USF Libraries
USF Digital Collections

Numerical simulation of thermal comfort and contaminant transport in air conditioned rooms

MISSING IMAGE

Material Information

Title:
Numerical simulation of thermal comfort and contaminant transport in air conditioned rooms
Physical Description:
Book
Language:
English
Creator:
Ho, Son Hong
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
heat and mass transfer
ventilation
computational fluid dynamics
relative humidity
multi-component flow
Dissertations, Academic -- Mechanical Engineering -- Masters -- USF
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Health care facilities, offices, as well as workshops and other commercial occupancies, require ventilation and air conditioning for thermal comfort and removal of contaminants and other pollutions. A good design of ventilation and air conditioning provides a healthy and comfortable environment for patients, workers, and visitors. The increasing developments of computational fluid dynamics (CFD) in the recent years have opened the possibilities of low-cost yet effective method for improving HVAC systems in design phase, with less experiment required. This work presents numerical simulations of thermal comfort and contaminant removal for two typical working spaces where these factors are critical: a hospital operating room with various configurations of inlet and outlet arrangements, and an office with two cases of air distribution systems: underfloor and overhead, also with alternative cases. The 2-D simulation approach was employed.Temperature, relative humidity, contaminant concentration, thermal sensation, predicted mean vote (PMV), and contaminant removal factor were computed and used for assessing thermal comfort and contaminant removal characteristics of the office room and operating room. The result shows good agreements with experimental data taken from related literature.
Thesis:
Thesis (M.S.M.E.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Son Hong Ho.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 108 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001498271
oclc - 57717801
notis - AJU6876
usfldc doi - E14-SFE0000548
usfldc handle - e14.548
System ID:
SFS0025239:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Numerical Simulation of Thermal Comfort and Contaminant Transport in Air Conditioned Rooms by Son Hong Ho A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Mechanical Engineering Department of Mechanical Engineering College of Engineering University of South Florida Major Professor: Muhammad Rahman, Ph.D. Ashok Kumar, Ph.D. Thomas Eason, Ph.D. Date of Approval: November 8, 2004 Keywords: computational fluid dynamics, multi-component flow, relative humidity, ventilation, heat and mass transfer Copyright 2004, Son Hong Ho

PAGE 2

i Table of Contents List of Tables iii List of Figures iv Abstract vii Chapter 1: Introduction 1 1.1 Overview on Simulation of Air Conditioning in Office Buildings 1 1.2 Overview on Simulation of Air Cond itioning in Hospital Operating Rooms 3 1.3 Thesis Outlines 4 1.4 Nomenclature 5 Chapter 2: Simulation Approach 7 2.1 Introduction 7 2.2 Governing Equations 8 2.3 Relative Humidity 10 2.4 Thermal Comfort Assessment 11 2.5 Contaminant Removal Effectiveness 12 2.6 Computation Procedures 13 Chapter 3: Simulation of Underfloor a nd Overhead Air Distribution Systems in an Office 15 3.1 Introduction 15 3.2 CFD Model 16 3.3 Results and Discussion 21 Chapter 4: Predictions of Thermal Comfort and Contaminant Removal in an Operating Room 46 4.1 Introduction 46 4.2 CFD Model 47 4.3 Results and Discussion 52 Chapter 5: Conclusions and Recommendations 69 5.1 Simulation of the Office Room 69 5.2 Simulation of the Operating Room 69 5.3 Recommendations 70 References 71

PAGE 3

ii Appendices 73 Appendix A: FIDAP Program for Office Room Simulation 74 Appendix B: FIDAP Program fo r Operating Room Simulation 88

PAGE 4

iii List of Tables Table 3.1 Dimension Parameters on Figures 3.1 and 3.2, meter(s) 19 Table 3.2 Inlet Boundary Conditions, Inlet Setup, and Simulation Cases 19 Table 3.3 Boundary Conditions for Office Room Simulation 21 Table 3.4 Comparison of AverageValues of Th ermal Comfort to Experimental Data for Underfloor Ai r Distribution System 37 Table 3.5 Comparison of Average Values of Thermal Comfort to Experimental Data for Overhead Air Distribution System 37 Table 3.6 Comparison of Contaminant Removal Effectiveness 40 Table 4.1 Dimension Parameters on Figure 4.1, meter(s) 49 Table 4.2 Inlet Angle, Outlet Sizes, Ou tlet Ratios, and Simulation Cases 50 Table 4.3 Boundary Conditions for Operating Room Simulation 51 Table 4.4 Average Air Speed, Temperature, and Relative Humidity vs. Inlet Angle for Basic Configuration 63 Table 4.5 Average Air Speed, Temperature, and Relative Humidity vs. Outlet Ratio for Two-Exhaust Configuration 63 Table 4.6 Comparison of Average Temp erature and Relative Humidity to Experimental Data 67

PAGE 5

iv List of Figures Figure 3.1 Model of Office Cubicle with Underfloor Air Distribution System 17 Figure 3.2 Model of Office Cubicle with Overhead Air Distribution System 18 Figure 3.3 Velocity Field for Simulation 1, m/s 22 Figure 3.4 Temperature Distribution for Simulation 1, oC 23 Figure 3.5 Relative Humidity Distribution for Simulation 1 25 Figure 3.6 Contaminant Concentration Di stribution for Simulation 1, kg/kg air 26 Figure 3.7 Vertical Distribu tion of Average Air Speed for Underfloor System 27 Figure 3.8 Vertical Distribu tion of Average Temperature for Underfloor System28 Figure 3.9 Vertical Distribution of Aver age Relative Humidity for Underfloo r System 29 Figure 3.10 Vertical Distribution of Av erage Contaminant Concentrationfo r Underfloor System 30 Figure 3.11 Velocity Field for Simulation 4, m/s 30 Figure 3.12 Temperature Distribution for Simulation 4, oC 32 Figure 3.13 Relative Humidity Distribution for Simulation 4 32 Figure 3.14 Contaminant Concentration Di stribution for Simulation 4, kg/kg air 33 Figure 3.15 Vertical Distribu tion of Average Air Speed for Overhead System 34 Figure 3.16 Vertical Distribu tion of Average Temperature for Overhead System 35 Figure 3.17 Vertical Distribution of Av erage Relative Humidity for Overhea d System 36

PAGE 6

v Figure 3.18 Vertical Distribution of Av erage Contaminant Concentrationfo r Overhead System 36 Figure 3.19 Thermal Comfort Factors vs. In let Location for Underfloor System 38 Figure 3.20 Thermal Comfort Factors vs. Inlet Angle for Overhead System 39 Figure 3.21 Vertical Distribution of Aver age Air Speed, Underfloor System vs. Overhead System 41 Figure 3.22 Vertical Distribution of Averag e Temperature, Underfloor System vs. Overhead System 42 Figure 3.23 Vertical Distribution of Av erage Relative Humidity, Underfloo r System vs. Overhead System 43 Figure 3.24 Vertical Distribution of Average Contaminant Concentration, Underfloor System vs. Overhead System 43 Figure 3.25 Distribution of Vertical Ve locity along Outlet Length, Underfloo r System vs. Overhead System 44 Figure 3.26 Distribution of Contaminan t Concentration along Outlet Length, Underfloor vs. Overhead System 45 Figure 4.1 Simplified Typical Operating Room 47 Figure 4.2 Model of Operating Room 48 Figure 4.3 Velocity Field for Simulation 1, m/s 53 Figure 4.4 Temperature Distribution for Simulation 1, oC 55 Figure 4.5 Relative Humidity Distribution for Simulation 1 56 Figure 4.6 Contaminant Concentration Di stribution for Simulation 1, kg/kg air 57 Figure 4.7 Contaminant Concentration Di stribution for Simulation 5, kg/kg air 58 Figure 4.8 Velocity Field for Simulation 5, m/s 58 Figure 4.9 Contaminant Concentration Dist ribution for Simulation 10, kg/kg air 59 Figure 4.10 Average Contaminant Concen tration vs. Inlet Angle for Basic Configuration 60

PAGE 7

vi Figure 4.11 Contaminant Removal Effectiveness vs. Inlet Angle for Basic Configuration 61 Figure 4.12 Average Contaminant Concen tration vs. Outlet Ratio for TwoExhaust Configuration 61 Figure 4.13 Contaminant Removal Effec tiveness vs. Outlet Ratio for TwoExhaust Configuration 62 Figure 4.14 Thermal Sensation Index vs. In let Angle for Basic Configuration 65 Figure 4.15 Predicted Mean Vote vs. In let Angle for Basic Configuration 65 Figure 4.16 Thermal Sensation Index vs. Outlet Ratio for Two-Exhaust Configuration 66 Figure 4.17 Predicted Mean Vote vs. Outlet Ratio for Two-Exhaust Configuration66

PAGE 8

vii Numerical Simulation of Thermal Comfort and Contaminant Transport in Air Conditioned Rooms Son H. Ho ABSTRACT Health care facilities, offices, as well as workshops and other commercial occupancies, require ventilation and air c onditioning for thermal comfort and removal of contaminants and other pollutions. A good design of ventilation and air conditioning provides a healthy and comfor table environment for patient s, workers, and visitors. The increasing developments of com putational fluid dynam ics (CFD) in the recent years have opened the possibilities of low-cost yet effective method for improving HVAC systems in design phase, with less e xperiment required. This work presents numerical simulations of thermal comfort and contaminant removal for two typical working spaces where these factors are critic al: a hospital operating room with various configurations of inlet and outlet arrangements, and an office with two cases of air distribution systems: underfloor and overh ead, also with alternative cases. The 2-D simulation approach was employed. Temp erature, relative humidity, contaminant concentration, thermal sensation, predicted mean vote (PMV), and contaminant removal factor was computed and used for assessi ng thermal comfort and contaminant removal characteristics of the office room and ope rating room. The result shows good agreements with experimental data take from related literature.

PAGE 9

1 Chapter 1 Introduction 1.1 Overview on Simulation of Air Conditioning in Office Buildings Within the last few years, underfloor air distribution (UFAD) systems have become popular design alternatives to c onventional air distribut ion (CAD) such as overhead air distribution systems for thermal a nd ventilation control [1, 2]. Underfloor air distribution is of increasing interest to those who own or design office buildings. Some industry-watchers predict that as many as 35 percent of future office buildings will include UFAD systems [3]. In comparison to cl assic overhead systems that deliver air at low velocities, typical UFAD systems deliver air through floor diffusers with higher supply air velocities [2]. The UFAD systems can have significant impacts on room air stratification and thermal comfort in occupied zone. Halza [4] introduced the advantages of UFAD system: improved air quality, lower life-cycle costs, as well as overhead system: better comfort, lower capital cost. Woods [1] did a review by literature searching and field investigations to assess the actual performance of UFAD system in real world. He showed that there are gaps in available data: valid and reliable field data are not fr om a sufficient population of existing facilities to conclude that underfloor system’s perfor mance is superior to overhead system; and that designers must be made aware that unde rfloor as well as overhead system requires

PAGE 10

2 more care in design, installation, and operatio ns. He also recommended that objective analysis should be made before choosing an HVAC system. Webster et al. [2] presented a series of full-scale laboratory experiments to determine room air stratification for a variety of design and operating parameters. Fuka o et al. [5] carried out comparative field measurements for both systems in an actual la rge-scale office buildi ng. Webster et al. [6] presented a study about a building that opera ted with an UFAD system. They showed little troubleshooting with the system operati on, pointing out the positive aspects of using well-designed UFAD systems. Bauman [7] of fered a work presenting a discussion about several advantages shown by the UFAD systems. In the design stage, CFD simulation can play an important role in improving the understanding of any particular system. The increasing developments of computa tional fluids dynamics (CFD) in recent years have opened the possibilities of lo w-cost yet effective method for improving HVAC system in design phase, with less ex periment required. One advantage of CFD modeling is that it allows specific entry details of a room that have relevant airflow. CFD models have been used to study indoor air quality (IAQ) problems, pollutant distributions, and performan ce of HVAC systems (Chow a nd Fung [8], Emmerich [9], Gadgil et al. [10]). Hirnikel et al. [11] investigated contam inant removal effectiveness of three air distribution systems for a bar/rest aurant by using CFD modeling. They showed that directional airflow systems could re duce people’s exposure to contaminants. Thermal comfort can be predicted base d on Fanger’s PMV model [12], which assumes a uniform thermal environment. Thermal sensation index from Rohles and Nevins’ work [13] is also widely used for assessing thermal comfort. Relative humidity can be computed by using the pr ocedure recommended in [14].

