Volume 100, Number 5, September–October 1995

Journal of Research of the National Institute of Standards and Technology

[J. Res. Natl. Inst. Stand. Technol. 100, 591 (1995)]

A Third Generation Water Bath Based

Blackbody Source

Volume 100 Number 5 September–October 1995

Joel B. Fowler A third generation water bath based black- to 7.8 mK. The calculated emissivity of
body source has been designed and these sources is 0.9997 with a relative
National Institute of Standards constructed in the Radiometric Physics standard uncertainty of 0.0003. With a
and Technology, Division at the National Institute of 50 mm limiting aperture at the cavity;
Gaithersburg, MD 20899-0001 Standards and Technology, Gaithersburg, entrance, the emissivity increases to
MD. The goal of this work was to design 0.99997.
a large aperture blackbody source with
improved temporal stability and repro- Key words: aperture; blackbody; cavity;
ducibility compared with earlier designs, as conical; emissivity; radiation; radiometry;
well as improved ease of use. These reflectance; source; temperature; ther-
blackbody sources operate in the 278 K to mometer; water bath.
353 K range with water temperature
combined standard uncertainties of 3.5 mK Accepted: July 21, 1995

1. Introduction
1
In 1985 NIST designed and built a first generation uncertainty (that is, estimated standard deviation as
water bath based blackbody source [1] with a wide stated in [1]), including the water temperature stability
aperture cylindrical-conical cavity design. The cavity and uniformity at any of the temperatures in its operat-
was coated on the inside with a specular black ing range. The second generation bath had temperature
gloss enamel paint and immersed in the water medium instabilities in the temperature range from 10 8C above
of a commercially available temperature-controlled to 10 8C below ambient room temperature, which neces-
water bath. Although the cavity used in this first-gener- sitated the use of an auxiliary cooling loop to maintain
ation source had the same large 10.8 cm diameter aper- the above combined standard uncertainty in this range.
ture as the one described here, it was shallower, result- The apparent source of these water temperature instabil-
ing in a lower effective emissivity. The bath used in this ities was the bath configuration and internal temperature
earlier blackbody design was unstable at temperatures control loop stability.
near ambient. In 1987 NIST designed and built a The third generation blackbody source described in
second generation water bath based on blackbody this paper has a wide (10.8 cm) diameter aperture and
source with the same large-area aperture as the first an extended conical cavity section similar to that of
generation, but with a 75 % deeper cavity which in- the second generation design. The water temperature
creased the number of reflections in the cone from three stability; of this new blackbody source is 62 mK or
to four with a resultant increase in the effective emissiv-
ity. The new bath had an increased temperature stability 1
All uncertainties given in this paper are standard uncertainties (i.e.,
resulting in a 20 mK to 50 mK combined standard 1 standard deviation estimates) unless otherwise specified.

591
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology
less over many days; the temperature uniformity of the controlled water bath with GPIB control capability to
water volume is 62.0 mK at the lowest temperature in heat and cool the water in which the cavity is immersed
it’s operating range and 65.0 at the high end of its and a Hart Scientific electronic thermometer Model
operating range of 278 to 353, as measured using the 1575 used in conjunction with a Thermometrics pre-
resistance thermometry detailed later in this paper. The cision thermistor probe model ES-210 ‘‘temperature
increased stability is due to state of the art control loop standard’’ to accurately measure the temperature of the
electronics and the thermo-mechanical configuration of water in the bath. Hereafter, the model 7008 water bath
the bath. The increased uniformity is due primarily to will be referred to as the bath, the model 1575 electronic
the physical configuration of the bath. thermometer as the electronic thermometer, and the
model ES-210 thermistor probe as the thermistor probe
unless otherwise specified. The entire apparatus is
2. Design shown in Fig. 1.
This blackbody source incorporates a specially The bath is well suited for this application because of
modified Hart Scientific Model 7008 2 temperature- the excellent temperature uniformity and stability of the

Fig. 1. Third generation water bath based blackbody source.

