by Karl D.D. Willis, 2013
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: The miniaturization of projection technology has enabled a new class of lightweight mobile devices with embedded projectors. Projection engines as small as a postage stamp are currently being embedded in thousands of mobile devices. Mobile projector-based devices differ in very fundamental ways from the display-based devices we commonly use. Mobile projectors can be carried with the user and project imagery into almost any space, projected content is visible to multiple users and supports social interaction, physical objects and surfaces can be augmented with projected content, and embedded projectors can enable new form-factors for mobile displays. 

This research investigates the potential of mobile projectors as a new platform for human-computer interaction. I aim to demonstrate that the unique affordances created by the miniaturization of projection technology can inspire new and compelling interaction with single-users, multi-users, the environment, and projector-embedded objects. This research presents a comprehensive survey of mobile projector-based interaction – documenting interaction with historic projection devices; introducing novel interaction techniques, metaphors, and principles for mobile projector-based systems; providing implementation details of functional prototype devices using mobile projectors; presenting technical innovations, such as the development of specialized projectors and custom marker tracking algorithms; and detailing results from preliminary user testing with the prototype systems created. This research forms a systematic investigation of the past, the present, and a possible future for interaction using mobile projectors.

by Mustapha A. Bello, 2012
PhD-Architecture-Engineering and Construction Management

Download Full Text [PDF]

ABSTRACT: The construction industry is notorious for having one of the worst safety records among all industries in the private sector (Bentil, 1990; and Behm, 2005). In the 
United States, the industry accounts for up to 18% of work-related deaths and
15% of all worker compensation cases with approximately 1,000 construction
workers killed annually (BLS, 2000-2009).
Towards minimizing safety hazards and incidents, construction companies
employ several strategies including safety planning, staffing and training among
many others (CII, 2003). Different strategies apply to different project phases.
However, as the early identification and elimination of potential safety hazards is
not only more effective but cheaper (Behm 2005; and Anumba, 1999), those
strategies applicable to the earlier project phases are likely to have a more
significant impact in improving construction worker safety. One of such
strategies, Design for Construction Safety (DFCS), has the ability to function
effectively in the current Architecture/Engineering/Construction (AEC) industry
environment without requiring any major changes in procedure or contractual
structure. DFCS is the explicit consideration of construction worker safety in the
design of a project (Toole and Gambatese, 2008). Besides the ultimate benefit of
decreasing site safety hazards, DFCS, through the proactive identification and
elimination of hazards is safer and more cost effective than reactive management
of the same hazards (Toole and Gambatese, 2008).
The most critical impediments to DFCS include designers' concern about
increased liability, increased cost, and designers' lack of safety expertise. Others
include concerns about schedule problems, diminished design creativity, and
designers’ lack of interest (Gambatese et al, 2005). To assist designers in DFCS
implementation, safety researchers sponsored by the Construction Industry
Institute (CII) developed over 400 design suggestions to minimize or eliminate
certain construction safety hazards (Gambatese et al, 1997). These suggestions
were incorporated in a computer program, the DFCS Toolbox. Besides this, other
research has been conducted and guidelines developed to aid DFCS
implementation.
However, as DFCS is still experiencing limited application (Toole and
Gambatese, 2008), this research presented a different paradigm. This paradigm
considered that the guidelines and tools provided to enable and aid DFCS
implementation were incomplete, inaccurate and/or inadequate to serve their
intended purpose. Through this research, some of the available guidelines and
tools were fine-tuned and detailed to better enable DFCS implementation.
Hence, the research produced certain deliverables.

Firstly, the research identified DFCS measures that meet all the criteria for being situated in the capital project design phase. Secondly, the research identified impediments to implementing each of these design-phase DFCS measures where applicable. Thirdly, the research obtained revisions of certain designphase DFCS measures based on their identified impediments to make them more viable, both for implementation and for improving construction safety.

Additionally, the safety benefits of implementing each of the design-phase DFCS measures were identified through the publicly accessible Occupational Safety and Health Administration (OSHA) database. These benefits refer to the construction hazard incidents that could have been prevented by implementing the DFCS measures. Lastly, a relational database application was developed to assist designers in making safety a consideration in the early phases of the capital project delivery process. This desktop software application was developed to have the functionality to provide the design-phase DFCS measures, their preventable safety incidents, their potential impediments, potential solutions to their impediments, and their tier of feasibility, based on project characteristics, design profession, and the stage of the design phase. The application also allows for the addition of new DFCS measures and accompanying data. It therefore incorporates the other research deliverables and thus, encapsulates the research findings to serve as a vehicle for utilizing the data to enhance DFCS implementation. In producing and validating these deliverables, a number of research tasks were executed including survey administration to AEC design professionals. Also, over 30 interviews were conducted with design professionals.

Besides the deliverables, there were a number of findings from the research results. Firstly, the results emphasized a key shortcoming of the DFCS concept. This is the effectiveness of DFCS depends on construction sequence. Secondly, it was determined that DFCS measures or modifications that not only improve construction worker safety but occupant and maintenance worker safety are more likely to be implemented by AEC design professionals and more likely to be accommodated by project owners as well. On this basis, a new dimension was identified towards increasing and improving DFCS implementation. Thirdly, this research further emphasized that the design-build project delivery method offers more opportunity and fewer barriers for DFCS implementation.

This research made a number of contributions. Firstly, the research characterized the design suggestions for construction worker safety yielded from earlier research. This research also brought focus to individual DFCS measures and their feasibility for implementation, as opposed to for the DFCS concept as a whole. Secondly, this research, through its deliverables, serves in fulfilling several earlier recommendations for DFCS research and some earlier identified information gaps. These research contributions are collectively intended to enhance and increase DFCS implementation on projects towards improving construction safety. There are a number of motivating factors for this. Firstly, professional, ethical and moral obligations require the safety of others to be protected. Secondly, the improvement of safety could potentially benefit every project stakeholder and participant by minimizing or eliminating the numerous costs associated with injuries to construction workers. Thirdly, all project participants may also benefit in that reducing the number of construction accidents and injuries could avoid disruption to work and avert delays in project completion and as a result, improve productivity (Huang, 2003). Additionally, poor safety performance and its resulting consequences such as court cases and lawsuits expose all project participants to bad publicity which could have such adverse impacts as preventing job awards or causing even more lawsuits from prior projects (Huang, 2003). These reasons collectively highlight the importance of improving construction worker safety and towards this goal, this research emphasized and enhanced DFCS as a strategy for reducing or eliminating construction hazard risks on capital projects.

by Yang Hu, 2012
PhD-Building, Performance and Diagnostics

Download Full Text [PDF]

ABSTRACT: Solar driven absorption chiller technology as an alternative mechanism for cooling has been the focus of tremendous recent interest due to its potential advantages for energy conservation and the environment benefits. However, the comparatively lower coefficient of performance (COP) of these systems compared to the COP of the electric chillers has prevented more widespread applications.

In this dissertation, a novel comparison between an electric and an absorption chiller will be presented (Chapter 3), including a method for calculating the true solar energy requirement for fossil fuel (coal) formation. The same comparison is then extended to the exergy domain. Compared to an electric chiller, in order to provide the same amount of cooling, a solar driven absorption chiller actually consumes a much smaller amount of both solar energy and exergy.

Beyond demonstrating this lower level of actual energy/exergy consumption, it is still important to increase the efficiency of the solar chiller system. Therefore, a detailed exergy destruction analysis (Chapter 4 and 5) is performed in this dissertation, both for the solar collector, which includes pumping exergy loss, and for a two-stage lithium bromide (LiBr)-water absorption chiller. These analyses allow for a better understanding of the exergy destruction due to a component’s own inefficiency and/or due to the remaining components’ inefficiencies. Given the limits of current technology, the results show that the COP of a two-stage LiBr-water absorption chiller can be improved from 1.1 to 1.38, a 25% efficiency increase, by recovering the avoidable exergy destruction.

Finally, based on the contributions described above and the conclusions that can be drawn from them, a range of potential future work is presented. Other comparative systems are discussed, and the challenges in properly assessing the performance of those systems are described.

by Tsung-Hsien Wang, 2012
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: A growing trend in contemporary architectural practice, pioneered by such avant-garde architects as Frank Gehry, Zaha Hadid and others, exploits NURBS (Non-Uniform Rational Basis Spline) surfaces to design and model intricate geometries for projects which otherwise would be impossible to realize. In doing so, they have liberally borrowed digital fabrication techniques developed in the automobile and aerospace industries (Kolevaric 2005a, 2008; Pottmann 2008). A NURBS surface is a mathematical model for freeform shapes. To manifest a NURBS surface, a discrete model, namely, mesh, is utilized. Transforming a NURBS surface into a mesh appropriate for application is computationally intensive, and generally, it is not an easy task for architects or designers who have no formal geometry training.

In order to design, model, and, subsequently, fabricate intriguing, sometimes intricate, freeform shapes, this research looks at the surface tessellation problem, which is an extension of the problem of meshing a NURBS surface, with an added consideration of incorporating constructible building components. There are close relationships and analogies between the elements of a mesh and the components of a freeform design, e.g., face to panel, edge to structural frame, etc.

