Physical and Thermal Properties of Pistachios

Transcription

1 J. ugric. Engng Res. (1991) 49, Physical and Thermal Properties of Pistachios M.-H. Hsu; J. D. MANNAPPERUMA: R. P. SINGH Department of Agricultural Engineering, University of California, Davis, CA 95616, USA (Received 2.2 November 1989; accepted in revised form 17 February 1991) Physical and thermal properties of pistachios were evaluated as functions of moisture content at room temperature. The moisture content of pistachios ranged from 40% wet basis at harvest time to a minimum of 5%. The length, width and thickness of pistachios increased with increasing moisture content as represented by third-degree regression equations. The bulk density, specific heat, thermal conductivity, and bulk thermal conductivity increased linearly with moisture content. The specific gravity, surface area, and bulk thermal diffusivity decreased linearly with moisture content. 1. Introduction Pistachio is the edible seed of the pistachio tree (Pisfucia ueru L.). Mostly the cultivar Kerman is grown in California. There has been a dramatic increase in production of pistachios in California during the past 20 years. A special feature of pistachios is its split shell. This feature makes roasted and salted pistachios an attractive snack food for eating out of hand. Splitting of the shell takes place while the nut matures on the tree. The harvested nuts undergo a wet hulling process to remove the mushy hull covering the shell. Unsplit nuts, which have a lower density, are separated from split nuts by floatation in water. The unsplit nuts have a lower market value hence they are processed separately from split nuts which have a prime market value. The moisture content of pistachios vary between 3540% (wet basis) after dehulling and separation processes. It must be reduced to 5% (wet basis) for safe storage and further processing. Despite an extensive literature search, no published information was found on properties of pistachios useful for the design of drying and storage systems. There is considerable information in the literature on measurement of properties of other agricultural materials (Morita and Singh; Kustermann ef al. ; Singh3). The objective of this study was to determine the physical and thermal properties, length, width, thickness, bulk density, specific gravity, surface area, specific heat, thermal conductivity, bulk thermal conductivity and bulk thermal diffusivity of pistachios as functions of moisture content. 2. Method and materials 2.1. Sample preparation Pistachios used in this study were of Kerman cultivar, obtained from Pistachio Producers of California located in Madeira, California, during the 1987 harvest season. Samples were packaged in double-layer plastic bags to avoid moisture loss, and kept in a controlled temperature chamber at 2 C for 5 d to allow moisture equilibrium within samples. The initial moisture content of the samples ranged from 35-40% wet basis. The desired moisture levels in pistachios were obtained by forcing unheated ambient air 311 OO Zl-8634/91/ Silsoc Research Institute

2 312 PHYSICAL AND THERMAL PROPERTIES OF PISTACHIOS A average surface area of a pistachio, cm2. (Y bulk thermal diffusivity of pistachios, m2/s. cp specific heat of pistachios, J/&g C) CW average specific heat of water, J/&g C) At, measured temperature rise in pistachios, C. Atw measured temperature water and calorimeter, drop in C. 4 temperature correction factor for heat loss surroundings, C. y specific gravity of pistachios. k thermal conductivity of pistachios, W/(m C). kb bulk thermal conductivity of pistachios, W(m C). L average length of a pistachio, mm. M moisture content, % wet basis. mp mass of pistachios in the calorimeter, kg. Notation m, mass of water in the calorimeter, kg. q power consumed by probe heater, W/m. 81, 02 temperature of probe thermocouple at times t, and t2, C. p bulk density of pistachios, r2 kg/m3. correlation coefficient. S the slope of the linear portion of the temperature versus In(time) curve, C. T average thickness of a pistachio, mm. t,, t2 two arbitrary times from the initiation of the line heat source, s. W average width of a pistachio, mm. IV, heat capacity of calorimeter, J/C. through shallow beds of pistachios. The moisture content of pistachios was determined by an oven method (120 C for 48 h). The emphasis of this study was on split-shell pistachios. Therefore, all the closed-shell samples were manually sorted and removed Measurement of length, width, and thickness One hundred pistachios were randomly selected and labeled for easy identification. The samples were conditioned in a manner described in Section 2.1 to obtain seven different moisture levels ranging from 7*6%-37.5% wet basis. The length, width, and thickness of 100 pistachios were determined using an electronic digital caliper (Sears Craftsman, Max-U) at each moisture level. The length, width, and thickness of pistachios were obtained by direct measurement of the distance from tip to tip, the maximum diameter, and the distance from the highest point to the lowest point of pistachios when piaced on a horizontal plate respectively of each of the one hundred pistachios Measurement of bulk density and specific gravity The bulk density of pistachios was determined by measuring the weight of a pistachio sample in a 80.2 ml container (Mohsenin4). An air comparison pycnometer (Beckman Model 930) was used to measure the true volume of pistachio sample. Samples conditioned at eight different moisture contents were used to determine the bulk density and specific gravity of pistachios. Ten replications were made for each moisture level.

