Abstract
Heat transfer by individual modes is estimated during baking of rice (Oryza sativa) pan cake (Dosa), a traditional food. The mathematical expressions proposed could be used to modify the baking oven for controlling the individual modes of heat transfer to obtain the desired product texture, colour and flavour. Conduction from the rotating hot plate is found to be the most prominent mode of heat transfer and is critical for obtaining the desired product characteristics such as texture and flavour. Temperature profiles along the thickness of Dosa are obtained and compared with those obtained experimentally. Heat transfer parameters such as thermal conductivity and emissivity of Dosa are determined (0.42 W/m K and 0.31, respectively). The effect of material of construction of the hot plate such as alloy steel, teflon coated aluminum, cast iron and stainless steel on product texture was studied and stainless steel was found to give good surface finish to the product, which was confirmed by scanning electron microscope. Sensory evaluation was carried out to evaluate the product acceptability. The thermal efficiency of the baking oven was 51.5%.
Keywords: Indian rice pan cake (Dosa), Modes of heat transfer, Thermal conductivity, Emissivity, Temperature profiles, Energy balance
Introduction
Indian traditional foods based on different cereals/grains are popular for their unique taste and flavor (Sarojini et al. 1996a, b; Roopa et al. 1998; Tripathy et al. 2003). Indian rice pan cake (Dosa) is one of the most popular traditional foods. Dosa is a breakfast food made from the batter of ground rice, blackgram/dhal and water, which is allowed to ferment (Vasan and Kousalya 1982; Patil et al. 2001) for about 15–17 h at a temperature between 313~318 K.
In case of a conventional Dosa, hot plate (tava) is used domestically for baking and the tava is heated at the bottom by any of the heat sources such as firewood, charcoal, coal, kerosene or liquefied petroleum gas (LPG). Conduction from hot plate and radiation from the hood/top cover are considered most important for the desired product quality. It is a general feeling that heat transfer by radiation significantly contributes for baking of Dosa. One of the objectives of the study was to verify how far this is true.
The process of bread preparation has been studied for its attributes by Adler (1958). Role of individual modes of heat transfer for baking of biscuits has been reported by Standing (1974) using band oven. Importance of quality of wheat flour for preparation of chapathi an Indian traditional food has been studied (Chaudhri and Muller 1970). Tandoori roti, a popular Indian traditional product is baked using a Tandoor oven, which is an in ground oven consisting an earthen pot surrounded by bricks. The analysis of modes of heat transfer in Tandoor oven has been reported (Saxena et al. 1995).
In all these studies, it was acknowledged that the mode of heat transfer is more important than just supplying the required quantity of heat for obtaining the desired product characteristics such as flavor, crustiness and color. Hence, it was thought desirable to analyze the effect of individual modes of heat transfer in baking of Dosa on these characteristics. Rheological properties of the Dosa batter was studied by Bhattacharya and Bhat (1997) which is important for batter standardization. A mathematical expression developed to analyze the contributions of the individual modes of heat transfer, which is expected to be useful in improving the Dosa characteristics as well as in modifying the design of the baking oven. The study also includes the analysis of the temperature profiles (experimental and theoretical) along the thickness of Dosa. The experiments were also conducted with different hot plate materials (tava) such as alloy steel, teflon coated aluminum, cast iron and stainless steel to study its effect on texture of Dosa. The product was examined under a scanning electron microscope to understand the structural changes and the crater density is attributed to the surface finish of Dosa. A sensory analysis was carried out from the trained panelists to evaluate Dosa made from baking oven.
