Weight loss abstract
The present invention provides a method and apparatus for accurately
determining weight loss of a sample during heating in a furnace.
The method includes the steps of placing a sample in a heated furnace,
heating the sample while measurements of sample weight are made,
determining rate function from the sample weight measurements, producing
a weight loss correction factor using the rate function and using
the weight loss correction factor to obtain a corrected weight loss
for the sample.
Weight loss claims
That which is claimed:
1. An apparatus for determining weight loss of a sample, comprising:
a furnace, said furnace comprising a combustion chamber and a door
providing access to said combustion chamber; a scale mounted within
said furnace for measuring sample weight; a data store operatively
connected to said scale for storing sample weight measurements;
means for determining a weight loss rate function from the sample
weight measurements in said data store; means for determining the
approximate time at which the onset of sample combustion occurs;
and means for generating a weight loss correction factor using the
approximate time of combustion onset and the weight loss rate function.
2. An apparatus according to claim 1, further comprising: a temperature
sensor mounted within said combustion chamber, said temperature
sensor operatively positioned to measure sample or combustion chamber
temperature; and a second data store operatively connected to said
temperature sensor for storing temperature measurements.
3. An apparatus according to claim 1, further comprising means
for generating a corrected weight loss measurement using a final
sample weight measurement from said data store and said weight loss
correction factor.
4. An apparatus according to claim 1, wherein said means for determining
the combustion onset time comprises means for determining the time
at which the weight loss of the sample departs from a linear function
by a threshold amount using the weight measurements in said data
store.
5. An apparatus according to claim 1, wherein said means for determining
the combustion onset time comprises: a temperature sensor mounted
within said combustion chamber, said temperature sensor operatively
positioned to measure sample or combustion chamber temperature;
a second data store operatively connected to said temperature sensor
for storing temperature measurements; and means for determining
the time at which the rate of change in the stored temperature measurements
departs from a linear function by a threshold amount using the measurements
in said second data store.
6. An apparatus according to claim 1, wherein said means for determining
the combustion onset time comprises: a temperature sensor mounted
within said combustion chamber, said temperature sensor operatively
positioned to measure sample or combustion chamber temperature;
a second data store operatively connected to said temperature sensor
for storing temperature measurements; and means for determining
the time at which the measured temperature reaches a predetermined
temperature using the measurements in said second data store.
7. An apparatus according to claim 1, further comprising a digital
filter operatively connected to said data store for filtering sample
weight measurements.
8. An apparatus according to claim 7, wherein said digital filter
comprises a low pass filter.
9. An apparatus according to claim 1, further comprising a delay
timer operatively connected to said scale to delay the measurement
of sample weight for a predetermined amount of time.
10. An apparatus according to claim 1, further comprising a plate
overlying said scale.
11. An apparatus according to claim 1, further comprising a pan
assembly operatively positioned for weighing on said scale.
Weight loss description
FIELD OF THE INVENTION
This invention is concerned with accurately weighing heated materials,
and is especially applicable to a pyrolysis furnace and to the measurement
of weight loss in such a furnace.
BACKGROUND OF THE INVENTION
Certain industrial processes require accurate measurement of the
weight or mass of a material before it is in a state of thermal
equilibrium. In some cases, it is necessary to achieve accuracy
on the order of tenths of a gram with samples larger than 3 kg.
For example, the construction industry requires the measurement
of asphalt content for quality control purposes. Asphalt is a mixture
of asphalt binder and aggregate and is used heavily in the construction
of roads. The mechanical properties of this mixture depend on many
parameters, such as the asphalt binder content by weight and the
gradation of the aggregate. In order to measure the quality of these
materials, the contractor needs a process to separate the binder
from the aggregate.
In the past, there have been several accepted methods to obtain
this information. Two such methods involved chemical solvents and
nuclear isotopes. The nuclear asphalt content gauge can be used
to accurately measure the binder content of asphalt in just a few
minutes. Although this method is fast, the drawback is that gradation
analysis cannot be obtained. Chemical solvents can give both asphalt
content and gradation analysis. However this method is laborious,
time consuming, and the waste solvent poses environmental problems.
In recent years, a method of igniting asphalt in order to measure
the weight loss due to combustion has become accepted. Although
this method is relatively slow as compared to the nuclear techniques,
gradation analysis can be obtained as soon as the ash has cooled.
With the advent of new technologies in the construction industry,
the standards have also become more stringent. Variations in weight
loss measurements from lab to production site to construction site,
and even furnace manufacturer, must be minimized.