PAGE 11

3 1.2 Overview on Simulation of Air Cond itioning in Hospital Operating Rooms The main purpose of the HVAC system de sign for operating rooms is to prevent the risk of infections duri ng surgical operations while ma intaining adequate comfort conditions for the patient and surgical sta ff. There are standards suggested for airconditioning systems for operating rooms ar ound the world. The American Institute of Architects (AIA) has guidelines for designi ng and construction of hospitals and health care facilities in the USA. The institute has presented its latest revision of its guidelines in 2001 [16]. Proper indoor comfort conditions a nd indoor air quality are prerequisites for securing a safe and suitable environment for operating rooms. Many experimental studies have been presented about infections and rela ted factors in a typical operating room [17, 18]. Lewis [19] studied the influence of room air distribution on in fection rate in an operating room. He concluded that optimal air distribution played an important role in environmental conditions within a surgical room. Memarzadeh [20] proposed a methodology fo r minimizing contamination risk in hospital rooms. Mora et al. [21] studied th ermal comfort in operating rooms. They based their analysis on the thermal comfort model proposed by Fanger [12]. They concluded that the only means to provide thermal comfort for the surgical staff was to eliminate or to minimize the heat transfer from the surgical lights. They realized that more research is needed to evaluate an acceptable thermal environment in operating rooms. It can be observed that there is a need to predict ambient conditions within an operating room. Numerical analysis is usually employed for simulating airflow and temperature distribution. Memarzadeh and Manning [22] studied the performance of a ventilation system in a typical patient room using CF D modeling. They were able to predict the

PAGE 12

4 necessity of using baseboard he ating in extreme weather conditions. Hirnikel et al. [11] investigated contaminant removal effectiven ess of three air distribution systems for a bar/restaurant by using CFD modeling. They s howed that directional airflow systems can reduce people’s exposure to contaminants. Memarzadeh and Manning [23] simulated contaminant deposition on an operating room using CFD air flow modeling. They showed that laminar flow conditions were th e best choice for ven tilation systems when contaminant deposition was considered. Health care facilities, as well as workshops and other commercial occupancies, require ventilation and air conditioning for th ermal comfort and removal of contaminants as well as other pollutions. A good design of ventilation an d air conditioning provides a healthy and comfortable environment for people such as patients, workers, and visitors. Poorly ventilated workspaces not only make pe ople feel uncomfortable but also can make them become infected or intoxicated since the likelihood of air borne pathogens or other kinds of toxic chemicals are quite high. 1.3 Thesis Outlines This thesis presents the CFD simulation of two problems of air-conditioned rooms. Chapter 2 reviewed the relevant details of simulation approac h. In Chapter 3, two different air distribution syst ems for office buildings were compared on thermal comfort and contaminant removal effectiveness. Each system has its own variation, such as inlet location for underfloor system and inlet a ngle for overhead system. In Chapter 4, an operating room in hospital was modeled, di fferent cases of inlet angle and outlet arrangement were investigated. The simulation results were compared for assessing

PAGE 13

5 thermal comfort and contaminant removal char acteristics. The CFD computations for all simulations were done on FIDAP (Fluent, Inc.), a finite element analysis CFD software package. The post-processing computations we re done on Matlab (The MathWorks, Inc.). The results from the simulations were also compared to experiment al data on operating rooms and office rooms, taken from literature. 1.4 Nomenclature C Mean contaminant concentration, kg of contaminant/kg of air mixture cp Specific heat of air, J/(kg.K) D Mass diffusivity of species in air, m2/s fcl Ratio of clothed surface area to nude surface area Gr Grashof number g Gravity acceleration, m/s2 h Heat transfer coefficient, W/(m2.K) I Thermal resistance, m2K/W k Thermal conductivity of air, W/(m.K) L Characteristic length m Concentration of species, kg of species/kg of air mixture M Metabolic heat generation flux, W/m2 of naked body area p Pressure; partial pressu re (with subscript), Pa Re Reynolds number T Temperature; mean temperature (with subscript), C U Characteristic velocity, m/s

PAGE 14

6 u velocity, m/s v Mean air speed rela tive to the body, m/s W External work, W/m2 of naked body area Y Thermal sensation index Greek Symbols Thermal expansion coefficient, 1/K Relative humidity Viscosity of air, kg/(m.s) Density of air, kg/m3 Subscripts 1 Water vapor 2 Contaminant a Air BZ Breathing zone c Convective cl Clothing E Exhaust r Radiant ref Reference S Supply s Saturated (water vapor) w Water vapor

PAGE 15

7 Chapter 2 Simulation Approach 2.1 Introduction To predict the indoor thermal environmen t, it is necessary to determine air velocity, temperature, and relative humidity in a room. The prediction was carried out by solving coupled equations for the conservation of mass (for the whole air mixture as well as for each species), momentum, and energy. Fo r most air conditioning applications in indoor environment assessing and designing, the solution of interest is steady state. Since the real problems are three-dimens ional (3-D) by nature, using 3-D models to simulate them would be the best approach. However, 3-D simulations require very large amount of computation memory and time, sometimes possibly exceed the available resources. Besides, from the design point of vi ew, it can be very difficult to locate and to assess the key parameters, which most signifi cantly affect the performance of a design, from a 3-D simulation where the interaction of space dimensions co mplicates the results. Two-dimensional (2-D) simulation requires le ss computation resources, but still can provide reasonable results on what parameters are important and how they affect the performance of a design. It de scribes the phenomenon of fluid flow and heat transfer in the local section of interest (e.g. near work ing people) but not for the entire region. As a basic approach for the problems at ha nd, this work employed 2-D simulations.

PAGE 16

8 The fluid properties are assumed constant s. They were taken at the reference temperature, Tref = 22oC = 295 K, as follows: = 1.1967 kg/m3 = 1.8273E-5 kg/(m.s) cp = 1.0043E3 J/(kg.K) k = 2.5776E-2 W/(m.K) = 3.3932E-3 K-1 D1 = 2.5448E-5 m2/s D2 = 2.5033E-5 m2/s 2.2 Governing Equations Consider a steady state, two-dimensional incompressible flow of air as a multicomponent fluid, which includes dry air, wa ter vapor, and contaminant gas. The fluid properties are considered as constants excep t the varying density for buoyancy term in the momentum equation. The equation for the conservation of ma ss applied for the air mixture as a whole or carrying fluid is given by 0 u (2.1) Assuming that the mass diffusivities of speci es in air are scalars, thermal diffusion (Soret effect) is negligible, and there is neither source nor chemical reaction, the equations for the mass conservation of wate r vapor and contaminant gas as carried species are 1 2 1 1D m m u (2.2)

PAGE 17

92 2 2 2D m m u (2.3) The buoyancy force term arising from dens ity variation is included by means of the Boussinesq approximation based on the as sumptions that variation in fluid density affect only the buoyancy term and fluid de nsity is a function of temperature and concentration only. For most HVAC applicatio ns, the species concentrations are very small such that the dependency of buoyancy term on them can be neglected. The equation for the conservation of linear momentum is given by refT T p g u u u2 (2.4) Assuming that there is no h eat generation, thermal conduc tivity is scalar, energy flux due to inter-diffusion and Dufour effect are negligible, and the equation for the conservation of energy is given by T k T cp 2 u (2.5) The equation Eq. 2.4 is a vector equation for velocity (and pressure). It is actually two coupled scalar equations of two velocity components (and pressure). This equation describes the mixed convection fluid flow, th at is both forced convection and natural convection exist. The last term in the right hand side of Eq. 2.4 is the buoyancy term, which represents the effect of natural convection. The buoyancy term couples the equations Eq. 2.4 and 2.5 through temperature va riable. If the buoyancy term is small, it can be discarded and thus decoupling the e quations except for the convection terms. For judging if the buoyancy effect is small en ough to be eliminated without causing significant errors, Reynolds number and Gras hof number are used to characterize the effect of forced convection and natu ral convection, respectively, given by

PAGE 18

10 UL Re 2 3 2Gr L T T gref where U and L are the characteristic velocity an d characteristic length, respectively. If Gr > Re2 then natural convecti on dominates. If Gr < Re2 then forced convection dominates. If Gr << Re2 then the effect of natural conve ction is very small and the flow can be considered to be forced convection only. The typical values of the dimensionle ss numbers for the air-conditioned rooms are Re ~ 104 and Gr ~ 109. Then Gr ~ Re2, the effects of forced convection and natural convection are generally the same; the flow is actually mixed convection and a little natural convection dominated. Therefore, the buoyancy term has a very strong effect on the solution and cannot be eliminated. The solution obtained from solving the equations Eq. 2.1 – 2.5, associated with their boundary conditio ns, gives six primary paramete rs: two velocity components, pressure, temperature, water vapor con centration, and contaminant concentration. 2.3 Relative Humidity From the primary parameters: temper ature, water vapor concentration, and pressure, relative humidity can be computed by using the procedure recommended in [14], which is summarized as follows: ws wp p (2.6) where

PAGE 19

11 1 137802 0 62198 0 101325 m m p pw (2.7) 15 273 ln 546 6 15 273 10 445 1 15 273 10 176 4 15 273 10 864 4 516 5 15 273 10 800 5 exp 10003 8 2 5 2 3 T T T T T pws (2.8) 2.4 Thermal Comfort Assessment One of the most frequently cited thermal comfort models is the Fanger model. The Fanger model is based on steady-state en ergy balance. This model was originally developed to predict human thermal comfort in office-like environments and has gained wide usage in the HVAC industry because it s simplicity [15]. Predicted mean vote (PMV) is a parameter for assessing thermal co mfort in an occupied zone based on the conditions of metabolic rate, clothing, air sp eed besides temperatur e and humidity. From the work of Fanger given in [12], the value of PMV is given by a cl c cl a cl cl a w wT T h f T T f T M p M W M p W M W M W M 4 4 8 5 3273 273 10 96 3 34 0014 0 5867 10 7 1 15 58 42 0 99 6 5733 10 05 3 028 0 036 0 exp 303 0 PMV (2.9) where a cl c cl a cl cl cl clT T h f T T f I W M T 4 4 8273 273 10 96 3 028 0 7 35 (2.10) greater is whichever v 1 12 or 38 25 0 25 0 c a cl ch T T h (2.11) K/W m 078 0 for 645 0 05 1 K/W m 078 0 for 29 1 00 12 2cl cl cl cl clI I . I I f (2.12)

PAGE 20

12 Thermal sensation index represents th e effect of enviro nmental and personal variables on thermal response a nd comfort level, such as temp erature, humidity, sex, and length of exposure. Thermal sensation can be predicted using empirical equations from the work of Rohles and Nevins given in [ 13]. The empirical equation for men and women combined with exposure period of 3 hours, conversed for SI units, is given by 802 6 000278 0 243 0 w ap T Y (2.13) Thermal sensation index values refer to the thermal sensation scale adopted by ASHRAE now known as the ASHRAE thermal sens ation scale. PMV valu es also refer to this scale. ASHRAE thermal sensation scale ranges from -3 to 3 as follows: 3 = hot 2 = warm 1 = slightly warm 0 = neutral -1 = slightly cool -2 = cool -3 = cold 2.5 Contaminant Removal Effectiveness For assessing the effectiveness of an occupied zone, the contaminant removal effectiveness (CRE) is used. The CRE was determined based on the mean contaminant concentration in the supply inlet, in the e xhaust outlet, and in the occupied zone [9]. S BZ S EC C C C CRE (2.14)

PAGE 21

13 where CE is the mean concentration in exhaust; CS is the mean concentration in supply air and CBZ is the mean concentration in occupied zone. Assuming that the supply airflow is contaminant-free, the contaminant removal eff ectiveness from (2.14) can be computed as BZ EC C CRE (2.15) 2.6 Computation Procedures The simulations were done on CFD software package FIDAP (Fluent, Inc.). For each simulation, two steps were performed. Fi rst, the strongly coupled problem of the equations Eq. 2.1, 2.4, and 2.5 was solved. Then the advection-diffusion problem of the species equations, Eq. 2.2 and 2.3, was solved with known velocity field from the first step. Source code for the typi cal cases of the two simulation sets in Chapters 3 and 4 are given in Appendices A and B. The output numerical solution includes velocity component, pressure, temperature, water vapor concentration, and contaminant concentration at every node of the computation region. In post-processing stage, available commands in FIPOST, the post-processing module in FIDAP package, were used when possible. The average values of speed, temperature, and contaminant concentration were computed directly by using the MEAN command, which is a weighted average based on the size of the elements. Similarly, the average contaminant on the outlet was computed by using the FLUX command. The relative humidity, which depends on temperature, pressure, and water vapor concentration, can be computed by two meth ods: (i) using user subroutines written by user and incorporated into FIDAP, and (ii) using Matlab. The first method is very

PAGE 22

14 convenient for getting the rela tive humidity directly from FIDAP as a primary variable, but require the to include the species equation fo r water vapor in the first step, i.e. solving the equations Eq. 2.1, 2.4, 2.5, and 2.2 simultaneously, which costs more computation resources. The second method, using Matlab gives a more flexible alternative. The FIDAP numerical solution was expo rted into neutral files, then read into Matlab. Several Matlab M-file were created to compute relati ve humidity at every nodes and mean values, as well as other relevant parameters such as thermal sensation index, PMV, CRE. Matlab also handles the 2-D contour and vector plots the variables of interest (velocity, temperature, relative humidity, and contaminant concentration).

PAGE 23

15 Chapter 3 Simulation of Underfloor and Overhead Air Distribution Systems in an Office 3.1 Introduction This study compares thermal environments and contaminant removal effectiveness (CRE) of two air distribution systems for an office setting by the use of computational fluid dynamics (CFD) modeling. The air supply dist ribution and exhaust arrangement were modeled for an underfloor air distribution (UFAD) system and an overhead air distribution system The study of a thermally co mfortable typical cubicle in a large office floor requires detailed inform ation about distribution of air velocity, air temperature and relative humidity in the indoor environment. The CRE of each system was determined for contaminant distributions The model included a typical cubicle in a large office floor in a steady-state condition wi th a chair, a desk w ith a PC on top, and heat sources such as seated people and light s. For underfloor air distribution system, air entered the occupied zone th rough an inlet located at floor level supplying a vertical upward inflow. Three different locations of inlet diffuser were considered. For overhead air distribution, the inlet is located on the ceiling with slower and cooler inflow. Three cases of inlet angle were considered. For both systems, the air return location is on the ceiling at the same place. Distributions of velocity, temperature, relative humidity, and contaminant concentration in various cases for both systems were computed. Thermal

PAGE 24

16 comfort factors and contaminant removal effectiveness were assessed for the two systems. The results were compared among cas es of each system, as well as between two typical cases of the two systems. A comparison with experimental data of an actual office building given in literature was also offered. The objective of this part of the work is to use CFD modeling to simulate airflow in two air distribution system s: underfloor and overhead, for a single cubicle on an office floor. The results can be related to thermal e nvironment, indoor air quality and ventilation effectiveness. Temperature and relative humid ity distributions as well as contaminant concentration and velocity patt erns are to be presented. Thermal comfort is predicted based on Fanger’s PMV model [12], which a ssumes a uniform thermal environment. Thermal sensation index from Rohles and Nevi ns’ work [13] is also used for assessing the thermal comfort of the cubicle. The results are to be compared to each other. CFD prediction results are also compared to experimental results reported in [5]. 3.2 CFD Model A cubicle in a large office floor was m odeled as a rectangu lar region. Two air distribution systems were considered in the present investigation: underfloor air distribution (Fig. 3.1) and ove rhead air distribution (Fig. 3. 2). These two figures show a typical set up and esse ntial dimensions of the cubicle. The essential dimensions are denoted in general forms as L1 to L16 fo r lengths and A1 for angle. Giving the dimensions in general form makes it flexible for further parameterized investigation, as the alternation of essential dimensions can be performed without significant changes in the CFD program. The numerical values of the lengths L1 to L14 used for the

PAGE 25

17 computations in this paper are given in Ta ble 3.1 for both systems. The inlet length denoted as L15 is different between two systems (underfloor and overhead) and takes corresponding numerical values from Table 3.2. For underfloor system (Fig. 3.1), various locations of the inlet diffuser can be considered by altering the length L16. In this paper, three such locations were taken for computation: close to the backside of the s eat (typical), under the desk, and facing the outlet. The values of L16 as the inlet lo cation parameter are given in Table 3.2. LIGHTSOUTLET(RETURN) MONITOR CPU DESK CHAIRPERSONINLET L8 L9 L5 L6 L4 L3 L2 L1 L11 PANEL SEPARATOR L13 L12 L10 L7 FLOOR CEILINGSYMMETRY SYMMETRY L14 L15 L16 Figure 3.1 Model of Office Cubicle with Underfloor Air Distribution System For overhead system (Fig. 3.2), although th e location of the inlet diffuser remains unchanged, various inlet angles can be c onsidered by setting the angle A1. The inlet angle is measured downward from the ceiling. For this paper, three inlet angles were