Initially, features of a NURBS surface and contemporary tessellation methods are examined. Mathematically, a NURBS surface is regulated by a set of control points and edges. The control points are used mainly to interpolate a continuous shape using a higher order equation, in most cases, usually cubic. The edges delineate the appearance of the freeform shape. For a surface, edges (also called boundaries) indicate where the surface analysis starts and where it ends, and thus, plays a significant role in the meshing process.

Two kinds of boundaries are examined in this research. The first are global boundaries, which form the overall appearance, e.g. exterior edges, or interior trimming edges. The second kind is a local boundary, which specifies how a discrete element is formed—namely, the pattern of a face, e.g. triangle or quadrilateral. By looking at given surface boundary conditions and tessellation patterns, this research presents an algorithmic approach to pattern-based surface tessellation and develops strategies to resolve issues that stem from the juxtaposition of computational geometry and freeform architectural design.

The contributions includes the technical implementation of boundary-driven mesh generation, which demonstrates the potential of utilizing featured boundaries for customizable polygon-based tessellation in comparison to conventional iso-parametric subdivision. This is described through examples by extending the optimized mesh network for various pattern generations. In addition, pedagogical implications are exemplified by solving the geometric constraints for surface tessellation within the parametric modeling paradigm. These contributions are expected to support future sustainable development in the field of freeform architectural design.

Keywords: Pattern-based surface tessellations, irregular boundary conditions, meshing

by Yuebin Yu, 2012
PhD-Building, Performance and Diagnostics

Download Full Text [PDF]

Building systems have unique features that make multivariate model-based predictive control (MPC) technique as a promising option in the future. Thermal comfort and energy conservation are the two basic reasons for having a building automation system in operation in buildings. However, neither of the two objectives can be explicitly included in conventional single-input-single-output (SISO) control laws. In addition, a building system is normally a nonlinear multiple-input-multiple-output (MIMO) system which could have coupling issues among the controlled objects. Conventional control laws, such as an proportional-integral-derivative (PID) controller, are linear time-invariant SISObased. Finally, there are hard and soft constraints present in building systems that cannot be explicitly handled in PIDs.

A literature review carried out in this research reveals that there are plenty of studies of model-assisted control conducted in the building industry in the past decades. However, the majority focused on utilizing nonlinear models and applied a simple "simulator plus optimization solver" approach. The advantage of this approach is that it simplifies the construction efforts since a building system model can be encapsulated in simulation tools, and it does not require much understanding on the optimization solver. However, since the simulator becomes a black-box to the solver, this method is usually computationally costly and not suitable for dynamic optimization and online implementation. The approach and benefits of applying linear classical MPC in, which is much faster and has mature analyzing theory, buildings have not been not fully explored.

This thesis research concentrates on three enabling techniques for utilizing linear classical MPC in building systems. The MPC is designed for the implementation in a dual-loop predictive structure to explore the full potential. More specifically, in this thesis, multivariate control-oriented dynamic models for diversified building energy systems are developed based on first principles. This includes constructing simulators for the various components included in building systems. Second, linear MPC controllers for the mechanical components in the diversified systems are investigated. Third, fast nonlinear MPC controllers utilizing linear classical MPC as the core for the space thermal conditioning for both thermal comfort and energy conservation are designed. The performances of the controllers are evaluated in the Intelligent Workplace north (IWn) test bed that includes a hydronic system.

The MPCs for complex primary equipment, including an enthalpy recovery unit, a vapor compression unit, and an active desiccant unit, are designed and evaluated in a decentralized architecture. It is found that, by applying MPCs, the multiple variables are coordinated properly toward the objective with minimum efforts. In an intensive swing season test, the MPC on the enthalpy recovery unit successfully manages the outdoor air intake and enthalpy recovery wheel across the three different operation conditions: full outdoor air cooling without recovering the enthalpy, economizer, and minimum outdoor with enthalpy recovering. The MPC for the vapor compression unit modulates the compressor speed, the condenser fan, and the supply fan toward the steady state optimal values with minimum movement when the step disturbances on the set point and outdoor air temperature are introduced.

Wiener nonlinear MPCs and Hammerstein-Wiener nonlinear MPCs are studied for the space thermal conditioning. The approach remains the advantages of linear classical MPC and solves the nonlinearity issues involved in thermal comfort oriented control. With a linear time-invariant first-principle model for the building and linear MPC as the core, the controllers achieve the optimal solution for IWn in less than five minutes for five days simulation in a full look-ahead scenario. The fastest scenario takes only thirty more seconds to accomplish, which makes the approach practical for online implementation.

The techniques of using MPC for achieving smooth day-night switch, band control, dynamic constraints, and dynamic weighting are discussed. In addition, the energy saving potential of applying the proposed MPCs is found to be between six to thirty six percent compared to a 24/7 temperature oriented control and seven to forty two percent compared to a temperature oriented conventional control with office schedule. The techniques are applicable to other space conditioning systems other than a hydronic system.

In addition, this thesis also addresses the under-sensing and modeling issues in hydronic system control. A "node-branch-state" based approach is proposed to simplify modeling efforts and facilitate thermal-hydronic coupled simulation. Unlike conventional "nodeloop" approach, the modeling method can handle the discontinuity caused by pump on/off and valves close/open operation in a large hydronic network simulation. With the approach, an advanced control with the integration of thermal and hydronic network becomes possible. Preliminary case studies, including two for water distribution and one for coupled thermal and hydronic network, are presented.

In conclusion, the work and findings presented in this thesis make important contributions to the research of model predictive control of air-conditioning systems in office buildings. The results can be partially utilized or as a whole for other dynamic system oriented research. Future work, such as automated model generation, model parameter acquirement, is briefly discussed.

by Sora Key, 2012
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: A method of programming descriptive design models is presented and illustrated with examples. Design models are seen as independent objects that encapsulate descriptions of physical layouts. The models are constructed and represented using a set of constructible functions that indicate geometric attributes and the characteristics of physical configurations. 

The subject of design models in this work is experiential qualities of architecture which we call spatial qualities. This document provides in-depth description of the process of computing terms that indicate spatial qualities. The terms are abstract, conceptual design objectives from the early phase design process. The process of computation – the analysis of common terms and patterns, translation of the terms into computable functions, and reconstruction of the functions into machine executable design models (descriptor) – all builds a method of how we can computationally represent spatial qualities. 

A computable language supporting the design analysis is proposed and its implementation is briefly discussed. The language, being implemented in a fully-working prototype system called Descriptor, is used to express compound qualities implied in the subset of Christopher Alexander’s A Pattern Language. More examples are also provided in expressing common visual language terms using the language. 

Descriptor, a software environment for programming architectural qualities, is built to support design representation at the level of abstract concepts based on the measurable attributes of the physical built environments. The work brings a discussion about places for digital tools and methods by offering a systematic method for logical construct of design concepts while maintaining its explorative nature.

by Peng-Hui Maffee Wan, 2011
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Wayfinding is a kind of spatial riddle that people encounter almost daily. Although it has been well documented that wayfinding elements – namely, environmental cues, people and time –significantly influence wayfinding, there has been little work done to examine the effectiveness of those influences. In particular, the notion of wayfindingly manageable environment is investigated in the dissertation. A wayfinding task is termed as manageable, whenever a way-seeker successfully navigates the task. When a significant number of wayfinding tasks in an environment are determined to be wayfindingly manageable, the whole environment can be claimed as wayfindingly manageable. “How cues are arranged so that an environment becomes wayfindingly manageable?” is the main research problem considered in the dissertation. This is illustrated through a consecutive three-step development process of theory, visualization and simulation, which looks at the issue from both environment-centered and people-centered perspectives.

The first step, engages in the environment-centered perspective, in developing a measure for calculating manageability for wayfinding, expressed in terms of variables at the level of decision points, pair of decision points, routes, tasks and environment, through a paper-based experiment and statistical analyses. The second step is to develop a visualization/simulation approach to apply the criterion of wayfinding manageability considered from both environment and people-centered views. As a first pass, the measure of manageability developed in the first step is employed in the visualization to explore the issue at the environment-centered level. From the people-centered view, an agent-based modeling simulation is considered in which peopleagents are have settings for wayfinding strategies for different wayfinding tasks and associated with time-based benchmarks. Wayfinding-decision processes involving people-agents, namely, wayfinding strategies and time-based benchmarks are retrieved from a computer-based experiment and statistical analyses.

The last step in the development process is implementing an agent-based modeling application in which environmental cues are behavior-control agents, and way-seekers are rulebased people agents. Both one-way passive and two-way interactive environmental cues are captured in the application. Different wayfinding tasks for a single environment with different settings for sign-placement are used as proof of concept for demonstrating wayfinding manageability. The ultimate goal, of course, is for this agent-based modeling application to be a design tool for navigable environments through generative manipulation of wayfinding elements, mostly, environmental cues.

