Impact damage testing on composite marine sandwich panels,part 1: Quasi-static indentation

Abstract

High fibre content composite laminate skins use marine grade orthophthalic polyester, POLYLITE® 440-M850 resin and chopped strand mat/woven E-glass for the thin outer face skins and DIVINYCELL® H100 closed-cell linear PVC foam as core. Marine grade hybrid sandwich panels are designed in accordance with BS EN ISO 12215-5:2008 for small craft hull construction, using a wet lay-up and cured under vacuum pressure. Impact damage testing followed ASTM D7766-11 and ASTM D6264-98 procedures A and B for rigidly supported and simply supported sandwich panels. A review of the current state of standard testing procedures of marine sandwich panels is described. Testing using the default hemisphere indentor also included other standard rock geometries – conical, square-based pyramid and flat-faced cylindrical. The damage incurred under each variation of indentation impact is described, in terms of force, absorbed energy and indentation displacement. New contact laws are suggested for the different rock geometries. Destructive sectioning of the panels provides the visual damage incurred through the thin face skins and core and the roles played by all members comprising the sandwich panels.

Keywords

Composite impact marine sandwich vacuum bagging quasi-static testing

Introduction

Composite material technology application within the general engineering field is rapidly increasing. Composite structures are only limited to the designer’s imagination. Reinforced plastic composites are now the material of choice in the marine craft manufacturing industry. Fibre-reinforced plastic (FRP) offers many advantages over more traditional materials (steel, aluminium, wood) such as resistance to corrosion and rot, high specific material properties, which can be engineered to one’s own requirements and the ease of forming composite complex shapes. FRP laminate skins in addition to a marine grade core material, produces sandwich panels. The variety of sandwich panels produced and method of fabrication ultimately depends upon its location and application on the marine craft. The sandwich panel construction is closely controlled by maritime classification society rules, enabling the marine craft to obtain certification.

Laminated composite marine sandwich panels are susceptible to impact damage events. Such events on the hull are not only limited to the normal hydrostatic loadings usually associated with high speed craft (HSC) and the associated out-of-plane events, but also limited to the impact events of collisions with foreign objects, which can occur at varying speeds of impact, such as floating debris, other craft, docks, falling objects and grounding.

Damage on a composite panel is a complex and a dynamic event owing to the many constituents making up the part. The large global deflections induce membrane and shear deflections. Damage modes in the laminate skin may induce internal delamination, matrix cracking, fibre fracture and ply shear. Damage, which may occur, is not always clearly visible; however, it can cause a significant reduction in strength that can lead to premature failure of the laminate of the overall sandwich panel. Understanding the damage characteristics of a composite sandwich laminate is therefore essential in order to optimise their design against such failure.

The investigation involves the quasi-static low-velocity indentation impact (QSLVII) on seven certifiable marine sandwich panels. Following a brief survey of literature, a description of the materials and method used in the production process is given whereby high quality vacuum bagged sandwiches may be fabricated. Reasons are given as to why these specific marine panels are engineered, following the strict small craft standards and maritime rules. The characterisation of such panels is highlighted to obtain their important material and mechanical properties. A review of the current state of standard testing of marine panels is described, which highlights that, presently, standards do not fully describe sandwich indentation impact since the impact standards commonly only relate to laminates, whilst testing specifically of marine sandwich panels is non-existent. The standards adopted here only describe what is expected for a hemisphere indentor. Marine hulls commonly suffer impact from pointed bluff and sharp edge objects not only from objects which fall on the deck, but also from grounding and impact incidents when underway. Therefore, these objects are standardised into three other shapes – conical, pyramid and cylindrical. The testing regime presented here specifically allows only 50–80% panel penetration, since through panel damage is not desired in craft hulls since this may result in water ingress and subsequent foundering. Furthermore, it is also necessary to compliment and consider further standard tests, such as instrumented drop weight testing reaching the same panel penetration and residual strength analysis of the damaged panels. The present damage analysis and the response will be described in terms of the force and absorbed energy during indentation displacement and how the standard is applied to all the test variations described here. New contact laws are proposed, which vary depending on the indentor geometry and ensuing damage. Throughout the discussion, various shortcomings in the standards will be presented: regarding the reporting of the findings; how one is to present the forces and points of first noticeable damage and localised maximum forces; to show the effect of the two support configurations; describe how the various indentor geometries effect the results obtained; destructive sectioning of the panels to report the findings on the skins and core in terms of fibre failure, fibre pull out, matrix cracking, delamination, core and face disbanding, core crushing and cracking.

Abrate [1 –4] carried out different studies on the impact behaviour of single skinned laminates presenting different indentation mechanisms and mathematical models for predicting the damage on composites. Low velocity impact (LVI), experimental and numerical analysis and the effect of residual strength have been discussed by Abrate [4], Choi et al. [5, 6], Choi and Chang [7], Finn and Springer [8], Lakshminarayana [9], Huang [10], Zheng and Sun [11], Zhou [12], Chun and Lam [13], De Moura et al. [14 –16], Hou et al. [17], Davis and Olsson [18], Hosur et al. [19], Li et al. [20], Bouvet et al. [21], Murthy et al. [22], Loresto and Caprino [23]. High-velocity impact using non-instrumented drop weight testing including the effect of variation of impact geometries has been investigated by Sevkat et al. [24] and Mitrovski et al. [25 –27].

Sandwich impact has been investigated by Abrate [28] where he discusses the relation between the skin displacement and the contact force when in contact with a spherical indentor. Olsson [29] and Chai and Zhu [30] consider different contact and impact models depending upon the impactor to panel mass ratios using static indentation experiments and LVI impacts on sandwich structures. Hazizan and Cantwell [31] studies LVI of PVC foam-based sandwiches. Wen et al. [32] used indentors of various geometries and different impact velocities. Borsellino et al. [33] investigated vacuum bagging sandwich panels with variations in indentor speed and various skin and core materials. Hasebe and Sun [34] investigated foam core sandwiches under static and dynamic point loads, whilst Foo et al. [35] used a modified energy-balance model to predict LVI responses on sandwich composites as well as applying an finite element analysis (FEA) to static indentation. Simply supported and rigidly backed sandwich beams of E-glass-epoxy and PVC foam cores using a single degree of freedom and two degree of freedom (TDOF) models together with an Abaqus analysis was reported by Sadighi and Pouriayevali [36]. Abot et al. [37] uses the Winkler foundation theory on Divinycell PVC foams with carbon/epoxy skins on rigid backed and simply supported beams. Hand lay-up and vacuum bagged carbon fibre skin and foam-filled honeycomb was investigated by Gustin et al. [38] who developed a TDOF spring/mass model. Williamson [39] impacted panels of graphite/epoxy skins with a honeycomb core. Wang et al. [40] used different indentor hemispherical diameters on PVC foam cores and carbon fibre skins. Li et al. [41] tested Rochacell polymethacrylimide (PMI) foam and unidirectional carbon fibre in a bismatermide (BMI) resin matrix and studied both quasi-static and dynamic impact. Li et al. [42] carried out further investigation on low-velocity perforation behaviour of aluminium foam core composite sandwich panels using conical nosed, hemispherical nosed and flat-ended geometry indentors.

Owing to the growing popularity of the use of composite laminates and sandwich panels in the fabrication of marine craft, it becomes important to understand the behaviour and properties of E-glass/polyester marine laminates and closed-cell PVC foam cores. As a result research within this area is starting to emerge. Sutherland and Soares [43] have investigated low energy impact on low fibre volume hand lay-up marine skin laminates and have also reported on scaling [44], impact behaviour [45 –47], characterisation [48], and laminate thickness, matrix resin and reinforcement type [49 –51]. Various authors [52 –59] have also investigated water slamming and solid body impacts. De Marco Muscat-Fenech et al. [60] have also investigated the impact on hand lay-up vacuum packed marine sandwich panels on design standard hulls.

