18 TH INTERNATIONAL CONFERENCE ON COMPOSITE MATERIALS NUMERICAL INVESTIGATION OF FIBRE-METAL LAMINATES SUBJECT TO IMPACT DAMAGE M. Rathnasabapathy 1 *, A.P. Mouritz 1 , A.C. Orifici 1 1 School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, GPO Box 2476 Melbourne, Victoria Australia 3001 * Corresponding author ( minoo.rathnasabapathy@rmit.edu.au ) Keywords : Fibre-metal laminates, Damage tolerance, Impact behavior 1 Introduction 2 Numerical Modelling Fibre-Metal Laminates (FMLs) are a hybrid of metal 2.1 Modelling Strategy and composite laminates that are increasingly being used in aerospace applications. Consisting of Numerical analysis was conducted using the alternating layers of thin metallic sheets and fibre- commercial finite element solver Abaqus/Explicit reinforced epoxy composite prepreg, the two main [1] for the evaluation of impact damage to the FML types of FML are aramid fibre-reinforced epoxy specimens. Hexahedral solid elements (C3D8R) /aluminium laminates (ARALL) and S-2 glass fibre- were used for the aluminium layers. The isotropic reinforced epoxy/aluminium laminates (GLARE). elastic-plastic properties of aluminium were The combination of mechanical properties of modeled using the isotropic plasticity model in monolithic metal and fibre-reinforced composite Abaqus. provides FMLs with mechanical advantages such as low density, high strength, and high damage Hexahedral continuum shell elements (SC8R), each tolerance. having eight nodes and three degrees of freedom, were used for the glass/epoxy composite ply layers. Impact damage is a key concern for aerospace In order to accurately analyse the through-thickness structures. The inability to visually detect interior shear stresses resulting from the impact, continuum damage to composite layers, sometimes extending shell elements were selected over standard 4-node well beyond the impacted area, remains an important shell elements [2]. Due to the excessive distortions safety issue. Therefore it is necessary to accurately of elements, an enhanced stiffness relaxation method predict internal impact damage to FMLs. Due to out- was applied for hourglass control [1]. The Hashin of-plane loads, such as impacts, FMLs may suffer [3] failure criterion was implemented in Abaqus to damage in the form of different mechanisms such as: model the progressive intralaminar damage of the (i) plastic deformation of the metal layers; (ii) matrix composite layers due to impact. cracking and fibre failure; (iii) delamination between composite plies; and (iv) debonding of the metal and The adhesive bond between the glass/epoxy plies composite layer. and aluminium sheets was modelled using the traction-separation cohesive law in Abaqus to A Finite Element (FE) model was developed to represent the mechanical response of the adhesive analyse the complex damage responses and under impact loading. The thickness of the adhesive deformation that lead to the strength and stiffness is considered negligible in this analysis, and loss of FML structures. The accuracy achieved therefore the surface-based cohesive contact through FE analysis increases the reliability of capability was used to model the delamination. A numerical models to simulate impact loads onto detailed element sensitivity study for the numerical FMLs, enabling the reduction of time and costs models was conducted using the different element associated with mechanical testing. selections available in Abaqus.
NUMERICAL INVESTIGATION OF FIBRE-METAL LAMINATES SUBJECT TO IMPACT DAMAGE 12.7 J Experimental [4] 2.2 Model Validation 8 16.3 J Experimental [4] 7 Wu [4] conducted a series of low velocity impact 24.2 J Experimental [4] tests to evaluate the deformation and damage 6 12.7 J Numerical responses of FMLs. A comparative study was Simulation Force (kN) 5 16.3 J Numerical conducted using the experimental impact results of Simulation Wu [4] for Glare 5 2/1 to validate the numerical 24.2 J Numerical 4 Simulation model predictions. Glare 5 2/1 consists of two layers 3 of Al 2024-T3 aluminium alloy sheet and one layer of [0/90/90/0] glass/epoxy composite. Fig. 1 shows 2 the Glare 5 2/1 lay-up configuration used by Wu [4]. A 76 × 76 mm 2 square test FML specimen was 1 clamped between two steel plates exposing a circular 0 central impact region with a diameter of 50 mm. A 0 1 2 3 4 5 6 7 8 steel spherical impactor of 16 mm diameter with a Time (ms) mass of 6.29 kg was used with impact energies varying from 7 to 40 J. Fig.2. Comparison of numerically predicted force- time history of Glare 5 2/1 and experimental results [4] for varying impact energies Further comparison was conducted on the impact damage region of glass-epoxy