3.1.1. Cylinder Liner Corrosion Analysis
Table 5 and
Figure 2 present the average corrosion depths with the standard deviation and error bar of the CRGCI, HCCI, and NCI cylinder liner specimens exposed to the three types of corrosive solutions. As the concentrations of formaldehyde and methanoic acid increased, the corrosiveness of these liquids toward the specimens gradually intensified. The average corrosion depths of CRGCI in the three corrosive solutions were measured as 0.042, 0.055, and 0.591 µm. Notably, the corrosion depth was lowest in corrosive solutions #1 and #2. With increasing concentrations of formaldehyde and methanoic acid in the corrosive liquid, the corrosion depth of NCI increased from 0.128 to 1.241 µm. In corrosive solution #3, the corrosion depth sharply increased, reaching 2.1 times that of CRGCI. The average corrosion depths of HCCI in the three corrosive solutions were 0.105, 0.352, and 0.232 µm, and in corrosive solution #3, which had high formaldehyde and methanoic acid content, the corrosion depth was the lowest compared with the other two materials.
As shown in
Figure 2, CRGCI exhibited optimal corrosion resistance in low-concentration corrosion solutions #1 and #2, with corrosion depths measuring 0.042 ± 0.005 and 0.055 ± 0.006 µm, respectively. The low standard deviations observed for CRGCI (e.g., ±0.005 μm in solution #1) confirmed the reproducibility of its superior corrosion resistance under low-concentration conditions. This was primarily attributed to the microalloying effect of 0.2–0.8% chromium elements within its pearlite matrix, as shown in
Figure 3a. In the corrosion environment of low-concentration methanoic acid (0.01–0.02%) and formaldehyde (0.5–1.5%), an oxide film predominantly composed of Fe
3O
4 preferentially formed on the surface of the Fe–Cr solid solution. Although its homogeneous density was lower than that of Cr
2O
3, the synergistic effect of Sn (0.05–0.2%) and Cu (0.1–0.5%) in the composition effectively inhibited the anodic dissolution reaction Fe→Fe
2+ + 2e
−. Furthermore, the directional distribution of flake graphite (
Figure 3b) mitigated the degree of continuous matrix damage and reduced the formation of local corrosion microcells. However, as the concentration of the corrosion solution increased to #3 (0.1% methanoic acid + 2.5% formaldehyde), formaldehyde and methanoic acid created an acidic environment in the solution. Methanoic acid (HCOOH), as a weak acid, could partially ionize to generate H
+ and formate ion (HCOO
−), according to Formula (2):
The increase in H
+ concentration led to the acidolysis reaction of the Fe
3O
4 film, as represented by Formula (3), which resulted in a sudden increase in corrosion depth to 0.591 μm:
The primary components of NCI were Fe, C, Si, and Mg. The graphite present existed in spherical form, as illustrated in
Figure 4, and it also contained a small amount of Cr and other alloying elements. Due to the low chromium content, ductile iron could not form a stable Cr
2O
3 film, unlike HCCI. Instead, it primarily generated Fe
2O
3 or Fe
3O
4 [
23]. These oxide films provided inadequate protection, leading to a significant increase in corrosion depth as the concentration of the corrosive liquid increased. Furthermore, spherical graphite in ductile iron could function as a cathodic phase, exacerbating local galvanic corrosion. The overall corrosion mechanism of ductile iron in the presence of formaldehyde and methanoic acid is shown in Formulas (4) and (5):
The corrosion kinetics of HCCI showed nonlinear characteristics. The average corrosion depths in the #1 and #2 corrosion solutions were 0.105 and 0.352 μm, respectively, which fell to 0.232 μm in corrosion solution #3. As previously mentioned, CRGCI and NCI exhibited increased corrosion depth in the #3 corrosive solution due to the acid concentration rise. This increase triggered the acidolysis reaction of the Fe
3O
4 oxide film on the specimen surface (see Equation (3)), thus diminishing its protective effect. However, HCCI demonstrated reduced corrosion depth under high-concentration #3 conditions, owing to the synergistic effects of dynamic passivation, selective precipitation, and microstructural stability. The corrosion mechanism was attributed to the 25–30% high chromium content, which promoted the formation of a continuous Cr
2O
3 passive film on the test piece surface. This formation followed the reaction 2Cr + 3H
2O→Cr
2O
3 + 6H
+ + 6e
− [
32]. Additionally, formaldehyde (HCHO) was oxidized to form methanoic acid or other intermediate products in an acidic environment, which further increased the solution’s acidity. Chromium (Cr) was oxidized in an acidic environment to generate chromium ions (Cr
3+), as shown in the reaction in Equation (6):
In low-concentration corrosion solution #1, the Cr
2O
3 passivation film remained stable through dynamic equilibrium. In medium-concentration solution #2, the HCOO
− generated by the ionization of methanoic acid formed a complex with Cr
3+, which triggered local pitting and led to an increase in corrosion. In solution #3, the increased H
+ concentration sped up the dissolution of metastable Fe-rich oxides, including Fe
3O
4, facilitating the dynamic reconstruction of a denser Cr
2O
3-Fe
2O
3 composite passive film. This film exhibited a reduced defect density and improved chemical stability owing to covalent Cr
3+-O bonding, which effectively obstructed the permeation of corrosive ions [
33]. Simultaneously, the increase in methanoic acid concentration aided in the formation of Fe(HCOO)
2 precipitate from Fe
2+ and HCOO
−. This reaction, in turn, inhibited the anodic reaction of corrosion and reduced the corrosion depth. In addition, the electrochemical potential difference between the chromium-alloyed ferrite matrix and (Fe, Cr) carbides in HCCI was minimized, thereby mitigating micro-galvanic corrosion [
34]. Contrastingly, the spherical graphite in NCI acted as a cathodic phase, which intensified localized galvanic corrosion at the graphite–matrix interface. The interconnected (Fe-Cr) carbide network in HCCI (
Figure 5) further enhanced corrosion-wear resistance, thus preserving the integrity of the passive film under corrosion–wear coupling conditions. These mechanisms collectively accounted for the reduced HCCI corrosion depth in high-concentration solution #3 compared with that of CRGCI and NCI.
