DETERMINATION OF THE SPECIFIC HEAT OF AGRICULTURAL MATERIALS: PART II. EXPERIMENTAL RESULTS Lambert Otten and George Samaan School of Engineering, University of Guelph, Guelph, Ontario NIG 2WI. Received 9 August 1978 Otten, Lambert and George Samaan. 1980. Experimental results. Can. Agric. Eng. 22: Determination of the specific heat of agricultural materials: Part II. 25-27. Experimental heat capacity measurements of a variety of agricultural products are presented as a function of their moisture content for the temperature range of 30-90°C. A number of values found in the literature are included for comparison. In many cases the agreement is not particularly good, which may be attributed to the method used by other investigators and differences in the samples of the products. INTRODUCTION The interest in heat capacity values of agricultural products resulted from work done by Otten (1974) on the thermal conductivity of such materials. The frequency response method used in that work required accurate values of heat capacity. A survey of the literature clearly showed a considerable lack of reliable data; it was therefore decided to initiate an experimental study to verify and com plement the literature data. After consid ering the various types of calorimeters available, the continuous adiabatic calorimeter was selected. It is relatively easy to design, construct and operate; furthermore, it can be used for a wide variety of products with sufficient accuracy. The first attempt at producing an accurate continuous adiabatic calorimeter was successful and was described by Otten and Ezeike (1976). Much was learned from the initial experiments and a second apparatus was designed and tested by Samaan (1978). EXPERIMENTAL CONSIDERA TIONS 1. Choice of Inert Fluid Water is a major constituent of agricultural materials which gain or lose moisture in response to their environ ment. During the experiments with inert, non-agricultural materials, the fluid in the sample cell was water. It was introduced in the cell to promote heat transfer from the heater to the sample. Use of water as the inert fluid with agricultural material was questioned because absorption could influence the results. A number of initial experiments confirmed this suspicion and it was concluded that another liquid that is not absorbed had to be used. After testing a number of liquids, mineral oil was found to satisfy all criteria; namely, it was not absorbed and did not react with the materials. It also has a high thermal diffusivity so that heat transfer is rapid. Since no specific heat value of the oil was available, four experiments were run. The average specific heat of the oil for the experimental range of 30-90°C was determined to be 2.189 J/g°C with a standard deviation of 0.034 and a coefficient of variation of 1.5%. Sub sequent experiments on aluminum pellets using the oil instead of water produced specific heat values within the experimen tal range reported by Otten et al. (1980). 2. Sample Preparation Each sample was cleaned from all impurities and broken seeds. To have samples at different moisture contents, a high moisture content batch was dried in a ventilated oven at 50°C and samples were removed at various time intervals. Each sample was then sealed until testing. All moisture content determinations were done using the ASAE Standard Method (1976). This method was adopted because Hart et al. (1959) have shown, by heating many types of unground seeds in an air oven at 130°C, that an error of less than 0.3% as compared with the Karl Fischer method is obtained. 3. Heating Rate The power input to the heater was adjusted to keep the heating rate below l°C/min. The resulting temperature difference between the cell wall and the center of the sample was always less then 3°C. EXPERIMENTAL RESULTS 1. Wheat The specific heat values of two samples of soft, white wheat and one sample of red, hard, spring wheat were obtained at four or five different moisture contents. For purpose of identification, the first two samples will be referred to as Ontario CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980 (1976) (crop year) and Frederick (1977) while the third one is Western (1976). Three experiments were run on each sample and each moisture content. In each case the reproducibility was within ± 1.5%. It was also observed that the variation of specific heat with tempera ture was 0.5% in the 30-95°C tempera ture range so that only the average specific heat is reported. The temperature-time plots were straight lines with less than 0.2 residual error. A regression analysis of the data of each of the three samples provided the relationships between specific heat and moisture content shown in Table I. The results showed only slight variations between the three samples and an average correlation equation was obtained for all the wheat data (Table I and Fig. 1). The variation between this equation and the other three is less than 5%. An attempt to express the specific heat ofwheat as the sum of contributions of the specific heats of dry matter and water provided no useful information. Figure 1 is a graphical comparison of our results with those found in the literature. Pfalzner (1951) used the method of mixtures on three samples of hard wheat for the temperature range 0-65cC. Although samples A and B were for the same year, there is a significant difference in the results. Sample C was a different year. Moote (1955) determined the specific heat of No. 1 Northern Manitoba grade wheat at 30°C. An ice calorimeter was used by Disney (1954) to measure the specific heat values for Manitoba and Bersee hard wheat for a temperature range of 0-25°C. The most frequently quoted results for wheat are those obtained by Kazarian and Hall (1965) with a drop calorimeter on white wheat (0 < T < 25°C). The figure shows that our experimental results compare favorably with those 25
Transcript
DETERMINATION OF THE SPECIFIC HEAT OF
AGRICULTURAL MATERIALS: PART II. EXPERIMENTAL RESULTS
Lambert Otten and George Samaan
School of Engineering, University of Guelph, Guelph, Ontario NIG 2WI.
