The study presents a novel hybrid deep learning and ensemble learning (DL-EL) model to predict the effects of chemical composition and thermo-mechanical processing on the properties of Cu–Ni–Si alloys. The model integrates various input parameters like chemical composition and thermo-mechanical processing parameters and aims to predict key output properties such as mechanical properties and electrical conductivity. This study addresses gaps in existing research by providing a comprehensive model for Cu–Ni–Si alloys and integrating thermo-mechanical processing parameters with chemical composition in the predictive model. The research demonstrates the model's superior predictive performance, with near-perfect R2 values on both training and test sets. The hybrid DL-EL model was compared with three other machine learning models and its efficacy was assessed using R2 value, offering new insights into the performance of these alloys. A feature importance analysis was conducted to identify the most influential features of the model. This work contributes to the material science field by providing insights into optimizing alloy composition and processing, leveraging machine learning to enhance material design.© 2017 Elsevier Inc. All rights reserved.

Polymer composites with high aspect ratio Na–Co oxide ceramic filler – polymer matrix was investigated in this study. They demonstrate high dielectric permittivity, low dielectric loss tangent values which is useful for capacitive field grading and reversible voltage dependent electrical conductivity properties under an applied electric field which can be utilized for resistive field grading. These unique properties are useful for electrical stress control applications in power components. This study employs novel ceramic platelets as the varistor filler component in a polymer composite form, introducing a new class of materials to the world of electrical stress control technology. By minimizing the amount of varistor filler material required to very low levels (15 wt%), it is possible to develop lightweight non-linear dielectric composites that not only exhibit improved electrical characteristics but also make processing easier and open up new design space for making various components for electrical power applications. By utilizing novel NaxCoO2 (x < 1) ceramic fillers in composite form, we aim to contribute to the development of innovative nonlinear dielectric composites that can enhance the performance and processibility of various power system components under various working environments.

Over the past years, a variety of energy harvesters have been used to scavenge energy from ambient vibrations, but most harvesters work effectively only near the resonant frequency. This work presents a soft piezoelectric energy harvester with a wide bandwidth by compounding lead zirconate titanate particles and shape memory polyurethane. To clarify the relation between energy harvesting performances and frequency, the developed harvester was tested from 20 Hz to 1300 Hz. The results show a bandwidth of 330 Hz, which is benefit from the microstructures of particulate composites. This work supplies an alternative method to develop broadband energy harvesters.

The majority of existing ionogels display weak toughness, which confines their potential applications. This manuscript presents a novel study on the development of ultra-stretchable, long-term stable and toughened composite ionogels manipulating hydration and hydrogen bond interactions. By leveraging the low evaporation rate of ionic liquids and their robust interactions with water, combined with the establishment of a hydrogen bond network between polyacrylic acid and fructose, the resulting optimized composite ionogel achieved a strain capacity of 2090% at room temperature. Remarkably, its toughness is further heightened under subzero temperatures (15.6 MJ/m³), representing a significant increase compared to its toughness at room temperature (4.4 MJ/m³). To exemplify its practicality, a wearable strain sensor is fabricated for the purpose of sensing and detecting motions in subzero temperature settings, potentially leading to the creation of freeze-yet toughened, ultra-stretchable materials for novel applications.

Joining titanium alloy with carbon fiber reinforced thermoplastic (CFRTP) is one of the hot subjects in the automotive and aerospace industries. To improve the joint strength of TC4/CFRTP PA6, this study combined the surface treatment of TC4 alloy, including laser ablation, Silane Coupling Agent (SCA) treatment, and interfacial PA6 adhesive. The results showed that the laser texturing induced the micro-scale porous and macro-scale groove grid structure, respectively. The superhydrophilicity of the laser texturing as well as amino polar groups induced by SCA treatment promoted the adhesion of the thermoplastic PA6 melt into the microstructure of the TC4 alloy surface. The mechanical interlocking and hydrogen bonding between TC4 alloy and CFRTP PA6 significantly improved the shear strength of the TC4/CFRTP PA6 joint (17.5 MPa). Furthermore, with the incorporation of interfacial PA6 adhesive, the joint strength was further enhanced, and the damage of the TC4/CFRTP PA6 joint was transformed from interfacial to interlaminar CFRTP PA6.

