Super-strong biomimetic bulk bamboo-based composites by a neural network interfacial design strategy

As a sustainable ecological material, bamboo has become a popular modern green building material because of its rich yield, lightweight, high strength and rich cultural heritage. However, due to the limitation of bamboo tube thickness, multiple thickness-direction laminations are usually required to achieve large-sized materials, which leads to a significant decrease in strength. Therefore, it is urgent to find a way produce high-strength bamboo engineering composites on a large scale. Herein, a neural network interface design strategy was proposed, and a mechanical dissociation and partial matrix removal pretreatment method was used to open the weak intercellular layer and bamboo cell wall layer to increase the resin permeation channels. This allowed the resin to form a multi-scale bonding interface between multiple dense bamboo layers, achieving the preparation of bulk bamboo-based composite with adjustable dimensions and properties. The neural network-like bonding interface could firmly fix the compressed bamboo cells and enhance the mechanical properties of the bamboo cell wall and intercellular layer of bamboo, resulting in a tensile strength of 853 MPa for the composite, which was nearly three times that of natural bamboo and significantly superior to many structural materials such as alloys and other bamboo-based composites. In addition, this material showed good mildew resistance, flame retardancy and dimensional stability. This large-size bamboo composites are easy to scale production, which can be used in fields such as wind turbine blades, building structures, and outdoor walkways in the future.