This dissertation, developed from studies in psychology, is an attempt to make an original contribution to our understanding of the design of wayfindingly navigable environments, in particular, in architecture and environmental graphic design. The methodology comprising experiments, analyses, visualization, and simulation employed in this dissertation is expected to be applicable to other behavior related studies in the fields of design, architecture and psychology.

by Yi Chun Huang, 2011
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Lighting simulation contribute readily to the synthesis of high performance lighting designs. Unfortunately there exist several issues impeding the pervasive use of lighting simulation, including:

1. Most of the time in preparing lighting simulations is spent towards the input of existing but non-interoperable information between different tools.
2. Lighting simulation tools do not complement integrated building design processes where the design solution is progressively developed in multiple disciplines concurrently; lighting simulation tools require design information (attributes) that may not yet be defined, and is non-interoperable with other tools.
3. Disparate tools with vastly different technical approaches available for different stages of the building design process do not allow consistent or meaningful performance comparisons between design versions, and similarly makes design performance progress tracking between design versions difficult.
4. Lighting simulation tools provide radiance and irradiance values as simulation results, and much time and manual effort is required to process these results into operative information, information that is directly applicable in making design decisions.
5. Lighting simulation tools employ outdated rendering techniques that are inadequate in evaluating highly-reflected irradiance, a typical feature in high performance building designs.

While there remain other shortcomings in lighting simulation tools as identified by contemporary research, the issues above relate closely to the overall effort and time-cost factors attributed to using simulation tools, which has been consistently identified as obstacles
towards using simulation tools. This research seeks to reduce the effort and time-cost required to conduct lighting simulation by addressing the issues above. Case studies of actual design scenarios are used to establish quantitatively the effort and time costs baselines for comparison.

The effort and time reduction goal is structured as the following objectives in a new lighting design support tool:

1. Reduce the time and effort to set up and conduct lighting simulation by using interoperable information (building information models) from design modeling tools.
2. Complement integrated design processes by supporting design models of varying completeness, in a format that is interoperable with tools from other disciplines in the design team. All information, including assumptions, must be consistent across all disciplines.
3. Provide ability to use performance metrics and consistent technical approaches throughout design stages, regardless of completeness of design model.
4. Provide operative information with minimum user effort.
5. Implement a first principle-based rendering technique that handles high performance building designs well, and produce simulation results within reasonable time constraints.

By meeting these objectives, the new lighting design tool is able to automate much of the previously manual, time-consuming, and disparate efforts in lighting simulation, thus reducing the effort and time-cost. By sharing interoperable information with other tools across the design team, the new lighting design tool is integrated. The new tool is also scalable in being able to support models of varying completeness throughout all design stages.

by Rui Zhang, 2011
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

Energy modeling approaches have continued to advance to cater for emerging new design concepts towards “greener” solutions. Notwithstanding the advances in system modeling, there are still limitations in the capability of representing the “real’” environmental behavior with different building spatial configurations. For example, most current energy models adopt a “nodal” approach to simulate the heat transfer process in the building, with simplified heat resistors and capacitors network. Finite Volume Method (FVM) such as Computational Fluid Dynamics (CFD) on the contrary will perform detailed computation on the temperature profiles and air flow fields, which could supplement the nodal model for the building energy simulation with heat transfer coefficients airflow rates and etc. With the rapid development of computation power, the computation time of CFD is decreasing, it is highly necessary to develop an integrated thermal simulation model that combines the advantages of both the nodal and the FVM model, which will improve simulation accuracy with acceptable computation time.

Mesh generation is a critical and probably the most manually intensive step in CFD simulations in the architectural domain. Mesh generation for CFD simulation in buildings poses special challenges. Firstly, the span of the dimensional scales encountered in design is large. Secondly, the geometry model of a building usually involves non-manifold surfaces. Thirdly, the lack of data interoperability between CFD simulation tools and major architectural Computer Aided Design tools makes the application of CFD in architectural design an expensive practice.

Research by NIST (National Institute of Standards and Technology) estimates the cost of inadequate interoperability in the U.S. capital facilities industry to be $15.8 billion per year. At the same time, Building Information Model (BIM) is being widely accepted by the AECO (Architecture, Engineering, Construction and Operation) industry as a data sharing and management platform supporting interoperability in order to increase productivity in building design and constructions.

In addressing the challenges, this dissertation work implemented a coupled simulation platform between a nodal model and a CFD tool with advanced mesh generation. Eight days of natural ventilation simulation for the live retrofit project of building 661 at the Philadelphia Navy Yard was conducted. Significant differences have been identified in heat transfer coefficients and airflow rates between the coupled simulation and the nodal model. The differences of heat transfer coefficients between the nodal model and the coupled model range from 7% to 188% for exterior surfaces and range from 10% to 80% for interior surfaces. The differences of airflow rates between the two models are from 100% to 770%. An interpolation model was developed based on the simulation period of June 1st to June 8th. The weather analysis shows that there are 854 hours annually, which are suitable for natural ventilation in Philadelphia. It is found that with the 8 days’ simulation results, the interpolation model is able to predict 266 (35% of 854) hours of airflow rates annually, with excepted error of 24%.

The mesh generation algorithm development has two stages. First, a prototype automatic two dimensional mesh generation tool is implemented to generate adaptive quadrilateral meshes from architecture drawings for CFD simulations. Starting from two dimensional image data, adaptive quadrilateral meshes are constructed automatically. The nearly orthogonal boundary layer is generated and the thickness is controlled to facilitate various simulation scenarios. Second, a three dimensional automatic mesh generation tool was developed to generate the adaptive hexahedraldominant mesh and the uniform all-hexahdral mesh from architecture conceptual design tools for CFD simulation in the architectural domain. Simulation experiments show that the adaptive 3D mesh reduces the number of elements significantly (> 90%) while maintaining the accuracy compared with uniform mesh. The three dimensional meshing tool will take as input the VRML model generated by Google SketchUp. The mesh generated shows good quality in terms of shape parameters, the scaled Jacobian, and the condition number of the Jacobian matrix. A prototype BIM platform that supports interoperability between the energy simulation tool and the CFD tool was developed. The manual time required to create the

EnergyPlus model and the Fluent Model on the prototype BIM platform is reduced by 87 _ 90% compared to that required on the non-BIM supported approach for similar models. Based on the developed prototype platform, key information requirement for the coupled simulation was identified. Dynamic information, which are the properties of an object that are dependent on changing geometry configurations, system operations, or boundary conditions, was proposed as Domain Object Model. The proposed dynamic information objects are heat transfer coefficients for building envelop components and airflow rates for openings.

by Yun Gu, 2011
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Current building design and engineering practices emphasizing on energy conservation can be improved further by developing methods focusing on building occupants’ needs and interests in conservation. Specifically, the resulting energy effective building performance improvements cannot reach the desired goals, if the resulting indoor environmental conditions do not meet thermal, visual and air quality needs of the occupants. To meet both energy conservation and human performance requirements simultaneously requires to give the occupants information regarding indoor environmental qualities and energy implications of possible individual decisions. This requires that building control components and systems must enable occupants to understand how the building operates and how their own actions meet both their needs and the energy and environmental goals of the building project.

The goal of the research and experiments of this dissertation is to explore if real-time information regarding visual comfort requirements to meet a variety of tasks and to simultaneously conserve energy, improves occupant behavior to meet both objectives. Two workplaces in Robert L. Preger Intelligent Workplace were equipped to test the performance of 60 invited participants in conducting computer based tasks and a paper based task, under three difference lighting controls:

1. Centralized lighting control with no user choice
2. User control of
   
   • blind positions for daylight shading
   • ceiling based lighting fixture luminance output level
   • task lighting: on/off
3. User control the three components (as listed under point 2 above), with provided simultaneous information regarding energy and related CO₂ emissions implications, appropriate light levels meeting tasks requirements, and best choices in order to meet both task requirements and energy conservation goals/objectives.

The main findings of the experiments are that real-time information (listed under point 3 above) enables users to meet the visual quality requirements for both computer tasks and the paper task, and to conserve significant amounts of electricity for lighting. Furthermore, the 60 invited participants were asked to identify the importance of the four types of provided information tested in point 3 above. While individual users identified the importance of different information categories, the overall assessment were considered to be significant.

by Yeonjoo Oh, 2010
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Critiquing is a fundamental part of design education, yet we lack a clear and systematic understanding of how effective teachers make decisions about how to critique students. Although there is a considerable literature on design education, little has been written about design critiquing, specifically about critiquing strategies.

The dissertation outlines a theoretical framework of design critiquing practice developed through a literature survey. It then describes a computational model based on this framework, implemented in the Furniture Design Critic program, a kind of constraint-based tutor. The Furniture Design Critic provides a basis for describing and articulating critiquing strategies. The program first assesses the conditions of critiquing: how much a designer knows, his or her weaknesses and strengths, what critiquing methods have been effective for the designer, and the history of interaction between critic and designer. Based on this the Furniture Design Critic then selects a set of critiquing methods.

