Application of aluminide coatings modified with yttrium for protecting corn post-harvest processing equipments against erosion in food industries

Recent studies demonstrate that aluminide coatings are beneficial to wear and erosion resistance and also oxygenactive elements such as yttrium can remarkably improve this resistance. In this research, the micro-hardness of the aluminide coatings with and without yttrium on carbon steel AISI 1045 was investigated using a micro-mechanical probe in post-harvest processing equipments in food industries. Wear of the samples was measured using a pin-on-disk tribometer. The erosion loss of specimens against five varieties of corn was also evaluated using a slurry erosion test machine. Experimental data were analyzed statistically using a 5-factor completely randomized design to study the effect of corn varieties, moisture content at three levels [13, 15, and 17% (wet basis)], and rotation velocity of the slurry erosion machine at three levels (150, 300, and 450 rpm) on erosion resistance. The results showed that the aluminide coatings improved the wear and erosion resistance of substrate carbon steel AISI 1045; yttrium markedly improved the hardness of the aluminide coating and its wear and erosion resistance. The erosion loss of materials was significantly (P < 0.01) influenced by the variety of corn, moisture content and rotation velocity. Both aluminide coating without yttrium and aluminide coating with yttrium showed higher wear and erosion resistance than carbon steel AISI 1045 substrate. Additional key words: carbon steel AISI 1045, corrosion, Indurata, storage bins, transportation, wear.


Introduction
Corn kernels are handled mechanically in storage bins and post-harvest processing equipments in food industries by augers and conveyors, airflow, or when it is allowed to flow by gravity.When corn kernels are transported through these equipments, they may impact on the body of equipments and may cause serious erosion damage.When corn is transported through pipelines, previous researches showed that the erosion rate of the bend could be 50 times higher than that of straight pipe (Fan et al., 2001).For this reason improving bend protection against erosion is an urgent task.
Wear of a solid surface by particle erosion has been the subject of many studies.For a similar resistance of the target pipe material and a similar condition of the impacting particles, different approaches can be used to reduce the erosion damage of the pipes such as: a decrease in the momentum of the impacting particles, a decrease in the range of the impact incidence angle (Tabakoff, 1984;Humphrey, 1990;Burakowski and Wierzchon, 1999;Fan et al., 2001), fixed ribs on the wall of the inside bend (Song et al., 1996;Yao et al., 2000;Fan et al., 2001), and a finned pipe erosion protection method (Fan et al., 1992).Surface damage due to wear and erosion may result in changes of the surface conditions and/or dimension of a mechanical component.As a result, these damages increase the risk of surface breaking, which might cause a disastrous failure of an entire mechanical system.Coating is one of the effective approaches against surface failure.Steel is often used as the core material coated with different materials for applications in the mining, chemical, petrochemical and food processing industries, and in transport pipelines to protect equipment and machinery from wear, corrosion and erosion.Recent studies have shown that wear and corrosive erosion of the aluminide coating can be markedly reduced by adding oxygen-active elements.For instance, yttrium improved the resistance of the aluminide coating to wear, corrosive wear, corrosion, corrosive erosion and dry sand erosion (Zhang et al., 1999;Zhang and Li, 2000a,b;Ahmadi and Li, 2002).The objective of this research was to investigate the protection against erosion of corn post-harvest processing equipments in food industries by applying aluminide coatings modified with yttrium.

