Improving user-interaction with material extrusion prototypes through emulation of mass properties

  • Harry Felton

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

Additive manufacturing is widely used as a method for prototyping products – especially in early-stage design. The most common method is Material Extrusion (MEX), which deposits filament using a heated nozzle. The method is widely used due to its low cost, ease of use, wide range of processable materials, and safety. While the method allows an accurate geometric form of a prototype to be fabricated relatively quickly, the mass properties of the fabricated product are often poorly represented. This discrepancy between the ‘as-designed’ mass properties and the fabricated prototype can, depending on use, have a significant impact on a user’s perception of the product and ultimately the quality and commercial success of the final design. To address this shortcoming, this thesis develops a method that enables the fabrication of MEX prototypes that better emulate mass properties of the as-designed product.

The thesis begins with a user study to investigate the influence of each mass property (mass, mass balance, rotational inertia) on user perception of mass. The study reveals that, for consumer products, the principal rotational inertia has very little influence. Given this, the strategy for mass emulation aims at matching the mass and centre of mass position. Physical methods for modifying the mass properties were reviewed and down-selected based on their potential advantages and disadvantages. The use of variable infill was selected as full automation is possible and the method can be performed on multi-material printers – a capability that is becoming commonplace. The method was initially developed on a simple primitive, with a range of process parameters adjusted and tested spanning computation and manufacturing constraints. The process uses a directed search method, that iterates through potential centres of mass within the internal volume. Mass distributions are then generated and evaluated through Monte Carlo optimisation, using an exponential probability distribution. The developed method was applied to three prototypical products – a games controller, electric hand drill and laser pointer – with results compared to a nominal fabrication. The case studies demonstrated that mass and centre of mass position can be emulated to high accuracy (in most instances to within 1% of the "as-designed" value, determined from the CAD model). Example prototypes were then manufactured using the method, further demonstrating encouraging results. Scope for improvement is, however, identified, and future work is discussed.

In summary, a novel process has been developed that allows mass properties to be emulated in MEX fabrications of prototypes of consumer goods. The process has been applied to specific use cases, demonstrating significant improvement in mass property accuracy compared to the typical fabrications. The scientific contribution of this thesis is three-fold:

1. Characterisation of the influence of mass properties on user perception of mass.
2. Creation of a method for emulating the as-designed mass properties of a 3D printed prototype.
3. Application and demonstration of the methodology in the MEX workflow for a range of consumer products.
Date of Award6 Dec 2022
Original languageEnglish
Awarding Institution
  • University of Bristol
SponsorsEngineering and Physical Sciences Research Council
SupervisorJason M Yon (Supervisor) & Ben J Hicks (Supervisor)

Cite this

'