In the conventional industry process, a sample is weighed to the
nearest tenth of a gram using an external scale and placed in a
basket assembly. The assembly is then placed in a preheated furnace,
which is outfitted with an internal scale assembly or load cell.
The door is secured, and the weighing process begins. During the
first few moments, a tare or beginning weight is measured. During
the next few minutes, the asphalt binder begins to burn and the
furnace automatically calculates a weight loss relative to the initial
weight and calculates the real time asphalt binder content. The
entire process may last from 20 to 60 minutes depending on the initial
sample weight and design of the furnace.
Since the asphalt is usually mixed at a temperature of about 150.degree.
C., and the furnaces are usually preheated to temperatures near
538.degree. C., thermal instabilities exist that make the process
of obtaining an accurate initial weight of the asphalt a very challenging
endeavor. Typically, the errors incurred are on the order of a few
grams, and decrease as the sample temperature approaches the temperature
of the furnace. The largest error in the weight loss determined
using this method is due to an erroneous tare weight obtained during
the first few minutes. Generally, the internal scale in the furnace
reports a higher basket assembly weight during the first few minutes
in the furnace than one would obtain from an external weighing.
This error is the direct result of the temperature differential
between the furnace and the sample and basket assembly. Furthermore,
the last few minutes in the furnace atmosphere are measured as lighter
in weight by the internal scale than one would expect externally.
Compared to external scale measurements at ambient temperature,
the furnace internal scale overestimates the actual weight loss
of the sample.
There have been several attempts to clarify the physics of this
effect. In one patent, U.S. Pat. No. 5,279,971 to Schneider, the
initial error in tare weight is reported as due to moisture absorbed
in the asphalt. However, an asphalt plant mixes these constituents
at 150.degree. C. and moisture accounts for a small percent by weight,
if any. Even where the sample is dried overnight and all moisture
is removed, the same errors occur. The Schneider patent reports
that samples should be preheated to 300.degree. C. before placing
them in the 550.degree. C. furnace. The Schneider patent states
that this reduces the "moisture" error. Actually, the
error in tare weight was reduced only because the temperature differential
between the sample and furnace was 200.degree. C. as opposed to
400.degree. C.-500.degree. C. with a typical sample removed from
the production line.
The temperature error caused by placing a relatively cool sample
into an extremely hot oven results in a complicated model involving
several external factors, such as air density, air flow, and bombardment
of the sample and pan assembly by high energy gas molecules. Furthermore,
these factors affect the measurement in different ways according
to the properties of the sample, such as mass, thermal capacity,
thermal conductance, voids or density, and specific gravity. Thus,
there are many different combinations of these variables that perturb
the initial measurement. There remains a need in the art for a method
of accurately weighing samples in a heated furnace that takes into
account the complex effects of thermal instability present during
the initial weighing process.
SUMMARY OF THE INVENTION
The present invention provides a method and apparatus for accurately
weighing samples in a heated furnace. More particularly, the present
invention provides a method and apparatus in which the weight loss
of a sample may be accurately determined as the sample is heated
in a furnace. In one specific embodiment, the sample is an asphalt
binder/aggregate paving mix and the method and apparatus are utilized
to accurately measure the asphalt binder content of the paving mix
by determining the weight loss resulting from pyrolysis of the asphalt
binder. Using the present invention, weight loss values calculated
using the internal scales of a furnace are within about 0.05% of
the weight loss values calculated with an external scale.
According to the invention, a correction factor is generated which
corrects for errors in the measurement of the tare weight of the
sample due to external influences and variables such as those noted
above. The invention may additionally correct for errors in the
end point weight, also due to external influences. A method in accordance
with the broad aspects of the invention, includes the steps of placing
a sample in a heated furnace, heating the sample while measurements
of sample weight are made, determining a rate function from the
sample measurements, producing a weight loss correction factor using
the rate function, and using the weight loss correction factor to
obtain a corrected weight loss for the sample.
In another aspect, the method includes the steps of placing a combustible
sample in a heated furnace, heating the sample while measurements
of the weight of the sample are made, determining a weight loss
rate function from the sample weight measurements, determining the
approximate time at which the onset of sample combustion occurs,
producing a weight loss correction factor using the time of combustion
onset and the weight loss rate function, and using the weight loss
correction factor to obtain a corrected weight loss for the sample.
