An experimental study on the fatigue performance of CFRP and BFRP repaired aluminium plates
Introduction
The repair of cracked aircraft structures with adhesively bonded composite reinforcements has been proven to be effective by numerous studies and applications over decades [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. The boltless repairs have several advantages and they can be used for temporary repairs and permanent crack or corrosion damage repairs.
The efficiency of composite repairs of cracked aluminium structures depends on several factors. These factors include the repair materials, the adhesive curing temperatures, the effective thermal expansion during the curing, the type of fatigue loading, and the structural support against the secondary bending of single-sided repairs. Some of these factors may be difficult to simulate with plate specimens in laboratory testing.
The main problems with the bonded repairs are the surface preparation of the metal and the residual thermal stresses with the composite reinforcements. Silane-based surface preparation methods are commonly applied. Their effectiveness has been previously reported by the authors [23].
The residual thermal stresses result from the mismatch of the coefficients of the thermal expansion (CTE) between the repaired aluminium plate and the composite reinforcement and from the temperature difference between the curing and operating temperatures (ΔT). The effect has been studied (or at least discussed) in almost all crack-patching investigations [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. However, none of them have applied a bonding rig (as presented in this investigation) in order to constrain the thermal expansion of the repaired plate during the adhesive curing.
Most authors have used unidirectional boron/epoxy (b/e) prepreg reinforcements. Only Sandow and Cannon [4] have tested multidirectional b/e patches. The unidirectional b/e plies have a high tensile stiffness (three times the aluminium stiffness), and their CTE in the fibre direction is one-fourth of the aluminium CTE.
Carbon/epoxy (c/e) prepreg patches have been used in some repair studies [6], [7], [8], [9], [10], [11]. The c/e patches can have a tensile stiffness ranging from the aluminium stiffness to the b/e patch stiffness. Usually it is in the middle of these two values. The CTE of the unidirectional c/e patches will cause more problems, since it is practically zero. The wet-laminated c/e repairs have been studied mainly for the repair of secondary structures [8].
GLARE material is made of aluminium and glass/epoxy plies. Consequently, the CTE of the GLARE is close to the CTE of aluminium but the stiffness is lower than the aluminium stiffness. The GLARE patches have been tested in Delft [18], [19].
The wet laminated glass/epoxy (g/e) reinforcements have been used mostly in secondary structures or in non-structural repairs. The g/e-reinforcements have been tested by Hosseini-Toudeshky [20] and by Rao [21].
Most authors have used the FM73 or other adhesive films cured at 120 °C for bonding the patch [1], [2], [3], [4], [5], [6], [7], [8], [10], [11], [13], [14], [15], [16], [17], [18], [19], [20], [21]. Since the operating temperature is usually considered to be room temperature (RT) or lower, the residual stresses can be remarkable. In some studies, the authors have tested lower temperatures for curing the adhesive films [1], [7], [22]. Some authors have also used two-part paste adhesives cured at lower temperatures in order to reduce the residual thermal stresses [4], [7], [9], [12].
Single-sided composite repairs have been found to be efficient for the repair of thin (less than 3.2 mm) structures [1], [2], [3], [4], [5], [6], [7], [8], [9], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. Repairs of thicker aluminium structures from 6 to 10 mm have also been analysed and tested [10], [11], [12], [13], [14], [15], [16]. A limited number of tests have been done with medium (3.2 mm) thick plates [1], [2], [3], [4], [5], [11], [14], [22]. Even fewer authors have compared single- and double-sided repairs [4], [5], [8], [16], [17].
The fatigue testing in most studies has been done with constant amplitude (CA) loading and a positive stress ratio R [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. Vlot [19] has also used a filtered flight spectrum without any compressive loads. Only a few authors have used a simulated flight spectrum loading containing compressive loads [2], [3], [4], [5], [6].
In most investigations, the stress ratio R of the fatigue loading has been positive and no structural support has been used. However, with a single-sided repair, the neutral axis of the repaired structure will shift, and the secondary bending effects on the fatigue life of an unsupported structure can be significant. Some authors have used a sandwich support during the fatigue loading in order to eliminate the bending effects [1], [2], [3], [5], [7], [8]. Sandow [4] and Kan [11] have used an anti-buckling support with edge rails.
It has been standard practice since the beginning of the crack-patching technique to not use stop-drilling of the crack tip before the composite repairs [1], [2], [3]. Therefore, the effect of stop-drilling has been tested in only a few papers [7], [21]. The reasons for omitting the stop-drilling are well known, but the simplified assumption based on the plastic zone in front of the crack tip is applicable only for a constant amplitude loading and does not take into account the compressive loads’ interaction effects [24].
