Explosive Blast Response of Marine Sandwich Composites

Abstract

This chapter presents an experimental investigation on the explosive blast response of four types of sandwich composites commonly used in naval ships, including hull and topside structures. Air blast tests were performed using plastic explosive charges on square target sandwich plates made of carbon or glass fiber-polymer facesheets with either an end-grain balsa or closed-cell PVC foam core. The sandwich composites were dynamically loaded by air shock waves of increasing pressure and impulse generated by plastic explosive charges, and the deformation, damage and post-blast mechanical properties were determined. The amount of out-of-plane deformation and blast-induced damage for a given shock wave impulse depends on both the type of fiber reinforcement used in the facesheet laminates and the type of core material. For a given shock wave impulse, glass sandwich composites have higher resistance to deflection and blast-induced delamination cracking and tow rupture then carbon fiber sandwich composites. Also, the balsa core composites are more resistant to out-of-plane deformation then the PVC core composites at low blast impulses. Damage initiates in the balsa core composites as core cracking and splitting, and facesheet-core debonding, whereas damage in the PVC core composites initiated as front facesheet compressive failure. Blast-induced damage to the skins and core causes large reductions to the mechanical properties of the sandwich composites.

1 Introduction

Fiber-reinforced sandwich composites are used in a wide variety of light-weight military assets at risk of attack using explosive munitions. These assets include armoured vehicles, tanks, ships, submarines, aircraft, helicopters and drones. The types of sandwich composite materials used in naval ship structures can consist of either fiberglass or carbon fibre-polymer laminate facesheets covering a thick core of polymer foam or balsa wood. A concern with sandwich composites instead of metals (e.g. steels, aluminium alloys) in naval vessels is their brittle-like response under the high strain rate loading caused by the shock wave. Metals often undergo large plastic strain deformation which absorbs a large amount of shock wave energy before catastrophic failure. The laminate facesheets do not usually plastically deform which results in brittle-type damage such as matrix cracking, delamination and fiber fracture. The core material to sandwich composites may plastically deform and crush under the impulse load exerted by a shock wave.

The deformation and damage response of sandwich composites due to explosive blast loading has been studied in detail [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17]. Researchers have identified that the explosive blast response of sandwich composites is largely governed by their geometry (facesheet-core thickness ratio etc.) [18,19,20], density of the core material [11, 14, 15, 22], and properties of the individual material constituents used in the laminate facesheets and core [11, 12, 14, 15, 21, 22].

It is known that sandwich composites fail under several damage modes when subjected to an explosive blast. Damage initiation due to blast is most commonly in the form of facesheet delamination [2, 3, 7,8,9, 14, 26], facesheet-core debonding [4, 5, 10, 15, 22, 23], core crushing [1, 5, 6, 10, 15, 21, 24] and core cracking [4, 5, 8,9,10,11, 22, 25, 26]. Failure via core fragmentation [7, 14], core disintegration [1,2,3, 24] and tearing failure at the supports [11] have also been reported.

Majority of the studies have tested the explosive blast response of sandwich composites with glass fiber facesheets [2, 3, 7, 14] and polymer foam core [4, 5, 8,9,10, 26]. Little published research is available on the explosive blast resistant properties of sandwich composites containing carbon fiber laminate facesheets or end-grain balsa wood core – which are both commonly used materials in naval structures.

The effect of the laminate facesheet material and the core material on the explosive blast resistance of sandwich composites commonly used in naval ship structures is investigated. The sandwich composites used have facesheets made of woven E-glass or woven carbon fiber reinforced vinyl ester laminate. The core materials were closed-cell PVC foam and end-grain balsa wood. The facesheet (2 mm thick) and core thickness (6 mm) was the same to provide a direct comparison of the blast response of the different types of sandwich composites. Air blast tests were performed on the sandwich composites under three shock wave conditions by varying the mass or stand-off distance of the explosive charge. The dynamic deformation, amount and types of damage, and post-blast mechanical properties are compared for the sandwich materials.