Theoretical aspects
Considering the baking of Dosa made from rice and blackgram batter spread on a rotating hot plate (tava), the main mechanisms of heat transfer to the product are conduction from the rotating hot plate and radiation from the hood. Convection heat transfer is not considered, as the air/flue gas flow is very less. Thus total heat transferred to Dosa (QT) can be written using expressions given elsewhere (Kern 1950; McCabe et al. 1995; Heldman and Singh 1993) as,
 |
1 |
Where,
 |
2 |
and qC is heat transferred by conduction, kJ; K is the thermal conductivity of Dosa, W/m K; Ab the area of the product bottom in contact with the hot plate, m2; TpbDosa bottom surface temperature which is in equilibrium with the hot plate temperature and equal to it K; Tpt the temperature of Dosa top surface at the end of the baking period, K and x the thickness of Dosa, m. qR is the heat transferred by radiation, kJ; σ the Stefan-Boltzman constant, W/m2 h K4; TH the hood (refractory surface) temperature, K, Fpr overall coefficient for radiation heat transfer (Charm 1975; Heldman and Singh 1993), fpr is geometrical factor; εp the emissivity of Dosa; εH emissivity of the hood and Ar area of the radiating refractory surface, m2.
The total heat transferred must be equal to the total heat absorbed by Dosa. Considering the gross temperature rise of Dosa (sensible heat increase) and the latent heat of vaporization of evaporated moisture, the total theoretical heat absorbed (QA) can be expressed as,
 |
3 |
where, m is average mass of Dosa—batter, kg; Cp average heat capacity, kJ/kg K; Td the Dosa—batter temperature at the beginning of the baking period, K; t the Dosa baking time, s; L the moisture loss during baking, kg; Tpt the temperature of the product top surface at the end of the baking period, K and λv latent heat of water evaporation, kJ/kg.
It would be of interest to obtain the temperature profiles across the thickness of the product. The time-temperature-distance relationship for the body of uniform temperature subject to sudden source of heat can be represented by the most general equation of this type (Perry and Green 1984).
 |
4 |
where, T is the temperature of product K; C1, C2, C3, p and q are constants. The initial and boundary conditions for the present case (an infinite slab heated on one face) are,
 |
where, Ts is the hot plate temperature, K.
 |
The above equation is valid only when, C2 = 0, otherwise Td would have to vary with x, whereas it is assumed to be uniform. Hence,
 |
Substituting in Eq. 4, results in
 |
5 |
where, f1 (x/2√αt) denotes the error integral and α thermal diffusivity, m2/s. For different values of x, α and t, the factor (x/2√αt) was calculated to obtain the corresponding error integral. The values of the integral are obtained from literature (Perry and Green 1984).
Material and methods
Baking oven
The baking oven employed in the present study was designed and developed at the Institute (Venkateshmurthy and Jayaprakashan 2000, 2002). The cross sectional view of the automatic Dosa machine is shown in Fig. 1. The baking oven consists of a geared motor, a spreader assembly and a revolving hot plate (hot plate rotates at 0.5 rpm having thermal conductivity of 54 W/m K) assembly. A circular burner with accessories is used for heating the hot plate and heating is from the bottom of the hot plate (hot plate has a diameter of 1.2 m and thickness of 0.12 m) and a scraper is used for scraping of Dosa from the hot plate. The main frame with necessary insulation houses all the parts of the machine and batter pump attached to a vessel is coupled to a solenoid valve. A set of sprocket assembly is used for the drive of the hot plate and the end of the main drive is housed inside anti-friction bearings. An electronic timer controls the volume of the batter and the temperature of the hot plate is controlled by a temperature controller. All the electrical parts of the machine are housed inside an electrical panel board. The whole assembly is mounted on swivel castors for easy movement of the device to the required place. The baking oven has a capacity to produce 360 Dosas/h. At any given time there will be around 12 Dosas on the hot plate of the baking oven.
Fig. 1.
Automatic pan cake (Dosa) machine
Preparation of Dosa
Batter was prepared by grinding soaked rice (Oryza sativa) and urd dhal/blackgram dhal (Phaseolus mungo) with water in known proportions (4:1:5) for a predetermined time to get the required fineness. The batter was allowed for auto fermentation for 15–17 h. The Dosa batter was charged (0.1 kg/Dosa or 100 ml/Dosa) to a batter dispenser and on to the preheated rotating hot plate (443~453 K) of the baking oven and the batter was spread to a size of 0.2~0.23 m in diameter with a thickness of 0.002~0.004 m or 2~4 mm by a spreader and oil was sprayed (0.008~0.010 kg/Dosa or 8~10 g/Dosa) on to the product which will not only improve the sensory properties of the product but also aid in rapid and uniform heat transfer. The baked products were scarped and discharged for consumption. The average baking time of the product was about 120 s. The average value of the moisture content of the batter and the baked product were 65 and 35%, respectively.
