Towards a Geometry Assurance Framework for Compliant Composite Assemblies

Licentiate thesis, 2026

The increasing adoption of carbon fiber reinforced polymer (CFRP) composites in aerospace structures introduces geometric challenges that must be managed to ensure dimensional quality in production. During manufacturing, process-induced deformations such as spring-in and warpage cause composite parts to deviate from their nominal geometry, and these deviations are systematic rather than random, forming a design-specific manufacturing signature. When deviated parts are assembled, gaps form at the mating interfaces. Installing fasteners across open gaps induces forced-fit stresses, so gaps are filled with shims before fastening. Shimming, however, introduces its own structural penalty, increasing the bending stress carried by the fasteners as shim thickness grows. The interface between mating parts must therefore satisfy two competing requirements at once: post-assembly geometric quality and structural joint integrity.

This thesis develops simulation-based methods to support geometry assurance in shimmed compliant CFRP assemblies, with focus on managing this interface from both the assembly and the design stage. Two research questions are addressed. The first establishes the state of the art in geometry assurance for composites, identifying the variation sources and simulation challenges that distinguish composite assemblies from metallic ones. The second examines how interface management can balance geometric quality and joint integrity, from both the assembly and the design stage.

The work is based on three appended papers. In Paper A, a structured literature review maps the open challenges in geometry assurance for composites and shows that the systematic manufacturing signature, which has no counterpart in metallic variation simulation, propagates into the assembly as the interface gap that must be managed. In Paper B, using a CFRP wingbox section as a virtual case study, fixture adjustment is then shown to reduce interface gaps by 43% while respecting the structural shim limit, but this gap reduction increases post-assembly geometric deviation in the majority of cases, revealing a conflict between local gap minimization and global geometric accuracy. In Paper C, at the design stage, a non-monotonic relationship between the design gap and post-assembly deviation is identified, with an optimal gap of approximately 0.3 mm that minimizes deviation and remains consistent across manufacturing populations with different signatures, while larger gaps add shim thickness, and with it fastener loading, without geometric benefit.

Together, these results show that interface management can balance geometric quality and joint integrity through complementary levers at the design and assembly stages, quantified rather than left to engineering judgment, and lay the groundwork for integrating process-induced deformation predictions into assembly geometry assurance.