Fadoul, Abdelaziz
(2020)
A BIM-based model for constructability assessment of buildings design.
PhD thesis, University of Nottingham.
Abstract
Implementation of constructability principles in the construction industry has a potential return on investment concerning time and money. Existing empirical studies demonstrate that incorporating these principles into the initial stages of design maximises outcomes for all stakeholders, including designers, contractors, and clients. However, constructability encounters many challenges in practical implementation. One of the main obstacles is knowledge acquisition and representation, leading to the lack of a knowledge-based tool to model design constructability. Current methods demand laborious efforts and resources to execute assessment calculations and interpret their outcomes. The dynamic design process and the need for ongoing modifications in designed products necessitate revision of the constructability assessment routine, whenever design changes are introduced, and it is highly desirable that these revisions should be automated.
This research, therefore, investigates design-stage assessment of design constructability by examining contemporary process and object-oriented models. The study reviews currently employed approaches for assessing design constructability and highlights their shortcomings. Based on this, it presents an original assessment framework to measure constructability of BIM-based design solutions. The proposed mechanism separates the formulation of construction knowledge from carrying out the assessment processes. The modelling framework is composed of three key parts: the Constructability Model (CM), which formulates user-based knowledge; the BIM Design Model, which provides required data for the assessment; and the Assessment Model (AM), which reasons the formulated knowledge into design features.
The model was implemented in a prototype, using object-oriented programming in a C# application. The prototype was developed using .NET Framework as a plug-in to BIM software, Revit, to operate on the design models created. The prototype was tested using typical design case studies, which have proved its usefulness in informing constructability decision-making. The process also enabled the exploration and evaluation of what-if scenarios in design iterations, and construction methods.
A developed BIM-based constructability assessment model was validated through different approaches, including interviews with experienced practitioners and a focus group comprising experts from industry and academia. As a result, the model has been found to provide the capability to represent constructability assessment knowledge within its Constructability Model. In addition, it demonstrated the ability to employ the knowledge-bases produced to reason about the constructability of alternative designs. Furthermore, practitioners have confirmed that the model is highly applicable in the industry and greatly needed to improve the practice of designing for constructability.
The research concludes that the introduced assessment framework effectively enables modelling of buildings’ design constructability. The implemented prototype is found to provide qualities lacking in current constructability tools. These include the qualities of being generic, scalable, flexible, comprehensive (both quantitatively and qualitatively), simple to use, accurate, and effective in delivering meaningful results that enable constructability improvement.
In addition, the separation between knowledge acquisition and knowledge reasoning processes simplifies the assessment procedure and saves the user time and effort. It allows for the reuse of formulated knowledge (i.e., CM) to model the constructability of multiple designs and at different stages. It also eliminates any potential bias that could arise during constructability assessment, given the subjectivity of the problem. Furthermore, the use of the BIM-based assessment tool automates the process and delivers an instant feedback on constructability performance.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)