The first objective of this study was to develop a representative laboratory simulation for bonded aluminium aircraft repairs with plate specimens. The surrounding aircraft structure will restrict the free displacement of the repair area in two ways. During the adhesive curing, the real aircraft structure will constrain thermal expansion of the heated area and thus affect the residual thermal stresses. After the repair, the aircraft structure will also support the area against secondary bending during the fatigue loading and thus affect the fatigue life. For the first issue we designed a bonding rig made of steel in order to constrain the effective thermal expansion during the adhesive curing to the same level as experienced in the real locally heated aircraft structures. For the second issue we used edge support rails during the fatigue loading.
The second objective was to test the effect of residual thermal stresses using the designed rig and to test the efficiency of different repair materials with typical aircraft aluminium grades and with typical aircraft repair configurations. We used thin centre-cracked aluminium plates when we simulated the single-sided aircraft skin repairs. Typical stringer and frame structures were simulated with the narrower but thicker edge-cracked plates. Both single-sided and double-sided b/e repairs were tested with these edge-cracked specimens. In this investigation, all repaired cracks were stop-drilled before the repair because it was found to be beneficial in these cases [25]. The effects of stop-drilling the cracks will be investigated further in a separate study.
The testing was started with clad aluminium 2024-T3 [26], since historically it has been used in skin plates prone to fatigue cracking. However, current aircraft operated by the Finnish Air Force (FINAF) use different aluminium grades. Therefore, it was found necessary to compare the results achieved with 2024-T3 aluminium plates to the results achieved with these other grades.
Typical prepreg and wet-laminated repair configurations were tested with the developed approach. Multidirectional patches were used in this investigation, because we wanted the patches to withstand the bi-directional loads that can occur in real-life applications. Also, as observed by Sandow and Cannon [4], multidirectional patches may withstand spectrum loading with compressive loads better than unidirectional patches. The repair materials included c/e prepreg, b/e prepreg and wet-laminated c/e (wet c/e) reinforcements. We varied the residual thermal stresses in the repairs by using the steel bonding rig, by varying adhesive and resin curing temperatures and by varying the testing temperatures.
A variable amplitude (VA) loading with tensile and compressive peaks was selected in order to realistically simulate the flight loading and to include the load interaction effects on fatigue crack growth. Because of the single-sided repairs and the compressive loads, a structural support against bending and buckling was used. The results achieved with the edge-supported specimens were compared to the results measured with the sandwich-supported and the unsupported specimens.
Section snippets
Aluminium plates
We did most of the testing with 1.6 mm thick 2024-T3 clad aluminium sheets. For medium thickness specimens we selected 3.2 mm thick French aluminium AU4G1 clad (equivalent to ASTM 2024-T3 [26]) due to its availability. Clad was not removed from the specimens prior to surface treatment.
A comparison between different aircraft aluminium grades was accomplished with the 1.6 mm thick plates. The aluminium grades compared were BAE Military Systems S07-1020 (equivalent to ASTM 2014A-T4 [27]) used in the
Bonding rig
A bonding rig was developed in order to control thermal expansion of the repaired plates during the elevated temperature adhesive curing. The purpose of the rig was to simulate a typical locally heated aircraft structure surrounded by cool structures. The rig restrains the expansion of the aluminium plate. The expansion coefficient was aimed at representing the typical effective thermal expansion attained during an on-the-aircraft repair.
The bonding rig was constructed from steel having a
Thermal expansion coefficients and residual thermal stresses
The effective CTE of an aluminium plate repaired in the steel rig and placed in a vacuum bag was measured with strain gages. The measured value was .
We made a reference measurement from a locally heated Hawk aircraft rear fuselage structure. The result was .
The CTE of a locally heated structure was also approximated using FEM. Simple 800 × 800 mm size NISA [35] and IDEAS [36] aluminium plate models were constructed using two-dimensional shell elements. The
Bonding rig and residual thermal stresses
The steel bonding rig provided a suitable constraint to the thermal expansion of the aluminium plate during the elevated temperature curing. The measured effective CTE was in good agreement with the value measured from the locally heated aluminium aircraft structure. Some amount of residual bending was observed with the repairs cured over 100 °C.
