Mechanical Property Model of Q620 High-Strength Steel with Corrosion Effects

Abstract

High-strength steel (HSS) is widely used in engineering structures, due to its superior material performance, but corrosion tends to occur in steel structures with time. The corrosion effects on mechanical performance of Q620 HSS were investigated experimentally. The electrochemical accelerated corrosion test was conducted to generate corroded Q620 HSS specimens (ρ = 0~60%). With increases in the corrosion degree, corrosion performance became more and more non-uniform. The tensile coupon test was conducted to clarify mechanical properties of corroded Q620 HSS specimens. With increases in corrosion degree, f_y, f_u, ε_u and E decreased, correspondingly. The effect of corrosion on ε_y could be ignored. With the deepening of corrosion, the necking of Q620 HSS specimens was weakened, which reduced their ductility. The simplified constitutive model consisting of nominal yield point (ε_y, f_y) and ultimate point (ε_u, f_u) was proposed to quantify the mechanical properties of Q620 HSS with different corrosion degrees. After the numerical fitting, relationships between the corrosion degree and mechanical properties were clarified. Based on the results of numerical fitting, mechanical properties of corroded Q620 HSS specimens were worse than those of specimens with idealized uniform corrosion. The adverse effect of corrosion on ε_uc was more obvious than that on strength properties. Comparison among different mild steels and HSSs was performed. Different indexes were chosen to clarify corrosion effects on the ductility of corroded Q620 HSS specimens. This study considers and discusses the research on corrosion rates, the relationships between service time, service environment, corrosion form and strength properties of Q620 HSS.

1. Introduction

With progress in the processing technology of steel, the structural steel with excellent mechanical properties, such as high strength, high toughness and good weldability, is produced [1,2]. For the high-strength steel (HSS), the yield and ultimate strength are better than those of mild steel [3]. For the elastic modulus, there is barely a difference between HSS and mild steel [4,5]. For mild steel, it should be noted that there is an obvious yield plateau in the stress–strain curve. However, no yield plateau is recognized in the stress–strain curve of the HSS. Furthermore, compared with mild steel, HSS performs a poorer ductile property. Even though there are some disadvantages with HSS, the wide application of HSS has been seen in the structures recently, because of its superior material performance [4,5,6,7]. The HSS has become one of the research hotspots in engineering, and much research has been conducted to investigate the resistant property of HSS components, such as beams, columns and plate girders [5,8,9,10].

For steel structures, the corrosion issue is hard to avoid during their service life [11]. In the corrosive environment such as the marine and coastal environment, the corrosion issue is prone to occur in steel components because of the high humidity and chloride [12,13,14,15,16,17]. Some high–performance reinforcing and structural steels were suggested to reduce adverse effects from corrosion [18,19,20]. However, the price of these high–performance reinforcing and structural steels was relatively high [21]. No practical design method was proposed, and these issues limited its engineering application. A paint coating is commonly used to prevent the steel component from corrosive factors [22,23,24]. However, after a period of service, the paint coating usually fails because of aging and spalling. After that, serious corrosion issues occur in steel structures, which reduce the mechanical properties of the structural steel [25]. For the steel component, the effective cross-sectional area will be reduced because of the corrosion issue. The resistant property of the component is weakened, correspondingly [26,27,28]. Besides this, corrosion results in a stress concentration, which reduces ductility of the steel component. Furthermore, the failure characteristics of the component with the corrosion issue might be different from those of the component without the corrosion issue [29,30,31]. Blikharskyy et al. [32] studied the fatigue properties of rebars with corrosion experimentally, which indicated that the corrosion had significant effects on the fatigue damage. Lipiński performed an experiment on 1.0721 steel corroded in in 20% NaCl [33]. Taylor et al. introduced the [34] corrosion mechanisms and models.

Based on the above statement, the adverse effect of the corrosion issue on engineering structures should not be ignored. Many researchers have conducted investigations on the adverse effect of the corrosion issue on the resistance of steel components [35,36,37]. The decreases in resistant properties were quantified through the experiment, numerical simulation and theoretical analysis. However, existing studies were mostly based on mild steel components [26,27,38,39,40,41]. The investigation of the adverse effect of the corrosion issue on HSS is relatively limited [42]. It should be noted that compared with that of the stainless steel and weathering steel, the corrosion resistant property of HSS is comparatively low [11,43,44]. Considering that the ductility of the HSS is poorer than that of mild steel, the negative effect of the corrosion issue on HSS mechanical properties might be more significant than that on mild steel mechanical properties. Furthermore, because of the higher strength properties, the use of the HSS usually results in reduction in the cross-sectional area of the component. The corrosion rate is quantified through the corrosion depth per unit time. Therefore, with the same corrosion depth, the effect of corrosion on the HSS component is more serious than that on a mild steel component, when the service time is the same. In this regard, when the service time is the same, corrosion effect on the component with thin-walled steel plate is more severe than that on the component with normal thickness steel plate. The effect of corrosion on mechanical properties of structural steel is important to clarify the effect of corrosion on resistance of the steel component. Hence, it is meaningful to study the mechanical performance of the corroded HSS through the tensile coupon test.

