Seawater Corrosion of the Anodized A1050 Aluminum Plate for Heat Exchangers

Abstract

To confirm the suitability of aluminum for the heat transfer surfaces as a heat exchanger material for ocean thermal energy conversion, the seawater corrosion resistance of aluminum plates in a plate heat exchanger was experimentally investigated. In this study, four different surface shapes with chevron angles of 45° and 60° and different treatment types of A1050 aluminum heat transfer surfaces were processed into herringbone patterns. Additionally, the surfaces of the test plates were either anodized or untreated. In continuously flowing deep ocean water, the surface conditions of the test plates were observed at 1, 3, 6, and 12 months using mass measurements, visual inspection, laser microscopy, and SEM. For the anodized A1050 plates, regardless of the surface shape, there was almost no change in the mass, laser microscopy, or SEM results even after 12 months. In contrast, the untreated plate mass decreased in the samples after 3 months or later, and the mass reduction rate was approximately 2–7%. In conclusion, untreated aluminum is not suitable for use in seawater and an anodizing treatment is necessary for its use in heat exchangers for ocean thermal energy conversion.

1. Introduction

A plate heat exchanger (PHE) is a type of heat exchanger with a high heat transfer efficiency due to its large heat transfer area per unit volume [1]. They are widely used in industrial fields, mainly for heating and cooling operations in air-conditioning equipment for heating and cooling, the food industry, and small-temperature-difference power generation plants [2]. A PHE consists of two frames, multiple thin metal plates for heat transfer, and clamping bolts. The multiple metal plates form several flow paths, allowing for heat exchange through the alternate flowing of hot and cold fluids. Steel is typically used as the frame material, whereas stainless steel and titanium are often used for the metal plates. PHEs are also used in geothermal power generation [3] and ocean thermal energy conversion (OTEC) [4], which utilizes the temperature difference between the surface and deep seawater as a heat source. In geothermal power generation, it is necessary to carefully select the material of the heat exchanger according to the geology, because highly corrosive hot spring water is used [3]. The evaporators and condensers used in OTEC systems are PHEs. In the PHEs used for OTEC, titanium, which is highly resistant to corrosion, is used for the metal plates to facilitate heat exchange between hot seawater and ammonia [5]. However, titanium has a low thermal conductivity, which hinders the heat transfer performance of the heat exchanger. Additionally, owing to its high processing cost, alternative materials to titanium are required. Aluminum, which has a lower material cost and higher thermal conductivity, is being considered as an alternative material. Aluminum is already used in other types of heat exchangers. One example is the plate–fin–tube heat exchanger used in automobile radiators, where the aluminum usage rate is nearly 100% [6]. Additionally, aluminum is also used in the mini-channel heat exchangers of air conditioners [7]. However, there are no examples of using aluminum in plate heat exchangers. To apply aluminum in PHEs for OTEC, it is necessary to preliminarily examine its corrosion resistance and heat transfer performance in ammonia and seawater. Regarding ammonia resistance, the authors previously conducted immersion experiments with ammonia [8] using aluminum alloy samples subjected to surface treatments, such as PEEK resin, DLC coating, anodization, and anodizing paint composite coatings. For the tested materials, anodization was found to yield the highest corrosion resistance. Combe [9] proposed polymer coatings on metals as a countermeasure against ammonia corrosion in OTEC applications. On the other hand, the corrosion of aluminum by seawater also needs to be considered. Blawert et al. [10] investigated the effects of the surface morphology, composition, and corrosion on AA6082 substrates with single-layer epoxy coatings in artificial seawater. Meanwhile, Kiepfer et al. [11] proposed using polypropylene–graphite composite materials for heat transfer plates to provide corrosion resistance without using metals. Srinath et al. [12] suggested Cr₂O₃ + 10% TiC coatings on Al-6061 plates for seawater corrosion resistance. Lai et al. [13] investigated the corrosion of aluminum alloys by seawater and microorganisms in seawater, pointing out the acceleration of corrosion by microorganisms. Additionally, Berdibekov et al. [14] revealed that the addition of iodine prevents corrosion caused by the incomplete oxidation of aluminum in seawater. In addition, to evaluate the seawater resistance of aluminum alloys in OTEC, Dexter [15] investigated the seawater corrosion of A5052 and pure aluminum materials when used in intake pipes. Foust [16] proposed enhancing the corrosion resistance of aluminum alloy brazed heat exchangers used in OTEC by adding a zinc protective layer. Panchal [17] considered the application of aluminum materials to OTEC and observed the corrosion of various aluminum alloys. In this study, the author focused on anodization, which demonstrated ammonia corrosion resistance. Yabuki et al. [18] investigated the seawater corrosion resistance of anodized aluminum materials with self-healing properties and confirmed their effectiveness. However, the authors considered it necessary to investigate the seawater corrosion resistance of aluminum materials used as heat transfer surfaces in plate heat exchangers. In this study, a demonstration test using a heat exchanger was conducted. To enable the use of aluminum as the heat transfer surface in a PHE, it is necessary to investigate the seawater corrosion resistance. In this study, a demonstration test was conducted using a PHE with aluminum heat transfer surfaces in seawater. The objective of this study was to evaluate the seawater resistance of aluminum treated by anodic oxidation. A long-term seawater flow test was performed in a deep seawater circulation facility using a PHE with the following two types of A1050 aluminum plates: herringbone plates with and without anodic oxidation treatment. These plates were exposed to deep seawater for 12 months and their corrosion conditions were observed.