Characterising composite laminate sandwiches

When designing any composite structure, the properties must be well-defined and understood. The designer can engineer a composite sandwich fit for its intended purpose, which results in a specific sandwich for its application. Material (mechanical) properties may be evaluated using theoretical equations; however, experimental verification (usually destructive) using standardised test procedures is essential for proper modelling and understanding of the characterisation of the composite sandwich materials. These composite materials usually have non-homogenous structures and their properties are dependent on many factors, such as the reinforcements, property of the matrix, volume/mass fraction of the phases (also dependent on the fabrication process) and core material.

The marine panels investigated here have undergone an in-depth testing regime to determine the tensile, compressive and flexural properties required for understanding the sandwich response to inflicted damage. The properties have been determined for the faces acting alone, when fabricated using traditional methods and cured under both natural curing conditions and when using the vacuum backed curing process. Improvement of the laminate skins, for the constituent laminates described here, is well-documented and recorded by De Marco Muscat-Fenech and Cortis [61] and Cortis et al. [62]. Additionally, the sandwich hull panels have also been characterised by De Marco Muscat-Fenech et al. [63] following the standard ASTM testing procedures to obtain the panel characteristics necessary to define the response of the panels to the indentation, impact damage and global working conditions. The material (mechanical) properties are presented during the description of the design of the hull sandwich panels.

Composite materials used and method of fabrication of hull panels

The application considered here is for manufacture of planning craft for a local boat builder. The reinforcement fabrics used for the construction of the composites in this research are mainly chopped strand mat (CSM) and woven roving. The glass fibre is E-glass. The CSM used is made up of PPG called Mat 92, having an average fibre diameter of 11 µm. The binder holding the fibre strands together is an emulsion and it is therefore very easy to work around with complex shapes. Woven roving cloths plain weave consisting of equal warp and weft, passes over each other in a symmetrical fashion. The resin used was the Reichhold POLYLITE® 440-M850 orthophtalic polyester resin. For an ambient temperature of 20℃, a 1% catalyst of methyl ethyl ketone peroxide (MEKP) hardener was employed to begin the polymerisation process, giving a gel time of about 45 min.

The foam core being used in this investigation, sandwiched between the outer thin skin layers, is a closed-cell cross-linked PVC foam called Divinycell® H100 by DIAB group (Figure 1). This PVC foam exhibits high strength-to-density ratio and is compatible with most of the resin systems used including those having high styrene content. It is suitable for both hand laminating and vacuum pressure, having a density of 100 kg/m³. The core has good chemical resistance, low water absorption and good thermal insulation properties. The foam comes in different finishes, including ‘double-cut’ finish which has been chosen for this research; this finish involves cutting the core in a 0 and 90° pattern on both sides of the core, with an offset equal to half the size of a cut between the upper and lower sides. This type of finish was chosen since it has two main uses. Firstly, it can be used in small curvatures without the need of using a scrim (which is an additional layer of fibre glass cloth attached to the foam). Secondly, since the cuts overlap each other, both air and resin can easily flow through the foam thereby providing a good bond between the skins and the core. The hand lay-up resin and foam core material is designed specifically for marine, industrial and transport application and is also certified by most classification societies (Lloyds, DNV, BV, RINA) for use in marine applications.

Figure 1.

The double cut closed-cell cross-linked PVC foam Divinycell® H100 by DIAB, thickness 10 mm, 20 mm and 30 mm.

Figure 2.

Four of the hybrid sandwich panels being cured under vacuum pressure during manufacturing.

Vaccum bagging fabrication is accomplished using the Vacmobile™ 20/2, which is a complete unit incorporating a vacuum pump, a resin catch pot and a pressure regulating valve (all in one unit). The vacuum pump can handle 18 m³/h of air, and can reach 99% vacuum, at a pressure of 10 mbar or 29.5 in. of mercury.

Vacuum-bagged sandwich panel fabrication procedure: The foam core sheets of the required thicknesses were cut into 600 mm × 600 mm. The edges were tapered at an angle of 45° so that the vacuum bagging consumables (mostly the peel-ply and breather fabric) would be easily removed after the curing stage.

The moulds were prepared by applying a couple of coats of mould release wax to the surface, which will aid in the removal of the panel once it has been cured. At least six coats were applied at intervals of between 15 and 30 min to avoid the laminate sticking to the mould. The reinforcement fabrics for the skins were cut into 600 mm × 600 mm and weighed, and the amount of fibre reinforcement in the panel was noted. The mass of the reinforcement fabric was multiplied by 1.5 in order to approximately determine the mass of resin required. The resin was catalysed with 1% of MEKP hardener so that sufficient time is allowed for the layup and the bagging consumables to be set-up.

The process started by applying resin on the mould using a brush, before placing the first layer of fibreglass fabric. The layers making up one skin were added one by one, whilst at the same time applying resin to wet out the fibres. The last fabric layer before the foam core was wet excessively with resin, so that when the vacuum is applied, this excess resin could travel into the grooves of the core. The foam core is then added and another resin-rich layer of fibreglass mat is then applied over it. Again the excess resin enters the grooves that are found inside the core, which are useful when the shape of the mould includes some simple contours and curves. Once the rest of the fabric layers were laid and wet with resin, the peel ply was applied on top of the last layer and squeezed with a plastic squeegee. This process helps the peel ply to absorb some resin and avoid any streaks, which would show up in the final laminate. Any air bubbles present were squeezed out to prevent wrinkles that may otherwise show up in the laminate, thus compromising on the final surface finish of the panel. The P3 perforated release film was then applied over the peel ply. This layer is placed on top of the peel-ply fabric and controls the volume of resin to be absorbed by the breather fabric, which is the next layer to be added. The vacuum bag was then attached to the periphery of the mould using mastic sealant tape and a vacuum connector was placed in the bag but outside the panel area. A small tear was made in the bag and the vacuum pipe was inserted into the vacuum bag. The pipe was then sealed to the bag using a couple of turns of mastic sealant tape wrapped around the vacuum pipe. A vacuum of 0.5 bar (or 14 in. of mercury) was applied to the laminate and left to cure under vacuum for about 4 to 6 h at room temperature. Once the laminate was cured, it was left for 24 h with the vacuum pump disconnected and the mass of the panel was measured, in order to determine the weight percentage of reinforcement used. The breather fabric absorbs excess resin that pass through the perforated release film. This process was repeated for three other different panels so that a total of four panels could be manufactured simultaneously, Figure 2.

Figure 3 shows the thicker traditionally made laminate panel with natural curing conditions with a resulting glass fibre content of 25 vol.% (35–40 mass%); a second laminate panel equivalent in fabric content as the first, fabricated using the hand lay-up and cured under vacuum pressure, showing the reduced thickness of such vacuum-bagged laminates reaching 45% fibre volume fractions (65% fibre by mass) and 40% reduction in thickness (Figure 4) and the marine hybrid sandwich panel also fabricated and cured under vacuum pressure. Cortis and De Marco Muscat-Fenech [62] and De Marco Muscat-Fenech and Cortis [61, 63] have reported the results of increased mechanical properties of the vacuum-cured laminates over the naturally cured laminates as well as of the vacuum-bagged sandwich panels.

Figure 3.

Skin laminate panels using traditional hand lay-up, natural curing (middle), hand lay-up vacuum bagging (top), vacuum bagged sandwich panels (bottom) [61].

Figure 4.

Illustration of the reduction in thickness due to the vacuum pressure acting on a laminate made up of woven roving and mat: (a) shows the hand lay-up which has much more resin (light grey areas) and less mechanical linking between the fibres (dark grey areas); (b) shows the effect of applying vacuum pressure across the laminate surface. Reduction in thickness occurs due to better compaction of the fibres and less resin content [62].

