A simple inexpensive continuously operating steam/water calorimeter
can be obtained using a combined barrel and flow calorimetry.
A water container, a barrel, or perhaps a trash can which is silicone
sealed for leaks, can be used to condense steam via a submerged
copper coil, preferably mostly located near the top of the barrel to
avoid imposing a steam pressure head on the tested device. This
water container can be insulated cheaply using construction foam
board and fiberglass. A stirrer can be driven via a shaft through
the foam board.
A secondary coil can be used for pumped coolant. A fixed flow rate
pump can be used to deliver the coolant flow. The coolant flow
circuit can be open or closed. A closed secondary coolant temperature
can be maintained via either water or air heat exchange or ice heat
exchange. The source of the coolant energy is not important if the
Tin and Tout are measured close to the water container, and any
tubing between the temperature measuring stations and the water
container is insulated. Ideally the secondary flow rate would be
measured by a digital flow meter, and driven by a variable speed
pump. The coolant flow rate can then be adjusted to suit the coolant
delta T and the thermal power of the device under test.
Alternatively, an accurate fixed flow rate pump can be chosen with a
flow rate approximately matching the expected thermal power of the
device under test given the expected coolant delta T. A reasonable
goal for the water container temperature is the range 50°C to 70°C.
Use of a large water container provides some degree of momentary heat
pulse energy integration and confidence in the device thermal power
measurements. It applies a significant time constant to the thermal
data that reduces the frequency temperature data must be taken. It
even permits manual temperature reading if a modestly stable
condition is established. This is at the cost of being able to see
instant response thermal and energy output curves. There is no need
to see such fast response curves if the primary goal is to measure
total energy in vs total energy out for a long run.
The primary circuit water flow can be pumped directly from the water
container. Ideally the primary water flow should be measured by
digital flow meter. If a low pressure head is presented to the
primary circuit flow pump, then a precision fixed flow rate pump can
be used. If precision digital flow meters are not used, and reliance
is placed on precision flow rate pumps, then at minimum simple (flow
integrating) water meters should be monitored periodically to verify
assumed pump mean flow rates. Calibration runs on dummy devices
should be used to verify the calorimeter over the thermal range
expected. A calibration control run should be used with the device
under test to determine the water capacity of the device so the
volume of water in the barrel is known in order to provide improved
intermediate time thermal power measurements. At the conclusion of
a run, the circuits should continue to be driven until thermal
equilibrium is obtained and essentially all thermal energy is drained
form the device under test. A water depth gage for the barrel may be
of use, calibrated to depth vs volume, in order to keep track of the
amount of water in the device under test.
The secondary circuit input and output temperature should be recorded
frequently. Alternatively, a direct delta T can be measured
frequently using an appropriate dual thermocouple arrangement, thus
providing improved data quality and reducing data acquisition
required. Flow stirrers should be used, if feasible, in the secondary
circuit prior to the thermometer wells. Barrel water temperature
should be monitored. Ideally primary circuit water input temperature
and room temperature should be monitored as well.
A thermal decline curve should be measured for the water container
when there is no primary circuit flow, and the water is stirred.
The calorimeter constant C(dT) as a function of the difference
between room temperature and water contained temperature (dT) should
be determined. The curve C(dT) can be fit to a polynomial using
regression analysis for convenient use in data analysis. Experience
shows this method is not very accurate if the water container is not
well insulated. This is due to room drafts, variations in humidity
and temperature during the day, etc. Ideally active insulation could
be used, whereby an extra envelope surrounds the water container
insulation and the temperature there is maintained at the temperature
of the water, thereby producing a dT = 0, and no heat loss. This is
excessive for this approach, however, the goals of which are "cheap",
"simple", and "good enough".
In summary, a minimum configuration then would consist of an
insulated barrel with copper condensing coil, and secondary heat
exchanger coil, a stirrer, and two precision volume pumps, one
primary, the other secondary. Temperatures would be monitored
frequently for the secondary in and out flows, perhaps less
frequently for room temperature, barrel temperature, and test device
input temperature. As a second level, ordinary integrating water
meters could be added, for flow confirmation, on both the primary and
secondary circuits. Ideally, precision digital flow meters should be
used for both the primary and secondary circuit input flows.
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