Determination of thermal conductivity (K)
Dosa was baked on the hot plate by discharging a known amount of batter of predetermined composition as described above. The top of the product was covered by an asbestos plate of 0.005 m or 5 mm thickness to prevent radiation from the stainless steel hood and to allow the product baking by conduction alone. The probe of the temperature indicator was positioned through a hole at the center of asbestos disc to measure the product surface temperature. Thermal conductivity was calculated from these test results by using appropriate terms in Eqs. 1 and 3.
Determination of product emissivity (εp)
Dosa batter was spread on the asbestos sheet and placed on to the hot plate with another asbestos insulation at the bottom (to avoid conduction heat transfer). The batter was baked by radiation alone from the hood of the machine. The temperature of the hood was recorded. Emissivity was calculated from these test results by using appropriate terms in Eqs. 1 and 3.
Determination of the temperature profiles (T)
The uniform thickness of Dosa batter was spread on to the hot plate, which was maintained at uniform temperature. A cylinder with a hole for insertion of the thermocouple was used for measuring the temperature at different depths of the Dosa. Thickness gauge having thickness varying from 0.0005 to 0.002 m or 0.5 to 2 mm was used for varying the depth of insertion of the thermocouple for measuring the temperature at different depths/thickness of Dosa. A digital temperature indicator (Model–TFF 200, Make–EBRO, Germany, PT-100, Range: −50 to 200 °C) was employed to measure the product temperature and had a resolution of 0.1 °C. Increase in temperature of the product batter was recorded at different time intervals from 30–120 s at different depths.
Determination of microstructure
Dosa was prepared using hot plates of different material such as cast iron, stainless steel, steel and teflon coated aluminum. The Dosa samples prepared on different hot plate materials (tava) were viewed under the scanning electron microscope (model-Leo 435VP of Leo Electron Microscope Ltd., Cambridge, UK) at a magnification of 500. The thermal conductivity of the different hot plate materials was in the range of 16~54 W/m K.
Sensory analysis
Dosa was prepared on the baking oven of the automatic Dosa machine. Sensory characteristics of the product were tested among 10 panelists using quantitative descriptive analysis (Stone et al. 1974).
Results and discussion
In order to test the exclusive effect of conductive heat transfer, Dosa was heated from the bottom only (while preventing radiation by covering the product with an asbestos disc) on the hot plate of the baking oven at three different preset temperatures. Thermal conductivity of product was calculated by setting the total heat absorbed by the product equal to the expression for conduction. The results of these experiments are given in Table 1 and the average thermal conductivity of the product was found to be 0.42 W/m K.
Table 1.
Thermal conductivity (K) as a function of hot plate temperature
| Trial No. | Hot plate temp K | Total heat W | Final temp. of Dosa K | Mass of batter kg (Initial) | Mass of Dosa kg (Final) | Thermal conductivity W/m. K | |
|---|---|---|---|---|---|---|---|
| (T s ) | (Q A ) | (T pt ) | (m) | (m 1 ) | (K) | ||
| Sensible Heat, (Q2) | Latent heat, (Q3) | ||||||
| 1 | 458.00 | 192.31 | 532.29 | 353.50 | 0.1 | 0.072 | 0.44 |
| 2 | 443.50 | 188.84 | 451.33 | 347.60 | 0.1 | 0.076 | 0.43 |
| 3 | 445.00 | 191.58 | 459.79 | 348.50 | 0.1 | 0.076 | 0.40 |
The expressions for radiation from hood/cover surfaces given in equations are simplified by assuming the value of both hood/cover emissivity (εH) and the geometric factor (fpr) to be unity, which are valid assumptions under the conditions of the present investigation, as
 |
6 |
When the product was baked only through radiation heat, as described earlier, there was no appreciable change on the product surface (except for the formation of a thin hard layer). The characteristic holes on the top of the product were also absent. This could be attributed to the fact that the product was not being in contact with the surface of the hot plate, indicating the importance of conduction heating in baking of product. The result of test runs is given in Table 2. Using Eq. 3, the total theoretical heat absorbed by the test product was calculated using temperature and moisture data given in Table 1. The value of specific heat (Cp) is taken as 1.83 kJ/kg for computing the sensible heat requirement of the product (Gupta 1990). The emissivity of the product surface (εp) calculated by setting the total theoretical heat absorbed by the product equal to the expression for emissivity in Eq. 6, was found to be 0.31. This value was used in subsequent calculations. Equation 6 can be written as,
 |
7 |
Table 2.
