Modern manufacturing industries are increasingly turning to cleanroom technology for the development of new products. Products get contaminated and either malfunction or become hazardous to people without clean conditions. Therefore, its critical to validate that the cleanroom is working correctly and achieving the contamination standards that it has been designed to fulfill. There has been a substantial increase in the working environment of cleanroom. Special garments are therefore dressed in all cleanrooms to control particles contamination dispersed from personnel in cleanrooms. It will result in thermal comfort dissatisfaction. Therefore, its significant to validate that the cleanroom is working correctly and achieving the contamination standards without the sacrifice of thermal comfort.
Field-testing is to establish that the cleanroom performs correctly and achieves the contamination standards set down at the design stage. These standards are specified comprehensively in ISO 14644-1 (1999). A lot of valuable information describing cleanroom tests to evaluate and characterize the overall performance of cleanroom and clean zone system can be found in IEST-RP-CC006.2 (1993). It also contains the latest data and information on cleanroom testing methods and procedures. Furthermore, NEBB (1996) provides essential information on design consideration, requirements, techniques, equipments and comprehensive procedures for certified testing of cleanrooms.
Large amounts of contamination are dispersed from people in cleanrooms. The cleanroom industry has set considerable emphasis on minimizing the particles dispersed from cleanroom clothing. Accordingly, cleanroom clothing can sometimes be hot and uncomfortable. The thermal comfort of cleanroom garments should be assessed by comfortable indices. ISO 7730 Standard (1994) describes the predicted mean vote (PMV) and predicted percentage of dissatisfied (PPD) indices and specifies acceptable conditions for thermal comfort. More information on thermal comfort standard could be found in the literature (Olesen and Parsons, 2002). The local differences between body segments caused by high radiant temperature have been investigated with their effect on thermal comfort (Atmaca et al., 2007). Besides, the evaluation of the optimum temperature based on PMV/PPD in cabins including human factors, thermal resistance of maritime police uniform and metabolic rate have been studied comprehensively (Jang et al., 2007).
Three-dimension air flow field for improving the ventilation performance of
a minienvironment has been investigated in literature (Cheng and Hung, 2005).
They found that the numerical and experimental approaches lead to useful information
and the ventilation performance in the minienvironment may be improved enormously
by adjusting the location of the HEPA filter. The influence of external perturbations
on a minienvironment by experimental investigations was examined carefully (Rouaud
et al., 2004). It showed that the minienvironment and particularly the
air curtain are strongly sensitive to perturbations. Besides, the detailed experimental
measurements of air flow characteristics in a full-scale cleanroom have been
extensively examined (Hu et al., 1996). Velocity vectors, turbulent intensity,
turbulence kinetic energy and velocity contours in the space domain were also
presented. The specific experimental data are not only beneficial for cleanroom
design but also helpful for evaluating of modeling in cleanroom air flow fields.
Computational Fluid Dynamics (CFD) simulation technique is a scientific technique that allows improvement of cleanroom configuration without interfering with normal manufacturing processes (Manning, 2005). A minienvironment for controlling the process area from ambient air contamination by a buffer zone through parametric has been studied extensively using CFD method (Hu et al., 2002). In addition, a deterministic CFD model that incorporated fan-performance characteristics was also applied to investigate the air-recirculation performance of unidirectional flow cleanrooms. The CFD codes were successfully used to simulate the air currents and contamination decay in a model room as well as comparison of indoor particle concentration in different rooms (Zhao et al., 2004). Besides, the STAR-CD code was employed to compute the airflow, temperature and concentration distribution in the lecture room with mixing ventilation system (Noh et al., 2007). They performed the experimental and numerical studied on thermal comfort in the lecture room with cooling loads when the operating conditions were changed. Moreover, the numerical simulation using STAR-CD code has been conducted to investigate the dispersion of pollutant in a fan-filter-unit cleanroom (Chen et al., 2007). The pollutant hot spots and peak pollutant concentration could be obtained from the simulation results.
In this study, field tests of a cleanroom have been carried out in our newly constructed Micro elector-mechanical system (MEMS) laboratory. The ASHRAE thermal comfort code was conducted to investigate thermal comfort of personnel based on field-testing data consequently. Furthermore, the effects of clothing on thermal comfort and contamination control have been assessed comprehensively. The results from computer simulation and field tests indicated that there existed compromise between the PMV and airborne particle counts under different cleanroom garments.
