Modeling Grain Losses in Mechanized Harvesting of Oily Sunflower

Sunflower planting is mostly carried out for two particular purposes; oil production and as nut. Harvesting is one of the biggest problems in both types of sunflower. The difficulty of harvesting and less scientific research have led us to study the mechanized harvesting of this kind of crops. In this research, head losses and grain losses for the inner section of combine were investigated during mechanized harvesting of oily sunflower and a regression model was used based on the experimental tests for head losses and grain losses in the inner section of the combine.

After preparing an especial head for harvesting sunflower, the head was set up on the combine for measuring the harvest losses. The cutting, threshing and clearing process for sunflower seeds were done during the tests. The design of the head is the same as the sunflower bushes are firstly bent by the bar and then sequentially the cutting, and transferring processes are done. The tests were implemented in an oily sunflower farm by a combine harvester (1055 john deer) in 3 replications. The farm performance was 2170 kg ha-1 and was located in Kermanshah province in Iran. A pre-test was done to define the best combine forward speed and finally 2.5 km h-1 was adjusted for combine forward speed. The bar height (BH) in two levels (20 and 70 cm) and head height (HH) in two levels (60 and 120 cm) were independent parameters to evaluate the head. The dependent parameters were the combine losses and head losses. For the analysis of variance of the variable parameters, a 2×2 factorial plot with 3 replications was used. A regression model was defined based on experimental tests.

Having done the experimental tests, data were analyzed and the effect of independent parameters on the head and combine grain losses were investigated. The effect of the bar height on the head grain losses was significant at 1% level and the effect of the head height and interaction between bar height and head height on the head grain losses was also significant at 5% level. Results showed that with increasing in bar height, the head grain losses increased. With a change in the bar height, the location of the cutting point is changed and this led to a change in the head grain losses. The effect of the bar height on the combine grain losses was significant at 5% level but the effect of the head height and interaction between bar height and head height was not significant on the combine grain losses. Increasing in the bar height led to increase in material other grain (MOG) which enters to the combine, and also resulted in increasing in combine grain losses. The coefficient of determination of head grain losses in the regression model was 0.97. The model was able to explain the relationship between the bar and head height with head grain losses due to the relationship between independent and dependent parameters. The amount of R-squared for the combine grain losses in the regression model was 0.53. Because of the effect of other parameters in the inner section of the combine, the output of the model predicted that increasing in the bar height and head height, resulted in increasing in head grain losses, and also increasing in the bar height and decreasing in head height let to increasing in combine grain losses. The output of model showed that regulating the bar height and cutting height could reduce the harvest losses by less than 3%. This R-squared is obviously less than R-squared of head grain losses model. The output of the regression model predicted that the increase in the bar height and head height was associated with increase in the head grain losses, and increasing in the bar height and decreasing in head height, resulted in increasing in combine grain losses. The output of the regression model showed that the harvest losses can be reduced less than 5% by regulating the bar height and cutting height.

One of the most important parameters for mechanized harvesting is the head mechanism which cuts the crops and transfers them to the threshing unit. The cutting height in the sunflower head was defined by the bar height and head height. According to the linear relationship between the head and combine losses with the bar height and head height, and the interaction between them, the regression model was able to predict the result successfully. This model of grain losses in the head and combine model can be used in the intelligent combine to minimize the harvest losses. The optimization of the bar height and head height for minimizing the harvest losses can be the subject of next researches.