PAGE 26

18 taken: 30 (typical), 45, and 60 degrees. The numerical values A1 corresponding to each simulation case are given in Table 3.2 as well. LIGHTS OUTLET(RETURN) MONITOR CPU DESK CHAIRPERSONINLET L8 L9 L5 L6 L4 L3 L2 L1 L11 PANEL SEPARATOR L13 L12 L10 L7 FLOORSYMMETRY SYMMETRY L14 L15 A1 Figure 3.2 Model of Office Cubicle with Overhead Air Distribution System The office floor layout can be thought of as including many aisles of cubicles. Each block of an aisle incl udes two cubicles symmetrically facing each other through a panel separator between them. The left boundary above the separator is considered as symmetry boundary. The open space on the right side models half of the walkway (perpendicular to paper’s pl ane) between cubicle aisles. The right boundary was also taken as symmetry boundary becaus e of the symmetric of floor layout. Right next to the separator is a desk and a pers onal computer (CPU a nd monitor) placed on it. A person is sitting on a chair, facing the computer. The lights are located on the ceiling, right above

PAGE 27

19 the person’s position. On the backside of the person is the right symmetry boundary at half of the walkway. The air return outlet is placed on the ceiling above that region for both cases. The top face of the monitor was defi ned in CFD model as the entity “hot top” for releasing heat flux to the surrounding. Th ere was also heat flux from the lights. The person was considered as constant temperat ure surface and also imposed a flux of water vapor. Contaminant gas as evaporating cleaning chemicals released from the rug on the floor, as a mass flux. Table 3.1 Dimension Parameters on Figures 3.1 and 3.2, meter(s) L1 L2 L3 L4 L5 L6 L7 L8 L9 L10L11 L12 L13L14 2.7 1.75 0.8 0.7 0.4 0.35 0.15 0.25 0.8 0.5 0.4 0.6 0.54 2.0 Table 3.2 Inlet Boundary Conditions, Inlet Setup and Simulation Cases Air distribution system Underfloor Overhead Inlet temperature 20oC 18oC Inlet speed 1.0 m/s 0.6 m/s Inlet direction Vertical upward Oblique downward Inlet width (L15) 0.16 m 0.2 m Inlet location, L16 Inlet angle, A1 Varying parameter 0.54 m Typical 1.54 m Under desk 0.18 m Face outlet 30o Typical 45o 60o Simulation case number 1 2 3 4 5 6

PAGE 28

20 The two models are almost identical except for the air supply inlet. Key differences between the systems arise with th e location and the size (and thus the flow rate) of inlets. For the underfloor system, the inlet is located near the back of the person on the floor, which is the typical setup. Tw o more locations of in let were considered: under the desk and facing the outlet, which make two limit locations. For the overhead system, the supply diffuser is located on the ceiling, right above the separator, symmetrically sharing for two opposite cubicles ; hence our model on Fi g. 3.2 takes a half of the diffuser size as its inlet size. The typi cal inlet angle is 30 de grees. By exploring how inlet angle affects thermal comfort and contaminant removal characteristics of the model, two more inlet angles we re considered: 45 and 60 degrees. Because the air is supplied directly into the occupied zone in the underfloor system, supply air temperatures can be highe r than that used for conventional overhead system. Higher supply air temperatures woul d suggest that higher supply air velocities are required. Inlet speed and temperature for each system are given in Table 3.2. The CFD simulations estimated variables such as pressure, velocity, temperature, and contaminant concentration for each cell, throughout the entire cubicle in accordance with mass and concentration conservation e quations. Six simulations were performed, three for the underfloor system and the other three for the overhead system. The simulation cases and associated inlet boundary conditions are given in Table 3.2. Details of boundary conditions are given in Table 3.3. For each simulation, velocity components and temperature were found first by solving the coupled equations Eq. 2.1, 2.4, and 2.5, then the species concentrations were solved from the equations Eq. 2.2 and 2.3 with known velocity field.

PAGE 29

21 Table 3.3 Boundary Conditions for Office Room Simulation No. Entity Velocity Temperature/ Heat flux Water vapor concentration Contaminant concentration 1 Inlet See Table 3.1See Table 3.1 0.011 kg/kg air 0 kg/kg air 2 Symmetry UX = 0 Flux =0 Flux = 0 Flux = 0 3 Hot top 0 Flux = 100 W/m2Flux = 0 Flux = 0 4 Lights 0 Flux = 75 W/m2 Flux = 0 Flux = 0 5 Person 0 Temp = 33oC Flux = 5E-7 kg/(m2.s) Flux = 0 6 Floor 0 Flux = 0 Flux = 0 Flux = 1E-6 kg/(m2.s) 7 Outlet Unknown Unknown Unknown Unknown 8 Others 0 Flux = 0 Flux = 0 Flux = 0 3.3 Results and Discussion Three cases were simulated for the underfl oor system (simulations 1, 2, and 3) and another three for the overhead system (sim ulations 4, 5, and 6). For each simulation, the governing equations Eq. 2.1 – 2.5, associated with the boundary conditions given in Tables 3.2 and 3.3, were solved by using fini te element analysis. Each solution included velocity field, pressure, temperature, wa ter vapor concentrati on, and contaminant concentration. Relative humidity distributi on was then computed by using equations Eq. 2.6 – 2.8. Predicted mean vote (PMV) was ca lculated based on solution average values using equations Eq. 2.9 – 2.12. Thermal se nsation index was calculated from equation Eq. 2.13 and contaminant removal effectiveness, equation Eq. 2.15.

PAGE 30

22 Figure 3.3 shows velocity distribution for typical case of underfloor system (simulation 1). The velocity vector field was plotted on color b ackground showing speed distribution. The cool airflow entered the c ubicle vertically through a floor-level diffuser at uniform full speed (1.0 m/s). The main flow slightly bent to the left then vertically swept along the local space near the person’s back, up to about 1.8 m height, spread and bent to the right toward the return outlet at reducing speed. Near the outlet, a fraction of the main flow did not go through the outlet but made a sharp U-turn and vertically went down along the symmetry boundary. The upward flow were dominated by forced convection from the imposed inlet velocity a nd also induced by natural convection due to higher temperature surface along the person’s back, while the downward flow was under the effect of natural convection only because of its lower temperature. The upward and downward flow created a region of circulation near the backsi de of the person (right hand side on Fig. 3.3). Figure 3.3 Velocity Field for Simulation 1, m/s

PAGE 31

23 There was also a small flow separated from the main flow, sweeping under the chair and the desk, creating some kind of ai r mixing flow, and moving up through the gap between the person and the desk. This sma ll flow might have some positive but little effect on the natural convection flow along th e front surface of the person, which caused a slight circulation in the region between the person an d the computer. The region above the person and the computer was a mixing z one mostly caused by natural convection, showing unclear gentle circulations. Figure 3.4 shows the temperature distri bution for typical underfloor case. The right hand side region, or backside of the pe rson, was a zone with temperature as low as inlet air temperature, since the backside circ ulation caused by the main stream was strong that make the air in that region well mixed with the cool air from inlet, inducing heat transfer, mostly by convection. Figure 3.4 Temperature Dist ribution for Simulation 1, oC

PAGE 32

24 On Fig. 3.4, in other regions outside th e main stream, temperature was higher but still fair-manner distributed due to both diffusion and natural convection. Slight circulation under the chair and the desk as we ll as the absence of heated surface kept the region colder than the worki ng space above where there are hot surfaces such as the person, the computer, and the lights. Ther e were small moderate temperature zones around the person model, which was of constant temperature, and high temperature zones close to the computer top and lights wher e there were heat fluxes coming into the occupied zone. A warmer region of about 25oC was formed in the region above the person. Figure 3.5 is the plot of relative humid ity distribution for typical underfloor case. Relative humidity is a function of absolute pressure, water vapor concentration, and temperature. Since the gage pressure in the whole region was found very small (at the order of 1 Pa) compared to the atmospheric pressure (at the order of 101 kPa), it does not affect the total absolute pressure significan tly. The water vapor concentration also does not change much, since the only water vapor supply was the person’s surface with very small flux. Therefore, the relative humid ity distribution was mostly dependent on temperature distribution, and their plots appear similar. In the backside circulation zone, it is observed a uniform distribution at about 75%. In the person’s working zone, relative humidity was about 50%-65% everywhere and higher at hot surfaces, such as computer top, and lights. Around the person area, relative humidity was between 50% and 60%; a zone of low relative humidity (around 50%) was found in front of the person and in the warmer region above.

PAGE 33

25 Figure 3.5 Relative Humidity Distribution for Simulation 1 Figure 3.6 shows the distribution of c ontaminant concentration for typical underfloor system. In this CFD models, it is supposed that contaminant to be releasing from the floor as a constant flux. Contam inant transport was driven by concentration gradient and by convection. On Fig. 3.6, it can be seen that there is almost zero contaminant concentration in the main flow and regions next to it, since the inlet flow of fresh air swept through the region and br ought the contaminan t to the outlet by convection. The region around, above and in fr ont of the person was almost contaminant free, by the effects of the small flow thr ough the gap and the natu ral convection flow along the front surface. The backside circula tion, while sweeping along the floor, kept the high contaminant concentration confined in the small zone right above the floor. Under the desk and the chair, air was moving very slowly, thus the main transport mean was by diffusion that made a uniform-like distributi on of higher concentration under the desk.

PAGE 34

26 Figure 3.6 Contaminant Concentration Di stribution for Simulation 1, kg/kg air The results from the simulations 2 and 3 show some different behavior of the airflow with respect to different inlet loca tions. We will comment on their affects on average parameters controlling thermal comfort and contaminant removal. For a convenient view of how thermal environment and contaminant concentration respond to an air distribution syst em, we consider the vertical distributions of air speed, temperature, relative humidity, and contamin ant concentration. At each different height, average values of the parame ter of interest were taken over all the width of the region. Figure 3.7 presents a comparis on of vertical distri bution of average air speed for three cases of the underfloor sy stem. The typical case shows a moderate vertical distribution of air speed in the range of 0.2-0. 3 m/s, while the under-desk-inlet case (simulation 2) and the inlet-facing-out let case (simulation 3) show significant changes of air speed to the height. Both cases give high air speed as high as 0.4 m/s at

PAGE 35

27 about 1.7 m, right on top the person, which mi ght not comfortable for the person. For the under-desk-inlet case, the air sp eed was higher than that for typical case in the region under the desk and the chair, as one might e xpect, which is not comfortable as well. The average air speed in that region was far higher for the case of inlet facing outlet, but most of the high air speed concentrat ed at the inlet region as it wa s coupled with the outlet to form a straight open flow channel, wh ich does not much affect the person. 0 0.5 1 1.5 2 2.5 00.10.20.30.40.5 Average air speed (m/s)Height (m) Simulation 1 Simulation 2 Simulation 3 Figure 3.7 Vertical Distribution of Aver age Air Speed for Underfloor System Figure 3.8 shows the vertical distribution of temperatur e for underfloor cases. All the three cases show similar distributions. Temperature was higher for typical case along most of the sitting height of the person but lower at height cl oser to the ceiling. Vertical average temperature was most uniform in under-desk-inlet case with a narrow band of

PAGE 36

28 temperature variation of 20oC-23oC. The temperature in the region under the desk and chair is important for comfort of the lower part of the person such as legs and feet. Typical case shows the best ch aracteristics in this aspect while the other cases might cause the “cold feet” effect on the person. 0 0.5 1 1.5 2 2.5 182022242628303234 Average temperature (oC)Height (m) Simulation 1 Simulation 2 Simulation 3 Figure 3.8 Vertical Distribution of Aver age Temperature for Underfloor System Figure 3.9 shows the vertical distribution of average relative humidity for three underfloor cases. The typical case had the mo st uniform distributi on and provided best comfort for the lower region, ranging in 63%-72%. The under-desk-inlet case shows more uniform relative humidity in the hei ghts occupied by the person but the higher in the lower region and lower in higher region ma de the over all perfor mance less satisfied.

PAGE 37

29 0 0.5 1 1.5 2 2.5 4550556065707580 Average relative humidity (%)Height (m) Simulation 1 Simulation 2 Simulation 3 Figure 3.9 Vertical Distribution of Average Relative Humidity for Underfloor System Figure 3.10 is a comparison among the three underfloor cas es in vertical distribution of average contaminant concentrat ion. All three cases s how uniform vertical distribution. The lowest leve l of contaminant concentratio n was found in inlet-facingoutlet case, resulted from the direct flow from inlet to outlet. The under-desk-inlet case had highest level of contam inant concentration (about 0.00015 kg contaminant/kg air mixture), while the typical case kept it as low as one-third of that level (0.00005 kg contaminant/kg air mixture). Figure 3.11 shows the velocity fields for typical overhead system (simulation 4). The fresh, cool airflow entered the re gion through supply inlet on the ceiling with uniform 0.6 m/s speed at 30o downward. Most of the main flow went down induced by natural convection because of its lower temperature.

PAGE 38

30 0 0.5 1 1.5 2 2.5 00.10.20.30.40.5 Average contaminant concentration (parts per thousand)Height (m) Simulation 1 Simulation 2 Simulation 3 Figure 3.10 Vertical Distri bution of Average Contaminant Concentration for Underfloor System Figure 3.11 Velocity Field for Simulation 4, m/s

PAGE 39

31 Under the effects of mixed convection cont ributed by the main flow itself and by natural convection due to hot surfaces of th e person and computer top, an upward flow was formed and went toward the outlet. That e ffect pushed the main flow bent to the left and they created a circulation in the regi on above the working space of the person. A small part of the main flow after sweep ing the computer top went through the gap between the desk and the person to the floor and rose up a gain on the backside caused a slight slow circulation. Generally, this veloci ty distribution shows mo re disturbances than that of the typical underfloor system. Figure 3.12 gives a view of temperatur e distribution for typical overhead case. Temperature was distributed more uniformly fo r this case compare to that of underfloor system, because of the better mixing as a result from the more perturbed velocity field. Lowest temperature region was in the main flow, above and in front of the person’s working space. The hot zone next to the co mputer top was reduced significantly, since the downward main flow sweeping through the z one with cool air removed most of the heat released. However, the higher temperat ure zone around the pe rson seems to be the same, as heat transport in this zone relied mainly on diffusion rather than convection due to lower air speed, but being compensated by lower inlet temperature. The warmer zone above the person was larger than in typical underfloor case but of lower temperature. Figure 3.13 shows relative humidity dist ribution for typical case of overhead air distribution system. For this case, similar to any other case, relative humidity distribution depends strongly on temperature distribution. The relative humidity seems a little higher over all compared to that in the UFAD case, and the zone of low humidity around the person remained the same, as for temperatur e. Humidity was 65%-75% all over the place.