Material and methods
The substrate material used in this study was commercial carbon steel AISI 1045.The dimension of the specimens was 13,612 mm.All specimens were polished with 600 grit SiC sand paper and cleaned in acetone, and then coated with Al to form an aluminide coating using a pack-cementation process (Soliman and El-Azim, 1999;Xiao and Mei 1999).FeAl (50%Fe-50%Al) alloy powder was used as the aluminium source, NH 4 Cl powder as an activator, and Al 2 O 3 powder as a filler material.About 3% yttrium powder (-40 mesh) was mixed with the FeAl powder (Ahmadi and Li, 2003).
Nitrogen gas was used to protect the specimens from oxidation.The coated specimens were mechanically polished using SiC sand papers and, finally, polished using 0.05-micron alumina.The specimens were cleaned in acetone.Micro-hardness of the specimens was determined using a micromechanical probe (Ahmadi and Li, 2003).
In order to evaluate wear resistance of the aluminide coated steel, specimens were also tested using a pinon-disk tribometer.Volume loss of samples after 1,000 laps was measured.The sliding speed and applied normal load were 10 mm s -1 and 3 N, respectively.The diameter of the silicon carbide ball pin was 6 mm.The width of the wear track (W) on the target specimen and diameter of the worn area of the ball pin (w) were measured using an optical-microscope.The crosssection area of a wear track on the target material was the trapezoidal shape area as shown in Figure 1.The volume loss of the material per unit length of the wear track is thus the trapezoidal shape area multiplied by a unit length.
Different varieties of corn were selected for the trials, from hard and soft quality classes.The moisture content of corn kernels (MC wb , % w.b.) was measured using the following formula: where M b = mass of kernels before drying; and M a = mass of kernels after drying.The sphericity percentage (S, %) of corn kernels was evaluated using the following formula: [2] where a = length of kernel, mm; b = width of kernel, mm and c = height of kernel, mm.The erosion loss of specimens against f ive varieties of corn (Amyalcea, Everta, Indentata, Indurata, Sacchatata) was evaluated using a slurry erosion test machine.The volume loss was estimated using the following formula: [3] where V = erosion rate, cm 3 m -2 hr -1 ; ∆W = weight loss, g; ρ = specimen density, gcm -3 ; A = eroded area, m 2 and t= testing time, hr.
The erosion loss of Y-containing coating specimens against different varieties of corn was evaluated using a slurry erosion test machine.The rotation velocities of slurry erosion test machine were 150, 300, and 450 rpm.The moisture contents were 13, 15, and 17% (wet basis).
Trials were carried out according to a 5-factor completely randomized design to study the effect on the erosion resistance of the five varieties of corn, moisture contents at three levels [13, 15, and 17 % (wet basis)], and the rotation velocity of the slurry erosion machine at three levels (150, 300, and 450 rpm).Five measurements were done for each test and results were analyzed and compared using the analysis of variance (F-test) and multiple linear regression (MLR).

Diagram of water absorption
To prepare the sample at different moisture levels (13, 15, and 17% wb), corn kernels were soaked in water for 5 to 120 min and then their moisture contents were determined by standard air oven method at 103°C for 24 hr.In order to determine the time needed to achieve these moisture contents for each corn variety, a previous experiment was conducted.This action was done four times for each variety and average of them was plotted in Figure 2 as corn-absorption diagram for five varieties of corn.Table 1 gives some physical properties of five varieties of corn that were used for the tests.

Micro-hardness properties
Figure 3 shows the hardness values of the Y-containing and Y-free aluminide coatings, at different loads from 25 to 750 mN.The results demonstrated that the hardness of the Y-free and Y-containing aluminide coatings was about 3 and 5 times higher than that of carbon steel AISI 1045, respectively.These results demonstrated that Y significantly increased the hardness of the aluminide coating.

Wear behaviour
The mean volume loss of f ive measurements of specimens showed that there were significant differences among carbon steel AISI 1045 (695 µm 3 µm -1 ) and aluminide coating without and with yttrium after 1,000 laps (307 and 96 µm 3 µm -1 , respectively) (P < 0.05).Both yttrium-free and yttrium-containing aluminide coating showed higher wear resistance than that of uncoated material.The wear resistances of aluminide coatings with yttrium and without yttrium were approximately 7 and 2.5 times as high as that of the substrate steel, respectively.

