thermoplastic Sandwich Structures with Continuous Fibre Reinforced Facesheets and integral Foam Cores. Dissertation, TU Clausthal. Clauthal, 2020.

19.

Passaro

Corvaglia

Manni

, et al. Processing-properties relationship of sandwich panels with polypropylene-core and polypropylene-matrix composite skins. Polym Composites 2004; 25: 307–318.

20.

McGarva

. Thermoplastic Composite Sandwich Components: Experimental and Numerical Investigation of Manufacturing Issues. Stockholm: Dissertation, 2002.

21.

Pappadà

Rametta

Passaro

, et al. Processing, mechanical properties, and interfacial bonding of a thermoplastic core-foam/composite-skin sandwich panel. Adv Polym Technol 2010; 29: 137–145.

22.

Sewell

Czarnecki

Pflug

. Continuous thermoplastic honeycomb sandwich panel process technology for cost optimal lightweight composites. Reinforced Plastics 2016; 60: 146–150.

23.

Grünewald

. Thermoplastic Composite Sandwiches for Structural Helicopter Applications. Bayreuth: Universität Bayreuth, 2018.

24.

Skawinski

Binetruy

Krawczak

, et al. All-thermoplastic composite sandwich panels-Part II: modelling of bending behaviour. J Sandwich Structures Mater 2016; 6: 423–446.

25.

Ward

. Analysis of Sandwich Structures. In: Miracle

Donaldson

(eds). Composites. ASM International; 2001, p. 0.

26.

Hara

Özgen

. Investigation of weight reduction of automotive body structures with the use of sandwich materials. Transportation Res Proced 2016; 14: 1013–1020.

27.

Steeves

Fleck

. Collapse mechanisms of sandwich beams with composite faces and a foam core, loaded in three-point bending. Part I: Analytical models and minimum weight design. Int J Mech Sci 2004; 46: 561–583.

28.

Wiedemann

. Leichtbau: Elemente und Konstruktion. 3. Auflage. Berlin, Heidelberg: Springer-Verlag Berlin Heidelberg, 2007.

29.

Allen

. Analysis and Design of Structural Sandwich Panels. Elsevier, 1969.

30.

Yan

Kortschot

. Light-weight honeycomb core sandwich panels containing biofiber-reinforced thermoset polymer composite skins: Fabrication and evaluation. Composites B: Eng 2012; 43: 2875–2882.

31.

Reyes

Rangaraj

. Fracture properties of high performance carbon foam sandwich structures. Composites Part A: Appl Sci Manufacturing 2011; 42: 1–7.

32.

Karlsson

TomasÅström

. Manufacturing and applications of structural sandwich components. Composites Part A: Appl Sci Manufacturing 1997; 28: 97–111.

33.

Åström

Åkermo

Carlsson

, et al. (eds). All-thermoplastic sandwich concept; 1998.

34.

Åkermo

Åström

. Modeling compression molding of all-thermoplastic honeycomb core sandwich components. Part B: Model verification. Polym Compos 2000; 21: 257–267.

35.

Åkermo

Åström

. Modeling compression molding of all-thermoplastic honeycomb core sandwich components. Part A: Model development. Polym Compos 2000; 21: 245–256.

36.

Vieille

Casado

Bouvet

. Comparative study on the impact behavior and damage tolerance of woven carbon fiber reinforced thermoplastic– and thermosetting– composites. In: ECCM15 - 15th European conference on composite materials. Venice, Italy. 2012, pp. 24–28.

37.

Cantwell

Davies

. A study of skin-core adhesion in glass fibre reinforced sandwich materials. Appl Compos Mater 1996; 3: 407–420.

38.

Carlsson

. Fracture mechanics analysis of tilted sandwich debond (TSD) specimen. J Compos Mater 2001; 35: 2145–2168.

39.

Adams

Nelson

Bluth

(eds). Development and Evaluation of Fracture Mechanics Test Methods for Sandwich Composites; 2012.

40.

Ratcliffe

. Sizing a Single Cantilever Beam Specimen for Characterizing Facesheet–Core Debonding in Sandwich Structure. NASA/TP–2010-216169. National Institute of Aerospace, Langley Research Center, 2010.

41.

Ratcliffe

Reeder

. Sizing a single cantilever beam specimen for characterizing fact sheets-core debonding in sandwich structure. J Compos Mater 2011; 45: 2669–2684.

