Study on Shear Fatigue Performance of Scarf Repaired Honeycomb Sandwich Structure

The shear static and shear fatigue tests of the honeycomb sandwich composite were carried out to study the effects of three different repair methods on shear fatigue performance: the unilateral panel scarf repair, the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair. The stress ratio R=0.1 was selected. The results show that the failure mode of the honeycomb sandwich structure is the destruction of the honeycomb core. The structure is torn into two halves from the core, and the core tends to squeeze to one side. There is no debonding between the panel and the core, and no delamination in the panel. The static strength, fatigue limit, fatigue life and residual strength of the specimens are significantly decreased by the scarf repair. The rank of reduction from most to least is: the unilateral panel scarf repair, the bilateral panel and core scarf repair, and the unilateral panel and core scarf repair.And the latter two are remarkably similar.


Introduction
Composite honeycomb sandwich structure has the advantages of light weight, high specific strength and specific stiffness, good wave transparency and corrosion resistance.They are widely used in aerospace fields, such as airborne radome, fairing, landing gear hatch and other parts [1,2].These structural parts are often damaged during service and need repair.Scarf repair is the main permanent repair method.After repair, the structure is often subjected to alternating shear stress under the action of complex aerodynamic loads and will suffer fatigue damage.Therefore, it is of great engineering significance to study the shear fatigue performance of scarf repaired honeycomb sandwich composite structures.
At present, some achievements have been made at home and abroad on the fatigue performance research of honeycomb sandwich structure about shear failure.Abbadi A et al. [3] conducted a fourpoint bending fatigue test on the honeycomb sandwich structure (aluminum/aramid paper) and did numerical simulation.The failure mode was shear failure of the honeycomb core.The fatigue cracks of the L-direction and W-direction [4] configuration honeycomb cores all propagate along the thickness direction of the core.The failure propagation of the L-direction configuration honeycomb core is always in the diagonal direction.The failure propagation of the W-direction configuration honeycomb core is always in the horizontal direction.Wahl L et al. [5] carried out a three-point bending fatigue test on the aluminum honeycomb sandwich structure.The tests were done respectively in the L direction, W direction [4] and 62° direction of the honeycomb core.The results show that under the three-point bending load, the core shear failure occurs in the honeycomb sandwich structure.The shear stress takes the maximum value in the 62° direction.The strength takes the minimum value in this direction and the maximum value in L direction.Burman M et al. [6] used an improved four-point bending fixture to conduct four-point bending fatigue tests on foam sandwich beam structures.The results show that the core shear failure occurs in the structure.The damage starts in the entire length direction of the high shear stress zone and the middle region of the test piece.Then the damages gradually expand into horizontal macroscopic crack and extend to the surface/core interface.Wahl L et al. [7] used ANSYS to simulate the static three-point bending test of the aluminum honeycomb sandwich structure [5], and compared the simulation results with the fatigue test results.They then did fatigue failure prediction.The results showed that the fatigue cracks were generated inside the honeycomb core and occurred in the honeycomb cells away from the location where the load was applied, which indicated that the failure mode was core shear failure.Wudi Cao [8] and Zilong Chai [9] respectively carried out three-point bending static and fatigue tests on aluminum honeycomb sandwich panels and aramid paper honeycomb sandwich panels.The tests showed that the static failure mode is local core collapse, and the fatigue failure mode is shear failure of the honeycomb core.
In summary, the current research on fatigue performance of honeycomb sandwich structures mainly focuses on three-point bending or four-point bending fatigue.There is few news on shear fatigue performance.G Bianchi et al. [10,11] studied the static mechanical properties and fatigue behavior of the aluminum alloy hexagonal honeycomb sandwich structure using the unilateral shear loading test method.The results showed that: for constant amplitude stresses, the fatigue life in the L direction of the honeycomb core is longer than that in the W direction.This is because the honeycomb core has a greater static ultimate strength in the L direction.However, the honeycomb core is more effective against fatigue damage in the W direction at the load level of 0-20% of the static strength.Yong Wang et al. [12] carried out tensile, compressive and shear static strength tests on the honeycomb sandwich structure before and after aging, as well as shear fatigue tests under different stress levels.The results show that the damages of the static and fatigue test specimens are caused by the failure of the honeycomb core.There is no damage to the panels.And no surface-core debonding occur.The standard aging environment reduces or eliminates the residual stress and makes the honeycomb core evenly stressed, thus prolonging the fatigue life.
To sum up, there is no news on shear fatigue of scarf repaired sandwich structures.This paper addresses the shear fatigue performance of the scarf repaired honeycomb sandwich structure with plain woven glass fiber laminates as the face plate and aramid paper honeycomb Nomex as the core.The impacts of scarf repair on failure modes, static strength, fatigue limit, fatigue life and residual strength are analyzed.