2 Sandwich Composite Materials and Experimental Methodology

2.1 Fabrication of Sandwich Composite Materials

The response to explosive blast loading of four types of sandwich composite commonly used in ship structures was evaluated:

• woven carbon fiber-end grain balsa core composite

• woven carbon fiber-PVC foam core composite

• woven glass fiber-end grain balsa core composite

• woven glass fiber-PVC foam core composite.

Top and side-views of these sandwich materials are shown in Fig. 1. The core material was 6 mm thick and was either end-grain balsa (Baltek SB) with a bulk density of 150 kg/m³ or Divinylcell PVC foam (Diab Group) with a density of 130 kg/m³. Both core materials were gently sanded and cleaned prior to manufacturing to ensure effective bonding to the laminate facesheets. The carbon and glass used in the facesheets were both single ply plain woven fabric (areal density of 600 g/m²). The carbon and glass fabrics were stacked on both sides of the sandwich core, with the warp tows aligned in the same direction, giving a cross-ply fiber [0/90] pattern for the facesheets. The dry fabric to the facesheets were impregnated with liquid vinyl ester resin at room temperature using vacuum bag resin infusion (VBRI) in a one-step process (i.e. there was no secondary process to bond the laminate facesheets to the core). The vinyl ester was catalysed using 0.8 wt% MEKP solution (40 wt% MEKP in dimethyl phthalate) (Norox from Nuplex Composites). Due to the closed cell structure of both the PVC and balsa wood, the vinyl ester did not penetrate the core during the VBRI process.

Fig. 1

Top and side views of the sandwich composite panels for blast testing

Following the VBRI process, the sandwich composites were allowed to gel and partially cure at 23 °C for 1 day and were then post-cured at 80 °C for 1 h. The glass and carbon fiber facesheets were about 2 mm thick each, which resulted in an overall thickness of about 10 mm for the sandwich composites.

2.2 Explosive Blast Testing of Sandwich Composite Materials

The blast response of the sandwich composites was performed using an experimental facility operated by the Defence Science and Technology Group (Australia) that consists of an enclosed steel plate-lined concrete chamber fitted with viewing windows to observe the explosion and the dynamic deformation of the targets (Fig. 2). The sandwich composite targets were flat 275 mm × 275 mm square panels held within a steel window frame having a 250 mm × 250 mm aperture. The frame was lined with soft rubber which allowed the target sandwich plate to bend freely under the pressure exerted by the shock wave without causing edge clamping damage.

Fig. 2

Explosive blast test chamber

A schematic of the explosive blast experimental setup is shown in Fig. 3. The window frame holding the target panel was bolted to a 1.075 m high steel stand that was placed in the center of the chamber. The out-of-plane deformation of the sandwich panels was captured using high-speed photography and measured using 3D DIC software ARAMIS. Two Photron SA5 cameras were positioned 1.7 m behind the target, at an angle of 22.5° from the center of the panel. 32 mm thick blast and fragment resistant flexi-glass was used in the viewing ports to protect the cameras from damage. A PF300 slow peak flashbulb was used to illuminate the rear of the target panel to capture the deformation. During the explosive blast testing, the SA5 cameras were operated at a frame rate of 7000 per second capturing the out-of-plane deformation of the target. A third Photron SAZ camera was positioned 1.7 m at an angle of 67.5° from the center of the plate. The SAZ camera was operated at frame rates ranging between 12,000–30,000 per second, depending on the proximity of the charge to the target. Near-field blasts required higher camera frame rate to adequately capture the explosive-panel interactions. Time-lapse images of the explosive detonation, fireball and shock wave interaction with the sandwich plate for far- and near-field tests are shown in Figs. 4 and 5, respectively.

