Host Institution: The University of Nottingham
Start Date: 1st April, 2017
Duration: 36 months
Lead Investigator: Thomas Turner
Co-Investigators: Adam Joesbury, Andreas Endruweit, Anthony Evans, Shimin Lu, Usman Shafique
This core project investigates the rate and quality limiting factors concerned with the manufacturing of components via Automated Dry Fibre Placement (ADFP). It has been established that the fibre deposition rate is typically less than 1m/s for commercial systems, governed by dynamic limitations such as heavy deposition heads and physical limitations, such as low material adhesion.
However, robustness of the technology is also a primary issue, as machine downtime contributes to limiting production rates. There is the potential to build on other work in related fields as some of these factors are not exclusively limited to the ADFP process and are being addressed by automation and machine design within the manufacturing industry, including the drive towards Industry 4.0.
The project work packages have been developed to align with each stage of the manufacturing process.
These include 4 key areas:
The overall aim is to understand the rate and quality limiting effects in the ADFP process, by developing numerical models to increase understanding of the critical factors. The project has the following objectives:
Previous development of the ADFP process has focussed on optimisation of materials to suit the existing deposition heads, this has resulted in a costly material which behaves much like a pre-impregnated slit fabric. This project attempts to rethink the process philosophy from first principles by developing deposition technologies which use materials in their lowest cost raw form. This has required the development of multi-physics models to describe the complex behaviour of the easily deformed materials.
Using a novel real-time control methodology the developed models are able to impact real-time operation of the developed sensor-rich ADFP test rig in order to create preforms with higher quality than existing methods and to build a digital twin of the preform to collect manufacturing data and inform downstream processes.
The developed test rig uses a post-processor-free approach where parts are manufactured direct from the CAD model of the part, this is facilitated by a new data format for storing ply / course data.
A novel method for determining the infusion characteristics of the manufactured preforms is also under development. The permeability of gapless preforms is very low and so the infusion process can be lengthy for complex parts. A detailed study is underway into the optimisation of gaps within the preform which serve to improve the permeability without compromising mechanical properties. In the longer term, computationally efficient models of actual as-deposited preforms will inform the infusion process on the shop floor.