PAGE 40

32 Figure 3.12 Temperature Dist ribution for Simulation 4, oC Figure 3.13 Relative Humidity Distribution for Simulation 4

PAGE 41

33 Figure 3.14 shows the di stribution of contaminant concentration for typical overhead case. The person’s wo rking space was almost contaminant free since the force convection flow from the ceilin g does not directly induced the contaminant from the floor as well as the strong circula tion in the above zone and th e through-gap flow drove the slight concentrated contaminant away and ke pt the high concentration stay close to the floor. The contaminant highly concentrated in a small portion at the right symmetry boundary as the result of a raising flow by natural convection. Figure 3.14 Contaminant Concentration Di stribution for Simulation 4, kg/kg air The overhead system alternative cases with different inlet angle affect the airflow response but not too significantl y. Figure 3.15 compares th e three overhead cases in vertical distribution of air speed. The t ypical overhead case with inlet angle of 30o shows most moderate distribution with air speed in the range of 0.07m/s-0.25m/s, while the 45o

PAGE 42

34 case (simulation 5) and the 60o case (simulation 6) show larg e changes in air speed along the height, 0.05m/s-0.25m/s and 0.07m/s-0.27m/s respectively. The air speeds at the feet were of the same order as we might expect th at the change of inlet angle would not affect the lower part of the whole region. The 60o case had lowest air spee d at the sitting height. 0 0.5 1 1.5 2 2.5 00.050.10.150.20.250.3 Average air speed (m/s)Height (m) Simulation 4 Simulation 5 Simulation 6 Figure 3.15 Vertical Distri bution of Average Air Speed for Overhead System Figure 3.16 is the plot of vertical average temperature for three overhead cases. It shows that the typical 30o case had higher temperature at the height of sitting person (22oC-24oC) while lower at the feet (19oC-20oC). The alternative cases had almost the same performance. Their temper ature distributio ns range in 19oC-23oC, slightly cold at feet, warmer at the person sitting height a nd getting colder toward the ceiling, just like the typical overhead case, but c ooler over all the total height.

PAGE 43

35 0 0.5 1 1.5 2 2.5 182022242628303234 Average temperature (oC)Height (m) Simulation 4 Simulation 5 Simulation 6 Figure 3.16 Vertical Distri bution of Average Temperature for Overhead System Figure 3.17 presents the ve rtical distribution of rela tive humidity for overhead cases. The typical overhead case displays lo wer relative humidity but more uniform, while the other cases shows highe r humid in the higher part of the region. In general, all three curves look very similar in thei r form. Figure 3.18 shows how contaminant concentrations were distributed vertically for overhead cases. The typical overhead case gave very good characteristic s of contaminant control with lowest and almost unchanged concentrated level. The alternative 45o case had higher concentra tion from the floor up to almost all the working height then redu ced toward the ceiling; while in the 60o case, the level increased gradually from the floor, c overed the working hei ght and then reduced toward the ceiling.

PAGE 44

36 0 0.5 1 1.5 2 2.5 5055606570758085 Average relative humidity (%)Height (m) Simulation 4 Simulation 5 Simulation 6 Figure 3.17 Vertical Distribu tion of Average Relative Humidity for Overhead System 0 0.5 1 1.5 2 2.5 00.20.40.60.81 Average contaminant concentration (parts per thousand)Height (m) Simulation 4 Simulation 5 Simulation 6 Figure 3.18 Vertical Distri bution of Average Contamin ant Concentration for Overhead System

PAGE 45

37 Table 3.4 compares summarize results from simulations of the underfloor system to experimental data from [5]. Table 3. 5 shows a similar comparison for the overhead system simulation results and those results f ound in experimental da ta from [5] for an analogous distribution system. The parameters of interest are the average values of air speed, temperature, relative humidity. The average values were taken for each parameter on all over the computation region. Table 3.4 Comparison of Average Values of Th ermal Comfort to Experimental Data for Underfloor Air Distribution System Simulation No. Parameter 1 2 3 Experiment results [5] Air speed (m/s) 0.276 0.360 0.328 0.12 Temperature (oC) 21.7 21.1 21.4 24.4 Relative humidity 66% 69% 68% 60% Table 3.5 Comparison of Average Values of Th ermal Comfort to Experimental Data for Overhead Air Distribution System Simulation No. Parameter 4 5 6 Experiment results [5] Air speed (m/s) 0.201 0.173 0.206 0.19 Temperature (oC) 22.3 21.6 21.9 24.9 Relative humidity 70% 72% 72% 65%

PAGE 46

38 From the obtained average values, thermal comfort factors as predicted mean vote (PMV) and thermal sensation index were also computed, using the equations Eq. 2.9 – 2.12 and 2.13, respectively. Figure 3.19 shows the change of PMV a nd thermal sensation as inlet location changes for the underfloor system model. PM V and thermal sensation values are very close to each other and to the lower limit of the comfort zone. They show a slightly change in thermal comfort level due to inlet location. The higher values of PMV and thermal sensation appear as the inlet is under the seat. 0.18 0.54 1.54 -3 -2 -1 0 1 2 3 Inlet Location (m)Predicted Mean Vote, PMV Comfort zone -3 -2 -1 0 1 2 3 Thermal Sensation, Y PMV Y Figure 3.19 Thermal Comfort Factors vs. Inlet Location for Underfloor System Figure 3.20 shows the change of PMV and thermal sensation as inlet angle changes for the overhead system model. Sim ilar to the underfloor case, PMV and thermal sensations values are very close to each othe r and slightly change as inlet angle changes. They lay completely inside the comfort zone but still close to its lower limit. The lower

PAGE 47

39 thermal comfort level is at the average inlet angle (45o). For both underfloor and overhead system models, there may be extra heat sources on the office floor unaccounted for; including their effects can raise the PMV and thermal sensation for both model and move them inside the comfort zone with a large adjusting margin for designing. 30 45 60 -3 -2 -1 0 1 2 3 Inlet Angle (degree)Predicted Mean Vote, PMV Comfort zone -3 -2 -1 0 1 2 3 Thermal Sensation, Y PMV Y Figure 3.20 Thermal Comfort Factors vs Inlet Angle for Overhead System The average air speed for under floor case was higher than that for overhead case, resulted from higher inlet air speed. The aver age temperatures and relative humidity were almost the same for both cases but sli ghtly lower for underfloor system. PMV and thermal sensation index were inside or cl ose to the comfort zone for both cases. In general, the two systems were satisfied in thermal comfort viewpoint. The simulation results agree with experiment al data on most of the rela tionships between UFAD and overhead system, such as lower average temperature and relative humidity, or higher PMV, for UFAD system compared to overh ead system. However, for the same air

PAGE 48

40 distribution system, there are differences be tween simulation and experimental results. For both air distribution systems, experimental data shows that the average temperature was higher than that of simulation results, thus lower relative humidity, and the thermal comfort indices were closer to neutral conditi on. This could be that the simulations were done for a small cubicle for individual use with symmetric assumed on its boundary to the rest of a large office floor, while the e xperiments were carried out in a more common zone where there are many more factors interf ered. Another reason could be the total heat load was underestimated for not taking into account of many kind heat loads in the common area on an office floor or particular us ed area, such as sunlight radiation through glass windows, photocopy machines or some ot her heat generated business equipments. Table 3.6 shows a comparison of the contaminant removal performance for both systems. On considering average contaminan t concentrations, it seems that contaminant concentration can be higher for underfloor syst em since the inflow at floor-level likely induced the convection of contaminant, also from the floor. The average contaminant concentrations over all as well as at outlet were about of the same order. CRE values were ranging from 0.2 to 0.4 for both systems. For each system, the typical set up shows best control of contaminant removal. Table 3.6 Comparison of Contam inant Removal Effectiveness Underfloor system Overhead system Simulation No. 1 2 3 4 5 6 CRE 0.26 0.33 0.27 0.20 0.21 0.38

PAGE 49

41 Figure 3.21 is the plot of vertical distri bution of average air speed for the typical cases of the two systems. The distributi on profiles for both unde rfloor and overhead system are similar, as expected the speed value was higher for underfloor system, mostly because of the higher speed at inlet. For bot h cases, the vertical average speed quickly increased from the floor level, then graduall y increased in the zone occupied by the person and the computer, and continued the tr end toward the ceiling. It shows that the average air speed was slow down s lightly at the person position. 0 0.5 1 1.5 2 2.5 00.050.10.150.20.250.3 Average air speed (m/s)Height (m) Simulation 1 (underfloor) Simulation 4 (overhead) Figure 3.21 Vertical Distri bution of Average Air Speed, Underfloor System vs. Overhead System Figure 3.22 shows the vertical distribution of average temperature. Both system show the average temperature was higher at the person’s position. Most parts of the distribution curves for both systems are id entical except the temp erature of underfloor

PAGE 50

42 system is slightly higher. The difference of vertical average temperature is about 1oC, while the difference of inlet temperature was 2oC. 0 0.5 1 1.5 2 2.5 182022242628303234 Average temperature (oC)Height (m) Simulation 1 (underfloor) Simulation 4 (overhead) Figure 3.22 Vertical Distribu tion of Average Temperatur e, Underfloor System vs. Overhead System Figure 3.23 shows the vertical distribution of average relative humidity. Relative humidity was lower at the person’s position, and it was lower for underfloor system than for overhead system, i.e. the person feel s “dryer” if using underfloor system. Figure 3.24 shows the vertical distribution of average contaminant concentration. Overhead system has better perf ormance in this aspect. Its di stribution profile was at low values but less uniformly distributed along the height, while for underfloor system, almost constant higher concentra tion distributed along the height.

PAGE 51

43 0 0.5 1 1.5 2 2.5 455055606570758085 Average relative humidity (%)Height (m) Simulation 1 (underfloor) Simulation 4 (overhead) Figure 3.23 Vertical Distributi on of Average Relative Humidity, Underfloor System vs. Overhead System 0 0.5 1 1.5 2 2.5 00.10.20.30.40.5 Average contaminant concentration (parts per thousand)Height (m) Simulation 1 (underfloor) Simulation 4 (overhead) Figure 3.24 Vertical Distributi on of Average Contaminant C oncentration, Underfloor System vs. Overhead System

PAGE 52

44 The performance of each system at th e outlet location was studied. These analyses are shown on Fig. 3.25 and 3.26. Figure 3.25 shows the distribution of the vertical component of the velocity vector, which is the normal velocity, along the outlet length. For both systems, the higher vertical velo city was concentrated on the right half of the outlet, i.e. toward the symmetry boundary. Vertical velo city was distributed much more uniformly for underfloor system than overhead system. -0.5 0 0.5 1 1.5 2 1.461.521.581.641.71.761.821.881.942 Position (m)Vertical air velocity component (m/s) Simulation 1 (underfloor) Simulation 4 (overhead) Figure 3.25 Distribution of Verti cal Velocity along Outlet Lengt h, Underfloor System vs. Overhead System Figure 3.26 shows the distribution of contaminant concentration along outlet length. From this figure, we can see that th e higher concentration was on the left side of the outlet for underfloor system and on almo st uniform over all the outlet length for overhead system.

PAGE 53

45 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 1.461.521.581.641.71.761.821.881.942 Position (m)Contaminant concentration (parts per thousand) Simulation 1 (undefloor) Simulation 4 (overhead) Figure 3.26 Distribution of C ontaminant Concentration along Outlet Length, Underfloor System vs. Overhead System From the above discussion, it is shown that in each system, the typical case is the best setup for that particular system. Both systems satisfy thermal comfort requirements. They have similar performance characterist ics in thermal comfort performance. The underfloor velocity field is ge ntle while the overhead syst em is more perturbed. The underfloor system has more risk of induced the contaminant at the floor, while the contaminant removal effectiveness of both systems are almost the same.

PAGE 54

46 Chapter 4 Predictions of Thermal Comfort and Cont aminant Removal in an Operating Room 4.1 Introduction This part of the work uses airflow simu lations to evaluate different ventilation systems on an operating room (OR). This st udy compares air distri bution systems for an operating room by use of computational flui d dynamics (CFD) modeling. The air supply distribution and exhaust arrange ments were modeled for a di rectional air flow system where air moves across the space from the high -pressure supply area to the low pressure exhaust area. A simplified model of a typical operating room (Fig. 4.1) was considered with inclusion of objects such as surg ical lights, operating table, h eat sources such as surgical staff (standing) and a patient (lying on opera ting table), side wall supply grilles and exhaust air grilles. Inlet angle and air return locations were both st udied. One and two airexhaust outlet sites inside the surgical suite were considered. For basic configuration, the model only has the exhaust grill s at lower positions on the ri ght wall. The discharge angle for the supply grilles was varied from 0 to 45 degrees. For the two-exhaust outlet configuration, one outlet position was low, cl ose to the floor and the other position was high on the right wall. Simulations with co mbinations of 30:5, 25:10, 20:15, 15:20, 10:25, and 5:30 flow rates between the tw o return locations were performed.

