Search

Geopolymers are an economical, durable and environmentally friendly building material that achieves ultimate strength superior to concrete within hours and can withstand extreme high temperatures without degradation, but they lack elasticity and have poor flexural strength. To tackle these challenges, this paper presents a novel nanotechnological and affordable method to produce an elastic geopolymer that can be used as a construction material. Specifically, we report a simple and straightforward strategy that combines electrospinning and the sol-gel process to construct geopolymer nanofibers with superelasticity, extreme durability, and fire resistance. In order to improve the safety level of building construction sites that experience extreme conditions, such as freeze-thaw deterioration, the geopolymer nanofibers can be further functionalized by incorporating various functional components. In this study, Fe3O4 nanoparticles (NPs) adsorbed by an electric double layer structure were dispersed in a geopolymer sol (SAPOs). Elastic geopolymer nanofibers embedded with Fe3O4 nanoparticles were then prepared (Fe3O4/SAPOs) by an electrospinning method. Nanofibers constructed by such a compound method could have superparamagnetic behaviors and stable chemical properties. When Fe3O4/SAPOs are exposed to an alternating magnetic field, eddy currents are generated that cause the geopolymer nanofibers to heat up. This self-heating behavior is a powerful feature that enhances the freeze-thaw resistance of building materials at high altitudes. The nanotechnological route of this study could provide novel insights into the development of advanced construction materials.

Natural-fiber-reinforced geopolymer composites are environmentally friendly materials that have garnered considerable attention. In practice, the drying shrinkage and uneven dispersion of fibers in geopolymers result in low flexural strength, which limits the use of large-scale composites. In this study, high-flexural strength geopolymer composites with self-assembled nanofiber networks were hybridized using in situ polymerization. Composites with different fiber contents were evaluated via morphology and mechanical strength assessments. A mixed nanofiber and geopolymer solution was created to achieve uniform dispersion, and then the geopolymers were optimized at the nanoscale level using a bottom-up approach to achieve self-assembled nanofiber networks with good interfacial bonding. Due to the water retention properties of the nanofibers and subsequent gel material, the networks facilitated internal curing of the geopolymers, thus preventing drying shrinkage. This ultimately reduced the weight loss and increased the density of the composites. In addition, high flexural strength properties were observed in composites with a fiber content of 0.5%, due to the combination of fiber reinforcement and porosity. Thus, this work introduces new considerations for geopolymer applications.

The widespread application of deep neural network (DNN) techniques is being challenged by adversarial examples, the legitimate input added with imperceptible and well-designed perturbations that can fool DNNs easily in the DNN testing/deploying stage. Previous adversarial example generation algorithms for adversarial white-box attacks used Jacobian gradient information to add perturbations. This information is too imprecise and inexplicit, which will cause unnecessary perturbations when generating adversarial examples. This paper aims to address this issue. We first propose to apply a more informative and distilled gradient information, namely integrated gradient, to generate adversarial examples. To further make the perturbations more imperceptible, we propose to employ the restriction combination of $L_0$ and $L_1/L_2$ secondly, which can restrict the total perturbations and perturbation points simultaneously. Meanwhile, to address the non-differentiable problem of $L_1$, we explore a proximal operation of $L_1$ thirdly. Based on these three works, we propose two Integrated gradient based White-box Adversarial example generation algorithms (IWA): IFPA and IUA. IFPA is suitable for situations where there are a determined number of points to be perturbed. IUA is suitable for situations where no perturbation point number is preset in order to obtain more adversarial examples. We verify the effectiveness of the proposed algorithms on both structured and unstructured datasets, and we compare them with five baseline generation algorithms. The results show that our proposed algorithms do craft adversarial examples with more imperceptible perturbations and satisfactory crafting rate. $L_2$ restriction is more suitable for unstructured dataset and $L_1$ restriction performs better in structured dataset.