2. Experiment

2.1. Experimental Apparatus

Figure 1 shows a schematic of the experimental apparatus. The experimental setup consisted of two PHEs equipped with deep seawater (DSW) piping systems and aluminum plate test specimens. The DSW was continuously supplied by branching off from the DSW piping of the OTEC demonstration facility installed at the Okinawa Prefectural Deep Water Research Center. The state quantities in the piping were measured using temperature, the volumetric flow rate, and pressure sensors. The temperature of the DSW at each inlet and outlet of the PHE was measured using K-type thermocouples (Hayashidenko, Tokyo, Japan: ST6-K-S1.6-150-3KX), the volumetric flow rate was measured using an ultrasonic flow meter (KEYENCE, Osaka, Japan: FD-Q10C, range 0–30 L/min), and the pressure was measured using a pressure transducer (Yokogawa, Tokyo, Japan: FP101-C31, range 0–1 MPa). The measurements from each sensor were collected using a data logger (GRAPHTEC, Kanagawa, Japan: midi LOGGER GL820) and recorded every second.

The test specimens were analyzed using mass measurements and surface observations. An electronic balance (Sartorius, Tokyo, Japan: TE1502S) was used for the mass measurement, a digital camera (RICOH, Tokyo, Japan: PENTAX K-S1) was used for the surface observation of the test specimens, a digital microscope (KEYENCE, Osaka, Japan: VHX-X1) was used for the cross-section of the test specimen observation, and a laser microscope (Olympus, Tokyo, Japan: OLS-4100) was used to analyze the depth of the pitting corrosion. A scanning electron microscope/energy dispersive X-ray spectroscope (SEM/EDX, Hitachi High-Tech, Tokyo, Japan: SU5000) was used to analyze the corrosion products through elemental mapping.

2.2. Test Sections and Test Specimens

Figure 2 shows the test section of a PHE. The PHE consisted of front and rear frames made of SUS304 (height, 380 mm; width, 140 mm; thickness, 15 mm), gaskets, and aluminum corrugated plates as test specimens. Additionally, the PHE formed a single flow path by sandwiching a gasket between two aluminum plates.

The PHE had inlet and outlet ports for both the high- and low-temperature sides. The DSW piping was connected to the inlet port of the PHE on the supply side and to the outlet port of the PHE on the discharge side.

Figure 3 shows the configuration of the PHE flow channels. The flow channels were composed of multiple specimens, and seawater flowed from the two inlets of the PHE, alternating through each flow channel, allowing both sides of the specimens to come into contact with the seawater. Figure 4 shows the cross-sectional structure of the flow channels. One flow channel was composed of two specimens, alternately arranged up and down. The cross-section of the flow channel was wavy, resulting in a complex flow of seawater through the flow channel.

The two PHEs are referred to as Test section-1 and Test section-2. Test section-1 included test specimens for 1-month and 3-month water flow experiments, while Test section-2 included test specimens for 6-month and 12-month experiments. Figure 5 shows photographs of the test specimens. The specimens were corrugated plates made of A1050 aluminum, 350 mm high, 100 mm wide, and 1 mm thick, pressed into a herringbone pattern. Four types of specimens (TP-1 to TP-4) with different surface shapes and treatments were used. The surface shapes included two chevron angles, 45° and 60°, and the surface treatments included anodization and no treatment.

Table 1. Shapes and treatment methods of the test specimens.

Test Plate	Material	Surface Shape	Surface Treatment
TP-1	A1050	Chevron angle 45°	Anodic oxidation
TP-2	A1050	Chevron angle 60°	Anodic oxidation
TP-3	A1050	Chevron angle 45°	Without treatment
TP-4	A1050	Chevron angle 60°	Without treatment

lists the four types of test specimens (TP-1 to TP-4), and

Table 2. Design of the test specimens and flow channels.