Design of the hull sandwich panels

In order to design strong and effective panels, the BS EN ISO 12215-5 [64] standard was used. This standard is the established small craft building standard for hull scantling determination. The builders must follow one of the widely available standards in order to obtain the CE certification required to sell the boats to the local and international customers. During the course of the investigation, a FACTS Hull Property Calculator [65] provides a framework for the boat designer to engineer the best combination of skin and core material, to produce a lightweight and certifiable hull and structure. This model uses boat craft hull dimensions, craft design speed and panel sizes, for which the hydrostatic pressures, in the displacement and planning mode, acting on each part of the hull (bottom and side) and deck can be determined. Knowing the pressures which act on the craft, values of minimum skin and core thicknesses with the correct material properties may be computed. This marine hybrid sandwich panel is then subjected to the computed pressure applied to the DNV Rules for high speed light craft (HSLC) [66]. The model investigates the face stress to skin ultimate strength ratio, the core shear stress to ultimate shear strength ratio and overall panel deflection. These ratios and deflection all have criteria, which must be passed according to the HSLC code for the craft to be certified – the final aim for any boat builder.

Two of the most popular boat models produced by the manufacturer were used to size the panels for the testing regime described here. However, the procedure can be equally adapted and applied to different sized craft. The design pressure and hence the thickness of the panels depends upon the panel size between two stiffeners and bulkheads of the hull. The new sandwich panels will be overall thicker than the single skin panels; however, the panel area may be increased in order to reduce the number of stiffeners and hence reduce the weight of the hull. Panel sizes ranged from the standard 700 mm × 300 mm (used in single skin hull construction) up to a maximum of 2100 mm × 900 mm. As the panel sizes get larger, the force on the panel also increases for the same pressure; this results in an overall increase in panel thickness. The required inputs to start the sandwich panel design include the boat’s length of hull, waterline length, beam, maximum cruising speed and dead rise angle at 0.4 times the waterline length. For hulls having a length less than 9 m, a design category factor needs to be determined, which in this case takes the value of 0.6 for CE design category C. The standard stipulates that this design category refers to boats, which are able to operate in inshore seas with significant wave heights of up to 2 m and a wind force of Beaufort Scale 6 or less. The dynamic load factor is calculated next, which is a value of the acceleration expressed in g’s that the hull experiences at a certain wave height. The design area is also input and the pressures are then calculated, at different locations along the hull for both displacement and planning modes. For calculations involving panel thicknesses, only the design pressures in planning mode are considered, since the planning pressures exceed the displacement mode pressures. Other calculated design pressures include the side and the deck pressures. The pressures resulting on the side and deck will result in values lower than the bottom planning pressure. These pressures are evaluated so that the side or deck panel thickness may be determined. When designing the sandwich panels there are many geometrical features of the craft that the designer must know and take into consideration for sandwich design. In addition to these geometrical attributes, the material characteristic of both the laminate skins and core are required. Typical required values are of tensile and compressive ultimate strength of the skin and its associated fibre volume fraction, skin and core tensile and compressive moduli, core shear modulus, core ultimate shear strength.

FACTS Sandwich Panel Property Calculator [65] utilises sandwich beam and plate theory. This concept model allows the user to input the face and core properties and geometrical information, whilst checking that thin skin sandwich criteria are fulfilled. This model besides evaluating the flexural rigidity, shear stiffness and deflections of these panel, also analyses possible potential failure modes when forces are applied to various sizes panel plates for point loading at the panel midpoint or when distributed loads are applied.

The sandwich panels were designed with symmetrical skins because both skins have their own importance. The outside skin is the side which directly faces an impact, whilst the inner skin is the skin on which the deck compartment will be mounted on and, therefore, it must resist impact loads when passengers move or something relatively heavy is dropped on the floor of the deck.

Seven different panels were designed, with core thicknesses ranging from 10 up to 30 mm. These panels represent the different design panel areas used in order to reduce the number of wooden stiffeners and hence total craft weight. Table 1 shows the seven panels that were fabricated, including the core thickness and the face sheet constituents. The face sheets comprise CSM and woven rovings (denoted by w next to the mass density of the fibres). The skin thickness was measured using a micrometer having an accuracy of ±0.01. Table 2 describes the sandwich panel material and mechanical properties, together with the ASTM standards used during the panel characterisation.

Table 1.

The sandwich panels constituents according to BS EN ISO 12215-5 [64].

Panel name	Core thickness (mm)	Fibre sheets in each skin	Skin thickness (mm)
A	10	450CSM, 450CSM	0.81
B	10	450CSM, 300CSM	0.67
C	15	300CSM, 450CSM, 400w, 450CSM	1.35
D	15	450CSM, 450CSM, 600w, 450CSM, 450CSM	2.03
E	20	450CSM, 600w, 450CSM	1.22
F	25	450CSM, 450CSM, 600w, 450CSM, 450CSM	2.03
G	30	300CSM, 450CSM, 600w, 450CSM, 300CSM	1.76

Table 2.

Sandwich panel material (mechanical) properties and ASTM standards used for the panel characterisation.

Panel
ASTM Standard		A	B	C	D	E	F	G
Effective sandwich flexural stiffness/rigidity, D (×10⁶ Nmm²)	D7250M-06	49.6	45.3	248	374	393	945	1160
Sandwich shear rigidity, S × 10 ^s (N)	D7250M-06	32.3	33.4	42.9	44.0	46.1	61.0	67.8
Facing ultimate strength, F^U (MPa)	D7249M-06	140	161	169	143	144	141	131
Core shear modulus, G MPa (manufacturer’s data = 35 MPa)	D7250M-06	46.0	37.3	31.1	29.1	26.7	27.3	26.5
Effective facing chord modulus, E^f (GPa) – tensile	D7249M-06	10.9	13.2	16.3	15.9	18.0	17.1	16.8
Effective facing chord modulus, E^f (GPa) – compresssive	D7249M-06	12.3	15.0	16.5	15.1	18.5	19.6	20.7
Core shear allowable strength (manufacturer’s data = 1.6 MPa) F_sult (MPa)	C393M-06	1.69	1.39	1.95	2.23	1.65	2.14	1.87
Core shear yield strength, F_Syield (MPa)	C393M-06	1.5	1.2	1.7	1.8	1.5	1.6	1.7
Facing stress at core shear failure, σ_f (MPa)	C393M-06	133.8	140.5	101.7	81.3	98.7	81.5	79.0

Impact damage resistance testing

The British Standard has a few standards describing impact on rigid plastic, but none on sandwich panels. Of notable interest for impact of rigid plastic composite panels are BS EN ISO 7765 [67] for laminates with thicknesses less than 1 mm, whilst for laminates of thickness greater than 4 mm testing falls within the scope of BS EN ISO 6603-1 (non-instrumented) [68] and BS EN ISO 6602-2 (instrumented) [69], using the falling dart technique. BS EN ISO 6603-1 [68] characterises the impact behaviour of plastics by an energy criterion of a failure based on a threshold value of impact failure-failure energy, such that the energy is estimated for 50% of the specimens to fail, and any failure or fissure on the surface of the specimen is visible by the naked eye without penetrating the full thickness of the material. This failure is achieved by either altering the falling mass or height of fall. Whilst the instrumented test procedure determines the puncture impact properties of these rigid plastic through the measurements of force and deflection throughout the impact time, as well as the impact velocity. The striker for such tests is hemispherical impactors with diameter of either 10 mm or 20 mm (default). However, no BS EN ISO standard exists to test any sandwich or even marine grade sandwich panels under impact loading scenarios.

The standards referring to impact test methods on sandwich constructions are provided by the American Society for Testing and Materials (ASTM), under the prerogative of D30.05 on structural test methods and D30.09 on sandwich construction.

Also as outlined above, there exists no standard that specifies what tests need to be followed to test marine hull panels or structures according to ASTM standards. The subcommittee introduced the D30.05.01 for Civil and Marine Testing, with the aim of introducing such testing procedures for these applications. Also specific standards to establish whether a composite or a marine sandwich structure is able to support impact loads during operation is not available and which is the focus of this present study.

Standards regulating the impact damage resistance testing of sandwich constructions

In 2011, ASTM published the standard ASTM D7766-11 [70] (under the jurisdiction of the subcommittee D30.09). This standard brings together the test methods found in the standards for single skin laminates ASTM D6264 [71] and D7136 [72], which are to be applied to sandwich constructions and QSLVII. It consists of three main procedures called A, B and C, each having a different purpose and use.

