Performance of mechanically shaken indirect contact atmospheric dryer in drying pastelike materials

Abstract

Pastelike materials are encountered in many technological processes in chemical, pharmaceutical, foodstuff and natural product industries. The most important factor in the drying of this type of materials is the nature of the moisture bonding that occurs. Because of the different characteristics of pastes, it is impossible to recommend a universal type of dryer for all of these materials. Some of the dryers available provide only indirect contact with the drying agent (heat) and also maintain constant moisture with a system of rotating paddles. We evaluated the performance of this type by studying the dryer kinetics curves for ground coffee under a variety of operational conditions of moisture load of material, temperature of the heating plate, intensity of the mechanical mixing of the moisture material, and initial moisture. The effects of these parameters (except for moisture) were studied using a 2³ factorial design. According the analyses of the kinetics drying curves, was observed that the increase in the temperature of the plate and rotation as well as the decrease in the load facilitates more effective removal of moisture. In statistical analysis was determined that the load of the material and the heating plate temperature influence the final moisture content of the material and plate temperature modifies the final temperature of the solid. Also, was suggested linear models from the factorial design to describe the process of drying coffee grounds satisfactorily.

contact dryer; performance; coffee grounds

TRANSPORT PHENOMENA AND UNITY OPERATION

Performance of mechanically shaken indirect contact atmospheric dryer in drying pastelike materials

K. P. Melo; M. A. Cremasco^*

School of Chemical Engineering, State University of Campinas PO Box 6066, 13083-970, Campinas - SP, Brazil. E-mail: [e-mail]

ABSTRACT

Pastelike materials are encountered in many technological processes in chemical, pharmaceutical, foodstuff and natural product industries. The most important factor in the drying of this type of materials is the nature of the moisture bonding that occurs. Because of the different characteristics of pastes, it is impossible to recommend a universal type of dryer for all of these materials. Some of the dryers available provide only indirect contact with the drying agent (heat) and also maintain constant moisture with a system of rotating paddles. We evaluated the performance of this type by studying the dryer kinetics curves for ground coffee under a variety of operational conditions of moisture load of material, temperature of the heating plate, intensity of the mechanical mixing of the moisture material, and initial moisture. The effects of these parameters (except for moisture) were studied using a 2³ factorial design. According the analyses of the kinetics drying curves, was observed that the increase in the temperature of the plate and rotation as well as the decrease in the load facilitates more effective removal of moisture. In statistical analysis was determined that the load of the material and the heating plate temperature influence the final moisture content of the material and plate temperature modifies the final temperature of the solid. Also, was suggested linear models from the factorial design to describe the process of drying coffee grounds satisfactorily.

Keywords: contact dryer, performance, coffee grounds.

INTRODUCTION

Drying is used in a variety of industrial operations, with the objective to remove moisture, as vapor, from the material (Okada., 1987). Removal of this moisture reduces the volume and the mass of the product, making its transport and packing economically more advantageous. In addition, increases this storage time (Cook and Dumont, 1991). Pastelike materials are among those submitted to drying processes; these can be found in a variety of technological processes of chemical, pharmaceutical, and natural products industries (Strumillo et ii., 1983). According to these authors pastelike materials include hard pastes (such as pies and precipitates), soft pastes and sludge, suspensions, emulsions, and solutions. However, it is worth mentioning, that, due to the characteristics of the pastes, it is to recommend a standard dryer for paste processing and related operations. On the other hand, some of the dryers that can be used - are of the indirect contact type with mechanical shaking where by the drying agent, heat, is supplied to a metallic surface and, consequently, to the material being dried. This type of equipment will be studied in the present work, in which coffee grounds after the first extraction were used, as the test material in a atmospheric drying chamber. The performance of the equipment is evaluated by analyzing the drying kinetics curves as a function of several operational parameters, such as wet load of material, temperature of the heating plate, mechanical shaking, and moisture of the material.

Drying by Indirect Contact

Contact drying in shaking bed is often used to dry granular materials such as pharmaceutical products, mineral powders, agricultural products or foodstuffs, and pastes (Gevaudan and Andrieu, 1991).

According to Sztabert (1989), the drying mechanism of this type of equipment depends on the temperature of the hot surface, the temperature and properties of the raw material and mechanical operation of the dryer and the pressure (vacuum or atmospheric).

MATERIALS AND METHODS

This work was conducted in the Laboratory of Processes in Porous Media (LPMP) of the School of Chemical Engineering (FEQ) at the State University of Campinas (UNICAMP).