Received 9 August 1978
Otten, Lambert and George Samaan. 1980.Experimental results. Can. Agric. Eng. 22:
Determination of the specific heat of agricultural materials: Part II.25-27.
Experimental heat capacity measurements of a variety of agricultural products are presented as a function of theirmoisture content for the temperature range of 30-90°C. A number of values found in the literature are included forcomparison. In many cases the agreement is not particularly good, which may be attributed to the method used by otherinvestigators and differences in the samples of the products.
INTRODUCTION
The interest in heat capacity values ofagricultural products resulted from workdone by Otten (1974) on the thermalconductivity of such materials. Thefrequency response method used in thatwork required accurate values of heatcapacity. A survey of the literature clearlyshowed a considerable lack of reliable
data; it was therefore decided to initiate an
experimental study to verify and complement the literature data. After considering the various types of calorimetersavailable, the continuous adiabaticcalorimeter was selected. It is relativelyeasy to design, construct and operate;furthermore, it can be used for a widevariety of products with sufficientaccuracy.
The first attempt at producing anaccurate continuous adiabatic calorimeter
was successful and was described byOtten and Ezeike (1976). Much waslearned from the initial experiments and asecond apparatus was designed and testedby Samaan (1978).
EXPERIMENTAL CONSIDERA
TIONS
1. Choice of Inert Fluid
Water is a major constituent ofagricultural materials which gain or losemoisture in response to their environment. During the experiments with inert,non-agricultural materials, the fluid in thesample cell was water. It was introducedin the cell to promote heat transfer fromthe heater to the sample. Use of water asthe inert fluid with agricultural materialwas questioned because absorption couldinfluence the results.
A number of initial experimentsconfirmed this suspicion and it wasconcluded that another liquid that is notabsorbed had to be used. After testing anumber of liquids, mineral oil was foundto satisfy all criteria; namely, it was notabsorbed and did not react with the
materials. It also has a high thermaldiffusivity so that heat transfer is rapid.
Since no specific heat value of the oilwas available, four experiments wererun. The average specific heat of the oilfor the experimental range of 30-90°Cwas determined to be 2.189 J/g°C with astandard deviation of 0.034 and a
coefficient of variation of 1.5%. Sub
sequent experiments on aluminum pelletsusing the oil instead of water producedspecific heat values within the experimental range reported by Otten et al. (1980).
2. Sample PreparationEach sample was cleaned from all
impurities and broken seeds. To havesamples at different moisture contents, ahigh moisture content batch was dried in aventilated oven at 50°C and samples wereremoved at various time intervals. Each
sample was then sealed until testing.All moisture content determinations
were done using the ASAE StandardMethod (1976). This method was adoptedbecause Hart et al. (1959) have shown, byheating many types of unground seeds inan air oven at 130°C, that an error of lessthan 0.3% as compared with the KarlFischer method is obtained.
3. Heating RateThe power input to the heater was
adjusted to keep the heating rate belowl°C/min. The resulting temperaturedifference between the cell wall and thecenter of the sample was always less then3°C.
EXPERIMENTAL RESULTS
1. Wheat
The specific heat values of two samplesof soft, white wheat and one sample ofred, hard, spring wheat were obtained atfour or five different moisture contents.
For purpose of identification, the first twosamples will be referred to as Ontario
CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980
(1976) (crop year) and Frederick (1977)while the third one is Western (1976).
