View a printable version of this program.
Last updated: Tuesday, Feburary 28, 2012
Professor Michelle Addington – Yale University
From the proliferation of sustainability consultants and green design firms to the sweeping adoption of energy building codes and LEED certification, a wide range of sustainable objectives and methods have entered the profession with the potential to profoundly reshape the practice and products of building design and operation. Committing to a target, adopting an array of best practices, using state-of-the-art evaluation tools, and installing advanced technologies are all aspects that are associated with the sustainable design of buildings, but do they indeed lead to effective results? Energy use by buildings is not only increasing faster than that of any other sector, and but also faster than the rate of construction. Of most concern, however, is that the largest increases are in buildings that adopted many of the accepted sustainable practices. We yearn for well-defined solutions, but perhaps we need to look beyond our normative context for the questions we should be asking instead of coming up with solutions for ill-defined problems. We frame our problems in terms of what we know and how we do things - but what would or could happen if we had the ability to step back and question the very construction of our assumptions? We have at our disposal an unprecedented array of advanced simulation tools, can we use them to ask fundamental questions about how energy phenomena behave and how buildings should be conceptualized?
Michelle Addington, Hines Professor of Sustainable Architectural Design at Yale University, is educated as both an architect and engineer whose teaching and research explore energy systems, advanced materials and new technologies. Building on her dissertation research on the discrete control of boundary layer heat transfer using micro-machines, she has extended her work to defining the strategic relationships between the differing scales of energy phenomena and the possible actions from the domain of building construction. Her articles and chapters on energy, system design, HVAC, lighting and advanced materials have appeared in several journals, books and reference volumes, and she recently co-authored a book titled "Smart Materials and Technologies for the Architecture and Design Professions." Addington also taught at Harvard University for ten years before coming to Yale in 2006. Her engineering background includes work at NASA Goddard Space Flight Center, where she developed structural data for composite materials and designed components for unmanned spacecraft, and she spent a decade at Dupont as a process design and power plant engineer as well as a manufacturing supervisor. In 2009, Architect magazine selected her as one of the country's top ten faculty in architecture.
Phillip G. Bernstein – Autodesk
Phillip G. Bernstein is a Vice President at Autodesk, a leading provider of software for the architecture, engineering and construction industry. An architect with twenty-five years of experience, he leads Industry Strategy and Relations for the AEC Division where he is responsible for setting the company's future vision and strategy for technology serving the building industry. Phil was formerly with Pelli Clarke Pelli Architects where he managed many of the firm's most complex commissions. Phil teaches Professional Practice at the Yale School of Architecture where he received his both his B.A. and his M.Arch. He is co-editor of Building (In) The Future: Recasting Labor in Architecture published in 2010 and BIM In Academia published in 2011. He is a Senior Fellow of the Design Futures Council and former Chair of the AIA National Contract Documents Committee.
Martin Tamke, Elisa Lafuente Hernández, Anders Deleuran, Christoph Gengnagel, Mark Burry and Mette Ramsgard Thomsen
Computational techniques in design and fabrication introduce a building practice where a structures shape and dimension is negotiated between architectural intent and performative aspects. Within the frame of the Dermoid demonstrator this discussion takes place on multiple levels. In this construction made of short plywood members it is the topology, structure and material that is equally informed. A practice is introduced were the behavior of material is as much a part of design as its specification. Where the concept and need of this process of integration is comprehensive the implications on the level of simulation and the process to feedback received results into the design level are ambiguous. The first level of questions appears when we simulate a structure with material active components. Here we are due to restraints on computational level as due to the nature of a necessarily open ended design process not able to simulate the structure in a high resolution. The Dermoid projects proposes a multi-scalar approach which is based on the calibration of the FE simulation with the help of physical prototypes. The information gained from this single component informs the behavior of the overall system. The measurement of the build structure and comparison of its simulated digital counterpart allowed an evaluation on the overall system level. The feedback between simulation and design poses the second layer of questions. This process is taking part in a chain of processes starting with the generation of shape employing the nucleus engine as constraint solver, followed by further steps in parametric software and finally the cnc fabrication. The Dermoid project discusses the inherent possibilities and complexities that come along when feedback is introduced at different steps of the computational design process.
