Additive Manufacture of Compliant Mechanisms

Abstract

Additive Manufacturing (AM) is a technology currently used for bespoke creation of difficult to manufacture forms allowing lightweight structures as well as use in rapid development and component creation. Compliant Mechanisms (CMs) are a mechanical technology where flexure (elastic deformation) is used in place of multiple interlocking moving parts to perform a useful task. Although the concept is longstanding, modern computational tools and advanced manufacturing have renewed interest in CMs, with applications ranging from everyday products to aerospace mechanisms. However, barriers to widespread adoption remain, particularly in the accurate prediction of stresses, manufacturability of specific geometries, and concerns regarding durability and fatigue.
This thesis investigates how AM can be used to overcome these limitations by enabling the direct manufacture of additively manufactured compliant mechanisms (AMCMs).
A design framework was developed in the form of a flowchart to simplify the modelling and analysis of compliant elements, allowing rapid exploration of design alternatives. The flowchart is subdivided into several categories including ‘design’ and ‘evaluation’; to aid in each, a simplified finite element analysis (FEA) approach was created, where a 2D shell mesh centreline with an applied thickness was used to model the stress in an AMCM’s operation. These stress outputs were used to predict the behaviour of 6 printed AMCMs representing 2 redesign processes (referred to throughout as ‘consolidated’ and ‘generated’) for 3 different existing aerospace fittings (referred to throughout as ‘hook’, ‘snap latch’ and ‘double stud’).
Experimental verification was conducted using two AM processes (FDM and PBF) and materials (rPETG and an Aluminium alloy) to compare predicted and observed behaviours for these 6 designs. The test that was performed was a functionality test of the compliant mechanism, applying a displacement to ensure a full flexure (compression or extension as required) and return of the AMCMs range of motion as required by the design, this was repeated for up to 5000 cycles. This approach demonstrated that AMCMs can replicate the
Page 5 of 188
functions of equivalent rigid-body mechanisms while enabling new design possibilities, such as integrated flexible structures and optimised geometries, and can perform with a degree of predictability compared to the modelled behaviour. The numerical findings were as follows: for the generated aluminium designs, 2 of 3 reached the maximum number of cycles (over 5000) without breaking, while the snap latch failed below expected at ≈3000 cycles. Furthermore, the consolidated designs all failed at much lower values of ≈1900, ≈1900 and ≈60 for the hook, double stud and snap latch respectively.
The FEA approach correctly predicted the failure for all the consolidated aluminium designs, where it estimated required thinner walls than could be printed (or a larger scale than could be printed). The only outliers from the predictions of the FEA were the aluminium generated double stud, which was predicted to fail yet did not, and the generated aluminium snap latch, which was not predicted to fail, yet did. All rPETG designs surpassed the 5000 maximum cycles.
The findings highlight the potential for AMCMs in industries where lightweight, robust, and resource-efficient solutions are required, for example, in aerospace. In particular, the research identifies opportunities for applications in extreme or resource-constrained environments, where the ability to manufacture single-part, functional mechanisms is highly advantageous due to the logistical constraints. Finally, the thesis recommends further investigation into the proposed thermally responsive structure with a negative coefficient of thermal expansion to confirm its feasibility and expand the design space of AMCMs.

Date of Award	20 May 2026
Original language	English
Awarding Institution	University of Bath
Supervisor	Vimal Dhokia (Supervisor), Antonio Pellegrino (Supervisor) & Elise Pegg (Supervisor)