		TP-1, 3	TP-2, 4
Plate length, L	mm	350
Plate width, W	mm	100
Port distance, Lp	mm	298
Thickness, t	mm	1
Chevron angle, β	deg	45	60
Wave height, h	mm	2.48	2.44
Wave pitch, p	mm	8.130	8.143
Surface area, S	mm2/plate	77,102	77,721
Cross sectional area of channel	mm2/channel	17.6	17.3
Density, ρ	g/mm3	0.2705

lists their specifications.

Anodization was performed according to the surface treatment specified in JIS H8601. Oxidation was performed using the sulfuric acid method, followed by post-treatment with boiling water and metal salt sealing. The film thickness was equivalent to AA15 (measured at 20 to 25 μm). Figure 6 shows photographs of the cross-sectional microstructure near the surface of the test specimen. Figure 6a,b show that there is an approximately 20 μm anodized layer on the surface of the base A1050. Additionally, Figure 6c,d show the surface of the A1050 exposed due to no treatment. Each PHE test section was constructed by inserting a total of 26 specimens, including 4 specimens each of TP-1 to TP-4 and 10 additional specimens used for the other experiments.

2.3. Experimental Conditions

The experimental conditions were as follows. The flow rate of the DSW periodically varied based on the characteristics of the DSW pumping system at the Okinawa Prefectural Deep Water Research Center. The measured flow rate ranged from 2.7 to 3.7 L/min (2.5 to 3.4 cm/s). Additionally, the DSW temperature fluctuated seasonally, with the inlet temperature of the PHE being approximately 17.5 to 18.7 °C. The DSW was continuously circulated for 12 months; however, the test specimens were removed and observed after 1, 3, 6, and 12 months.

2.4. Pre-Processing

The test plates removed from the test section were pre-treated before the analysis according to the following procedure:

• Rinse off the remaining seawater and deposits on the surface using tap water.
• Air-dry for one day.

3. Results and Discussion

3.1. Mass Reductions in the Test Plate

Mass measurements were taken at 1, 3, 6, and 12 months to determine the corrosion rate of each specimen from the start of the seawater flow. The corrosion rate V_c [mm/y] was determined based on the measured mass changes.

The corrosion rate V_c [mm/y] was defined as shown in Equation (1) in terms of the mass of the specimen before continuous water flow m_B [g], the mass after seawater flow m_A [g], the surface area of the specimen S [m²], and the specimen density ρ [g/mm³]:

$$V_c=(m_B-m_A)/\rho S$$

(1)

Table 3. Corrosion rates of the test specimens at each month of measurement.

	Months	Original Mass mb [g]	Mass Difference ma − mb [g]	Corrosion Rate Vc [mm/y]
TP-1	1	90.29	0.03	1.53 × 10−3
	3	90.40	0.00	0.00 × 100
	6	91.05	0.03	3.20 × 10−4
	12	90.75	0.20	9.59 × 10−4
TP-2	1	92.05	0.02	9.51 × 10−4
	3	91.97	0.02	4.44 × 10−4
	6	91.87	0.05	4.44 × 10−4
	12	91.02	0.49	2.31 × 10−3
TP-3	1	90.46	0.55	3.16 × 10−2
	3	90.37	0.62	1.18 × 10−2
	6	90.60	1.88	1.80 × 10−2
	12	90.86	1.88	9.01 × 10−3
TP-4	1	91.32	0.60	3.44 × 10−2
	3	91.59	1.96	3.72 × 10−2
	6	91.38	3.11	2.96 × 10−2
	12	91.51	6.45	3.07 × 10−2

lists the mass changes and corrosion rates of specimens TP-1 to TP-4 at 1, 3, 6, and 12 months. As shown in Table 1, the mass changes in the anodized TP-1 and TP-2 were less than 1 g for all months. Additionally, the maximum corrosion rates were 1.53 μm/y for TP-1 and 2.3 μm/y for TP-2. The highest measurements were observed at 1 month for TP-1 and 12 months for TP-2, indicating that the changes in the corrosion rates over time varied depending on the shape of the specimen. On the other hand, the maximum corrosion rates were 31.6 μm/y for TP-3 and 37.2 μm/y for TP-4. The rate is approximately 20 times compared to the treated specimens.