2
Certain commercial equipment, instruments, or materials are identi-
fied in this paper to foster understanding. Such identification does not
imply recommendation or endorsement by the National Institute of
Standards and Technology, nor does it imply that the materials or
equipment identified are necessarily the best available for the purpose.

592
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

water in the bath. The cooling and heating sources are

laminated into a single plate covering the bottom of the
bath well. The water in the bath well is agitated by a low
speed double stirrer. The cavity is mounted in the side of
the bath allowing deflection of the warm air from the
evaporator coil of the refrigeration unit away from the
cavity mouth; this reduces air currents in the vicinity of
the cavity. Without the cavity installed, the bath is stable
to better than 61 for extended periods of time and the
water uniformity is better than 61 throughout the
volume over the entire operating range. With the cavity
installed, the instability values increase to 62 or better
over periods of days with a 65 maximum nonuniformity
at any temperature in the operating range.

2.1 Cavity Design and Construction

3. Control and Measurement
The cavity was constructed using oxygen-free cop- The bath electronics controls the heating cycle through
per. The conical portion was machined from a solid the use of a proportional-integral-derivative (PID) con-
round bar and the cylindrical portion was machined trol loop. The cooling cycle is preset at one of two levels
from a section of tubing. These two parts, along with a depending on the cooling requirements. Cooling and
mounting ring, were then brazed together in a vacuum heating modes are controlled from the front panel or by
oven using a high-copper-content-alloy brazing mate- the computer interface. The heating and cooling cir-
rial. The oxygen-free copper has a very high thermal cuitry allow achievement of bulk water temperature
conductivity of 3.88 W/(cm) [2], which improves the stability of better than 62 with the cavity installed. The
thermal uniformity of the cavity and decreases the ther- temperature of the water in the bath is measured with
mal resistance of the cavity wall. The outer surface of the electronic thermometer in conjunction with the
the cavity was plated with a thin layer of gold over nickel thermistor probe. The electronic thermometer con-
to retard oxidation of the copper surface. tributes a temperature measurement uncertainty of
Enamel paint was applied to the interior surface of the 1 mK and the thermistor probe contributes an additional
cavity wall by setting the cavity in a ring stand with the temperature measurement uncertainty of 1.5 mK. The
tip of the cavity down and introducing approximately temperature setpoint of the bath is controlled by, and the
1 fluid ounce ( four 0.25 fluid ounce bottles) of Testor’s temperature data from the electronic thermometer is
black model paint into the cavity. The cavity was then read by, a digital computer utilizing a GPIB interface
rotated axially while being slowly tipped from vertical and software developed at NIST.
to horizontal; thereby coating the entire interior surface The bath may be controlled by either control panel
with the paint. Once coated, the cavity was placed on a settings or the GPIB interface. The method of setting the
sheet of paper with the tip pointing up, thereby permit- temperature is similar either way. When under program
ting the excess paint to drain out of the cavity while at control, the desired temperature is input to the program.
the same time keeping the paint fluid by trapping the The bath is then commanded by the program to go to the
vapors from the paint inside the cavity. Once the excess desired temperature setpoint in the coarse setting mode
paint had drained out of the cavity, in about 4 h, the which has a 1 K setting accuracy. The bath temperature
cavity was supported approximately 12 mm above the is then read by the electronic thermometer until the
paper and the paint permitted to harden. Support in this temperature has stabilized close to the setpoint, with an
manner reduces the chance of distortion of the paint due instability of 65 mK. The bath is then commanded in
to changes in the force acting on the paint due to gravity the high resolution setting mode (which is capable of
and keeps dust away from the inside cavity surface while setting the bath temperature relative to the measured
the paint is hardening. The paint was tack free and firm bath temperature to better than 61.0 mK) to change the
after 2 days. The cavity was then baked at 50 8C for 8 h. temperature by the difference between the desired tem-
Useful dimensions and other properties of the cavity perature and the actual measured temperature. After this
section are shown in Fig. 2. final temperature setpoint adjustment, the bath will at-
tain a temperature setpoint well within the requirements
of this application. The computer program monitors and
logs the temperature of the bath continuously while at
this setpoint.