CRGCI exhibited a slow corrosion rate and minimal corrosion depth in environments with low concentrations of formaldehyde and methanoic acid (#1 and #2). Under actual working conditions, when the methanol engine was operational, the surface of the cylinder liner was coated with a lubricating oil film, which served to isolate the corrosive medium produced by the acidic byproducts of methanol combustion. In addition, the high-temperature environment (>100 °C) accelerated the formation of the lubricating oil film and promoted the partial densification of the iron oxide layer, indirectly enhancing short-term corrosion resistance. Due to its balanced performance and cost advantages, CRGCI could serve as an ideal material for cylinder liners.
3.1.2. Piston Corrosion Analysis
Table 6 presents the average corrosion depths observed during the corrosion tests of the F38MnVS steel and cast aluminum alloy ZL109 pistons. The F38MnVS steel primarily experienced reduction in depths in the three types of corrosion solutions, recorded at 0.155 ± 0.02, 0.166 ± 0.03, and 0.676 ± 0.04 µm. Notably, the reduction in depths due to corrosion significantly increased with a higher concentration of formaldehyde and methanoic acid in the corrosion solution. The corrosiveness of corrosion solution #3 toward the steel was the highest, indicating that the dissolution effect of the high-concentration methanoic acid formaldehyde corrosion medium on the steel matrix was more pronounced. The cast aluminum alloy ZL109 exhibited increases in depths after the corrosion test, with gains of 0.746 ± 0.05, 1.622 ± 0.05, and 1.201 ± 0.04 µm. The F38MnVS steel primarily exhibited weight loss in three types of corrosion solutions, with losses recorded at 1.4, 1.5, and 6.1 mg, respectively. In contrast, the ZL109 demonstrated weight gains of 2.3, 5.0, and 3.7 mg following the corrosion tests. As illustrated in
Figure 6, the weight loss change rates of the steel piston specimens were 0.0095%, 0.0102%, and 0.0414%, while the weight gain change rates of the aluminum piston specimens were 0.0457%, 0.0994%, and 0.0736%. In general, the weight change rate of the aluminum piston specimens was greater than that of the steel pistons. The corrosion data are shown in
Table 6 and
Figure 6, indicating that the corrosion resistance of the aluminum piston in the methanol engine was inferior to that of the steel piston.
The corrosion behavior of the F38MnVS steel piston in formaldehyde/methanoic acid solutions was characterized by a localized electrochemical dissolution mechanism [
21,
22]. In the presence of formaldehyde/methanoic acid, the iron matrix functioned as the anode, undergoing a dissolution reaction represented by Fe→Fe
2+ + 2e
−. Concurrently, the H
+ ions in the corrosion solution participated in a hydrogen evolution reaction at the cathode, described by 2H
+ + 2e
−→H
2↑. As the concentration of the corrosion solution increased from #1 to #3, the acidity rose, which accelerated the oxidation of Fe
2+. This reaction could be expressed as 4Fe
2+ + O
2 + 6H
2O→4FeOOH + 8H
+. The resulting FeOOH was subsequently transformed into soluble Fe
3+ in the more acidic environment, leading to the formation of a porous non-protective oxide film. The sulfur content (0.035–0.075 wt%) in the steel led to the formation of MnS inclusions, which functioned as micro-cathodes, thus accelerating pitting corrosion at the interfaces between the inclusions and the matrix [
35]. As a result, the weight loss of the steel specimens in the #3 corrosion solution reached 6.1 mg, representing a 336% increase compared with that in corrosion solution #1.
