From:
http://www.mtaonline.net/~hheffner/Caloristors.pdf
OBJECTIVE
The objective here is to define and describe the nature and use of
caloristors in the
making of calorimeters.
DEFINITION
A caloristor is (defined here as) a high thermal conductivity rod,
possibly copper or
aluminum, surrounded by very good insulation, with a metal heat
transfer surface
exposed at both ends. About 1/3 from either end of the insulation a
thermistor (or
thermocouple or RTD) is located under the insulation and in contact
with the metal
rod, possibly imbedded in the rod. The two temperature measuring
points are used
to determine the heat flux through the metal rod. Such caloristors
can be made in
all sizes and can accommodate and measure any sized heat flow.
DESCRIPTION AND DISCUSSION
It is likely useful to pre-make and calibrate caloristors for general
purpose use.
Thermistors provide accurate and cheap temperature measuring at boiling
temperatures or less. Thermistors can be read by very fast pulses to
avoid heating
the rods. Use of 100K ohm thermistors can also help avoid heat
contamination from
the measuring current. See:
http://www.mtaonline.net/~hheffner/Thermo.pdf
for a discussion of thermistors vs thermocouples. Alternatively,
thermocouples or
differential thermocouples can be used to measure the delta t, or
RTDs used.
The "cold" end of the caloristors can be attached to plumbing/hoses
that carry cooling
water (or air flow) and exchange heat with the cold end of the rod.
In that manner a
dual form of calorimetry can be made by using the (thermal mass flow)*
(delta t)
from the cooling water (or air flow). The caloristors provide
localized heat flow data
while the cooling water (or air) flow can provide fully integrated
flow calorimetry
using only two additional thermistors in the flow.
Alternatively, the cold ends of the caloristors can be attached to
peltier devices using
thermal grease, and the dual calorimetry performed by calibrating the
current used
vs heat flow into the peltier. Another variation is to use cooling
fins on the hot ends
and then measure the (thermal mass flow)*(delta t) for airflow into
and out of an
outer chamber. This method has the advantage of providing a measure
of heat flow
through the insulation in addition to that measured through the
caloristors. That
airflow can be temperature controlled using peltiers, or possibly by
just a tube in a
bucket of ice water, since ice is available in most hotels, etc.
The "hot" end of the caloristors can be pre-attached to metal plates
so as to make
building a calorimeter box easy, or can simply be made so it is easy
to make thermal
contact of the "hot" ends of the rods with any metal structure,
possibly using
machine screws.
Hot and cold are in quotes above because the heat flow can actually
be in either
direction.
With a supply of caloristors on hand, and a multiplexing A/D
converter, making an
ad hoc calorimeter becomes fairly simple. Surround the volume of
interest with a
highly conductive metal envelope, say copper, aluminum, titanium,
magnesium, or
thick steel, which is in turn surrounded by a blanket of insulating
material through
which the caloristors protrude to thermally connect with the
envelope. A second
blanket of insulation, say fiberglass, should cover the plumbing or
air duct system if
it is used. A thermally controlled jacket can form an outer layer if
needed.
The accuracy of the calorimeter so built then seems to depend
primarily on how good
the innermost surrounding insulation is, and if that insulation is
good, to a much
lesser degree on how good the surrounding innermost metal box
conducts heat and
avoids hot spots where there are no caloristors. The ideal insulation
material is
aerogel, but that is expensive. One advantage to this design is it
can be used with
any internal operating temperature desired - limited only by the
melting point of the
metal and insulation chosen and the limits on the temperature
measuring device
chosen.
If enough caloristors are used a picture can be gained of any hot
spots that develop or
move around and a determination can be made as to the thermal
stability and the
accuracy of the data. The main problem with accuracy may be a long
stabilization
time, but copper or aluminum rods conduct heat pretty fast and a well
insulated
thick internal metal box should at least provide a good integrated
total heat flow.
The other problem is the cost and complexity of building the
thermistor multiplexer.
One advantage is that enormous variations in scale are available by
simply varying
the size of the rods used in the caloristors.
Also of interest is that a single caloristor and "hot-end" plate
could be placed on the
outer surface of the inner layer of insulation in order to measure
the heat flow lost
through the insulation per area of insulation. This should be placed
at the location
of greatest heat flow through the insulation so as to establish a
maximal error
value.
One key to making all this work is obtaining or building a good and
cheap analog
multiplexer with lots of channels, and the driver software to
calculate and sum the
heat flows. Once you have those, and a bunch of pre-made caloristors
handy,
throwing together a custom dual method calorimeter of any size or
operating
temperature becomes fairly quick. It is just a matter of building a
right sized metal
box, insulating the box, and inserting caloristors through that
insulation, and
calibrating using control heat sources.
Horace Heffner
http://www.mtaonline.net/~hheffner/