SYSTEM DESCRIPTION AND FIELD TESTS
It becomes obvious that a cleanroom enclosure is essential for both research
projects and student training. The layout of cleanroom in MEMS laboratory of
Mechanical Engineering Department is shown in Fig. 1. This
facility with cleanliness level class 1000 (ISO class 6) consists of process
area, etching area and lithography area. The measured cleanroom with the dimension
of length(L)xwidth(W)xheight(H) = 20x6x2.4 m. Another change area with air shower/air
lock and visiting aisle is included in ancillary clean zone adjacent to the
main process area. Fan-Filter Unit (FFU) type cleanroom was designed with HEPA
filter of 36% coverage rate in process area. The specified design conditions
are temperature 22±2 (°C), humidity 55±5 (%RH) and pressurization
15 Pa for main process area of cleanroom.
To verify the effects of clothing type on particle counts and thermal comfort,
an environmental test section 1.3 m (L) x 1.3 m (W) x 2.2 m (H) along with an
air shower were set up at the upper left corner of the cleanroom (Fig.
1). The schematic diagram of environmental test section was shown in Fig.
2. To remove particles from clothing and reduce dispersion in test section,
personnel entered the air shower before testing. Figure 3a
represents the snapshot of test section and the full-scale geometric model of
the test section for CFD simulation has been established as shown in Fig.
3b. Types of garment system for tests including (a) attached hood coverage
(b) two piece suit and (c) cleanroom coat were shown in Fig. 4.
All of the different cases of cleanroom apparel combination with facemask and
gloves were presented in Table 1. The effect of clothing type
on contamination control and thermal comfort will be tested extensively. A certain
amount of trade-off between contamination control and thermal comfort may be
Field-tests including the particle counts at different occupancy state, airflow
volume of FFU, temperature level, turbulent intensity and pressurization were
carried out comprehensively. The airborne particle concentration tests were
performed to determine the actual particle count level within the cleanroom
at different occupancy state (at-rest and operational). Quantities measurements
of airborne particle counts were made with a Met-One Model 3313 particle counter,
sensitive to particles 0.3 μm, 0.5 μm or larger. The minimum number
of sampling locations was determined by the square root of clean zone area (m2)
according to ISO standard 14644-1.
||Layout of cleanroom in MEMS Laboratory
||Schematic diagram of environment testing section
Consequently, 13 sampling locations were evenly distributed at the process
area (Fig. 1). The airflow volume of each FFU was determined
by an ALNOR Model 720 flow measuring hood for accuracy and repeatability.
To provide reliable measurement data as the boundary conditions of CFD simulation,
the temperature and face velocity of the FFU has been tested with an ALNOR Model
8585 thermal (hot-wired) anemometer. Additionally, the environmental conditions
needed to evaluate the thermal comfort indices could be determined by an INNOVA
1221 thermal comfort data logger. Air temperature, mean radiant temperature
(MRT) and relative humidity could on-site be recorded simultaneously.
||ISO thermal sensation scale
||Snapshot and geometric model of test section (a) test section
(b) geometric model of test section for simulation
Furthermore, turbulence intensity is a very efficient parameter depicting the
degree of turbulence which is defined as the ratio of standard variation of
velocity fluctuation to the local mean velocity. Tests of turbulence intensity
were carried out using an ANEMOSONIC Model UA6 ultrasonic anemometer to assess
and verify the level of air flow disturbance.
||Types of garment system for testing (a) attached hood coverage
(b) two piece suit and (c) cleanroom coat
THERMAL COMFORT MODELING AND NUMERICAL SIMULATION
According to the ISO 7730 Standard, the PMV index specifies acceptable conditions
for thermal comfort. The PMV predicts the mean value of the votes of a large
group of people on ISO thermal sensation scale, as shown in Table
1, i.e., +3 = hot, +2 = warm, +1 = slightly warm, 0 = neutral, -1 = slightly
cool, -2 = cool, -3 = cold. PMV is derived from the physics of heat transfer
combined with an empirical fit to sensation. A commercial code developed by
ASHRAE (Fountain, 1995) was adapted to evaluated thermal comfort of test section.
The program predicts human thermal response to the environment using PMV-PPD
thermal comfort model and allows to calculating the predicted thermal comfort
for a human in space. Basic thermal comfort model parameters were shown in Fig.
5. The right-hand side of the screen allows access to the input variables
while the left-hand sides present the output from the model. All of the environment
conditions including air temperature, MRT, air velocity and relative humidity
were adopted base on field testing data. The personal condition of activity
defined by metabolic rate can be obtained easily from physiological data built
in program. Clothing is defined in terms of clo unit (1 clo = 0.155 m2-K/W).
The selection and generation of clothing ensembles also can be calculated by
programmed clo calculator automatically.
To asses the effect of different environment condition on thermal comfort index
under various face velocity of FFU, another commercial CFD code, STAR-CD (2001),
was used to simulate the temperature contours, velocity vector and turbulent
kinetic energy of test section accordingly. The governing equations solved by
STAR-CD include the three-dimensional time-dependent incompressible Navier-Stokes
equation and k-ε turbulence equations.
||Typical input and output ASHRAE thermal comfort model
These formulated equations can be found in the STAR-CD users manual
as well as any CFD text books and will not be repeated here. In the present
study, the full-scale geometric model of cleanroom is shown in Fig.
