Mechanical Behavior of Sandwich Panels with Hybrid PU Foam Core

+e traditional composite sandwich structures have disadvantages of low shear modulus and large deformation when used in civil engineering applications. To overcome these problems, this paper proposed a novel composite sandwich panel with upper and lower GFRP skins and a hybrid polyurethane (PU) foam core (GHP panels). +e hybrid core is composed of different densities (150, 250, and 350 kg/m) of the foam core which is divided functionally by horizontal GFRP ribs. +e hard core is placed in the compression area to resist compressive strength and improve the stiffness of the composite sandwich structure, while the soft core is placed in the tension area. Six GHP panels were tested loaded in 4-point bending to study the effect of horizontal ribs and hybrid core configurations on the stiffness, strength, and failure modes of GHP panels. Experimental results show that compared to the control panel, a maximum of 54.6% and 50% increase in the strength and bending stiffness can be achieved, respectively. GHP panels with the hybrid PU foam core show obvious secondary stiffness. Finally, analytical methods were proposed to predict the initial stiffness and peak load of the GHP panels, and the results agree well with experimental results.


Introduction
Composite sandwich panels with two high stiffness skins and a middle light core are increasingly applied in civil engineering applications [1][2][3][4][5][6]. However, there have been very limited attempts to use these structures for large-scale structural elements. e main reason is that the core material currently used is foam or light wood, and the deformation of the composite panels is large for the cause of its low Young's modulus.
Till now, researchers have done a lot of research on these problems. Steeves and Fleck [7,8] investigated the mechanism of the composite panel with GFRP skin and PU foam using experimental and analytical methods and gained its typical failure modes. Umer et al. [9] investigated the bending properties of composite panels with various core densities. e study showed that the load bearing capacity of sandwich panels increased with the increase of foam density. Sharaf et al. [10] researched the flexural properties of ten sandwich panels, and the results showed that the shape and density of the sandwich plate play an important role in the failure mode, load bearing capacity, and stiffness of the panel.
Dweib et al. [5], Keller et al. [11], and Fam and Sharaf [12] studied the bending properties of composite sandwich panels reinforced with GFRP ribs. e results showed that the longitudinal ribs can significantly increase the bending stiffness and strength of the structure. Zi et al. [13], Moon et al. [14], and Mohamed et al. [15] researched the bending behavior of sandwich panels with transverse ribs, which shows that the reinforcement can alter the failure mechanism of the composite panels. Wang et al. [16] studied the bending properties of foam-filled sandwich plates using 4point bending test. Compared to the reference material, the final bending strength is greatly increased by the introduction of lattice ribs. Manalo et al. [17,18] and Awad et al. [19] studied the bending behavior of fiber composite sandwich panels with horizontal ribs in the middle. e study also illustrated that the horizontal rib can alter the failure mechanism of the composite panels. In order to further increase the stiffness and the final bearing strength of sandwich structures, authors have developed a novel composite sandwich panel with GFRP skins, lattice ribs, and a PU foam core. Test results indicated that the lattice ribs can increase the ultimate load bearing capacity of sandwich panels greatly [20].
is paper detailed analysis of the flexural properties of the composite sandwich panel (GHP panels) with GFRP skins and a hybrid PU foam core (Figure 1). e hybrid foam core is functionally designed with different densities (150, 250, and 350 kg/m 3 ) of the PU foam core and divided by horizontal ribs. e hard core (350 kg/m 3 ) is located at the compression area to resist the compressive strength and improve the stiffness of the composite sandwich structure, while the soft core (150 kg/m 3 ) is located at the tension area.
is paper studied the flexural properties of the GHP panels to evaluate their possibility as structural panel elements. Six panels with the same size (1400 × 120 × 80 mm 3 ) were tested to evaluate their ultimate bending strength, failure mechanism, and bending stiffness. An appropriate analysis model was proposed to predict the bending stiffness and strength of the proposed GHP panels.

