1、2008 ASHRAE 431ABSTRACT The sensations of thermal discomfort in the near-windowregions of rooms may be significant. Close to windows occu-pants may be directly exposed to both transmitted solar irra-diation and enhanced long wave radiation exchange due towindow surfaces that are noticeably hotter or
2、 colder than otherroom surfaces. The superior insulating qualities of modernhigh performance glazing systems result in relatively highersurface temperatures in wintertime. This may reduce the localdiscomfort experienced by occupants and increase the utilityof glazed perimeter spaces. In evaluating g
3、lazing systems onewould like to quantify such benefits. Prediction of comfort perception in this asymmetric radi-ant environment is challenging. Being able to account forlocal, and not just overall, sensations of discomfort is partic-ularly important. In this work a multi-segment dynamiccomfort mode
4、l has been employed that incorporates recentlydeveloped models of local thermal comfort response. The workrequired the development of simulation methods able to predictthe detailed long-wave and convective exchanges to thesurrounding space and the absorbed solar irradiation. Thishas been done in an
5、efficient and generic manner so that para-metric studies of local comfort responses have been possible.Such studies have been used to examine the relationshipsbetween local discomfort and room and window temperaturesas well as the role solar irradiation and clothing may play indetermining satisfacto
6、ry winter environmental conditions.INTRODUCTIONIn real buildings, there are a number of reasons why occu-pants find themselves seated, or standing, close to windows.Spatial planning may oblige some occupants must have desksnear windows, or it may be that occupants choose to be neara window to gain t
7、he benefits of daylight or external view. Thethermal conditions to which such occupants are exposed arecomplex asymmetrical and highly dynamic. They can bevery different from central room positions because occupantsmay be directly exposed to both transmitted solar irradiationand enhanced long wave r
8、adiation exchange. As a result occu-pants in near-window regions may have quite different percep-tions of the thermal comfort of the space than other occupants.Long-wave radiation can be enhanced due to windowsurfaces that are noticeably hotter or colder than other roomsurfaces. At the same time, so
9、lar irradiation of the bodysurface can be a significantly larger than both convection andlong-wave radiation. These radiant conditions are highlyasymmetric and dependent on body position, posture andorientation; some parts of the body may be exposed to largeheat fluxes while others may be completely
10、 shielded. The radi-ant fields are also highly dynamic as solar irradiation can varyrapidly and by up to two orders of magnitude. At certain times,the effects of solar irradiation may provide some compensa-tion for cold glass temperatures. At other times solar irradia-tion may serve to raise glass s
11、urface temperatures higher thanthe room air temperature. These effects both the absolute magnitudes and thedynamics are partly under the control of the buildingdesigner; glazings are available with a wide range of insulationand solar transmission properties and the window size andshape, and the room
12、 geometry and spatial planning, can bemanipulated. However, the information available about thethermal comfort implications of these options is limited.Thermal comfort research, over several decades, hasenabled engineers to determine what room average tempera-Predicting Local Thermal DiscomfortAdjac
13、ent to GlazingSimon J. Rees, PhD Kevin J. Lomas, PhD Dusan Fiala, PhDMember ASHRAESimon J. Rees is a Research Fellow and Kevin J. Lomas is Director of the Institute of Energy and Sustainable Development, De MontfortUniversity, Leicester, United Kingdom. Dusan Fiala is Deputy Director of the Institut
14、e for Construction Economics at the University ofStuttgart, Stuggart, Germany.NY-08-0532008, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 114, Part 1. For personal use only. Additional reproduction, distrib
15、ution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.432 ASHRAE Transactionstures should be maintained to provide general overall thermalcomfort, and research into comfort under asymmetric radiantconditions has provided guidance on the limi
16、ts of asymmetrythat should be allowed. There is however, little guidance relat-ing to the complex conditions near windows. This is largelybecause the prediction of human comfort responses nearwindows is complicated by the radiant asymmetry, whichmeans that the comfort impacts are very localized: the
17、rmalconditions vary between one body part and another (adjacentone); and the perceived impact on comfort differs because(adjacent) body parts can have different sensitivities. Onlywith the recent development of detailed multi-segmentalmodels has local thermal comfort prediction become possible.A det
