Some physical properties of apricots and testing apricot sorting machines

Examinations were carried out in the manipulating and packaging plant of Gyumolcsert Ltd, in Boldogkovaralja, to determine some physical properties of five apricot cultivars and to test the work quality of the apricot sorting machines. The size and the weight of the fruits were measured and two sorting machines were tested. The results are given in tables and diagrams. The conclusions are also summarised.


Introduction
The main objectives of this research work were to determine some physical properties of apricots and the effect of the physical characteristics of the apricot cultivars on the quality of grading.The classification accuracy of the sorting machines with different apricot cultivars and the classification performance of the machines with different apricot cultivars were also examined based on our previous works (Ancza et al. 2011;Polyák and Csizmazia, 2003;Polyák et al. 2010 and2011).

Material and methods
The physical characteristics of the following five domestic and foreign apricot cultivars were determined: 1. Hungarian apricot 2. Jumbo Cot 3. Bergerouge 4. Bergeron 5. Late Jumbo From each cultivar 50-50 pieces of fruit (3-4 typical fractions were analyzed) were chosen and then measured.
The following data were recorded: 1. the main sizes of the fruits in three orthogonal dimensions with a digital slide gauge, with an accuracy of 0.01 mm; 2. the weight of each fruit with an accuracy of 0.01 g; 3. the weight ratio of the total, the sorted and the graded fruits; 4. the operating characteristics of the machines.

The sorting machines
There are two Compac type (New Zealand) sorting machines in the plant: a one-line, with a weight capacity of 0,8-1,2 t/h and a three-line, with a weight capacity of 2-3 t/h.Before the procedure the fruit is chilled to 8 o C, and then it is loaded by hand, tilting the container gradually (Figure 1).There is a selector in front of the machine to select the undersized fruit and the contamination.During the manual selection the overripe, the green, the deteriorated and the damaged fruits are selected.The roller sorting table (Figure 2) makes possible the careful selection of the fruit by turning it around.
From the sorting machine the fruit gets to a multi-line roller feeder through rotating brush rollers (Figure 3), which organizes the fruits to a classifying belt (Figure 4).
The speed of the multi-line roller feeder and the classifying belt is adjustable and is consistent with the chain speed moving the roller cars.The sorting machine is started at a small start-up performance, to check the correct operation of the machine, and then the performance will increase as long as it does not go to the expense of quality, or even what the packing staff can handle.
In order to exploit the sorting machine the row has to operate at an optimal capacity (Figure 5), but there cannot be double fruit at one measuring point.The movement of the roller carriers also ensures that, as the excess fruit drop-down at the side of the transport line and a belt delivers it back to the starting point of the line for repeated measurements.The roller carriers pass over an electronic grading scale.Two load cells per lane then gather weight information from each weigh point and process approximately 250 readings in less than 1/10th of a second for each fruit.Unique mathematical algorithms are an important but hidden part of the electronics that provide high precision.The machines are controlled by a computer.The speed, the weight categories and the output point of each category are adjustable.
The output can be at one or both sides of the machine and the packing can take place from the round tables or from the conveyor belt (Fig. 6 and 7) manually.The kerning is carried out using a balance.The customer demand determines how the fruit is presented: -5 kg bulk packaging in a timber bin (export), by 5 mm size-category; -M10-type crates of 10 kg (domestic), by 5 mm sizecategory; -In lines in a paper box; -In lines in plastic crates (Figure 8); -1 kg in cardboard boxes; -1 kg in plastic box (Figure 9); -10×1 kg in carton box; -10×1 kg collapsible plastic crates; -2.3 kg in a carton box (Figure 10).
A Sorma-type net bagger is available for packaging the 1 kg units (Figure 11).A template is used to adjust the weight categories (Figure 12).The weight categories are determined by the size defined by the template and a suitable correlation.The following diameter categories are used when categorizing the fruit:

Some physical properties of apricots and testing apricot sorting machines
-

Results
The results of our examinations are summarized in tables and shown in figures by cultivars.