Effect of corn variety, moisture content, and machine velocity on volume loss of materials
The analysis of variance test showed that, in the slurry erosion test machine, erosion loss of carbon steel AISI 1045 (11.55 mm 3 m -2 ) (Fig. 4) was significantly different from other materials (P < 0.05), but there was no signif icant difference between erosion loss of Y-containing coating and Y-free coating (3.71 and 4.52 mm 3 m -2 respectively) (P > 0.05).According to Figure 4, for the first measurement of Y-free coating and Y-containing coating the volume loss was not significantly different (P = 0.01).It was observed that when these two samples held high moisture content in the corn kernels, corrosive erosion occurred during the test.Also, the corrosive erosion resistance of Y-containing coating was higher than that of Y-free coating.Figure 4 shows that the slope of volume loss of Y-containing coating between the third and fourth measurements was lower than that of between previous steps, it can be showed that this coating had a good erosion resistance.Both aluminide coatings showed higher erosion resistance than that of uncoated material.The results showed that yttrium significantly improved the erosion resistance of the aluminide coatings.The erosion resistances of aluminide coatings with and without yttrium were approximately 4.5 and 3 times as high as that of the steel substrate.Results of MLR analysis showed that effect of time on the sample volume loss for carbon steel AISI 1045, Y-containing and Y-free coatings was significant (P < 0.01).According to Table 2,    where VL is the mean of sample volume loss in mm 3 mm -2 , t is time in hr, Z steel is an indicator variable that takes the value 1 for the data corresponding to carbon steel AISI 1045, and zero otherwise.Similarly, Z Y-free is an indicator variable that takes the value 1 for data corresponding to Y-free coatings and zero otherwise.The reference level of this equation is Y-containing coating.Data given represent the mean of sample volume loss of Y-containing coating against different corn varieties, moisture content and machine velocity.
Figure 5 shows that erosion loss against varieties Amyalcea, Indurata, Indentata was higher than that against varieties Everta and Sacchatata (P < 0.05).This can be the result of a lower sphericity percent of Amyalcea, Indurata, Indentata kernels than that of Everta, Sacchatata kernels.According to the results, there was no significant difference between average volume loss of Y-containing coating specimens against Indurata, Amyalcea, and Indentata (13.2, 13.42, and 13 mm 3 m -2 respectively) after 48 hours (P > 0.05).Similar result was found for Everta and Sacchatata (11.47 and 11.37 mm 3 m -2 respectively).
Table 3 shows the results of MLR analysis for sample volume loss, time, and type of corn variety.Results showed that effect of corn variety type (other than In-dentata and Amylacea) on the volume loss of Y-containing coating specimens was signif icant (P < 0.01).Relationship among these parameters was derived as below: , R 2 = 0.95 where VL is the mean of sample volume loss in mm 3 mm -2 , t is time in hr, and Z Indentata , Z Sacchatata , Z Everta , and Z Amylacea are indicator variables.The reference level of this equation is Indurata.
MLR was done for sample volume loss as dependent variable and time and kernel sphericity as independent variables.Results showed that effect of these two variables on the sample volume loss was significant (P < 0.01).According to Table 4, relationship among these variables is linear as: , R 2 = 0.73 where S is spericity in %.
The results demonstrated that, by increasing the moisture content, the erosion loss increased (P < 0.05) (Fig. 6) but for 15% and 17% of moisture content there was not significant differences between erosion loss in these moisture contents (11.28 mm 3 m -2 for 13%, 12.44 mm 3 m -2 for 15%, and 12.87 mm 3 m -2 for 17%) (P > 0.05).It can be shown that, during the test, when the corn kernels had high moisture content, the speci-   Figure 7 shows that by increasing the rotation velocity, the erosion loss increased (9.83, 13.06, and 14.74 mm 3 m -2 for 150, 300, and 450 rpm, respectively after 48 h) (P < 0.01).It can be shown that when the rotation velocity of the slurry erosion test machine increased, so did the impact force of corn kernels on specimens and so, the erosion loss increased.Effect of time and machine rotation velocity on the sample volume loss, according to the results of MRL analysis, was significant (P < 0.01) (see Table 6) and relationship among these parameters was obtained as below: , R 2 = 0.99 where Z 150 RPM and Z 300 RPM are indicator variables.The reference level of this equation is 450 rpm.

Conclusions
This research was conducted to investigate the beneficial effect of aluminide coatings on the resistance of carbon steel AISI 1045 against erosion.Results of the present research demonstrated that: 1. Yttrium remarkably improved the wear and erosion resistance of the aluminide coating.
2. Both aluminide coatings had higher hardness than that of substrate material.
3. The erosion loss of materials was significantly (P < 0.05) influenced by moisture content (at 15 % and 17% of moisture content) but there wasn't significantly different between erosion loss in these moisture contents (P > 0.05).
4. Interaction effect of the time and kernel moisture content on the sample volume loss was significant (P < 0.01).
5. The erosion loss of materials was significantly (P < 0.01) influenced by rotation velocity.
6. Interaction effect of the time and machine rotation velocity on the sample volume loss was significant (P < 0.01).
7. Erosion loss corresponding to varieties Amyalcea, Indurata, and Indentata was higher than that of varieties Everta and Sacchatata.This can be the result of a lower sphericity percent of Amyalcea, Indurata, Indentata kernels than that of Everta, Sacchatata kernels.
8. Interaction effect of the time and kernel sphericity on the sample volume loss was significant (P < 0.01).9. Erosion loss of different materials was significantly (P < 0.01) different.

Figure 1 .
Figure1.Schematic view of the wear track cross-section area (trapezoidal shape are).W: width of the wear track on a target specimen.w: diameter of the worn area of the ball pin. w

Figure 3 .
Figure 3. Hardness value of carbon steel AISI 1045, Y-free coating and Y-containing coating (all values were obtained by averaging at least three measurements).

Figure 4 .
Figure 4. Volume loss of the aluminide coatings with and without yttrium, and carbon steel AISI 1045 (slurry erosion test machine).1045 steel Y-free coating Y-containing coating

Table 2 .
Results of multiple linear regression for sample volume loss, time and type of material SE: standard error.
Mean volume loss of Y-containing coating specimens against different corn varieties.

Table 3 .
Results of multiple linear regression for sample volume loss, time and type of corn variety SE: standard error.

Table 4 .
Results of multiple linear regression for sample volume loss, time and kernel sphericity

Table 5 .
Results of multiple linear regression for sample volume loss, time and kernel moisture content Figure 7. Mean volume loss of specimens for different rotation velocities.

Table 6 .
Results of multiple linear regression for sample volume loss, time and machine rotation velocity