Effect of the cleaning process on physical properties for different malting barley seed varieties

Abstract The aim of the study was to investigate the influence of the barley cleaning process in relation to physical properties. The knowledge of the range of changes in the physical parameters of processed material and their mutual relationships is required for the design and implementation of various technological processes. In this study were compared the input and output commodities in the primary postharvest cleaning process of undesirable components—occurring as admixtures of fine and coarse barley impurities as well as the barley component itself. An efficient cleaning process ensuring barley grain quality is a basic step in beer production. Therefore, seven bred varieties of brewing barley (Malz, Sebastian, Francin, KWS Irina1, KWS Irina2, Bojos, and Laudis) were tested for the qualitative assessment of the cleaning process. Physical parameters such as granulometry, bulk and tapped density, angle of repose, internal and wall friction angle, and flow functions were determined for all samples. In order to identify whether the barley variety or the sample cleaning significantly influences the determined physical properties, two‐way ANOVA was applied. The results imply that barley cleaning had the main influence on wall friction angle, while the barley variety had a significant effect on effective internal friction. Moreover, the mechanical postharvest cleaning process reduces the overall wall friction.

The pursuit of an objective and comprehensive opinion on the malting quality of barley varieties has led to the development of different evaluation systems and procedures. At present, the malting quality of barley varieties is evaluated according to the "malting quality indicator," which was created based on the requirements of the processing industry (Rhodes, 2008). The key indicators of malting barley quality are mainly germination and germination energy (Daneri-Castro, Svensson, & Roberts, 2016). Germination is given as the percentage of germinated seeds over a specified period (6 days) under laboratory conditions. It is the quantity expressed as a percentage of all live cereals in the sample. The germination energy is expressed as the percentage of caryopses in the sample that germinate under natural conditions for 3 days at the time of determination. Causes of reduced germination may be due to disturbances in seed development, defects in fertilization, uneven maturation, weather conditions, harvest damage, poor postharvest treatment, and inappropriate storage (Rhodes, 2008). A very important feature is the content of nitrogenous substances-the protein and starch content (Balet, Gous, Fox, Lloyd, & Manley, 2020).
Other important quality criteria are the physical characteristics, such as the proportion of grain above the 2.5 mm sieve, at a basic value of 80%, and the basic value of 20% at a size of 2.8 mm.
In the cleaning process, dust, light, and metal particles are removed. The removed particles could be subject to further research (i.e., recycling them by torrefaction and ecological disposal). The sorting of barley grains by size is of technological importance for achieving uniform steeping and germination and obtaining a perfectly homogeneous malt (Chladek, 2007). It is evident that its detailed analysis will help us understand the basic principles and to find critical areas with the possibility of technical innovation. After the cleaning process, the barley seeds are stored in silos.
Therefore, the aim of the study was to determine the physical properties of various barley varieties in their raw state and in the postcleaning state (uncleaned and cleaned samples, respectively).
This was mainly the bulk angle, bulk density, tapped density, internal and wall friction angle, and the resulting flow properties and granulometry, as well as to compare the possible effect of the cleaning process on these barley grains and their interconnection with physical factors (influences).
Information on the flow characteristics of barley grains has to be verified from the perspective of determining the capacity of the storage space, transport, emptying or filling silos and for the correct design of the cleaning facility (Aghajani, Ansaripour, & Kashaninejad, 2012). The magnitude of internal and wall friction angles is the basis for designing the correct construction of containers and other storage equipment (Vilche, Gely, & Santalla, 2003).
Knowledge of the particle size distribution of the input barley is one of the primary properties for designing machinery for cleaning, sorting, separation, and storage (Kachru, Gupta, & Alam, 1994).
Another important and interesting feature, often neglected, is the morphology of barley grains, which allows the particle surface to be calculated, which is an important aspect, for example, in the design of drying technology (Sologubik, Campañone, Pagano, & Gely, 2013). In addition, the friction parameters of barley seeds are important in order to design the storage structures and the selection of equipment used in the whole technological line (Hamdani et al., 2014;Tavakoli, Tavakoli, Rajabipour, Ahmadi, & Gharib-Zahedi, 2009).
However, the information on the systematic determination of the basic physical properties of barley varieties currently available in literature has not been comprehensively processed. Therefore, the systematic characterization and comparison of individual varieties is a necessity, not only to achieve an optimum design of technological line parts but also to ensure sufficient product quality.
The aim of the study was to investigate the possible influence of barley cleaning processes in relation to physical properties, including statistical data analysis, which proves the interconnection of the tested parameters.
Samples of seven barley varieties were collected at the time of silo delivery, in both their raw, uncleaned state containing impurities of clay, stones, grass blades, damaged grains, metal particles, dust, and barley material samples after the cleaning process.
For the cleaning process, the KDC 4000 combi-cleaner (Kongskilde, Denmark) was used (Figure 1 right). It is a combined screen and aspiration cleaner with capacities of up to 40 t/h, commonly used for the cleaning of malting barley and seed. Motor outputs are 1.5 kW for the screen drum, 4 kW for the fan, and 0.75 kW for the auger. The material enters the cleaning device in the upper part of the machine (black arrow). In the first cylinder, the material mixtures are separated from coarse/large impurities. In order to obtain optimal efficiency, a cylinder screen with the correct slot/hole dimensions has to be used properly.
The inner cylinder has in its wider end round meshes with an 11 mm diameter-and the same mesh diameter is up to 2/3 of the cylinder length; in the last 1/3 of the cylinder length, the round mesh diameter is 9 mm. This is according to a standard recommendation, when at the beginning, the process capacity is increased, and afterward, optimal separation of coarser impurities is achieved, and simultaneously, the mesh capacity is decreased (finer meshes in the last 1/3).
In the outer cylinder, the remaining material particles fall through, with barley being separated from fine impurities. Thus, it is the slot/hole dimensions of the outer screen that determine the size of the impurities arrested. The size of oblong holes is 2.5 × 16.5 mm.
This size is appropriate for the very few good grains removed.
Subsequently, the fine impurities fall through (green arrow) and the finest particles are removed (brown arrow).