This program offers a computational model to describe design critiquing and to model inference about critiquing, and an environment for exploring and investigating alternative critiquing strategies. The dissertation contributes to an ongoing discussion of critiquing in design, design education, and intelligent tutoring systems.

by Bing Dong, 2010
PhD-Building Performance and Diagnostics

Download Full Text [PDF]

ABSTRACT: Current research studies show that building heating, cooling and ventilation energy consumption account for nearly 40% of the total building energy use in the U.S. The potential for saving energy through building control systems varies from 5% to 20% based on recent market surveys. In addition, building control affects environmental performances such as thermal, visual, air quality, etc., and occupancy such as working productivity and comfort. Building control has been proven to be important both in design and operation stages.

Building control design and operation need consistent and reliable static and dynamic information from multiple resources. Static information includes building geometry, construction and HVAC equipment. Dynamic information includes zone environmental performance, occupancy and outside weather information during operation. At the same time, model-based predicted control can help to optimize energy use while maintaining indoor set-point temperature when occupied. Unfortunately, several issues in the current approach of building control design and operation impede achieving this goal. These issues include: a) dynamic information data such as real-time on-site weather (e.g., temperature, wind speed and solar radiation) and occupancy (number of occupants and occupancy duration in the space) are not readily available; b) a comprehensive building energy model is not fully integrated into advanced control for accuracy and robustness; c) real-time implementation of indoor air temperature control are rare. This dissertation aims to investigate and solve these issues based on an integrated building control approach.

This dissertation introduces and illustrates a method for integrated building heating, cooling and ventilation control to reduce energy consumption and maintain indoor temperature set-point, based on the prediction of occupant behavior patterns and weather conditions. Advanced machine learning methods including Adaptive Gaussian Process, Hidden Markov Model, Episode Discovery and Semi-Markov Model are modified and implemented into this dissertation. A nonlinear Model Predictive Control (NMPC) is designed and implemented in real-time based on Dynamic Programming. The experiment test-bed is setup in the Solar Decathlon House (2005), with over 100 sensor points measuring indoor environmental parameters such as temperature, relative humidity, CO2, lighting, motion and acoustics, and power consumption for electrical plugs, HVAC and lighting. The outdoor environmental parameters, such as temperature, relative humidity, CO2, global horizontal solar radiation and wind speed, are measured by the on-site weather station. The designed controller is implemented through LabVIEW. The experiments are carried out for two continuous months in the heating season and for a week in cooling season. The results show that there is a 26% measured energy reduction in the heating season compared with the scheduled temperature set-points, and 17.8% energy reduction in the cooling season. Further simulation-based results show that with tighter building façade, the cooling energy reduction could reach 20%. Overall, the heating, cooling and ventilation energy reduction could reach nearly 50% based on this integrated control approach for the entire heating/cooling testing periods compared to the conventional scheduled temperature set-point.

by Joon Ho Choi, 2010
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Current existing thermal control systems are operated based on thermal comfort models generated by regression formulas averaging the thermal responses over data collected during extensive experiments involving panels of human subjects. These models may not be appropriate for an individual whose physiological characteristics happen to be located outside of the main stream from the experimental sample of occupants. By necessity, existing automatic control systems disregard individual characteristics such as health, age, gender, body mass, etc., which may affect physiological responses. Thereby these systems have serious limitations in ensuring individual thermal satisfaction.

While there have been many efforts to overcome the limitations of current technology and to improve individualized control, most of the attempts to make smart controllers for buildings have dealt primarily with optimizing mechanical building components to deliver uniform conditions, largely ignoring whether a generated thermal environment by building systems meet actual users’ comfort and satisfaction. Over-cooling and over-heating are common unnecessary results.

Thermal control innovations for building mechanical systems are critically needed to demonstrate that meeting the physiological needs of occupants can actually save energy and improve environmental quality while enhancing user satisfaction.

The thermoregulation of the human body has a biological mechanism, homeostasis, which enables it to maintain a stable and constant body temperature by changing physiological signals including skin temperatures and heart rate. These signal patterns have the potential to provide information about each individual’s current thermal sensation.

The goal of this research is to establish an adaptive thermal comfort control driven by ongoing human physiological responses or bio-signals. Confirming the optimum driver of skin temperature, and location of sensors, the bio-sensing adaptive control logic is developed to support the optimum control of HVAC terminal units. The bio-sensing controllers offer major opportunities for office, healthcare and residential buildings, especially where environmental quality and control can be linked to productivity and health, and where energy savings are critical. The CoBi bio-sensing adaptive HVAC systems control research would substantially improve occupant comfort, health, and well-being while advancing environmental sustainability with energy savings, at a small first cost for existing or new buildings.

by Viraj Srivastava, 2010
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: This thesis examines the development, demonstration, and evaluation of an advanced control system specifically designed for cooling, heating, and ventilation in office buildings. It focuses on providing thermal comfort in both open and closed offices through the use of state-of-the-art fan coil units with advanced instrumentation and control. The premise of the work is that environmental effectiveness and energy performance can be significantly enhanced by a control system that makes use of real-time data from distributed sensors deployed in the building. Specifically, the performance of control systems, i.e. maintenance of a setpoint and reduction of energy usage, can be improved by combining predictive (feed-forward) control operations with feedback control and techniques to incorporate data from sensors installed in the space and outdoors measuring parameters such as air temperature, radiant temperature, humidity, occupancy, and solar radiation. The data are used to estimate thermal loads in the space.

The control algorithm operates cooling, heating and ventilation functions with the potential for the control of windows and blinds. The control problem is to set the fan coil unit valve position, fan speed and ventilation damper position in individual office spaces with the objectives of providing a healthy and comfortable environment for individual occupants while minimize energy usage.

The control algorithm developed in this work is comprised of three major components – a model-based feed-forward component that generates a control signal based on measurements of disturbances that influence the thermal conditions of the space (solar radiation incident on the façade, outdoor air temperature, and the window surface temperature); a conventional feedback component that generates a control signal based on measured system variables (for example, space air temperature); an occupant model that assists in computing the space air temperature required for comfort based on sensor measurements of current conditions in the space (for example, air temperature, radiant temperature, air flow etc.) including input from the occupant. The feed-forward algorithm comprises two sub-components – the Building Space Model and an Equipment Model. These models are mathematical equations that characterize the building space and the equipment.

The control system is limited by factors such as the size of the equipment and the temperature of the chilled water. Complications arise from interactive systems and multiple sources of thermal energy, for example, fan coil units, the ventilation system and outdoor conditions.

This work has included the installation of fan coil units, sensing and control hardware and software in two functioning offices to demonstrate the proposed control algorithms for indoor thermal environmental control. The test space used in this work – the Robert L Preger Intelligent Workplace (IW), at Carnegie Mellon University – involves a large number of variables and hence a complex control task. Studies performed include a verification of the performance of the fan coil units, a study of the performance of the fan coil units in influencing the air temperature in the space and a verification of the performance of the control algorithms in maintaining comfort and reducing energy consumption – electric as well as chilled and heated water.

The results obtained include an assessment of the performance of the fan coil units, a study of advanced control strategies applied to building fan coil units, and the influence of specific variables – fan coil unit sizing, thermal disturbances in the space, number, distribution and variety of sensors on thermal control. Tests performed with controlled sources of thermal energy demonstrate that the use of the integrated feed-forward and feedback strategy provide a 40% improvement in the maintenance of an air temperature setpoint and a 12% reduction in energy consumption compared to conventional feedback control strategies. Measurements taken provide information on thermal transfer – the ability to maintain an air temperature setpoint in a space as well as the operation of setback temperatures. Measurements further indicate a need for avoiding thermal stratification in the space during the heating season and the need to provide sufficient air mixing in the occupied zone.

It is expected that the results obtained for a system of fan coil units in two offices will lead to future work in operating a group of offices as well as including the operation of the shading devices (blinds and reflectors). This work examined feed forward control using measurements of current disturbances; further benefits may be obtained by incorporating anticipatory control that will make use of predictions of future disturbances, such as weather forecasts, occupant schedules. To further increase energy effectiveness, it is necessary to integrate the control of the energy supply system and of the fan coil units (or of other controllable devices to deliver thermal
energy to the space).

by Yun-Shang Chiou, 2009
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Human behavior has long been understood as the key determinant of building energy consumption. Until recent years, however, occupancy parameters have often been represented at a generalized level in term of fraction of some assumed maximum occupancy over time. To further refine the energy modeling approach, a more rigorous theoretical framework for mapping human behavior to the prediction of energy consumption is necessary.

Recent efforts on integrating the human element into residential building energy simulation have focused on application of the Markov Chain Monte Carlo (MCMC) technique with national Time Use Survey (TUS) data to generate various types of schedules as inputs for building energy simulation. While this method is straightforward in implementation, a MCMC model requires far more data than a TUS database can provide to construct a high resolution transition matrix, and requires enormous computational resource to use the matrix to generate energy use schedules. These constraints limit the resolution of a TUS-data driven MCMC model, and thus the applicability of the MCMC simulated schedules in building energy simulation. In addition, differences between the driving factors of environment-dependent human time-use behaviors and that of the environment-independent MCMC process reduce the validity of results generated by applying MCMC technique on TUS data.