Mansour Jadid and Mustafa Badrah
The objective of this paper is to implement a decision support system for materials selection for projects under design or construction by consultants and owners. The paper focuses on issues of materials approval, selection criteria and materials information management. The proposed system includes database and decision support components. The database can enhance the functionality of the selection process as it provides a source of information to feed into the decision support component. The decision support component relies on the quantitative methods of value engineering. Construction involves processes that deal with materials in all stages of a project. Decision making in project meetings requires the application of knowledge from many disciplines, by all parties in the project teams to select materials and avoid delays for the project under construction. Acquiring and recording information (in a database) regarding material quality, durability, specifications, and maintenance requirements is a vital part of good project management practice. It is anticipated that using such a system for materials selection will provide an improved tool that systematically enhances the selection process and enables upgrading the system during project life cycle.
Design Optioneering: Multi-disciplinary Design Optimization through Parameterization, Domain Integration and Automation of a Genetic Algorithm
David Gerber, Shih-Hsin "Eve" Lin, Bei "Penny" Pan and Aslihan Senel Solmaz
The overall performance of buildings is heavily impacted by design decisions made during the early stages of the design process. Design professionals are most often unable to explore design alternatives and their impact on energy profiles adequately during this phase. Combining parametric modeling with multi-objective design optimization has been previously identified as a potential solution. By utilizing parametric design and multi-objective design optimization to influence design at the schematic level in the interest of exploring more energy efficient design configurations, the H.D.S. Beagle 1.0 tool was developed. The tool enables the generation of design alternatives according to user defined parameter ranges; automatically gathers the energy analysis result of each design alternative; automatically calculates three objective functions; and uses Genetic Algorithm to intelligently search, rank, select, and breed the solution space for decision making. Current case studies demonstrate our tool's ability to reduce design cycle latency and improve quality. However, the future work is to further investigate how to acclimate this process to accommodate early design stages and processes.
Ben Doherty, Dan Rumery, Ben Barnes and Bin Zhou
We have developed a piece of software for finding shortest paths in buildings. It consumes data produced by building designs documented in Revit. It aids design creating simulations of where people would travel if they were omniscient and there were no flow restrictions; this allows for circulation volumes to be predicted more quickly than through comparable agent simulations. It also provides distance related information for regulatory checks, mean euclidean distance calculation and a number of other types of analysis. Our paper will introduce this tool and give some worked examples of its use.
Julien Nembrini, Steffen Samberger, Andre Sternitzke and Guillaume Labelle
The paper presents an approach which combines parametric geometric description techniques, performance simulation, and sensitivity analysis to achieve a quantitative analysis of the influence of selected design parameters on performance. By using a scripting environment for parametric design linked to the EnergyPlus building simulation tool, it becomes possible to automatically screen and simulate thermal behaviour over a selected design parameter space. Then by defining one - or several - performance metrics, their sensitivity to modification of design parameters can be computed through established statistical techniques for sensitivity analysis.
This approach provides quantitative insight into the level of influence of each parameter under scrutiny, informing about changes altering, enhancing, or having only minor effect on performance. This application of sensitivity analysis to parametric performative design is presented through the help of a case study.
Results and implications are discussed respective to their relevance to the early design phase of an architectural project. The framework in definitive provides the design team with means to balance design intentions with performance aims.