593
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

4. Temperature Measurements 4.2 Cavity Lip Temperature

4.1 Measurement of Water-Volume Thermal A differential thermocouple thermometer was used to
Uniformity Around the Cavity measure the temperature drop between the immersed
cavity components and the outside cavity lip. One
The water surrounding the cavity wall was measured thermocouple was imbedded in white heat sink com-
with the electronic thermometer, using two thermistor pound and pressed against the outer lip of the cavity with
probes, one on each of the two input channels of the a 20 cm length of its cable in contact with the cavity lip.
instrument. One thermistor probe was positioned at a The other probe was covered with the same compound
fixed reference point near to the cavity tip and the other and pressed against the wetted portion of the conical
thermistor probe was movable around the perimeter of section of the cavity while immersed in the bath water
the cavity. The movable thermistor probe was positioned with approximately the same 20 cm cable length in the
at ten locations , at three different levels around the water with the sensor. Each channel of the differential
outline of the cavity using a special fixture. The three thermometer was calibrated using a bath whose temper-
levels correspond to the centerline of the cavity, bottom ature was measured and set with the Hart thermometer
edge of the cavity, and top edge of the cavity. Immersion to read exactly 353.00 K. A difference of 4.1 K was
effects were minimized by encasing the area above the measured between the cavity lip and the immersed tip of
thermistor probes within an insulated dome. This per- the cavity. Both channels were then rechecked at the
mitted the entire volume above the thermistor probes to conclusion of the measurement and they still indicated
be engulfed by water vapor or air at nearly the same 353.00 K. The standard uncertainty of the differential
temperature as the immersed portion, thus minimizing thermometer was 50 mK, as both indicated in the
heat conduction along the probe casing and thereby manufacturer’s data sheet and verified by the electronic
reducing the immersion loss errors to a negligible level thermometer. The temperature of 353.00 K was chosen
compared to the other errors in the temperature as the worst case as it represents the largest deviation
measurement system. The resulting measured values for from ambient temperature.
the water temperature uniformity are given in Table 1.
4.3 Water Temperature Uncertainties
Table 1. Water temperature nonuniformity and standard uncertainty The water temperature uncertainties can be divided
values used in the uncertainty analysis at four set-point temperatures into two categories. The first is the combined contri-
bution from the external temperature measurement
Water Nonuniformity Standard uncertainty
temperature (K) (mK) value used (mK)
system consisting of the electronic thermometer and the
thermistor probe, along with the possible temperature
278 –1 to +4 4 control errors due to the water bath temperature control
303 –1 to 0 0
333 0 to –5 5
characteristics. The second consists of the uncertainties
353 0 to –7 7 of the thermodynamic properties and the behavior of the
cavity, the effects of the cavity interior coating, and the
environmental effects due to such things as convection,
These measurements indicate that the maximum stray air currents, etc.
deviation of the water temperature at each of the three The uncertainties associated with the control and
levels around the periphery of the cavity varies from external measurement of the temperature of the water in
0.0 mK near the cavity tip to +4 mK near the bath wall the bath are straightforward in that only the contribu-
adjacent to the cavity at 278 K, has no variation between tions from the water bath temperature control and the
the cavity tip and a position near the bath wall adjacent electronic thermometer-thermistor probe combination
to the cavity at 303 K, varies from 0.0 mK near the need to be considered. The bath control circuitry con-
cavity tip to –5 mK near the bath wall adjacent to the tributes an absolute temperature setting uncertainty of
cavity at 333 K, and varies from –2 mK near the cavity 1 K without external temperature measurement and less
tip to –7 mK near the bath wall adjacent to the cavity at than 1.0 mK with external temperature measurement
353 K referenced to the fixed probe. The highest values when using the high-resolution-setting mode as stated
of the deviations of the water temperature at each of the by the manufacturer. The external temperature mea-
above four temperatures were chosen as the conservative surement uncertainty is due to the thermistor probe and
values of the water temperature nonuniformity at those the electronic thermometer. The thermistor probe con-
temperatures. tributes 1.5 mK uncertainty as calibrated at the factory