Procedure A involves an indentation test of a rigidly backed specimen. An out-of-plane concentrated force is applied so that damage is imparted on the specimen surface using a hemispherical indentor at a low speed. The size of the dent, location and type of damage resulting from the applied indentation gives an idea of the damage suffered by the panel.

Procedure B involves the low-speed indentation impact of a specimen, which is supported at the edges. This procedure is very similar to procedure A, but the specimen is now supported at the edges such that it is allowed to deflect. Damage on the specimen is again quantified by the depth and size of the indentation, the location and the type of failure observed at the skin and/or core.

Procedure C of the standard involves performing a drop weight impact test on a sandwich specimen supported at the edges. This is different from the previous two methods which are quasi-static, i.e. performed at low speeds.

The standard specifies that the most suitable procedure to evaluate and compare different damage characteristics for different sandwich panels is procedure A, since the rigidly supported specimen prevents out-of-plane deflection of the specimen. Therefore, the sandwich flexural stiffness does not influence damage initiation in the specimen. A hemispherical indentor is specified as having a geometry, which has, over the years, generated a larger amount of internal damage for a given amount of external damage, used mainly in applications for plastics. However, in addition to the standard hemisphere other ‘rock’ geometries also produce damage of interest to both hull and sandwich manufacturers. In this investigation, all the procedures outlined in standard ASTM D7766 [70] and ASTM D6264 procedures A and B [71] shall be followed in order to determine quasi-static indentation impact loading damage assessments. The drop weight impact testing following ASTM D7136, procedure C [72] is the subject of another investigation by De Marco Muscat-Fenech et al. [73].

Experimental design requirements

The testing apparatus for carrying out a quasi-static indentation test was designed and built in-house according to the ASTM D6264 [71] standard, using procedures A and B. A force is applied to the panel by pushing a hemispherical (default) indentor into the marine sandwich panel. The displacement of the indentor is increased until a required damage state is reached. The contact force and indentor displacement are noted so that a force–indentation displacement curve can be plotted and analysed.

The apparatus for testing follow the standard comprising a lower support fixture consisting of a flat rigid plate of dimensions 300 mm × 300 mm and thickness of 30 mm. The flat support is fixed to the machine base using the recommended metal T lower support fixture. For procedure A, the sandwich hull panels sit on the flat plate to perform the rigidly back tests. A plate of dimensions 200 mm × 200 mm of thickness 40 mm (the thickness of the plate was set to 40 mm, since as specified in the standard; the thickness must be greater than the expected maximum indentor displacement, panel G is of thickness approximately 35 mm) with an opening of 127.0 mm, with the tip rim of the opening rounded to a radius of 0.75 mm sits on top of the rigid plate to allow the simply supported configuration for procedure B. An upper fixture connects the indentor to the load cell of the Instron 4206 tensile testing machine via the indentor housing. The indentor which is to strike, indent and penetrate the sandwich panel is hemispherical with a diameter of 12.7 ± 0.1 mm (default), as shown in Figure 5.

Figure 5.

Hemispherical indentor, 12.7 mm diameter.

In order to simulate the different indentation impact loads that can be experienced by the sandwich panels during operation, different indentors were designed to simulate varying and different damage requirements, as shown in Figure 6. Common grounding scenario implies that the boat hull can be hit by sharp pointed rock edges, which define the need to indent the sandwich panels with a sharp object.

Figure 6.

Indentors – cylindrical, conical, square based pyramid.

A conical indentor simulates the rocks type of a sharp cutting tip and bluff edges, which push and open through the composite laminate and core material. The apex angle is 30°, with an indent depth of 46.7 mm. A pyramid-shaped indentor was designed, having a sharp point and four edges, leading to a square base having sides of 28.28 mm of diagonal 40 mm. The sharp tip pierces and penetrates the composite surface and edges cuts through the laminate and core as the damage is being widened. The apex angle was kept fixed at 30° to give an overall length indent depth of 52.8 mm from the square base to the apex. A cylindrical indentor was also designed to simulate the rock impact. The cylinder is of diameter 15 mm and an overall length of 40 mm. The blunt flat front compresses the composite surface and core materials. The sharp circular edge cuts through the surface and foam. All indentors are made up of EN24 (molybdenum nickel chromium) alloy steel, with a surface hardness of 60 to 62 HRC. This steel possesses high shock resistance and good impact properties at low temperatures. The complete upper assembly is shown in Figure 7.

Figure 7.

(a) Assembled testing rig for the rigidly backed specimen configuration – procedure A; (b) assembled testing rig for the simply supported specimen configuration – procedure B.

Simulating slow speed indentation impact damage scenarios of sandwich hull panels

The test procedure is in accordance with procedures A and B of standard ASTM D7766-11 [70]. The preparation and test procedure is in accordance with standard ASTM D6264-98 [71], but applied and modified for sandwich constructions.

Procedure A: Rigidly backed specimens. The specimens were cut into a size of 75 mm²× 75 mm² since larger specimens as stated in the standard are not required (Figure 8(a) and (b)). The size of the specimen chosen, after careful consideration, ensured that the dent periphery was concentrated at the centre and there was sufficient area around the dent periphery where no damage was visible.

Figure 8.

Quasi-static indentation testing using the different indentor shapes: (a) procedure A – hemispherical indentor; (b) procedure A – conical indentor; (c) procedure B – pyramid indentor; (d) procedure B – cylindrical indentor.

Procedure B: Edge supported specimens . The specimen size was fixed at 152 mm²× 152 mm², as per the standard so that the specimens fit onto the 200 mm × 200 mm plate containing the 127 mm hole, which allows the specimens to deflect under load (Figure 8(c) and (d)).

The centre of the specimen is marked to ensure that the specimen’s centre is in line with the centreline of the indentor, which is not to exceed a 1 mm difference. Load is applied to the specimen at a constant loading rate of 1.25 mm/min for cores with low compressive strength such as foams and honeycombs. The maximum force to act on the specimen during the test should ideally be reached within 1 to 10 min after initial application of the load. The machine cross-head was stopped within 50–80% through-thickness indentation, but definitely before penetration of the inner bottom skin, to ensure that no damage occurs to the indentor and to the rigid plate, particularly in the case of the rigidly backed specimen configuration. Hulls need to be designed to resist impacts, which are not through thickness. Such damage would allow water ingress and could result in the loss of the vessel and potential loss of life at sea. Furthermore, CAI testing according to ASTM [74] should also be undertaken on such panels to evaluate the residual strength of the concerned hull panels. The tensile testing machine recorded values of force (kN) against crosshead displacement (mm), which enabled a graph of indentation force (kN) versus indentor displacement (mm) to be plotted for each of the samples. The maximum force required to fully puncture the upper skin and the core (except for the lower skin) was also recorded so that the indentor that causes the greatest amount of damage was noted.

The periphery of the dent is measured using a depth gauge. The depth gauge is zeroed at 25 to 50 mm away from the centre of the specimen, where the specimen is flat. When the depth gauge is moved to the centre, the peripheral point can be identified as being that point at which the depth gauge starts giving a non-zero reading. The maximum dent diameter can be determined using the distance between the two points which are furthest from each other. Although the dent diameter measurement is recommended in the standard, it is reported that depending upon the indentor causing the damage, the dent diameter measurement as described here reveals no significant importance except visual observation. It is more productive to view and report on the face (and bottom) skin and internal damage in terms of fibre failure, fibre pull out, matrix cracking, delamination, core and face disbanding, core crushing and cracking.

Each of the different designed indentors represent a particular impact scenario. Different indentations types and sizes are expected using different shapes. Using this test, a boat designer/builder may determine what failure mode is expected when the outer hull made up of a sandwich panel will be subjected to grounding or hitting of a sharp-edged object during service. The designer may also alter the design to see what core/skin thickness combination (within the strictly controlled maritime guidelines) would be best to resist such loads and where failure is likely to occur.

All four indentors and seven panels were tested following both procedures and each test was repeated for a minimum of five times.