Guided by Integrative Household System Theory, this dissertation presents a novel TUS-data driven residential-load-schedule generating method that replaces the MCMC data sampling technique with Bootstrap sampling to address the shortcomings of current applications of national TUS data. In this method, American Time Use Survey (ATUS) data are used to represent the daily activities of residents. Family composition is used as a sampling criterion in the Bootstrap process. The causal relationship among occupants’ daily activities, use of space and the dwelling’s energy demand is defined using the framework of Integrative Household Theory. The appliance load and occupants’ heat gain are represented by multiple sets of sub-house schedules derived from Bootstrap sampling of ATUS data. These schedules represent the variation of energy loads for given family compositions. Finally, this load-schedule-generating method is integrated with the Energy Plus building-energy- simulation engine to create the Occupant-Driven Residential Building Energy (ODRBE) simulation model. This model calculates the heating, cooling and lighting loads of the dwelling using the Bootstrap process-generated appliance load and occupants’ heat gain schedules. The validity of this approach is supported by extensive ODRBE simulations as well as the comparison between simulated and utility metering data-generated energy load profiles.

The ODRBE simulation method developed in this dissertation is most salient to retrofitting existing houses. In existing houses, household demography is known and improvements to the building are made incrementally, allowing the construction of the ODRBE simulation model. This dissertation uses ODRBE simulations to study the impact of sub-house zoning on energy-efficiency residential building retrofit for 2- and 4-occupant family compositions in a heating-load-dominant climate. Computational results indicate that much higher heating-load reductions can be achieved with the introduction of multi-zone HVAC systems as compared with building envelope improvements alone.

According to the U.S. Energy Information Administration 2006 report and the U.S. 2000 Census, the residential housing sector, comprising 105 million housing units, is responsible for approximately 21% of total U.S energy consumption. Less than 2% of all residential units are constructed each current year. Thus, existing dwellings are a primary contributor to U.S. energy consumption. An energy simulation tool to assist the energy-efficiency improvement of these dwellings can have substantial societal impact. Moreover, among competing building energy simulation approaches, the method developed in this dissertation is the only one that incorporates in a holistic manner the human and physical dimensions of residential building energy consumption.

by Kui Yue, 2009
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: A shape grammar is a formalism that has been widely applied, in many different fields, to analyzing designs. Computer implementation of a shape grammar interpreter is vital to both research and application. However, implementing a shape grammar interpreter is hard, especially for parametric shapes defined by open terms.

This dissertation explores the problem of implementing a shape grammar interpreter, which arose in the context of the AutoPILOT project, in which we were seeking an algorithm to determine the interior layout of a building given an input of building features and a shape grammar describing the building style. A general approach was adopted based on the fact that when applied exhaustively, a shape grammar can generate, as a tree, the entire layout space of the building style. The approach essentially requires a parametric shape grammar interpreter that caters to a variety of building types.

As extensions to the fact that shape grammars can simulate any Turing machine, three corollaries are found that significantly impact the implementation of a shape grammar interpreter. They are the following: a shape grammar may not halt; the language space of a shape grammar may be exponentially large; and parsing of a configuration against a shape grammar is generally unsolvable. The problem of interpreting a general parametric shape grammar is thus in general intractable; even parametric subshape recognition of twodimensional, rectilinear shapes may require a high-degree polynomial time.

In reality, there are distinct but large classes of shape grammars, with differing underlying characteristics, for which interpreters are known to be computationally tractable. In this dissertation, I present a practical, ‘general’ paradigm for ensuring the tractability of designed shape grammars and implementing such shape grammars.

Even these tractable shape grammars may significantly differ from one another. A further way of classifying these tractable shape grammars, optimally in my belief, is by the types of data structure capable of carrying out their rule application. There are, of course, other factors that influence the computation of shape grammars; these include the description language and all adopted underlying manipulations. As a consequence, the paradigm is augmented so that every interpreter is supported by an application programming interface-wise (API-wise) framework, which comprises an underlying data structure, basic manipulation algorithms, and a description meta-language. The paradigm specifies an overall framework comprising a series of sub-frameworks. The overall framework is capable of ensuring the computation for a specified shape grammar interpreter. Shape grammars, which follow such a framework, are termed as computation-friendly.

The concept of the overall framework is detailed by examining three sub-frameworks. These include one over 2D rectangular shapes (Rectangular sub-framework), one over 2D polygonal shapes (Polygonal sub-framework), and one for shapes describable by a graph structure (Graph sub-framework). The issue of how to develop a computation-friendly shape grammar is explained and illustrated by using the Baltimore rowhouse grammar as the exemplar.

The rectangular sub-framework, which has direct application to the AutoPILOT project, is examined in detail. This includes an investigation of estimating an initial interior layout from the feature input by constraints solution, and the application of spatial constraints from an estimated initial layout to prune the layout tree and ‘fix’ the open terms of the intermediate configurations. The building feature input for the AutoPILOT project is typically difficult to obtain. The technical feasibility of automatically extracting building features from image data is examined by comparing two pipelines, an ideal pipeline and a realistic pipeline.

In summary, in this dissertation, I develop a general approach for determining building interior layouts from exterior features with the aid of shape grammars. Central to the general approach, issues of implementing a shape grammars interpreter are formally investigated by complexity analysis. Subsequently, a practical ‘general’ paradigm is developed and demonstrated by sub-framework examples. The paradigm facilitates the development of a practical shape grammar interpreter and is readily extensible to future development.

by Fred Betz, 2009
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Combined heat and power (CHP) systems are frequently used to reduce energy consumption in a facility due to the increased energy efficiency. System efficiencies range between 65% and 85%, whereas the average utility efficiency for electric power supply is 31% and for heating from a natural gas supply is around 80%, which yields a combined efficiency of approximately 50% for all energy supplied in the United States of America. Buildings use 70% of all electricity generated in the U.S., 40% of all U.S. primary energy to heat, light, ventilate and cool facilities. Therefore, it makes sense to site power plants near both the electrical and thermal loads to make use of the nearly 70% of energy that is annually wasted by large central power plants.

CHP systems are frequently reserved for larger facilities due to high first costs and complex operations, however 75% of all buildings in the U.S. have an area of less than 10,000 ft² (929 m²). There have been several attempts made by various corporations to break into the micro CHP market with limited success. Studies commissioned by the Department of Energy show that two of the key barriers to the adoption of CHP systems in smaller facilities is the high first cost and the lack of packaged plug-and-play systems.

To address this challenge, the Center for Building Performance and Diagnostics (CBPD) has designed, installed, operated and evaluated a 25 kWe biodiesel fueled CHP system that is integrated with an absorption chiller system and an enthalpy recovery ventilation system with solid desiccant dehumidification in a single system that provides all of the electric, cooling, heating, and ventilation needs of the Intelligent Workplace, IW. The chiller and ventilation systems are well understood with three published dissertations in the last four years.
This dissertation integrates elements of each subsystem through the use of calibrated simulations to determine the effectiveness of operating such a system in a commercial office building as well as potentially in a data center.

Key contributions of this work include:
• A complete accounting of how the CHP system is setup and how it operates withboth Diesel and biodiesel fuel.
• A generic preliminary design procedure for the CHP system of a building as wellas the specific design procedures for the biodiesel fueled CHP system.
• A simplified TRNSYS CHP system performance model that can be easilyadjusted to be used for different buildings and/or for different prime movers.
• A conceptual systems integration model, which identifies how components andsub-systems may fit together.

Key results in this dissertation include:
• The results show that for efficient and effective performance of a CHP system in a high performance building it is essential to have electrical and thermal grids available to export and/or import CHP energy. The grids allow the CHP system to operate continuously at the design load. The grids also provide back up in case of system outage.
• The results of operating the biodiesel fueled CHP system in the IW yields anaverage annual efficiency of about 66% and a peak of 78%.
• A scaled up system for the Building As Power Plant (BAPP) will achieve similar efficiencies unless a larger load for the coolant energy can be found.
• Data centers offer an ideal location for CHP systems as they do not have such highly variable loads such as office buildings. Furthermore, data centers do not have latent cooling or heating loads, which simplifies systems integration, as the only components required for the system are an engine or turbine, heat recovery equipment, and absorption chillers. A CHP system with absorption chillers has been calculated in this dissertation to achieve an average efficiency of 78% in data centers.

There are many possible next steps; however the three most important steps in the development of the CCHP/V technology are to complete the automation and integration of the CHP system with the rest of the IW.

Second, to refine the BAPP data for the TRNSYS simulation and to create a modular CHP system in TRNSYS so the development of BAPP mechanical system can proceed and provide a future testing ground for packaged CCHP/V systems.

Third, to conceive and develop the means for reducing equipment and installation costs by a factor of 10 to 20 must be developed.

by Eric Schweikardt, 2009
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: This thesis is about heterogeneous modular reconfigurable robotics: a class of physical systems made up of a number of tangible, computational components that operate concurrently. Although this is a new type of system and not in widespread use for any purpose, the current research interest and potential benefits of future effective systems suggest that it will be soon. Two contributions related to the design of these systems are described.

In an attempt to inform robot designers, this thesis analyzes the properties and parameters of roBlocks, a modular reconfigurable robotic construction toy. Design issues related to mechanics, data flow, and power are addressed, and the system, its use, and its affordances are described in detail. Design of the roBlocks system provides a valuable case study; design decisions and their alternatives are discussed and the system properties that emerge from them are analyzed.