PAGE 55

47 Figure 4.1 Simplified Typical Operating Room Calculations were done for the operating room’s 2D model (Fig. 4.2) in steadystate condition. Predictions for the air m ovement, room temperature, room relative humidity, and concentration of contaminan ts within the operating room are shown. Analysis of these predictions is discussed. The supply and exhaust conditions of the ventilation airflow are shown to play an important role in the control of air quality. Results show good agreement with experimental data. 4.2 CFD Model The operating room was modeled as a 2-D rectangular region with its four boundaries present floor, ceiling and two wall s as shown on Fig. 4.2. The essential

PAGE 56

48 dimensions are denoted in general forms as L1 to L18 for lengths and A1 for inlet angle. Giving the dimensions in general form ma kes it flexible for further parameterized investigation, so that alte rnating essential dimensions can be performed without significant changes in the CFD program. A1 L10 HIGH OUTLET L18 WALL L7 L8 L9 L12 L13 L5 L4 L3 L6 L4 L4 L11 L10 L2 L17 L16 L1 L14 L15 FLOOR CEILING WALL LOW OUTLET INLET (SUPPLY) SURGICAL LIGHTSSTAFF 2 STAFF 1PATIENT Figure 4.2 Model of Operating Room The numerical values of the lengths L1 to L16 used for the computations in this paper are given in Table 4.2. The inlet angle A1, and the low and high outlet length, L17 and L18, respectively, are the varying parame ters whose effects are to be considered. The air supply inlet of the room is locat ed at high position on the left wall. For one-exhaust (basic) configuration, there is only one outlet placed at low position on the

PAGE 57

49 right wall (low outlet, L17). For two-exhaus t outlet configuration, there is also an additional outlet placed at high position on the ri ght wall, at the same height level of and facing the inlet across the room length (high outlet, L18). The sizes of both outlets can be changed but the total size of the two outlets is kept the same as that for the outlet of basic case (and equal to the inlet size, i.e. L15). For the two-exhaust configuration, the highoutlet-to-total-outlet ratio HTR is defined as 15 L 18 L 18 L 17 L 18 L area outlet Total area outlet High HTR (4.1) Table 4.1 Dimensions Parameters on Figure 4.1, meter(s) Name Length Name Length Name Length Name Length L1 6.00 L5 2.00 L9 0.20 L13 2.70 L2 3.50 L6 1.80 L10 0.65 L14 0.30 L3 1.75 L7 0.80 L11 0.30 L15 0.35 L4 0.25 L8 1.75 L12 0.60 L16 0.20 To investigate the effect of the supply inlet angle, five cases of inlet angle A1 were considered: 0 5 15 30 and 45o for the basic configur ation (one-exhaust). The inlet angle was measured clockwise from the hor izontal direction, i.e. the inlet flow was directed level (0 ), down (5 15 30 and 45o, toward the floor). For the two-exhaust configuration, six combinati ons of different sizes of hi gh outlet and low outlet (with unchanged total size) were studied, wh ile the inlet angle was kept at 0 These simulation cases are summarized in Table 4.2.

PAGE 58

50 Table 4.2 Inlet Angle, Outlet Sizes, Outlet Ratios, and Simulation Cases Simulation number Inlet angle, A1 (degree) Low outlet, L17 (m) High outlet, L18 (m) HTR (%) 1 0 0.35 0 0 2 5 0.35 0 0 3 15 0.35 0 0 4 30 0.35 0 0 5 45 0.35 0 0 6 0 0.30 0.05 14.3 7 0 0.25 0.10 28.6 8 0 0.20 0.15 42.9 9 0 0.15 0.20 57.1 10 0 0.10 0.25 71.4 11 0 0.30 0.05 85.7 The two walls were kept at constant te mperature. The lying patient was modeled as the horizontal rectangle at the middle of the room. Its bottom edge facing the floor modeled the operating table, which is heat and mass insulated. The other three edges modeled the patient’s body, which was kept at constant temperatur e and releasing heat, water vapor, and contaminant as constant fl uxes. The standing staffs were modeled by two vertical rectangles at both of the patient’s ends. Similar to the patient’s model, these two staff models were consid ered surface at constant temp erature and constant water vapor flux. The surgical light was also mode led as a rectangle above the patient, whose

PAGE 59

51 bottom edge (facing the patient) was defined as “lamp face” entity, on which the major heat flux went through; and th e other three edges were defi ned as “lamp back” entity, on which a smaller heat flux went through. The boundary conditions on outlet was unknown and to be solved for, as part of the soluti on of the flow over the whole region. The other boundary conditions left unmentioned were a ssumed to be zero velocity and totally insulated to heat and mass (e.g. zero velocity and neither heat flux nor mass flux at solid surfaces such as walls, floor, and ceiling; no contaminant flux from the staffs’ body, etc.). Details of boundary conditions are given in Table 4.3. Table 4.3 Boundary Conditions for Operating Room Simulation No. Entity Velocity Temperature/ Heat flux Water vapor concentration Contaminant concentration 1 Inlet V = 0.4 m/s, (See Table 4.2) T = 17oC m1 = 0.01018 kg/kg air m2 = 0 kg/kg air 2 Walls 0 T = 22oC Flux = 0 Flux = 0 3 Lamp face 0 Flux=100W/m2 Flux = 0 Flux = 0 4 Lamp back 0 Flux=5 W/m2 Flux = 0 Flux = 0 5 Patient 0 T = 33oC Flux = 5E-7 kg/(m2.s) Flux = 1E-5 kg/(m2.s) 6 Staff 0 T = 33oC Flux = 8E-7 kg/(m2.s) Flux = 0 7 Outlet Unknown Unknown Unknown Unknown 8 Others 0 Flux = 0 Flux = 0 Flux = 0

PAGE 60

52 For each simulation, velocity components and temperature were found first by solving the coupled equations Eq. 2.1, 2.4, and 2.5, then the species concentrations (water vapor, contaminant gas) were solved from the equations Eq. 2.2 and 2.3 with known velocity field. The solutions of the finite element analysis generated velocity field, pressure, temperature, water vapor concentra tion, and contaminant concentration. From pressure, water vapor concentration, a nd temperature, the relative humidity was computed by using Eq. 2.6 – 2.8. From rele vant average parameters, predicted mean vote (PMV) was computed by using Eq. 2.9 – 2. 12, thermal sensation by Eq. 2.13, and contaminant removal effec tiveness (CRE) by Eq. 2.14. 4.3 Results and Discussion Figure 4.3 presents velocity distribution for the basic case (simulation 1 in Table 4.2). On Fig. 4.3, velocity field was plotte d on the filled speed contour background. It gives the image of how the direction (veloc ity vector) and magnit ude (speed) of the velocity field are distributed. The flow ente red the room through the inlet located high on the left wall at 0o, with full speed (0.4 m/s). If ther e is negligible buoyancy effect, the main stream will flow straight forward at first as shown in [24]. However, for this problem the buoyancy effect is quite strong, wh ich caused the most of the inflow to bend down sharply right at the inle t because of its lower temper ature and thus, higher density, compare to the average temperature in the ro om. The resistances of the stream against this sharp turn created a complicated perturbed region in the higher part of the left end of the room, where the strong separation of stre ams took place. Most of the inflow went down along the wall and swept al ong the floor to the outlet. Its top layer mixed with the

PAGE 61

53 warmer air next to it, rose at warm surf aces as staffs’ bodies under the influence of buoyancy effect, causing upward flows. Th e warmer upward flows along the staffs’ bodies combined with the cooler downward flows along the wall, a nd supposed to form some slight circulations in the unocc upied space between them. However, these circulations were influenced and deformed by the perturbed region right above it. They combined to make a complicated mixing region. At the inlet, a smaller part of the inflow was pushed up to the ceiling, instead of going down like most part of the inflow. This stream mixed up with the hot air coming up fr om the surgical site and the lights, swept along the ceiling, and went dow n at the wall on the right, th en exit at the low outlet. Combined with the natural c onvection flow along the right staff, it created slight circulations in the unoccupied spac e at the right end of the room. Figure 4.3 Velocity Field for Simulation 1, m/s

PAGE 62

54 At the surgical site, stronge r circulations were formed as the results of the a complicated mixing of natural convection al ong the hot surfaces of people bodies and lights inside the half closed region, interfacing with the co ld air around the site from outside of that region. The natural convective flow along the staff on the left went up and got warmer, until it hit the perturbed mixing re gion above this person, it went down to the lying patient. The buoyancy effect in the ga ps between the staffs and the patient was weak because of the resistance of small gaps and the lack of temperature difference, but it could push the flow a little to the right. There, the flow swept along the patient and raised up at the staff on the right, mixed up with the natural convection flow along this person from outside and moved up. This stream wa s strengthened by the mixed convection flow in the region near the lights a nd moved to the left. There, it was affected by the perturbed region on its left to form a small region of circulation as an in tersection of several streams, then mixed up with the ceiling st ream and ran to the right along the ceiling, down the right wall, and exit, as described for the ceiling stream. Figure 4.4 is the plot of temperature distribution for the basic case. The low temperature of the supply air from inlet is concentrated mostly along walls, ceiling, and floor. There are people, considered as surfaces of constant temperature, and surgical light – as surfaces of constant heat flux. Near these surfaces, temperature changes very steep. In the far surroundings, the temp erature distribution seems uniform in general, as this problem is natural convection dominant and this type of convect ion has better mixing capability than force convection. Another obser vation is that the te mperature distribution mostly looks like the distribution of velo city, i.e. the convective terms are more significant than the diffusive terms in the energy equation.

PAGE 63

55 Figure 4.4 Temperature Dist ribution for Simulation 1, oC Figure 4.5 is the plot of relative humidity distribution, a key factor of thermal comfort. Relative humidity is a function of ab solute pressure, water vapor concentration, and temperature. Its distribution was comput ed from pressure, temperature, and water vapor concentration, using ASHRAE proce dure as mentioned above. Since the room gage pressure was found very small (at the or der of 1 Pa), compared to the atmosphere pressure (as high as 101 kPa), then it does not significantly affect the total (absolute) pressure, and thus almost does not affect the values of relative hum idity. Wherever low temperature and high water vapor concentration exist, relative humidity is high also. Near the surgical light, the relative humidity is very low because of the high temperature. There is a high humidity region on the right side of a staff, thus the staff on the right hand side has a more humid surrounding.

PAGE 64

56 Figure 4.5 Relative Humidity Distribution for Simulation 1 Figure 4.6 shows the plot of contaminant concentration distribution. Contaminant supposedly releases from the body of the patient at a constant rate. It is driven by the concentration gradient, i.e. from patient to the surrounding, especially to the flow of “fresh air”. Then the airflow carries contaminant to the outlet. We can see that the process is very effective: the flow swept thr ough the patient from left to right and wash the contaminant away, rose up and carried it to the outlet. Near the patient, the higher concentration is on the right end; theref ore, the staff on the right gets a higher contaminant concentration in front of him/her. Figure 4.7 shows how the contaminant c oncentration changed for the case of 45 inlet angle on basic configur ation (simulation 5). The aver age contaminant concentration increased significantly, a bout 30 times compare to that of simulation 1.

PAGE 65

57 Figure 4.6 Contaminant Concentration Di stribution for Simulation 1, kg/kg air The significant reduction of contaminan t concentration in simulation 4 happens because the inlet angle down directed the fl ow sweeping along the wall, in favor of the downward orientation of the cold air in buoyancy dominant region, whose effect was mostly sweeping the floor rather than going ove r the surgical site and washed away the contaminant to the outlet. The velocity field is shown on Fig. 4.8 for simulation 5 (basic configuration, 45o-inlet angle). Figure 4.9 presents the contaminant conc entration for the case of two-exhaust configuration with HTR = 71.4% (simulation 10). It can be observed that the contaminant concentration will exit the room by the high outlet if there is one. Although it shows a worse case than the basic case in contaminan t removal, it suggests that the use of an additional high outlet may improve the contaminant removal performance of the room. 10

PAGE 66

58 Figure 4.7 Contaminant Concentration Di stribution for Simulation 5, kg/kg air Figure 4.8 Velocity Field for Simulation 5, m/s 10

PAGE 67

59 Figure 4.9 Contaminant Concentration Di stribution for Simulation 10, kg/kg air To evaluate the contaminant removal performance of the room, the average contaminant concentration and the contaminant removal effectiveness (CRE) were considered. The average value for each simu lation was taken for all over the computation region. Figure 4.10 shows that the average contam inant concentration increases as inlet angle increases. Thus for this kind of buoyancy dominant airflow in operating room, increasing the inlet angle may induce higher contaminant concentration. If the buoyancy effect is negligible (force convection dominant), the incr ease of inlet angle (directed down) may help reduce the contaminant concen tration [24], because it directs the main flow to wash through the surgi cal site. However, if buoyancy effect cannot be neglected, which is now considered in the simulations, the basic case of 0o-inlet angle itself is 10

PAGE 68

60 equivalent to a case of se veral tens of degree down in problem with only force convection, since the buoyancy effect tends to pull the inflow stream down already. 0 0.5 1 1.5 2 0153045 Inlet Flow Angle (degree)Contaminant concentration (parts per thousand) Figure 4.10 Average Contaminant Concentratio n vs. Inlet Angle for Basic Configuration On Fig. 4.11, CRE decreases sign ificantly from the basic case (0o-inlet angle) to higher inlet angle then increases a little after 30o. Increasing the inlet angle (down) always causes negative effects on contaminant level control. Figure 4.12 shows how contaminant concentr ation changes as a function of outlet ratio HTR. It does not show a clear relatio nship between the two, with the average contaminant concentration going up and down as HTR increases, and all values were higher than the basic case w ith one low outlet only. How ever, the contaminant removal effectiveness is affected by the outlet ratio as shown on Fig. 4.13.

PAGE 69

61 0 0.5 1 1.5 2 2.5 0153045 Inlet Flow Angle (degree)CRE Figure 4.11 Contaminant Removal Effectivene ss vs. Inlet Angle for Basic Configuration 0 0.5 1 1.5 0102030405060708090 High Outlet to Total Outlet Ratio (HTR), %Contaminant concentration (parts per thousand) Figure 4.12 Average Contaminant Concentra tion vs. Outlet Ratio for Two-Exhaust Configuration

PAGE 70

62 Figure 4.13 shows that as HTR increases, the contaminant removal effectiveness (CRE) increases, after a slight drop-down at first from the basic case (HTR = 0%). This response can be used for controlling the c ontaminant concentrati on level, but with caution to reduce the negative effect of increasing contaminant level. 0 0.1 0.2 0.3 0.4 0.5 0.6 0102030405060708090 High Outlet to Total Outlet Ratio (HTR), %CRE Figure 4.13 Contaminant Removal Effectiven ess vs. Outlet Ratio for Two-Exhaust Configuration Table 4.4 shows a comparison of average essential thermal comfort parameters (air speed, temperature, and relative humidity ) for different inlet angles. As inlet angle increases, average air speed ranges in 0. 2 m/s – 0.4 m/s, average temperature, 20oC – 23oC, and relative humidity, 61% – 69%. Table 4.5 shows a comparison of air speed, temperature, and relative humidity for different high outlet to total outlet ratio (HTR).

