Friction surfacing is a new variation of friction stir processing for surface property modification of metallic substrates. There is an increasing body of literature about friction surfacing by deposition of metal from a consumable tool to a solid substrate. Friction surfacing has many potential applications in joining, coating for corrosion resistance, and repair of degraded components. This article presents a review of the basic principles and latest research organized by processing techniques and variations, thermomechanical transfer and deposition of material, and finally metallurgical, mechanical, and chemical properties of the resulting deposition. Different friction surfacing processes are reviewed of novel tool–substrate configurations for material deposition for noncoating purposes like keyhole filling and joining dissimilar materials. Possible future topics of study for this area are discussed, which include deeper understanding of material transfer through metallurgy, FEM, and scale up of the technique for practical application.

The friction surfacing technique is an advanced method for creating coatings of various materials onto the surface of a similar or dissimilar material substrate. In this method, there is no external source of heat energy, and all the heat energy required in this method is generated by friction. In this paper, a novel method of friction surfacing from the side of the consumable tool is introduced. The most significant difference in this technique is that material transfer will occur from the radial surface of the consumable tool as opposed to the end of the tool as in conventional friction surfacing. In lateral friction surfacing, the side of the rotating consumable tool is pressed against the substrate surface, which generates frictional heating and shear forces at the interface between tool and substrate. A layer of tool material is transferred from the consumable rod to the substrate surface as the tool moves across. In this study, 6063 aluminum alloy and AISI 1018 carbon steel are used as the materials of consumable tool and substrate, respectively. The impact of process factors, surface roughness values, tool mass loss, and deposition thickness are discussed in detail. The experimental results of this study reveal that lateral friction surfacing produces a very smooth ultra-thin deposition with full coverage, with coating layers with roughness values in the order of 1 µm. Additionally, there is no flash formed in this technique which reduces material consumption. Moreover, temperatures at the interface between the consumable tool and workpiece were measured to be lower than for that in friction surfacing from the end of the tool, which is beneficial for the metallurgical characteristics of the deposited material.

In transportation sector, there is an increasing need for joining dissimilar materials for lightweight structures; however, substantial barriers to the joining of dissimilar materials have led to an investigation and development of new joining techniques. Friction stir blind riveting (FSBR), a newly invented method, has shown great promise in joining complex structures with dissimilar materials. The process can be utilized more effectively if knowledge regarding the failure mechanisms of the FSBR joints becomes available. This research focuses on investigating the different mechanisms that lead to a failure in FSBR joints under lap-shear tensile tests. An in situ, nondestructive, acoustic emission (AE) testing method was applied during quasi-static tensile tests to monitor the initiation and evolution of damage in FSBR joints with different combinations of dissimilar materials (including aluminum, magnesium, and a carbon-fiber reinforced polymeric composite). In addition, a fractographic analysis was conducted to characterize the failure modes. Finally, based on the analysis, the distinct failure modes and damage accumulation processes for the joints were identified. An AE accumulative hit history curve was found to be efficient to discriminate the deformation characteristics, such as the deformation zone and failure mode, which cannot be observed through a traditional extensometer measurement method. In addition, the AE accumulative hit history curve can be applied to predict the failure extension or moment of FSBR joints through an identification of the changes in curve slope. Such slope changes usually occur around the middle of Zone II, which is defined in this study.

The model-based, rapid-prototyping-enabled design and manufacture of a pulsatile blood vessel (PBV) for high-fidelity mannequin-based clinical simulations is presented. The PBV presented here is a pressurized, flexible tube with alternating fluid pressure created by a pump to mimic the behavior of a human vessel in response to pulsatile pressure. The use of PBVs is important for the fidelity of a clinical simulator that requires residents to palpate and/or access the vessel. In this study, a PBV is presented which features the integration of 3D modeling using patient-specific computed tomography (CT) data, mold fabrication using rapid-prototyping, and finite element method for estimating the required pumping pressure to generate the same level of force (about 1.5 N) experienced by the user through palpation. The relationship between this palpation force and the vessel pressure is studied using two strategies: finite element analysis (FEA) and experiments in a femoral arterial access simulator with a pump, artificial vessel, and surrounding phantom tissue. The experimental results show a discrepancy of 8.7% from the FEA-predicted value. Qualitative validation is done by exposing and surveying 19 interventional cardiology residents at four major educational institutions to the simulator for accuracy of its feel. The overall survey results are positive.

Friction stir forming (FSF) is a new environmentally friendly manufacturing process for lap joining of dissimilar materials. Fundamentally, this process is based on frictionally heating and mechanically stirring work material of the top piece in a plasticized state to form a mechanical interlocking joint within the bottom material. In this research, the significant process parameters were identified and optimized for Al 6014 alloy and mild steel using a design of experiments (DOE) methodology. The overall joint structure and grain microstructure were mapped as the FSF process progressed and the aluminum work material deformed through different stages. It was found that the work material within the joint exhibited two layers, thermomechanical affected zone, which formed due to the contact pressure and angular momentum of the tool, and heat affected formation zone, which was composed of work material formed through the hole in the steel sheet and into the anvil cavity. Two different geometries of anvil design were employed to investigate geometrical effects during FSF of the aluminum. It was found that the direction and amount of work material deformation under the tool varies from the center to the shoulder.

Cold rolling is widely recognized as a key industrial process for enhancing the mechanical properties of materials, particularly hardness, through strain hardening. Despite its importance, accurately predicting the final hardness remains a challenge due to the inherently nonlinear nature of the deformation. While several studies have employed artificial neural networks to predict mechanical properties, architectural parameters still need to be investigated to understand their effects on network behavior and model performance, ultimately supporting the design of more effective architectures. This study investigates hyperparameter tuning in artificial neural networks trained using Resilient Backpropagation by evaluating the impact of varying number of hidden layers and neurons on the prediction accuracy of hardness in 70-30 brass specimens subjected to cold rolling. A dataset of 1000 input–output pairs, containing dimensional and hardness measurements from multiple rolling passes, was used to train and evaluate 819 artificial neural network architectures, each with a different configuration of 1 to 3 hidden layers and 4 to 12 neurons per layer. Each configuration was tested over 50 runs to reduce the influence of randomness and enhance result consistency. Enhancing the network depth from one to two hidden layers improved predictive performance. Architectures with two hidden layers achieved better performance metrics, faster convergence, and lower variation than single-layer networks. Introducing a third hidden layer did not yield meaningful improvements over two-hidden-layer architectures in terms of performance metrics. While the top three-layer model converged in fewer epochs, it required more computational time due to increased model complexity and weight elements.