594
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

and referenced to a standard traceable to NIST. The Hart referenced to a NIST standard when used in combina-
electronic thermometer contributes an uncertainty of tion with the thermistor probe ( same type used with the
1 mK or less as calibrated by the factory and traceable model 1575 thermometer). The substitution of the
to a NIST standard, resulting in a combined standard model 1506 thermometer will increase the combined
uncertainty of 1.8 mK for the combination of the elec- standard uncertainty of the water temperature measure-
tronic thermometer and the thermistor probe. The un- ment to 9.8 mK, which is near the maximum 10 mK
certainty in the bath temperature stability is due to the standard uncertainty required for this application. The
bath control loop which has an instability of less than more expensive model 1575 thermometer was used in
62 mK and has no contribution from the external ther- these tests only to determine how stable and accurate
mometry for a given setpoint temperature within the this blackbody source could be with high accuracy com-
operating range of the instrument. ponents.
Uniformity of the bath temperature, as discussed in
Sec. 4.1, is typically 61 mK without the cavity inserted. 5. Cavity Emissivity
Upon insertion of the cavity, this value degrades slightly
to 62 mK for the range 78 K to 313 K and 65 mK for Ideally, when the walls of the cavity are in local
the range 313 K to 353 K. Table 2 shows the uncertain- thermodynamic equilibrium, the emissivity of the
ties associated with the temperature of the water in the cavity e is 1 minus the cavity reflectance. Using this
bath. fact, we may calculate an approximation for the emissiv-
ity e based on the assumption that the reflectance of the
interior wall of the cavity is the sum of a perfectly
Table 2. Water temperature measurement and control errors and
standard uncertainties used in uncertainty analysis specular component r s and a perfectly diffuse compo-
nent r d. The specular reflectance is divided into two
Source of Range of possible Standard components: one to account for the specular reflectance
uncertainty temperature error uncertainty (mK) at normal incidence and one to account for the specular
values (mK) reflectance at lower angles of incidence. This is neces-
Electronic thermometer 61.0 1.0 sary, as the angle of incidence of each reflection after
Thermistor probe 61.5 1.5 entering the cavity varies. For radiation entering the
Stability of bath 62.0 2.0 cavity at near normal incidence, the emissivity e is
Bath setting 61.0 1.0 expressed by [3]
Est. immersion loss 62.0 2.0

e = 1–(r ss 2 r sh 2 )–dF 13 r d (1)

The combined standard uncertainty of the bath water
temperature may be calculated by adding in quadrature where dF 13 is the differential configuration factor which
the standard uncertainties associated with the water describes the fraction of the radiation emitted from a
temperature nonuniformity, bath temperature setting differential area dA on the cavity wall which exits
error, possible thermistor probe immersion error, tem- through the opening of the cavity, r sh is the specular
perature measurement error due to the thermistor probe, reflectance of the cavity surface at high angles of inci-
and temperature measurement error due to the elec- dence for the first two reflections, and r ss is the specular
tronic thermometer, yielding a standard uncertainty of reflectance of the cavity surface at smaller angles of
5.3 mK at 278 K, a standard uncertainty of 3.5 mK at incidence for the remaining two reflections. We have
303 K, a standard uncertainty of 6.1 mK at 333 K, and ignored the small variation of e with wavelength l . Four
a standard uncertainty of 7.8 mK at 353 K as shown in reflections inside the cavity were chosen due to physical
Table 1 and Table 2. These are worst case standard un- limitations of the bath and the manufacture of the cavity
certainties of the water temperature at any point in the and diminishing returns from additional reflections.
bath surrounding the cavity, but does not take into These four reflections equate to an absorption of
account the interface between the water and the cavity 99.998 % of the radiation entering the cavity. Equation
and the thermodynamics of the cavity. These last two (1), though not a worst case approximation, may be used
factors are taken into account in the calculations for a worst case analysis by choosing conservative esti-
described later in this paper. mates for r d , r sh , and r ss .
These blackbody sources will generally be used with The measured reflectance of a witness sample of the
a lower cost Hart Scientific Model 1506 electronic ther- black gloss coating used inside the cavity but applied to
mometer with a temperature measurement standard un- the same copper material from which the cavity was
certainty of 6 mK, as stated by the manufacturer and machined is 5 % total reflectance up to 10 mm, 7 %

