MATHEMATICAL MODELING OF THIN LAYER DRYING OF SALTED YELLOWTAIL FISH UNDER OPEN SUN AND IN GREENHOUSE DRYER

The thin layer dryer model was used to describe the characteristics of changes in water content of salted yellowtail dried under the open sun and by using a greenhouse dryer. Thirteen different thin layer dryer models were used to predict fish water content values. The results of the conducted experiments validated the values. From the results of modeling for open sun drying, it was discovered that Modified Henderson and Pabis were the most suitable models to be used. While for drying using greenhouse dryers, it was suitable to use the Diffusion Approach model and Verma et al. From the performance of modeling indicator, it was shown that the value of the correlation coefficient (R) approaching 1, where the mean square of deviation between experimental, predicted values and root mean square error analysis (RMSE) have infinitesimal values.

Muhfizar (2018) conducted an experimental study of yellowtail fish drying under an active greenhouse dryer.
The thin layer equation describes the overall drying phenomenon, regardless of the control mechanism. This equation has been used to estimate the drying time of some products and to generalize the drying curve (Akpinar & Bicer, 2008). Several studies have used the thin layer model for the fish drying process. Kituu et al. (2010) used mathematical thin layer models for the Tilapia fish drying process in the solar tunnel dryer. Guan, Wang, Li, and Jiang Guan, Wang, Li, and Jiang (2013) used thin layer modeling for fresh tilapia fillets using hot air convection. Sobukola and Olatunde (2011) conducted thin layer modeling for the drying process of African catfish with different brine concentrations and temperatures. Jain and Pathare (2007) used thin layer modeling for the drying process of shrimp and chelwa fish (Indian minor carp) in open sun drying. Bai, Li, Sun, and Shi (2011) used mathematical modeling for drying fish slices using the electrohydrodynamic (EHD) drying layer method. Darvishi et al. (2013) conducted thin layer modeling for sardine fish dried in microwave heaters.
Yellowtail fish is a type of consumption fish that has essential economic value and is one type of reef fish that lives in warm waters around the Indo Pacific. In this study, yellowtail drying was carried out under open sun drying and greenhouse solar dryer. The mathematical thin layer dryer for drying yellowtail fish has not been done yet. Therefore, this study aims to model salted yellow fish as thin layer dryers under open sun drying and solar greenhouse dryers.

METHODS OF RESEARCH
Material and Experimental Procedure. The greenhouse dryer shown in Figure 1 has a parabolic roof with paving blocks as a base. The surface area is 6.5 m 2 , with dimensions of length and width are 3.25 m x 2 m. The greenhouse length sides lead to the north and south in order to suit the movement of the sun. Polyethylene plastic was used for the material of greenhouse cover. The greenhouse framework made of galvanized pipes. Six 12 VDC exhaust fans with 100 WP solar cell as the electric energy supplier are used to circulate air inside the greenhouse. The tests were carried out with two methods, i.e., open sun drying and under the greenhouse dryer. The drying was done at Politeknik Kelautan and Perikanan Sorong for three days in December 2017 to reduce the water content of fish products. The yellowtail fish was used in this study. It was dried for 8 hours every day. The fishes were given salt of 0.2 gr NaCl / gr of fish mass on the dry salted method. The mass of fish used during the experiment about 0.3 kg either for open sun drying or under the greenhouse dryer. The dried fishes for three days were dried further by using the oven for 24 hours at a temperature of 105 °C (Mujaffar & Sankat, 2005). The method was done to find out the initial water content of the fish.
In this study, DHT 22 sensors were used to measure air humidity and temperature both in the environment and inside the greenhouse by using a calibrated data logger microcontroller. The data recording of air humidity and the temperature were set once every hour during the drying process. A digital scale was used to measure the changes of fish mass during the drying process. The fish mass data was manually recorded to measure changes in mass.
Model   Eq.

RESULTS AND DISCUSSION
In the drying process for three days, as shown in Figure 2, the air temperature in the greenhouse dryer ranges from 32 -54°C. The air temperature in the greenhouse tends to be higher than the ambient air temperature which only ranges from 29.8 -39°C. Unlike the temperature, the value of air relative humidity in the greenhouse tends to be lower compared to ambient relative humidity. The range of air humidity values in the drying greenhouse ranges from 24.90 -58.71%, while the air humidity value of the environment ranges from 43.26 -82.15%.  Figure 3 shows the change in water content of fish (dry basis) with two drying methods are with open sun and in the greenhouse dryer. The drying process in the greenhouse dryer looks faster to reduce fish water content compared to the open sun drying method. It proves that the drying rate of fish products in the greenhouse is higher than open sun drying, which is in line with the high air temperature and low relative humidity in it. Moisture ratio data from experimental results and calculations using thin layer drier modeling was processed using a statistic program on the computer. Each thin layer model was evaluated using the correlation coefficient (R), the mean square of predicted and predicted values (χ 2 ), and the root means square error analysis (RMSE) in (Eq. 2-4). The results of statistic computer program analysis obtained shown in Tables 2 and 3. The suitable model for describes open sun drying of yellowtail fish with dry salting method of 0.2 gr NaCl / gr mass of fish as shown in Table 2 Table 3. The value of R = 0.9975, χ 2 = 0.0002, and RMSE = 0.0121 for both models have the same value even though the constants in the two equations are different.     The modeling results in the drying of yellowtail fish with open sun and drying greenhouses were validated using the results of each experiment. Figures 4 and 5 shows a comparison between predictive values from modeling that match the experimental results for both drying methods. This is evidenced by the experimental data indicated by an asterisk (figure 4) or a circle (figure 5) generally around a straight line which is predictive data from modeling.

CONCLUSION
Modeling of thin layer dryer in this study was used to model the drying process of salted yellowtail fish (0.2 gr NaCl / gr of fish mass) in the open sun and greenhouse dryer. Thirteen models were used to describe changes in moisture content characteristics in both drying methods. Modified Henderson and Pabis is a suitable model to describe the drying process with an open sun with a value of R = 0.9934, x 2 = 0.00053, and RMSE = 0.01896. Diffusion approach and Verma et al. are two suitable models describing the drying process in a greenhouse dryer with values R = 0.9975, x 2 = 0.0002, and R.MSE = 0.0121.