PAGE 71

63 Table 4.4 Average Air Speed, Temperature, Relative Humidity vs. Inlet Angle for Basic Configuration Inlet angle, A1 0o 5o 15o 30o 45o Air speed (m/s) 0.25 0.43 0.35 0.36 0.27 Temperature (oC) 22.1 20.6 22.9 20.9 23.0 Relative humidity (%) 61.2 67.7 66.0 68.5 61.6 Table 4.5 Average Air Speed, Temperature, Relative Humidity vs. Outlet Ratio for Two-Exhaust Configuration Outlet Ratio, HTR 14.3% 28.6% 42.9% 57.1% 71.4% 85.7% Air speed (m/s) 0.41 0.46 0.59 0.35 0.35 0.33 Temperature (oC) 23.7 22.5 21.9 23.1 20.6 23.6 Relative humidity (%) 60.1 60.9 62.3 69.2 68.7 67.0 As HTR increases and 0o-inlet angle, average air speed ranges in 0.3 m/s – 0.6 m/s, average temperature, 20oC – 24oC, and relative humidity, 60% – 69%.These parameters are in the reasonable range for an operating room in hospital [18, 21]. It can be expected that the change of inlet angle and HTR does not affect the thermal comfort of the room very significantly, and HTR has str onger affect than that of inlet angle. For assessing the thermal comfort level of the room, we consider its thermal sensation index and predicted mean vote (P MV). Figures from 4.13 to 4.16 show the thermal sensation and PMV as functions of in let angle and outlet ratio. Both of them are parameters for assessing thermal comfort, but at slightly different viewpoints. Thermal

PAGE 72

64 sensation expresses the correlation between co mfort level, temperature, humidity, sex, and length of exposure, wh ile predicted mean vote (PMV ) provides a measure of how people are likely to respond to different e nvironments based on the conditions of a particular individual including metabolic ra te, clothing, and air velocity besides temperature and humidity. Thermal sensation and PMV use the ASHRAE scale, which is an index from -3 (very cold) through 0 (neutr al) to +3 (very hot). The comfort zone can be taken from -1 (slightly cool) to +1 (s lightly warm), which is shown on Figures 4.9 through 4.12 as two level dash-dotted lines. Figure 4.14 shows how the thermal sensat ion changes as inlet angle changes. There is a decrease at first and then the thermal sensation slightly increases. It suggests that the thermal sensation does not depend mu ch on inlet angle. The thermal sensation curve was in the limit of the comfort zone, left a wide margin for design, which may raise the curve deeper inside the comfort zone. Figure 4.15 shows the predicted mean vote (PMV) for patient and staff as inlet angle changes. It decreases at first as inlet angle increases up to about 5 then increases as inlet angle increases, ranging from cold to slightly cool for patient and slightly cool to slightly warm for staff. The thermal sensa tion index from Fig. 4.13, as a factor for assessing the environment generally can be observed to be laying between PMV curves for patient and staff. At any in let angle, staff was always in comfort zone while patient is not comfortable (cold). Figures 4.16 and 4.17 present the therma l sensation and PMV as functions of outlet ratio HTR. Although they are not very se nsitive to outlet ratio, it can be found that their slight variations are quite interesting.

PAGE 73

65 -3 -2 -1 0 1 2 3 0153045 Inlet Flow Angle (degree)Thermal Sensation Comfort zone Figure 4.14 Thermal Sensation Index vs. Inlet Angle for Basic Configuration -3 -2 -1 0 1 2 3 0153045 Inlet Flow Angle (degree)PMV Staff Patient Comfort zone Figure 4.15 Predicted Mean Vote vs. Inlet Angle for Basic Configuration

PAGE 74

66 -3 -2 -1 0 1 2 3 0102030405060708090 High Outlet to Total Outlet Ratio (HTR), %Thermal Sensation Comfort zone Figure 4.16 Thermal Sensation Index vs. Ou tlet Ratio for Two-E xhaust Configuration -3 -2 -1 0 1 2 3 0102030405060708090 High Outlet to Total Outlet Ratio (HTR), %PMV Staff Patient Comfort zone Figure 4.17 Predicted Mean Vote vs. Outle t Ratio for Two-Exhaust Configuration

PAGE 75

67 As HTR increases, thermal sensation incr eases at first then decreases, then increases again. The same response was also observed on PMV curves of patient and staff. Similar to the case of varying inlet an gle, staff was always in comfort zone while patient is not comfort (cold). The average temperature and relative hum idity were compared with those from experimental data given in [18, 21] in Table 4.6. Table 4.6 Comparison of Average Temper ature and Relative Humidity to Experimental Data Simulation results Simulation number 1 2 3 4 5 6 7 8 9 10 11 Temperature (oC) 22 21 23 21 23 24 23 22 23 21 24 Relative humidity (%) 61 68 66 69 62 60 61 62 69 69 67 Experimental data from [21], based on 2 operating rooms Temperature (oC) Ranges from 19.5 to 25 Relative humidity (%) Ranges from 24% to 63.5% Experimental data from [18], based on 20 operating rooms Temperature (oC) Ranges from 18.6 to 24.5 Relative humidity (%) Ranges from 27% to 53% The data from [21] and [18] are colle cted from 2 and 20 operating rooms, respectively. The average temperature from the numerical simulati on shows reasonably good agreement with experimental data. Th e average relative humidity from CFD

PAGE 76

68 solution is higher than that of experimental data. The relative humidity is very sensitive to the changes of temperature and water va por concentration. The averaging for CFD solution included all regions inside the room, while the experimental data were collected at some specifics locations near the work ing spaces, where the temperature was higher thus relative humidity was lower. Besides, th ere are some points in experimental data where relative humidity values were higher than the CFD average value.

PAGE 77

69 Chapter 5 Conclusions and Recommendations 5.1 Simulation of the Office Room Two air distribution systems widely us ed for office rooms with typical and alternative cases were consider ed, showing the responses of ai rflow with different system and setup. Both simulated cases showed co mparable thermal sensation. PMV was close to comfort zone in both cases. The comparison of results from two simulations shows that the UFAD system has some advantages to overhead system, especially in contaminant removal. Improvement in indoor air quality wa s expected by delive ring the fresh supply air near the occupant at floor level, allowing an overall floor -to-ceiling airflow pattern to more efficiently remove contaminants from the occupied zone of the cubicle. Comparison to experimental data shows good agreem ent among systems of similar airflow characteristics. The simulation results sugge st that CFD modeling can be satisfactory used for predicting airflows in an office. 5.2 Simulation of the Operating Room The CFD simulations gave a good understanding of multi-component flow in an operating room. From the above discussion, it was found that the chan ge of inlet angle down could have negative effects on the cont aminant removal characteristics of the OR.

PAGE 78

70 The basic case of 0 inlet angle might be the best choi ce. Two-exhaust configurations can be employed for improving CRE with careful considerations since it may increase the total level of contaminant concentration. A bad choice of higher outlet size may cause more trouble than benefit. The lower outle t should still be the main outlet while the higher one can be considered as a regulating mechanism. A ratio of higher outlet-to-lower outlet area at about 0.7 would improve the c ontaminant removal characteristics without raising the contaminant level too high. Inlet a ngle and outlet ratio are two main factors to control the contaminant level and need to be selected concurrently. Thermal comfort factors, however, are not greatly affected by inlet angle and outlet ratio. In OR’s, it seems that the patient always feel colder than the staff. Since thermal comfort for patient and staff vary in a narrow range, the inlet temperat ure can be raised a few degrees to make the thermal sensation and PMV go into the comfor t zone for both, with the patient at the lower limit and the staff at the higher limit of the comfort zone. 5.3 Recommendations For improving the CFD modeling to simulate better the real phenomena of airflow and heat transfer in real life air conditioned rooms, the following approach can be considered: Three-dimensional modeling: 3-D model w ill show better the space interaction of the fluid flow and heat transfer phenomenon. Taking into account the equipments in the r ooms as obstacles to the fluid flow as well as heat transfer surfaces where needed. This will give distribution of the parameters of interest closer to the real environment.

PAGE 79

71 References [1] Woods J.E., 2004, “What real-world expe rience says about UFAD alternatives,” ASHRAE Journal, 46 (2), pp. 3-15. [2] Webster T.W., Bauman F., and Reese J., 2002, “Underfloor air di stribution: thermal stratification,” ASHRAE Journal, 44 (5), pp. 28-33. [3] Stanke D., 2001, “Turning air distribution upside down… underfloor air distribution,” Engineers Newsletter, 30 (4), http://www.trane.com/ [4] Halza J.M., 2003, “Underfloor & overhead ductless VAV systems,” ASHRAE Journal, 45 (11), pp. 43-48. [5] Fukao H., Oguro M., Ichihara M., and Tanabe S., 2002, “Comparison of underfloor vs. overhead air distribution systems in an office building,” ASHRAE Transactions, 104 (1), pp. 64-76. [6] Webster T., Bannon R., and Lehrer D., 2002, “Teledesic broadband center field study,” Center for the Built Environmen t (CBE), Summary Report April 2002. [7] Chow W.F., and Fung W.F., 1996, “Numeri cal studies on indoor air flow in the occupied zone of ventilated and air-c onditioned space,” Building and Environment, 31, pp. 319-344. [8] Bauman F.S., 1999, “Giving occupants wh at they want: guidelines for implementing personal environmental control in your build ing,” Proceedings at World Workplace 99, Los Angeles, CA. [9] Emmerich S.J., 1997, “Use of computa tional fluid dynamics to analyze indoor air quality issues,” NISTIR 5997, Building and Fi re Research Laborator y, National Institute of Standards and Technology, Gaithersburg, MD. [10] Gadgil A.J., Finlayson E.U., Hong K.H ., and Sextro R.G., 1999, “Commercial CFD software capabilities for modeling a pulse re lease of pollutant in a large indoor space,” Proceedings of Indoor Air ‘99, Edinburgh, 4, pp. 749. [11] Hirnikel D.J., Lipowicz P.J., and Lau R.W., 2002, “Predicting contaminant removal effectiveness of three air distribu tion systems by CFD modeling,” ASHRAE Transactions, 108 (1), pp. 350-359.

PAGE 80

72 [12] Fanger P.O., 1970, Thermal Comfort analysis and app lications in environmental engineering, McGraw-Hill, New York. [13] Rohles F.H. Jr. and Nevins R.G., 1971, “T he nature of thermal comfort for sedentary man,” ASHRAE Transactions, 77 (1), pp. 239-244. [14] ASHRAE, 1997, ASHRAE Handbook of Fundamentals, American Society of Heating, Refrigerating and Air Conditioni ng Engineers, Inc., Atlanta, Georgia. [15] Guan Y., Hosni M., Jones B.W., and Giel da T.P., 2003, “Liter ature review of the advances in thermal comfort m odeling,” ASHRAE Transactions, 109 (2), pp. 908-916. [16] AIA, 2001, Guidelines for Design and Construction of Hospitals and Health Care facilities, American Institute of Architects, Washington, D.C. [17] Woods J.E., Brayman D.T., Rasmussen R.W., Reynolds P.E., and Montag G.M., 1986, “Ventilation requirements in hospital operating rooms Part I: Control of airborne particles,” ASHRAE Transactions, 92 (2), pp. 396-426. [18] Balaras C.A., Dascalaki E., Argiriou A. A., and Gaglia A., 2002, “HVAC Systems in indoor conditions in Hell enic hospital operating rooms,” ASHRAE Transactions, 108 (2), pp. 23-38. [19] Lewis J.R., 1993, “Operating room air distribution effectiveness,” ASHRAE Transactions, 99 (2), pp. 1191-1199. [20] Memarzadeh F., 2000, “Methodology for minimizing risk from airborne organisms in hospital isolation room s,” ASHRAE Transactions, 106 (2), pp. 731-742. [21] Mora R., English M., and Athienitis A., 2001, “Asse ssment of thermal comfort during surgical operations,” ASHRAE Transactions, 107 (1), pp. 52-62. [22] Memarzadeh F. and Manning A., 200 0, “Thermal comfort, uniformity, and ventilation effectiveness in patient room s: performance assessment using ventilation indices,” ASHRAE Transactions, 106 (2), pp. 748-761. [23] Memarzadeh F. and Manning A., 2002, “C omparison of operating room ventilation systems in protection of the surgic al site,” ASHRAE Transactions, 108 (2), pp. 3-15. [24] Ho S.H., Rosario L., Rahman M.M., 2004, “Predictions of relative humidity and temperature in an operating room,” Proceedings of IMECE 2004: ASME International Mechanical Engineering Congre ss and RD&D Expo, on publishing.

PAGE 81

73 Appendices

PAGE 82

74 Appendix A: FIDAP Program for Office Room Simulation / file name: unf01.txt TITLE AIR FLOW in OFFICE ROOM UNDERFLOOR AIR DISTRIBUTION // FI-GEN FI-GEN( ELEM = 1, POIN = 1, CURV = 1, SURF = 1, NODE = 0, MEDG = 1, MLOO = 1, MFAC = 1, BEDG = 1, SPAV = 1, MSHE = 1, MSOL = 1, COOR = 1 ) $CNT = 1 / Lengths in X and Y direction $NLX = 13 DECLARE $LX[1:$NLX] $LX[1] = 0.05 $LX[2] = 0.20 $LX[3] = 0.20 $LX[4] = 0.25 $LX[5] = 0.10 $LX[6] = 0.25 $LX[7] = 0.05 $LX[8] = 0.20 $LX[9] = 0.16 $LX[10] = 0.54 $LX[11] = 0.60 $LX[12] = 0.35 $LX[13] = 0.40 $NLY = 12 DECLARE $LY[1:$NLY] $LY[1] = 0.50 $LY[2] = 0.10 $LY[3] = 0.10 $LY[4] = 0.10 $LY[5] = 0.15 $LY[6] = 0.05 $LY[7] = 0.35 $LY[8] = 0.05 $LY[9] = 0.35 $LY[10] = 0.95 $LY[11] = 0.50 $LY[12] = 0.40 / Generate numbers of intervals DECLARE $MX[1:$NLX] DECLARE $MY[1:$NLY] $ALPHA = 1.35 $L1 = 0.002 DO( $CNT = 1, $CNT .LE. $NLX ) $MX[$CNT] = 2*INT(1+LOG(1+($ALPHA-1)*0.5*$LX[$CNT]/$L1)/LOG($ALPHA)) ENDDO $MX[10] = INT(1+LOG(1+($ALPHA-1)*$LX[10]/$L1)/LOG($ALPHA))