Emissivity of rice pan cake (Dosa) (ε p ) as a function of refractory surface temperature
| Trial No. | Hood temp K | Final temp. of Dosa K | Mass of batter kg (Initial) | Mass of Dosa kg (Final) | Emissivity of Dosa |
|---|---|---|---|---|---|
| (T H ) | (T pt ) | (m) | (m 1 ) | (ε p ) | |
| 1 | 414.0 | 313.5 | 0.1 | 0.103 | 0.29 |
| 2 | 433.3 | 323.2 | 0.1 | 0.093 | 0.31 |
| 3 | 440.0 | 330.6 | 0.1 | 0.092 | 0.33 |
The complete heat transfer model as expressed by Eqs. 1 and 3 was then applied to the baking in the baking oven. The total heat absorbed [Eq. 3] was taken as a sum of latent heat and sensible heat. On conducting several experiments each term of Eq. 1 was calculated for this purpose. It can be noted that approximately 532 W of the heat absorbed by the product was latent heat of evaporation (Q3) of water while sensible heat (Q2) was about 192 W. It is interesting to note that, of the total heat transferred (QT) to the product, about 98% was transferred by conduction from the rotating hot plate and about 1.5% was transferred by radiation from the hood/cover. The heat transfer distribution clearly indicates the significant contribution from conduction, which was crucial for the characteristic flavor and crustiness of the Dosa. Thermal efficiency of the baking oven was found to be about 51%.
The temperatures across the thickness of the product were estimated using the Eq. 5. The temperature profiles, both experimental and theoretical are shown in Fig. 2 and it may be noted that the trend of these profiles are similar. However, the experimental temperatures are much lower than the theoretical ones. The reduction in the temperatures of the experimental value could be attributed to the evaporative cooling that occurs during baking of Dosa.
Fig. 2.
Temperature profiles across the thickness of rice pan cake (Dosa) (Experimental/Theoretical)
Dosa was prepared using different hot plate material such as cast iron, stainless steel, steel and Teflon coated aluminum. The product was viewed under the scanning electron microscope to see the pattern of craters formed due to the evaporation of moisture during the baking process. It was clearly seen from the Fig. 3 that, material such as stainless steel and teflon coated aluminum had smaller and uniform craters of large numbers as compared to the cast iron and alloy steel material. This can be attributed to the fact that the material with lower thermal conductivity has uniform distribution of heat and also lower rate of heat transfer compared to alloy steel and cast iron. The difference in density of the craters was clearly seen in better surface finish of the product prepared using stainless steel and teflon coated aluminum hot plate materials.
Fig. 3.
Microstructure of rice pan cake (Dosa) made on different hot plate material
The sensory analysis was carried out among the 10 panelists using quantitative descriptive analysis. The profilogram (Fig. 4) indicated high overall acceptability of Dosa by the panelists (scored 11 out of 15).
Fig. 4.
Profilogram of product made using automatic Dosa machine (n = 15 panelists)
Conclusion
In baking of Dosa made from rice and blackgram batter conduction heat transfer is found to play the most prominent role in determining its characteristic flavour and texture. Hence, the amount of heat supplied by individual modes of heat transfer is to be controlled rather than the total heat supplied to the product in order to achieve the desired product characteristics. The mathematical expressions are of significant use in estimating the magnitude of the heat transfer and contribution from each individual heat transfer modes, which in turn is useful for design modifications of the burner of the hot plate of the baking oven. The experimental temperature profiles across the thickness of the product indicated similar trend of the theoretical ones, however, the values are lower than the theoretical ones which could be attributed to evaporative cooling that takes place during baking of Dosa. The sensory analysis indicated a reasonably good over all quality of the product baked on the baking oven.
