Search

Composite materials with complex internal microstructures, such as the flax nonwoven bio-composite studied in this work, require advanced numerical models in order to predict their mechanical performance. Otherwise, the micro-structural interactions that take place between their components makes very difficult to obtain their mechanical properties and failure mechanisms. This paper presents a novel methodology that couples two homogenization formulations: a phenomenological one, the serial-parallel mixing theory; and a numerical multiscale procedure. The resulting methodology has a minimal computational cost, while it is capable to account for the different interactions that take place among the composite constituents. With the proposed approach, it is possible to characterize the mechanical response of nonwoven composites and to predict their structural failure. The methodology developed is applied to a flax nonwoven biocomposite manufactured and tested by the German Aerospace Center (DLR). The good results obtained from the simulation, when compared with the experimental values, allow considering the proposed procedure an excellent approach for the analysis of large structures made with complex microstructures, such as nonwoven biocomposites.

The objective of this work is the tune of a methodology for the determination of the fracture toughness in ductile materials at high strain-rate by means of dynamic 3 Point Bending (3PB) tests. The specimens made in high strength steel have been pre-cracked in fatigue with a single edge notch. The dynamic loading conditions have been reached by means of a modified Split Hopkinson pressure Bar setup, in which two small diameter bars are used as transmitted bars instead of the standard single one. In order to obtain the maximum amount of information from each test, different measuring systems have been adopted, in addition to the strain measurements on the bars performed using semiconductor strain-gages. On the specimen surface, one or two crack-gages (one for each side for specimens) have been bonded for obtaining data about the crack propagation and the crack tip position vs. time curve. Moreover, a high speed camera has been used for obtaining the measurements of the Crack Opening Displacement (COD) and the opening velocity: digital images have been processed using tracking techniques following some points on the specimen notch and input bar tip. Starting from the experimental measurements in term of force, displacement and crack-gage data, a protocol for fracture toughness calculation at the initiation stage of the crack growth in dynamic condition has been defined and a hybrid experimental – numerical methodology has been used. One of the key variables for the determination of dynamic initiation fracture toughness is the crack time to initiation. Many efforts have been done in order to better determine a precise method to experimentally measure this entity: this datum has been obtained from the analysis of both the force and crack-gage vs. time signals and the COD measurement on high speed video. Good agreement has been found between the two methodologies. At this stage, a 3D model, which simulates the SE(B) specimens during the Hopkinson 3PB test, has been done in order to determine the toughness (J integral) as a function of time. Knowing the initiation time, the fracture toughness at crack initiation can directly be determined.

Static and fatigue properties are important factors when developing design rules to optimize the reduction of damage in the middle of a composite component loaded under a three point bending test. The progressive damage which occurs in a composite laminate during fatigue will affect the mechanical properties of the component to an extent which depends on the material type and lay-up of the composite and on the mode of testing. By choosing an appropriate combination of ply materials and ply stacking sequence for the composite laminate, a better static condition, as well as better fatigue behavior of the component is expected under in-service conditions. In order to minimize the through-thickness damage in the middle of a composite laminate under a short beam shear test, the present work focuses on determining the best stacking sequence, combining two different materials through-thickness of a composite component. The use of an algorithm that has been created to run all possible combinations of angles and materials was the key to finding the best stacking sequence for a composite component. Through the optimization method, two different hybrid laminate lay-ups were obtained for a maximum of 40% of glass plies within a composite laminate. The laminates developed through the algorithm, which are, the Hybrid Design 1([(0c)2/(±45c)2/(±45g)2/(±45c)]s ) and Hybrid Design 2 ([(0c)2/(±45g)/(90c)4/(±45g)/(90c)2]s ), have shown that the best ply orientations to minimize the damage in the middle of a composite component occur when the value of the variable angle θ in ±θ plies has a maximum value of 90 degrees. The algorithm used showed that best laminates were obtained when glass plies were not used as the central plies in a glass/carbon composite laminate, with a maximum 40% of glass plies. Static and fatigue tests were performed on both hybrid designs and on a quasi-isotropic composite laminate ([(0/45/90/-45)3 ]s) made of carbon plies. The results of the static and fatigue tests have demonstrated that the best design, giving the best static and fatigue behavior is the Hybrid Design 1. Also, it has been demonstrated through experimental tests that 90 degree plies within a composite laminate subject to a three point bending condition will cause the component to be weaker in terms of fatigue and static properties due to the tendency to develop inclined cracks in the through-thickness direction on the 90 degree plies of the composite laminate, causing delamination and loss of stiffness of the component.

The main objective of this study was the assessment of the stress-strain fields of the sandwich core during bending tests. Finite element models (FEM) using Siemens NX10 and digital image correlation (DIC) were used to predict the behavior of sandwich beams in 3 and 4 point bending. Two different core thickness and two different sandwich length (short and long beams) with basalt fiber reinforced polymer (BFRP) faces were simulated. The numerical simulation using 3D finite element were compared and validated with experimental results obtained previously by DIC. This work validated innovative numerical simulations and experimental techniques for determining stress-strain fields in composite sandwich beams subjected to bending. The correlation between experience and analysis allowed a better knowledge and understanding of the complex stress-strain fields along the core thickness of sandwich beams in bending.