The weight loss rate function may be suitably determined from the
sample weight measurements using regression analysis, such as least
squares regression analysis, or other known techniques. During the
initial heating of the sample prior to combustion, the weight loss
rate may be suitably modeled by a linear function, although other
functions could be employed. The time at which the onset of sample
combustion occurs can be ascertained in a number of ways. In one
embodiment or aspect, combustion onset may be determined by observing
the time at which the weight loss rate ceases to be linear, or departs
from linear by some threshold amount. In another embodiment or aspect,
combustion onset may be determined by monitoring the rate of change
in sample temperature or combustion chamber temperature and determining
therefrom the projected time at which the sample will reach a known
combustion temperature for the particular sample or some other selected
temperature. Still another approach involves monitoring the rate
of change in sample temperature or combustion chamber temperature
and determining the time at which the temperature change rate ceases
to be linear, or departs from linear by some threshold amount. Instead
of relying upon combustion onset time, it is possible to use other
values, such as combustion onset time less 10% or even a fixed time
interval. The appropriate method of determining the combustion onset
time or other value may depend, in part, on the design of the furnace.
The present invention also provides an apparatus for determining
weight loss of a sample, comprising a furnace, a scale mounted within
the furnace for measuring sample weight, a data store operatively
connected to said scale for storing sample weight measurements,
and a weight loss correction factor generator for generating a weight
loss correction factor using the sample weight measurements in the
data store. The apparatus may also include means for generating
a corrected weight loss measurement using a final sample weight
measurement from the data store and the weight loss correction factor.
Preferably, the weight loss correction factor generator comprises
means for determining a weight loss rate function from the sample
weight measurements in the data store, means for determining the
approximate time of combustion onset, and means for generating a
weight loss correction factor using the time of combustion onset
and the weight loss rate function.
Additional features and aspects of the invention will become apparent
from the detailed description which follows and from the accompanying
drawings, which are intended to be illustrative of the invention,
but not restrictive as to the scope and breadth of the invention.
BRIEF DESCRIPTION OF THE DRAWINGS
Having thus described the invention in general terms, reference
will now be made to the accompanying drawings, which are not necessarily
drawn to scale, and wherein:
FIG. 1 illustrates a preferred design of an analytical furnace
useful in the present invention;
FIG. 2 illustrates a typical burn cycle for a 1,500 gram sample;
FIG. 3 illustrates the mass or weight loss profile and temperature
profile of the first ten minutes of a typical burn cycle;
FIG. 4 illustrates the effect of heating on the weight of an empty
pan assembly; and
FIG. 5 illustrates a preferred design of a pan assembly useful
in the present invention.
DETAILED DESCRIPTION OF THE INVENTION
The present invention now will be described more fully hereinafter
with reference to the accompanying drawings, in which preferred
embodiments of the invention are shown. This invention may, however,
be embodied in many different forms and should not be construed
as limited to the embodiments set forth herein; rather, these embodiments
are provided so that this disclosure will be thorough and complete,
and will fully convey the scope of the invention to those skilled
in the art. Like numbers refer to like elements throughout.
FIG. 1 illustrates an analytical furnace 10 with a combustion chamber
12 and a door 20 that provides access to the combustion chamber.
The furnace 10 further includes a sample support 14 that is operatively
connected to an internal scale assembly (not shown) that measures
the weight of the sample during the combustion cycle. In one embodiment,
the support 14 comprises a pair of rails positioned to receive a
sample. The internal scale may be any known weighing device in the
art, such as a load cell.
A heat transfer plate 16 may be placed above the sample support
14. The sample, such as asphalt, is placed in a sample pan assembly
18 and placed in the combustion chamber 12 such that the pan assembly
rests on the plate 16 during heating. A temperature sensor (not
shown), such as a thermocouple, is strategically located near the
sample to measure chamber or sample temperature.
The plate 16 increases the transfer of heat into the sample and
aids in preheating the air as it enters the chamber 12 through holes
below the plate. To this end, the plate 16 is preferably made of
a material having high thermal heat capacity and good heat conductivity.
Particularly suitable are materials such as silicon carbide, aluminum
oxide and some metals. Quick transfer of heat into the pan assembly
18 assists in reducing the settling time and duration of thermal
instability.
Advantageously, a processor 26 is operatively connected to the
furnace 10. The processor 26 may be any combination of computer
hardware and software capable of performing calculations arid manipulating
data as needed to practice the present invention. Preferably, the
processor 26 includes one or more data stores for storing data such
as weight and temperature readings. Additionally, the processor
preferably includes a digital filter for filtering data measurements,
such as weight and/or temperature measurements, to dampen or reduce
oscillations and noise caused by the mechanical vibrations of a
thermally expanding system. The digital filter smoothes the response
of the weight and/or temperature data. The digital filter could
be as simple as an N pole low pass Butterworth-type filter, or even
an adaptive filter as known to one skilled in the art.