595
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

up to 20 mm and rises rapidly past 20 mm at near normal 5.1 Temperature Distribution Over the Interior
incidence. The reflectance value is increased from 5 % Cavity Surface
to 10 % for the larger angles of incidence. Though no
The worst case approximations used to estimate the
data were available for the diffuse reflectance beyond
temperature drops in regions 1 and 2, shown in Fig. 2,
2.5 mm, the diffuse reflectance is known to be <0.2 %
for this the cavity are as follows:
between 800 nm and 2.5 mm and become less diffuse
and more specular with increasing wavelength. 1) Region 3 is at a uniform temperature throughout
The reflectance for this cavity was chosen to be a (T 3).
conservative value of r sh = 10 % specular for the larger 2) Region 2 is at a uniform temperature throughout
angles of incidence r ss = 5 % specular for the smaller (T 2).
angles of incidence, and r d = 0.2 % for the diffuse 3) The worst case value for the temperature in region
reflectance. The value of the differential configuration 2 is the temperature at the very edge of the
factor dF 13 was calculated for nominal cavity dimen- cavity lip.
sions and varies from 0.03 near the tip of the conical
4) Region 0, the surface in contact with the bath
section of the cavity to 0.07 near the cylindrical-conical
water, is at a uniform temperature T 0 which is the
intersection. A conservative value of dF 13 = 0.07 was
same temperature as the bath water.
chosen.
Utilizing the above values, Eq. (1) yields 0.9997 for a For high accuracy measurements, the temperature of
lower bound for the emissivity e with 0.999760.0003 the bath water must be very stable and accurately
for a conservative estimate for the spectral range measured. The water in this bath was accurately
of 1 mm to 30 mm. If a 50 mm diameter aperture with measured for stability and absolute temperature as
high infrared reflectance on the side facing the cavity outlined in Sec. 4.1 and meets this requirement in excess
were added to the front of the cylindrical portion of of the extent necessary to achieve the desired quality of
the cavity and the calculation repeated, the emissivity the source. The term quality will be described later.
increases to near 0.99997 at normal incidence to the
cavity.
5.2 Temperature Drop Across the Cavity Wall and
As a check of the above calculations, the emissivity
the Black Paint
was recalculated utilizing a computer program written
by Prokhorov and Sapritsky [3] for the calculation of The differential heat conduction across the cavity
blackbody emissivity. For the same parameters as used wall in region 1 and radiating out of the cavity is given
in the above calculations, this computer program yields by:
an effective emissivity of 0.9998 at normal incidence. dP = (T 0–T 1) / (d cu /K cu+d bp /K bp), (2)
Recalculating with the addition of a 50 mm aperture
in front of the cylindrical portion of the cavity, the where d cu is the cavity wall thickness, K cu is the thermal
emissivity normal to the cavity increased to 0.99996. conductivity for the copper wall of the cavity, d bp is the
This program also can account for nonuniform tempera- thickness of the paint, and K bp is the thermal conductiv-
ture distributions over the cross sectional and longi- ity of the enamel paint. For the assumed thermal equi-
tudinal dimensions of the cavity. Randomly varying the librium, the above quantity must balance the net differ-
temperature uniformity input to the computer program ential radiant power leaving the surface of the paint on
by as much as 6100 mK, a much worse case than the inside of the cavity wall at any point in region 1 as
our 5 mK maximum measured nonuniformity, the shown in Fig. 2. This quantity dP is given by
normal emissivity calculated was never less than 0.9991
without the 50 mm aperture or less than 0.99991 with dP = dF 13 s (T 14–T 34 ) + (dF 12–dF 13 ) s (T 14–T 24 ) (3)
the 50 mm aperture. This confirms our assumption
that the nonuniformity of the surface temperature where s is the Stefan-Bolzmann constant, and dF ij is the
of the cavity as measured in this instrument is not differential configuration factor from the point of inter-
significant. est in region i to all of region j . Because the temperature
Effects such as air currents and the consequences difference between T 0 and T 1 is small, the error intro-
of off-axis viewing have been ignored and will duced by approximating T14 by
be addressed in a report on the radiometric testing
of the blackbody currently being performed at NIST. T14 = T04 + T03 D T (4)

