Article - Issue 17, October/November 2003
Building simulation: virtual prototyping for construction projects
Virtual prototyping for construction projects
The latest computer simulation techniques are revolutionising the construction industry through mathematical modelling. Previously, scale models were limited as a design aid and used mainly for aesthetic purposes. Now architect and builder can assess a full-scale version of a finished building, incorporating external physical processes and simulations of its intended use, before the foundations are laid. David Stribling explains the power of virtual prototyping.
In many engineering disciplines, building a prototype product has become commonplace prior to investment in full-scale production. This is an expensive process, though, and is therefore normally done some way down the design timeline when many design constraints have been fixed. In the construction industry, on the other hand, mock-ups of certain spaces do occasionally happen but there is little point in prototyping in the traditional sense. Virtual prototyping, however, is a more recent innovation in the construction industry where a computer model of a building can be constructed and tested through a variety of external conditions and operating scenarios. The added value of building simulation in terms of accuracy and design input is largely dependent on the experience and capability of a growing engineering discipline, building simulation engineering.
The role of the building simulation engineer lies somewhere between the disciplines of architecture and building services engineering and is being recognised more and more as a core function in the delivery of a successful and energy-efficient building. The architect deals with the building aesthetics, form and layout to meet the practical needs of the client as well as his/her aspirations for the building. The services engineer deals with the practical aspects of how the building is heated, cooled, powered, lit, etc. This diverse group of talents works together to co-ordinate and deliver a building, with ultimate success largely dependent on the level of integration at the early stages of the project. The simple fact is that at this stage there are many issues in the drive towards lower energy buildings and sustainable design which fall in between the traditional roles of architect and engineer.
Historically the make-up of the building design team has meant that the architect plans the site and designs the building form, space planning and orientation. Some time further down the project timeline this building concept is handed down to the building services engineer who provides the layout of heating, cooling, ventilation, lighting and power. New trends in issues such as solar shading, natural ventilation, night-time cooling, double façades, etc. are fundamental design issues which affect the architecture as much as how the building is serviced, and hence they should be considered at the architectural concept design stage. Technically, though, they fall outside the training and expertise of the architect and traditionally outside of the plant, or ‘pipes and wires’, orientated view of the building services engineer.
Virtual prototyping is ideally placed to fill this gap in the building design process. A building simulation engineer can be employed by the architect to advise on the consequences of building orientation, façade type and so on, or by the services engineer to assess the feasibility of natural ventilation, a mechanical ventilation scheme or other system. They may even be employed by the building end user for an independent assessment of a design or to identify problems with an existing building through the simulation tools at their disposal backed up by experience on a diverse range of projects.
Dynamic thermal modelling
Building simulation falls broadly into four technologies. The first of these and the most holistic in terms of virtual prototyping is dynamic thermal modelling (DTM).
The traditional way of calculating heating and cooling loads on buildings is a steady-state heat-loss or heat-gain calculation based on the thermal conductivity of the building material, the external temperature and the internal temperature. This approach is fine when designing a structure such as a tent, where a one-degree change in the external temperature is matched almost immediately by a one-degree change in the internal temperature. This is because the fabric of the tent has negligible thermal mass and hence a very low time constant. If one considers the other extreme, however, such as a traditional church, then this structure is characterised by heavy, thick stone walls which store heat with a long time constant that can be of the order of days or weeks. The majority of buildings though, are of steel and/or concrete construction, where the time constant of the structure lies somewhere in between these two extremes. This means that the thermal characteristics of a building are not merely three-dimensional but become four-dimensional with the addition of a time-based thermal lag.
A dynamic thermal model accounts for this significant effect with a three-dimensional model of the building where the internal temperatures and loads are calculated hourly based on historical weather data for the region. Furthermore, the usage patterns of the building can be applied and changes in internal heat gains accounted for during the course of working and nonworking days. This type of analysis predicts the peak loads on the building, as well as when they are likely to occur, much more accurately than conventional steady-state calculations. For example, it accounts for thermal storage in the structure overnight, meaning a reduction in heating load (and hence boiler size) during early morning warm-up periods. Similarly, early afternoon peaks in summer temperatures and solar loads may not reach the occupied space until early evening when there is no need for comfort cooling, thus reducing chiller sizes.
The global attempts to reduce CO2 emissions mean that sustainability is gaining increased popularity, leading to a move towards more naturally ventilated buildings. This raises the question of precisely what level of summer time temperature will be achieved in the space with no mechanical cooling. A DTM simulation of the building can go a long way to answering this question as well as testing innovative strategies such as night-time cooling, core activation or automatically opening vents. Other possible areas for testing may also be strategies for solar shading such as high-performance glazing or brise-soleil. Here a comparison can be made of the peak solar gain and a calculation of the number of hours a certain temperature criterion may be exceeded during the course of a year. This information can then be combined with financial information in order to arrive at a cost-effective design solution.
The latest versions of DTM software packages even allow the testing of a building’s compliance with the Building Regulations which place constraints on the energy consumption of new buildings. DTM analysis means that potential problems can be identified early on in the design and remedial measures taken to satisfy the building control officer of compliance.
Another important application of DTM is the EU Directive 2002/91/EC on the energy performance of buildings which came into force in January 2003. A fundamental part of this directive is an energy-labelling scheme for new buildings which will be translated into UK law from January 2006. The exact methodology for implementing these requirements is still in a state of flux, although it looks as though using a virtual prototype of the building to predict energy usage and hence rate the energy performance is a natural progression of the technology.