Typically, the furnace 10 is utilized to measure the weight of
a combustible portion of a sample by measuring the weight of the
sample before and after combustion of the combustible portion of
the sample. A commercially available furnace suitable for use with
the present invention is Model 4155B available from Troxler Electronic
Laboratories. As used herein, combustion refers generally to the
boiling, evaporation, thermal degradation and thermal decomposition
of the combustible portion of a sample.
FIG. 2 illustrates a typical burn cycle of a combustible sample
with time as the abscissa and weight loss in grams on the ordinate.
The burn cycle shows the weight loss that occurs during a combustion
cycle, from the time the sample is placed in the furnace until combustion
of the combustible portion of the sample is complete. At t-0, the
sample is loaded into the chamber and the process begins. Point
A indicates the beginning of the burn cycle. Between t=0 and point
A the system marks a beginning point commonly referred to as the
tare. Point B marks the end of the flame, and point C indicates
the end of the cycle, as determined by a slope of less than 0.1
g per minute. In an automated data collection system, the asphalt
content by weight is calculated using the difference between the
weight loss of paint C and the tare.
FIG. 3 shows the expanded view of the burn cycle concentrating
on the first 10 minutes. The material weight loss is shown in FIG.
3A while the chamber temperature is illustrated in FIG. 3B. When
the sample is first loaded into the chamber, the chamber temperature
begins to drop. This is evident from the response between points
A and B. The temperature then increases in a linear fashion to point
C where a positive identification of the burn is evident by the
deviation from the linear response.
As shown in FIG. 3A, at t=0, the weight loss increases in a linear
fashion up to point D. Between points D and E the function describing
loss deviates from a linear relationship. The ripple in the weight
curve is due to oscillations in the pan assembly 18. During the
initial heating of this assembly 18, the pans may warp and begin
to rock on the platform at a frequency determined by the rocking
moment and mass of this assembly and the sample.
It has been hypothesized that all the loss from t=0 to point E
of FIG. 3A was due to moisture evaporation from the sample. However,
FIG. 4 shows a similar weight loss response even with an empty pan
assembly 18. Note that FIG. 4 is a graph of total weight rather
than weight loss. This indicates that the weight loss during the
first few minutes is a phenomenon linked with temperature instabilities
and not moisture. The moisture error theory lead to the practice
of reading the weight loss directly off the curve of FIG. 3A at
point E, as taught by U.S. Pat. No. 5,279,971 to Schneider. Unfortunately,
the weight loss obtained by interpreting the correction factor exclusively
by the weight indicated by E also contains the boiling, evaporation
and thermal degradation of the asphalt material. Thus, the Schneider
method actually underestimates the weight of the asphalt binder
by incorporating some of the weight lost due to combustion of the
binder in the correction factor.
One of the largest effects pertaining to an inaccurate tare is
due to the difference in air densities and void content of the asphalt
sample. During the initial linear portion of the weight loss curve,
the relatively cold asphalt is out-gassing and becomes lighter.
The bombardment of the pan assembly 18 by energetic gas molecules
creates a transfer of momentum or force, which also decreases as
thermal equilibrium approaches. Furthermore, during this period
the airflow and turbulence created in the combustion chamber 12
begins to settle. When this process is near completion, evaporation
of the petroleum-based bitumen of the asphalt begins, as indicated
by point D of FIG. 3A. This is the beginning of the nonlinear response
of the weight loss curve. However, even though the outgassing has
decreased by point D, it is not complete, but merely accompanied
by evaporation up until the point of combustion located at point
E.
As previously stated, a problem arises when the tare is obtained
before thermal stability has been achieved. Unfortunately, with
a combustible sample, thermal stability is not achieved until after
point B in FIG. 2. Hence a weight loss correction factor, typically
measured in grams, is necessary to obtain the proper beginning weight.
The present invention provides a method of accurately determining
the weight loss of a sample during heating, wherein the method includes
using a weight loss correction factor to obtain a corrected weight
loss for the sample. The method includes placing a sample, such
as a combustible sample, in a heated furnace and heating the sample
while measurements of sample weight are made. The sample weight
measurements are used to determine a rate function. The rate function,
in turn, is used to produce a weight loss correction factor. Thereafter,
the weight loss correction factor may be used to obtain the corrected
weight loss of the sample.