Zheng Yang, Nan Li, Burcin Becerik-Gerber and Michael Orosz
In the U.S., 40% of energy consumption is from buildings, approximately 48% of which is consumed by heating, ventilation, and air conditioning (HVAC) systems. Implementing demand driven HVAC operations is a way to reduce HVAC related energy consumption, and ultimately to achieve sustainable building operation and maintenance. This relies on the availability of occupancy information, which determines peak/off-hour modes and impacts cooling/heating loads of HVAC systems. This research proposes an occupancy monitoring system that is built on a combination of non-intrusive sensors that can detect indoor temperature, humidity, CO2 concentration, door status, light, sound and motion. Sensor data is communicated wirelessly, and processed in real time using a back-propagation neural network algorithm to estimate the number of occupants in a room. Field tests are carried out in a shared office space over two month. The test results show the effectiveness of each sensor in occupancy estimation, and report an overall detection rate of 73%, which indicates the ability of the proposed system to monitor the occupancy information of multi-occupancy spaces in real time to support the implementation of demand driven HVAC operations.
Preliminary Investigation of the Use of Sankey Diagrams to Enhance Building Performance Simulation-Supported Design
The diversity of building performance simulation (BPS) tools and the technologies and phenomena they support are continually growing. However, the vast majority of developments in the field are aimed at detailed design, or even post-design analysis, and such, do not place emphasis on influencing design. The characteristics of this are numerous and detailed inputs and aggregated outputs. The current research addresses the latter point by facilitating the communication of BPS results to building designers and their clients with the express intent on identifying predict building designs'. Specifically, this paper will discuss a methodology for generating Sankey diagrams using data from BPS tools. Specifically, Sankey diagrams demonstrate a method for tracking every Joule of energy from the time it enters the building (e.g., electrical supply) to the time is leaves (e.g., heat loss though the building envelope). Sankey diagrams, despite being over 100 years old , have rarely been applied to communicate building energy performance. This is likely due to complexities related to defining useful (and adverse) heat gains are well-defined. For example, the solar gains through a window may contribute to reducing heat loads, increase the cooling loads, or neither. To complicate matters, energy flows in buildings are not instantaneous - particularly for passive solar buildings, which are specifically intended to passively store solar gains for periods of as long as overnight. This paper will introduce a methodology for creating Sankey diagrams to track all energy flows through a building and demonstrate it with a high-performance passive solar house (simple example shown below).
Nathaniel Jones, Kevin Pratt, Lars Schumann, David Bosworth and Andrew Heumann
In the early stages of building design, architects rapidly produce, explore, analyze, and eliminate design options. Ideally, energy analysis should accompany and inform each design option, but in reality, the creation of building models for thermal simulation is too slow to keep up with the pace of design exploration. This paper describes a framework for rapid energy analysis of architectural early design models. The framework consists of a flexible modeling protocol to be followed by the user, a standard communication protocol that may be implemented in virtually any architectural modeling software as a script or plug-in, and a translation protocol for automated production of energy models to be run in a stand-alone program simultaneously with the modeling environment. A prototype implementation of these protocols has successfully performed EnergyPlus analysis of early design stage architectural models created in SketchUp, Grasshopper for Rhinoceros, and 3ds Max. This allows timely feedback on building energy consumption to be displayed side-by-side with an actively changing architectural model.
Siobhan Rockcastle and Marilyne Andersen
Daylight is a dynamic source of illumination in architectural space, creating diverse and ephemeral configurations of light and shadow within the built environment. It can generate contrasting levels of brightness between distinct geometries or it can highlight smooth gradients of texture and color within the visual field. Although there are a growing number of studies that seek to define the relationship between brightness, contrast, and lighting quality, the dynamic role of daylight within the visual field is underrepresented by existing metrics. This study proposes a new family of metrics that quantify the magnitude of contrast-based visual effects and time-based variation within daylit space through the use of time-segmented daylight renderings. This paper will introduce two new annual metrics; Annual Spatial Contrast and Annual Luminance Variability. These metrics will be applied to a series of abstract case studies to evaluate their effectiveness in comparing annual contrast-based visual effects.