Fig. 3

Schematic representation of the explosive blast test setup

Fig. 4

Time-lapse photographs of a far-field explosive blast event. The times after detonation are given

Fig. 5

Time-lapse photographs of a near-field explosive blast event. The times after detonation are given

The blast was generated using a spherical plastic explosive Type 4 (PE4) charge made of RDX (cyclotrimethylenetrinitramine). The explosive was fired using a 3.8 g RP-80 EBW electric detonator. The peak overpressure and impulse of the shock wave was controlled by varying the explosive charge mass and stand-off distance from the sandwich panel. The overpressure-time response of the incident and reflected shock waves were measured using two free-field pressure transducers (Kulite XTL-190). The near-field blast pressures and impulses were predicted using a two-dimensional, axially symmetric coupled Euler-Lagrangian model in ANSYS AutoDyn 16.2. The explosive blast overpressures and impulses used in this study are given in Table 1. One sandwich composite sample was tested at each blast test condition.

Table 1 Conditions used for explosive blast testing of the sandwich composites. The symbol ∗ indicates the blast conditions when the peak pressure and impulse were calculated because they could not be precisely measured using pressure transducers

2.3 Mechanical Property Testing of Sandwich Composite Materials

The flexural properties were determined for the laminate facesheet materials. The stress-strain curves are shown in Fig. 6, and indicate that the response of both facesheet materials was linearly elastic followed by progressive failure at approximately 450 MPa and 650 MPa for the carbon and glass fiber laminates, respectively. The flexural modulus of the carbon fiber facesheet was higher than the glass fiber facesheet, although the flexural failure stress, failure strain and strain energy density (area under the stress-strain curve) are substantially lower.

Fig. 6

Flexural stress-strain curves for the carbon- and glass-vinyl ester laminate facesheets

Three-point flexure tests were also performed on long, slender beams of the sandwich composites according to ASTM C393 [27] and D7249 [28] specifications. Bending tests were performed on 20 mm wide rectangular coupons of the sandwich materials. A span-to-thickness ratio of 16:1 was used, and therefore testing involved a support span of 160 mm and a total beam length of 200 mm. The coupons were loaded at a displacement rate of 5 mm/min until failure. The relation for the calculation of facesheet failure stress in three-point bending is given by:

$$ {\sigma}_{facing}=\frac{P_{max}S}{4 bt\left(d-t\right)} $$

(1)

where P_max is the maximum force, S is the support span length, t is the facesheet thickness, d is the sandwich thickness, and b is the specimen width. The core shear failure stress under three-point bending was calculated using:

$$ \tau =\frac{P_{max}}{\left(d+c\right)b} $$

(2)

Flexural stress-strain curves measured for the sandwich materials are shown in Fig. 7. The balsa core sandwich composites exhibited linear elastic behavour up to a peak stress of ~120 MPa and ~170 MPa for the glass fiber- and carbon fiber-balsa sandwich composites, respectively. Damage initiated as core shear failure at the supports, resulting in a gradual reduction of the applied load. The shear cracking in the core propagated to the bond line, where the balsa core composite failed abruptly due to facesheet-core debonding. The PVC core sandwich composites initially displayed linear elastic behavour, followed by a large non-linear region due to progressive compression failure of the core. The PVC sandwich beams ultimately failed in compression of the facesheet under the central loading point. The test results in Table 2 show that the carbon-balsa sandwich composite had the highest face sheet and core shear strengths. In general, the flexural facing and core shear strength properties of the PVC foam sandwich composites were lower than the balsa core sandwich composites. Although, the highest flexural strain energy density (area under stress-strain curve) was measured for the glass fiber/PVC sandwich composite.

Fig. 7

Flexural stress-displacement curves for the (a) balsa-core and (b) glass fiber sandwich composites

Table 2 Flexural properties of sandwich composites

The mechanical properties measured for the sandwich composites are given in Table 2.

3 Results and Discussion

3.1 Blast-Induced Deformation of Sandwich Composite Materials

The dynamic deformation response of the sandwich composites when impulsively loaded by the shock wave and detonation products (in the case of near-field blast tests) was measured using the high-speed DIC technique. Typical DIC generated images showing the displacement and surface strain maps of the back surface of a target panel are shown in Fig. 8. In this case, the panel was tested using a 100 g PE4 explosive charge at 0.4 m stand-off distance, which generated a blast impulse of ~350 Pa.s. The images are shown directly after the explosive was detonated (0.1 ms), and at the point of maximum center-point deflection (2.1 ms). The impulse from the explosive detonation caused the panel to deform in the direction of the shock wave. As expected, the point of maximum deflection was in the center of the plate.