The sample weight measurements define a weight loss curve, such
as illustrated in FIG. 3A, wherein the curve includes an initial
substantially linear portion and a subsequent non-linear portion.
Preferably, the rate function is a weight loss rate function comprising
a function corresponding to the substantially linear portion of
the weight loss curve. The weight loss rate function may be determined
by applying a regression analysis, such as a least squares regression
analysis, to the sample weight measurements. In a preferred embodiment,
the weight loss rate function comprises a linear function corresponding
to the substantially linear portion of the weight loss curve.
Once the rate function is determined, the initial substantially
linear portion of the weight loss curve may be linearly extrapolated
beyond the linear portion of the weight loss curve. Since it is
believed the linear portion of the weight loss curve is mainly attributable
to weight measurement errors caused by thermal instabilities as
discussed above, extrapolation of the linear portion of the weight
loss curve to the approximate point of combustion will provide a
weight loss correction factor that will negate the effect of thermal
instability on the measurement of weight loss of the sample. The
remaining weight loss (occurring after the approximate onset of
combustion) should be attributable to combustion of the combustible
portion of the sample, such as asphalt binder.
Thus, in a preferred embodiment of the invention, the onset of
combustion or approximate onset of combustion is determined in order
to ascertain the point at which changes in weight are no longer
attributable to thermal instabilities present in the furnace. As
shown in FIG. 3A, once this time is known, the weight loss correction
factor may be calculated by extrapolating the rate function to that
time.
For example, in one embodiment, when the approximate combustion
time has been determined, the calculated time is inserted into the
weight loss rate function to determine the weight loss correction
factor. Where a linear function is used, the time is multiplied
by the slope of the linear equation derived from the substantially
linear portion of the weight loss curve, and the intercept of the
linear equation is added to this result to obtain the weight loss
correction factor. Notice that this calculated value is much less
than the measured weight loss at this point in time, as the measured
weight loss has partially incorporated the combustion of the asphalt
binder, or other combustible portion, of the sample.
The approximate onset of combustion may be determined in a number
of ways. For example, as shown in FIG. 3A, the approximate onset
of combustion results in a departure of the weight loss curve from
a linear response. Thus, the onset of combustion may be determined
by determining the time at which the weight loss rate departs from
a linear function by a threshold amount. The threshold amount may
vary from zero to any suitable amount, such as about three grams.
In other words, the approximate point of combustion may be determined
as the time at which the actual measured weight of the sample deviates
from the extrapolated weight calculated using the rate function
by a threshold amount.
Similarly, the approximate onset of combustion may be determined
using the temperature profile as shown in FIG. 3B. As shown, the
chamber temperature or sample temperature departs from a linear
function at the approximate point of combustion. Thus, the onset
of combustion may be determined by monitoring the rate of change
in sample or combustion chamber temperature and determining the
time at which the rate of change of the monitored temperature departs
from a linear function by a threshold amount. For example, a rate
function for the substantially linear portion of the temperature
curve may be calculated and extrapolated. The actual temperature
of the sample or chamber may be compared to the extrapolated temperature
calculated using the rate function and the onset of combustion may
be determined as the time at which the two values diverge by a threshold
amount, such as about 10.degree. C.
Alternatively, the onset of combustion may be determined as the
time at which the sample temperature or combustion chamber temperature
reaches a predetermined temperature. Since asphalt typically ignites
at a temperature of about 460.degree. C., the onset of combustion
may be approximated by simply determining the time at which 460.degree.
C. is reached in the combustion chamber and using that time, and
the rate function discussed above, to calculate the weight loss
correction factor. Further, repeated experimentation with samples
of a known initial weight using the same furnace would enable the
user to determine the appropriate onset of combustion time without
reference to temperature data. For example, if several 1500 gram
samples are burned in a particular furnace, the user could estimate
the time to combustion for that sample size in that furnace type.
Once the weight loss correction factor is calculated, the final
corrected weight loss of the sample may be calculated by subtracting,
or otherwise applying, the weight loss correction factor from the
final measured weight loss of the sample. In this manner, the approximate
weight loss attributable to thermal instability or other external
factors is removed from the final weight loss calculation, resulting
in greater accuracy.
The integrity of this method is associated with the slope of the
weight curve. In practice, small masses reach equilibrium at a faster
rate than larger masses. Likewise, different installations and systems
will achieve thermal equilibrium at differing rates. In this case,
the response of the sample is related to the thermal capacity of
the furnace, the sample, and airflow of the installation. However,
these parameters are naturally accounted for through the slopes
of the linearized weight and temperature curves.