Several authors have used Eq. (3) in estimating residual thermal stresses [1], [2], [3], [8], [37], [38]. They have also shown that the values
Conclusions
An experimental investigation was conducted to characterize the effect of several parameters on fatigue crack growth of pre-cracked aluminium plates repaired with multidirectional c/e and b/e composite patches. We draw the following conclusions from the study:
- •
The steel bonding rig we developed in order to control the thermal expansion of the repaired plates restricted the expansion to the level measured from locally heated real aircraft structures.
- •
The residual thermal stresses we varied using
Acknowledgements
This research was sponsored by the Finnish Air Force. The authors acknowledge the FINAF Air Materiel Command for the possibility to accomplish this work. The support and help of Patria Aviation is also highly appreciated.
References (44)
- et al.
Fatigue crack growth in patched plates under constant amplitude and flight simulation loading
- et al.
The reinforcing effect of composite patch repairs on metallic aircraft structures
Int J Adhes Adhes
(1992) - et al.
Fatigue crack behavior of cracked aluminium plate repaired with composite patch
Compos Struct
(2002) - et al.
A study on the fatigue crack growth behavior of thick aluminum panels repaired with a composite patch
Compos Struct
(2003) - et al.
Modeling and validation of composite patch repair to cracked thick and thin metallic panels
Compos A
(2009) - et al.
Investigation into cracked aluminium plate repaired with bonded composite patch
Compos Struct
(2007) - et al.
Comparison between double and single sided bonded composite repair with circular shape
Mater Des
(2011) - et al.
Finite element analysis of cracks in aging aircraft structures with bonded composite-patch repairs
Compos B
(2011) - et al.
Residual strength and fatigue life assessment of composite patch repaired specimens
Compos B
(1999) Design and validation of composite patch repairs to cracked metallic structures
Compos A
(2009)
Silane based field level surface treatment methods for aircraft and naval grade aluminium. Titanium and steel bonding
Int J Adhes Adhes
Numerical analysis of the effect of thermal residual stresses on the performances of bonded composite repairs in aircraft structures
Compos B
Design, analysis and performance of adhesively bonded composite patch repair of cracked aluminum aircraft panels
Compos Struct
Effects of pre-moulded woven CFRP patches on fatigue crack growth and residual strength in 7475-T761 aluminium alloy sheet
Fatigue Eng Mater Struct
Composite patch repair of thick aluminium structures
AIAA
Bonded composite patch geometry effects on fatigue crack growth in thin and thick aluminum panels
Technol Sci Press SL
Cited by (18)
Quasi-static and fatigue analyses of aluminium structures repaired by CFRP patch bonding
2023, Composite StructuresThe fatigue failure study of repaired aluminum plates by composite patches using Acoustic Emission
2019, Engineering Fracture MechanicsCutting force modelling in machining of fiber-reinforced polymer matrix composites (PMCs): A review
2019, Composites Part A: Applied Science and ManufacturingCitation Excerpt :According to the types of fibers, fiber-reinforced PMCs mainly include carbon fiber reinforced polymer (CFRP) [8], glass fiber reinforced polymer (GFRP) [9], Kevlar fiber reinforced polymers (KFRP), or Aramid fiber reinforced polymers (AFRP) [10]. Besides, Basalt fiber reinforced polymers [11] and Boron fiber reinforced polymers [12] are also widely used in industry. CFRP and AFRP have higher specific strength, higher specific stiffness and lighter weight than GFRP, but they are more expensive.
Boundary element analysis of fatigue behavior for CFRP-strengthened steel plates with center inclined cracks
2018, Thin-Walled StructuresCitation Excerpt :Chen et al. [8] conducted both numerical and experimental studies on the fatigue behavior of non-load-carrying cruciform welded joints reinforced by CFRP sheets, and it was demonstrated that the overlays had considerable effects on retrofitting. Aakkula et al. [9] proposed a series of fatigue experiments on boron fiber-reinforced polymer (BFRP) and CFRP-repaired aluminum plates, and the results indicated that CFRP and BFRP patching enhanced the fatigue life up to 6 times. In terms of steel members, Yu and Wu [10] carried out fatigue tests on cracked steel beams with various strengthening techniques.
Buckling behavior of CFRP-aluminum alloy hybrid tubes in axial compression
2017, Engineering StructuresCitation Excerpt :Efficiently combining FRP (fiber-reinforced polymer) and traditional materials, such as concrete and metals, to create hybrid structural members can provide favorable properties, including higher bearing capacities [1], greater buckling resistances [2–7], longer fatigue lives [8–12], and greater energy absorption [13].