Fig. 8

DIC generated images for the carbon-balsa sandwich composite tested at a 353 Pa.s blast impulse. The times following detonation are given. The upper and lower images show the out-of-plane displacement and strain, respectively

The center-point deflection versus time curves measured for the glass fiber- and carbon fiber-balsa sandwich composites are compared in Fig. 9. When subjected to a ~350 Pa.s impulse, the sandwich composites rapidly deformed to reach the maximum center-point deflection. The maximum positive displacements were similar for both materials, however the glass fiber composite deformed ~10% less than the carbon. Also, the maximum negative deformation was also less for the glass fiber composite. This was due to the development of front facesheet damage in the carbon panel, whereas no facesheet damage occurred in the glass panel. At a higher blast impulse (~470 Pa.s), the maximum center-point deflection of the carbon fiber-balsa sandwich composite was much higher (~50%) than the glass fiber-balsa composite. Also, the residual deformation of the carbon fiber-balsa sandwich composite was higher, which was due to the more severe damage experienced under blast loading (as described later). The same result was found for the glass fiber- and carbon fiber-PVC sandwich composites subjected to ~470 Pa.s shock wave impulse, as shown in Fig. 10.

Fig. 9

Comparison of the center-point deflection versus time histories for the glass fiber- and carbon fiber-balsa sandwich composites subjected to (a) 348 Pa.s and (b) 472 Pa.s blast impulse

Fig. 10

Comparison of the center-point deflection versus time histories for the glass fiber- and carbon fiber-PVC composites subjected to 472 Pa.s blast impulse

The center-point deflection versus time curves of the carbon fiber-PVC and balsa core sandwich composites are compared in Fig. 11. When subjected to a ~350 Pa.s blast impulse, both carbon sandwich panels reached maximum out-of-plane displacement at approximately the same time (~1 ms post detonation). However, the maximum center-point displacement of the carbon fiber-balsa composite was lower than the carbon fiber-PVC composite. This is expected given the higher stiffness of the balsa core composite. Both panels suffered from front facesheet tearing at the edges, however this damage was more extensive in the PVC core composite. The PVC foam compressed under the force exerted by the shock wave, which limited the load transferred to the back facesheet. Similar behavour was observed in the PVC foam composite under three-point bend loading (substantial core compression resulting in failure of the top facesheet). At a higher blast impulse (~470 Pa.s), the deflection response and maximum center-point displacement of the carbon balsa and PVC sandwich composites were similar. Despite the higher flexural stiffness of the balsa core panel, the amount of damage sustained by the core reduced the panels rigidity during deformation.

Fig. 11

Comparison of the center-point deflection versus time histories for the carbon fiber-balsa and PVC sandwich composites subjected to (a) 348 Pa.s and (b) 472 Pa.s blast impulse

The center-point deflection versus time curves of the glass fiber-PVC and balsa core sandwich composites are compared in Fig. 12. When subjected to a 472 Pa.s blast impulse, both sandwich composites reached their maximum out-of-plane deflection at approximately the same time (~0.7 ms post detonation). However, similar to that found for the carbon composites, the balsa-core sandwich composite had a lower maximum displacement than the PVC-core composite owing to its higher bending stiffness. Comparatively, at the highest shock wave impulse (500 Pa.s), both sandwich materials reached similar maximum out-of-plane deflections (~40 mm). The glass-balsa composite reached maximum out-of-plane deflection 1.7 ms post detonation, which was later than the glass-PVC composite (~0.86 ms). Also, the glass-balsa composite failed to return to its initial starting position, both of which were representative of extensive damage and permanent deformation.