Quasi-static test results – procedure A: Rigidly backed configuration

The complete set of curves for all indentors and panels are shown in Figure 9. All plots show that each indentor geometry produce similar trends for all the seven panels.

Figure 9.

Graphs of force (kN) versus indentor displacement (mm) for: (a) hemispherical; (b) conical; (c) pyramid; and (d) cylindrical indentors in a rigidly backed specimen configuration, procedure A, for all panels.

In the following discussion panel F, having a skin thickness of 2.04 mm and a core thickness of 25 mm is chosen to be indicative of all the panels. The force–displacement graph shown in Figure 10 for the hemispherical indentor indicates that at the increase in load, the graph shows that a first noticeable damage point is always present. This position occurs at an indentation depth of t_s to 1.3t_s, (where t_s is the skin thickness). This point indicates that damage is occurring at the interface between the top skin and core. The force continues to increase and peaks produce a localised maximum at an indentation of approximately 2.5t_s. This position indicates that the upper skin is completely punctured.

Figure 10.

Force (kN), absorbed energy (J) versus indentor displacement (mm) for hemispherical indentor – panel F.

Figure 11 shows the average first noticeable damage and localised maximum points for each panel. A separate study is currently underway, analysing the contact law for the initial force increase before the localised maximum is reached, F = kαⁿ (where F is the contact force, k is the contact stiffness, α the indentation displacement and n is a constant) and additionally employing finite element analysis methods has shown that the force–indentation displacement is not linear, i.e. n = 1 as often suggested for sandwich panels [1, 4, 29, 30, 36] nor is n = 1.5 for monolithic laminates [3, 4, 30, 37] but follows the power law of n = 1.3. Whilst Abrate [1, 2] suggests n = 0.8 for a hemisphere. The power index has been validated throughout experimental testing. The energy absorbed during the indentation impact process was also analysed. From the start of indentation to the localised maximum peak, the absorbed energy follows an n = 2.3 power law. The foam is soft compared to the composite skin. The foam material under the advancing indentor offers reducing resistance as the foam thickness decreases, as a result the force value also decreases.

Figure 11.

Force of first noticeable damage and localised maximum for the hemisphere indentor.

Figure 12.

Force (kN), absorbed energy (J) versus indentor displacement (mm) for conical indentor – panel F.

The force–indentation plot during the conical indentation process is shown in Figure 12. The graphs show that two contact curves are required to describe the whole process. The crossover occurs at an indentation depth of approximately t_s. Initially as the force increases during indentation of the face, the power index is n = 0.8, following which a linear trend line fits the remaining experimental data. The energy absorbed during the whole process is obtained from data acquisition during the tests. The energy trend curve, fits a power index, n = 1.8, which compliments the findings just described for the force–indentation plots adopting an n = 0.8 power law. The composite skin and core offer little resistance as the point pierces its way first through the skin and sub-sequentially through the core. The steady increase in force value is due to the increasing cross-sectional area penetrating the specimen at a uniform rate, and therefore the damage imparted to the specimen also increases uniformly. Reporting maximum force, as dictated by the standard has no significant meaning for such an indentor. The force will keep on increasing so long as the advancing cross-sectional area increases.

Figure 13.

Force (kN), absorbed energy (J) versus indentor displacement (mm) for pyramid indentor – panel F.

The graph in Figures 9(c) and 13 show the pyramid indentor results. These are very similar to the conical indentor, following the initial power law and linear trend after top skin penetration. The cross-sectional area penetrating the specimen also increases with indentor travel producing the increase in force trend with indentation displacement. However, the damage mechanisms vary; the conical indentor opens the fibres around the smooth circular surface area, while the pyramid cuts its way through. These effects described here are shown when comparing the force–indentation plot of the conical versus pyramid plot of Figures 14 and 15. The cutting action of the pyramid requires a greater force. While the skin to interface penetration force damage is less for the conical indentor as it separates the fibre, which for the pyramid requires a cutting process.

Figure 14.

Comparison between the conical and pyramid indentor – panel F.

Figure 15.

Comparison of forces between the conical and pyramid indentor at the skin core interface – panel F.

The force–indentation travel for the cylindrical indentor is shown in Figures 9(d) and 16, in which the greatest damage occurred. During the initial stages during increase in force, for all the seven panels, before reaching the local maximum, there are two noticeable damage points occurring on average at 0.6t_s and 1.5t_s, whereas the local maximum occurs within a 2t_s to 3t_s range. The force values at the two damage points and at the local maximum when the skin fractures and starts to penetrate and compress the core is shown in Figure 17. During the initial force increase a best power trend line fits, n = 0.8 confirming the experimental findings that the absorbed energy also follows a power law curve of n = 1.8. Immediately following the local maximum the force drops abruptly, which is when the upper skin is perforated.

Figure 16.

Force (kN), absorbed energy (J) versus indentor displacement (mm) for cylindrical indentor – panel F.

Figure 17.

Force of noticeable damage and local maximum for the cylindrical indentor.

In Figure 17, the forces experienced by all panels are compared with the cylindrical indentor at each of the identifiable damage positions. The skin face is punctured and produces a disc-shaped plate. The force required during this disc-forming process results in the local maximum force far greater than for any other indentation impact scenario, as shown in Figure 18. As the disc advances, the core is compressed. Finally, as the indentor approaches the bottom face and the core is crushed against the rigid back supported by the bottom plate, which does not allow any deflection of the lower skin, the force increases rapidly.

Figure 18.

Graphs of force (kN) versus indentor displacement (mm), panel F – procedure A.

The maximum force comparison required to penetrate the upper skin and up to 80% of the core thickness in the rigidly backed configuration is shown in Figure 19, for the hemispherical, pyramid, conical and cylindrical indentors, respectively.

Figure 19.

Maximum force required to penetrate approximately 80% of the sandwich panel with a rigidly backed configuration – procedure A.

Figure 19 shows the maximum force attained to indent and penetrate the skin thickness. These force values give an indication of the importance of skin thickness, laminate composition and layup. Table 1 gives the combination layups, which required a CSM layer as a final outside layer to ensure a good cosmetic appearance of the finished hull. Considering that these panels and skin layup result from the BS EN ISO 12215-5 design criteria, whereby these panels have to support the hydrostatic and hydrodynamic pressures and have adequate bending stiffness according to span support as dictated in the standard. The maximum force values, for all the indentors generally follow the same trend. Panels D and F have the greatest thickness of 2.03 mm with the 600w middle layer and two 450CSM outer layers. Panel G of thickness 1.76 mm also has five layers; however, the outer 300CSM influences the force values and results in a slightly lower value. Panel E of three layers and thickness 1.22 mm follows. Panel C comprises four layers of thickness 1.35 mm with a 400w, 300CSM and 450CSM/2, followed by the thinnest skin sandwiches A (0.81 mm) and B (0.67 mm) comprising only CSM. The central woven layers play a crucial role of the maximum force attained; however, when designing the hull skins, the designer has to be aware that the woven pattern must not be visible on the finished hull surface. Following this, it is also recommended that the outer skin is thicker than the inner, thus ensuring a greater impact resistance with less overall weight.

Visual damage discussion

Figure 20(a) shows the damage that has occurred as a result of the 12.5 mm diameter hemispherical indentor. The surface surrounding the material, roughly stretching to an additional 10% radius, shows the whitening characteristic of skin delamination, fibre detaching from the resin; however, due to the low resin content due to the vacuum bagging process, this area is very contained around the indentor periphery. The face skin fractures in a very characteristic method of stretching around a hemisphere and producing a cap head. The edges of the skin laminate are clearly seen to have suffered delamination and are pushed into the core material. The indentor is pushed into the panel to approximately half the panel thickness. The sectioned core material readily shows that the core has been compressed, shown by the darker shade of core colour; however, the core has also shown to have sufficient resistance to the indentor, since the panel groves which are resin-filled during the manufacturing process do not show any visual failure. Finally, very little recovery, if any, of the foam is noticed in the time delay between testing and visual inspection.

Figure 20.