Three experiments were run on eachsample and each moisture content. Ineach case the reproducibility was within± 1.5%. It was also observed that the
variation of specific heat with temperature was 0.5% in the 30-95°C temperature range so that only the averagespecific heat is reported. Thetemperature-time plots were straight lineswith less than 0.2 residual error.
A regression analysis of the data ofeach of the three samples provided therelationships between specific heat andmoisture content shown in Table I. Theresults showed only slight variationsbetween the three samples and an averagecorrelation equation was obtained for allthe wheat data (Table I and Fig. 1). Thevariation between this equation and theother three is less than 5%.
An attempt to express the specific heatofwheat as the sum of contributions of the
specific heats of dry matter and waterprovided no useful information.
Figure 1 is a graphical comparison ofour results with those found in theliterature. Pfalzner (1951) used themethod of mixtures on three samples ofhard wheat for the temperature range0-65cC. Although samples A and B werefor the same year, there is a significantdifference in the results. Sample C was adifferent year.
Moote (1955) determined the specificheat of No. 1 Northern Manitoba gradewheat at 30°C.
An ice calorimeter was used by Disney(1954) to measure the specific heat valuesfor Manitoba and Bersee hard wheat for atemperature range of 0-25°C.
The most frequently quoted results forwheat are those obtained by Kazarian andHall (1965) with a drop calorimeter onwhite wheat (0 < T < 25°C).
The figure shows that our experimentalresults compare favorably with those
25
presented by Moote (1955) and Kazarianand Hall (1965) even though ourtemperature range is different (30 < T <95°C).
2. Corn
Specific heat values for yellow dent(Pioneer 3965) corn were measuredexperimentally at different moisturecontents. The variation of specific heatwith temperature was ± 0.4% in theexperimental range 30-95°C. Thus theaverage value of the specific heat in theabove range was considered. Thetemperature-time plots were straight lineswith less than 0.1 residual error. The
experiments were repeated three times ateach moisture content and the reproducibility of these experiments was found tobe ± 1.4%.
A polynomial of the formY =AB + Atxl + A2x2
was found to provide the best fit to thedata. The least squares method was usedto obtain the coefficients.
Therefore:
Cp = 1.178 + 0.0627 M - 8.7 x 10"4M2
Kazarian and Hall (1965) found thespecific heat of yellow dent corn to be(0.0 < T < 25°C)
C„ = 1.466 + 0.0356 M
3. Rye, Oats, Barley, Peas, Soybeansand Peanuts (Grown in Ontario)
The temperature-time plots were againstraight lines with less than 0.5 residualerror. The specific heat as a function oftemperature was found to be less than0.8% in the temperature range 30-95°C.The reproducibility of the experimentswas found to be ± 1.5% by repeating theexperiment three times for each moisturecontent.
The relationship between specific heatand moisture content for the above
products is given in Table II and Figs. 2and 3.
Haswell (1954), using the samemethod and procedure as Disney (1954),reported the specific heat of oats as (0 < T< 25°C)C„ = 1.277 + 0.0326M 11.7 <M< 17.8%
Watts and Bilanski (1970) reported thespecific heat of soybeans at 7.4% w.b. as1.891 J/g°C for the temperature range of30-129°C.
Alam and Shove (1973) used themethod of mixtures and obtained the
following correlation for soybeans in thetemperature range of 12-28°C.C„ = 1.6369 + 1.927 x 10"2M 0 < M < 38%
Young and Whitaker (1973) used thedifferential scanning calorimeter (DSC)and found the specific heat of peanuts
26
TABLE I. THE REGRESSION EQUATIONS FOR THE THREE TYPES OF WHEAT
Moisture
Regression equation Residual content No ofM.C.
