Preparation and Performance Optimization of Original Aluminum Ash Coating Based on Plasma Spraying

As an industrial solid waste, the original aluminum ash (OAA) will cause serious pollution to the air and soil. How to reuse the OAA has been a research difficulty. Thus, a method of preparing a plasma spray powder using OAA is proposed. The OAA was hydrolyzed and ball milled, and the flowability of original aluminum ash spray powder (OAASP) was evaluated by the angle of repose. The coating properties were determined via analyzing the microstructure and the phase of the coating, and the effects of plasma spray parameters on the coating properties were investigated by the orthogonal experiment to optimize spray parameters. The results show that the angle of repose of OAASP after granulation was less than 40◦, which met the requirements of plasma spraying. When the spraying current was 600 A, the spraying voltage was 60 V, the main gas flow was 33 slpm, and the powder flow rate was 22 g/min, and the prepared original aluminum ash coating (OAAC) had excellent comprehensive performance. After the spraying process parameters were optimized, the microhardness of the coating was 606.54 HV, which is about twice the hardness of the substrate; the abrasion rate was 12.86 × 10−3 g/min; the porosity was 0.16%; and the adhesive strength was 16 MPa. When the amount of Al2O3 added was 50%, the hardness of the coating was increased by 17.61%.


Introduction
Aluminum ash, also known as aluminum slag, is a by-product of aluminum electrolysis and aluminum smelting.According to statistics, 1 ton of aluminum will produce 180-290 kg of aluminum ash during the entire technological process [1][2][3][4].In line with the different components, the aluminum ash can be split into primary and secondary aluminum ash.The primary aluminum ash color is grayish white, which is produced in the primary aluminum electrolysis and casting process without adding salt flux [5].It is a mixture of aluminum and aluminum oxides with an aluminum content of 30%-70%.The primary material used in this study, the original aluminum ash (OAA), can be acquired by ball milling and screening of the primary aluminum ash [6,7].Furthermore, the ultimate aluminum ash (UAA) can be obtained after OAA is processed with the hot aluminum ash separator, but the OAA and UAA differ greatly in performance.
Due to technical limitations, there is currently no efficacious way to recycle large quantities of aluminum ash.The treatment of aluminum ash is still dominated by accumulation and landfill in most enterprises, which will generate environmental pollution.Aluminum nitride (AlN) in aluminum ash reacts with water vapor or rain in the air at a certain temperature to form NH 3 with a strong The main experimental raw material was OAA, bought from China Jiangsu Haiguang Metal Co., Ltd.(Suqian, China), which collected aluminum ash used in the experiment from multiple aluminum processing factories.Therefore, the source of aluminum ash is complicated.The chemical composition of OAA is shown in Table 1.According to X-ray fluorescence (XRF, MiniPal4, PANalytical Company, Almelo, The Netherlands) analysis, the key elements in OAA are Al, Fe, and Si, among which Al has the most content of 61.802%.Due to the fact that the OAA was not merely intricate in source, but also prone to introduce new impurities during production and transportation, the OAA contained various trace elements such as Zn, Mn, and Ti.The 45-steel without any heat treatment was employed as the substrate, which was sandblasted using Al 2 O 3 before spraying.The OAA was subjected to X-ray diffraction (XRD, D/Max 2500PC Rigaku, Japan Science and Technology Co., Ltd., Tokyo, Japan) semi-quantitative analysis.The phase composition was as shown in Table 2, and the XRD pattern is shown in Figure 1.The main phase of OAA was 21 wt.% ± 3 wt.%Al and 44 wt.% ± 3 wt.%AlN.In addition, OAA also comprised some fluoride salts and chloride salts, which were not detected by XRD because their content was generally less than 5%.A mass of AlN was contained in the OAA because N 2 was introduced during the refining of the aluminum alloy.Due to the presence of AlN, the aluminum ash was prone to hydrolysis at temperatures above 30 • C, resulting in the production of pungent ammonia gas, so that the aluminum ash could not be stored for a long time, and the OAA needed to be denitrified.temperatures above 30 °C, resulting in the production of pungent ammonia gas, so that the aluminum ash could not be stored for a long time, and the OAA needed to be denitrified.

