Combined Temperature and Moisture Effect on the Monotonic and Fatigue Strengths of Sandwich Beams with Glass-Polypropylene Faces and Aluminum Foam Cores

Abstract

Recently, the sandwich structures with thermoplastic faces and metal foam cores have been received much attention in the automobile, aerospace and naval industries. Since the material properties of the polymer-based faces and adhesive films employed in the sandwich structures are sensitive to the temperature and humidity, the knowledge of the environmental effect on the mechanical properties are important to the design and application of such structures. Therefore, the hygrothermal effect on the static and fatigue bending strengths of the sandwich beams with glass-polypropylene faces and aluminum foam cores were experimentally analyzed in the present study. The monotonic and cyclic four-point bending tests were conducted under four environmental conditions, i.e., 25°C/45% RH, 25°C/75% RH, 50°C/45% RH, and 50°C/75% RH, to evaluate the influence of combined temperature and humidity on the strengths against the static and cyclic flexural loads. Experimental results show that the humidity has tiny effect on the static and fatigue strengths when the specimens were tested at fixed temperature. However, the temperature plays an important role in the environmental effect because the monotonic and fatigue strengths of the studied sandwich specimens decrease significantly when the ambient temperature rises from 25 to 50°C. Furthermore, under four considered environmental conditions, two crack systems, the core shear ones and the face/core interfacial ones, were observed both in the monotonic and cyclic tests. The development of interfacial cracks strongly depends on the environmental variables. Accordingly the interfacial cracks play an important role in the static and fatigue strengths of the studied sandwich structures.

1. Introduction

Due to the characteristics of high specific stiffness and high specific strength, the sandwich structures have been widely applied in the fields where the lightweight of structures is strictly required, such as automobile, aerospace and shipbuilding industries etc. Furthermore, the excellent abilities of heat insulation, sound insulation, and energy absorption extend the application of sandwich structures. The sandwich plates are generally prepared by bonding two faces on the core material. The faces and the cores are designed to resist the bending stress and the transverse shear stress respectively when the sandwich structures are subjected to the bending loading. Traditionally, the polymer foams are employed as the core materials because of their easy of manufacturing and low cost.

Recently, the improvement of productive techniques makes the metal foams become the alternatives of the core materials for their higher melting points, stronger mechanical properties, and better capability of heat insulation. Especially, the high melting point of metal foam core makes it possible to utilize the thermoplastic composites as the face materials. Accordingly, the sandwich structures with thermoplastic composite faces and metal foam cores have received much attention in the related applications of sandwich structures.

Since the sandwich structures are widely employed in various fields, the related mechanical properties have been paid much attention by the researchers recently. Among the properties investigated, the fatigue characteristics of the sandwich structures were seldom analyzed. However, the knowledge of fatigue resistance is important for the design and application of sandwich structures because in practice the sandwich panels are frequently subjected to the cyclic loading. Reviewing the past investigations on concerning the fatigue behaviors of the sandwich structures subjected to the cyclic loading demonstrates that lots of studies were dedicated to predicting the fatigue life of the sandwich structures. Some works have shown that the fatigue strengths of the sandwich structures strongly depend on the failure modes^1–4. The fatigue-life curves corresponding to various failure modes can be theoretically or experimentally obtained. The failure modes are dependent on the geometries of the sandwich structures, and the mechanical properties of the employed core/face materials. Accordingly, the design map can be developed to obtain the decisive predicted fatigue life of the sandwich structures by comparing the fatigue strengths with different failure modes. Besides, Jen et al. found that the face/core interfacial debonding was the major failure mode of the aluminum honeycomb sandwich beams subjected to the cyclic bending loading^5–7. The finite element based local parameters were proposed to evaluate the effects of core density ⁵ , amount of adhesive ⁶ , and face thickness ⁷ on the bending fatigue strength of the sandwich beams. Moreover, the stiffness degradations of the sandwich structures under cyclic loadings are another indicators frequently employed to evaluate the fatigue strength of the sandwich structures^8–10.