Test specimens and test method (1) Test specimens
The ASTM (American Society of Testing Materials) C273/C273M-16 standard [13] and C394/C394M-13 standard [14] are referred to design the honeycomb sandwich structure shear static strength and shear fatigue test specimen.The test specimen is 150 mm long, 50 mm wide, and 7 mm thick, as shown in figure 1.The panels consist of 3 single-layer boards, each with a thickness of 0.267 mm.The single-layer board is composed of 7781 type glass fiber plain weave cloth and PR381 type epoxy resin matrix.The woven cloth is laid in the 0° direction.The honeycomb core uses Nomex aramid paper honeycomb model CMAG-CNC1-1.83-96.The side length of the core cell is 2 mm.BMS5-101 TYⅡ type adhesive film is used for bonding between the face sheets of the honeycomb sandwich structure.The intact test specimen is cured and molded by autoclave process.The curing temperature is 127 °C and kept for 2 hours.The pressure was 0.345 MPa.
In this paper, three repair methods are considered: the unilateral panel scarf repair, the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, as shown in figure 2. A simplified one-dimensional stepped scarf repair model [15,16] is adopted.The angle of inclination is 1:30 [17].
The material of the patch of the panel is the same as that of the intact panel.The replacement honeycomb core is the same as the intact core.The patch, the replacement honeycomb core and the original specimen are bonded with special foam glue.The repaired specimens are all cured in a vacuum bag oven.The curing pressure is 0.1 MPa.The curing temperature curve is the same as that of the prepreg.The test matrix is shown in table 1, where the intact test specimen is denoted by A, the unilateral panel scarf repaired specimen is indicated by B, the unilateral panel and core scarf repaired specimen is represented by C, and the bilateral panel and core scarf repaired specimen is represented by D. S represents the shear static strength test, while F represents the shear fatigue test.The static shear performance test of the honeycomb sandwich structure was carried out in accordance with the ASTM C273/C273M-16 standard [13].The test adopts the tensile loading mode.The upper and lower panels of the sandwich structure are glued to the fixture made of 6061 aluminum alloy with J-78B adhesive film.The shear performance test was carried out on an LF5504 50 kN electro-hydraulic servo testing machine with a speed of 0.5 mm/min.In order to reduce the influence of the deformation of the fixture, the deformation is measured with an independent dial indicator.The ambient temperature is 18 °C and the humidity is 51% RH.The fixture is shown in figure . 3(a).The shear fatigue performance test of the honeycomb sandwich structure was carried out in accordance with the ASTM C394/C394M-13 standard [14].The test also uses the LF5504 50 kN electrohydraulic servo testing machine.The stress loading method was adopted.The loading frequency was set to 5 Hz.The stress ratio R= 0.1.The waveform is a sine wave.The maximum number of stress cycles is NT=10 6 .The environmental conditions are the same as the static shear test, and the fixture is shown in figure 3(b).

