Designing-in performance: Energy simulation feedback for early stage design decision making

With the continued advancement of computational tools for building design, performance has gradually been allowed to claim a more prominent role as a driving force behind design decisions. However, there is currently only limited direct energy performance feedback available for designers early in the design process where such decision making has the highest potential impact on the overall design's energy performance. Therefore, this research aims to propose a design process framework that can provide designers a "designing-in performance" environment, where design decisions can be influenced by energy performance feedback during the early stages of the design process.;In response to the overall aim of this research, the first objective is to identify the most potentially suitable method through investigating current and past efforts. Extensive literature review revealed that time constraints and interoperability issues between tools and expert domain knowledge are the primary obstacles faced by designers in exploring design alternatives with consideration for energy performance. Moreover, evidence suggests that Multidisciplinary Design Optimization (MDO) methodology presents the most potential to overcome these obstacles. This determination stems from the aerospace and automobile industries successfully integrating multiple engineering domains through MDO to optimize and identify the best-fit design among various competing objectives during the design process. As a result, it is the position of this research that providing the designers with a designer-oriented MDO framework during the early stages of design will allow energy performance feedback to influence their design decision-making based on their design goals, thereby resulting in higher-performing designs. However, the applications of MDO to the building industry are still in infancy, especially in relation to bridging energy performance and design form exploration during the early stages of the design process. As the applicability of this approach during the design process is yet to be fully explored, this task is the second objective of this research.;More specifically, the second research objective is to identify the proposed framework characteristics that would assist the designers during the early stage design process and enable a "designing-in performance" environment. Also included in the second objective is the validation of the proposed framework against the identified criteria. In order to achieve this objective, this research first synthesizes the pertinent research findings in order to isolate the criteria for "designing-in performance" and identify the gaps in the extant approaches, which hindered their applicability to the design process. Based on these results, the research presents the theoretical structure of the proposed early stage designer-centered MDO framework, entitled Evolutionary Energy Performance Feedback for Design (EEPFD), which incorporates conceptual energy analysis and design exploration of complex geometry through an automated evolutionary searching method. EEPFD is a semi-automated design exploration process, enabled by a customized genetic algorithm (GA)-based multi-objective optimization (MOO) approach, to provide energy performance feedback in assisting design decision-making. In order to realize EEPFD for the purpose of validation and evaluation against the previously identified criteria, a prototype tool, H.D.S. Beagle, is developed to host the customized GA-based MOO algorithm. In H.D.S. Beagle, energy use intensity (EUI) is selected as the energy objective function. Also included are spatial programming compliance (SPC) and a schematic net present value (NPV) calculation for consideration in performance tradeoff studies. A series of hypothetical cases are used to form the initial framework, as well as obtain and evaluate the technology affordance of H.D.S. Beagle. These hypothetical cases are also used as a means to assess whether EEPFD demonstrates the potential to meet the needs of early stage design, where rapid design exploration and accommodation of varying degrees of geometric complexity are needed. Based on these results, EEPFD can be considered as suitable for further exploration in early stage design applications. Finally, the hypothetical cases are used to reaffirm the need of incorporating energy performance feedback during the early stages of the design process.;The last research objective is to evaluate the impact of the proposed framework and availability of energy performance feedback on the early stage design process. To achieve this objective, evaluation metrics are first established to provide the means and measurements by which to conduct process evaluation and comparative studies in both the design profession and pedagogical case-based experiments. In the design profession case-based experiment, EEPFD is applied to a design competition open to professional design firms. In this case study, the chosen design firm utilizes three approaches in pursuing higher performance design: (1) collaboration with mechanical electrical and plumbing (MEP) consultants; (2) in-house analysis through available analytical tools; and (3) EEPFD application. While this case revealed that the output of these three approaches is not directly comparable, it is observed that EEPFD provides exploration of a greater number of design alternatives and tradeoff studies compared to the two conventional processes. The pedagogical case-based experiments conducted as a part of this study are divided into three sets, based on the setting---a computational classroom, a design studio, and a computational workshop setting. These experiments utilize the established benchmark process to (1) compare the human decision process against the automated EEPFD; (2) observe the ability of students to translate their design intent into a parametric model; and (3) gauge the impact of the availability of energy performance feedback on the early stage design process. The results obtained in these experiments indicate that EEPFD is capable of generating a solution space with a higher Pareto designated solution rate than the students can achieve. Moreover, it eliminates up to 50% of the human error rate, as observed in the manual exploration process. It is also observed that students are able to use the EEPFD-provided feedback to identify a design alternative that is not only consistent with their design exploration intent, but also improves upon the energy performance of their initial design. However, in all these case-based experiments, designers---both students and design professionals---encountered difficulties in translating their design intent into a parametric model compatible with the use of EEPFD. While this is acknowledged, it is likely that increased familiarity with parametric modeling techniques would overcome these observed difficulties.;The significant contribution of this research is the EEPFD---a designer-oriented MDO framework enabling the combination of complex geometric form exploration via energy performance feedback. This tool addresses the observed gaps in the currently available solutions, thereby enabling further exploration of MDO application to the early stage design. Subsequently, this research provides the means and measurements to explore and evaluate the application of EEPFD to the early stages of design, identifying potentially advantageous adjustments from previously implemented MDO frameworks and the early stage design process. Considering the nature of early stage design in the architectural field---which consists of subjective and objective design requirements, time constraints, uncertainty regarding design components, and the unique conditions of each design problem---a best practice is proposed in applying EEPFD to early stage design, which significantly differs from the more common MDO application method of seeking a mathematically-defined convergence "best fit" solution set. Finally, this research provides a "designing-in performance" environment in which designers can use available energy performance feedback to influence their design decisions during the early stages of the design process, where such decisions have the greatest impact on the overall design's energy performance.