On the other hand, many efforts have been made to experimentally or numerically study the fatigue behavior of sandwich panels with face/core interfacial cracks^11–18, core cracks ¹⁹ , inserts ²⁰ , face wrinkle defects ²¹ , artificial defect ²² , or weld notches ²³ . Moreover, the influences of some loading variables, such as variable amplitudes of cyclic loading^24–25, load ratios^26–27, load frequencies ²⁸ and low-velocity impact ²⁹ on the fatigue strength of sandwich structures have been also investigated.

Surveying the aforementioned studies reveals that almost no fatigue studies of the sandwich structures with thermoplastic faces and metal foam cores are available. Moreover, in spite the metal foams were employed as the core materials for the innovative sandwich structures, the polymer materials were still used in the specific occasions, such as the matrix of the face composites and the adhesive films between the face and core. Since the polymer materials are sensitive to the temperature and humidity, the knowledge of the hygrothermal effect on the mechanical behavior of these sandwich structures is important to the future employment of these sandwich structures under various environmental conditions. In the past, the environmental effect on the monotonic mechanical properties of the sandwich structures are rarely studies. In 2004, the effect of sub-ambient temperature on the compressive strength of the S2-glass–vinylester foam core sandwich specimens was studied by Thomas et al. ³⁰ . The specimen showed higher compressive strength when tested at low temperature than that tested at room temperature. However, the significant influence of low temperature on the failure mode of the studied sandwich structure was observed.

In the same year, Veazie et al. studied the effects of marine environments on the critical strain energy release rates of the sandwich structures with face/core interfacial pre-cracks ³¹ . The sandwich structures displayed the core degradation evidently after high temperature exposure, and the sea-water exposure made the specimens fail with interfacial damage. Furthermore, the high temperature is detrimental to the mechanical properties of the studied sandwich structures with face/core pre-cracks.

In 2006, Gates et al. performed the toughness tests on the sandwich structures with face/core interfacial cracks at cryogenic temperatures ³² . Experimental results demonstrate that the toughness of the specimen increases when the ambient temperature decreases. Furthermore, Siriruk et al. investigated the effect of sea water exposure on the delamination behavior of the sandwich specimens with a closed cell polymeric foam layer and carbon/glass fiber reinforced polymeric composite faces in 2009 ³³ . The cracks of the wet specimens propagated along the face/core interface, while the cracks grew into the cores when the specimens were tested in dry environments. Moreover, the fracture toughness of the wet specimen decreased significantly when compared with the specimen tested in the dry environment.

In 2014, Zhibin et al. performed the quasi-static three-point bending tests on the aluminum foam sandwich specimens at various temperatures ³⁴ . The failure modes of the studied specimens were found to change with the test temperatures. When the test temperature increases, the leading failure mode of the studied sandwich specimen transfers from core shear to face yield. A design map was proposed to predict the temperature-dependent failure modes of the studied specimens.

Compared with the monotonic studies, fewer investigations on concerning the environmental effects on the fatigue strength of the sandwich structures are available. In 2003, the sandwich structures with PVC foam cores and S2-glass/vinylester faces were fatigue tested at various temperature by Kanny et al. ³⁵ . The fatigue lives of the studied sandwich specimens were found to decrease with increased temperatures.

Soni et al. performed the four-point bending fatigue tests on the sandwich specimens with foam cores and carbon-epoxy/glass-epoxy faces at room temperatures and sub-zero temperatures in 2009 ³⁶ . The core shear is the main failure modes observed both in the room-temperature and sub-zero temperature tests. Longer fatigue life was observed when the ambient temperatures change from room temperatures to sub-zero ones. The room-temperature specimens experienced a long-term gradual loss of stiffness before the final failure. However, the catastrophic failure mode of the low-temperature specimens was observed without any early warning.

In 2013, Jen and Lin studied the temperature effect on the monotonic and fatigue bending strengths of the aluminum honeycomb sandwich beams ³⁷ . The monotonic and fatigue strengths of the studied specimens decrease as temperature increases. The failure modes of the specimens change from the face indentation to the face/core debonding. An analytical procedure was proposed to predict the fatigue life and the corresponding failure mode of the sandwich specimens by comparing the temperature-dependent fatigue strengths of various failure modes.