Preparation of Original Aluminum Ash Spray Powder
Since aluminum ash comes from disparate factories and is apt to introduce impurities during transportation, OAA contained many large-particle impurities, such as broken glass, stone and plastic, and so a 50-mesh standard sieve was utilized for impurity removal.The hydrolysis experiment was conducted at room temperature; the screened OAA was placed in a beaker, ultrapure water was added at a solid:liquid ratio of 1:10.Then the beaker was placed in a constant-temperature water bath with a temperature of 90-100 °C, and a stirring device was applied to carry out hydrolysis treatment to remove salt and nitrogen in the aluminum ash.
The nitrogen removal process lasted for 3-6 h, and phenolphthalein test paper was placed in the beaker mouth.If the test paper did not change color, it indicated that the nitrogen removal was over.Then we stopped heating and let the beaker stand for 1-2 h.The aluminum ash liquid in the beaker stratified, as shown in Figure 2. The upper layer was scum which was hardly soluble in water; the middle layer was turbid liquid; the lower layer was a precipitate which was also hardly soluble in water; the middle layer gradually became transparent as the standing time extended.

Preparation of Original Aluminum Ash Spray Powder
Since aluminum ash comes from disparate factories and is apt to introduce impurities during transportation, OAA contained many large-particle impurities, such as broken glass, stone and plastic, and so a 50-mesh standard sieve was utilized for impurity removal.The hydrolysis experiment was conducted at room temperature; the screened OAA was placed in a beaker, ultrapure water was added at a solid:liquid ratio of 1:10.Then the beaker was placed in a constant-temperature water bath with a temperature of 90-100 • C, and a stirring device was applied to carry out hydrolysis treatment to remove salt and nitrogen in the aluminum ash.
The nitrogen removal process lasted for 3-6 h, and phenolphthalein test paper was placed in the beaker mouth.If the test paper did not change color, it indicated that the nitrogen removal was over.Then we stopped heating and let the beaker stand for 1-2 h.The aluminum ash liquid in the beaker stratified, as shown in Figure 2. The upper layer was scum which was hardly soluble in water; the middle layer was turbid liquid; the lower layer was a precipitate which was also hardly soluble in water; the middle layer gradually became transparent as the standing time extended.
The turbid liquid of the middle layer and the sediment of the lower layer were placed in an oven at 120 • C respectively, and dried for 6-12 h.The dried aluminum ash was taken out from the beaker and subjected to a ball milling treatment.The instrument used was a laboratory horizontal planetary ball mill, the DECO-PBM-H-0.4Ltype, produced by Deco Instruments (Changsha, China).The grinding ball was made of stainless steel and the radius of the ball was 5 mm.The grinding method was vacuum grinding, and the rotation speed of the ball grinding tank was 1100 rpm.After being ground for 5-10 min, the aluminum ash was taken out and sieved with a 100-mesh sample sieve to obtain the OAASP.Figure 3 is a particle size distribution diagram of OAASP, and the primary particle size distribution range is 120-150 mesh.The turbid liquid of the middle layer and the sediment of the lower layer were placed in an oven at 120 °C respectively, and dried for 6-12 h.The dried aluminum ash was taken out from the beaker and subjected to a ball milling treatment.The instrument used was a laboratory horizontal planetary ball mill, the DECO-PBM-H-0.4Ltype, produced by Deco Instruments (Changsha, China).The grinding ball was made of stainless steel and the radius of the ball was 5 mm.The grinding method was vacuum grinding, and the rotation speed of the ball grinding tank was 1100 rpm.After being ground for 5-10 min, the aluminum ash was taken out and sieved with a 100-mesh sample sieve to obtain the OAASP.Figure 3 is a particle size distribution diagram of OAASP, and the primary particle size distribution range is 120-150 mesh.

Preparation of Original Aluminum Ash Coating
The Al-Ni powder was chosen to prepare the bonding layer for its wide application and low prices.The Al-Ni alloy bonding layer was prepared on the surface of 45-steel substrates by plasma spraying equipment (FH-80, Fahan Spraying Machinery Co., Ltd., Shanghai, China), and its thickness was 0.08-0.18μm.The material was Al-Ni powder commonly used for the alumina ceramic coating.Through the preparation of the Al-Ni bonding layer, it assisted in improving the bonding strength of the coating [32].Table 3 is the bonding layer process parameters.The turbid liquid of the middle layer and the sediment of the lower layer were placed in an oven at 120 °C respectively, and dried for 6-12 h.The dried aluminum ash was taken out from the beaker and subjected to a ball milling treatment.The instrument used was a laboratory horizontal planetary ball mill, the DECO-PBM-H-0.4Ltype, produced by Deco Instruments (Changsha, China).The grinding ball was made of stainless steel and the radius of the ball was 5 mm.The grinding method was vacuum grinding, and the rotation speed of the ball grinding tank was 1100 rpm.After being ground for 5-10 min, the aluminum ash was taken out and sieved with a 100-mesh sample sieve to obtain the OAASP.Figure 3 is a particle size distribution diagram of OAASP, and the primary particle size distribution range is 120-150 mesh.