Materials

For the purpose of this paper, coffee grounds after the first extraction on an express machine were used as the test material. The principal characteristics physics of the coffee are in Table 1.

The density was determined in gas picnometer Accupyc 1330 and porosity in porosimeter Autopore III 9400, both were producing for Micromeritics Instrument Corp. The particle diameter was used such as Maialle and Menegalli (1996).

Equipment

The apparatus used in this work is illustrated in Figure 1. This unit consists of a stainless steel cylindrical drying chamber exposed to the atmosphere with 0.24 m diameter, 1.5 cm thickness and 0.11 m² drying area having an internal heater device, through which silicone oil is circulated at 350cP.

The equipment is isolated with rock wool and corrugated aluminum. The drying chamber is connected to the oil container (1) (Figure 1), which is coupled to two electrical resistance of 750W each (2). The oil is heated by these electrical resistances, which are controlled by the control panel (3), and a gear pump (4) circulates the heating fluid. Two valves (5a, 5b) control the pressure of the circulation device pump; one of these (5b) operates partway-open and is located after the gate valve by-pass. This by-pass (6) sends oil to the container or heater device (7). The equipment has a replacement oil container (8) and an oil distribution container (9). The shaking system consists of a continuous flow motor (11), a connection shaft, and a scraper, which is controlled by a voltmeter. The drying system also contains type T thermocouples (12a, 12b), installed at the input and output of the oil container. Measurement of pressure is monitored by a manometer (13) coupled to the line.

The wet material to be used in the dryer is weighed on model C&F semi-analytical balance, with a precision of 0.1 g. Loads used for determination of moisture were weighed on a model BG400 Gehaka analytical balance, with a precision of 0.001g. Moisture content was determined by the oven method at 104 ±5ºC for 24 hours. This method was carried out in a model 315SE Fanem oven. The temperature of the heating plate was recorded by a type T digital thermometer from IOPE Therm with a precision of 0.1ºC.

Experimental Procedure

Initially the temperature of the heating surface was adjusted by a resistance monitored by a control panel. If took two hours to reach the stationary regime in the drying chamber. Then, when the variation in the temperature of the heating surface was about 0.5°C, the regime was considered permanent. The criterion for drying time take into account the time for total removal of the moisture with weight 0.5 kg, i.e., when the moisture is keep constant.

Once the stationary regime was achieved, the wet material, whose mass had been previously measured, was added to the drying chamber. At the same time the shaking system, whose revolution had also been previously fixed, was activated. Samples were collected in the drying chamber every 5 min in same place, because the less moisture until the material temperature was practically constant. At the end of the period an some of the material homogenous was collected in a small, previously weighed container and the small container with material was weighed and put in the oven at 104±5ºC. After 24 hours the sample was taken out and weighed again and the moisture removed (d.b.) determined by the difference in mass as follows:

Factorial Design

The strategy of the experiments was defined by means of a 2³ factorial design generating a total of eight experiments. This design was used to verify to what extent the three independent variables, load of the material, temperature of the heating plate, and agitation speed, would be influence the response variables X_final and T_p,final.

The three independent variables used in this design were coded in the following way:

Var cod = +1 (highest level of the variable)

Var cod = -1 (lowest level of the variable)

The originals variables chosen to experimental and their levels are show in Table 2.

To analyze the trust worthiness and application of the models, three approaches were used: standard errors, probability (ANOVA), and regression coefficients.

RESULTS

Nine assays were carries out in the laboratory. Drying kinetics was studied using five selected experiments. The experimental conditions adopted to this studied are presented in Table 3 for further analysis, where T_p,0 is the initial temperature of the particle, T_w ,the temperature of the heating plate, W_s the material load, X₀, the initial moisture of the material, and n the rotation of the mixing device.

Figures (2) to (5) are the experimental kinetics drying curves in the form of moisture of the material on a dry weight basis expressed in kg of H₂O/kg of dry material for the time (fixed at 1h for all the experiments). It can be verified that for all curves only surface moisture is removed from the material. It can also be observed that initially the material is cohesive (until the fourth point of the curves). Starting at the fourth points the material behaves like powder. It can be seen in Figure 2 that this modification is faster as the temperature difference between the drying agent and the material to be dried becomes higher (when the temperature of the plate is 95ºC). So, this should result in a larger removal moisture. Considering the effect of material load (Figure 3), it can be seen that the moisture reaches about 0.13 d.b with the lighter load, while for the heavier load it reaches around 0.6 d.b. The difference between these values is due to the smaller amount of processed material. When the mixing velocity is increased, a small tendency towards a quicker drying appears, which tends to show that shaking provides on time of contact of the material with the warm surface (Figure 4). However, it can be observed that moisture removal with time varies in the same proportion for different values of initial moisture (Figure 5).