Damage sustained on the sandwich panels with a rigidly backed configuration – procedure A. (a) Hemisphere, (b) Conical, (c) Pyramid and (d) Cylindrical.

The conical indentor has, as described above, also produce an outer ring of the whitened skin, as the fibres within this region are detached from the resin (Figure 20(b)). The different diameters visible on the specimens are due to the fact that the panels have various thicknesses and the indentor travel into the material is a function of the panel thickness in order to reach the same 80% through-thickness penetration. The point of the indentor pierces its way through the skin and as it enters the core pulls down the skin opening up the fibres, which eventually fracture in a globally brittle tensile manner. The delaminated fibres push the core below the immediate surface apart as the cross-sectional area increases as the indentor advances. Virtually no core compression immediately surrounding the advancing indentor is noted. Again the resin-filled groove shows no sign of visual failure due to the core’s resistance to the damage.

The pyramid has a pointed tip and sharp cutting edges. The point pierces the face skin and the edges cut and slice through the face. Although delamination is noted in the final cross section, this is due to the fact that the skin turns down onto the core, in fact the turned down skin shows a near plane face indicating the cutting action rather than the fibres being pulled out as described for the hemispherical and conical indentor. The sharp edges are noted both on the skin and in the core (Figure 20(c)).

Figure 20(d) shows the 15 mm diameter cylindrical indentor damage. The face whitened area extends slightly further out than in the previous cases to approximately an additional 25% of the radius. This increase is due to the cutting rim and flat face of the cylinder. The face and rim act as a punch, which forms a skin flat cap and a reasonable smooth edge with minimal fibre filaments considering the aggressive nature of the skin facture. The core below the flat cap is pushed down and the core closed cells are compressed. Interestingly, as the core moves down, the crack path is diverging outwards. A characteristic that is commonly encountered with crack paths are those that are made to advance by pushing with the material collapsing in front of the progressing damage force. This has been reported in sheet material crack path directions [75, 76]. Also the resin-filled groove withstood the advancing flat cap and it can clearly be seen that the cap moved sideways to avoid the stiffened resin-filled region. Finally, the right side below the flat cap the core is seen to be compressed even further and actually breaking and flaking apart.

Quasi-static test results – Procedure B: Simply supported configuration

Figure 21 shows the complete set of curves for all indentors and panels, which were repeated five times. It is readily observed that in procedure B, simply supported configuration graphs are very similar to that of the rigidly backed in Procedure A. In this set up, the panel being simply supported will allow some overall panel deflection, and the differences in the graph as reported for procedure A can be attributed to the support variation.

Figure 21.

Graphs of force (kN) versus indentor displacement (mm) for: (a) hemispherical; (b) conical (c) pyramid; and (d) cylindrical indentors in a simply supported configuration, procedure B, for all panels.

Figure 21(a) for the hemispherical indentors shows that the point of first noticeable damage is also present. However, this damage point occurs within a narrow range of indentation, between t_s and 1.15t_s, while the local maximum force is delayed compared to the rigid backed configuration on an average to between 2.7t_s and 3.2t_s indentation displacement. The contact law for the initial stages of force increase, to the point of first noticeable damage shows a power contact law is applicable, n = 1.3. From this point to the local maximum, a linear trend line fits, n = 1. However, it must be stated that these two equations cannot be replaced by one single linear, n = 1 trend line, as often stated in various literatures [1, 4, 29, 30, 36]. The simply support configuration allows the panel to flex within the circular opening and this produces a reduced stiffness effect. Figure 22 compares the important force values of the rigid backed and simply supported configurations when impact indented with the hemisphere. Due to the absence of flexure in the rigid backed, the first noticeable damage is less, while for the localised maximum higher forces ensue when compared to the simply support configuration.

Figure 22.

Force comparison between hemisphere force indentation for the rigid backed (RB) and simply supported (SS).

Figure 23 describes the indentation of the conical and pyramid indentors, in the simply supported configuration. The pyramid produces higher force–indentation graphs as obtained in the rigid backed consideration, due to the cutting action of the pyramid edges which require a greater force/energy for this process as opposed to the opening mechanism of the conical indentor.

Figure 23.

Force comparison between conical versus pyramid force indentation for the simply supported configuration, procedure B – panel F.

Figures 24 and 25 compare the differences between the rigid backed and simply supported. The rigid back does not allow flexure, therefore as the indentor sharp tip points are pushed through the skin and core, the material cannot move out of the way fast enough and offer more resistance, hence higher force values for the rigid backed tests. Maximum force for this testing has no significance since the force increases linearly with the increases in indentor cross section and indentor displacement.

Figure 24.

Force comparison between the conical and pyramid indentors for the rigid backed (RB) and simply supported (SS).

Figure 25.

Comparison between (a) conical, (b) pyramid force-indentation for the rigid backed (RB) and simply supported (SS) – panel F.

Figure 26.

Comparison between indentation for the rigid backed (RB) and simply supported (SS) – cylindrical, panel F.

Figure 27.

Comparison of forces between the rigid backed (RB) and simply supported (SS) procedures – cylindrical, panel F.

In the simply-supported configuration, the same turn in the initial force–indentation curve is noted and the experimental force contact plot can be described by two trend lines. Similar to the rigid backed, the power law index n = 0.8 describes the initial increase in force; however, there is a similar change in slope and shape of the force curve, a linear n = 1 describes the process up to the end of the test. However, the crossover of these two curves differs than in the previous case and is delayed to 1.1t_s and 1.5t_s for the conical and pyramid indentors, respectively. The absorbed energy follows a 1.8 power law for the full test duration.

Figure 26 and 27 show the comparison of the two methods on the cylindrical indentor. The flexure in the panel of the simply supported configuration is seen in the shift of the graphs showing the delay of the same occurance. The simply supported equally shows that there are two noticeable damage points during the initial stages. During the end stages of the cylindrical tests, the unsupported bottom face flexes and core foam is not as compressed and crushed as with the rigid support, therefore the forces do not rise to the extent of the rigid backed.

Visual observation of damage

The damage caused by the various indentor geometries for procedure B is shown in Figure 28. The simply supported loading condition show very similar findings to the rigidly backed tests. During the tests for both methods, the indentor was displaced by the same distance into the panels. However, during these simply supported tests the panel is able to deflect into the circular opening.

Figure 28.

Damage sustained on the sandwich panels with a simply supported configuration – Procedure B: (a) hemisphere – face and through-thickness damage perforations, panel F; (b) conical – face and through-thickness damage perforations; (c) pyramid – face and through-thickness damage perforations; (d) cylindrical – face and through-thickness damage perforations.

The hemispherical indentation damage again shows the annular whitening, but with the formation of the cap head and the delamination of the fibres into the indent to a lesser extent. The significant difference is that there is no visible core compression. Although the indentor displacement is half the panel thickness, it is readily seen that the hemispherical indentation left behind shows that penetration did not occur to half the thickness.

The conical indentor damage shows the same fibre pull out and tensile brittle fracture after the sharp point pierced its way through the skin. The core compression is only highly localised on the resulting increasing cross-sectional edge layer.

The pyramid point and sharp edges slice through the skin and core. The edges of the skin are fairly straight and clearly show the cut fibres. Whilst the cut fibres push into the core as the indentor advances the core is compressed below the fibre turn down and once again highly localised on the edges which have been cut through.

The flat face and cutting rim of the cylindrical indentor produced the same flat cap, with a few fibre pull outs in the lower fibre glass layers in the upper skin. Again, the flat cap pushes the core down and compresses it, however, to a lesser extent than in the rigid backed configuration. The diverging crack path as described for procedure A is also visible; coincidentally the indentor was pushed through two resin-filled grooves which can clearly be seen to add a stiffening effect and remain reasonably intact even with the unavoidable manufacturing air voids in the resin in the grooves. Due to the panel deflection, there is no presence of the severe crushing of the foam core.

Conclusion

Quasi-static impact testing according to ASTM D7766 and procedures A and B from ASTM D6264 have been completed, whereby different marine sandwich panels of composite skins and closed-cell PVC foam have been tested.