PAGE 83

75 Appendix A: (Continued) DO( $CNT = 1, $CNT .LE. $NLY ) $MY[$CNT] = 2*INT(1+LOG(1+($ALPHA-1)*0.5*$LY[$CNT]/$L1)/LOG($ALPHA)) ENDDO / Generate coordinates $NX = 12 DECLARE $XP[1:$NX] $XP[1] = 0 DO( $CNT = 1, $CNT .LT. $NX-1 ) $XP[$CNT+1] = $XP[$CNT] + $LX[$CNT] ENDDO $XP[12] = 0.10 $NY = 12 DECLARE $YP[1:$NY] $YP[1] = 0 DO( $CNT = 1, $CNT .LT. $NY-1 ) $YP[$CNT+1] = $YP[$CNT] + $LY[$CNT] ENDDO $YP[12] = 1.00 // ADD POINTS POINT( ADD, COOR ) $XP[1] $YP[1] $XP[1] $YP[5] $XP[1] $YP[6] $XP[1] $YP[7] $XP[1] $YP[8] $XP[1] $YP[9] $XP[1] $YP[10] $XP[1] $YP[11] $XP[2] $YP[5] $XP[2] $YP[6] $XP[2] $YP[7] $XP[2] $YP[8] $XP[2] $YP[9] $XP[2] $YP[10] $XP[2] $YP[11] $XP[12] $YP[1] $XP[12] $YP[2] $XP[12] $YP[3] $XP[12] $YP[4] $XP[3] $YP[8] $XP[3] $YP[11] $XP[4] $YP[5] $XP[4] $YP[6] $XP[4] $YP[7] $XP[4] $YP[8]

PAGE 84

76 Appendix A: (Continued) $XP[4] $YP[9] $XP[4] $YP[10] $XP[4] $YP[11] $XP[5] $YP[1] $XP[5] $YP[2] $XP[5] $YP[3] $XP[5] $YP[4] $XP[5] $YP[5] $XP[5] $YP[11] $XP[6] $YP[3] $XP[6] $YP[4] $XP[6] $YP[5] $XP[6] $YP[6] $XP[6] $YP[7] $XP[6] $YP[8] $XP[6] $YP[9] $XP[6] $YP[10] $XP[6] $YP[11] $XP[7] $YP[12] $XP[7] $YP[9] $XP[7] $YP[11] $XP[8] $YP[1] $XP[8] $YP[2] $XP[8] $YP[12] $XP[8] $YP[9] $XP[8] $YP[10] $XP[8] $YP[11] $XP[9] $YP[1] $XP[9] $YP[11] $XP[10] $YP[1] $XP[10] $YP[11] $XP[11] $YP[1] $XP[11] $YP[2] $XP[11] $YP[12] $XP[11] $YP[9] $XP[11] $YP[10] $XP[11] $YP[11] // ADD LINES POINT( SELE, ID) 1 8 CURVE( ADD, LINE ) POINT( SELE, ID) 9 15

PAGE 85

77 Appendix A: (Continued) CURVE( ADD, LINE ) POINT( SELE, ID) 16 19 CURVE( ADD, LINE ) POINT( SELE, ID) 22 28 CURVE( ADD, LINE ) POINT( SELE, ID) 29 33 CURVE( ADD, LINE ) POINT( SELE, ID) 35 43 CURVE( ADD, LINE ) POINT( SELE, ID) 44 45 CURVE( ADD, LINE ) POINT( SELE, ID) 47 52 CURVE( ADD, LINE ) POINT( SELE, ID) 57 62 CURVE( ADD, LINE ) POINT( SELE, ID) 1 16 29 47 53 55 57 CURVE( ADD, LINE ) POINT( SELE, ID) 30 48 CURVE( ADD, LINE ) POINT( SELE, ID) 31 35 CURVE( ADD, LINE ) POINT( SELE, ID) 19 32

PAGE 86

78 Appendix A: (Continued) CURVE( ADD, LINE ) POINT( SELE, ID) 2 9 CURVE( ADD, LINE ) POINT( SELE, ID) 22 33 37 CURVE( ADD, LINE ) POINT( SELE, ID) 10 23 CURVE( ADD, LINE ) POINT( SELE, ID) 11 24 CURVE( ADD, LINE ) POINT( SELE, ID) 44 49 CURVE( ADD, LINE ) POINT( SELE, ID) 12 20 25 CURVE( ADD, LINE ) POINT( SELE, ID) 41 45 50 CURVE( ADD, LINE ) POINT( SELE, ID) 8 15 21 28 34 43 46 52 54 56 62 CURVE( ADD, LINE )

PAGE 87

79 Appendix A: (Continued) // ADD SURFACES POINT(SELE, ID ) 8 62 1 57 SURFACE( ADD, POIN, ROWW = 2 ) // ADD MESH EDGES CURVE( SELE, ID ) 55 65 MEDGE( ADD, FRTL, INTE = $MX[1], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 58 59 MEDGE( ADD, FRTL, INTE = $MX[13], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 61 66 MEDGE( ADD, FRTL, INTE = $MX[2], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 62 67 MEDGE( ADD, FRTL, INTE = $MX[3], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 56 68 MEDGE( ADD, FRTL, INTE = $MX[4], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 53 57 69 MEDGE( ADD, FRTL, INTE = $MX[5], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 63 70 MEDGE( ADD, FRTL, INTE = $MX[6], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 60 64 71 MEDGE( ADD, FRTL, INTE = $MX[7], RATI = $L1, 2RAT = $L1, PCEN = 0 )

PAGE 88

80 Appendix A: (Continued) CURVE( SELE, ID ) 49 72 MEDGE( ADD, FRTL, INTE = $MX[8], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 50 73 MEDGE( ADD, FRTL, INTE = $MX[9], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 51 74 MEDGE( ADD, FRTL, INTE = $MX[10], RATI = $L1, 2RAT = 0, PCEN = 0 ) CURVE( SELE, ID ) 47 54 MEDGE( ADD, FRTL, INTE = $MX[11], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 48 52 MEDGE( ADD, FRTL, INTE = $MX[12], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 14 23 36 41 MEDGE( ADD, FRTL, INTE = $MY[1], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 15 24 MEDGE( ADD, FRTL, INTE = $MY[2], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 16 25 27 MEDGE( ADD, FRTL, INTE = $MY[3], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 26 28 MEDGE( ADD, FRTL, INTE = $MY[4], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 2 8 17 29

PAGE 89

81 Appendix A: (Continued) MEDGE( ADD, FRTL, INTE = $MY[5], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 3 9 18 30 MEDGE( ADD, FRTL, INTE = $MY[6], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 4 10 19 31 MEDGE( ADD, FRTL, INTE = $MY[7], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 5 11 20 32 MEDGE( ADD, FRTL, INTE = $MY[8], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 6 12 21 33 39 44 MEDGE( ADD, FRTL, INTE = $MY[9], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 7 13 22 34 40 45 MEDGE( ADD, FRTL, INTE = $MY[10], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 37 42 MEDGE( ADD, FRTL, INTE = $MY[11], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 35 38 43 MEDGE( ADD, FRTL, INTE = $MY[12], RATI = $L1, 2RAT = $L1, PCEN = 0 ) // ADD MESH LOOPS

PAGE 90

82 Appendix A: (Continued) CURVE( SELE, ID ) 55 2 7 65 13 12 11 10 9 8 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 6, EDG3 = 1, EDG4 = 6 ) CURVE( SELE, ID ) 58 9 59 18 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) CURVE( SELE, ID ) 62 61 11 13 66 67 22 21 20 MLOOP( ADD, MAP, EDG1 = 2, EDG2 = 3, EDG3 = 2, EDG4 = 3 ) CURVE( SELE, ID ) 57 56 17 22 68 69 34 33 32 31 30 29 MLOOP( ADD, MAP, EDG1 = 2, EDG2 = 6, EDG3 = 2, EDG4 = 6 ) CURVE( SELE, ID ) 53 25 26 57 28 27 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 2, EDG3 = 1, EDG4 = 2 ) CURVE( SELE, ID ) 64 63

PAGE 91

83 Appendix A: (Continued) 33 34 70 71 40 39 MLOOP( ADD, MAP, EDG1 = 2, EDG2 = 2, EDG3 = 2, EDG4 = 2 ) CURVE( SELE, ID ) 60 35 64 38 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) CURVE( SELE, ID ) 47 14 16 54 25 24 23 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 3, EDG3 = 1, EDG4 = 3 ) CURVE( SELE, ID ) 48 23 52 36 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) CURVE( SELE, ID ) 51 50 49 36 40 72 74 45 44 43 42 41 MLOOP( ADD, MAP, EDG1 = 3, EDG2 = 5, EDG3 = 3, EDG4 = 5 ) // ADD MESH FACES DO( $CNT = 1, $CNT .LE. 10 ) SURFACE( SELE, ID = 1 ) MLOOP( SELE, ID = $CNT ) MFACE( ADD ) ENDDO // GENERATE MESH MFACE( SELE, ALL )

PAGE 92

84 Appendix A: (Continued) MFACE( MESH, MAP, ENTI = "air" ) // BOUNDARY ENTITIES ELEMENT( SETD, EDGE, NODE = 2 ) MEDGE( SELE, ID = 24 ) MEDGE( MESH, MAP, ENTI = "outlet" ) MEDGE( SELE, ID = 21 ) MEDGE( MESH, MAP, ENTI = "inlet" ) MEDGE( SELE, ID ) 62 32 69 72 61 67 MEDGE( MESH, MAP, ENTI = "symmetry" ) MEDGE( SELE, ID ) /ceiling 2 8 10 22 6 /lights 13 15 18 20 /floor 25 27 19 23 /person 11 37 39 43 47 51 55 14 70 /chair 16 68 28 34

PAGE 93

85 Appendix A: (Continued) /desk 29 33 35 26 38 9 1 /computer 41 42 3 4 49 50 5 7 /panel 40 44 48 52 56 MEDGE( MESH, MAP, ENTI = "nonslip" ) MEDGE( SELE, ID ) 11 37 39 43 47 51 55 14 70 MEDGE( MESH, MAP, ENTI = "person" ) MEDGE( SELE, ID ) 5 7 MEDGE( MESH, MAP, ENTI = "hottop" ) MEDGE( SELE, ID ) 13 15 18 20 MEDGE( MESH, MAP, ENTI = "lights" ) MEDGE( SELE, ID ) 25 27 19 23 MEDGE( MESH, MAP, ENTI = "floor" ) END

PAGE 94

86 Appendix A: (Continued) FIPREP / SI units / Reference temperature: 22 oC = 295 K / Underfloor air distribution: $U2 m/s $U2 = 1 DENSITY( CONS = 1.1967 ) VISCOSITY( CONS = 1.8273E-5 ) SPECIFICHEAT( CONS = 1.0043E3 ) CONDUCTIVITY( CONS = 2.5776E-2 ) VOLUMEX( CONS = 3.3932E-3, REFTEMP = 22 ) GRAVITY( MAGNITUDE = 9.8 ) DIFFUSIVITY( SET = "H2O", CONS = 2.5448E-5 ) DIFFUSIVITY( SET = "NH3", CONS = 2.5033E-5 ) ENTITY( FLUI, NAME = "air", SPEC = 1, MDIFF = "H2O", SPEC = 2, MDIFF = "NH3" ) ENTITY( PLOT, NAME = "outlet" ) ENTITY( PLOT, NAME = "inlet" ) ENTITY( PLOT, NAME = "symmetry" ) ENTITY( PLOT, NAME = "nonslip" ) ENTITY( PLOT, NAME = "person" ) ENTITY( PLOT, NAME = "hottop" ) ENTITY( PLOT, NAME = "lights" ) ENTITY( PLOT, NAME = "floor" ) BCNODE( VELO, ENTI = "inlet", CONS, X = 0, Y = $U2 ) BCNODE( VELO, ENTI = "nonslip", ZERO ) BCNODE( UX, ENTI = "symmetry", ZERO ) BCNODE( TEMP, ENTI = "inlet", CONS = 20 ) BCNODE( TEMP, ENTI = "person", CONS = 33 ) BCFLUX( HEAT, ENTI = "hottop", CONS = 100 ) BCFLUX( HEAT, ENTI = "lights", CONS = 75 ) BCNODE( SPEC = 1, ENTI = "inlet", CONS = 0.011 ) BCFLUX( SPEC = 1, ENTI = "person", CONS = 5E-7 ) BCNODE( SPEC = 2, ENTI = "inlet", CONS = 0 ) BCFLUX( SPEC = 2, ENTI = "floor", CONS = 1E-6 ) CLIPPING( MINI ) 0 0 0 0 20 0 0 0 1.E-20 1.E-20 CLIPPING( MAXI ) 0 0 0 0 0 0 0 0 1. 1. RENUMBER( PROFILE ) DATAPRINT( NONE ) PRINTOUT( NONE, NOBO ) OPTIONS( UPWI ) EXECUTION( NEWJ )

PAGE 95

87 Appendix A: (Continued) PRESSURE( PENA = 1.E-7, DISC ) PROBLEM( 2-D, NONL, MOME, BUOY ) SOLUTION( S.S. = 1000, VELC = 0.02, RESC = 0.02, ACCF = 0.5 ) /ICNODE( VELO, READ, ALL ) /EXECUTION( NEWJ ) /PROBLEM( 2-D, NONL, NOMO, SPEC = 1, SPEC = 2 ) END CREATE( FISOLV ) RUN( FISOLV, IDENT = "unf01a", BACK ) /RUN( FISOLV, IDENT = "unf01z", REST = "unf01a.FDPOST", BACK )

PAGE 96

88 Appendix B: FIDAP Program fo r Operating Room Simulation / Input file for mixed convection in OR, SI units / file name: 3500.txt / basic configuration, single outlet OL = 35, angle 0 deg. TITLE AIRFLOW in O.R. Simulation No. SI-35-00 // FI-GEN FI-GEN( ELEM = 1, POIN = 1, CURV = 1, SURF = 1, NODE = 0, MEDG = 1, MLOO = 1, MFAC = 1, BEDG = 1, SPAV = 1, MSHE = 1, MSOL = 1, COOR = 1 ) $NX = 10 DECLARE $X_VALS[1:$NX] $X_VALS[1] = 0 $X_VALS[2] = 1.75 $X_VALS[3] = 2.00 $X_VALS[4] = 2.10 $X_VALS[5] = 2.70 $X_VALS[6] = 3.30 $X_VALS[7] = 3.90 $X_VALS[8] = 4.00 $X_VALS[9] = 4.25 $X_VALS[10] = 6.00 $NY = 11 DECLARE $Y_VALS[1:$NY] $Y_VALS[1] = 3.50 $Y_VALS[2] = 3.20 $Y_VALs[3] = 3.00 $Y_VALS[4] = 2.85 $Y_VALS[5] = 2.55 $Y_VALS[6] = 1.75 $Y_VALS[7] = 1.05 $Y_VALS[8] = 0.80 $Y_VALS[9] = 0.55 $Y_VALS[10] = 0.20 $Y_VALS[11] = 0 $NL = 18 DECLARE $LEN[1:$NL] DECLARE $MSH[1:$NL] $LEN[1] = 1.75 $LEN[2] = 0.25 $LEN[3] = 0.10 $LEN[4] = 0.60 $LEN[5] = 0.60 $LEN[6] = 1.80 $LEN[7] = 0.30 $LEN[8] = 0.20 $LEN[9] = 0.15 $LEN[10] = 0.30 $LEN[11] = 0.80 $LEN[12] = 1.55