Experimental Program
e properties of the GFRP skins and ribs were shown in the paper by Zhang et al. [6]. e hybrid PU foam core with different densities (150, 250, and 350 kg/m 3 ) is divided functionally by horizontal GFRP ribs. e GFRP face sheets and ribs were composed of [0/90]°symmetric E-glass woven fiber (800 g/m 2 ) and HS-2101-G100 unsaturated polyester resin. e GFRP laminates were manufactured by Vacuum Infusion Process (VIP). e GFRP fiber and the resin were provided by Nanjing Spare Composites Co., Ltd. e mechanical properties of GFRP laminates and ribs were examined by tensile, compression, and shear testing according to the ASTM standards. Table 1 summarizes the detail material property result.
In this study, six panels with the same dimensions (1400 × 120 × 80 mm 3 ) were fabricated. Table 2 shows the summary of the test parameters. Specimen GHP-CON, a controlled sandwich panel, is composed of GFRP skins and a 150 kg/m 3 density PU foam core. Specimens GHP-1-1 and GHP-2-1 were fabricated with horizontal ribs and with a kind of 150 kg/m 3 density PU foam core to evaluate the bending properties of sandwich panels with different spaces of horizontal ribs. Specimens GHP-1-2 and GHP-2-2 were fabricated with two different foam core densities (150 and 350 kg/m 3 ), divided by horizontal ribs. Specimens GHP-2-3 with a functionally multilayered PU foam core (150, 250, and 350 kg/m 3 ) and horizontal ribs. Composite sandwich structure is composed of two 3.6 mm-thickness GFRP skins. e thickness of the horizontal ribs is 2.4 mm, and the detailed thickness of each layer of the PU foam core is shown in Table 2.
e manufacture process and the tested specimens are shown in Figure 2. e detailed manufacture process was described in the paper by Zhang et al. [6].
Four-point bending tests were conducted on each panel according to ASTM C393 [21]. e net loading span L c of the sandwich beam is 1200 mm, and the spacing between loading points is 300 mm. e panel deflections are measured using three 100 mm linear variable differential transducers (LVDTs, with a precision of 25 micrometers). e resistance distortion meter is applied to the upper and lower sides to test the longitudinal tension and compression strain of GFRP materials.

Failure Mechanism.
e sample failures can be divided into two main types (see Figure 3): (1) completely core shear failure: the panels lost their bearing capacity completely when the foam core shear failure happened, which occurred in the control specimen (panels GHP-1-1 and GHP-2-1) (see Figures 3(a)-3(c)); (2) core shear failure occurred step by step. When the soft core failed in shear failure, the panel still has the ability to carry the load because of the existence of the hard PU foam core, which can continue to provide the bending stiffness and strength. Finally, specimens collapsed in core shear failure when the strength reached the peak strength of the hard PU foam core, which occurred in specimens GHP-1-2, GHP-2-2, and GHP-2-3 (see e reason of the corresponding failure mode is as follows: (1) the shear strain beyond its maximum shear failure strain of the PU foam; (2) the shear strain of the soft foam core exceeds its maximum failure strain, and the cracks developing path were prevented due to the existence of the horizontal GFRP ribs and the hard PU foam core. e ribs, the hard foam core, and the GFRP skins formed a new sandwich panel. us, the GHP panels can continue carrying the load until the shear strain of the hard core goes beyond its maximum value. e results indicated that the GHP panels with the hard PU foam core in the compression area can alter the failure mechanism of the composite panels.
Comparing panels GHP-CON, GHP-1-1, and GHP-2-1, one can get that the horizontal GFRP ribs have no effect on the failure mode of the composite panel. Comparing panels GHP-1-2, GHP-2-2, and GHP-2-3 with panel GHP-CON, one can get that the hybrid PU foam core made the panel show a more ductility behavior.