3 M.-H. HSU ET AL Measurement of surface area The surface area of pistachios was determined by the coating method proposed by Hedlin and Collins. In this method, uniform surface coatings are applied on pistachios and on plastic spheres and the weight of the surface coatings, and the known surface area of the plastic spheres are used to determine the surface area of pistachios. A slow drying varnish was used to bind a fine nickel powder ( mm particle size, manufactured by Fisher Scientific Company) on the surface of the objects to produce a uniform surface coating. A number of precautions were taken to ensure uniformity of the surface coating. Seven pistachios and four plastic balls, weighed separately, were used in each experiment. The pistachios and plastic balls were vigorously shaken with 1.5 g of slow drying varnish for 20 s in a 300 ml flask. A mixture of 500 g of fine sand (100% through 20 mesh) and log of varnish was prepared separately by shaking in another flask. This pre-coated sand was mixed with the pistachios and plastic balls and shaken for another 30 s. This process removed any excess varnish from the objects. The pistachios and plastic balls were then transferred to an empty plastic jar and shaken to remove all the sand. The pistachios and plastic balls were then mixed with 2 g of nickel powder in a 300 ml Pyrex flask and shaken for 60 s to apply a pre-coating. The mixture was then transferred to a cylindrical metal box about three-quarters full of nickel powder. The final coating was done by shaking the mixture in the box for 30s. To prevent the nickel powder from sticking in the cracks and holes of split pistachios, the well-coated mixture was placed in an empty cylindrical metal box and tapped gently. The coated pistachios and plastic balls were then weighed separately to obtain the final weights. The surface area was calculated using Eqn (1). 4 Weight of coating on seven pistachio nuts X Surface area of a plastic ball A=? Weight of coating on four plastic balls The surface area measurements were conducted at seven different moisture levels with six replications at each moisture level. (1) 2.5. Measurement of specific heat The specific heat of the pistachios was determined by the method of mixtures (Maglic et al6). The pistachios of known mass and temperature were dropped into a calorimeter of known heat capacity containing water. Equilibrium temperatures for the mixture of water and pistachios were recorded and the enthalpy change of pistachios was calculated from the heat exchange between water, calorimeter and pistachios. The average specific heat of pistachios was calculated using Eqn (2). (Muir and Viranvanichai7). m,c,(at, + $I= mwcw(atw - $I+ w4&., (2) The temperature of pistachios and water was monitored using a 27-gauge copperconstantan thermocouple and a Hewlett Packard 3497A data acquisition/control unit. Since the calorimeter was a composite of glass, metal, and insulation material, it was found easier to determine its heat capacity experimentally (Wratten et al8). The heat capacity of the drop calorimeter was experimentally determined to be 104 J/OC. The temperature correction factor was also determined experimentally. It was found that the temperature change inside the calorimeter was O*l C in 1 hour. The specific heat of pistachios was determined at five moisture levels, and five measurements were repeated for each moisture level.