Result analysis and discussion
(1) Failure modes When the honeycomb sandwich structure test specimens were damaged under the action of shear static load and fatigue load, the structural failures were always caused by the destruction of the honeycomb cores.There was no debonding between the panel and the core.And there was no delamination in the panel.The test specimen was torn into two halves from the core.The damaged honeycomb core can be seen clearly, as shown in figure 4. Since the mechanical performance test of the honeycomb sandwich structure test specimen adopts the tensile loading mode, it can be seen clearly from the failure mode diagram that the honeycomb core is squeezing to one side.The appearance of the failure is centrally symmetrical on the two halves of the torn test specimen.The honeycomb core at the upper and lower edges always breaks close to the panel.(

2) Static strength
The static shear test results of the intact specimens and the three kinds of repaired test specimens are shown in table 2. The coefficient of variation of the failure strength of the four kinds of test specimens are all less than 1%, indicating that the test results are acceptable.The failure strength of the intact specimens is the highest, while the failure strength of the unilateral panel scarf repaired specimens is the lowest.The three scarf repair methods all significantly reduce the static shear strength.The average recovery rate of the mechanical performance of repaired specimens is defined as follows: In the formula: α is the average recovery rate of mechanical performance of repaired specimens; Ix is the average value of mechanical performance indexes of repaired specimens; Iw is the average value of mechanical performance indexes of intact specimens.
The average recovery rate of the failure strength corresponding to the three repair methods is calculated according to the above formula.As shown in table 2, the recovery concerning the static strength is the best for the unilateral panel and core scarf repaired specimens, followed by the unilateral panel and core scarf repaired specimens, and then the unilateral panel scarf repaired specimens.The scarf repair of the panel can lead to stress concentration near the bonded area between the patch and the original specimen.For the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, foam glue is used between the replacement core and the original core.The bonding strength will be higher than the strength of the aramid paper honeycomb after the foam glue cures [18].Thus, the failure strengths of these two kinds of repaired specimens are similar and higher than that of the unilateral panel scarf repaired specimen.The fatigue strength at a given life (NT=10 6 ), that is, the conditional fatigue limit, is determined by the staircase method.The determination of the conditional fatigue limit of the intact specimens and the three kinds of repaired test specimens was carried out under three load levels, as shown in figure.5. table 3, table 4, table 5 and table 6 show the fatigue test results of intact specimens and three kinds of repaired test specimens under different load levels.The test results intended for the staircase method are used to calculate the conditional fatigue limit.The test results intended for the group method are used to draw the S-N curve.4) is roughly the same for the three kinds of repaired specimens and that the value of the material parameter a is changed by specialized scarf repair.Under the same load level, the fatigue life of the intact specimens is the greatest, followed by the fatigue life of the unilateral panel and core scarf repaired specimens, and the bilateral panel and core scarf repaired specimens.The fatigue life of the unilateral panel scarf repaired specimens is the shortest.Since there is no damage and thus no obvious stress concentration, the intact test specimens are most resistant to fatigue damage.The scarf repair changes the load transmission path and produces more stress concentration near the glued surface.Therefore, the fatigue life of various scarf repaired test specimens is shorter than that of the intact test specimens under the same load level.For the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, the mechanical properties of the core improve when the foam glue between the replacement core and the original core cures [18].Consequently, the fatigue life of the two kinds of structures is greater than that of the other type. (

5) Residual Strength
The test specimens that do not fail after reaching the maximum stress cycle (NT=10 6 ) are also tested for residual strength.The test results of static strength and residual strength after cyclic loading of intact specimens and three kinds of repaired specimens are summarized in table 8.In the table, SA-0, SB-0, SC-0 and SD-0 are respectively represented as the average value of the static strength of the intact specimens, the unilateral panel scarf repaired specimens, the unilateral panel and core scarf repaired specimens, and the bilateral panel and core scarf repaired specimens.While FA-0, FB-0, FC-0 and FD-0 are represented as the average value of the residual strength.It can be found from table 8 that, after fatigue loading of the intact specimens and three kinds of scarf repaired specimens, the fatigue residual strength of most of the test specimens is less than their static strength.The residual strength of the intact specimens is the highest.For the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, the residual strengths are almost equal and greater than that of the unilateral panel scarf repair.1) For intact specimens and repaired specimens of composite honeycomb sandwich structure under the static and fatigue shear loads, the failure mode is the failure of the honeycomb core.The structure is torn into two halves, and the core is squeezed to one side.There is no debonding between the panel and the core.And there is no delamination in the panel.
2) The static strength, fatigue limit, fatigue life and residual strength of the specimens are significantly decreased by the scarf repair.The most obvious reduction is observed in the unilateral panel scarf repair.The more obvious reduction is observed in the bilateral panel and core scarf repair.The reduction is similar in the unilateral, and bilateral panel and core scarf repair.