596
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

is negligible, where DT = T 1 –T 0 is the temperature drop Table 4. Calculated temperature drop across the cavity wall from the
across the cavity wall and paint. Equations (3) and (4) bath water to the inside cavity wall (ambient temperature 298 K)
may be solved simultaneously for DT in closed form
Water Temperature Standard
temperature (K) drop (mK) uncertainty (mK)

– bT 0 (dF 13 [1–(T 3 /T 0 ) 4 ]+[dF 12 –dF 13 ][1–(T 2 /T 0 ) 4 ]) 278 +0.9 0.9

DT= , 283 +0.5 0.5
1+4dF 12 b 293 0 0
(5) 303 –0.5 0.5
313 –1.1 1.1
323 –1.7 1.7
333 –2.4 2.4
where 343 –3.2 3.2
353 –4.0 4.0
363 –4.9 4.9

b = s T 03 ( (d cu /K cu+d bp /K bp). (6)

6. Blackbody Quality
Table 3 enumerates the nominal values used in the
evaluation of DT in Region 3. The above analysis is The blackbody quality accounts for the effects of
similar to the analysis presented in NBS Technical Note temperature gradients between the water in the bath and
1228 [1] and has been modified to reflect the changes in the cavity surface, and the cavity wall reflectance in a
the design of the new blackbody design. single quantity [4,5]. Quality is defined here in terms of
a reference temperature, which is conveniently the
temperature which is actually being measured during
the operation of the blackbody, the water temperature in
Table 3. Values of parameters used to calculate DT from Eq. (5)
this case. It is the ratio of two radiances that are impor-
Parameter Value
tant: the actual cavity radiance, and the ideal Planck-law
radiance at this reference temperature.
Cavity internal diameter 10.7 cm A simple expression for the quality of a blackbody of
Length of cylindrical cavity section 10.9 cm
Full angle of cavity 38 8
this type can therefore be expressed [1] by
Thickness of cavity wall 0.4 cm
Thickness of black paint 0.005 cm Q = e [exp(C2 / lT 0 )–1] / [exp(C 2 / lT )–1] (7)
Thermal conductivity of cavity wall 3.8 W/(cm K)
Thermal conductivity of black paint 0.0018 W/(cm K) where e is the emissivity (again ignoring the small vari-
ations of e with l ), l is the wavelength of interest, T is
the effective cavity temperature, T 0 is the reference tem-
perature and C 2 is the the second radiation constant. Q
The paint thickness was measured by taking the dif- is simply the calculated emissivity modified by the ratio
ference between the thickness of the coated metal blank of the ideal and actual Planck law radiances.
used for the witness sample in the measurement of the Taking the first two terms of a Taylor series expansion
reflectance of the black paint before and after coating. of the right hand side of Eq. (7) and substituting
Although the method of coating the sample was per- DT = T–T 0 when the second term is small compared to
formed to closely approximate the inside of the cavity, unity, Eq. (7) may be approximated [1] by
the estimate of 0.005 cm may be incorrect by up to
50 %, therefore d bp = (0.005 6 0.0025) cm has been
Q ≈ e [1+(DT /T 0 ) (C 2 / lT 0 ) / [1–exp(– C 2 /DT 0 )]. (8)
chosen as a conservative estimate.
Values for the temperature drop DT across the cavity
wall at the intersection of the conical and cylindrical
6.1 Overall Blackbody Quality
portions of the cavity which are totally immersed in the
temperature controlled water were calculated and are We can use Eq. (8) to calculate the quality of the
shown in Table 4 for several water temperatures, along blackbody at any wavelength and to calculate the uncer-
with the associated uncertainties. tainty of the quality using the values calculated for DT

597
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

and e . An equation for the uncertainty in the blackbody Table 5. Parameters used in the calculation of the blackbody quality
quality is [1] and uncertainty using Eq. (1) and Eqs. (8) and (9).