4 314 PHYSICAL AND THERMAL PROPERTIES OF PISTACHIOS 2.6. Measurement of thermal conductivity The thermal conductivity of pistachios was determined using the line heat source method which is the most commonly used transient state methods (Maghc et al. ). This method is based on the assumption of an infinitely long and infinitesimally thin line heat source. Once embedded in the test material, the line heat source is energized and the temperature rise at a short distance from the source is monitored over a brief time period. Thermal conductivity is then calculated using equation (3) which was developed by Hooper and Lepper. k = 4 ln WJ (3) 4Jr(& - 6) The temperature is plotted against the natural logarithm of time. The resultant graph is a straight line with a curved portion at the beginning due to initial transient and towards the end due to the temperature profile reaching the outer boundary of the object. The equation is valid only within the linear portion. The thermal conductivity is calculated by Eqn (4) using the slope of the linear portion. k=& (4) A thermal conductivity probe with known heater resistance ( ) and thermocouple resistance ( ) was used in the experiment. It consists of a hypodermic needle, within which there is an axial constantan electrical resistance wire insulated over its length and grounded to the tube at the lower end. An electrical current can pass through this wire to provide a heat source of constant power per unit length. Within the tube near the centre of its length is the hot junction of an E-type thermocouple. The thermal conductivity was determined at seven moisture levels and five measurements were repeated for each moisture level. Three pistachios were pierced through and threaded onto the probe and then pressed together. A power supply was used to provide constant power output at 5 V. A. Hewlett Packard 44468A data acquisition & control ROM was connected to the thermocouple and used to monitor the temperature for 10 min at 5 s interval. The temperature was plotted against the natural logarithm of time. Linear regression analysis was conducted for the linear portion of the curve using Statview software (Brain Power Inc.). The slope of the line was used to calculate the thermal conductivity Measurement of bulk thermal conductivity The bulk thermal conductivity of pistachios was determined using the line heat source theory of transient heat transfer analysis (Sweat and Haugh ). The apparatus (Fig. I) used in this experiment consisted of a 102 mm inside diameter and 305 mm long aluminum cylinder. This cylinder was insulated with 25 mm thick polyurethene on the circumference and two 38 mm thick Styrofoam pads at the ends. The heating element was a 203 mm long single strand of 27 gau e chrome1 resistance wire soldered to 17 gauge copper extension wire (Morita and Singh $ ). The heating element was stretched and fixed on two holders (a and b in Fig. 1) to maintain its position at exactly the central point along the aluminum cylinder which effectively reduced the error in temperature measurement. Four other thermocouples placed along the holder (b) were used to indicate the equilibrium status to begin the experiment. The power to the heating wire was supplied by a low voltage regulated DC power supply. The current was measured by a Fluke 8024A multimeter. The temperature was

5 M.-H. HSU ET AL. 315 t 305mm. _.. -f f.::. -!5mm ::... :..C Y...., YY. : , :. :. : ::: (.I. -.::..::. :::. :: :. :,, _ _. :::..:..,.... _ :.!: + 102mm dla - + 6mm r y Styrofoam / Holder (a) A Chrome1 resistance - Holder (b) wire t DC power supply \ Aluminium cylinder.l _.., Thermocouple plugs / Fig. 1. Apparatus for bulk density measurement of pistachio nuts measured using a 27-gauge copper-constantan thermocouple attached to the centre of the heating wire. A Hewlett Packard 3497A Data Acquisition/Control Unit was connected to the thermocouple and used to monitor the temperature change of the heating wire. The pistachios were poured into the cylinder. Several gentle tappings were required during filling to obtain uniform density. After the temperature inside the cylinder reached equilibrium, the power was turned on and the temperature rise during the trial was recorded for 20 min at 5 s intervals. The temperature was plotted against the natural logarithm of time and the best fitting straight line portion was chosen. The slope of the line was determined by linear regression on the chosen linear portion and used to calculate the bulk thermal conductivity of pistachios. The bulk thermal conductivity measurements for pistachios were conducted at five different moisture levels with five replications at each moisture level Calculation of bulk thermal diffusivity The bulk thermal diffusivity of pistachios was calculated from experimentally determined values of bulk thermal conductivity, specific heat and bulk density using Eqn (5).