Performance Driven Design and Simulation Interfaces: A Multi-Objective Parametric Optimization Process
Angelos Chronis, Martha Tsigkari, Evangelos Giouvanos, Francis Aish and Anis Abou Zaki
Despite the continuous development and integration of simulation interfacing tools in current architectural research, the availability and operability of off-the-shelf tools has still not met the timeframes and performance requirements of current architectural practice. The complexity and demanding performance goals of contemporary large-scale projects often require innovative approaches, as well as the development of novel simulation interfacing tools to meet these requirements. This paper reports on a multi-objective optimization process, aiming at reducing incident solar radiations whilst optimizing daylight penetration, for the façade of a large-scale office building. This was achieved through the combined use of a parametric model and a genetic algorithm, along with an integrated data set of pre-computed results. To minimize the resources demand of analyzing the plethora of potential configurations of the façade, a number of strategically defined modular cases were modeled and simulated using bespoke interfacing tools to produce a database of results. This database was then linked to a parametric model, providing real time feedback and allowing for an exhaustive search of design configurations. To further explore potential optimal solutions, a multi-objective optimization process using a genetic algorithm, also linked to the results database, was implemented. The overall optimization process provided invaluable insight to the design problem at hand.
Climatic based consideration of Double skin Façade system- comparative analysis of energy performance of a Double Skin Facade Building in Boston
This paper focuses on investigation of climatic based design for double skin façade system in a case study- addition to cambridge public library- in Boston. The double-skin façade is an architectural phenomenon driven by the aestheticdesire for an all-glass façade and the practical desire to have naturalventilation for improved indoor air quality in buildings. Until recently theuse of double-skin façades had become more popular in many buildings in Europe. While a great deal of interest exists in learning how to integrate DSFs into our current architecture, there is a little knowledge or demonstration of how the concept might work in a city such as Chicago. The primary goal of this research is to clarify the state-of-the-art performance of DSFs, so that designers can assess the value of these building concepts in meeting design goals for energy efficiency, ventilation, productivity, and sustainability. In this study building envelope performance is investigated by modeling energy performance of the building in comparison with actual energy consumption.
Sotirios D Kotsopoulos, Federico Casalegno, Guglielmo Carra, Wessley Graybill and Bob Hsiung
An examination of the aesthetic and performative advantages in optimizing electrochromic technology in residential architecture is presented. The association between a specific architectural element of a prototype house - a glass façade involving 100 electrochromic windows - and the adjustment of the natural lighting conditions in the house interior is treated algorithmically. The façade operates as a dynamic filter between exterior and interior: it filters solar radiation and heat by allowing the modification of the chromatism and light transmittance of each individual window. Varying the number and the distribution of the active windows on the façade permits the regulation of the incoming natural light and affects the visual presence and the energy performance of the house. A generative grammar producing the language of the optimum - in terms of interior illumination performance - façade patterns, is presented. The novelty of this research is that it concerns real time permutation of two distinct computational systems: qualitative (visual) and quantitative (performative). The described system yields optimum patterns for the South façade of the prototype house, through mutual adjustment of electrochromic window activation and interior illumination intensity. Shape grammar formalism and digital simulation software are used for the task.
Andres Sevtsuk and Michael Mekonnen
We introduce an open-source Urban Network Analysis (UNA) toolbox for ArcGIS, which can be used to compute five types of centrality measures on spatial networks - Reach; Gravity Index; Betweenness; Closeness; and Straightness. Though primarily developed for the analysis of urban street- and building-networks, the toolbox is equally suited for other spatial networks, such as building layout-, highway-, or utility networks. Unlike previous network analysis tools that operate with two network elements (nodes and edges), the UNA tools include a third network element - buildings - that are used as spatial units of analysis for all measures. Neighboring buildings on the same street segment can therefore obtain different accessibility results. UNA tools allow buildings to be weighted according to their characteristics - more voluminous, more populated, or otherwise more important buildings can be specified to have a proportionately stronger effect on the analysis outcomes. The paper demonstrates its application in Cambridge and Somerville, MA.