Reviewing the previous mechanical-property investigations with the consideration of environmental factors indicates that the environmental effects on the sandwich panels with thermoplastic faces and metal foam cores have not been explored. In the present study, the aluminum foam will be selected as the core material because the environmental variables have slight effect on the mechanical properties of the metal core materials. Moreover, the metal foam core can withstand the high operating temperature when preparing the sandwich structures with thermoplastic faces. Hence, the purpose of this investigation is to experimentally study the combined influences of temperature and moisture on the monotonic and fatigue strengths of the sandwich beams with glass-polypropylene composite faces and aluminum foam cores. Monotonic and cyclic four-point bending tests will be performed under various hygrothermal conditions to understand the environmental effects on the static and fatigue properties of the innovative sandwich structures.

2. Experimental Procedures

2.1 Materials and Specimen Preparation

The faces of the studied sandwich structure were obtained by hot-pressing four layer Twintex balanced ±45° stitched glass-polypropylene fabric prepregs at 5 bar and 200°C for 30 min. The prepregs were supplied by Owens Corning Reinforcements Co., France. The glass-polypropylene faces employed in the present study were prepared by hot-pressing the Twintex prepregs. The glass-polypropylene prepregs have been commercialized for many years and the resin of the product has been functionalized to improve the adhesion ability with other metals. Accordingly this composite material has been applied frequently as the thermoplastic faces of the sandwich structures for various applications. The aluminum foam plates provided by the Sichuan Yuantaida Non-Ferrous Metals Co., China, were employed as the core material of the studied sandwich structures. The aluminum foam plates were manufactured by heating the foaming agency and the powder of aluminum metal in a hermetic space. The composition of pure aluminum accounts for more than 95 wt.%. The average value and standard variation value of diameter of the cell are 3.09 and 1.35 mm, respectively. The density of the received foam plate is 0.49 g/cm³, and the corresponding relative density is 0.183.

The sandwich plates were obtained by hot-pressing the faces and the aluminum foam core at 5 bar and 165°C for 15 min with the polypropylene-based adhesive films adhesive placed between the core and the faces. The adhesive films were supplied by the Collano Co., Switzerland, and coded by Collano 23.111. According to the datasheet provided by the supplier, the yield stress, tensile strength and elastic modulus of the employed adhesive film are 17.2, 30.3, and 870 MPa, respectively. Moreover, the lap shear strength on aluminum is 1050 MPa, which indicates that the adhesive film presents excellent bonding performance with the aluminum alloys. During the manufacturing process, the spacers with 25-mm height were employed to ensure the thickness of the sandwich plates. Next, the sandwich plates were cut to the required sizes of the beam specimens (300 × 50 × 25 mm) using a water-jet.

2.2 Moisture Absorption Tests

To understand the characteristics of the moisture absorption for the polymer materials employed in the preparation of studied sandwich specimens. The moisture absorption test of the aluminum foam materials was skipped because the environmental variables considered in the present study have slight influences on the mechanical properties of metal materials. The moisture absorption tests under high humidity conditions were performed on the glass-polypropylene composites and polypropylene-based the adhesive films. The standard specimens were prepared and dried first at 120°C in a chamber for 24 h to eliminate the water and the weight of the dry specimen were measured and recorded. The dried specimen was subsequently kept in the chamber under specific hygrothermal condition. Two hygrothermal conditions, i.e., 25°C/75% RH and 50°C/75% RH, were considered in the experimental program. The weight of the specimen was measured and recorded per hour till the weight reached the saturated value. The water absorption (in percentage) is defined as the ratio of the increased weight due to the moisture absorption to the weight of the originally dried specimen.

2.3 Monotonic and Cyclic Tests

All monotonic and cyclic bending tests were performed using an MTS 810 servo-hydraulic material testing system with a four-point bending jig. Figure 1 shows the schematic illustration of the employed four-point bending tests. The jig consists of one pair of loading cylindrical rollers and one pair of supporting cylindrical rollers. The diameters of all cylindrical rollers are 10 mm. The spans of loading and supporting rollers are 110 and 220 mm, respectively. The hygrothermal bending tests were performed within a chamber, where the temperature and relative humidity were precisely controlled. The hygrothermal effects on the monotonic and cyclic bending strengths of the studied sandwich beams by performing the tests under four hygrothermal conditions, i.e., 25°C (298 K)/45% RH, 25°C (298 K)/75% RH, 50°C (323 K)/45% RH, and 50°C (323K)/75% RH. The specimen was pre-heated in the chamber under the required environmental condition for 3 h before the bending test to reach the thermal and humid equilibrium state.