In contrast, the mass changes in the untreated TP-3 and TP-4 increased with the months of measurement, and the changes were greater for TP-4. The values at each month of measurement, except for the 1-month value of TP-3, were almost equal. The values for TP-4 were higher than those for TP-3, which can also be attributed to the differences in shape. The chevron angles of the corrugated plates, viewed from the front, were 45° for TP-3 and 60° for TP-4. The mass flow rate of the seawater flowing through each specimen was the same. However, as listed in Table 2, the cross-sectional area of the TP-4 flow path was smaller than that of TP-3, resulting in a relative increase of 1.7% in the average velocity within the flow channel. Additionally, the surface area of TP-4 was 0.8% larger, resulting in a greater impact of the seawater on the surfaces of the specimens.

3.2. Surface Conditions

To confirm the changes in the surface conditions, photographs of the surface conditions of the specimens captured after seawater flow for 1, 3, 6, and 12 months are shown in Figure 7. For the anodized TP-1 and TP-2, no changes in the surface conditions due to differences in the specimens were observed. Moreover, no corrosion was observed. In contrast, in the untreated TP-3 and TP-4, white deposits along the surrounding grooves were observed for all months. Surface discoloration and pitting corrosion were also observed each month. Additionally, pitting corrosion occurred along the grooves, and, in the 6- and 12-month measurements, this corrosion had completely penetrated the specimen.

3.3. Laser Microscope Analysis

A laser microscope was used to observe the surface conditions of the specimens. Figure 8a–d show the laser + color images and 3D surface height maps for each month of measurement. The observed area was a representative 2.5 mm square region on the surface of each specimen. In specimen TP-1, as shown in Figure 8a, the results for 1 and 3 months showed no pitting corrosion on the surface. Partial surface changes were observed in the 6- and 12-month results; however, no corrosion was noted in the other areas. In contrast, in specimen TP-2, as shown in Figure 8b, no surface changes were observed over the months.

In specimens TP-3 and TP-4 without the surface treatment, as shown in Figure 8c,d, pitting corrosion was observed at 1 month, and surface alteration was observed in more than half of the observed area at 12 months.

From the 3D surface height maps, shown in Figure 8a–d, the pitting area ratio and maximum pitting depth were determined. First, the height with the highest frequency distribution was obtained from the height values of the 3D surface height maps, and its average value was taken as the reference height. Next, the regions with heights where the frequency distribution was less than 1% were considered as pitting. The results are shown in Figure 9a. On the other hand, the maximum pitting depth was determined from the difference between the reference height and the average height in the pitting regions. The results are shown in Figure 9b.

As shown in Figure 9a, TP-4 had the highest pitting area ratio in all of the months of measurement, with a maximum of 12% at 12 months. TP-3 exhibited a maximum of 3%. For the anodized specimens, TP-1 and TP-2, the pitting area ratio was less than 1%, with a maximum of 0.3% for TP-1 and 0.2% for TP-2 at 12 months. The mass changes, shown in Table 3, indicate that the value for TP-4 at 12 months was the largest, showing a trend similar to that of the pitting area ratio.

As shown in Figure 9b, the maximum pitting depth for TP-1 and TP-2 was less than 50 μm. On the other hand, the maximum pitting depths for TP-3 and TP-4 were 80 μm and 125 μm, respectively, which is up to approximately 2.5 times deeper than TP-1 and TP-2. However, this does not align with the area ratio trends shown in Figure 9a. It is considered that TP-1 and TP-2 had locally deeper regions, which were detected as the pitting depth.

3.4. Component Analysis of Corrosion Products

SEM/EDX analysis was conducted to confirm the presence of corrosion products on the surfaces of the specimens. Figure 10 shows the SEM photographs and

Table 4. Mass distribution of the elements analyzed by EDX (APEX™ 2.0 EDS, EBSD software).

	Elements	C	O	Mg	Al	S	Cl	Ni	Br
	Month (s)								[wt%]
TP-1	1	9.2	55.9	0.3	18.0	3.1	-	1.8	11.1
	3	11.0	55.7	0.3	19.1	3.0	0.2	1.8	8.1
	6	18.0	53.3	0.3	17.4	2.6	-	1.6	6.7
	12	12.7	54.3	0.3	20.4	2.9	0.3	1.6	6.8
TP-2	1	7.5	40.8	-	13.5	2.0	-	-	5.9
	3	10.1	55.5	-	19.0	0.0	-	1.6	9.5
	6	12.4	54.9	-	18.6	2.8	-	1.4	9.9
	12	21.1	50.0	-	17.9	2.5	-	1.3	7.2
TP-3	1	7.1	23.1	1.6	43.7	0.5	-	-	24.0
	3	-	13.6	-	58.6	-	-	-	25.9
	6	4.8	49.7	0.4	25.8	6.5	-	-	11.7
	12	3.8	3.7	1.9	64.6	0.1	-	-	22.8
TP-4	1	12.0	28.8	-	58.4	-	-	-	-
	3	5.2	38.9	2.1	37.3	2.1	-	-	13.4
	6	4.9	23.6	1.5	48.9	0.8	-	-	19.5
	12	4.8	22.2	-	50.9	0.8	-	-	20.6