The parametric exploration of spatial properties - Coupling parametric geometry modeling and the graph-based spatial analysis of urban street networks
Sven Schneider, Martin Bielik and Reinhard König
Parametric modeling is a powerful tool that allows designers to explore a wide range of variants of their design concept. However, when evaluating the spatial properties of such variants, the tools for parametric modeling offer little support. While on the other hand a wide range of methods exists for analyzing geometry in terms of its spatial properties via graph-based-measures, the available software for analyzing these properties offers limited capabilities for changing the geometry. After analyzing the geometry, design changes then have to be made in another program, which hinders the design workflow and with it the desinger's willingness to explore a wide range of variants. This paper describes an approach that attempts to bridge the gap between parametric geometry modeling and methods for measuring the spatial properties of this geometry. To this end we developed a component for a well-known parametric modeling software package for graph-based analysis. This component is able to interpret segment maps from parametric line drawings (e.g. urban networks) and calculates centrality (integration) and betweenness (choice) for different radii. We demonstrate the applicability of the method in a test scenario.
The paper introduces a dynamic spatial model of urban retail system in which the 'retail units' of different sizes and types compete with each other and the households continuously re-evaluate their shopping orientation. The simulation model has been built within network city theory framework - especially Gabriel Dupuy's (1991) three level model of urban networks. In the model spatial approach is emphasized through morphological types of retail units and transportation network as primary spatial structure of an urban agglomeration. The model enables observation how different boundary conditions - set by the typology and transportation networks - effect on the spatial organization of retail units in urban environment. The model is based on three major elements: (1) infrastructural level: topological transportation network, (2) level of production and consumption: retail units and (3) individual level: households. The simulation model utilizes both modern agent based modeling techniques and methodologies of traditional spatial interaction models (e.g. Wilson-Bennett 1985). The interest lies on the process in which the competing retail units act as independent agents and locate themselves in urban structure. One of the objectives is to compare different accessibility measures that essentially influence consumer utilities and also on the clustering degree of retail units. Particularly the focus is on urban phase transitions and resulting morphological change. The case study has been executed in Helsinki city region in Finland.
Negin Behboudi, Fouad Butt and Abdolreza Abhari
It is our goal to develop an automatic system that employs soft computing techniques to automate the design process of a building. Neural networks replace simulation processes. Structural characteristics, occupant behavior and also external factors influence the thermal comfort level perceived by the occupants of a building. Conflicting design objectives of thermal comfort, structural properties and cost form a complex search space of design solutions highlighting the utility of multi-objective genetic algorithms as an attractive technique for searching the solution space.
Pascal Goffin and Arno Schlueter
The integration of building energy simulation tools into the architectural workflow at an early stage in the design process depends on the information and the usefulness it provides, and on how this information is visualized to the technically experienced architect. In our work, we focus on the visualization of this simulation data for decision support. A problem-driven approach is used to create new visualizations of this data to support sustainable building design, where the focus is on reducing CO2 emission, improving energy efficiency, and simplifying building technology systems. In this context, we investigate the visualization of thermal comfort. A lot of work has been done on modeling and analyzing thermal comfort, but currently we are lacking appropriate thermal comfort visualization methods for the architect. For engineers, different variations of diagrams exist that visualize comfort values, but we argue that there are better and more appropriate ways to visualize thermal comfort. By using methods from information visualization we can provide a better understanding of thermal comfort in buildings allowing the architect to take better decisions for designing sustainable buildings easier and faster. Our work follows the guidelines of the information visualization mantra that favors, first an overview and then through interaction techniques like zooming and filtering more information is presented.
Manolis Patriarche and Dominique Dumortier
Global urbanisation has reached 50% by 2010; urban areas are getting bigger and denser which implies a reduced access to daylight and solar radiation. The impact of urban topology on daylight availability on facades has to be evaluated. The process of shortwave radiation exchange between surfaces is well known and software has been recently developed to model it at urban scale. This study aims at assessing the accuracy of 3 of them: Citysim, Daysim and Heliodon. For this study, we built a model of a street canyon in 1:50 scale using moveable facades. In the middle section of the street canyon, we placed 13 illuminance sensors: 6 on each façade and 1 on the road. The model was placed outside under real sky conditions close to our CIE-IDMP daylight measuring station. Illuminances were measured every minute during a year, for various combinations of canyon geometry, orientation, façade reflection factors and sky types defined according to CIE. We modeled every combination using the 3 software. For each of them, the sky condition was described using the direct and diffuse components of daylight measured by our CIE-IDMP station. We compare the modeled illuminances to the measured ones. We discuss the differences between the 3 software and we provide recommendations for their use. For instance, we look at the tradeoff between accuracy and computation time. Further work will focus on comparing modeled illuminances to illuminances measured in real street canyons of Lyon.