Preparation of Original Aluminum Ash Coating
The Al-Ni powder was chosen to prepare the bonding layer for its wide application and low prices.The Al-Ni alloy bonding layer was prepared on the surface of 45-steel substrates by plasma spraying equipment (FH-80, Fahan Spraying Machinery Co., Ltd., Shanghai, China), and its thickness was 0.08-0.18μm.The material was Al-Ni powder commonly used for the alumina ceramic coating.Through the preparation of the Al-Ni bonding layer, it assisted in improving the bonding strength of the coating [32].Table 3 is the bonding layer process parameters.

Preparation of Original Aluminum Ash Coating
The Al-Ni powder was chosen to prepare the bonding layer for its wide application and low prices.The Al-Ni alloy bonding layer was prepared on the surface of 45-steel substrates by plasma spraying equipment (FH-80, Fahan Spraying Machinery Co., Ltd., Shanghai, China), and its thickness was 0.08-0.18µm.The material was Al-Ni powder commonly used for the alumina ceramic coating.Through the preparation of the Al-Ni bonding layer, it assisted in improving the bonding strength of the coating [32].Table 3 is the bonding layer process parameters.

Main Gas Flow/slpm
Al/Ni 55 550 24 33 The process parameters of the plasma spraying equipment were adjusted, and OAASP was sprayed on the surface of the bonding layer.An orthogonal experiment with four factors and three levels was used in the spraying experiment, and there were nine groups of L9 (3 4 ) (O1-O9).In addition, the spray current, spray voltage, the powder flow rate, and the main gas flow were selected as the principal factors of the spraying process, and the gas used was argon.The designed orthogonal test table is shown in Table 4.

Experiment Equipment and Testing Methods
The coating was prepared by the plasma spraying complete set of equipment, namely the FH-80 type, produced by Fahan Spraying Machinery Co., Ltd.(Shanghai, China).The composition analysis was performed by X-ray fluorescence, which was the MiniPal4 type produced by PANalytical Company (Almelo, The Netherlands).Phase analysis was performed by X-ray diffraction analysis, the D/Max 2500PC Rigaku type, which was produced by Japan Science and Technology Co., Ltd.(Tokyo, Japan).
The microstructure of the coating was observed by scanning electron microscope (SEM), the S-3400 type, produced by Hitachi, Ltd. (Tokyo, Japan) The type and the content of elements in the coating micro-area were analyzed by energy-dispersive X-ray spectroscopy (EDS), the Quantax75 type, produced by Japan Hitachi, Ltd.The main parameters used to evaluate the performance of OAAC were porosity [33], adhesive strength, microhardness, and abrasion rate.
The porosity was measured by the simplified Archimedes drainage method and laboratory-made platform.The test piece of coating was dried and weighed to obtain a mass m 0 , then put in distilled water to get the mass of the discharged water m 1 .The test piece was removed from the water and weighed to obtain the mass m 2 .Finally, the porosity of the OAAC could be obtained according to the formula (m 2 -m 0 )/m 1 .The smaller the porosity, the denser the coating.
The abrasion rate was measured by a ring three-body wear tester, MMH-5 type, produced by Hansen Precision Instrument Co., Ltd.(Jinan, China).According to the international standard ISO7784.2-97[34], the weight loss method was adopted in the experiment.The sample after abrasion was ultrasonically cleaned, dried and weighed, and the abrasion rate was calculated according to the formula (m 1 -m 2 )/t, where, m 1 is the mass before grinding, m 2 is the mass after grinding, and t is the grinding time.The type of the electronic balance for weighing was FA-2004N, produced by Grand Instrument and Equipment Co., Ltd.(Shanghai, China), and its accuracy was 0.1 mg.
According to the Chinese national standard GB/T 8642-2002 [35], the adhesive strength was measured by the stretching method.The used equipment was a universal mechanical testing machine, CMT5105 type, produced by MTS Industrial Systems Co., Ltd.(Shenzhen, China).The adhesive strength was used to characterize the mechanical combine ability between the bonding layer and the OAAC.
Due to the thin thickness of the aluminum ash coating, the Vickers hardness was used to measure the hardness of the OAAC based on Chinese national standard GB4342-84 [36].The digital micro hardness tester type was TMV-1, produced by Time Group Inc. (Beijing, China), which is suitable for hardness measurement of materials such as surface coatings, integrated circuit (IC) sheets, and ceramics.The principle was to calculate the Vickers hardness by measuring the diagonal length of the indentation.Then 10 relatively smooth areas selected on the surface of the coating were tested, and the results were averaged.