6 316 PHYSICAL AND THERMAL PROPERTIES OF PISTACHIOS Once the linear regression equations of thermal conductivity, bulk density and specific heat were established, six moisture levels, 35, 25, 20, 15, 10 & 7% were chosen as evaluation points. These moisture contents were substituted into the linear regression equations to derive values of bulk thermal conductivity, bulk density and specific heat. The bulk thermal diffusivity was readily calculated by substituting the derived values into Eqn (5). 3. Results and discussion 3.1. Length, width, and thickness Average values of 100 measurements for the length, width, and thickness of pistachios were obtained at seven different moisture contents ranging from 7.6% wet basis to 37.5% wet basis. These values are given in Table 1. Each physical dimension was fitted as a third degree polynomial function of moisture content and the regression Eqns (6), (7) and (8) were derived for the length, width and thickness. L = M x 10-3M x lo- ji# (r* = 0.982) 16) W = 13XKl+ OGI30M + O-5995 x 10-3M x 10P6M (r = 0.985) (7) T = M x 10e3M x 10e6M3 (r = O-976) (8) The average length of pistachios decreased from mm as moisture content decreased from 37.5% wet basis to 7.6% wet basis. This was a 6% decrease in the initial length. The total decrease in width was 0.5 mm or about 4% of the initial width. The total decrease in thickness was O-93 mm which is 7% of the initial thickness. The thickness decreased sharply in the initial drying phase but the rate of decrease was small as the moisture content decreased Bulk density Results of bulk density measurement conducted at eight different moisture contents ranging from 8.2% to 35.0% (wet basis) are shown in Fig. 2. The bulk density was found to be a linear function of moisture content [Eqn (9)]. p = M (r* = 0.959) (9) The opening in the pistachio shell increases as moisture content decreases. Due to the combined effects of the void space inside the pistachio she11 and the shrinkage of linear dimensions, the number of pistachios within a constant volume container remains Table 1 Variation of lengtb, width nod tbicknes of pistachio nuts with moisture content Moisture content, % wb Length, mm Width, mm Thickness, mm Average Std. deu Average Std. dev Average Std. dev , , a M

7 M.-H. HSU ET AL Moisture content (M).% w b Fig. 2. Bulk density on a function of moisture content practically the same. However, the kernel weight decreases. Therefore, the bulk density decreases with decreasing moisture content Specific gravity The specific gravity was found to be a linear function of moisture content as expressed by the regression Eqn (10). Fig. 3 shows the experimental values and linear relationship between specific gravity and moisture content. y = x lo- M (r2 = O-975) (10) The specific gravity of pistachios increases as the moisture content decreases. In measuring the specific gravity, the air comparison pycnometer measures the true volume of the sample without including the void space. The decrease in true volume was more significant than the decrease in weight. Therefore, the specific gravity of pistachios increases as the moisture content decreases. Moisture content (MI,% w.b. Fig. 3. Specific graviv as a function of mosture content

8 318 PHYSICAL AND THERMAL PROPERTIES OF PISTACHIOS Moisture content (M I,% w. b. Fig. 4. Surface area as a function of mokture content 3.4. Surface area The results of surface area measurements performed at seven different moisture contents ranging from 35-O-6.2% (wet basis) are shown in Fig. 4. The values of surface area may seem high at the first glance. However, the surface area in this study refers to the area which is actually exposed to the heating or cooling medium which includes the inside shell area and the kernel surface area. Therefore, the effective surface area is about twice as large as the surface area of the outer shell. The surface area was found to be a linear function of moisture content as expressed by the regression Eqn (11). A = M (r2 = 0.969) (11) The surface area of pistachios increases as the moisture content decreases. There are two major factors which are involved in the change of surface area, namely, the shrinkage of the physical dimensions and the opening of the pistachio shell. The shrinkage of length, width and thickness results in the decrease of the surface area. The opening of the shell extends as the moisture content decreases which actually exposes more of the kernel Moisture content (M 1.I w.b. Fig. 5. Specific heat as a function of moisture content