| Particle size distribution and particle shape
For measuring the size and shape of grains of free-flowing bulk materials ranging from 30 µm to 30 mm, a CAMSIZER optoelectronic device was used, consisting of a planar light source, feeder for the measured material, two CCD cameras for generating images of the loose material, a cleaning unit, and a PC with an evaluation program. Particles of bulk material fall over the end edge of the vibrating trough, with rotation occurring during their fall.
The rotating particles pass through the instrument through the measuring space-images are then generated using one or two CCD cameras, which are then analyzed. Granulometric analysis of the test material is generated automatically from the acquired images (Schwedes, 2003). After completion of the measurement, the results can be displayed in many formats, which are directly available for use. The measurements were made in a single cycle.

| Bulk and tapped densities
A sample of barley of the determined weight was poured into the measuring cylinder, and the value of the apparent volume was subtracted. The bulk density was established as the share of weight and apparent volume (Mellmann, Hoffmann, & Fürll, 2014). The result is the average value of ten measurements. Subsequently, the sample was subjected to tapping vibrations (amplitude 2 mm, 250 strokes per minute) at rates of 10x, 500x, and 750x. The apparent volumes were subtracted for each given tapping number and settling of the sample. The tapped density was determined as the ratio of the sample weight to the sample volume after the last tapping at 750x if no change above 2 ml occurred. Otherwise, 1,250 taps were performed. The result is an average of 10 measurements.

| Angle of repose
The barley sample of the specified weight was poured into a hollow cylinder, from where it was evenly poured out onto a horizontal surface into a cone shape (ASTM International, 2000). Upon completion of the pouring, the cone was divided into 8 parts, of which the angle was subtracted from each location using an enclosed calibrated spirit level to one decimal place. The bulk angle was determined as the average of the 8 parts. The measurement was repeated 5 times. (1) F I G U R E 1 Barley seeds' images and scheme of their cleaning. Left-sample of the Laudis variety-from uncleaned to cleaned. Rightdiagram of the cleaning principle

| Angle of internal friction and wall friction
The principle of measuring the shear properties of bulk materials consists in weighting the bulk material with a normal load of a preset size and then shifting layers of the bulk material in a direction perpendicular to the normal load. A shear stress is generated between the particles due to friction, which is detected. Specifically, the force needed to overcome the shear forces between the particles is measured by the action of the normal force (Zegzulka, 2013). These forces are further converted to stress (Scieszka & Adamecki, 2013). The measurement of the internal and wall friction angles was performed on a Schulze (Schulze Ring Shear Tester RST-01, Wolfenbuttel, Germany) rotary shear machine. The wall friction angle is a measure of the sliding friction at the grains/wall interface. Stainless steel was used as a contact material.
This material represents the wall surface of the hopper, silos, conveyor, etc. It is determined by shear cell testing where the grains under the load slide across a pad representing the equipment's wall surface, and it is expressed in degrees.
The measurement was performed in three cycles. The result is an average value.

| Flow properties of barley grains (flowability)
The Hausner ratio ( Carr's Index (CI) can be obtained using the Equation (6).
The smaller the value of HR and CI, the better the flow properties of the material. Inclusion of the tested material in the flow modes according to individual evaluation criteria is shown in   (Harrell, 2015). The significance level was p = .05 for all the analyses.