ASTM D7766 directs researchers to test sandwich panels according to ASTM D6264. However, ASTM D6264 focuses on FRP laminates and describes the process of testing and interpretation on these laminates. Response of composite sandwich panels is different from the laminates owing to the overall panel rigidity, constituents and support configurations. ASTM D6264 needs to reflect the differences in these responses as will be highlighted below.

The default and standard indentor is a 12.5 mm diameter hemisphere. Although providing adequate damage to laminates, falls short to describe marine craft indentation impact occurrences. Further, indentor geometries suitable to describe such marine vessel scenarios are: sharp but smooth fibre piercing and opening indentors – conical shaped; sharp pointed and cutting edge pyramids; cylindrical flat faced with a cutting edge – providing suitable ‘standard rocks’. Each standard rock produces a characteristic force-indentation and damage response in both support configurations.

Marine panels only require 50% to 80% panel penetration, since through-thickness penetration may cause foundering of the vessel; subsequent and next stage testing of such panels is normal, such as adopting procedure C of ASTM D7766, i.e. LVI instrumented drop weight testing to the same penetration depths and further compression CAI residual strength testing is required. Continuity of the standards is necessary.

The support plate for procedure B – simply supported is required to be thicker and specific to the panels tested, and this has to be stated in the standard.

Each force–indentation response has been analysed and shown to be different in all cases (however, the conical and pyramid show similar trends, but each provide essential damage information). Noticeable first (and some instances second) damage points and localised maximum points are seen and have been shown to occur are repeatable %skin thickness indentations.

The initial contact force plots together with the experimental absorbed energy are shown to follow power law. The power index for contact of the standard hemisphere is n = 1.3, whereas the conical and pyramid indentors follow, n = 0.8 until face skin penetration and subsequently follow a linear law. Similarly, the cylinder power index is n = 0.8 up to the point of the flat disc being cut in the face sheet.

Forces at important points, such as noticeable damage points and localised maximum must be reported in the context of the indentor causing the damage and the position at which it occurs.

Comparison between the two support configurations has been described and the major differences between the two conditions are described in terms of panel flexure and delayed damage.

Proper understanding of the internal damage can only be achieved by destructive testing of the panel and describing the process in terms of the damaged sustained by the face and core as the specific indentor penetrates through the thickness whilst relating this to the force plot.

Measuring the dent diameter is not an efficient means of describing the history and damage incurred in the panel material, although it has been noted that there is a whitened annular region in very close proximity to the indentor periphery. A depth gauge reveals (as suggested by the standard) that the panel face remains flat even in this minimal region thereby minimising the importance or even the usefulness of measuring the dent diameter.

Footnotes

Funding

This research project was financed by the Malta Council for Science and Technology through the National Research & Innovation Programme 2011 [Grant number: R&I-2011-002]: Fabrication of Advanced Hybrid Composite Sandwich Panels – Testing and Simulation (FACTS).

Conflict of interest

None declared.

References

Abrate

. Impact of composite laminates. Appl Mech Rev 1991; 44: 155–190.

Abrate

. Impact on laminated composites: Recent advances. Appl Mech Rev 1994; 47(11): 517–544.

Abrate

. Modelling of impacts on composite structures. Compos Struct 2001; 51: 129–138.

Abrate

. Impact on composite structures, Cambridge: Cambridge University Press, 2005.

Choi

Downs

Chang

. A new approach towards understanding damage mechanisms and mechanics of laminated composites due to low velocity impact: part 1 - Experiments. J Compos Mater 1991; 25: 992–1011.

Choi

TH-Y

Chang

. A new approach towards understanding damage mechanisms and mechanics of laminated composites due to low velocity impact: part 2 - Analysis. J Compos Mater 1991; 25(8): 1012–1038.

Choi

Chang

. A model for predicting damage in graphite epoxy laminated composites resulting from low velocity point impact. J Compos Mater 1992; 26(14): 2134–2169.

Finn

Springer

. Delaminations in composite plates under transverse static or impact loads. Compos Stuct 1993; 23: 177–204.

Lakshminarayana

. Impact response of laminated composite plates: Rediciton and verification. Compos Struct 1994; 28: 61–72.

10.

Huang

. Predicition of the residual strength of laminated composites subjected to impact loading. J Mater Process Technol 1995; 54: 205–210.

11.

Zheng

Sun

. A double plate finite element model for the impact induced delamaination problem. Compos Sci Technol 1995; 53: 111–118.

12.

Zhou

. Damage mechanisms in composite laminates impacted by a flat ended impactor. Compos Struct Technol 1995; 54: 267–273.

13.

Chun

Lam

. Dynamic response o fully clamped laminate composite plates subjected to low velocity impact of a mass. Int J Solids Struct 1998; 35(11): 963–977.

14.

De Moura

Goncalves

Marques

. Modelling compression failure after low velocity impact onlaminated composites using interface elements. J Compos Mater 1997; 31: 1462–1479.

15.

De Moura

Marques

. Predicition of low velocity impact damage in carbon epoxy laminates. Compos Part A: Appl Sci Manuf 2002; 33: 361–368.

16.

De Moura

Goncalves

. Modelling the interaction between matrix cracking and delamination in carbon expoxy laminates under low velocity impact. Compos Sci Technol 2001; 61: 2069–2074.

17.

Hou

Petrinic

Ruiz

. A delamination criterion for laminated composties underlow impact velocity. Compos Sci Technol 2001; 61: 2069–2074.

18.

Davies

Olsson

. Impact on composite structures. Aeronaut J 2004; 108: 541–563.

19.

Hosur

Adbullah

Jeelani

. Studies on the low velocity impact response of woven hybrid composites. Compos Struct 2005; 67: 253–262.

20.

Reid

Zou

. Modelling damage of multiple delamiantions and transverse matrix cracking in laminated composites dueo to low velocity lateral impact. Compos Sci Technol 2006; 66: 827–836.

21.

Bouvet

Castine

Bizeul

. Low velocity impact modelling in laminate composite panels with discrete inerface elements. Int J Solids Struct 2009; 46(14–15): 2809–2821.

22.

Murthy

Reddy

Harinathareddy

. FRP composite laminate behaviour under low velocity impact load. Int J Mech Prod Eng 2013; 2(1): 102–106.

23.

Lopresto V and Caprino G. Damage mechanisms and energy absorption in composite laminates under low velocity impact loads. In: Dynamic behaviour of composite and sandwich structures, Naples, Italy, pp. 209–289.

24.

Sevkat

Liaw

Delale

. Drop wight impact response of hybrid composites impacted by impactors of various geometries. Mater Des 2013; 52: 67–77.

25.

Mitrevski

Marshall

Thomson

. Low velocity impacts on preloaded GRRP specimens with various impactor shapes. Compos Struct 2006; 76: 209–217.

26.

Mitrevski

Marshall

Thomson

. The influence of impactor shape on the damage to the composite laminates. Compos Struct 2006; 76: 116–122.

27.

Mitrevski

Marshall

Thomson

. The effect of impactor shape on the impact response of composite laminates. Compos Struct 2005; 67: 139–148.

28.

Abrate

. Localized impact on sandwich structures with laminated facings. Appl Mech Rev 1997; 50(2): 69–82.

29.

Olsson

. Engineering method for prediction of impact response and damage in sandwich panels. J Sandwich Struct Mater 2002; 4: 3–29.

30.

Chai

Zhu

. A review of low-velocity impact on sandwich structures. Proc IMechE, Part L: J Materials: Design and Application 2011; 225: 207–230.

31.

Hazizan

Cantwell

. The low impact response of foam based sandwich structures. Compos Part B: Eng 2002; 33(3): 193–204.

32.

Wen

Reddy

Reid

. Indentation, penetration and perforation of composite laminates and sandwich panels under quasistatic and projectile loading. Key Eng Mater 1998; 141–143: 501–552.

33.

Borsellino

Calabrese

Di Bella

. Windsurf board sandwichpanels under static indentation. Appl Compos Mater 2008; 15: 75–86.

34.