Hungarian apricot
This cultivar was sorted into three fractions A, AA and AAA.The results of the measurements are given in Table 1.
Since the determination of the weight categories is based on the width values of the fruit, taking into account an appropriate formula, our analysis was also performed according to the width values.
From the measured data (3 × 50), we present the change of the width in the function of the mass, thickness and height values (Figure 13).
There is a close, second-degree correlation (R 2 = 0.943) between the width and weight.The thickness and height change linearly with the increase of the width, almost the same way.With the width-to-weight relationship the limit weight values to the category can be calculated (Figure 14).
The standard deviation of the weight values (4.41 to 7.66) is acceptable, which shows a good shape fidelity.The standard deviation of the width values (1.78 to 2.94) is good, which indicates the good sorting accuracy of the machine.
The distribution of values for each width category can be also studied with graphs (Figure 14).The distribution of the width values of category A (40-45 mm) is shown in Figure 15.
The figure shows that the smallest apricot is 3 mm smaller than the lower limit for the category.80% of the fruits is within the size category.The standard deviation (2.94) is acceptable.
The smallest width value of AA category (45-50 mm) remained slightly below the lower limit for the category (Figure 16).The maximum value is slightly above the upper limit of class size.The average value is within the boundaries of the category which shows the good selection of category limits.
The standard deviation value (1.78) is favourable.92% of fruits are between the boundaries of the category.
The smallest width value of AAA category (50-55 mm) was 3 mm below the lower limit for the category (Figure 17).The maximum and the average value are within the boundaries of the category.The standard deviation value (1.79) is favourable, so the sorting accuracy of the machine appeared to be good.72% of fruits are between the boundaries of the category, so the selection of category limits was not appropriate.

Jumbo Cot
This cultivar was sorted into four fractions B, A, AA and AAA.The results of the measurements are given in Table 2.
From the measured data (4 × 50), we present the change of the width in the function of the weight, thickness and height values (Figure 18).
There is a close, second-degree correlation (R 2 = 0.9516) between the width and weight.The standard deviation of the weight values (3.78-5.88) is acceptable.
The width and height is changing linearly with the increase of the width, but in a different way.With the widthto-weight relationship the limit weight values to the category can be calculated (Figure 19).
The distribution of the width values were analyzed in detail and the distributions are presented in a chart.
The distribution of width values for category B (35-40 mm) is shown in Figure 20.Only one apricot was below the minimum size of the category.The largest apricot was 4 mm above the upper limit of the category.

Some physical properties of apricots and testing apricot sorting machines
The average value was within the boundaries of the category.94% of the fruits were within or close to the size category.The standard deviation (1.59) was favourable.
The smallest width value for category A (40-45 mm) remained within the category limits (Figure 21), however the largest value was 4 mm above the maximum value.
74% of the fruits were within the size categories, which shows that the selection of the category boundaries was not accurate enough.The standard deviation (1.94) was acceptable.
The smallest width value for category AA (45-50 mm) remained slightly below the minimum value, the largest value was a bit above the maximum value (Figure 22), and the average value remained between the boundaries of the category.
94% of the fruits were between the boundaries of the category, so the boundaries were chosen correctly.The standard deviation of this size category is favourable (1.43), which shows a good sorting accuracy.
The smallest width value for category AAA (50-55 mm) remained slightly below the minimum value (Figure 23), the largest value was slightly above the maximum value and the average was between the category boundaries.

57
78% of the fruits were between the boundaries of the category, so the boundaries were chosen correctly.The standard deviation of this size category is favourable (1.96).

Bergarouge
This cultivar was sorted into three categories, B, A and AA.The measuring results are summarized in Table 3.
From the measured data (3 × 50) we present the change of the width in the function of the weight, thickness and height values (Figure 24).
There is a close, second-degree correlation (R 2 = 0. 9114) between the width and weight.
The standard deviation of weight values (3.51-7.84) is acceptable.The thickness and height is changing linearly with the increase of the width.With the width-to-weight relationship the limit weight values to the category can be calculated (Figure 25).The distribution of the width values for each category is presented in the following diagram.
The smallest value for category B (35-40 mm) was below the minimum value.The largest value was above the maximum value of the next category.The average was also above the maximum of the category.24% of the fruits were between the category boundaries, which indicate a faulty machine setting.The standard deviation was 2.69.The distribution of the width values for category B is shown in Figure 26.
The smallest value for category A (40-45 mm) was below the minimum of the category (Figure 27).
The largest value was slightly above the category maximum.The average was within the category boundaries.82% of the fruits were between the category boundaries.The standard deviation (1.56) was favourable.
The smallest width value for category AA (45-50) was significantly below the lower category limit (Figure 28).The largest value was slightly below the category maximum, while the average was between the category boundaries.88%   of the fruits were between the category boundaries, which show that they were chosen correctly.