As mentioned previously, a 2³ factorial design was adopted in this work. In agreement with this design, the original variables, T_w, W_s and n, coded at their respective levels for each experiment, as well as the values obtained experimentally for response X_final (final moisture of the material) and T_p,final (final temperature of the particle) in the drying process, can be founds in Table 4.

Starting from experimental results, a statistical analysis was conducted with the aid of the software Statistica 5.0â with the purpose of evaluating which of the variables, independent or combined, influences X_final and T_p,final in a period of 1 h of drying. As a result of the analysis the values of the estimated effects of each variable and their iterations with the respective deviation pattern, as shown in the Tables 5 and 6, were obtained.

Analyzing the p. levels values in Tables 5 and 6, it can be verified that for the response variable X_final, the independent variables (T_w and W_s) are statistically significant, i.e., their values are into of the significant level (p<0.055), while, for variable T_p,final, only the temperature of the plate is significant. These results, however, have to be confirmed for a confidence level of 94.5%. That is why, the statistical parameter Pr (probability) from the variance analysis is used, as shown in Tables 7 and 8.

Based on the statistical parameter Pr (probability), the influence of variables T_w and W_s on X_final and T_w on T_p,final to be confirmed and the no significant effects are considered to be aleatory experimental errors.

According to the results in Tables 7 and 8, a linear statistical model for the responses, X_final and T_p,final, is proposed. The coefficients of these models can be found in Tables 9 and 10.

These models are described by equations (2) and (3) as a function of the variables that had a significant influence on the responses X_final and T_p,final, in accordance with the regression coefficients presented in Tables 8 and 9. The correlation coefficients of 0.9972 for X_final and 0.9942 for T_p,final suggest that these linear models represent the experimental data satisfactorily.

CONCLUSIONS

Conclusions drawn on the influence of the operational parameters (load of the material, temperature of the heating plate, and rotation) on the drying process, based on the statistical analysis and on the results of the experimental drying kinetics curves, are as follows:

1) In accordance with the analyses of the kinetics drying curves, it can be observed that the increase in the temperature of the plate and rotation as well as the decrease in the load facilitates more effective removal of moisture.

2) By means of the statistical analysis one can determine that the load of the material and the heating plate temperature influence the final moisture content of the material and plate temperature modifies the final temperature of the solid.

3) It can be verified that the linear models from the factorial design describe the process of drying coffee grounds satisfactorily.

NOMENCLATURE

d_p particle diameter, m; n rotation of the mixing device, rpm; t drying time, min; T_p,0 initial temperature of the particle, ºC; T_p,final final temperature of the particle, ºC; T_w temperature of the heating plate, ºC; X removed moisture content of the material, kg/kg; X₀ initial moisture content of the material on a dry weight basis, kg/kg; X_final final moisture content of the material on a dry weight basis, kg/kg; W_s material load, kg; r_s density, kg/m³ e_p porosity.

Received: January 6, 2003;

Accepted: February 20, 2004

• Cook, E.M. and Dumont, H.D. Process Drying Practice. New York: Editora McGraw-Hill, 1991.
• Gevaudan, A. and Andrieu, J. Contact Drying Modelling of Agitated Porous Alumina Beads, Chem. Eng. Process, vol. 30, pp.31-37, 1991.
• Maialle, K.G.; Menegalli, F.C. Secagem de Borra de Café em Leito de Jorro Fluidizado Bidimensional: estudo preliminar. Anais do XXIV Congresso Brasileiro de Sistemas Particulados. UFSCar, Săo Carlos, p. 301-306, 1996.
• Okada, M., Fundamentos Sobre a Secagem de Sólidos. Manual de desidrataçăo de frutas e hortaliças do ITAL, vol. 18, pp. 1-38, 1987.
• Strumillo, C. Markowski, A. and Kaminski, W, Modern Developments in Drying of Pastelike Materials, Advances in Drying, vol. 2, pp. 193-231, 1983.
• Sztabert, Z. T, Size Selection of Vacuum Contact Dryer with Mechanically Mixed Particulate Material, Drying Technology, vol. 7, nş. 1, pp. 71-85, 1989.

*

To whom correspondence should be addressed

Publication Dates

• Publication in this collection
  14 July 2004
• Date of issue
  Sept 2004

History

• Received
  06 Jan 2003
• Accepted
  20 Feb 2004