Abstract
Curtain walls are lightweight, weathertight, exterior facades. They are capable of resisting wind loads, but provide no support for the building it to which they are attached. Although they are used to enclose many different types of modern building, and although they may be designed to carry any of the outward-facing materials an architect might wish to specify, the stereotypical curtain wall is a skyscraper's fully-glazed outer skin.
The materials used in these wall systems, particularly their structural aluminium frames, are produced by energy intensive methods. Even though there is an environmental motive to reduce the embodied energy by minimizing aluminium content, and despite the obvious commercial incentive, it is a difficult mathematical challenge to find optimal extrusion shapes. The authors believe that because of the inherent complexity of the optimization task, in the curtain walls of real buildings, metal is used inefficiently.
This paper describes the way in which near-optimal shapes for any particular building's curtain wall extrusions may be found using a parametrically-controlled geometric model in conjunction with a numerical search routine – in this case, a genetic algorithm.
When the curtain walls designed for large and recently-constructed buildings by experienced facade engineers are compared with designs developed using the algorithmic techniques described herein, it is consistently the numerically-optimized solutions which are more efficient. The magnitude of the metal savings achieved by applying computational methods will vary from building to building, but this study suggests that in many cases aluminium mass may be reduced by 20% or more.
The materials used in these wall systems, particularly their structural aluminium frames, are produced by energy intensive methods. Even though there is an environmental motive to reduce the embodied energy by minimizing aluminium content, and despite the obvious commercial incentive, it is a difficult mathematical challenge to find optimal extrusion shapes. The authors believe that because of the inherent complexity of the optimization task, in the curtain walls of real buildings, metal is used inefficiently.
This paper describes the way in which near-optimal shapes for any particular building's curtain wall extrusions may be found using a parametrically-controlled geometric model in conjunction with a numerical search routine – in this case, a genetic algorithm.
When the curtain walls designed for large and recently-constructed buildings by experienced facade engineers are compared with designs developed using the algorithmic techniques described herein, it is consistently the numerically-optimized solutions which are more efficient. The magnitude of the metal savings achieved by applying computational methods will vary from building to building, but this study suggests that in many cases aluminium mass may be reduced by 20% or more.
Original language | English |
---|---|
Pages (from-to) | 147-156 |
Number of pages | 10 |
Journal | Structures |
Volume | 10 |
Early online date | 18 Mar 2017 |
DOIs | |
Publication status | Published - 31 May 2017 |
Keywords
- facade design
- genetic algorithm
- embodied energy
- green building
- structural optimization
ASJC Scopus subject areas
- Engineering(all)
Fingerprint
Dive into the research topics of 'Optimizing the Cross-sectional Shapes of Extruded Aluminium Structural Members for Unitized Curtain Wall Facades'. Together they form a unique fingerprint.Profiles
-
Mark Evernden
- Department of Architecture & Civil Engineering - Senior Lecturer
- Centre for Regenerative Design & Engineering for a Net Positive World (RENEW)
Person: Research & Teaching, Core staff
-
Paul Shepherd
- Department of Architecture & Civil Engineering - Reader
- Made Smarter Innovation: Centre for People-Led Digitalisation
- Centre for Digital, Manufacturing & Design (dMaDe)
Person: Research & Teaching, Core staff