18、ailed understanding of, and model of, the complexconditions near windows would help to address a number ofquestions, such as:Can uncomfortable conditions be avoided by carefulselection of the glazing system?Are there window geometries and shading details thatcan be adopted to minimize discomfort and
19、 maximizethe utility of near window spaces?Can suitable system controls be devised to provide opti-mal local and overall comfort when solar fluxes varyrapidly?What forms of adaptation will best help occupantsmaintain comfort shading, orientation, clothing etc.Can simple window performance metrics be
20、 devisedthat would help designers and building owners appreci-ate the comfort benefits of certain window designs?The traditional mathematical models of human thermalcomfort, such as those of Fanger (1973) and Gagge (1986) arefounded on calorimetric principals but are only responsive tothe bodys over
21、all thermal state. A modeling approach that canaddress questions of local discomfort near glazing must incor-porate the following:1. A thermo-physiological model that is able to predict thethermal conditions at individual body parts.2. A model of local thermal comfort response.The first requirement
22、can be met by using a multi-segmental model such as the Berkeley Comfort Model (Huiz-enga et al., 2001) or the IESD-Fiala Model (Fiala et al., 1999).For this work, the IESD-Fiala model, which is familiar to theresearchers, was used. The model uses a detailed representa-tion of the human body and has
23、 the capability to reliablypredict both the overall and local temperature responses andregulatory behaviors for a wide range of environmental condi-tions (Fiala et al., 1999, 2001 and 2003). The second require-ment was met by extending the model so that the localdiscomfort response could be predicte
24、d from the thermalconditions of the individual body parts; and not just the overallthermal response (global PPD). This was done by incorporat-ing into the IESD-Fiala model the local discomfort modelsdeveloped by Fiala and Kubaha (2005). Calculation of local variations in the thermal environmentrequi
25、res a modelling methodology that incorporates:a detailed representation of body geometry preferably indifferent postures;a means of accurately calculating longwave radiationbetween each surrounding surface and each body part;a means of accurately calculating the shortwave irradia-tion on each body p
26、art for a particular window geome-try, orientation and solar position; Accordingly, a detailed representation of an upright male(consisting if some 10,000 polygons) has been used to deriveview factors for long-wave radiant exchange calculations. Aradiosity and simplified ray tracing method was devel
27、oped tocalculate the incident diffuse and direct short-wave radiation.The reported work was motivated by an interest in thecomfort benefits of high performance glazing. It has however,led on to generic parametric studies and analysis of the rela-tionship between local comfort and annual energy deman
28、ds.The results of these studies could also be used to examine therelationship between window comfort performance and otherparameters such as U-value.MODELING LOCAL THERMAL DISCOMFORTTo be able to examine local discomfort in a range of envi-ronmental conditions any thermoregulatory model of thehuman
29、body must be sufficiently discretized to allow thecondition of particular body parts to be established not justthe overall thermal condition of the body. Multi-segmentedthermophysiological models, of different levels of sophistica-tion, have been developed over a number of years (e.g. Stol-wijk 1971
30、, Konz et al., 1977, and Wissler 1985). They allowthe overall thermal condition of the body to be calculated andhave been useful in providing insight into the physiologicalprincipals of thermal comfort. More recent models (e.g.Huzenga et al., 2001 and Fiala et al., 1999, 2001 and 2003)allow dynamic
31、responses to complex heterogeneous environ-mental boundary conditions to be examined in considerabledetail. The model used in this work was based on the IESD-Fiala model of human physiology developed at De MontfortUniversity and the Egle-Institut at the University of AppliedSciences, Stuttgart. The
32、model has been described in detailelsewhere and so only a brief description follows.The Thermo-Physiological ModelThe IESD-Fiala model uses a multi-segmental, multi-layered representation of the body with a detailed representa-tion of its geometry. The overall geometry represent an aver-age person o
33、f weight 73.5 kg, body fat content 14% andDubois-area 1.83m2. The body is divided into 19 elementsconsisting of multiple tissue layers. Each element of the bodyis subdivided into multiple sectors, e.g. the upper leg is dividedASHRAE Transactions 433into interior, anterior, exterior and posterior sec