Observation and Analysis of Microstructure
The microstructure of the OAA and the OAASP was observed by scanning electron microscope, and the results are shown in Figure 4. Figure 4a is the SEM image of OAA, which had extremely irregular particle size and morphology and contained sharp edges and corners.Figure 4b is the SEM image of OAASP obtained after the hydrodynamic ball milling treatment.The particles were comparatively small and uniform, and the shape was approximately spherical.Meanwhile, during the hydrolysis process, some salt particles were dissolved in water and filtered off, and thus the content of salt particles was greatly reduced, which helped to enhance the flowability of the spray powder and facilitate the preparation of the OAAC.

Observation and Analysis of Microstructure
The microstructure of the OAA and the OAASP was observed by scanning electron microscope, and the results are shown in Figure 4. Figure 4a is the SEM image of OAA, which had extremely irregular particle size and morphology and contained sharp edges and corners.Figure 4b is the SEM image of OAASP obtained after the hydrodynamic ball milling treatment.The particles were comparatively small and uniform, and the shape was approximately spherical.Meanwhile, during the hydrolysis process, some salt particles were dissolved in water and filtered off, and thus the content of salt particles was greatly reduced, which helped to enhance the flowability of the spray powder and facilitate the preparation of the OAAC.

Composition Analysis
The chemical composition of OAASP was analyzed by XRF, and the results are shown in Table 5.Compared with OAA, OAASP had the lesser changed chemical composition and the unvaried type of the element.Some salts were soluble in water and were removed after hydrolysis and filtration.Figure 5 shows the XRD pattern of OAASP.In comparison with OAA, OAASP had significantly increased Al and Al2O3 content, but also had decreased content of AlN.AlN and H2O could react at room temperature to form Al(OH)3 and NH3, and heating and stirring in a water bath accelerated the chemical reaction.After the NH3 overflowed, the filtered Al(OH)3 was heated in the oven, to be further decomposed into Al2O3 and H2O.Therefore, the nitrogen content in OAASP was reduced, while the Al2O3 content was increased.The results shown in Figure 5 signify that the method of removing nitrogen in aluminum ash by hydrolysis was effective.However, OAASP still contained a handful of AlN.The reason lies in that when AlN is hydrolyzed, Al(OH)3 is formed on the surface of AlN particles to prevent further reaction.

Composition Analysis
The chemical composition of OAASP was analyzed by XRF, and the results are shown in Table 5.Compared with OAA, OAASP had the lesser changed chemical composition and the unvaried type of the element.Some salts were soluble in water and were removed after hydrolysis and filtration.Figure 5 shows the XRD pattern of OAASP.In comparison with OAA, OAASP had significantly increased Al and Al 2 O 3 content, but also had decreased content of AlN.AlN and H 2 O could react at room temperature to form Al(OH) 3 and NH 3 , and heating and stirring in a water bath accelerated the chemical reaction.After the NH 3 overflowed, the filtered Al(OH) 3 was heated in the oven, to be further decomposed into Al 2 O 3 and H 2 O. Therefore, the nitrogen content in OAASP was reduced, while the Al 2 O 3 content was increased.The results shown in Figure 5 signify that the method of removing nitrogen in aluminum ash by hydrolysis was effective.However, OAASP still contained a handful of AlN.The reason lies in that when AlN is hydrolyzed, Al(OH) 3 is formed on the surface of AlN particles to prevent further reaction.

