COMPUTER SIMULATION OF FOLIAGE SHADING IN BUILDING ENERGY LOADS Marc Schiler and Donald P. Greenberg Program of Computer Graphics - Cornell University Ithaca, New York 14853 I. INTRODUCTION The calculation of building thermal loads using the computer has been an accepted practice for several years. A substantial amount of research and theoreti- cal investigation has been expended in attempts to accurately quantify a building's thermal behavior. A number of existing simulation packages acceptably model this behavior.l,2,3,~,5, 6 The advantages of simulation are obvious, not only for the predictive information with respect to operating costs or fuel consumption, but as a potential aid for preliminary design. The influence and tradeoffs of a large number of design variables such as siting, orien- tation, window area, thermal resistivity, surface/ volume ratios, and cost can all be examined at an early stage in the design process. Unfortunately, the use of these simulation packages in the design stage has been limited, primarily by the difficulty and time requirements of the numerical input tasks. In addition, most of the available pro- grams do not have the capability for including shadows in the analysis, therefore depriving the architect of a valuable strategy with respect to orientation and the use of shading devices and limiting the applicability to sites not effected by shadows created by neighbor- ing buildings. None of the existing methods includes the effects of natural shading from foliage. The latter two restrictions are particularly important since it is then difficult to correlate the results of the simula- tions with measured data. At Cornell University, an interactive computer graphics analysis system (GLAS) has been developed to simplify and accelerate the input task for thermal analysis pro- grams. 7,8 By using visual input methods, not only can a designer work with a familiar communication medium, but the time of the data input process is reduced and less error-prone. The capability of graphical descrip- tion of shading devices as well as shadow calculations has been included. 9 A recent addition has been the development of the graphic input methods required to in- clude the effects of foliage on the load determination analyses. These routines consist of the definition of the tree silhouettes, the acquisition of the time-vary- ing density information of the foliage, and the location of the trees on the site. The techniques rely heavily on interactive computer graphics and image processing methods. m This paper briefly describes the total system. A descri- ption of the organization of the analysis is presented in Section II. The interactive routines for the foli- age definition are discussed in Section III. Proceed- ures for calculating shadows in general are described in Section IV with particular emphasis on the method used for quantifying the effect of foliage. This is followed by an example depicting the effects of foliage on reducing the amount of radiation. II. DESCRIPTION OF THE GRAPHIC LOAD ANALYSIS SYSTEM A. Organization and Graphic Input The interactive computer graphics energy analysis sys- tem developed at Cornell University, 7,°,9 the Graphic Load Analysis System, GLAS, is composed of an analysis section, a weather simulation model, and a number of separate graphics programs, each completing a differ- ent input or output task for specifying the information required for a building energy analysis. The system can be described as being subdivided into four distinct sections (Figure I). The first two sec- tions provide the input information necessary for an energy analysis. The analysis and output graphing op- tions comprise the last two sections. Although each section must be completed sequentially before starting the~n~xt, the sections themselves contain a number of prograhs which may be used in any order, eliminating the rigidity of a single path process. Each graphic task in the GLAS system permits the user to specify quickly the necessary analysis information through simple visual commands. The function of these tasks is to act as an interpreter in turning symbolic commands into the actual numerical information neces- sary for a building evaluation. The system has proved to be efficient in its operation and to be extremely easy to use. icl llci i . . . . .... of the of the Building Building C ~ o ~ t 1-y Geo~try "PRIMITIVES" Alternate Program 1 __k , Plotting Package Daily O~tput negy II r .... A~alysis Model $ $ Package Hour Output Figure i. GLAS Program Diagram 142