The rapid expansion of space activities requires manufacturing strategies that align environmental performance with engineering functionality, yet sustainability assessments of additive manufacturing (AM) remain affected by significant data uncertainty. This study presents an uncertainty-informed Life-Cycle Assessment (LCA) framework to evaluate the environmental and performance trade-offs between Laser Powder Bed Fusion (LPBF) and conventional CNC machining for a satellite mounting bracket. The assessment adopts a process-based cradle-to-gate approach and integrates a hybrid uncertainty propagation methodology combining Dempster–Shafer theory for epistemic uncertainty with Monte Carlo simulation for aleatory variability. Environmental impacts are represented as interval-valued outcomes with associated belief–plausibility measures, enabling explicit quantification of epistemic uncertainty. In parallel, a performance-based benefit metric based on stiffness-to-mass ratio is introduced and propagated under uncertainty using a consistent framework. Environmental and performance indicators are normalised and combined into a composite trade-off metric, allowing the evaluation of manufacturing alternatives across a range of environmental weighting scenarios. Decision outcomes are expressed in terms of belief and plausibility, capturing both support and indeterminacy under uncertainty. Results indicate that CNC machining exhibits lower midpoint environmental impacts and narrower uncertainty intervals across key categories, while LPBF shows higher potential impacts and substantially wider epistemic uncertainty, primarily driven by powder production and limited inventory data. However, when performance benefits are considered, LPBF may become preferable under specific trade-off conditions. These findings highlight the importance of explicitly accounting for epistemic uncertainty and performance considerations when evaluating sustainability trade-offs in aerospace manufacturing. The proposed framework supports early-stage eco-design by enabling robust decision-making under incomplete knowledge.
Alberti et al. (Tue,) studied this question.