42.

Müller

Herák

. Dimensioning of the bonded lap joint. Res Agr Eng 2010; 56: 59–68.

43.

Farmand-Ashtiani

Cugnoni

Botsis

. Monitoring and characterization of the interfacial fracture in sandwich composites with embedded multiplexed optical sensors. Compos Structures 2013; 96: 476–483.

44.

Adams

Bluth

Braegger

. Development and Evaluation of Fracture Mechanics Test Methods for Sandwich Composites. Utah, 2013.

45.

Gross

Seelig

. Bruchmechanik: Mit einer Einführung in die Mikromechanik ; mit 8 Tabellen. 4., bearb. Aufl. Berlin: Springer, 2007.

46.

Grouve

WJB

. Weld Strength of Laser-Assisted Tape-Placed Thermoplastic Composites. Dissertation, University of Twente. Enschede, 2012.

47.

Williams

. The stresses around a fault or crack in dissimilar media. Bull Seismological Soc America 1959; 49: 199–204.

48.

Erdogan

. Stress distribution in bonded dissimilar materials with cracks. J Appl Mech 1965; 32: 403–410.

49.

Rice

Sih

. Plane problems of cracks in dissimilar media. J Appl Mech 1965; 32: 418–423.

50.

England

. A crack between dissimilar media. J Appl Mech 1965; 32: 400–402.

51.

Cantwell

. Interfacial fracture in sandwich laminates. Composites Sci Techn 1999; 59: 2079–2085.

52.

ASTM D5528 . Standard Test Method for Mode I Interlaminar Fracture Toughness of Unidirectional Fiber-Reinforced Polymer Matrix Composites. West Conshohocken, PA: ASTM International, 2022.

53.

Hutchinson

. Mixed mode fracture mechanics of interfaces. Metal-ceramic Interfaces 1990; 4: 295–306.

54.

Suo

Hutchinson

. Sandwich test specimens for measuring interface crack toughness. Mater Sci Eng A 1989; 107: 135–143.

55.

Hutchinson

Suo

. Mixed Mode Cracking in Layered Materials. Adv Appl Mech, 29. Elsevier, 1991, pp. 63–191.

56.

Wang

De Freitas

Poulis

, et al. A review of experimental and theoretical fracture characterization of bi-material bonded joints. Composites Part B: Eng 2021; 206: 108537.

57.

Rice

. Elastic fracture mechanics concepts for interfacial cracks. J Appl Mech 1988; 55: 98–103.

58.

Obreimoff

. The splitting strength of mica. Proc R Soc Lond Ser A, Contain Pap Math Phys Char 1930; 127: 290–297.

59.

Gilman

. Direct measurements of the surface energies of crystals. J Appl Phys 1960; 31: 2208–2218.

60.

Prasad

Carlsson

. Debonding and crack kinking in foam core sandwich beams-I. Analysis of fracture specimens. Eng Fracture Mech 1994; 47: 813–824.

61.

Ramantani

de Moura

MFSF

Campilho

RDSG

, et al. Fracture characterization of sandwich structures interfaces under mode I loading. Composites Sci Techn 2010; 70: 1386–1394.

62.

Shivakumar

Chen

Smith

. An evaluation of data reduction methods for opening mode fracture toughness of sandwich panels. J Sandwich Structures Mater 2005; 7: 77–90.

63.

Williams

. The fracture mechanics of delamination tests. J Strain Anal Eng Des 1989; 24: 207–214.

64.

Carlsson

Sendlein

Merry

. Characterization of face sheet/core shear fracture of composite sandwich beams. J Compos Mater 1991; 25: 101–116.

65.

Carlsson

Prasad

. Interfacial fracture of sandwich beams. Eng Fracture Mech 1993; 44: 581–590.

66.

Avilés

Carlsson

. Analysis of the sandwich DCB specimen for debond characterization. Eng Fracture Mech 2008; 75: 153–168.

67.

Carlsson

. Elastic foundation analysis of tilted sandwich debond (TSD) specimen. J Sandwich Structures Mater 2000; 2: 3–32.

68.

Liechti

Marton

. Delamination of a high-temperature sandwich plate. Exp Mech 2002; 42: 206–213.

69.

Kanninen

. An augmented double cantilever beam model for studying crack propagation and arrest. Int J Fracture 1973; 9: 83–92.