Hasebe

Sun

. Performance of sandwich structures with composite reinforced core. J Sandwich Struct Mater 2000; 2: 75–100.

35.

Foo

Chai

Seah

. A model to predict low-velocity impact response and damage in sandwich composites. Compos Sci Technol 2008; 68: 1348–1356.

36.

Abot

Daniel

Gdoutos

. Contact law for composite sandwich beams. J Sandwich Struct 2002; 4: 157–173.

37.

Sadighi

Pouriayevali

. Quasi-static and low-velocity response of fully backed or simply supported sandwich beams. J Sandwich Struct Mater 2008; 10: 499–524.

38.

Gustin

Mahinfalah

Nakhaie Jazar

. Low velocity impact of sandwich composite plates. Exp Mech 2004; 15: 75–86.

39.

Williamson JE. Response mechanisms in the impact of graphite/expoxy honeycomb sandwich composite plates. Master of Science Thesis, MIT, USA, 1989.

40.

Wang J, Waas AM and Wang H. Experimental study on the lowvelocity impact behaviour of foan core sandwich materials. In: AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics and materials conference, Hawaii, pp. 1–13.

41.

Xuefeng

. Comparison with low velocity impact and quasi static indentation testing of foam core sandwich composites. Int J Appl Phys Math 2012; 2(1): 58–62.

42.

Zheng

. Low velocity perfforation behaviour of composite sandwich panels with aluminium foam core. J Sandwich Struct Mater 2012; 15(1): 92–109.

43.

Sutherland

Guedes Soares

. The use of quasi-static testing to obtain the low velocity impact damage resistance of marine GRP laminates. Compos Part B: Eng 2012; 43(3): 1459–1467.

44.

Sutherland

Guedes Soares

. Scaling of imapct on low fibre volume glass polyester laminates. Compos Part A: Appl Sci Manuf 2007; 38(2): 307–317.

45.

Sutherland

Guedes Soares

. Contact indentation of marine composites. Compos Struct 2005; 70: 287–294.

46.

Sutherland

Guedes Soares

. Impact on low fibre volume, glass/polyester rectangular plates. Compos Struct 2005; 68: 13–22.

47.

Sutherland

Guedes Soares

. Impact behaviour of typical marine composite laminates. Compos Part B: Eng 2006; 37: 89–100.

48.

Sutherland

Guedes Soares

. Impact characterisation of low fibre volume glass reinforced polyester circular plates. Int J Impact Eng 2005; 31: 1–23.

49.

Sutherland

Guedes Soares

. Effect of laminate thickness and of matrix resin on the impact of low volume fibre, woven rovening E-glass composites. Compos Sci Technol 2004; 64: 1694–7000.

50.

Sutherland

Guedes Soares

. Effects of laminate thickness and reinforcement type on the impact behaviour of E-glass/polyester laminates. Compos Sci Technol 1999; 59: 2243–2260.

51.

Sutherland

Guedes Soares

. Impact tests on woven roving E-glass/polyester laminates. Compos Sci Technol 1999; 59: 1553–1567.

52.

Choquese D, Baizeau R and Davies P. Experimental studies of impact on marine composites. In: Proceedings of ICCM12, Paris, 1999. Paper 193.

53.

Baral

Cartie

Partridge

. Improved impact performance of marine sandwich panels using through thickness reiinforcement: Experimental results. Compos Part B: Eng 2010; 41: 117–123.

54.

Collombet F, Guillaumat L, Lataillade JL, et al. A study of impacted composite structures by means of the response surface methodology. In: Proceedings of ICCM12, Paris, 1999. Paper 227.

55.

Davis P. Scale and size effects in the mechaincal characterisation of composite and sandwich materials. In: Proceednigs of ICCM12, Paris, 1999. Paper 1033.

56.

Atas

Sevim

. On the impact reponse of sandwich composites with cores of balsa wood and PVC foam. Compos Struct 2010; 93: 40–48.

57.

Charca

Shafiq

Just

. Repeated slamming of sandwich panels on water. J Sandwich Struct Mater 2009; 11: 409–424.

58.

Charca

Shafiq

. Damage assessment due to single slamming of foam core sandwich composites. J Sandwich Struct Mater 2010; 12: 97–112.

59.

Charca

Shafiq

. Damage assessment due to repeated slamming of foam core sandwich composites. J Sandwich Struct Mater 2011; 13: 97–109.

60.

De Marco Muscat-Fenech C, Cortis J, Buhagiar R, et al. Impact and grounding damage resistance analysis of marine grade sandwich hull panels. In: ICCS17, 17th international conference of composite structures, 17–21 June, Portugal, 2013.

61.

De Marco Muscat-Fenech C and Cortis J. Fabrication of advanced hybrid composite sandwich panels - testing & simulation – FACTS. Malta: University of Malta and Malta Council for Science and Technology, R&I-2011-002, 2013.

62.

Cortis J, De Marco Muscat-Fenech C and Cassar C. Hybrid composite hull panels fabricated using vacuum bagging techniques. In: NAV2012 international conference on ships and shipping research, Naples, Italy, 2012.

63.

De Marco Muscat-Fenech CM and Cortis J. Fabrication of advanced hybrid composite sandwich panels - testing & simulation - FACTS - 3. Strength testing of the fabrication proces. Malta: University of Malta and Malta Council for Science and Technology, R&I-2011-002-WP3, 2013.

64.

British Standards Institution. Small craft - Hull construction and scantlings - Part 5: Design pressures for monohulls, design stresses, scantlings determination. International Organisation for Standardisation, 2008. BS EN ISO 12215-5:2008.

65.

De Marco Muscat-Fenech CM and Cortis J. Fabrication of advanced hybrid composite sandwich panels - testing & simulation - FACTS - 6. Analytical evaluation. Malta: University of Malta and Malta Council for Science and Technology, R&I-2011-002-WP6, 2013.

66.

DNV. Rules for classification of high speed, light craft and naval surface craft - Part 3 Chapter 1, Design principles, design loads. DNV.

67.

British Standard Institution. Plastics film and sheeting. Determination of impact resistance by the free-falling dart method. Stair case methods. London: BIS, 2004. BS EN ISO 7765-1:2004.

68.

British Standards Institution Plastics. Determination of puncture impact behaviour of rigid plastics - Part 1: Non-instrumented impact testing. London: BIS, 2000. BS EN ISO 6603-1:2000.

69.

British Standards Institution. Plastics. Determination of multi-axial impact behaviour of rigid plastics - Part 2: Instrumented impact testing. London: BIS, 2001. BS EN ISO 6603-2:2000.

70.

ASTM. Standard practice for damage resistance testing of sandwich constructions. Philadelphia: American Society for Testing and Materials, 2011. ASTM D7766/D7766M-11.

71.

ASTM. Standard test method for measuring the damage resistance of a fiber-reinforced polymer-matrix composite to a concentrated quasi-static indentation force. Philadelphia: American Society for Testing and Materials, 2004. ASTM D6264-98.

72.

ASTM. Standard test method for measuring the damage resistance of a fiber reinforced polymer matrix composite to a drop weight impact event. Philadelphia: American Society for Testing and Materials, 2005. ASTM D7136/D7136M-05.

73.

De Marco Muscat-Fenech CM and Cortis J. Fabrication of advanced hybrid composite sandwich panels - testing & simulation - FACTS - 5. Testing of impact scenarios. Malta: University of Malta and Malta Council for Science and Technology, R&I-2011-002-WP5, 2013.

74.

ASTM. Standard test method for compressive residual strength properties of damaged polymer matrix composite plates. Philadelphia: American Society for Testing and Materials, 2007. ASTM D&137/D7137M-07.

75.

Muscat-Fenech C. The tearing of ships’ pating upon grounding (SHP155). Part 1 – ‘Blunt’ underwater obstacles. University of Reading, UK, 1996. EPSRC/MTD/MoD. (No. GR/J42014).

76.

Muscat-Fenech

Atkins

. Denting and fracture of sheet steel by blunt and sharp obstacles inglancing collisions. Int J Impact Eng 1998; 21: 499–519.