Some physical properties of apricots and testing apricot sorting machines
The standard deviation in category AA was acceptable (2.69).

Bergeron
This cultivar was sorted into three categories, B, AA and AAA.There was not available sample from size category A. The measuring results are summarized in Table 4.
From the measured data (3 × 50) we present the change of the width in the function of the weight, thickness and height values (Figure 29).
There is a close, second-degree correlation (R 2 = 0.9745) between the width and weight.
The mass variance is significant (6.56 to 10.64), which can be explained by shape errors.
The thickness and the height changed almost equally, linearly with the increase of the width.
With the width-to-weight relationship the limit weight values to the category can be calculated (Figure 30).
The smallest value in category B (35-40 mm) was slightly below the minimum of the category, while the largest value was above the maximum of the next size category.The medium value was within the category boundaries.Only 74% of the fruits was between the category boundaries, which shows sorting inaccuracy.The value of standard deviation (3.28) enhances that.
The smallest value in category B (35-40 mm) was slightly below the minimum of the category, the largest value was above the maximum of the next size category.The medium value was within the category boundaries.Only 74% of the fruits was between the category boundaries, which shows sorting inaccuracy.The value of standard deviation (3.28) enhances that.
The distribution of width values of category B is shown in Figure 31.
The smallest width value of category AA (45-50 mm) was significantly below the minimum value (Figure 32).The largest and the average values remained within the category boundaries.
84% of the fruits were between the category boundaries, which means that they were selected correctly.The standard deviation (2.48) is also acceptable.
The smallest width value of category AAA (50-55 mm) was below the minimum value (Figure 33).
Both, the largest and the average values were above the maximum value, and only 26% of the fruits were between the category boundaries, which indicates their incorrect selection.The standard deviation (2.78) in this size category was acceptable.

Late Jumbo
This cultivar was sorted into three fractions, B, A and AAA.There was not available sample from size category AA.The measuring results are summarized in Table 5.
From the measured data (3 × 50) we present the change of the width in the function of the weight, thickness and height values (Figure 34).There is a close, second-degree correlation (R 2 = 0.9793) between the width and weight.The mass variance is normal (4.33-7.08).
The thickness and the height changed linearly, but in a different way with the increase of the width.The width-toweight relationship is suitable to calculate the values of the weight size categories (Figure 35).
The smallest measured width was significantly below the minimum of the size category, while the largest value was close to the maximum of the next size category.The average was between the minimum and maximum values.The distribution of the width values of this size category is shown in Figure 36.80% of the fruits remained within the category boundaries or close to them.The standard deviation was acceptable, 2.42.

Some physical properties of apricots and testing apricot sorting machines
The smallest width value for category A was significantly below the minimum value of this category (Figure 37).
The larges value only slightly exceeded the upper limit of the category.The standard deviation was acceptable, 1.59.
96% of the fruits were between the category boundaries, which proves the good selection of the category limits.The standard deviation was favourable, 1.96.
The smallest width value in category AAA was also significantly below the minimum value of this category (Figure 38).The largest value was substantially higher than the upper category limit.The average value was slightly above the maximum.
Only 38% of the fruits was between the category boundaries, which shows sorting inaccuracy.The standard deviation of width values of AAA category was favourable, 1.96.

Fig. 9 :
Fig. 9: 1 kg fruit in a plastic box Fig. 10: 2.3 kg fruit in a carton box

Fig. 15 :Fig. 17 :
Fig. 15: Distribution of the width values of category A of Hungarian apricot

Fig. 18 .Fig. 22 :
Fig.18.The connection of the measured data with the width for Jumbo Cot apricot

Fig. 24 :
Fig. 24:The connection of the measured data with the width for Bergerouge apricot

Fig. 27 :Fig. 30 :
Fig. 27: Distribution of the width values of category A of Bergarouge

Fig. 31 :
Fig. 31: Distribution of width values of category B of Bergeron apricot

Fig. 34 :
Fig. 34:The connection of the measured data with the width for Late Jumbo apricot

Fig. 35 :
Fig. 35: Weight values calculated from the width-weight relationship

Fig. 37 :
Fig. 37: Distribution of width values of category A of Late Jumbo apricot

Table 1 .
Characteristics of Hungarian apricot

Table 2 .
Characteristics of Jumbo Cot

Table 4 .
Characteristics of Bergeron apricot cultivar

Table 5 .
Characteristics of Late Jumbo apricot cultivar

Table 6 .
Some operating characteristics connected to sorting