The second contribution involves the design of robotic constructions using the modules of these systems as media. Instances of this class of systems, with distributed, concurrent intelligence, exhibit emergent behavior that can be difficult or impossible to predict. This makes the
designer’s task of creating a functional robot out of a kit of parts very challenging. This problem is addressed by showing how evolutionary algorithms, in combination with accurate physical simulation, can create viable robots out of a modular kit. An evolutionary, automated design
program is presented that evolves robots in a manner based on genetic programming techniques and tests them in 3D physical simulation against user-specified objectives.

The complex, emergent behavior that is produced by combinations of modular robots (and there are lots of combinations) is hard to design without the aid of tools. This thesis offers tools in the form of (a) an analysis and case study for describing and designing this class of systems; (b) a physical kit that can be used to explore modular robots composed of smaller, heterogeneous, concurrent robots; and (c) a series of algorithms that evolve, or design, candidate configurations in order to satisfy design requirements.

by Sang Hoon Lee, 2009
PhD-Architecture-Engineering-Construction Management

Download Full Text [PDF]

ABSTRACT: There is significant potential for improvement in the performance of Operation and Maintenance (O&M) fieldwork. O&M occurs throughout the lifecycle of a building; the majority of expenses in a building’s lifecycle are incurred during O&M. Many strategies have been developed to enhance the O&M environment. However, it is well-known that the maintenance industry adapts new technologies more slowly than other industries. Although the industry’s O&M support systems have been enhanced considerably, its overall style of O&M fieldwork has remained essentially unchanged for decades. Furthermore, tradespeople, whose primary roles are O&M fieldwork, vastly underutilize information in the field due to problems with information accessibility and reliability.

This research investigates current practices from the initial phase of assigning O&M requests through the completion of the requests in order to identify inefficiency in O&M fieldwork and to develop strategies to improve the environment from the perspective of computational support. As the first step, shadowing tradespeople was conducted to better understand current O&M fieldwork and pinpoint bottlenecks in the workflow. Statistical analyses (F-test, Analysis of Variance and R2-Test) were conducted to see the correlation among O&M activities as well as the similarity of the collected data.

An Augmented Reality (AR)-based Operation and Maintenance Fieldwork Facilitator (AROMA-FF) is developed to computationally support O&M fieldwork. An O&M information model is developed by enhancing an existing Building Information Model with the data collected from O&M fieldwork practice. An Augmented Reality-based interface is developed for an intuitive user interface. BACnet protocol is used to get sensor-derived operation data in real time from Building Automation Systems.

A series of experiments was conducted in order to quantitatively measure improvement in O&M efficiency by using a software prototype of the AR-based O&M Fieldwork Facilitator. The key metric was time spent on O&M activities. The most impressive finding from the experiment is that while the subjects were trying to locate the target area, they spent, on average, 49% less time with the prototype than conventional strategies in addition to an 8% decrease in time spent getting operation-related data. These results show that the prototype is capable of improving O&M fieldwork efficiency.

by Chaoqin Zhai, 2008
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Desiccant coated enthalpy recovery and dehumidification devices have the potential to enhance the dehumidification performance of HVAC systems, and thus meet the needs of improved indoor air quality. Meanwhile, they can reduce or eliminate the energy penalty, including peak electricity demand and overall energy consumption, associated with higher ventilation rate and better humidity control. They make it possible for water based cooling devices such as radiant cooling panels, water mullion and fan coils to function under humid climatic conditions, by mitigating the impact on indoor humidity conditions.

In the past, there have been several attempts in modeling the operating performance of the desiccant wheels. These published models are either applicable to the enthalpy recovery wheel or the active desiccant wheel. Only a few are claimed to be applicable for both. In addition, there has been a lack of physical understanding of the desiccant materials despite the different moisture transport models presented in the previous publications. The practical issues encountered in the wheel operation, such as the wheel purge, the residual moisture contained in the desiccant materials and the impact of the wheel supporting structure, have not been considered. Furthermore, very limited validation information has been provided for the existing models.

This thesis presents the development of an equation based model to predict the operating performance of both the enthalpy recovery and the active desiccant wheels, based upon fundamental scientific and engineering principles. This model has related the desiccant wheel’s performance to its design parameters and operating conditions. The moisture transfer processes have been developed based on the physical analysis of desiccant materials. The effect of the practical issues on the operating performance of desiccant wheels has also been considered.

The model has been validated using the experimental data collected from a field installation of the enthalpy recovery and the active desiccant wheels. Reasonable agreements between the simulated and the measured performance parameters have been obtained, which indicates that the model well represents all significant mechanisms occurring in the desiccant wheels. This model can be used in selecting design parameters and operating variables, and in diagnosing experimental data for the desiccant wheels.

This model has been applied to evaluate the costs and benefits provided by the enthalpy recovery wheel. It is shown that the enthalpy recovery is an economic design feature. When properly applied, its payback is immediate for most climatic conditions.

As a thermally activated device, the active desiccant wheel represents a good candidate for CHP integration. This performance model has been used to develop operating strategies for the active desiccant wheel integrated CHP system.

by Ming Qu, 2008
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: The work presented in this thesis deals with the question of how solar energy might most effectively and efficiently be used in supplying energy for the operation of a building. The approach to dealing with this question has involved a specific building space, Carnegie Mellon’s Intelligent Workplace; a specific solar system, parabolic trough solar thermal receivers, Parabolic Trough Solar Collector’s; and a specific building energy use, space cooling and heating. The work has involved the design, installation, and test of a system incorporating PTSC’s, an absorption chiller, a heat recovery exchanger, auxiliary equipment, instrumentation and controls. Mathematical models based on fundamental scientific and engineering principles have been developed and programmed for both the PTSC’s and the overall IW cooling and heating system, These models have been improved and validated through comparisons of predicted and measured PTSC and IW cooling and heating system performance. The work reported in this thesis has developed suggestions and methods for the effective design and evaluation of PTSC’s and also for the optimized design and operation of solar absorption cooling and heating systems, so that the system is able to reduce building energy consumption, and achieve environmental benefits in the operation of buildings by the use of renewable, solar energy.

by Ying Hua, 2007
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Today's white-collar work becomes increasingly interactive and knowledge intensive. The prevalence of open-plan workplace setting today and the increasingly intense, open and informal interactions in workplaces create a tension between the needs for distraction-free concentrated work and rich interaction that presents the research question: "how should we design open-plan workplaces for effective collaboration?" Compared to a rich body of literature on workplace spatial impact on occupant reactions to work and work environment, there have been few attempts to build the understanding of workplaces' spatial impact on collaboration-related spatial perception and work performance.

Based on an extensive literature study and a two-year multiple-site field study with 308 participants, this thesis introduces to the workplace studies a new set of spatial variables that effectively describe workplace folio plan layout characteristics and the features of a wide spectrum of formal and informal places at work where interactions and collaboration take place. Findings from this theoretically informed and empirically grounded research identify a mole of workplace spatial impact on occupant spatial perception and collaboration effectiveness. Based on the model open-plan workplace design guidelines are generated to inform workplace design and reengineering, in order to create work environments that facilitates collaboration effectiveness by enhancing performance in both concentrated tasks and interactive team work.

by Hongxi Yin, 2006
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Developments in absorption cooling technology present an opportunity to achieve significant improvements in microscale building cooling, heating, and power (BCHP) systems for residential and light commercial buildings that are effective, energy efficient, and economic. However, model based design and performance analysis methods for micro scale absorption chillers and their applications have not been fully developed; particularly considering that thermal energy from a wide variety of sources might be used to drive the chiller in a residential or light commercial building. This thesis contributes important knowledge and methods for designing and integrating absorption chillers in BCHP systems that reduce energy consumption, decrease operational costs, and improve environmental benefits in residential and light commercial buildings.

To be more specific, this thesis contributes the development and application of absorption chiller and the computational model in the following areas:

1. establishment of a unique experimental environment and procedures for absorption chiller tests under various conditions
2. conduct of a comprehensive testing program on a microscale absorption chiller
3. construction of a comprehensive chiller model based on the pertinent scientific and engineering principles adapted to the design of a chiller and to the analysis of extensive, detailed test data obtained from the test program
4. analysis of the measured data, refinement of the model, and improvement of the chiller design on the basis of the data analysis process

The model is now being used as a tool to adapt the chiller to various heat sources and sinks and to carry out performance simulations of micro BCHP system.

by Muhsine Tanyel Türkaslan-Bülbül, 2006
PhD-Architecture-Engineering-Construction Management

Download Full Text [PDF]

ABSTRACT: In the context of architectural evaluation this research investigates the computability of building commissioning. The objective is showing that a data model that represents building commissioning information and operations can be used to facilitate standardization efforts in this domain and support the development of computational tools.

Architectural evaluation is multifaceted. A building can be considered either as a single object or as a complex system that is composed of independent units. This creates the need for multiphase, multi-agent evaluation procedures. Ordinarily, isolated measurements with discrete objectives determine our evaluation of buildings. However, the goal of the AEC (Architecture Engineering Construction) industry should be the persistent improvement of the quality of buildings through continuous evaluation. This goal can be achieved: (1) if evaluation is done systematically, (2) if the evaluation data can be kept and updated for future use, and (3) if there is continuity of information throughout the different phases of building lifecycle.