70.

Williams

. End corrections for orthotropic DCB specimens. Composites Sci Techn 1989; 35: 367–376.

71.

Ural

Zehnder

Ingraffea

. Fracture mechanics approach to fact sheets delamination in honeycomb: measurement of energy release rate of the adhesive bond. Eng Fracture Mech 2003; 70: 93–103.

72.

Perrin

Bureau

Denault

, et al. Mode I interlaminar crack propagation in continuous glass fiber/polypropylene composites: temperature and molding condition dependence. Composites Sci Techn 2003; 63: 597–607.

73.

Shivakumar

Smith

. In situ fracture toughness testing of core materials in sandwich panels. J Compos Mater 2004; 38: 655–668.

74.

Gill

. Development and Evaluation of Fracture Mechanics Test Methods for Sandwich Composites. Seattle, 2008.

75.

Bragagnolo

Crocombe

Ogin

, et al. Investigation of skin-core debonding in sandwich structures with foam cores. Mater Des 2020; 186: 108312.

76.

Weaver

. Evaluation of Mode I Fracture Mechanics Test Methods for Sandwich Composites. Salt lake City: Master Thesis, University of Utah, 2009.

77.

Burlayenko

Pietras

Sadowski

. Influence of geometry, elasticity properties and boundary conditions on the Mode I purity in sandwich composites. Compos Structures 2019; 223: 110942.

78.

Maleki

Toygar

. The fracture behavior of sandwich composites with different core densities and thickness subjected to mode I loading at different temperatures. Mater Res Express 2019; 6: 45314.

79.

M-Y

Hutchinson

. Kinking of a crack out of an interface. J Appl Mech 1989; 56: 270–278.

80.

Prasad

Carlsson

. Debonding and crack kinking in foam core sandwich beams-II. Experimental investigation. Eng Fracture Mech 1994; 47: 825–841.

81.

Cantwell

Broster

Davies

. The influence of water immersion on skin-core debonding in GFRP-balsa sandwich structures. J Reinforced Plastics Composites 1996; 15: 1161–1172.

82.

Carlsson

. The tilted sandwich debond (TSD) specimen for face/core interface fracture characterization. J Sandwich Structures Mater 1999; 1: 60–75.

83.

Cantwell

Davies

. A test technique for assessing core-skin adhesion in composite sandwich structures. J Mater Sci Lett 1994; 13: 203–205.

84.

Glaessgen

Reeder

Sleight

, et al. Debonding failure of sandwich-composite cryogenic fuel tank with internal core pressure. J Spacecraft Rockets 2005; 42: 613–627.

85.

Grenestedt

. The Tilted Sandwich Debond Specimen (Personal Communication). Stockholm, Sweden: Royal Institute of Technology, 2020.

86.

Kessler

Adams

. Development and Evaluation of Fracture Mechanics Test Methods for Sandwich Composites. Wichita, 2009.

87.

Adams

. Development and Evaluation of Fracture Mechanics Test Methods for Sandwich Composites. Baltimore, MD, 2011.

88.

McGarva

Åström

. Experimental investigation of compression moulding of glass/PA12-PMI foam core sandwich components. Comp Part A: Appl Sci Manufac 1999; 30: 1171–1185.

89.

Adams

Czabaj

Stanfield

. Fracture Mechanics Test Methods for Sandwich Composites, 2017.

90.

Saseendran

Berggreen

Schäuble

, et al. Disbond of Sandwich Structures - DoSS: EASA, 2016.

91.

Viana

Carlsson

. Influences of foam density and core thickness on debond toughness of sandwich specimens with PVC foam core. J Sandwich Structures Mat 2003; 5: 103–118.

92.

Spies

. The peeling test on redux‐bonded joints. Aircraft Eng Aerospace Techn 1953; 25: 64–70.

93.

Moore

. An introduction to the special issue on peel testing. Int J Adhes Adhesives 2008; 28: 153–157.

94.

Sacchetti

Grouve

WJB

Warnet

, et al. Interlaminar fracture toughness of 5HS Carbon/PEEK laminates. a comparison between DCB, ELS and mandrel peel tests. Polym Test 2018; 66: 13–23.

95.

Kok

Grouve

Warnet

, et al. (eds). Effect of ply orientation on bond strength in fiber-placed composites; 2015.

96.