Figure 1

Schematic illustration of the setup of the four-point bending tests

The monotonic bending tests were conducted by controlling the speed of loading rollers at 0.01 mm/s. The applied load F and the displacement of the loading rollers δ were recorded during the monotonic bending tests. The peak load values of the load-displacement curves are the ultimate loads, F _ult , which will be used as the reference data of subsequent cyclic tests.

The cyclic bending tests were performed by controlling the applied loads with the load ratio R equal to 0.1. The load ratio is defined as the minimum applied load F _min to the maximum applied load F _max . Note that the downward loads are considered to be positive in the present study. The waveform of the applied load was sinusoidal and the frequency of the fluctuated load was 3 Hz. Five loading levels r were selected in the cyclic tests for the specimens tested under each hygrothermal condition. The loading level is defined as the ratio of the maximum applied load in the cyclic test F _max to the ultimate applied load in the monotonic test F _ult under the same hygrothermal condition. The fatigue life N _f is defined as the number of cycles when the maximum displacement reaches the one corresponding to the ultimate load obtained in the monotonic test. The maximum displacements of the upper rollers corresponding to the ultimate loads obtained in the monotonic tests under the four hygrothermal conditions, i.e., 25°C/45% RH, 25°C/75% RH, 50°C/45% RH, and 50°C/75% RH, are 12.431, 12.221, 11.453, and 11.696 mm, respectively. After the monotonic and cyclic tests, the failure modes of the specimens were observed using an optical microscope.

3. Results and Discussion

3.1 Moisture Absorption Tests

Figures 2(a) and (b) show the variations of the water absorption with the exposure time tested under various hygrothermal conditions for the face composites and the adhesive films, respectively. It is evident that the water absorption increases with time rapidly and reach a saturated state after a period of exposure time for both the face and adhesive film materials. Furthermore, the increase rate of the water absorption of the studied material tested at 50°C is higher than that tested at 25°C, and the value of the saturated water absorption obtained at 50°C is larger than that tested at 25°C. Furthermore, in spite that the saturated values of water absorption for both the studied materials are less than 0.3%, it is shown that the water absorption ability of adhesive film is slightly higher than that of the face composite under the identical hygrothermal condition.

Figure 2

Variations of the water absorption with the exposure time for the (a) face composites, and (b) adhesive films

3.2 Monotonic Tests

Figure 3 shows the load-displacement curves of the loading rollers obtained in the monotonic bending tests under various hygrothermal conditions. As seen in this figure, the relationships between the loads and displacements for all curves display smooth nonlinear characteristics from the beginning of the tests. Moreover, it is evident that the temperature plays a more important role than the humidity in the monotonic behavior of the studied sandwich beams since these load-displacement curves can be divided into two groups according to the test temperatures. The stiffness and the ultimate load of the studied specimen obtained at room temperature are much higher than those obtained at higher temperature. Moreover, the load-displacement curves obtained at the same temperature are almost identical and the influence of moisture on the monotonic bending strengths of the studied sandwich beams is slight. Figure 4 shows the ultimate loads F _ult obtained under various hygrothermal conditions. When the ambient humidity changed from 45% RH to 75% RH, only 2.6% and 1.3% losses in the ultimate loads were observed when the specimens were monotonic tested at 25°C and 50°C, respectively.

Figure 3

Relationships between the applied load F and the displacement of the loading roller δ obtained in the monotonic bending tests performed under various hygrothermal conditions

Figure 4

Ultimate applied loads F _ult obtained in the monotonic bending tests performed under various hygrothermal conditions

Two crack systems were observed for the studied specimens subjected to the monotonic bending loading under various hygrothermal conditions. The core-shear crack system was found to initiate at the face/core interface and grow into the core area between the loading and supporting rollers. The other crack system initiated and propagated along the face/core interface. Figures 2a and 2b show the examples of the failure modes for the studied sandwich beams monotonically bended under 25°C/45% RH, and 50°C/75% RH conditions. Note that the interfacial crack of the specimen tested at high temperature is longer and obvious than that tested at lower temperature. It also indicates that the development of the interfacial crack system is related to the environmental variables, especially the temperature. Accordingly, the interfacial crack plays an important role in the monotonic properties. Furthermore, the failure mechanisms for the specimens tested under the same temperature but different humidity conditions are almost identical.