Computational fluid dynamics
Whereas DTM provides a holistic view of a building’s interaction with its environment, occupancy patterns and servicing strategies, computational fluid dynamics (CFD) focuses on particular issues in specific areas of the building. CFD is a simulation technique which grew up in the 1970s largely as a result of research work at Imperial College in London and the enormous growth in the aerospace and nuclear industries at the time. Similar in theory to finite element modelling for structural analysis, CFD models generally divide a three-dimensional model up into many hundreds of thousands of finite volumes of fluid. The Navier–Stokes equations describing fluid flow are then solved across the small cells of the fluid to calculate velocity components, pressure and temperature at every node point in the space. This enables a picture of the fluid flow patterns, pressure and temperature distributions to be built up and visualised by the simulation engineer.
CFD can be applied to virtually any system which involves the flow of a fluid, including heat transfer, species transport, combustion and chemical reactions and it is employed extensively in the automotive and aeronautical industries, to name just two. In the construction industry, however, CFD is a relatively recent discovery. Here, the physics of the problem are relatively simple, i.e. for most of the time the fluid is air, and low velocities mean that the air can be considered incompressible. Specifying the boundary conditions of a problem, though, is much more difficult because of the dynamic nature of the building described above. This is where the expertise and experience of the building simulation engineer are imperative: CFD tools nowadays make it relatively easy to create a very sophisticated time-dependent model of a building which could be almost impossible to solve. An in-depth knowledge of the technology as well as the individual tool’s capabilities is therefore required as well as the nature of the analysis which is to be carried out. This having been said, there is now a small but growing group of CFD expertise in the construction industry largely employed by the building services consultants to carry out detailed airflow analysis to test for thermal comfort, contaminant distribution, smoke control and all manner of other applications related to buildings.
Although recent legislation is aimed at reducing the environmental impact of buildings through energy efficient design, normally less than 2% of annual operating costs are attributable to utility bills. The lion’s share of this comes from salaries which can contribute around 75% of the annual costs. It therefore makes sense for a building owner/operators to take steps to provide a comfortable and pleasant environment, in order to maximise the working efficiency of their employees. CFD is the only way to make this comfort assessment prior to occupation. In the simplest terms this may be a prediction of localised air temperatures and air speeds which might cause draughts. On a higher level, the values calculated in a CFD analysis can be used for more objective calculations of occupant comfort such as the indices of predicted mean vote (PMV) and percentage of people dissatisfied (PPD). This is of an enormous advantage at the design stage of a building as it enables analysis results to be presented to non-engineering members of the design team such as the architect and the client. Furthermore, it enables them to relate to the performance of the proposed scheme with something immediately recognisable, i.e. a prediction of the occupants likely to be dissatisfied with a space. Although this number will never be 0% it can allow easy comparison between one design and the next in getting to the lowest possible value.
One of the main challenges for the simulation engineer is therefore not in the software capability, but more in taking raw data and summarising this into a simple assessment of the building performance through the use of the software tools at his or her disposal. This is even more important in the other areas, such as lighting simulation and people flow modelling. Lighting simulation Lighting simulation has several applications, such as assessing ‘right to light’ issues, light pollution, daylighting effectiveness and visualisation. Some of these analyses would involve communication with the lighting designer who wants information on lux levels, uniformity, glare, etc. In many cases, however, it would be the architect, client or planning authority who, in short, wants to know what the lighting scheme will look like.
The modelling of daylighting has an obvious environmental impact, with electric lighting accounting for typically 15% of annual energy use and 30% of annual energy cost in an office building. A well daylit space can reduce this by up to 90% when used with automatic lighting controls and therefore provides a very good case for considering daylighting simulation during design. Early consideration of this can impact on the glazed area, solar shading and even the conceptual shape of the building if good daylighting is a prime objective.
People flow modelling
The other area where simulation is probably more focused on the needs of the architect or space planner is in people flow modelling. These techniques were originally developed to evaluate evacuation routes and escape times for fire engineering where it can be used to compare evacuation times with the time for hazard development. Nowadays they are becoming increasingly used in predicting general people circulation. In this regard, the sorts of projects it has been applied to in the past include schools, airports, theatres and stadia.
Football grounds, for example, have many historical issues associated with the efficient movement and dispersion of crowds, particularly in the light of disasters in recent years. Although rules of thumb can be applied here, people flow modelling takes the analysis to a new level and is incredibly powerful at providing visualisation which can be easily interpreted by all members of the design team. Using people flow modelling, the width and number of stairs, corridors and exits can be optimised and comparisons easily made between different designs.
There is no doubt that with increased emphasis on our health and comfort both at work and leisure, and a more focused approach to energy efficiency and sustainability, performance-based design is the future of the construction industry. Rules of thumb are great tools for checking calculations are in the right ball park but as buildings evolve, each one becoming unique, design methods must follow suit. Building simulation bridges the gap between the architect and traditional building engineering and is a commonsense use of the technology at our fingertips to deliver the best, most efficient and suitable designs for our clients.
Simulation & Analysis Group, Buro Happold
Dr David Stribling is a consultant with Buro Happold’s Computational Simulation & Analysis (CoSA) group. He has ten years’ experience in the application of simulation techniques in the construction industry and has been closely involved in the evolution of the techniques and their integration into the design process over this period. Since joining Buro Happold he has worked on many projects including the Danish National Opera, University of Michigan Biomedical Sciences Research Building, Perth Concert Hall, BBC White City and Syddansk University.