Type of wheat Cp (J/g°C) R2 error range (%) samples
Ontario 1.272+0.0531 M 0.953 0.007 0<M<15.9 5
Western 1.377 + 0.0443 M 0.986 0.001 0<M<I3.8 4
Frederick 1.311 +0.0482 M 0.997 0.001 0<M<13.1 4
Avg of the three types 1.317+0.0491 M 0.991 0.001 0<M<15.9 13
2.2
2.0 -
1.8 -
o
•P1 1.6
I i 1 1 1 0 1B
-
A\s
/ /sSs
K
B_ Best fit to:
0 ONTARIO
-
D ,•^ Pic) EJ WESTERN
<•> FREDERICK
l 1 1 1
"Ptal
1 1
K KAZARIAN 119651
M_ MOOTE (1955)
D DISNEY 11954]
PFALZNER 11951)
P(al sample AP(bl B
Pic) .. C-
1.4
1.2
1.0
0 3 6 9 12 15 18
MOISTURE CONTENT (%w.b.)
Figure 1. Specific heat versus moisture content for wheat. (30 to 95°C).
TABLE II. REGRESSION RESULTS FOR CORN, RYE, OATS, PEAS, BARLEY ANDSOYBEANS
Barley 1.186 - 1.28 x lO"4 M + 0.0023M2Soybeans 1.296 + 0.075M - 0.0016A/2
0.990
0.957
0.998
0.995
0.998
0.995
8 n 16 20 24 28
MOISTURE CONTENT (% w b.)
Figure 2. Specific heat — moisture content results for barley, peas and rye (30-95°C).
CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980
0<A/<30 9
0<M<13 4
0<M<13 4
0<M<22 6
0<Az<15 4
0<A/<I7 5
32 36
O SOYBEANS
o CORN
O OATS
H HASWELL 119541
K KAZARIAN 11965)
A ALAM (1973)
12 16 20 24 28 32
MOISTURE CONTENT f/.wb)
36
Figure 3. Specific heat versus moisture content for soybeans, corn and oats (30-95°C).
(unspecified variety) at 57°C to be 1.739J/g°CatM = 5.4% w.b. and 1.630 J/g°Cat 0.0% w.b. Their results differ from
ours by about 15%.
DISCUSSION
The experimental results show that inthe temperature range of about 30-95°C,the temperature variation of specific heatof the grain tested is always less than 1%.It is therefore possible to use an averagevalue for this temperature range.
In contrast to the temperature dependence, the specific heat of all grainsincreased considerably with an increasein moisture content. This is, of course, asexpected because water absorbs heatreadily. Except for barley, all other grainsshowed a linear or nearly linear relationship between heat capacity and moisturecontent. Additional experiments on barley confirmed the trend shown in Fig. 2.
The reproducibility of the results wasfound to be always within 1%. Thisindicates that the apparatus and techniqueprovide an acceptable degree of precision; however, aside from the results
obtained by Otten et al. (1980) onaluminum, copper and ceramic pellets, itis very difficult to comment on theaccuracy of the results, especially theagricultural ones. A comparison with theavailable literature values is of little
assistance because many of them wereobtained in a questionable manner.Furthermore, there is a significantdifference in the results obtained from
different varieties of the same type ofgrain. For example, Pfalzner's (1951)results are not only different for thesamples obtained in successive years butalso for those of the same year.
The obvious conclusion that must bedrawn from the above comments is that
one must be careful when using a singlehandbook or literature value of specificheat in situations, such as dryingcalculations, where the selected value hasa significant effect on the outcome. Insuch cases it is necessary to use the fullrange or a representative average of allavailable data or to perform experimentalmeasurements on the actual material to be
dried.
CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980
ACKNOWLEDGMENTSThe financial support of the National
OTTEN, L.,G.Y. I. SAMAAN, andG. O. I.EZEIKE. 1980. Determination of thespecific heat of agricultural products: Part I.Continuous adiabatic calorimeter. Can.Agric. Eng. 22(1): 21-24.
PFALZNER, P.M. 1951. The specific heat ofwheat. Can. J. Tech. 29: 261-264.
SAMAAN, G. Y. I. 1978. Determination ofthe specific heat of granular and agriculturalproducts. M.Sc. thesis, School of Engineering, University of Guelph, Guelph,Ont.
WATTS, K. C. and W. K. BILANSKI. 1970.Calorimetric determination of the specificheat of soybeans. Can. Agric. Eng. 12(1):45-47.
YOUNG, J. H. andT. B. WHITAKER. 1973.Specific heat of peanuts by differentialscanning calorimetry. Trans. Amer. Soc.Agric. Eng. 16(3): 522-524.