Load-Deflection Curves.
e load-deflection curves of GHP panels are shown in Figure 4. e figure shows that the load-deflection curve of GHP-CON shows linear behavior before 6 kN, and when it reached the peak load of 10.8 kN, the curve dropped sharply. A similar load-deflection curve was obtained in panels GHP-1-1 and GHP-2-1 (see Figure 4(a)), failed at 11.1 kN and 10.6 kN, respectively, which was approximately equal to that of specimen GHP-CON. e results indicated that the horizontal ribs have little effect on the load-deflection curve of GHP panels. is is because the ribs distributed in the hybrid PU foam core have small contribution to the bending stiffness rather than placed at the top or bottom GFRP skins.     respectively. e failure load of panels GHP-2-2 and GHP-2-3 was approximately 45.4% and 54.6%, higher than that of panel GHP-COM. Comparing panels GHP-1-2, GHP-2-2, and GHP-2-3 to panel GHP-CON, one can get that the hybrid PU foam core can significantly improve the ultimate load and the bending stiffness of the GHP panels. Figure 5 gives the load-strain relationship of the GHP specimens. e figure shows that the load-strain curve of the GHP panels exhibited almost linear behavior. Table 3 shows the tested mean tensile and compressive distortion of the midspan top and bottom GFRP skins. e ultimate tensile and compressive strains are 0.36% and 0.43%, respectively, which are lower than the values gained from the material tests (Table 1). For panels GHP-CON, GHP-1-2, and GHP-2-2, the number of the compressive strain of the upper GFRP skin is nearly the same with the number of the lower GFRP skin. For panel GHP-2-3, the average number of the lower skin is 20% higher than that of the upper skin, which indicated that the high tensile strength of the GFRP laminates can be fully utilized when panels are with a hybrid PU foam core.

Analysis and Discussion
e following part gives the results of the bending stiffness, ultimate load bearing capacity of the tested panels, and a comparison between experimental and analytical results.

Stiffness Analysis.
Manalo et al. [17] illustrated that the bending stiffness EI ex of the sandwich panels can be gained according to the bending formula of the composite panel under 4-point bending tests. Using the linear elastic curve segment (Figure 4), EI ex can be written as where (ΔP/Δδ) is the initial linear part of the test load-deflection curves and L c is the net loading span of the test panel. e predicted bending stiffness is calculated by equivalent bending stiffness EI eq which can be written as where E f and E ic are the GFRP laminates and PU foam core elastic modulus, respectively. And t s means the GFRP skin thickness, and c i and c j mean the thickness of the PU foam core and GFRP ribs, respectively. Moreover, d s , d i , and d j are the distances from the panel's neutral axis to the GFRP skins, PU foam core, and GFRP ribs' centroids, respectively. e equivalent shear rigidity GA eq can be obtained by e total midspan deformation Δ of the GHP panels in the 4-point bending test can be written as where Δ b means the deformation attributed from bending and is written as Δ s means the shear deformation and can be written as where a is the distance of the loading points from the supports (450 mm).

Ultimate Load Bearing
Capacity. e stress distribution along the thickness (z-axis) of the GHP composite panel can be described as [22] where j means either skins or PU foam core layers and S j means the first moment of area of each part of the GHP panel.

4.3.
Discussion. e predicted bending stiffness of GHP panels can be obtained by equations (1) and (2), as presented in Table 3. For specimen GHP-CON, the predicted bending stiffness EI eq was 50% higher than that obtained experimentally. For panels with GFRP ribs and different laminated PU foam core configurations, the predicted bending stiffness EI eq was 7-53% larger than that obtained experimentally, which indicated that the contribution of shear deformation cannot be neglected in the GHP panels. e predicted midspan deflection and ultimate bending strength can be calculated according to equations (3)-(8), as presented in Table 4. Compared to the experimental results, there is an average error of 13%. e calculated deflections contributed by bending and shear deformation are shown in Table 4, respectively. e analytical method can conservatively estimate the experimental ultimate load of GHP panels with a maximum error of 22%. e maximum variation between the analytical and the test results is 3.53 kN, which occurred in specimen GHP-2-3 because the analytical method neglected the contribution of the GFRP skins and the horizontal ribs on the ultimate shear bearing capacity.

Conclusions
is paper researched the bending properties of a novel composite sandwich panel composed of the upper and lower GFRP skins and a hybrid PU foam core (GHP panels). According to the test and analysis results, the following conclusions can be obtained: (1) Compared to the control specimen, panels with a hybrid functional PU foam core exhibited the highest bending stiffness and ultimate load bearing capacity, a maximum of 50% and 54.6% increase, respectively. (2) e panel with the hybrid PU foam core shows obvious secondary stiffness, the deformation ability can be increased substantially, and the hybrid PU foam core made the panel show a more ductility behavior. (3) e horizontal GFRP rib shows little effect on the failure mode of the composite panel, and it also has little effect on the ultimate load bearing capacity of the sandwich structures.  in this study has been examined. However, the minimum weight sequence is recommended after establishing the appropriate measurement basis and testing more specimens.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.