Air Force Institute of Technology

Experimental research was conducted on the fatigue life of Selective Laser Melted Ti6Al-4v. A thorough understanding of the fatigue life performance for additively manufactured parts is necessary before such parts are utilized as production end-items for real-world applications such as the rapid, on-demand, 3D-printing of aircraft replacement parts. This research experimentally examines the fatigue life of Ti-6Al-4v material specimens built directly to net shape and then either stress-relieved or Hot Isostatically Pressed (HIP). Experimental results will help determine whether HIP effectively reduces porosity and increases fatigue life when the specimen surface is not machined to remove surface roughness from the additive manufacturing process.

Although additive manufacturing presents a novel and customizable approach to fabricating tungsten (W) alloys, the layer-by-layer production technique presents a unique set of processing challenges which are further exacerbated by the metal’s refractory properties. In this study, additions of niobium (Nb) were found to improve the mechanical properties and oxidation resistance of additively manufactured W. Microstructural investigation revealed the effect of Nb on the cracking, grain size, and texture in the as-built material. Mechanical compression testing showed significant improvement in ductility with increasing Nb content. The low concentrations of Nb used did not induce large changes in the yield strength of the material. The work shows that Nb may be a useful alloying element for improving additive manufacturing of W.

We show that there does not exist a complex d× n equiangular tight frame with d^2-d+1<n<d^2. The proof, which originated from an internal model at OpenAI, mimics the relationship between real equiangular tight frames and strongly regular graphs.

Reconnaissance is a vital component in a comprehensive planetary defense mitigation strategy—aimed at preventing or reducing the impact threat posed by celestial objects on close-approach trajectories to Earth. It provides decision-makers with critical information on the physical and orbital properties of a near-Earth object (NEO), enabling better assessment of size, composition, and trajectory. This study investigates how to optimally pre-position a fleet of reconnaissance spacecraft prior to the discovery of a specific hazardous NEO. It evaluates response timelines and mission success rates for combinations of spacecraft launched from Earth and those maneuvering from pre-deployed locations within the Sun-Earth system. A synthetic population of asteroid threats is used to generate performance metrics for each candidate architecture. This study employs a nested multi-objective optimization framework coupling the fast elitist non-dominated sorting genetic algorithm (NSGA-II) for reconnaissance architecture design with particle swarm optimization (PSO) for transfer trajectory optimization. The developed Pareto-optimal architecture designs have trade-offs among response time, cost, and flyby success rate. Mission constraints consider approach lighting, flyby velocity, and Δ v . Results indicate that Earth-launched spacecraft provide the most cost-effective solution for rapid-response NEO reconnaissance. For a single-spacecraft configuration, an Earth-launched vehicle increases the success rate by 8.4

Tieyi Road, Haidian District, Beijing 100038, ChinaFederated Learning (FL) enables joint training over distributed devices without data exchange but is highly vulnerable to attacks by adversaries in the form of model poisoning and malicious update injection. This work proposes Secured-FL, a blockchain-based defensive framework that combines smart contract-based authentication, clustering-driven outlier elimination, and dynamic threshold adjustment to defend against adversarial attacks. The framework was implemented on a private Ethereum network with a Proof-of-Authority consensus algorithm to ensure tamper-resistant and auditable model updates. Large-scale simulation on the Cyber Data dataset, under up to 50% malicious client settings, demonstrates Secured-FL achieves 6%-12% higher accuracy, 9%-15% lower latency, and approximately 14% less computational expense compared to the PPSS benchmark framework. Additional tests, including confusion matrices, ROC and Precision-Recall curves, and ablation tests, confirm the interpretability and robustness of the defense. Tests for scalability also show consistent performance up to 500 clients, affirming appropriateness to reasonably large deployments. These results make Secured-FL a feasible, adversarially resilient FL paradigm with promising potential for application in smart cities, medicine, and other mission-critical IoT deployments.