Figure 2 .
Figure 2. Schematic diagram of repair method of specimen.Table 1. Test matrix.Classification mark Repair method Load type Number of specimens SA Intact Static 3 SB Unilateral panel scarf repair Static 3 SC Unilateral panel and core scarf repair Static 4 SD Bilateral panel and core scarf repair Static 4 FA Intact Fatigue 28 FB Unilateral panel scarf repair Fatigue 20 FC Unilateral panel and core scarf repair Fatigue 28 FD Bilateral panel and core scarf repair Fatigue 22 (2) Test methodThe static shear performance test of the honeycomb sandwich structure was carried out in accordance with the ASTM C273/C273M-16 standard[13].The test adopts the tensile loading mode.The upper and lower panels of the sandwich structure are glued to the fixture made of 6061 aluminum alloy with J-78B adhesive film.The shear performance test was carried out on an LF5504 50 kN electro-hydraulic servo testing machine with a speed of 0.5 mm/min.In order to reduce the influence of the deformation of the fixture, the deformation is measured with an independent dial indicator.The ambient temperature is 18 °C and the humidity is 51% RH.The fixture is shown in figure.3(a).

Figure 4 .
Failure mode diagram of typical specimens.

Figure 7 .
Figure 7. S-N curve of specimens.The absolute value of the slope of the S-N curve of the intact specimen is the largest, while the absolute values of the slopes of the other three kinds of repaired specimens are smaller than that of the intact specimens.The slopes of the three kinds of repaired specimens are roughly equal.It indicates that the value of the material parameter b in formula (4) is roughly the same for the three kinds of repaired specimens and that the value of the material parameter a is changed by specialized scarf repair.Under the same load level, the fatigue life of the intact specimens is the greatest, followed by the fatigue life of the unilateral panel and core scarf repaired specimens, and the bilateral panel and core scarf repaired specimens.The fatigue life of the unilateral panel scarf repaired specimens is the shortest.Since there is no damage and thus no obvious stress concentration, the intact test specimens are most resistant to fatigue damage.The scarf repair changes the load transmission path and produces more stress concentration near the glued surface.Therefore, the fatigue life of various scarf repaired test specimens is shorter than that of the intact test specimens under the same load level.For the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, the mechanical properties of the core improve when the foam glue between the replacement core and the original core cures[18].Consequently, the fatigue life of the two kinds of structures is greater than that of the other type.(5)Residual StrengthThe test specimens that do not fail after reaching the maximum stress cycle (NT=10 6 ) are also tested for residual strength.The test results of static strength and residual strength after cyclic loading of intact specimens and three kinds of repaired specimens are summarized in table 8.In the table, SA-0, SB-0, SC-0 and SD-0 are respectively represented as the average value of the static strength of the intact specimens, the unilateral panel scarf repaired specimens, the unilateral panel and core scarf repaired specimens, and the bilateral panel and core scarf repaired specimens.While FA-0, FB-0, FC-0 and FD-0 are represented as the average value of the residual strength.It can be found from table 8 that, after fatigue loading of the intact specimens and three kinds of scarf repaired specimens, the fatigue residual strength of most of the test specimens is less than their static strength.The residual strength of the intact specimens is the highest.For the unilateral panel and core scarf repair, and the bilateral panel and core scarf repair, the residual strengths are almost equal and greater than that of the unilateral panel scarf repair.Table8.Residual strength results of intact specimens and three kinds of repaired specimens.

Table 2 .
Results of static shear test.

Table 3 .
Fatigue test results of FA intact specimens.

Table 4 .
Fatigue test results of FB repaired specimens.

Table 8 .
Residual strength results of intact specimens and three kinds of repaired specimens.In this paper, the shear static strength and shear fatigue tests of the intact specimens and the three kinds of repaired specimens of honeycomb sandwich structures are carried out.And the following conclusions can be drawn: 10