3b. It was assumed that the air flow field is homogenous, isotropic and
three-dimensional. For the k-ε turbulence equation, the empirical turbulence
coefficients were assigned as: σk = 1.0, σε
= 1.22, σε1 = 1.44, σε2 = 1.92 and
Cμ = 0.09, respectively. These values were widely accepted in CFD k-ε
model. Full-scale simulation has been conducted and compared under different
RESULTS AND DISCUSSION
Figure 6 shows the airborne particle counts at different
sample locations (Fig. 1) for particle size 0.5 μm and
larger. The flow rate of air sampled at each measurement is 1 cfm (28.3 L m¯3)
and the mean particle counts are recorded at each location for 3 times. It represents
that the cleanroom meets the specified class 1000 (ISO class 6) cleanliness
level at the state of at rest. However, at the specific sample location 13 in
right-hand side of cleanroom, the particle counts exceed the specified level
at operational state. Besides, to highlight the degree of turbulence especially
at high particle concentration area, an ultrasonic anemometer was employed to
the measurement of turbulence intensity in process area. The field testing data
of turbulence intensity in percentage at interval of 1.2 m along x-axis are
presented in Fig. 7. The legend of solid circle and hollowed
circle represents the depth of y = 1.5 and y = 2.7 m, respectively, while the
diameter of circle denotes the magnitude of airflow disturbance. The turbulence
intensity displays higher percentage at right-hand side of process area and
it corresponds with the highest particle counts at location 13.
Figure 8 represents the room air temperature about 20.3°C
and mean radiant temperature (MRT) or global temperature about 20.8°C, respectively.
The relative humidity (about 55%) and airflow velocity fluctuation (0.2 m sec¯1
approximately) are shown in Fig. 8 as well.
||Particle counts of sampling location at different occupancy
Different cases of cleanroom apparel type (Table 2) were
examined to verify the particle counts in the test section. Clothing ensembles
were defined and generated in terms of clo unit based on ASHRAE thermal comfort
code. Figure 9 shows that more tightly-woven fabrics of cleanroom
garment presents less particle counts which reveals better contamination control
in test section. However, the PMV values vary from -0.07 to 1.31 which characterizes
the dissatisfaction of thermal comfort. Personnel will prefer garments that
give minimum of protection inevitably. Although contamination controls are preeminent,
a certain amount of trade-off on thermal comfort may be necessary.
||Field testing of turbulence intensity at the process area
||Field measurement of environmental condition at testing section
||Different cases of cleanroom apparel
||The effects of clothing type on particle counts and thermal
||Temperature contours and velocity vector at the test section
(a) velocity vector (b) temperature contours and (c) turbulent kinetic energy
The full-scale CFD simulation of velocity vectors at x = 0.65 and y = 0.65
m in test section are displayed in Fig. 10a, which reveals
the obvious eddies arisen in the vicinity of foot under the FFU face velocity
of 0.3 m sec¯1. As shown in Fig. 10b, the
temperature contours are observed under the assumption of human body as wall
function with the thermal resistance of cleanroom clothing at 0.465 m2-°C/W.
It also demonstrates that the highest temperature contour occurs in the vicinity
of human body with cleanroom garment, which reveals dissatisfaction of thermal
comfort. Besides, Fig. 10c presents the turbulent kinetic
energy contours, which corresponds the velocity vector in Fig.
||Effect of face velocity of FFU on testing section environment
condition and thermal comfort
To verify the effect of face velocity of FFU on turbulent intensity, maximum
garment surface temperature of test section and PMV indices, many cases were
simulation based on field test data. Figure 11 shows the
maximum garment surface temperature decrease as the FFU face velocity varies
from 0.1-0.5 m sec¯1. It also reveals the improvement of cooling
effect at the range of 0.2-0.4 m sec¯1, while the PMV value
varies unapparently. However, the turbulent intensity increased with an increase
in air velocity, which also specifies the worse contamination control at the
This study investigated the trade-off of thermal comfort and contamination
control assessment of a cleanroom in a MEMS laboratory. The field tests have
been carried out comprehensively using many delicate instruments. The ASHRAE
thermal comfort code was conducted to investigate thermal comfort of personnel
based on field-testing data consequently. Furthermore, the effects of clothing
on thermal comfort and contamination control have been assessed comprehensively.
Cleanroom garment can sometimes be hot and uncomfortable. More tightly-woven
fabrics will achieve better contamination control and result in thermal comfort
dissatisfaction. Results in this study should provide valuable information to
the facility engineer facing comprise between thermal comfort and contamination
control in the cleanroom. The contamination control could be achieved by proper
types of garments with satisfied thermal comfort of predict mean vote between
The authors would like to express their great appreciation to the financial support by the National Science Council under the grant No. NSC-96-2628-E-167-016-MY2.