34、tors and thebody as a whole has 59 such sectors (Figure 1). The model of the human thermoregulation that is associ-ated with the geometry, can be thought of as having two inter-acting systems: the controlling active system and thecontrolled passive system. The active system is simulated bymeans of c
35、ybernetic models that predict responses such asshivering, vasomotion and sweating (see, for example, Fiala etal. 2001). The passive system, which is of particular interesthere, is constructed on the basis of the dynamic heat balancesat each tissue element to account for: conduction of heatthrough th
36、e body sector; conduction to adjacent sectors; trans-port of heat by the blood stream; and metabolic heat produc-tion. Heat balances, and heat transfer processes, establishedfor each body sector are coupled (from a thermal or mathe-matical point of view) by a model of the blood circulatorysystem. He
37、at rejection via respiration is also considered.Conduction heat transfer at each body sector is modeled usinga one-dimensional finite-difference method and the wholesystem of equations is discretized in time using a Crank-Nich-olson approach. Solution of the model equations providespredictions of bo
38、dy core and surface temperatures, heat fluxes,evaporation rates and wetted areas and other secondary quan-tities.When considering asymmetric long-wave and short-waveradiant conditions as we must when modeling near windowconditions the treatment of the body surface heat balance isof particular intere
39、st. In the IESD-Fiala model the surface heatbalance at each body sector is formulated as,(1)Where qsk= the net heat loss from the skin surface (W/m2)qc= the heat loss by convection to the air (W/m2)qr= the longwave radiation loss to the surrounding surfaces (W/m2)qrs= the absorption of direct and di
40、ffuse solar irradiation (W/m2)qe= the latent heat loss from the skin due to moisture evaporation (W/m2)Kubaha (2005) extended the original IESD-Fiala model sothat, rather than using a linear radiant heat transfer coefficient,the following formulation was used,(2)Where= the Stefan-Bolzmann constant (
41、5.67 x 10-8W/m2.K4) = the emissivity of the body sector, (-) = the emissivity of the surrounding surface (-)Fi,j= the view factor of the body sector with respect to the surrounding surface (-)Tb,i= the absolute temperature of body sector (K) Ts,j= the absolute temperature of surrounding surface sect
42、or (K)i = the body sector index (1 to 59)j = the surrounding surface index (n in total) In this study the representation of the body geometry,surrounding surfaces and associated view factors was treatedin some detail and is discussed in a later section.In many models of body heat transfer, shortwave
43、 radiantfluxes are approximated by multiplying the flux by aprojected area factor. In other words, shading and reflectionfrom surrounding surfaces is not treated explicitly. In thiswork the model has been adapted in order to allow a complexrepresentation of shortwave radiation a each body sector. Th
44、isis described in detail in a later section.A MODEL OF LOCALCOLD AND WARM DISCOMFORTUnderstanding and modelling human perceptualresponses to asymmetric radiation has been the subject ofexperimental investigation over a number of decades (e.g.Chrenko 1953, McNall and Biddison 1970, Olesen et al.,1972
45、, Fanger et al., 1985, Zhang 2003). Most experimentalwork has been carried out in climate chambers under well-controlled conditions for vertical and horizontal surfaces. Cold and warm cutaneous thermal receptors are distrib-uted in a heterogeneous manner over all parts of the bodysurface. Consequent
46、ly, skin surface temperature response hasbeen found to correlate well with perceived global or whole-body thermal comfort (Gagge et al., 1967, Gonzales et al.,1973). Furthermore it has been found that perceived localdiscomfort can be similarly correlated with, and predictedusing, local skin temperat
47、ures (Issing and Hensel 1982). Thisis the conceptual basis of the model of local cold and warmdiscomfort developed by Kubaha (2005) and is the platformfor the extension of the IESD-Fiala model used in this work. The basic approach, for each body part, is to use the differ-ence between the local skin
48、 temperature and a reference value,as a measure of the thermal stimulus and then to correlate thisdifference with the responses of subjects tested under care-fully controlled experimental conditions. The basic temperature difference is given by, (3)whereTsk,i= the local skin temperature (K)Tsk,ref=
49、a reference skin temperature (K)= negative values of indicate local cold stimulus and positive values of indicate local warm stimulus (K)However, not all body parts are equally sensitive to skintemperature stimuli, so a sensitivity parameter must be intro-duced. Furthermore, the choice of reference temperature, andthe values of the sensitivity coefficients, varies depending onqskqcqeqrqsr+=qri,ijFij,Tbi,4Tsj,4()j 1=n=ijTsk i,Tsk i,Tsk ref,=Tsk i,434 ASHRAE Transactionswhether local cold stimulus (LCS) or local warm stimulus(LWS) is being considered.In modelling