As explained above, a linear function is believed to adequately
model the weight loss due to thermal instability. However, other
functions known in the art could be used for the rate function.
For example, an exponential function, such as Equation 1, is believed
to accurately model weight loss of a sample due to thermal instability.
Equation 1 could be used to determine a weight loss correction
factor without determining onset of combustion where a single type
or model of furnace is utilized. If all furnaces that will utilize
the rate function are of the same type, so that characteristics
such as heat-up time, airflow and physical size of the chamber and
pan assembly are the same, a statistical sampling of the furnaces
can be used to determine the exponential coefficient, B. Likewise,
the manufacturer of a particular furnace type could calculate the
B coefficient for each individual furnace. For example, a non-combustible
material could be inserted in the furnace and several weight loss
versus time measurements could be taken in order to determine the
B coefficient.
Once the B coefficient is known for the particular furnace, only
two weight loss measurements are required in order to solve Equation
1 for the remaining two constants, A and C. Once A and C are known,
the weight loss correction factor may be calculated as the limit
of Equation 1 as t goes to infinity. Thus, using this method, the
weight loss correction factor is equal to the C constant.
When the pan assembly 18 is first placed into the chamber 12, it
undergoes rapid expansion, which results in disproportionate instabilities.
Hence, a delay is preferably incorporated into the method of the
present invention to ignore this period. For example, a delay of
about 20 to about 40 seconds may occur before sample weight measurements
are begun. The delay may be incorporated into the furnace 10 using,
for example, a delay timer incorporated into the processor 26. Following
the delay, sample weight measurements are begun.
Preferably, the sample or combustion chamber temperature is also
monitored to determine when the minimum occurs. This allows further
settling time to the pan assembly 18, and more importantly, generally
signals the beginning of the linear range of the chamber or sample
temperature curve. Preferably, temperature data collection begins
after the chamber temperature reaches the minimum. Typically, the
chamber temperature reaches the minimum value about two minutes
after the sample is placed in the furnace, but the time may vary
depending on furnace type. Once the temperature minimum occurs,
both weight and temperature data are collected and preferably continue
to be stored until the weight loss curve begins to become nonlinear.
Typically, for small samples, data collection can occur up to about
4 or 5 minutes, while larger masses remain linear for as much as
6 to 8 minutes. Preferably, the weight data is collected for at
least about two minutes to ensure that sufficient data is taken
to accurately determine the weight loss rate function. One way to
detect when the weight curve becomes nonlinear is to calculate the
residuals between the actual data and the corresponding linear curve-fitted
data. When the residuals become greater than some predetermined
value, then data collection is ceased. It is also possible that
each curve (T and weight) are individually analyzed, as the temperature
response remains linear long after the weight has deviated.
The weight and temperature measurements discussed above, and calculations
utilizing those measurements, may be stored and manipulated manually
or using processor 26. Preferably, the temperature and weight data
are fed into processor 26 and the processor performs all calculations
and curve-fitting functions.
A preferred design of the pan assembly 18 is shown in FIG. 5. During
the data collection period, rapid thermal expansion takes place
in the pan assembly 18. This expansion causes oscillations and reduces
the signal to noise ratio of the weight measurement. To reduce these
effects, a preferred design of the pan assembly 18 incorporates
a material with a low thermal expansion coefficient, such as stainless
steel. The pan assembly 18 is also perforated to allow oxygen to
flow into the sample, and the lower catch pan 22 has a cross break
to add mechanical strength and rigidity. The air gaps between the
sample pans 24 and catch pan 22 aid in oxidizing the asphalt while
decreasing the total burn time. Preferably, the pan assembly 18
further includes a perforated top cover 28 and a bail strap 30 to
hold the assembly in place.
The present invention provides a method of measuring weight loss
in an analytical furnace capable of consistently measuring weight
loss regardless of the furnace type, thermal capacity of the furnace,
thermal conductance and capacity of the sample, weight and void
ratio of the sample, installation variances, the temperature difference
between the sample and furnace and volume of the basket assembly.
Many modifications and other embodiments of the invention will
come to mind to one skilled in the art to which this invention pertains
having the benefit of the teachings presented in the foregoing descriptions
and the associated drawings. Therefore, it is to be understood that
the invention is not to be limited to the specific embodiments disclosed
and that modifications and other embodiments are intended to be
included within the scope of the appended claims. Although specific
terms are employed herein, they are used in a generic and descriptive
sense only and not for purposes of limitation. |