Satheesh

Tonejc

Potakowskyj

, et al. Peel strength characterisation on ply/ply interface using wedge and T-peel/pull-type tests. Polym Polym Composites 2018; 26: 431–445.

97.

de Rooij

Grouve

, et al. Characterisation of metal-thermoplastic composite hybrid joints by means of a mandrel peel test. Composites Part B: Eng 2016; 95: 293–300.

98.

de Rooij

Grouve

, et al. The effect of titanium surface treatment on the interfacial strength of titanium-Thermoplastic composite joints. Int J Adhes Adhesives 2017; 72: 98–108.

99.

Arouche

Budhe

Alves

, et al. Effect of moisture on the adhesion of CFRP-to-steel bonded joints using peel tests. J Braz Soc Mech Sci Eng 2018; 40.

100.

Reincke

Kreling

Dilger

. Adhesion of continuously manufactured fusion bonded multi-material structures consisting of steel and carbon fibre reinforced Polyamide 6. Int J Adhes Adhesives 2017; 79: 73–82.

101.

de Freitas

Sinke

. Adhesion properties of bonded composite-to-aluminium joints using peel tests. J Adhes 2014; 90: 511–525.

102.

Breslauer

Troczynski

. Experimental determination of the peel adhesion strength for metallic foils. J Adhes Sci Techn 1998; 12: 367–382.

103.

Kawashita

Kinloch

Moore

, et al. The Influence of Bond Line Thickness and Peel Arm Thickness on Adhesive Fracture Toughness of Metal Polymer Laminates, 2007.

104.

Martiny

Lani

Kinloch

, et al. Numerical analysis of the energy contributions in peel tests: a steady-state multilevel finite element approach. Int J Adhes Adhesives 2008; 28: 222–236.

105.

Breslauer

Troczynski

. The peel adhesion test for metal-ceramic interfaces. Key Eng Mater 1997; 132-136: 452–455, https://www.scientific.net/KEM.132-136.452.

106.

Sexsmith

Troczynski

. Peel adhesion test for thermal spray coatings. J Therm Spray Techn 1994; 3: 404–411.

107.

Sexsmith

Troczynski

. Peel strength of thermal sprayed coatings. J Therm Spray Techn 1996; 5: 196–206.

108.

Simlissi

Martiny

Mercier

, et al. Elastic-plastic analysis of the peel test for ductile thin film presenting a saturation of the yield stress. Int J Fracture 2019; 220: 1–16.

109.

Dolah

Matsuda

Miyagi

. L18 Optimization of T-Peel Test Method for Flexible Packaging Film, 2012.

110.

Lamnawar

Maazouz

. Rheological study of multilayer functionalized polymers: characterization of interdiffusion and reaction at polymer/polymer interface. Rheologica Acta 2006; 45: 411–424.

111.

ISO 4578 . Adhesives-Determination of Peel Resistance of High-Strength Adhesive Bonds - Floating-Roller Method. Geneva, Switzerland: International Organization for Standardization (ISO), 1997.

112.

DIN EN 1464 . Adhesives - Determination of Peel Resistance of Adhesive Bonds - Floating Roller Method. Germany, 2010.

113.

ASTM D3167-10 . Standard Test Method for Floating Roller Peel Resistance of Adhesives. West Conshohocken, PA: ASTM International, 2017.

114.

ASTM D6862 . Standard Test Method for 90 Degree Peel Resistance of Adhesives. West Conshohocken, PA: ASTM International, 2021.

115.

ASTM D3330 . Standard Test Method for Peel Adhesion of Pressure-Sensitive Tape. West Conshohocken, PA: ASTM International, 2018.

116.

DIN EN . Adhesives - Peel Test for a Flexible-Bonded-To-Rigid Test Specimen Assembly - Part 1: 90° Peel. Germany, 2013.

117.

DIN EN . Adhesives - Peel Test for a Flexible-Bonded-To-Rigid Test Specimen Assembly - Part 2: 180 Degree Peel. Germany, 2010.

118.

ASTM D1876 . Standard Test Method for Peel Resistance of Adhesives (T-Peel Test). West Conshohocken, PA: ASTM International, 2008.

119.

DIN EN . Packaging - Determining the Sealed-Seam Strength of Sealings Made of Flexible Packaging Material. Germany, 2012.

120.