Evaluation of Flowability
The angle of repose is the minimum angle between the slope and the horizontal surface; when the object on the slope is in a critical state of sliding along the slope.It is a common way to evaluate the flowability of the powder.The smaller the angle of repose, the smaller the friction and the better the flowability.It is generally considered that the θ ≤ 40° can meet the needs of production flowability performance [37].In this experiment, the angle of repose of the different samples is shown in Figure 6.The OAA was irregular particles with an angle of repose of 41.17° and poor flowability performance.After granulation, the OAASP had a smaller angle of repose of 35.70°, which belongs to regular particulates and has ideal flowability.After hydrolysis and ball milling, the angle of repose of OAASP was reduced by 13.28%.It can be seen that the granulation process of hydrolysis and ball milling was conducive to improving the flowability performance of aluminum ash. 41.17

Evaluation of Flowability
The angle of repose is the minimum angle between the slope and the horizontal surface; when the object on the slope is in a critical state of sliding along the slope.It is a common way to evaluate the flowability of the powder.The smaller the angle of repose, the smaller the friction and the better the flowability.It is generally considered that the θ ≤ 40 • can meet the needs of production flowability performance [37].In this experiment, the angle of repose of the different samples is shown in Figure 6.The OAA was irregular particles with an angle of repose of 41.17 • and poor flowability performance.After granulation, the OAASP had a smaller angle of repose of 35.70 • , which belongs to regular particulates and has ideal flowability.After hydrolysis and ball milling, the angle of repose of OAASP was reduced by 13.28%.It can be seen that the granulation process of hydrolysis and ball milling was conducive to improving the flowability performance of aluminum ash.

Evaluation of Flowability
The angle of repose is the minimum angle between the slope and the horizontal surface; when the object on the slope is in a critical state of sliding along the slope.It is a common way to evaluate the flowability of the powder.The smaller the angle of repose, the smaller the friction and the better the flowability.It is generally considered that the θ ≤ 40° can meet the needs of production flowability performance [37].In this experiment, the angle of repose of the different samples is shown in Figure 6.The OAA was irregular particles with an angle of repose of 41.17° and poor flowability performance.After granulation, the OAASP had a smaller angle of repose of 35.70°, which belongs to regular particulates and has ideal flowability.After hydrolysis and ball milling, the angle of repose of OAASP was reduced by 13.28%.It can be seen that the granulation process of hydrolysis and ball milling was conducive to improving the flowability performance of aluminum ash. 41.17

Microstructure of Coating
The microstructure of the OAAC was observed by SEM, and the results are shown in Figure 7. Figure 7a shows the surface morphology of the coating.The surface of the OAAC was relatively rough, because of manual spraying, and different deposition rates of the powders with various phases of OAA.The surface of the coating was dispersed with variety of areas, namely light regions, dark regions, and gray regions.As shown in Figure 7b, the cross section of the sample had obvious stratification.The upper layer was the coating portion formed by the OAASP through the atmospheric plasma spraying process; the middle layer was the bonding layer formed by the Al/Ni powder; the lower layer was a 45-steel substrate.Generally, the thickness of the coating is required to be about 500 µm in the industry [38].As the coating was hand-sprayed, the thickness of the OAAC was 450-550 µm, which basically met the requirements of industrial production.

Microstructure of Coating
The microstructure of the OAAC was observed by SEM, and the results are shown in Figure 7. Figure 7a shows the surface morphology of the coating.The surface of the OAAC was relatively rough, because of manual spraying, and different deposition rates of the powders with various phases of OAA.The surface of the coating was dispersed with variety of areas, namely light regions, dark regions, and gray regions.As shown in Figure 7b, the cross section of the sample had obvious stratification.The upper layer was the coating portion formed by the OAASP through the atmospheric plasma spraying process; the middle layer was the bonding layer formed by the Al/Ni powder; the lower layer was a 45-steel substrate.Generally, the thickness of the coating is required to be about 500 μm in the industry [38].As the coating was hand-sprayed, the thickness of the OAAC was 450-550 μm, which basically met the requirements of industrial production.

Energy Spectrum Analysis
Different color regions existed on the surface of the OAAC, and the SEM and the energy spectrum were used for analysis.The diverse color regions are shown in Figure 8, which are segmented into the gray region, the light region, and the dark region.