9 M.-H. HSU ET AL Moisture content (M1,Xw.b. Fig. 6. Thermal conductivity as a function of moi&ure content surface area to the drying or cooling medium. The second factor (opening of shells) was found to be more dominant than the first factor (shrinkage of the pistachios). Therefore, the overall result was an increase of surface area along the drying process Thermal properties The experimental data for the various thermal properties measured in this study showed a linear relationship with moisture content. Both the experimental points and the linear relationships are plotted in Figs 5, 6, 7, and 8 for specific heat, thermal conductivity, bulk thermal conductivity and bulk thermal diffusivity, respectively. The statistical by Hsu. 2 arameters obtained for each of the experimental measurements are presented The linear relationships are presented by Eqns (12)-(15). The regression i 0.025) I I 1 m Moisture content (MI,% w.b. Fig. 7. Bulk thermal conductivity as a function of moisture content

10 320 PHYSICAL AND THERMAL PROPERTIES OF PISTACHIOS Fig Moisture content CM),% w.b. Bulk thermal diffusivity as a function of moisture content coefficients show a good fit of the data. cp = M (r2 = 0.920) k = O x 10-3M (r2 = 0.963) kb = x 10-3M (r* = 0.922) cr = 51.1 x 1O x 10-9iW (r2 = 0.983) (12) (13) (14) (15) 4. Conclusions All the physical and thermal properties of pistachios studied in this work varied significantly with moisture content. The direction of change of most of the properties agree with the trends reported by Morita and Singh and Wratten ef al.* for rough rice. The only differences were in surface area and specific gravity. In both cases, Wratten et al.* reported an increase in their values with increase in moisture content while this study found decreasing trends. This discrepancy can be attributed to the behaviour of the split shell in the pistachios. The assistance appreciated. This to Agriculture. Acknowledgements provided by MS F. H. Guo in conducting experiments is greatly project was supported by California Committee on Relation of Energy References Morita T., Singh R. P. Physical and thermal properties of short-grain rough rice. Transactions of the American Society of Agricultural Engineers 1979, 22(3): * Kustermann M.; Scherer, R.; Kutzbach, H. D. Thermal conductivity and diffusivity of shelled corn and gram. Journal of Food Process Engineering 1981, 4: Sigh, R. P. Thermal diffusivity in food processing. Food Technology 1982, 36(2):87-91 Mobsenin, N. N. Thermal Properties of Foods and Agricultural Materials. New York: Gordon and Breach Science Publishers, 2nd edition pp , 1980

11 M.-H. HSU ET AL. 321 HecHin, C. P.; CoUins, S. H. A method of measuring the surface area of granular material. Canadian Journal of Chemical Engineering 1961, 2: a Maglic, K. D.; Cezaidiyan, A.; Peletsky, V. E. Compendium of thermophysical property measurement methods. Vol 1. Survey of Measurement Techniques New York,: Plenum Press 1984 Muir, W. E.; Viianvanichai, A. Specific heat of wheat. Journal of Agricultural Engineering Research 1972, 17(4): Wratten, F. T.; Poole, W. D.; Chesness, J. L.: Bal, S.; Ramarao. Physical and thermal properties of rough rice. Transactions of American Society of Agicultural Engineers 1969, l2(6): hoper, F. C.; Lepper, F. R. Transient heat flow apparatus for determination of thermal conductivities. ASHVE Transactions 1950, 56: lo Sweat, V. E.; Haugh, C. G. A thermal conductivity probe for small food samples. Transactions of the American Society of Agicultural Engineers 1974, 17(l): Hsu, M. Physical and thermal properties of pistachio. M. S. Thesis, University of California, Davis, CA, 1988