A morpho-energetic tool to model energy and density reasoned city areas: description of the methodology (Part I)
Laetitia Arantes, Olivier Baverel, Daniel Quenard and Nicolas Dubus
This paper describes a "morpho-energetic" study that considers the morphology of low energy cities. Nowadays, to suit the imperatives of sustainable development, many French urban planners recommend dense and compact cities, because they are said to be low energy and land saving, contrary to spread cities. In the last twenty years, many studies showed how a building's compactness can impact its heating loads: the more compact a building is, the less its heating consumption is. But, at the urban scale, this "morpho-energetic" relation is less obvious, since cities' forms and morphologies are complex. Thus a question remains: are compact cities really sustainable and low energy? The study we present in this paper is following a previous one focused on a simple calculation method to consider the whole energy balance in the design of residential buildings. This balance takes into account every consumption post (from heating to appliance electricity, including materials embodied energy) and passive and active gains (since the building produces electricity and thermal energy thanks to photovoltaic and thermal panels). Here, this simplified energy calculation method at the building scale is used in a broader model at the urban scale, aiming to find the optimal form(s) of a low energy and dense city area. At the urban scale, by gathering buildings together, a new issue arises since minimizing sun shading is required to produce as much energy as possible. Thus, the broader model we develop gathers three stakes at urban scale: density, energy performance and sun penetration into the city. More specially, it proposes to find the optimal buildings layout (their sizes, positions and orientations) in a city area, from sun shading and energy performance criteria, by optimizing the population density. The optimization is made thanks to a genetic algorithm tool called BIANCA, developed by Paolo Vannucci, Angela Vincenti and Marco Montemurro at the French Institute Jean Le Rond d'Alembert (Vincenti 2010). In this way, we aim to assess the link between a building's energy efficience and its urban and environmental insertion.
The SimAUD DEVS Workshop is the first step toward a new open framework for collaborative simulation development that we call DesignDEVS. The goal of the SimAUD Community is to design and define the DesignDEVS API in support of a modular building and urban simulator, in support of sustainability, where academia, government, and industry can participate and contribute. For this reason, we have set the SimAUD Symposium within the SpringSim Conference to engage directly with the DEVS research community. DEVS is a mathematical formalism broadly defining simulation. As such, the DEVS community has already considered many aspects of simulation that will be helpful like parallelism, distributed simulation, hierarchical encapsulation, etc.
We will open the workshop with a DesignDEVS introduction followed by an EasyDEVS tutorial to give participants a better understanding of how a modular simulation would operate.
EasyDEVS teaches people how to develop simulations using a well-established set of conventions known as DEVS. The application is essentially an Integrated Development Environment (IDE) intended for designing simulation models and quickly testing them. Users are encouraged to decompose complex real-world systems into simpler systems that can be modeled independently by domain experts. This results in libraries of models that can be shared between users and combined in various ways. By taking the theory behind DEVS and making it practical and accessible, EasyDEVS strives to help communities of researchers and practitioners collaborative effectively on large-scale simulation projects.
We will then form groups to focus on particular areas (e.g. Materials, Algorithms, Energy Flow, Early Design, Occupancy, Lighting, Optimization, Façade, Urban Design, Analysis). Finally, we will present our thoughts to the main group.
At the end of the workshop, we will begin a SimAUD DesignDEVS Subcommittee to help define, develop, and deliver this framework to the SimAUD Community, as well as to the broader research, development, and practice communities, over the coming years. Please join us at this workshop and contact us if you would like to participate in this multidisciplinary effort. Email Azam Khan at email@example.com.