Acknowledgement
Authors thank Prakash V, Director, CFTRI, Mysore for his keen interest in design and development of machinery for Indian foods. The authors also wish to thank Jayaprakashan SG, Shiva kumara M and Mahesh I for their help during fabrication of automatic Dosa machine and the trials on the baking oven. The authors also wish to thank the sensory panelists and also the staff of CFTRI departmental canteen.
References
- Adler L. Bread making in the land of Bible. Cereal Sci Today. 1958;3:28–31. [Google Scholar]
- Bhattacharya S, Bhat KK. Rheological properties of Dosa batter. J Food Eng. 1997;34:429–435. doi: 10.1016/S0260-8774(98)80034-4. [DOI] [Google Scholar]
- Charm SE. Fundamentals of food engineering. 2. Westport: AVI Publ Co; 1975. [Google Scholar]
- Chaudhri AB, Muller HG. Chapaties and chapati flour. Milling. 1970;152:22–25. [Google Scholar]
- Gupta TR. Specific heat of Indian unleavened flat breads (Chapati) at various stages of cooking. J Food Process Eng. 1990;13:217–227. doi: 10.1111/j.1745-4530.1990.tb00069.x. [DOI] [Google Scholar]
- Heldman DR, Singh PR. Introduction to food engineering. 2. New York: Academic; 1993. [Google Scholar]
- Kern DQ. Process heat transfer, International Student edn. Tokyo: McGraw-Hill; 1950. [Google Scholar]
- McCabe WL, Smith JC, Harriot P. Unit operations of chemical engineering. 3. New York: McGraw-Hill; 1995. [Google Scholar]
- Patil RT, Krishna J, Jaswant S, Bargale PC. Testing of a convection type cylindrical dryer for production of instant Soy-Dosa mix. J Food Sci Technol. 2001;38:111–115. [Google Scholar]
- Perry RH, Green D. Perry’s chemical engineer’s handbook. 6. New York: McGraw-Hill; 1984. [Google Scholar]
- Roopa MR, Urooj U, Shashikala P. Rate of invitro starch hydrolysis and digestibility index of Ragi- based preparations. J Food Sci Technol. 1998;35:138–142. [Google Scholar]
- Sarojini G, Nirmala G, Chaturvedi A. Studies on utilization of Rajmah (Phaseolus Vulgaris)flour in some Indian traditional products. J Food Sci Technol. 1996;33:159–161. [Google Scholar]
- Sarojini G, Nirmalamma N, Chaturvedi A. Product development and evaluation of whole grain and flour blends of Amaranthus paniculatus. J Food Sci Technol. 1996;33:345–347. [Google Scholar]
- Saxena DC, Haridas Rao P, Raghavarao KSMS. Analysis of modes of heat transfer in Tandoor oven. J Food Eng. 1995;26:209–217. doi: 10.1016/0260-8774(94)00057-G. [DOI] [Google Scholar]
- Standing CN. Individual heat transfer modes in band oven biscuit baking. J Food Sci. 1974;39:267–271. doi: 10.1111/j.1365-2621.1974.tb02872.x. [DOI] [Google Scholar]
- Stone H, Sidol J, Oliver S, Woolsay A, Single ton RC. Quantitative descriptive analysis: developments, application and the future. Food Technol. 1974;28(11):28–34. [Google Scholar]
- Tripathy S, Vijayalakshmi D, Jayaprakasha HM. Utilization of whey protein concentrate in ragi (Eleusine coracana) based food products. J Food Sci Technol. 2003;40:157–161. [Google Scholar]
- Vasan BS, Kousalya K. Use of rice germ in the common food preparations. J Food Sci Technol. 1982;19:161–162. [Google Scholar]
- Venkateshmurthy K, Jayaprakashan SG (2000) Device useful for automatic discharge of Dosa and other similar Indian traditional Foods. Indian Patent IN230984
- Venkateshmurthy K, Jayaprakashan SG (2002) Improved device useful for making Dosa and other similar products. Indian Patent IN 226241