Energy Spectrum Analysis
Different color regions existed on the surface of the OAAC, and the SEM and the energy spectrum were used for analysis.The diverse color regions are shown in Figure 8, which are segmented into the gray region, the light region, and the dark region.The microstructure of the OAAC was observed by SEM, and the results are shown in Figure 7. Figure 7a shows the surface morphology of the coating.The surface of the OAAC was relatively rough, because of manual spraying, and different deposition rates of the powders with various phases of OAA.The surface of the coating was dispersed with variety of areas, namely light regions, dark regions, and gray regions.As shown in Figure 7b, the cross section of the sample had obvious stratification.The upper layer was the coating portion formed by the OAASP through the atmospheric plasma spraying process; the middle layer was the bonding layer formed by the Al/Ni powder; the lower layer was a 45-steel substrate.Generally, the thickness of the coating is required to be about 500 μm in the industry [38].As the coating was hand-sprayed, the thickness of the OAAC was 450-550 μm, which basically met the requirements of industrial production.

Energy Spectrum Analysis
Different color regions existed on the surface of the OAAC, and the SEM and the energy spectrum were used for analysis.The diverse color regions are shown in Figure 8, which are segmented into the gray region, the light region, and the dark region.The energy spectrum analysis was applied to different regions, and the EDS images are shown in Figure 9.In the light of the proportion of each element, it can be inferred that the main phase in the light region was Al 2 O 3 , and the main phases in the gray region and the dark region were Al 2 O 3 , SiO 2 , and Fe 3 O 4 .The light region had the highest hardness, about 500-900 HV, and the dark region had the lowest hardness, about 100-200 HV.Therefore, in order to improve the hardness of the coating, dark regions should be avoided during the spraying process.
Coatings 2019, 9, x FOR PEER REVIEW 9 of 15 The energy spectrum analysis was applied to different regions, and the EDS images are shown in Figure 9.In the light of the proportion of each element, it can be inferred that the main phase in the light region was Al2O3, and the main phases in the gray region and the dark region were Al2O3, SiO2, and Fe3O4.The light region had the highest hardness, about 500-900 HV, and the dark region had the lowest hardness, about 100-200 HV.Therefore, in order to improve the hardness of the coating, dark regions should be avoided during the spraying process.Figure 10 shows the XRD pattern of the OAAC.It is observed that the phase of the OAAC lessened compared with OAASP, for the residual Al(OH)3 in the sprayed powder was completely converted to Al2O3 under the high temperature of plasma spraying, which could enhance the stability of the coating.The main phases were Al, different forms of Al2O3, AlN, etc. Al2O3 in the OAAC had three phases, which were α-Al2O3, γ-Al2O3, and δ-Al2O3.Among them, α-Al2O3 is the most stable form of the Al2O3 crystal form, which was the key to the high performance of the alumina ceramic coating.γ-Al2O3 and δ-Al2O3 are transitional alumina with unstable properties, and when they are heated to above 1200 °C, most of them can be converted into α-Al2O3, which could enhance the performance of the coating.The plasma arc outer flame temperature was above 2000 °C, and theoretically γ-Al2O3 and δ-Al2O3 should have been completely converted into α-Al2O3.However, in the actual spraying process, due to the excessive powder feed speed and the transition state alumina powder being wrapped by other types of powder, it could not be fully converted into α-Al2O3.For small particles, the metastable forms of Al2O3 are normally retained and also γ-Al2O3 is usually formed from molten particles by reason of low interfacial energy.

Orthogonal Experiment
The spray process had a large impact on the coating properties, while the spray process and coating properties all contained multiple parameters.An orthogonal experiment is a commonly used multi-factor and multi-level experimental design method, which can reduce the workload, and the experimental results are evenly dispersed and neatly comparable.This experiment mainly verified the feasibility of optimizing the performance of the coating by orthogonal experiment.Therefore, the spray current, spray voltage, the powder flow rate and the main gas flow were selected as the research variables.Other spraying process parameters such as spray distance (100 mm) and spray angle (60°) remained unchanged.Finally, the best coating process parameters for OAAC was determined by testing the coating properties.This study conducted an L9 (3 4 ) orthogonal experiment, and the parameter orthogonal table is shown in Table 4.The microhardness, adhesive strength, porosity, and abrasion rate of the obtained coating samples were taken as the indicators, and the orthogonal test results were analyzed by comprehensive scoring method.