Figure 5

Examples of the failure modes obtained in the monotonic bending tests performed under (a) 25°C/45% RH, and (b) 50°C/75% RH

3.3 Cyclic Tests

Table 1 lists the experimental results of the studied sandwich beams obtained in the cyclic tests. Figure 6 shows the relationship between the maximum applied load of the loading rollers F _max and the cycles to failure N _f obtained under various hygrothermal conditions. In this figure, the relationship between the maximum applied load and the fatigue life was described by fitting the data points using a power law equation, that is

$$F_{\max }\left(\text{N}\right)=a{\text{N}}_f^b$$

(1)

where a is the fatigue strength coefficient, and b is the fatigue strength exponent. These materials constants are dependent on the employed environmental conditions. These fitted fatigue life curves present as the straight lines in the log-log scale diagram. The material constants a and b obtained from the fitting results are listed in Table 2 . Furthermore, all the coefficients of determination, R-squared, for the fitted results obtained under different hygrothermal conditions are larger than 0.98, indicating that the power laws can describe the relationship between the maximum applied loads and the fatigue lives of the studied specimens appropriately. Similar to the trends observed in the monotonic tests, these fatigue-life curves can be divided into two groups according to the ambient temperature. The lower fatigue strength was observed evidently when the specimen was tested at higher temperature. The effect of the moisture on the fatigue strength of the studied sandwich beams is slight since the fatigue-life curves obtained under different humidity conditions are almost identical.

Figure 6

Relationships between the maximum applied loads F _max and the fatigue lives of the studied specimens obtained in the cyclic bending tests performed under various hygrothermal conditions

Table 1

Experimental results of the cyclic bending test for the studied sandwich specimens under various hygrothermal conditions

Hygrothermal condition	Applied maximum load, F max (N)	Loading ratio, r (%)	Fatigue life, N f (cycles)
25°C/45% RH	4259.5	63.7	19,949
	3931.8	58.8	35,197
	3604.2	53.9	157,262
	3276.5	49.0	452,605
	2948.9	44.1	>1,000,000
25°C/75% RH	4259.5	65.4	15,641
	3931.8	60.4	26,571
	3604.2	55.4	95,727
	3276.5	50.3	425,415
	2948.9	45.3	>1,000,000
50°C/45% RH	3931.8	79.6	4,048
	3604.2	72.9	18,898
	3276.5	66.3	44,330
	2948.9	59.7	222,654
	2621.2	53.1	>1,000,000
50°C/75% RH	3931.8	80.6	4,608
	3604.2	73.9	12,921
	3276.5	67.2	40,992
	2948.9	60.5	243,523
	2621.2	53.7	>1,000,000

Table 2

Fitting results of the fatigue-life curves obtained under various hygrothermal conditions

Hygrothermal condition	Fatigue strength coefficient, a	Fatigue strength exponent, b	Coefficient of determination, R 2
25°C/45% RH	9999.93	-0.0869	0.980
25°C/75% RH	8623.32	-0.0764	0.980
50°C/45% RH	7368.58	-0.0746	0.993
50°C/75% RH	7238.99	-0.0735	0.995

Similar to the crack behavior observed in the monotonic test, the core shear cracks were found first to form at the interface and grow into the core area between the inner and outer rollers in the cyclic tests. The other crack system subsequently developed along the interfaces and played a decisive factor in fatigue strength due to the environment-sensitive characteristics of the polymer adhesive film. The interfacial cracks of the specimens tested in the severe environments are more evident than those tested in the normal environments. Since the properties of the adhesive film are sensitive to the temperature and humidity, the degradation of the adhesive is expectable. Accordingly, the interfacial crack system develops rapidly in the hot-wet environments, which weakens the fatigue strength significantly.