shows elemental map, which includes the results of the composition analysis via EDX. The red frame in the SEM image indicates the EDX measurement area. The SEM images of TP-1 and TP-2 in Figure 10 show that surface irregularities of approximately 1 μm to 5 μm were observed. This is indicative of the sealing process of anodization. No defects due to pitting corrosion were observed on the surface condition each month. On the other hand, the SEM observations of the corroded areas of TP-3 and TP-4 revealed that fine particles adhered to the surface. Compared to TP-1 and TP-2, the surface was smoother, and no fine irregularities were observed. Additionally, large cracks were observed in TP-3 at 6 months, and in TP-4 at 3 and 6 months. Based on the subsequent EDX elemental analysis, these are considered to be Mg crystals. However, no defects due to corrosion were observed on the crystals or the surface. In contrast, the EDM analysis detected mainly aluminum, oxygen, and Mg in all of the specimens as shown in Table 4. The oxygen peaks were stronger in TP-1 and TP-2, indicating the detection of oxygen from the anodized layer. It is considered that the oxygen peaks in TP-3 and TP-4 are due to oxides formed by corrosion.

3.5. Discussion

In this study, the corrosion effects of aluminum in a seawater flow environment, with or without surface treatment and considering the differences in shape, were compared. For the surface treatment, the anodized specimens had a pitting area approximately 1/400 to 1/500 times that of the untreated specimens, and a pitting rate about 1/20 times that of the untreated specimens. Even untreated aluminum has a thin passive film (Al₂O₃) that is a few nanometers thick in the atmosphere, which acts as a barrier to prevent corrosion [19]. However, in seawater, this passive film is locally destroyed by Cl⁻, promoting anodic reactions with the aluminum substrate and leading to corrosion. On the other hand, anodized aluminum has an artificially created oxide film of about 20 μm on the surface, making local destruction less likely. Therefore, anodized aluminum was more resistant to corrosion in a seawater flow environment.

Next, regarding the impact of the surface shape on corrosion, comparing TP-1 and TP-3 with a chevron angle of 45 degrees, and TP-2 and TP-4 with a chevron angle of 60 degrees, the differences in geometric conditions were considered. Due to the difference in the surface shape, the surface area of TP-4 is approximately 0.8% larger. Additionally, due to the differences in the flow path cross-sectional area, the seawater flow rate is approximately 1% higher for TP-2 and TP-4. At 12 months, the pitting rate was approximately twice as high for TP-2 and three times as high for TP-4. This can be inferred to be a direct result of the contact time and surface velocity of the seawater.

4. Conclusions

To determine the effect of the surface treatment on A1050 aluminum in seawater, a one-year seawater flow experiment was performed. The test plates were anodized and untreated A1050 aluminum alloy.

(1)

The anodized A1050 aluminum exhibited no surface changes after one year of seawater flow. The maximum corrosion rates were 1.53 μm/y for TP-1 and 2.3 μm/y for TP-2. In addition, there was no mass change due to pitting corrosion.

(2)

The A1050 exhibited no pitting corrosion area over the seawater flow period, as shown by the laser microscopy results. The highest pitting area ratio of TP-1 exhibited a maximum of 0.3%, and that of TP-2 exhibited a maximum of 0.2%, at 12 months.

(3)

The anodized A1050 also showed no surface changes due to the seawater flow period, as observed using SEM.

(4)

The untreated A1050 exhibited surface discoloration and pitting corrosion in the one-month analysis. Additionally, the mass reduction was larger than that of the anodized sample because of the increase in the mass from the surface deposits. The maximum corrosion rates were 31.6 μm/y for TP-3 and 37.2 μm/y for TP-4.

(5)

The untreated A1050 exhibited an increase in the pitting corrosion area over the seawater flow period, as shown by the laser microscopy results. The highest pitting area ratio of TP-3 exhibited a maximum of 3%, and that of TP-4 exhibited a maximum of 12%, at 12 months.

(6)

Owing to the presence of crystals on the surface during the SEM observation, the corrosion state could not be observed. However, the SEM images did not reveal any surface conditions indicative of corrosion.