You may have heard about ‘parametric design’, ‘computational design’ and ‘generative design’, especially in connection with BIM or prefab projects. According to some leaders in architecture, they are the way of the future. So, what are they, how do they work, and how can they improve the construction process?
Curves, spirals and convolutions make for dramatic design, but when a builder first sees them on the project plans, they can cause some consternation. Parametric modelling and other computational approaches can help reduce risks associated with such projects while delivering a major boost to efficiency, productivity and cost management.
Parametric Monkey is a digital design consultancy that offers specialist BIM, computational design, and digital fabrication expertise. Founder Paul Wintour has a background in architecture and has worked for leading practices in Australia and overseas.
The consultancy works with project owners, architects, contractors, developers, engineers and local governments to improve the way buildings are designed, built and operated through the use of digital design tools.
Many Benefits of Computational Design
Wintour believes the whole industry—not just architects—could be benefiting from parametric and computational approaches.
“Manufacturing and construction contractors are still doing things the way they did them 30 or 40 years ago,” Wintour says.
However, he added, digital tools can make everyone in the industry more productive. In fact, they can even change the dynamic between clients, architects, builders and subcontractors for the better.
Parametric Monkey is currently developing ‘MetricMonkey’, software that is focused on project feasibility and enables architects, urban designers, and property developers to evaluate designs in real-time via an intuitive user interface. It is due to be released in early 2021.
Digital Terms You Need to Know
Wintour has published an essential guide to the terminology involved in the new digital toolkit. While often many terms are used interchangeably, there are some important distinctions.
Computational Design is an over-arching term which defines an algorithmic design process. So instead of manually defining a specific element, a set of rules defines a system.
Generative Design starts with the design goals—what the building, system or structure must achieve in terms of performance—and then the software iterates through possible solutions to find the best option.
Parametric Modelling uses rules or parameters defined by the design team to generate a model. As elements are associative, they are capable of quickly and automatically adapting to changes. So if a rule changes, a parametric model will automatically adjust every element where the rule applies.
Design for Manufacture and Assembly (DfMA)is a specific digital design and delivery approach where an element is designed to facilitate manufacturing of parts off-site in a manner which can be efficiently and safely assembled on-site.
Huge Time-Savings for Construction
An example of parametric modelling and DfMA in action can be seen in UTS Central in Sydney, a 17-storey glass-encased tower with operable sun shading delivered for Richard Crookes Construction by TILT Industrial Design. Parametric Monkey provided the parametric design services for the project, with digital fabrication utilised for the welded aluminium sunshade components.
According to Wintour, the parametric-enabled approach delivered “huge time savings.” Instead of the modelling and shop drawing process taking 12 weeks, it took just three.
He points out that for a builder, saving just a week on a program reduces wasted time and money, including site costs from crane hire to site amenities.
On UTS Central, parametric modelling and digital fabrication generated substantial man hour cost savings. Actually, Wintour believes, the approach was “critical to the success of the project” in terms of its tight timeframes.
The modelling can be undertaken by just one person and creates a model that can be sent “from file to factory.” The computer produces the shop drawings—automatically unfolding each designed element, reporting the exact dimensions of each part, and where folds are required. This file then goes directly to the CNC machines in the factory.
There’s another serious advantage to parametric modelling: Once the parameters are established, if there is a change, the software will automatically populate the model with the changes. By contrast, a BIM model will still require an operator to manually adjust the model, which takes time.
Wintour has seen how this nimble and rapid adjustment capability can pay off.
On the UTS Central project, when the design team went to the site, they discovered the tolerances and setout were different from those in the design model. In a business as usual scenario, this would result in significant time being used to alter the model.
However, the parametric software can quickly adjust the model to suit site conditions while retaining all the rules that define the design. This allows the design and construction teams to switch between the two setout conditions as necessary.
Parametric Modelling Helps Reduce Rework
Wintour says change is common in construction projects—clients may want changes, construction systems may change, products may be unavailable and substitutions needed, or tolerances may vary.
These situations can lead to conflict and stress, and the time and cost associated with variations affect the project. Changes can generate a significant amount of rework, another cause of margin erosion and program problems.
The rapid adjustments offered by parametric modelling can help avoid these common issues.
It “smooths the communication path,” Wintour said.
This is particularly true of communication between the architect, client and builder. With conventional approaches, the architect’s exact design is constructed “very rarely.” Along the way, elements get changed by contractors and subcontractors because of buildability, available materials, site constraints or development consent conditions.
With a parametric model, there is “no ambiguity” and “no need for interpretation of design intent.” Everything is right there in a digital model which was defined through rules including constructability.