Fig. 12

Comparison of the center-point deflection versus time histories for the glass fiber-balsa and PVC sandwich composites subjected to (a) 472 Pa.s and (b) 500 Pa.s blast impulse

The effect of increasing blast impulse loading on the maximum center-point displacements for the different sandwich panels is shown in Fig. 13. The maximum center point deflection of the carbon fiber sandwich composites was greater than the glass fiber composites. Also, up to the blast impulses of 353 Pa.s and 472 Pa.s for the carbon and glass, respectively, the maximum out-of-plane displacement of the PVC core composites was higher than the balsa core composites. When subjected to greater respective blast impulses, the maximum displacements of the balsa and PVC core composites were similar due to the extensive damage to the balsa core.

Fig. 13

Effect of blast impulse on the maximum center-point displacement of the sandwich composites

3.2 Blast-Induced Damage to Sandwich Composite Materials

Following blast testing, the sandwich composites were examined using visual inspection, through-transmission ultrasonics and X-ray computed tomography to determine the amount and types of damage. A front-view (blast side) comparison of the sandwich panels for all blast test conditions is shown in Fig. 14. The regions highlighted with red denote the location of through-thickness rupture of the front facesheet. Front facesheet rupture in the carbon sandwich composites initiated at a lower blast impulse compared to the glass sandwich panels. This was consistent with the recent study performed by Gargano et al. [29] which revealed that fiberglass laminates are more blast resistant than carbon fiber laminates. Therefore, for the sandwich composites, the higher flexural failure strain and flexural strain energy density of the glass fiber facesheet (given in Table 2) resulted in the fiberglass laminate facesheets withstanding larger out-of-plane displacements without damaging. The results in Fig. 14 also show that facesheet damage was influenced by core type, with the PVC-core composites having greater amounts of front facesheet cracking than the balsa-core composites for a given blast impulse.

Fig. 14

Comparison of blast-induced surface damage to the sandwich composites. Note the red markings represent front facesheet cracking. The images show the front facesheet of the sandwich panels

Figure 15 presents C-scan ultrasound images of the sandwich composites after blast testing. The images show the facesheet-core interface on the side that was exposed to the explosive blast. The dark-blue ‘sparkle’ region in the C-scan images for the balsa core represent regions where the facesheet has debonded, which is a common failure mode of sandwich composites exposed to shock wave loading [7,8,9, 14, 22]. The effect of increasing blast impulse loading on the amount of front facesheet-core debonding damage sustained by the balsa-core composites is summarised in Table 3. The percentage debonded area was measured from the C-scan ultrasound images shown in Fig. 15, and was defined as a percentage of the total surface area of the sandwich target. The C-scan images were analysed via a colour contrast technique using Adobe Photoshop CC software to determine the facesheet-core debonding area. After testing at the 219 Pa.s blast impulse, the majority of the front facesheet-core interface of the carbon fiber-balsa composite was undamaged. At a blast impulse of ~350 Pa.s, both the glass- and carbon-balsa composites exhibited substantial front facesheet-core debonding, however it was more extensive in the carbon-balsa composite than in the glass-balsa composite. Damage types such as front and rear facesheet fiber failure, longitudinal shear failure of the balsa core and cracking and splitting of the balsa were found in both the carbon- and glass-balsa sandwich composites (as shown for example in Fig. 16(a)). However, the magnitude of the blast impulse needed to cause these damage types in the glass-balsa composite was higher than the carbon-balsa composite. Similar observations were made by Rolfe et al. [10], who found that a carbon fiber sandwich panel had substantially greater amounts of debonding damage compared to the glass fiber sandwich panel when subjected to the same explosive blast load.

Fig. 15

Effect of increasing blast impulse on ultrasonic C-scan images for the sandwich composites. The images show the front skin-core interface

Table 3 Effect of increasing blast impulse loading on the debonding damage area to the balsa-core sandwich composites

Fig. 16

Side-on X-ray computed tomography images of the damage to the (a) carbon-balsa and (b) carbon-PVC sandwich composites subjected to 472 Pa.s blast impulse