DT Standard uncertainty u (DT ) a Restrictions

(mK) (mK) (K)

u c,r (Q ) = Î Q [(u (e )/e ) +(F (C /lT )u (T )/T ) ] ,

2
2 0 0
2 +0.9
+0.5
5.3
4.9
T 0 = 278
T 0 = 283
0.0 3.7 T 0 = 293
–0.5 3.5 T 0 = 303
(9)
–1.1 3.8 T 0 = 313
–1.7 4.9 T 0 = 323
where –2.4 6.1 T 0 = 333
–3.2 7.0 T 0 = 343
–4.0 7.3 T 0 = 353
F (x ) = x /[1–exp(–x )]. (10) –4.9 7.8 T 0 = 363
e Standard uncertainty u (De ) Restrictions

may be derived from Eq. (7). 0.9997 0.0003 No aperture

Table 5 shows the relevant values used in the calcula- 0.99997 0.00003 50 mm aperture
a
tion of the quality and the uncertainty of the quality, and These values are only valid when viewing the conical section of the
Figs. 3 and 4 graph the quality and it’s uncertainty cavity.
versus wavelength.

Fig. 3. Blackbody quality.

598
 Volume 100, Number 5, September–October 1995
Journal of Research of the National Institute of Standards and Technology

Fig. 4. Blackbody quality uncertainty.

7. Conclusion Acknowledgments

A high quality thermometer was used in the evalua- The author would like to acknowledge Dr. Jon Geist
tion and in the operation of the water bath during testing. (retired) of NIST Gaithersburg for his assistance in
In the normal use of this instrument, a thermometer only understanding the theory and use of differential config-
slightly better than the expected performance need be uration factors and their application to black body
used. The Hart Scientific Model 1506 electronic cavities, Victor Sapritsky of VNIIOFI for his assistance
thermometer with a thermistor probe suits this need in obtaining and using the EE31 emissivity program,
very nicely. Substitution of the Model 1506 only and Tom O’Brian of the Radiometric Physics Division at
degrades the performance by the increased uncertainty NIST for his assistance in the application of the EE31
of the thermometer. The temperature measurement computer program.
combined standard uncertainty of the Model 1506 elec-
tronic thermometer when used in conjunction with the 8. References
thermistor probe is 6.2 mK. The blackbody quality
would only decrease by 0.1 % at long wavelengths and [1] Jon Geist and Joel B. Fowler, A Water Bath Blackbody for the
0.01 % at the shorter wavelengths if used with the lower 5 8C to 60 8C Temperature Range: Performance Goal, Design
Concept, and Test Results, NBS Technical Note 1228, October
accuracy thermometer.
1986.
The uniformity and stability of this new generation [2] Materials Science and Engineering Handbook, CRC Press 1994.
water-bath-based blackbody shows definite improve- [3] V. I. Sapritsky and A. V. Prokhorov, Calculation of the Effective
ment over past designs, both in ease of use and overall Emissivity of Specular-diffuse Cavities by the Monte Carlo
quality. The design exceeds our goal of 10 mK com- Method, Metrologia 29, 9-14 (1992).
[4] Jon Geist, Theoretical Analysis of Laboratory Blackbodies, Appl.
bined standard uncertainty of the water temperature,
Opt. 12, 1325 (1973).
whether used with the Hart Super Thermometer or the [5] J. Geist, Note on the Quality of Freezing Point Blackbodies, Appl.
Model 1506 Metrology thermometer. The calculated Opt. 10, 2188 (1971).
emissivity is very high and we expect excellent radio-
metric characteristics. Radiometric measurements are
currently being conducted at NIST and will be the About the author: J. B. Fowler is an electrical
subject of a future paper. engineer in the Radiometric Physics Division of the
Physics Laboratory at NIST. The National Institute of
Standards and Technology is an agency of the Tech-
nology Administration, U.S. Department of Commerce.

599