Conversely, the ZL109 cast aluminum alloy exhibited dynamic passivation–dissolution corrosion [
36]. Initially, the aluminum surface formed a protective Al
2O
3 layer through the reaction 4Al + 3O
2→2Al
2O
3. This precipitated the Al
2O
3 passivation film adhering to the surface of the test piece, resulting in an increase in mass. In the experiment, the weight gains observed for samples #1 and #2 were 2.3 and 5.0 mg, respectively. However, the acidic medium (H
+) continuously dissolved this oxide film through the reaction Al
2O
3 + 6H
+→2Al
3+ + 3H
2O, which exposed the underlying matrix to further corrosion. The regenerated Al
3+ ions hydrolyzed to form Al(OH)
3 precipitates through the reaction Al
3+ + 3H
2O→Al(OH)
3↓ + 3H
+, and subsequently reacted with formate ions (HCOO
−) to produce insoluble aluminum formate complexes (Al(HCOO)
3↓). Although the Al
2O
3 film was continuously regenerated, high H
+ concentrations in solution #3 sped up Al
3+ dissolution (Al→Al
3+ + 3e
-), reducing solid deposition and explaining the lower weight gain (3.7 mg in solution #3 versus 5.0 mg in solution #2).
3.1.3. Piston Ring Corrosion Analysis
Figure 7 illustrates the corrosion weight loss of coated rings (DLC-, CKS-, and PVD-coated) with identical geometric specifications but different surface coatings when immersed in three distinct corrosive fluids. The test data indicated that the DLC coating exhibited the lowest weight loss in corrosion solutions #1, #2, and #3, measuring 0.283, 0.597, and 0.982 mg, respectively. By contrast, the CKS coating exhibited slightly higher weight loss than the DLC coating in low-concentration corrosion solutions #1 and #2, with measurements of 0.948 and 2.048 mg, respectively. However, the weight loss of the CKS coating increased sharply to 8.406 mg in high-concentration corrosion solution #3. The PVD coating demonstrated the highest weight loss in corrosion solutions #1 and #2, with values of 2.869 and 2.205 mg, respectively. However, in the high-concentration #3 corrosion solution, the weight loss fell between that of the DLC and CKS coatings, measuring 3.357 mg.
The weight loss of the DLC coating in the three corrosion solutions was significantly lower than that of the PVD and CKS coatings, and its excellent corrosion resistance was attributed to the physical shielding effect of the amorphous carbon structure. The dense carbon layer inhibited contact between the corrosive medium, H
+ and formate ion (HCOO
−), and the substrate metal, effectively hindering the penetration of H
+ generated by methanoic acid ionization. However, in the high-concentration #3 corrosion solution, the weight loss of the DLC coating increased by 64.5% compared with that in the #2 corrosion solution. This increase was due to the increasing concentration of methanoic acid and formaldehyde, which triggered a micro-region dissolution reaction (Fe + 2H
+→Fe
2+ + H
2↑) in the iron substrate of the test piece, leading to a gradual increase in the corrosion weight loss of the DLC coating. The weight loss of the PVD coating in low-concentration corrosion solutions #1 and #2 was greater than that of the DLC coating. This phenomenon could be attributed to the formation of micron-sized channels at the grain boundaries in the CrN phase, which accelerated the penetration of the corrosion solution. In addition, the micro-galvanic effect generated by the CrN phase (potential of +0.74 V) and the Cr
2N phase (potential of −0.42 V) enhanced the anodic dissolution of the Cr
2N phase at the grain boundaries, represented by the reaction Cr→Cr
3+ + 3e
− [
37]. Conversely, in the #3 high-concentration corrosion solution, the weight loss was lower than that of the CKS coating. This could be explained by the fact that H
+ ions promoted the dynamic reconstruction of the Cr
2O
3 passivation film, as indicated by the reaction 4CrN + 3O
2 + 6H
2O→2Cr
2O
3 + 4NH
3↑ [
32]. The corrosion behavior of the CKS coating exhibited a significant dependence on concentration. The weight loss observed in the #3 corrosion solution was 8.406 mg, representing a 310% increase compared with the weight loss in corrosion solution #2. This phenomenon could be attributed to the hydrolysis reaction of the Al
2O
3 ceramic phase under acidic conditions, which produced a loose coating structure. Once the Cr matrix was exposed, a localized corrosion micro-battery formed. The potential difference between the Cr-based anode and the residual Al
2O
3 cathode further accelerated the dissolution of the anode.
The superior corrosion resistance of DLC coatings indicates their potential applicability to other engine components exposed to methanol-derived corrosive environments, including valve stems, fuel injectors, and camshaft bearings [
38]. However, extending DLC coatings to cover the entire piston surface poses several challenges. First, while the amorphous carbon structure effectively blocks corrosive ions, large-area deposition on complex geometries (e.g., piston skirts or pin bosses) could result in localized coating delamination due to the thermal expansion mismatch between the DLC layer (coefficient of thermal expansion (CTE ≈ 4–6 × 10
−6/°C)) and the steel/aluminum substrate (CTE: 12–24 × 10
−6/°C) [
39]. Additionally, the high hardness (15–40 GPa) and low ductility of DLC coatings could worsen abrasive wear on softer mating surfaces, including aluminum cylinder liners. For full-piston applications, hybrid strategies include combining DLC-coated rings with selectively hardened piston skirt regions or optimizing interlayer adhesion through plasma-enhanced chemical vapor deposition (PECVD). Such strategies could mitigate these drawbacks while maintaining corrosion resistance [
40].