PAGE 97

89 Appendix B: (Continued) $LEN[13] = 0.20 $LEN[14] = 0.70 $LEN[15] = 0.25 $LEN[16] = 0.60 $LEN[17] = 1.20 $LEN[18] = 0.35 $ALPHA = 1.1 $L1 = 0.01 DO( $I = 1, $I .LE. $NL ) $MSH[$I] = 2*INT(1+LOG(1+($ALPHA-1)*0.5*$LEN[$I]/$L1)/LOG($ALPHA)) ENDDO // Add Points //1st row $Y_VAL = $Y_VALS[1] DO( $I = 1, $I .LE. $NX ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //2nd row $Y_VAL = $Y_VALS[2] POINT( ADD, COOR, X = $X_VALS[1], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[$NX], Y = $Y_VAL ) //3rd row $Y_VAL = $Y_VALS[3] POINT( ADD, COOR, X = $X_VALS[1], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[$NX], Y = $Y_VAL ) //4th row $Y_VAL = $Y_VALS[4] DO( $I = 1, $I .LE. $NX ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //5th row $Y_VAL = $Y_VALS[5] DO( $I = 1, $I .LE. $NX ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //6th row $Y_VAL = $Y_VALS[6] DO( $I = 1, $I .LE. $NX ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //7th row $Y_VAL = $Y_VALS[7] DO( $I = 3, $I .LE. 8 )

PAGE 98

90 Appendix B: (Continued) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //8th row $Y_VAL = $Y_VALS[8] POINT( ADD, COOR, X = $X_VALS[3], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[4], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[7], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[8], Y = $Y_VAL ) //9th row $Y_VAL = $Y_VALS[9] POINT( ADD, COOR, X = $X_VALS[9], Y = $Y_VAL ) POINT( ADD, COOR, X = $X_VALS[10], Y = $Y_VAL ) //10th row $Y_VAL = $Y_VALS[10] DO( $I = 1, $I .LE. 4 ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO DO( $I = 7, $I .LE. 10 ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO //11th row $Y_VAL = $Y_VALS[11] DO( $I = 1, $I .LE. 4 ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO DO( $I = 7, $I .LE. 10 ) POINT( ADD, COOR, X = $X_VALS[$I], Y = $Y_VAL ) ENDDO // Add Lines POINT( SELE, ID) 1 10 CURVE( ADD, LINE ) POINT( SELE, ID) 15 24 CURVE( ADD, LINE ) POINT( SELE, ID) 25 34 CURVE( ADD, LINE ) POINT( SELE, ID) 35 44 CURVE( ADD, LINE ) POINT( SELE, ID) 45 50 CURVE( ADD, LINE )

PAGE 99

91 Appendix B: (Continued) POINT( SELE, ID) 51 54 CURVE( ADD, LINE ) POINT( SELE, ID) 57 64 CURVE( ADD, LINE ) POINT( SELE, ID) 65 72 CURVE( ADD, LINE ) POINT( SELE, ID) 1 11 13 15 25 35 57 65 CURVE( ADD, LINE ) POINT( SELE, ID) 10 12 14 24 34 44 56 64 72 CURVE( ADD, LINE ) POINT( SELE, ID) 19 29 CURVE( ADD, LINE ) POINT( SELE, ID) 20 30 CURVE( ADD, LINE ) POINT( SELE, ID) 46 52 CURVE( ADD, LINE ) POINT( SELE, ID) 49 53 CURVE( ADD, LINE )

PAGE 100

92 Appendix B: (Continued) POINT( SELE, ID) 37 45 51 59 CURVE( ADD, LINE ) POINT( SELE, ID) 42 50 54 62 CURVE( ADD, LINE ) POINT( SELE, ID) 36 58 CURVE( ADD, LINE ) POINT( SELE, ID) 43 55 63 CURVE( ADD, LINE ) //Add Surfaces POINT(SELE, ID ) 1 10 65 72 SURFACE( ADD, POIN, ROWW = 2 ) //Add Mesh Edges CURVE( SELE, ID ) 1 10 19 28 45 52 9 18 27 36 51 58 MEDGE( ADD, FRTL, INTE = $MSH[1], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 2 11

PAGE 101

93 Appendix B: (Continued) 20 29 46 53 8 17 26 35 50 57 MEDGE( ADD, FRTL, INTE = $MSH[2], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 3 12 21 30 37 42 47 54 7 16 25 34 41 44 49 56 MEDGE( ADD, FRTL, INTE = $MSH[3], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 4 13 22 31 38 MEDGE( ADD, FRTL, INTE = $MSH[4], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 6 15 24 33 40 MEDGE( ADD, FRTL, INTE = $MSH[4], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 5 14 23 32 39

PAGE 102

94 Appendix B: (Continued) MEDGE( ADD, FRTL, INTE = $MSH[5], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 43 48 55 MEDGE( ADD, FRTL, INTE = $MSH[6], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 59 66 MEDGE( ADD, FRTL, INTE = $MSH[7], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 60 67 MEDGE( ADD, FRTL, INTE = $MSH[8], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 61 68 MEDGE( ADD, FRTL, INTE = $MSH[9], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 62 74 75 69 MEDGE( ADD, FRTL, INTE = $MSH[10], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 63 70 MEDGE( ADD, FRTL, INTE = $MSH[11], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 64 84 MEDGE( ADD, FRTL, INTE = $MSH[12], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 65 73 MEDGE( ADD, FRTL, INTE = $MSH[13], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 78 81 MEDGE( ADD, FRTL, INTE = $MSH[14], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 79 76

PAGE 103

95 Appendix B: (Continued) 77 82 MEDGE( ADD, FRTL, INTE = $MSH[15], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 80 83 MEDGE( ADD, FRTL, INTE = $MSH[16], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 85 71 MEDGE( ADD, FRTL, INTE = $MSH[17], RATI = $L1, 2RAT = $L1, PCEN = 0 ) CURVE( SELE, ID ) 86 72 MEDGE( ADD, FRTL, INTE = $MSH[18], RATI = $L1, 2RAT = $L1, PCEN = 0 ) //Add Mesh Loops CURVE( SELE, ID ) 61 60 59 1 9 66 68 18 17 16 15 14 13 12 11 10 MLOOP( ADD, MAP, EDG1 = 3, EDG2 = 9, EDG3 = 3, EDG4 = 9 ) CURVE( SELE, ID ) 62 10 13 74 22 21 20 19 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 4, EDG3 = 1, EDG4 = 4 ) CURVE( SELE, ID ) 75 15 18 69 27

PAGE 104

96 Appendix B: (Continued) 26 25 24 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 4, EDG3 = 1, EDG4 = 4 ) CURVE( SELE, ID ) 63 19 27 70 36 35 34 33 32 31 30 29 28 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 9, EDG3 = 1, EDG4 = 9 ) CURVE( SELE, ID ) 65 45 51 73 58 57 56 55 54 53 52 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 7, EDG3 = 1, EDG4 = 7 ) CURVE( SELE, ID ) 64 28 84 45 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) CURVE( SELE, ID ) 86 85 36 71 72 51 MLOOP( ADD, MAP, EDG1 = 2, EDG2 = 1, EDG3 = 2, EDG4 = 1 ) CURVE( SELE, ID ) 78 30 34 81 41

PAGE 105

97 Appendix B: (Continued) 40 39 38 37 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 5, EDG3 = 1, EDG4 = 5 ) CURVE( SELE, ID ) 80 42 44 83 49 48 47 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 3, EDG3 = 1, EDG4 = 3 ) CURVE( SELE, ID ) 79 37 76 42 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) CURVE( SELE, ID ) 77 41 82 44 MLOOP( ADD, MAP, EDG1 = 1, EDG2 = 1, EDG3 = 1, EDG4 = 1 ) //Add Mesh Faces DO( $I = 1, $I .LE. 11 ) SURFACE( SELE, ID = 1 ) MLOOP( SELE, ID = $I ) MFACE( ADD ) ENDDO //Meshing Mesh Faces ELEMENT( SETD, QUAD, NODE = 4 ) MFACE( SELE, ALL ) MFACE( MESH, MAP, ENTI = "air" ) //Mesh Map (Boundary Entities ) ELEMENT( SETD, EDGE, NODE = 2 ) MEDGE( SELE, ID ) 61 63 MEDGE( MESH, MAP, ENTI = "inlet" ) MEDGE( SELE, ID ) 86 MEDGE( MESH, MAP, ENTI = "outlet" ) MEDGE( SELE, ID ) / walls 59, 60

PAGE 106

98 Appendix B: (Continued) 62 64, 65 68, 71 84 73, 74 / ceiling 1 13 25 41 51 46 33 19 7 / floor 6 18 32 58 40 24 12 / surgical lights 66, 67 52, 53 / patient + table 78, 79 45 55, 56 50 / staffs 16, 17 72 75 77 81 22, 23 76 80 82, 83 85 MEDGE( MESH, MAP, ENTI = "nonslip" ) MEDGE( SELE, ID ) 59, 60 62 64, 65 68, 71 84 73, 74 MEDGE( MESH, MAP, ENTI = "walls" ) MEDGE( SELE, ID ) 66, 67

PAGE 107

99 Appendix B: (Continued) 52 MEDGE( MESH, MAP, ENTI = "lamp_back" ) MEDGE( SELE, ID ) 53 MEDGE( MESH, MAP, ENTI = "lamp_face" ) MEDGE( SELE, ID ) 78, 79 45 55 50 MEDGE( MESH, MAP, ENTI = "patient" ) MEDGE( SELE, ID ) 16, 17 72 75 77 81 22, 23 76 80 82, 83 85 MEDGE( MESH, MAP, ENTI = "staffs" ) END FIPREP $Un = 0.4 $Ang = 0 DENSITY( CONS = 1.1967 ) VISCOSITY( CONS = 1.8273e-05 ) SPECIFICHEAT( CONS = 1004.3 ) CONDUCTIVITY( CONS = 0.025776 ) VOLUMEXPANSION( CONS = 0.0033932, REFT = 22 ) GRAVITY( MAGN = 9.80665 ) DIFFUSIVITY( SET = "H2O", CONS = 2.5448e-05 ) DIFFUSIVITY( SET = "NH3", CONS = 2.5033e-05 ) ENTITY( FLUI, NAME = "air", SPEC = 1, MDIF = "H2O", SPEC = 2, MDIF = "NH3" ) ENTITY( PLOT, NAME = "outlet" ) ENTITY( PLOT, NAME = "inlet" ) ENTITY( PLOT, NAME = "nonslip" ) ENTITY( PLOT, NAME = "walls" ) ENTITY( PLOT, NAME = "lamp_back" ) ENTITY( PLOT, NAME = "lamp_face" ) ENTITY( PLOT, NAME = "patient" ) ENTITY( PLOT, NAME = "staffs" ) BCNODE( VELO, ENTI = "inlet", CONS, X = $Un*COS($Ang), Y = $Un*SIN($Ang) )

PAGE 108

100 Appendix B: (Continued) BCNODE( VELO, ENTI = "nonslip", ZERO ) BCNODE( TEMP, ENTI = "inlet", CONS = 17 ) BCNODE( TEMP, ENTI = "walls", CONS = 22 ) BCNODE( TEMP, ENTI = "patient", CONS = 33 ) BCNODE( TEMP, ENTI = "staffs", CONS = 33 ) BCFLUX( HEAT, ENTI = "lamp_back", CONS = 5 ) BCFLUX( HEAT, ENTI = "lamp_face", CONS = 100 ) BCNODE( SPEC = 1, ENTI = "inlet", CONS = 0.01018 ) BCFLUX( SPEC = 1, ENTI = "patient", CONS = 5e-07 ) BCFLUX( SPEC = 1, ENTI = "staffs", CONS = 8e-07 ) BCNODE( SPEC = 2, ENTI = "inlet", CONS = 0 ) BCFLUX( SPEC = 2, ENTI = "patient", CONS = 1e-05 ) CLIPPING( MINI ) 0 0 0 0 17 0 0 0 1.E-20 1.E-20 CLIPPING( MAXI ) 0 0 0 0 0 0 0 0 1. 1. DATAPRINT( NONE ) PRINTOUT( NONE ) OPTIONS( UPWI ) EXECUTION( NEWJ ) PRESSURE( PENA = 1e-07, DISC ) PROBLEM( 2-D, NONL, MOME, BUOY ) SOLUTION( S.S. = 1000, VELC = 0.02, RESC = 0.02, ACCF = 0.5 ) /ICNODE( VELO, READ, ALL ) /EXECUTION( NEWJ ) /PROBLEM( 2-D, NONL, NOMO, SPEC = 1, SPEC = 2 ) END CREATE( FISO ) RUN( FISOLV, IDENT = "3500a", BACK ) /RUN( FISOLV, IDENT = "3500z", REST = "3500a.FDPOST", BACK )


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001498271
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 041209s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000548
035
(OCoLC)57717801
9
AJU6876
b SE
SFE0000548
040
FHM
c FHM
090
TJ145 (ONLINE)
1 100
Ho, Son Hong.
0 245
Numerical simulation of thermal comfort and contaminant transport in air conditioned rooms
h [electronic resource] /
by Son Hong Ho.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (M.S.M.E.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 108 pages.
520
ABSTRACT: Health care facilities, offices, as well as workshops and other commercial occupancies, require ventilation and air conditioning for thermal comfort and removal of contaminants and other pollutions. A good design of ventilation and air conditioning provides a healthy and comfortable environment for patients, workers, and visitors. The increasing developments of computational fluid dynamics (CFD) in the recent years have opened the possibilities of low-cost yet effective method for improving HVAC systems in design phase, with less experiment required. This work presents numerical simulations of thermal comfort and contaminant removal for two typical working spaces where these factors are critical: a hospital operating room with various configurations of inlet and outlet arrangements, and an office with two cases of air distribution systems: underfloor and overhead, also with alternative cases. The 2-D simulation approach was employed.Temperature, relative humidity, contaminant concentration, thermal sensation, predicted mean vote (PMV), and contaminant removal factor were computed and used for assessing thermal comfort and contaminant removal characteristics of the office room and operating room. The result shows good agreements with experimental data taken from related literature.
590
Adviser: Rahman, Muhammad.
653
heat and mass transfer.
ventilation.
computational fluid dynamics.
relative humidity.
multi-component flow.
690
Dissertations, Academic
z USF
x Mechanical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.548