Figure 7 shows the variations of stiffness with the fatigue cycle ratios for some studied specimens under various hygrothermal conditions. The stiffness of the specimens is defined as the ratio of the range of the applied load ΔF to the range of the displacement of the loading roller Δδ. The fatigue cycle ratio is defined as the number of the applied cycles n to the fatigue life N _f . It is seen that from the beginning of the cyclic tests, the stiffness of the specimen tested at high temperature decreases more rapidly than that tested at lower temperature. It implies that the development of interfacial crack is affected by the temperature significantly, accordingly the stiffness degradation is markedly observed in the severe environments. Furthermore, the obvious drop in stiffness was found during the cyclic test performed under high humidity conditions. The stiffness drop occurred earlier for the specimen tested at the higher temperature, revealing that the effect of the humidity on the degradation of the mechanical properties is dependent on the exposure time and the high temperature accelerates the influence of the moisture on the fatigue strength of the studied specimens.

Figure 7

Examples of the variations of specimen stiffness with the fatigue cycle ratios

Figure 8 shows the relationships between the loading levels r and the fatigue-life data N _f of the studied sandwich specimens under various hygrothermal conditions. The concentration of data points indicates that a unique model can be developed to describe the relationship between the loading level and fatigue lives no matter which environmental condition is considered in the cyclic tests. The power law is utilized again to model the relationship between the loading levels and the fatigue lives, which can be expressed as follows:

$$r\left(\%\right)=142.53N_f{}^{-0.0778}$$

(2)

The high value of the coefficient of determination, R-squared (R ² =0.716) for the fitting result indicates the power-law model can appropriately describe the characteristics of the scattered data points. Furthermore, the concentrated trends of the data points observed in the various environments imply that the r-N _f curve obtained under the normal temperature and humidity condition (25°C/45% RH) can be considered as the reference design curve to predict the fatigue lives of the studied sandwich specimens under other hygrothermal conditions. The application of the fitting curve of all data points in the fatigue life prediction is not practical because the establishment of this reference curve needs large amount of data points obtained under various environmental conditions. Oppositely, the acquisition of the curve under 25°C and 45% RH is much more convenient and easier than those under other environmental conditions. Figure 9 shows the comparison between the predicted results and the experimental data by using the loading level parameters. The r-N _f relationship for the studied specimens obtained under 25°C/45% RH condition, r (%) = 149.65 N _f ^−0.0869, was employed as the reference baseline to predict the fatigue lives under other three hygrothermal conditions. The diagonal solid line depicted in this figure indicates the exact prediction, and the scatter bands of factors two and three are also provided as the dashed lines in this figure to evaluate the of the prediction performance. It is shown that among the 15 data points, only six points locate within the scatter band of factor three. However, the majority of the predicted fatigue-life values for the tested specimens are lower than the experimental ones, indicating that the prediction using the loading level parameter provides a conservative evaluation for the fatigue life of the studied sandwich structures tested under various hygrothermal conditions.

Figure 8

Relationships between the loading levels r and the fatigue lives of the studied specimens obtained in the cyclic bending tests performed under various hygrothermal conditions

Figure 9

Prediction performance of the loading level parameter for fatigue-life evaluation of the studied sandwich specimens under various hygrothermal conditions

4. Conclusions

In the present study, the hygrothermal effect on the static and fatigue bending strengths of the sandwich beams with glass-polypropylene faces and aluminum foam cores were experimentally analyzed by performing the monotonic and cyclic four-bending tests under various temperature-humidity conditions. Some conclusions can be drawn from the experimental observations:

1. 1.

   The static and fatigue bending strengths of the studied sandwich panels decrease remarkably as the ambient temperatures increase. However, the effect of humidity on the static and fatigue strengths is slight.

2. 2.

   Two crack systems, the core shear cracks and the face/core interfacial cracks were observed both in the monotonic and cyclic tests. The development of the interfacial cracks strongly depends on the hygrothermal conditions, accordingly affects the fatigue lives obtained in the hot-wet tests.

3. 3.

   The loading level parameter can be used to provide a conservative fatigue-life prediction for the studied specimens cyclic tested under various hygrothermal conditions.