| Particle characterization
The particle size distribution values for the measured barley samples in the raw state and after cleaning are summarized in Figure 3. This distribution was made with the aim of data orientation.
And greater dispersion of the measured parameters (fractiond 10 , d 50, and d 90 ) is evident from the graph for uncleaned samples.
The most significant reduction is logical for the larger grains. When determining the distribution of the sizes of the grains of the uncleaned sample, the impurities that were removed by the cleaning process were also determined. A relatively larger shift to lower val- (4) ρ tp = m V750 ρ tp F I G U R E 2 Schematic output of yield locus obtained from the Schulze Ring Shear Tester RST-01 samples after cleaning ( Table 2). The greater dispersion in the case of uncleaned material is likely to be due to the variety of impurities mixed in. Increased value after cleaning suggests approaching a symmetrical grain shape. The edges of the barley grains are gently abraded by cleaning. Thus, the value of sphericity increases.
The width-to-length ratios of b/l 3 are found to be between 0.492 and 0.528 for uncleaned specimens and between 0.509 and 0.533 for specimens after cleaning, which is 6.9% greater than after the specimen cleaning.
It is evident that the values characterizing the grain shape, in particular SPHT 3 (sphericity rate) and the width-to-length ratio b/l₃ for the samples after cleaning, are greater than those measured before cleaning. This may be due to the cleaning of a greater number of nonhomogeneous or otherwise inadequate grains (fragments, fractions) and impurities and the remaining quality grains of approximately the same width-to-length ratio. In this respect, the cleaning process improved homogeneity.

| Effect of the cleaning process on seed flowability
The dependence of the bulk density, Carr's index (CI), and the angle of repose (AOR) on the cleaning process in different barley varieties is shown in Figures 4 and 5 It can be said that the cleaning of the tested materials does not in any significant way influence their flow properties ( Figure 4, Table 1).
In all cases except KWS Irina1, the flow character was defined as excellent according to their CI (Table 3). The angle of repose (AOR) values show a change in the flow after cleaning for all the samples and in the framework of an excellent character ( Figure 5, Table 3). Despite the fact that the changes of the angle of repose values for cleaned and uncleaned samples achieved up to 35%, according to the flowability AOR assessment (Table 1,   Table 3), all samples belong to the group "Excellent."

TA B L E 1 Flow character of material according to individual evaluation criteria
The values of the Hausner ratio (HR) and tapped density, together with the overall assessment of the cleaning process of seed flowability, are listed in Table 3.
The Hausner ratio (HR) ranged from 1.0 to 1.11 for all the tested samples (before and after cleaning), except for the KWS Irina1 sample, which changed from the character of excellent to adequate after cleaning. Increased values of tapped density were observed TA B L E 2 SPHT 3 symmetry (sphericity) and b/l 3 (width/length ratio) values for barley samples in almost all studied varieties-after the cleaning process. With the constant weight, the particle density decreases-which is logical (debris fall of, abrasion).
As it is obvious from Table 3, that flowability remains "excellent" for all the samples after the cleaning process, except the variety    Lower values of the internal friction angle for uncleaned samples of some varieties could imply that the uncleaned material contains fine dust or other fine impurities that contribute to reducing friction.

| Effect of the cleaning process on barley seed friction parameters
The wall friction angle is greater for samples before the cleaning process. From the point of view of flow, the same conclusion was reached as measured by densities, that is, the degree of purity of the barley received in the tested cases does not significantly affect their flow. In this sense, the same conclusion was reached as in the HR and CI experiments.
The results of two-way ANOVA are displayed in In the case of WFA, cleaning was found to have the main effect, although the effect of the barley variety was significant as well.
Moreover, a significant effect of the interaction of the variety and cleaning was observed.

| Correlation analysis of physical parameters
The data of all the physical parameters were subjected to correlation analysis. The results-in the form of a correlation matrix-were then plotted ( Figure 7).
As expected, a strong positive correlation was observed be-  tion b/l 3 and ρ tp , it is clear that particles with a higher proportion of b/l 3 , that is, with a more spherical particle shape, show a lower volume at vibrations arising, for example, during transport of the sample along the production line. One example may be the sample of Francin-cleaned b/l 3 = 0.533 and ρ tp = 778 kg/m 3 .

| Linear regression of physical parameters
A linear regression analysis was performed for the significantly correlated pairs, with the outcomes presented in Table 5. A significant linear relationship was observed for all the tested pairs; nevertheless, the coefficient of determination was not strong.

| CON CLUS ION
In the present study, the influence of the barley cleaning process in relation to physical properties was investigated for seven barley varieties.
Results showed that the cleaning process slightly increased barley grain sphericity as well as width-to-length ratio. Thus, higher grain homogeneity was achieved. Aforementioned parameters could be used for the relatively easy and rapid characterization of barley grains in practice. The values of WFA were greater for precleaning samples than for postcleaning samples. The lower wall friction angle value after mechanical cleaning facilitates the movement (flow) of the cleaned grains over the contact material within the processing line. This is also confirmed by the lower energy loss during material transport after the cleaning process within the production line. It is also desirable to provide a suitable structural contact material with a varying wall friction angle of the particles. Techniques.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.

E TH I C A L S TATEM ENT
This study does not involve any human or animal testing.