Commissioning is a new area of practice and research in the industry that promotes evaluation of buildings during several points in the delivery process. It is a multi-phase process that ensures that the interacting systems in a building are properly installed and operating. Building commissioning has the potential to make architectural evaluation a persistent part of building lifecycle. However, current practice lacks standardization in commissioning procedures, computational support is not fully utilized, and building information produced during the commissioning process is rarely used as a reference for maintenance activities during phases of occupancy.

In order to address these issues this study aims to develop a product model, which will provide a basis for computational support in the digital environment. It will allow standardization of the commissioning processes and promote seamless data exchange, which can create continuity for commissioning information in different phases of the building lifecycle.

by Wen-Jaw Jonah Tsai, 2005
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Collaboration among discipline-specific design teams in the building design process is not only a common practice but also a vital aspect for the success of any non-trivial design. While many design task are now supported by, if not automated, by individually developed computer applications, conflicts in the world views, practices, and methods among disciplines have created islands of information. The current belief among researchers in this area is that shared resources must have a shared description to facilitate the information exchange between applications while allowing disciplines to maintain their own views on the same design artifact. In addition to a shared description, a communication infrastructure supporting collaboration among multiple disciplines not only must support literal exchanges of information but also must allow disciplines to maintain their internal belief system and data model that are vital for their respective design tasks and methodology. At the same time, the shared description cannot reflect a universal belief system and data model that is consistent and efficient for all the participating discipline-specific teams because it is simply not attainable with our current understandings of the design process, in particular, when human subjective judgement is involved. This thesis relies instead on a shared description that is minimally sufficient for design teams involved in a specific project in the sense that it captures just the information that has to be exchanged in this particular context.

Although at first sight, it seems that straight translations/mappings among different data models/formats would suffice, research in this area is, in general, limited to translations of finished design products in a batch manner. For collaborations on design products that are still in the design process, a finer grain of mapping and communication method is needed. Recent research in the area of information exchange has, in general, left the mapping issue unaddressed and handles the process in an ad-hoc manner. The consequence is not only duplication of efforts spent on similar tasks, but also and most importantly, any slight modification to the shared model will result in full reviews and modifications of all the mapping modules.

To address these issues, a series of experiments are conducted in several stages in conjunction with James Snyder’s thesis work using the Object Modeling Language (OML) and the SPROUT language. The insights gained from these experiments are then formalized in a communication infrastructure called GLYL that includes a persistent data storage and a routing server, several test clients (all utilizing the GLYL software library, which is designed to be integrated into new clients and servers), and a formalized concept mapping mechanism using the SPROUT language as the data modeling language such that the infrastructure can be targeted toward different software and hardware platforms. This allows software developed by different disciplines to select software and hardware platforms deemed most suitable for their tasks.

by Moustapha Hoda, 2005
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Design is often described as an exploration: a search for an adequate solution amongst as pace of alternatives (Simon 1969). It involves the development and transformation of alternatives. Composition of spatial relations such as symmetry, hierarchy, grid-alignment, and proportion, termed “design structures,” help in the development of these architectural alternatives, and are used as compositional principles during their transformation.

During the exploratory, early phases of design, configurations continually evolve. Both configuration elements and their relational structure are subject to change. Defining structures, exploring variations within structures, and redefining structures are the common means of transforming alternatives. Such transformations are necessary for developing design configurations and improving their quality.

Transformations of design structures, which often yield intellectually stimulating results, are labor intensive; they require individual modification of related elements. Such repetitive interaction considerably slows down the exploration, and often discourages it completely, particularly when configurations are complex and interrelations are numerous.

My research is motivated by the following factors: (i) the necessity of flexible geometry for early design exploration; (ii) the intellectual stimulation provided by the exploration of structure; (iii) the difficulty involved in transforming design structures; (iv) the lack of computational support for design exploration; and (v) the lack of comprehensive representations for architecturally significant design structures.

I have developed a framework of strategies that allows designers to explore complex configurations by manipulating their organizational structure. This framework, named Interactive Configuration Exploration (ICE), consist of two parallel endeavors: a notation and a computer implementation. The ICE notation is a formalism for describing shapes and configurations, by means of their generative and relational structures. The ICE implementation is a 3D modeling system that supports the exploration of such shapes and configurations through the transformation of their structures.

The approach used is to separate the structures from configuration elements. In this manner, we can use structures to summarize configurations in the ICE notation, and use structures as manipulation handles to control the configuration in the ICE system.

The principal vehicle in ICE is the regulator, which is an abstraction that captures a single unit of structure (i.e., a single relationship within a configuration). For instance, a grid structure is captured by alignment lines; a symmetry structure is captured by a reflection axis or center of rotation. Regulators, which are inspired by regulating lines, encapsulate a mathematical formula that determines the relationship between elements.

The ICE notation enumerates and classifies the various types of regulators. It defines composition strategies and generation methods in order to represent the widest possible range of configurations. Furthermore, it captures a method for generation as well as a set of applicable transformations for any given configuration, based on its organizational structure. The ICE notation is not merely a geometric descriptor. It allows the derivation of additional geometric information, such as subshapes, boundaries, lengths, areas, volumes, and midpoints by means of simple computations on the notation strings. Additionally, it is possible to derive steps for transforming one configuration into another, by means of a simple algorithm.

The ICE system, offers a higher-level of interaction with design configurations though regulators. These regulators maintain control over configuration elements, thus imposing relational constraints and propagating changes within the configuration. The parameters of regulators are manipulation handles; therefore, a user can transform the configuration, either completely or partially, by applying simple changes to the regulator. Such explorations yield significant transformations with relatively short paths, since manipulating a spatial relation results in the simultaneous transformation of multiple elements.

by Ipek Özkaya, 2005
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Requirement traceability is the commonly used term to refer to the management of the relationships that exist between design requirements and design solutions, throughout the design life-cycle. Current design computing practice makes only limited use of these relationships. Most of the current methods of computational requirement traceability rely on sequential or ad hoc processes. Such processes force the designers to either juggle the requirement information among many applications, or customize generic applications, such as databases or spreadsheets, to fit their needs. The resulting applications are not only inefficient in their use of requirement information, but also create problems in change management, consistency checking, and design compliance verification. I believe, augmenting architectural design computing environments with requirement traceability functionalities offers opportunities to overcome these drawbacks.

This research seeks to answer the following questions: 1) how can requirement traceability be integrated with computer-aided design at a process level; 2) how can a requirement traceability capability be created within computer-aided architectural design; and 3) how can the potential impact of requirement traceability be measured.

In addressing the research questions, a computational, hybrid assistance framework for the requirement management process, CHARM in short, is presented. The process captures activities both in requirement specification and design exploration phases to support traceability. Based on the observations in CHARM, a prototype application called DesignTrack, was developed. The capabilities of DesignTrack were evaluated by applying them to Leadership in Energy and Environmental Design (LEED) standard requirements. This evaluation provides evidence of the practicality and usefulness of CHARM and DesignTrack.

This research advances design computing by highlighting the requirement-driven design thinking process and traceability enabled computer-aided design support. CHARM introduces requirement-driven design thinking into computer-aided design. It introduces information traceability as a design task which assists designers to manage information change. The computational structures for requirement traceability utilize relationships between requirements and designs. DesignTrack as a prototype introduces an integrated design development environment for requirement specification and form exploration in the same design session. This effort provides a navigation environment for complex design information spaces.

by Michael James Cumming, 2005
PhD-Architecture

Download Full Text [PDF]

ABSTRACT: Collaborative design is a complex cognitive and social activity that requires coordination of both processes and products between its participants. Information required for this coordinative activity are descriptions of the various tasks and products found within a design project, and of the current state of these entities. State descriptions can arise from technical analysis, perhaps employing automated, machine-based methods, or can arise from a social process of consensual, collaborative assessment that results in design team members applying informal linguistic descriptions to processes. In the event that no automated process exists for state determination, then members of the design team must work together and find a mutually agreeable assessment of state. With this information designers are better able to determine the progress and status of a design process, and to assess their roles and responsibilities within a design team.

This research describes the design and implementation of a design support tool that enables distributed teams to collaboratively determine the state of design entities, such as tasks and products. The tool is role-based, and enables users to communicate simple looped state-transition models that they feel suitably describe the possible states and transitions that a design entity could experience. These state models can describe the degree of completion, degree of acceptance within a team, or progress with respect to a series of milestones. By attaching entities to simple state-transition loops, users make input based on simple questions about the state of individual entities, rather than complex ones arising from the interaction of entities. Complex branching process structures can be created by composing entities. The tool automatically handles state assessment of complex, linked compositions of entities, while users handle assessment of simple, non-linked entities. It provides users with information regarding design state and structure, and supports a form of bottom-up design coordination that requires no centralized policies or inputs, prior to deployment.

by Kamal Mubarak, 2004
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: This thesis presents a process model for design composition using an abstraction of form and function characteristics of design solutions. The design process model is based on adapting previous solutions to generate new designs using derivational analogy. The model relies on a geometric representation of design compositions. This representation encodes functional and form attributes of design and is used to build compositional characteristics of the design. The research uses Solution Traces (Sol-Traces) as constructs for recording and reusing the design composition. The trace consists of compositional steps. The trace – the sequence of steps - can be replayed to generate design solutions.