ASTM D1781 - 98 . Test Method for Climbing Drum Peel for Adhesives. West Conshohocken, PA: ASTM International, 2021.

121.

Kinloch

Lau

Williams

. The peeling of flexible laminates. Int J Fracture 1994; 66: 45–70.

122.

Grouve

WJB

Warnet

Akkerman

. Critical assessment of the mandrel peel test for fiber reinforced thermoplastic laminates. Eng Fracture Mech 2013; 101: 96–108.

123.

Kawashita

Moore

Williams

. Comparison of peel tests for metal-polymer laminates for aerospace applications. J Adhes 2005; 81: 561–586.

124.

Nettles

Gregory

Jackson

. Using the climbing drum peel (CDP) test to obtain a GIC value for core/face sheet bonds. J Compos Mater 2007; 41: 2863–2876.

125.

Behrens

B-A

Bouguecha

Bonk

, et al. Finite element analysis regarding the forming behaviour of symmetric hybrid structures consisting of two sheet metal outer layers and a thermoplastic core. IOP Conf Ser Mater Sci Eng 2017; 179: 12004.

126.

Cañas

Távara

Blázquez

, et al. Overview of Gc tests used to evaluate composite–composite adhesive joints. J Multiscale Model 2019; 10: 1842002.

127.

Long

Mataira

Talantsev

, et al. Mode I delamination testing of REBCO coated conductors via climbing drum peel test. IEEE Trans Appl Superconductivity 2018; 28: 1–5.

128.

Rao

Das

Bhattacharyya

. Investigation of bond strength and energy absorption capabilities in recyclable sandwich panels. Compo Part A: Appl Sci Manufact 2013; 45: 6–13.

129.

Grünewald

Parlevliet

Matschinski

et al. Mechanical performance of CF/PEEK–PEI foam core sandwich structures. Jnl Sandwich Struc Mater 2017; 21: 2680–2699.

130.

Daghia

Cluzel

. The climbing drum peel test: an alternative to the double cantilever beam for the determination of fracture toughness of monolithic laminates. Comp Part A: Appl Sci Manufac 2015; 78: 70–83.

131.

Okada

Kortschot

. The role of the resin fillet in the delamination of honeycomb sandwich structures. Composites Sci Techn 2002; 62: 1811–1819.

132.

Cañas

Távara

Blázquez

, et al. A new in situ peeling test for the characterisation of composite bonded joints. Composites Part A: Appl Sci Manufacturing 2018; 113: 298–310.

133.

Kawashita

Kinloch

Moore

, et al. A critical investigation of the use of a mandrel peel method for the determination of adhesive fracture toughness of metal-polymer laminates. Eng Fracture Mech 2006; 73: 2304–2323.

134.

Sexsmith

Troczynski

Breslauer

. Plastic work in the peeling of work hardening foils. J Adhes Sci Techn 1997; 11: 141–154.

135.

Arouche

Budhe

Banea

, et al. Interlaminar adhesion assessment of carbon-epoxy laminates under salt water ageing using peel tests. Proc Inst Mech Eng L: J Mater Des Appl 2019; 233: 1555–1563.

136.

Kawashita

Moore

Williams

. The development of a mandrel peel test for the measurement of adhesive fracture toughness of epoxy-metal laminates. J Adhes 2004; 80: 147–167.

137.

Grouve

WJB

Warnet

Akkerman

, et al. Weld strength assessment for tape placement. Int J Mater Forming 2010; 3: 707–710.

138.

de Freitas

Sinke

. Test method to assess interface adhesion in composite bonding. Appl Adhes Sci 2015; 3.

139.

Grouve

WJB

Warnet

Rietman

, et al. On the weld strength of in situ tape placed reinforcements on weave reinforced structures. Composites Part A: Appl Sci Manufac 2012; 43: 1530–1536.

140.

Sacchetti

. Interlaminar Toughness of Fusion Bonded Thermoplastic Composites. Dissertation, University of Twente. Enschede, 2017.

A review of mode I dominant interfacial fracture toughness test methods of skin-core bonding for thermoplastic composite sandwich structures

Abstract

Keywords

Introduction

Requirements for test methods

Interfacial fracture toughness test methods

Double cantilever beam

Single cantilever beam

Peel tests

Climbing drum test

Floating roller peel test

Mandrel peel test

Direct comparison of mode I dominant interfacial fracture toughness test methods

Conclusion

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

Note

References