For the carbon-balsa composite subjected to a 353 Pa.s blast impulse, the center region remained bonded to the front facesheet. This was believed to be due to compaction of the front facesheet and core when impulse loaded by the shock wave. Similar damage patterns have been observed in carbon and glass fiber sandwich panels using X-ray computed tomography; extensive debonding around the edges with core crushing and compaction in the center [10]. Andrews et al. [30] calculated failure mode maps for sandwich panels with PVC and balsa wood cores, whereby based on the core type, core density and facesheet thickness-length ratio the expected damage modes were calculated (either face sheet wrinkling, face sheet failure or core shear failure). For a core density of 130–150 kg/m³ and a facesheet thickness-length ratio of 0.007, it would be expected that balsa core composites fail by core shear at the supports when subjected to air blast loading. Core shear failure would then propagate to facesheet-core debonding once the cracks reached the bondline, as found previously from the three-point bend flexural tests. This failure map confirms the debonding pattern observed around the edges of the blast loaded balsa core panels.

The same damage pattern (debonding around the boundaries with a center region intact) was observed for the carbon- and glass-balsa composites subjected to a 472 Pa.s blast impulse. Also, at this shock wave impulse, the amount of facesheet-core debonding was similar for both the glass- and carbon-balsa sandwich composites. Unexpectedly, at the highest blast impulse (500 Pa.s), the debonding area in the glass-balsa composite panel was less than that of glass-balsa panel tested at 472 Pa.s blast impulse. It was possible that the region of compaction could increase with larger charge sizes due to the larger shock wave radius generated from a bigger charge. However, this is difficult to confirm considering only one panel was tested per explosive blast test condition. Further experimental tests would be needed to confirm this finding.

From the ultrasound images shown in Fig. 15, damage to the facesheet-core interface from explosive blast loading could not be determined for the PVC composites. The cells of the PVC foam are air-filled, which has a similar acoustic impedance value as debonding damage thereby making it difficult to isolate the debonding damage from the air in the PVC cells. Damage such as front and rear facesheet fiber failure, facesheet delamination, longitudinal shear and fragmentation failure of the core were instead observed using X-ray computed tomography, as shown in Fig. 16. Similar damage modes were observed by Langdon et al. [7, 14], who subjected glass fiber-PVC sandwich composites to localised air blast loading. The damage types occurred regardless of the type of fiber reinforcement (glass or carbon). However, the blast impulse to initiate damage in the carbon-PVC composite was lower than the glass-PVC, similar to the results for the balsa-core composites. This was due to the glass fibers having greater energy absorption capacity than the carbon fibers, as mentioned previously.

Figure 17 shows X-ray computed tomography images taken at different through-thickness locations of the carbon-balsa and carbon-PVC sandwich composite plates after loaded by a 472 Pa.s shock wave impulse. The dimensions below each image represent the through-thickness distance with respect to the top (blast side) surface (i.e. 0 mm is the top surface, 5 mm is directly in the middle and 10 mm is the rear surface). Both materials experienced fiber rupture on the front and rear facesheets. However, damage sustained by the balsa core was significantly greater than the PVC core. Large portions of the balsa core were fragmented throughout the core thickness, and thin cracks had propagated perpendicular to the grain direction, as shown in Fig. 18. Under the same blast impulse, the PVC-core in the carbon-PVC sandwich composite sustained localised core fragmentation. The greater damage in the balsa core composites was believed to be due to the balsa core’s lower shear ductility compared to the PVC, which has been previously observed to be a weakness in balsa sandwich composite design when subjected to blast loads [11]. Another consideration is that the greater amount of damage in the balsa core composites was due to the weaker face/core interfacial bond strength. Truxel et al. [31] compared the face/core fracture resistance of sandwich panels with glass fiber-vinyl ester facesheets with either end-grain balsa or PVC foam cores using various surface preparation techniques. They found that when specimens were prepared with no infusion aid (i.e. no grooves or interface mat) the glass fiber-PVC face/core interface was ~50% higher than the glass fiber-balsa face/core bond strength.