Orthogonal Experiment
The spray process had a large impact on the coating properties, while the spray process and coating properties all contained multiple parameters.An orthogonal experiment is a commonly used multi-factor and multi-level experimental design method, which can reduce the workload, and the experimental results are evenly dispersed and neatly comparable.This experiment mainly verified the feasibility of optimizing the performance of the coating by orthogonal experiment.Therefore, the spray current, spray voltage, the powder flow rate and the main gas flow were selected as the research variables.Other spraying process parameters such as spray distance (100 mm) and spray angle (60 • ) remained unchanged.Finally, the best coating process parameters for OAAC was determined by testing the coating properties.This study conducted an L9 (3 4 ) orthogonal experiment, and the parameter orthogonal table is shown in Table 4.The microhardness, adhesive strength, porosity, and abrasion rate of the obtained coating samples were taken as the indicators, and the orthogonal test results were analyzed by comprehensive scoring method.
First, the weight of each indicator was given according to the different importance, and then the weighted indicator was calculated for each test, in order to convert into a single indicator problem.The comprehensive score was 100 points, in which microhardness, porosity, abrasion rate, and adhesive strength are 25 points respectively.Because the porosity and abrasion rate of the coating were negatively related to the coating quality, it was a negative value.The microhardness, adhesive strength, porosity, and abrasion rate listed in Table 6 were calculated from the test results.In the lower half of Table 6, k 1 , k 2 , and k 3 are the average values of the first, second, and third levels of the respective factors.R is the extreme difference (i.e., the difference between the maximum and minimum values in k 1 , k 2 , and k 3 ), which reflected the degree of influence of the listed factors on the indicators of the sample; that is, the larger the R, the greater the influence of the listed factors on the indicators.The primary and secondary indicators affecting the performance of OAAC could be obtained by orthogonal test, i.e., spray current > spray voltage > main gas flow > powder flow rate.When the spray current was 600 A, the spray voltage was 60 V, the main gas flow rate was 33 slpm, and the powder flow rate was 11 V, the performance of obtained coating was superior.The OAAC was prepared according to the preferred parameters, and the performance test results of OAAC are shown in Table 7.Compared with the coating prepared under the O9 process which had the best comprehensive performance in nine groups, the adhesive strength was increased by 33%; compared with the 45-steel substrate, the hardness was increased by about 100%.Spray was carried out using OAASP as raw materials under the preferred process parameters of Table 7. Considering the cost of spray powder, adding 0%-50% high-purity alumina as modifier, the change of comprehensive performance of OAAC was investigated.The specific proportion of alumina in spray powder is shown in Table 8.

•
The spray process parameters were optimized by orthogonal experiments, but simply four factors were considered in our research.In the future, factors like spray distance and spray angle can be taken into account to obtain an optimal solution.

Figure 2 .
Figure 2. Stratification of aluminum ash after hydrolysis and resting.

Figure 2 .
Figure 2. Stratification of aluminum ash after hydrolysis and resting.

Figure 2 .
Figure 2. Stratification of aluminum ash after hydrolysis and resting.

Figure 6 .
Figure 6.Angle of repose of different samples.

Figure 6 . 6 .
Figure 6.Angle of repose of different samples.6. Angle of repose of different samples.

Figure 8 .
Figure 8. SEM image of the coating surface.

Figure 8 .
Figure 8. SEM image of the coating surface.

Figure 8 .
Figure 8. SEM image of the coating surface.

Figure 9 .
Figure 9.The SEM images of the coating surface with corresponding energy-dispersive X-ray spectroscopy (EDS) spectra: (a) light region; (b) grey region; (c) dark region.

Figure 9 .Figure 10
Figure 9.The SEM images of the surface with corresponding energy-dispersive X-ray spectroscopy (EDS) spectra: (a) light region; (b) grey region; (c) dark region.

Table 1 .
Chemical components of the original aluminum ash (OAA).
Figure 1.The XRD pattern of OAA.

Table 3 .
Spray process parameters of bonding layer.

Table 3 .
Spray process parameters of bonding layer.

Table 3 .
Spray process parameters of bonding layer.

Table 4 .
The factors and their levels for L9(3 4) orthogonal test.

Table 5 .
The X-ray fluorescence (XRF) analysis results of OAASP.

Table 5 .
The X-ray fluorescence (XRF) analysis results of OAASP.

Table 7 .
The performance of coating prepared under preferred parameters.