Physics-based modelling and optimisation of shimming operations in the assembly process of aircraft skin panels

Assembly process of aeronautical skin panels deals with large, thin and compliant components, which are usually joined with rivets. A leading challenge is the control of part-to-part gaps prior to riveting operation, which must be maintained below tight design specification limits to avoid excessive pretensions of the rivets which, if exceeded, impair the durability of the whole skin assembly. Gaps are compensated by number of time consuming and costly manual inspection-repair quality loops, which involve measuring gaps, disassembling parts, adding be-spoke shims, re-assembling parts. This paper proposes a novel methodology to support the inspection-repair quality loops with the aim to model and optimise the shape of the shims with the ultimate goal of reducing/eliminating manual and trial-and-error measurements as per today best practice. The methodology will be discussed in two steps: (1) physics-based variation simulation to model generation and propagation of dimensional and geometrical variations (by using scanning data o morphing mesh model) during multi-stage assembly operations; (2) virtual shimming simulator to model and optimise shimming condition between parts being assembled. The proposed methodology is presented and validated using the assembly process of the vertical stabiliser for commercial aircrafts.