Thompson, Samuel
(2025)
Transmissive solar sailing.
PhD thesis, University of Nottingham.
Abstract
Uniquely, solar sails are a means of accelerating a spacecraft via ambient space radiation rather than onboard propellant. In principle, this allows for continuous acceleration, and for the attaining of higher velocities than a reaction engine may achieve. These sails are also amongst the most sustainable orbit control systems, being associated with reduced satellite weight, mission cost and space debris footprint. However, they also have weaknesses that constrain how and where they can be used --- most notably, their poor performance in low Earth orbit (LEO), where the majority of satellite missions occur. {\textit{Transmissive}} solar sails may mitigate these weaknesses and thereby promote the broader adoption of solar sails. Advocating for this future, this thesis seeks to collate and expand the existing literature pertaining to these sails across three core domains: orbital dynamics, optical design, and manufacturing processes.
A flight model was developed to characterise the behaviour of differently designed and differently steered solar sails in LEO, with consideration given to atmospheric drag, eclipse and orbital precession during orbit-raising manoeuvres. These designs included transmissive sail proposals of both {\it{refractive}} and {\it{diffractive}} varieties and a contemporary {\it{reflective}} solar sail for comparison. Simplified `zero-\(\alpha\)' steering was found to constrain the flight envelope of transmissive sails. When optimally steered, these sails unanimously demonstrated lower performance sensitivity to altitude but greater sensitivity to orbital inclination than contemporaries. Greater synergy with certain Sun-synchronous orbits (SSO) was found, but also greater sensitivity to the orbital plane changes that arise in non-SSOs. Sails were sorted according to performance: some refractive and diffractive {\it{metasails}} surpassed reflectors in every flight regime ({\textit{Type A}} transmissive sails); simpler refractive and diffractive sails surpassed reflectors only in low altitude, high inclination orbits, particularly SSO ({\textit{Type B}}); other diffractive sails surpassed reflectors in low altitude orbits more generally, and were otherwise comparable to reflectors ({\textit{Type C}}). A case study with a $36\;\mathrm{m}^2$ Type A sail demonstrated transit times from LEO that were comparable to mid-range electric thrusters.
\par % Run this through Geebee and then move on to #3 and then properly move on:
The optical design of refractive sail patterns using single-index materials was explored in-depth via numerical optimisation: a ray tracing simulation was developed to calculate the solar radiation pressure (SRP) and torque per unit area generated by refractive objects illuminated in vacuum. A model-free reinforcement learning optimiser was developed to iterate upon geometries according to a user-defined fitness function. These tools were integrated to generate numerically optimised patterns ranging from high acceleration micro-prisms to passively stable freeform {\it{lightfoils}}. The optimiser was shown to substantially improve the performance of optical elements, usually by harnessing total internal reflection. In one notable example, a simple pattern of cylinders evolved into a pattern of freeform elements that achieved \(74\%\) of the theoretical maximum tangential SRP --- compared to the \(42\%\) efficiency reported for single-index refractive designs in prior literature. In another case, a single lightfoil was optimised for stability, and had its maximum corrective torque increased by \(50\%\) over that of the initial proposal. However, the optimisation process was shown to be highly sensitive to the configuration of both the simulation and the optimiser AI. For example, identical fitness functions applied to different initial geometries often resulted in very different solutions. Furthermore, geometries were seen to lose resolution during optimisation as a result of the optimiser removing `low reward' vertices from convex hull calculation. These findings demonstrate the validity --- but also the complexity --- of using model-free reinforcement learning for the design of non-imaging optical devices.
\par
The manufacturing of transmissive solar sails was investigated with a focus on feasibility, cost, and accessibility from the perspective of a small satellite developer. The viability and cost of executing processes in-house was compared with the cost of commercial outsourcing, and several in-house trials were conducted. For fabricating the unpatterned sail sections, thin film fabrication trials were carried out using doctor blading and a water-float method. For fabricating patterned moulds, greyscale lithography and dry plasma etching comprised the conventional manufacturing methods trialled, while digital light processing and two-photon polymerisation comprised the additive methods trialled. Finally, pattern transfer trials included silicone moulding, linear thermal nanoimprint lithography, and roll-to-plate UV nanoimprint lithography. The evaluation favoured a commercial solution for the thin film fabrication stage. For mould manufacture, several viable conventional and additive methods were discussed, but a clear winner was not identified. For the pattern transfer stage, procuring a rolling nanoimprinter for in-house use was favoured above a certain volume or size of sail. A prototype solar sail payload was also designed for the University of Nottingham CubeSat {\it{JamSail}} in preparation for a future in-orbit demonstration mission.
| Item Type: |
Thesis (University of Nottingham only)
(PhD)
|
| Supervisors: |
Cappelletti, Chantal
Tuck, Christopher
Pushparaj, Nishanth |
| Keywords: |
Solar sail, refractive, transmissive, ray tracing, numerical optimisation, optics, orbital dynamics, additive manufacturing |
| Subjects: |
T Technology > TL Motor vehicles. Aeronautics. Astronautics |
| Faculties/Schools: |
UK Campuses > Faculty of Engineering > Department of Mechanical, Materials and Manufacturing Engineering |
| Item ID: |
81160 |
| Depositing User: |
Thompson, Samuel
|
| Date Deposited: |
31 Jul 2025 04:40 |
| Last Modified: |
31 Jul 2025 04:40 |
| URI: |
https://eprints.nottingham.ac.uk/id/eprint/81160 |
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