Fig. 17

Through-thickness X-ray computed tomography images of (a) carbon-balsa and (b) carbon-PVC sandwich composites subjected to 472 Pa.s blast impulse. The numbers below each image represent the distance from the top surface with 0 mm, 5 mm and 10 mm for the top, middle and bottom of the panel

Fig. 18

X-ray computed tomography image showing longitudinal shear cracking between the balsa wood grains. This image is taken for carbon fiber-balsa sandwich composite after subjected to a 472 Pa.s blast impulse

Following post-blast damage assessment, the residual flexural properties of the sandwich composites were measured, and the results are shown in Figs. 19 and 20. Five flexural coupons were cut from each blast panel. The coupons were cut from the same locations across all blast panels to ensure the strength retention comparisons between materials was consistent. When subjected to the lowest intensity blast (219 Pa.s), the flexural ultimate facing stress decreased by 13% and 10% for the carbon-balsa and carbon-PVC composites, respectively. Similar reductions to the core shear strength were also recorded. The reduction in strength was expected to be due to fiber-matrix interfacial cracking, transverse inter-tow cracking and matrix cracking in the carbon facesheets, as previously found to occur in carbon laminates subjected to 219 Pa.s blast impulse load [29]. The residual flexural properties of the carbon sandwich composites were significantly reduced after loaded by a 353 Pa.s shock wave impulse. The facing strength and core shear strength were both reduced, on average, by 75% and 50% for the carbon-balsa and carbon-PVC, respectively. However, the magnitude of the strength properties of both carbon sandwich composites is similar, and statistically there is no difference between in the structural integrity of PVC and balsa core carbon composites when subjected to any shock wave impulse tested in this study. At the highest blast impulse tested on the carbon composite (472 Pa.s), the flexural properties were similar to those after the 353 Pa.s blast impulse.

Fig. 19

Post-blast (a) facing strength and (b) core shear strength of the carbon fiber-balsa and carbon fiber-PVC sandwich composites. The error bars represent one standard deviation

Fig. 20

Post-blast (a) facing strength and (b) core shear strength of the glass fiber-balsa and glass fiber-PVC sandwich composites. The error bars represent one standard deviation

For the lowest blast impulse tested on the glass fiber sandwich composites (353 Pa.s), the ultimate facing stress decreased by 63% for the glass-balsa composite, whereas there was no change for the glass-PVC composite. Similar results were observed for the ultimate core shear strength, with a 64% and 5% decrease for the glass-balsa and glass-PVC composites, respectively. The reduction in ultimate facing stress and core shear strength for the glass-balsa composite was due to significant damage sustained to the balsa core as mentioned previously. Similar observations have been reported by Tagarielli et al. [11], in that the sandwich beams with end-grain balsa wood cores failed at a lower blast impulse than the PVC core composite, despite being the better performing sandwich construction at low projectile impulses. The balsa-core composites suffered less front facesheet compressive failure then the PVC core composites, yet much more extensive core damage. The post-blast flexural results show that the core damage in the balsa-core composites was much more detrimental to the structural integrity of the sandwich construction than the front facesheet damage. However, when the shock wave impulse is high enough (>450 Pa.s), there is no significant difference in the post-blast flexural strength properties of the glass-fiber PVC and balsa core composites.

4 Conclusion

The explosive blast resistance of sandwich composites used in naval ship structures is influenced by both the type of fiber reinforcement used in the laminate facesheets and the type of core material. For the explosive blast test conditions used in this work, the resistance to out-of-plane deformation was higher for the glass fiber sandwich panels than the carbon fiber sandwich panels, and the balsa core composites were superior to the PVC core composites at low blast impulses. The blast impulse required to initiate front facesheet failure in the PVC composites was lower than the balsa composites. However, above a threshold impulse of ~350 Pa.s, the resistance of the balsa core sandwich composites to out-of-plane deformation was heavily reduced due to extensive core cracking and splitting, which developed into facesheet-core debonding over the majority of the panel. This resulted in large reductions to the flexural facing and core shear strengths of the balsa core composites. Also, the reduction in flexural properties was similar between glass fiber and carbon fiber balsa core sandwich composites as the threshold impulse to cause extensive core damage and facesheet-core debonding was the same (353 Pa.s). However, the impulse required to initiate damage in the carbon sandwich composites was lower than the glass sandwich composites, which was due to the lower energy absorption capacity of the former.