Through construction of a three-dimensional continuous ceramic network, the dielectric performance of polymer composites can be enhanced with a reduced amount of ceramic loading. However, the commonly employed methods, such as freeze casting or sacrificial template, fail to precisely control the structure of the ceramic skeleton, leading to significant randomness in content control. In this work, a Digital Light Processing (DLP) 3D printing technique is utilized to precisely fabricate ordered three-dimensional ceramic structures, which are further sintered at high temperatures to obtain a continuous barium titanate (BT) skeleton (BTS). This is then compounded with cyanate ester based polymers (CP) to prepare BTS/CP composites. Results show that when the content of BT is 61.2 vol%, the dielectric constant of the composite is 657 at 100 Hz, which is about 173 times higher than CP, while maintaining a lower dielectric loss of 0.026. Interestingly, the distribution of BTS in the composites has a significant impact on the dielectric performance of the composite. When the BT content is about 50 vol%, the dielectric constant of the “fine and dense” skeleton composite is 2.1 times higher than that of the “coarse and sparse” skeleton composite. This study proposes a novel strategy for building a structurally and compositionally controllable three-dimensional barium titanate (BT) network within polymers, demonstrating the noteworthy influence of macrostructure on the dielectric performance of polymer composites. This offers a fresh approach to the mass application of ordered three-dimensional ceramic skeleton enhanced polymer composites in the fields of electronics and electricity.

Three-dimensional woven composites (3DWC) have emerged as a promising lightweight solution for aviation structures prone to impact loads during operational life. However, effectively simulating the impact response and failure process of 3DWC presents a challenging task, requiring a delicate balance between computational efficiency and accuracy. In this study, a local-homogenization-based finite element modeling strategy for simulating the impact behavior of 3DWC panels was proposed. The model is constructed based on the fabric's architectural design, involving the decomposition of the unit cell into multiple homogeneous subcell elements aligned with the local yarn direction. The calibration exercises are conducted by simulating the mechanical responses of 3DWC under various monotonic loads and comparing the results with experimental data. Subsequently, the model is applied to predict the response of a 3DWC panel to low-velocity impacts, further corroborating the predictions through experimental validation. The modeling strategy proposed in this study provides an efficient method for analyzing and designing 3DWC structures.

Pore defects are inevitable in the manufacturing process and significantly affect the mechanical properties of composite materials. In this work, a novel algorithm capable of concurrently generating both ellipsoidal and gourd-shaped pores in composite materials is proposed and the interference detection procedure is optimized for efficient interference detection between the generated geometrical bodies. Besides, the algorithm is integrated with RVE-based finite element (FE) simulation to explore the effects of various pore types, volumes, aspect ratios, and porosity on the elastic properties of unidirectional C/C composites. The present work provides a valuable reference for generating general pore defects and investigating their resulting effects on mechanical properties of related composite structures.

The poor thermal shock resistance of coated C/CA composites limits the applications with rapid temperature fluctuations due to the huge thermal mismatch between the coating and the substrate. A facile strategy of tailoring coefficient of thermal expansion (CTE) of C/CA is proposed to match their thermal deformation behaviors. The optimized C/CA by heat treatment has a positive correlation between CTE and temperature, and thus the as-prepared SiC coating has a decreased macroscopic residual strain of 80.7% due to their coordinated thermal response compared to the original C/CA. Meanwhile, benefited from the structural improvement of larger pore size and higher degree of graphitization of C/CA, a significantly decreased tensile stress of 60.3% in the coating is achieved due to the formation of small SiC spheres in the carbon network. Resultantly, the coated C/CA shows remarkably improved thermal shock resistance with narrower and fewer microcracks, and lower mass and linear cyclic ablation rates of 0.252 mg/s and 0.607 μm/s, respectively.

We design auxetic 3D interlocking brick-and-mortar composites inspired by nacre - a natural strong and resilient high-performance material. The influences of geometrical parameters and base material properties on their effective elastic properties are systematically investigated by finite element simulations and computational homogenization. The numerical results indicate that the composites exhibit tunable negative Poisson’s ratios by properly tailoring the underlying microstructures. It is revealed that the auxetic behavior therein is attributed to the relative sliding between adjacent bricks with waviness, i.e., the sliding mechanism. These materials, with their auxetic properties and high fracture toughness, are well-suited for a wide range of engineering applications, such as serving as robust auxetic materials.