Development of fibre and particulate filled aerogel composites for subsea pipeline applications

Student thesis: Doctoral Thesis


The aim of this project is to develop an analytical tool through Finite Element Method to understand and guide the design of thermal insulation materials. In particular, fibrous mat and aerogel particle-filled resin are focused in the context of subsea pipe-in-pipe (PiP) application for oil and gas extraction. The former has become part of a novel combination with super-insulating material - aerogel, and found its commercial use in the annulus region of the PiP. The latter is considered to be a form of upgrade to the current nylon-based centraliser stabilising the PiP configuration under hydraulic pressure from the deep seawater. The use of porous materials (e.g. fibrous matt and foam) as thermal insulation has a long history, A randomly orientated fibre mat is effective insulation due to the lack of a straight heat conduction pathway and relatively small pores in the fibrous structure. Recent development of super insulating materials involves combining fibrous mat with aerogel, which has resulted in a flexible and less compressible aerogel-fibre blanket with lower thermal conductivity (14 - 20 mw/mK) than typical conventional fibrous counterparts (35 - 50 mw/mK). However, a fundamental understanding of how the fibrous mat interacts the aerogel has not been extensively studied. The use of porous particle-filled resin composites, particularly with aerogel particles, has received less studies in thermal cases and a similar understanding of such integration must be developed in order to inform the design process. The rationale for carrying out this research was to create a range of simulations that can be easily used to predict the effect of changing properties of these above composites on their thermal and mechanical performance. The use of simulations is less time consuming and allows more parameters to be easily varied to see what the optimal fibre design is for specific applications. In order to simulate the effects of varying the composition of the mats, an algorithm to generate a random fibre mat was produced, based on the work of Arambakam and Tafreshi [1]. This fibre positioning was done in MathWorks' MATLAB, which was used to generate a script for producing a 3D geometry that can be incorporated in ANSYS APDL. This geometry was then used as the basis for a 3D finite element method model that utilised ANSYS Mechanical to mesh and solve the simulation. The same fibre generation process was used for both the mechanical and the thermal simulations, but with the bounding region being differently shaped to accommodate representing the standard experimental techniques used to measure these values. Simulations were carried out to investigate the effect of various parameters of the fibre mat. Specifically, the fibre orientation and the volume fraction of the mat were the main parameters of importance, with the orientation both in and out of the plane of heat transfer being investigated. The ratio of straight fibres to sinusoidal fibres was also investigated as a key parameter affecting the heat transfer, with the volume of straight fibres to total fibre volume being used to determine the "straight fibre fraction" of the fibres. The effect of the fibre length and diameter were also investigated, though it was found that on the scales investigated, they had very little effect on the thermal conductivity. A study into using fully random fibres, where the fibre can be represented by a continuous curve in 3D space that varies in direction throughout the length, was carried out. However, the high level of complexity of this fibre configuration along with the large number of fibre intersections meant that meshing the geometry produced was very difficult. Some successful meshes were produced; however, they included a number of elements that was too large to successfully
Date of Award28 Sep 2021
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorLiu Yang (Supervisor) & Ashleigh Fletcher (Supervisor)

Cite this