Design reasoning with architectural cases has a long history in architecture. Generative case based reasoning using derivational analogy is a powerful problem solving technique that enables new designs to be created by utilizing the generative path of prior designs. This technique is adapted for use in developing a design assistance system for design composition in architecture. Solution Traces (Sol-Traces) are used within a CBR methodology for a design composition assistant: the TRACE system. Cases in TRACE include architectural representations such as floor plans, form diagram, function diagram, and the sequence of design composition steps that lead to a particular solution. The TRACE system utilizes two strategies, transformations and formative ideas, to generate forms. Cases and their components are classified and indexed in the case base using both form and function attributes. The thesis presents three worked examples using the TRACE system.

The main contributions of the research are: the abstraction language for design composition using the Sol-Traces representation and the process model for design composition using this approach. Another contribution of the research is the application of the generative CBR, using derivational analogy, in architectural design composition at the early phases. The research also provides the development of the TRACE system as a CBR system utilizing the research findings.

by Jayada Boonyakiat, 2003
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: The prevailing paradigm in indoor environment control (particularly in office buildings) often excludes contextual forces such as daylight and natural ventilation, partly because they could interfere with the "smooth" operation of mechanical conditioning systems for indoor air. Given the current indoor environmental control expectations in modern buildings, and given the outdoor climate constraints in various regions of the world, it is not likely that all buildings could be operated with passive technologies only. This necessitates the integration of natural ventilation into the design and operation of indoor environmental control systems. An obstacle in the way of such integration is the difficulty of predicting the behavior of passive environmental systems. For examples, natural ventilation models for building are often either too crude to be useful for reliable control algorithms, or prohibitive in view of required computational resources. This research presents an on-site calibrated simulation-based approach to treat natural ventilation as an integral component of the operation buildings' thermal systems. Toward this end, a multi-zone air-flow model is used to predict the fresh air volume flows into the building interior as a function of building geometry, status of passive ventilation devices, and micro-climatic parameters. The predictions of this model are then compared with actual measurements in the target building, resulting in a statistical "wrapper" to improve the model's predictive capability for the target building. The resulting statistically augmented numeric multi-zone model is integrated as part of a simulation-based control algorithm for the operation of the buildings' thermal systems.

by Halil I. Erhan, 2003
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Developing design requirements specification or an architectural program for a recurring building type offers unique opportunities for the programming phase in design. Recurring building types are repeated in different contexts, yet their general functional aspects do not change.

However, current practice makes only limited use of these opportunities. The use of passive programming media and manual methods together with non-standardized representation formats create problems with continuity, with upgrading programming knowledge, and with handling complex design requirements. Computer-based tools are generally limited to simple database or spread-sheet applications. Only few of these tools provide some generative mechanisms for formulating design problems separate from solution generation.

I believe that computer-assisted generative tools can assist programmers in partially alleviating the bottlenecks mentioned above and reduce the cognitive loads posed by using traditional manual techniques for handling complex programming information. Based on case studies and an extensive literature survey, I developed a general and flexible framework that models architectural programming knowledge as a generalized (extended) means-ends analysis; it can be made operational in the form of a computer-based support tool that can be adapted by users to any building type and is particularly suited to support programming recurring building types.

RaBBiT is a first prototype application. It is distinguished by the following features: (a) the ability to  computationally capture reusable programming knowledge based on a set of concepts that are general enough to accommodate various programming styles while remaining operational; (b) simplification of the designer-computer interaction to make the application usable, even programmable to a degree, for non-computer programmers; and (c) the ability to generate design requirements as output that can be used by other generative design and decision support tools. The first prototype consists of an object-oriented application that is highly integrated with a direct-manipulation user interface.

by Kristie Bosko Mertz, 2003
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Modern buildings tend to be very dynamic and complex entities. Dynamic exterior and interior environments, occupants, and systems cause a range of reactions and interactions that influence occupant comfort and impact the global environment. Control systems must respond to these dynamic changes in buildings but also must respond to innovations in building  design and advanced operational strategies.

Due to the shift away from blanket and centrally controlled systems and toward local environmental systems, control systems are challenged. They must be able to not only integrate local and central system processes but also to integrate local processes across multiple domains, such as lighting, heating, ventilation, and cooling. At the local level, multiple devices and environments influence each spatial parameter, yet control systems are not equipped to consider all influencing factors together. Without integration, systems may work against each other, preventing occupant comfort in every space or reducing potential energy savings to the building as a whole.

Control systems have traditionally been designed separately for each environmental system. Attempts to create an interoperable system that responds to the requirements of the modern control system generally involve protocols to enable communication among the central control systems of the various domains. This approach proposes integration at the local level by considering the impact of multiple devices on the same spatial parameter and of one device on multiple control zones. Such protocols do not define the method by which integration is achieved, where algorithms of multiple control systems cooperate to achieve a common goal. Essential to the implementation of such methods is the development of a control system structure to represent and coordinate the complex processes that enable integration.

However, such a structure cannot easily be represented by the geometric models commonly utilized to describe organizational and environmental system components. Since control system software is a series of hierarchical processes, the structure must represent these processes and not the spatial areas that they influence. On the other hand, this structure must provide a method to map the various geometric models into the process model to enable communication, data transfer, and system operational effectiveness. In addition, the algorithms that animate the model must be accurate, dynamic, and systematic to enable integration. To that end, rule-based and simulation-based control methods are examined and employed in an implementation of a two-zone space that considers device impacts across domains and shared devices between zones.

by ZhengChun Mo, 2003
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: Two of the primary objectives of building operations are maximizing occupancy comfort and minimizing energy costs. While research effort has been focused on concept development, design decision support and systems advancement, little attention has been paid to operational decision support. Most commercial buildings are operated under a central control scheme, in which a building operator makes control decisions without in-depth information about individual preference. Widely used set points represent generalized human requirements that do not sufficiently address individual differences. Energy costs, on the other hand, are easier to measure. As a result, operational decisions tend to favor cost savings at the expense of individual occupancy comfort.

Personal control systems have enabled individual occupants to customize their local environments. It is argued that individual occupants and building operators have different motivations for environmental controls. They access to different scopes of information and represent partial knowledge for operational solutions. Such a new control environment suggests a bi-lateral control scheme that cannot be offered by existing central control schemes or distributed control schemes. There is a critical need for methods that support the bi-lateral control scheme, in which building operators and individual occupants coordinate to make balanced control decisions. Toward this end, an agent-based simulation-assisted computational framework has been proposed and prototypically implemented in the lighting controls domain. The prototype supports bi-lateral building operations by offering concurrent evaluation of alternative control strategies. The experimental results showed that, by utilizing the proposed framework, the energy use is greatly reduced without undue increase in individual visual discomfort.

by Beran Gürtekin Çelik, 2001
PhD-Building Performance & Diagnostics

Download Full Text [PDF]

ABSTRACT: It is generally believed that the utilization of computational building performance simulation tools can contribute to the improvement of building designs. Accordingly, many such tools have been developed. Yet, their application in (and thus their impact on) the building delivery process has been rather limited. This thesis focuses on one possible contributing factor, i.e., insufficient support for navigation in the design-performance space. While many efforts have been invested in algorithms and models that help generate building performance data, much less has been done to support the process of organizing, exploring, and evaluating such data. To address this shortcoming, this thesis presents, for the building design domain, an approach to generation and exploration of the design-performance space. In this approach, an initial design is used to generate a set of alternative designs that collectively constitute the design space. One way of doing this relies on the "scalarization" of design variables. The scalarization leads to the representation of a building as a point in a d-dimensional design space. Each coordinate of such a space accommodates a salient (semantic or geometric) design variable. Subsequently, the entire corpus of design alternatives is subjected to performance modeling. Based on the modeling results, an n-dimensional design-performance space is constructed, where n = d + p (d: number of design variables; p: number of performance indicators). Once constructed, this space can be visualized and used by the designer to explore the relationship between design variables and corresponding performance
attributes.

by Mustafa Emre Ilal, 2001
PhD-Computational Design

Download Full Text [PDF]

ABSTRACT: Due to its potential impact on the AEC industry, integration of CAD environments with building performance analysis tools has attracted many research efforts over the last 20 years. During this time frame there has been a significant lack of standardization in building performance research as numerous analysis applications have been developed that are powerful yet cannot communicate with any other tool. The S2 system approaches integration with a performance perspective and provides integration tools for analysis applications based on distributed computing technologies, forming a network of building performance computing that is independent of any CAD environment. The S2 system embodies the requirements of a finite set of analysis domains for a certain range of design resolution that is useful for designers especially in the design development phase. This thesis presents contributions to the S2 framework and components emphasizing issues of representation, mapping, and software architecture.