composite and plastic deformation of the non-impacted side of the aluminium. Fig. 3. shows the increase in damage area for Glare 5 2/1 with an increase in impact energy. The numerical model shows good agreement with the experimental results of Wu [4]. From this comparison, the FE model satisfied the validation Fig.1. Glare 5 2/1 lay-up configuration [5] process showing the capability to accurately analyse the impact event and subsequent damage area of Comparison for contact force-time histories for FML structures. Glare 5 2/1 subject to impact showed good agreement between the experimental results [4] and numerical predictions, as shown in Fig.2. The force- 400 time history for Glare 5 2/1 shows the characteristic 350 Damage Area (mm 2 ) initial rise in force to a maximum value, followed by 300 a sudden load drop. The sudden load drops after 250 peak force indicates the local strength of the FML 200 Experimental structure was reached and energy starts to dissipate Results [4] 150 in the form of irreversible damage to the aluminium 100 Numerical and glass/epoxy layers [6]. At 24.2 J, the Simulation 50 experimental results show a sharp load drop at 3 ms. Despite the incorporation of the Hashin failure 0 criteria in Abaqus, the numerically predicted force- 0 10 20 30 40 time history did not replicate the sharp load drop, Impact Energy (J) characteristic of the stiffness degradation at this impact energy. This may be due to the cracking of Fig.3. Comparison of experimental results [4] the non-impacted aluminium side of the FML and numerically predicted damage area as a structure which is not explicitly modelled in the FE function of impact energy model [5]. 2
NUMERICAL INVESTIGATION OF FIBRE-METAL LAMINATES SUBJECT TO IMPACT DAMAGE 3. Numerical Analysis Table 3 Fracture energies for glass/epoxy [7] A FE model was developed to analyse the impact Parameter value, N/mm event and resulting damage of FMLs. Studies were G fl,t G fl,c G ft,t G ft,c conducted on two FML variants: FML 3 12.5 12.5 1.0 1.0 [Al/0/90/Al/0/90/Al] and FML 5 [Al/0/90/90/0/Al]. Specimens were modelled using Al 2024-T3 Table 4 aluminium alloy and S2 glass/epoxy prepreg. The Material properties of adhesive [7] average thickness of each aluminium layer was � t f t f E n G s G n s 0.406 mm and each ply of S2 glass/epoxy prepreg MPa MPa MPa N/mm N/mm was 0.3 mm thick. Two sides of the specimen were 2000 0.33 50 50 4.0 4.0 clamped. Taking advantage of the symmetry, only one-quarter of the FML test specimen was modelled 4 Results and Discussion with appropriate boundary conditions applied. The mass and diameter of the spherical steel impactor 4.1 Impactor force versus time history were 6.15 kg and 12.7 mm, respectively. The impact energies ranged from 10 to 30 J. The impactor force-time history calculated using the FE model is shown in Fig.4 for FML 3 and FML 5. Isotropic elastic properties for the Aluminium 2024- After impact initiation, the impactors velocity is T3 are: Young’s Modulus E = 73800 MPa and reduced as it comes in contact with the FML Poisson’s ratio � = 0.33. The unidirectional structure. This deceleration of the impactor is glass/epoxy composite exhibits transversely associated with the reaction-force on the impactor as isotropic behavior [7]. The corresponding elastic kinetic energy is transferred to the FML structure. properties, ultimate stresses and fracture energies for Small changes in the slope of the force-time the glass/epoxy are given in Table 1-3. The material histories for FML 3 and FML 5 are suspected to be properties of the adhesive used in the FE analysis are caused by the start of delamination and fibre failure shown in Table 4. The traction-separation cohesive [11]. law is characterised in terms of peak failure strengths, t f n and t f s and fracture energies G n and G s , The small load drops seen at 3.9 ms in Fig. 4b. where n and s refer to the normal (Mode 1) and correspond to fibre fracture in the glass/epoxy plies shear (Mode 2) directions [7]. shown in Fig. 5. Although no sharp load drops indicating cracking of the outer non-impacted Table 1 aluminium were seen to occur for both variations of Elastic material properties of glass/epoxy [7-10] FML, analysis of the plastic strain in Fig. 6. shows Materials Parameter Values, GPa significant damage of the outer aluminium layer in E 22 G 12 G 23 � 12 � 23 E 11 the impact region. No full impactor penetration was S2 glass/epoxy 55 9.5 5.5 3 0.33 0.33 recorded at these impact energies. Table 2 Ultimate strength properties for unidirectional glass/epoxy [7-10] Parameter value, MPa � L u,t � L u,c � T u,t � T u,c � LT u = � TT u 2430 2000 50 160 50 3
Recommend
More recommend