The Q620 HSS developed in the Chinese market was selected to perform the experimental research on mechanical properties of corroded HSS. The electrochemical accelerated corrosion test was the most commonly used method to obtain the corroded test specimens [30,42,44,45,46], and it was also selected for this study. Then, the tensile coupon test was performed to investigate the adverse effects of corrosion on mechanical properties of the Q620 HSS. Based on the test results, a simplified constitutive model was proposed, which quantified the relationship between corrosion degree and mechanical properties. To clarify the mechanical performance of corroded Q620 HSS comprehensively, the comparison among mild steels and HSSs was conducted. Different indexes were introduced to clarify the influence of corrosion on ductility of the Q620 HSS. Considering the investigations on corrosion rate, the relationships between service time, service environment, corrosion form and strength properties of Q620 HSS were discussed.

2. Materials and Methods

2.1. Test Specimen

The Q620 HSS has been widely used in the engineering structures, whose nominal yield strength should not be less than 620 MPa. The chemical composition of the Q620 HSS was listed in

Table 1. Element composition information for Q620 HSS.

Element	Ti	C	Si	Mn	Nb	P	S	V	Ni	Cr	Cu	Mo	B
wt.%	0.09	0.07	0.10	1.71	0.003	0.012	0.001	0.001	0.02	0.02	0.02	0.001	2

. Based on requirements in GB/T 228.1-2010 [47], the geometrical dimensions of the tensile coupon test specimen were determined (see Figure 1). Q620 HSS specimens were cut in the longitudinal direction of a Q620 HSS sheet of 5 mm thickness. The electrochemical accelerated corrosion test was conducted to have corroded Q620 HSS specimens. To avoid the corrosion in the ends of the Q620 HSS specimen, the epoxy resin and electric tape were used to separate the ends of the specimen from electrolyte solution (see Figure 2). To quantify corrosion degree, the mass loss ratio ρ = (m₀ − m_c)/m_e0 was selected, where the m₀ denoted the original mass of the Q620 HSS specimen without corrosion, the m_c denoted the mass of the corroded Q620 HSS specimen and the m_e0 denoted the original mass of the exposed area. Before the electrochemical accelerated corrosion test, each Q620 HSS specimen was treated to remove the mill scale and corrosion product on the surface. Then, the original mass of each Q620 HSS specimen was obtained through an electronic balance with an accuracy of 0.1 g. For the existing studies [42,48], the mass loss ratio of the corroded specimens was comparatively low (20%). With increases in corrosion degree, the corrosion weakening effect might become more and more non-uniform, which resulted in the more significant effect on the mechanical properties of structural steel. Therefore, it is meaningful and necessary to investigate the mechanical properties of HSS with large corrosion degrees. In this section, Q620 HSS specimens with comparatively large corrosion degrees (30~60%) were considered in the experimental investigation to fill the gap of existing researches. Q620 HSS specimens with ρ = 0%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50% and 60% were included in this experiment. Considering the uncertainty of corrosion test, three test specimens were prepared for each predetermined corrosion degree.

2.2. Electrochemical Accelerated Corrosion Test Procedure

After the preparation of Q620 HSS specimens, the accelerated corrosion test setup was built up (Figure 3). The conductor was used to complete the circuit. Test specimens and carbon plate were linked to the positive and negative terminals of the external power supply, separately. To build the current loop, the NaCl solution with 5% mass fraction was selected, whose pH was 7.2. Test specimens and carbon plates were placed in the tank filled with the 5% NaCl solution. It is worth noting that the electrochemical accelerated corrosion test was selected by many researchers [42,44,49]. The direct electric current was impressed on Q620 HSS specimens to accelerate the corrosion. The duration of the electrochemical accelerated corrosion test was calculated through the Faraday’s law (see Equation (1)), where ∆m denoted the mass loss of the corrosion, M denoted the molar mass of the iron (i.e., 56 g/mol), T denoted the duration time of the corrosion test, z denoted the valency of the iron element (i.e., 2) and F denoted the Faraday constant (i.e., 96,500 C/mol). The I denoted magnitude of the electric current determined by the Equation (2), where i denoted the current density and S denoted the exposed surface area of the Q620 HSS specimen. When the above corrosion test was over, the electricity supply was cut off and test specimens were removed from the tank. Before the tensile coupon test, specimens were cleaned based on requirements in the code [50], as shown in Figure 4.

$$\Delta m=MIT/zF$$

(1)

$$I=i·S$$

(2)

2.3. Tensile Coupon Test Procedure

After the electrochemical accelerated corrosion test, the tensile coupon test was conducted to investigate the mechanical properties of corroded Q620 HSS specimens, as shown in Figure 5. The MTS 809 testing system was used to perform the tensile coupon test, which was produced by MTS Systems Corporation (Eden Prairie, MN, USA) in 02/2018. A total amount of 30 Q620 HSS specimens were tested in this study. It should be noted that the tensile coupon test was conducted based on the relevant regulations in the GB/T 228.1-2010 [47]. The monotonic load was applied on Q620 HSS specimens until the failure. The extensometer was selected to measure the displacement of the Q620 HSS specimen during the tensile test, whose gauge distance was 100 mm. The loading was controlled by the displacement, whose rate was 1 mm/min. During the tensile coupon test, the load vs. displacement curve from the extensometer was recorded by the control system.

3. Results

3.1. Corrosion Test Result

After the above corrosion test and the cleaning treatment, the mass of each corroded Q620 HSS specimen was obtained through the aforementioned electronic balance. After that, corrosion degrees of test specimens were obtained (see

Table 2. Corrosion degrees and mechanical properties of Q620 HSS specimens.

Number	ρ (%)	E (MPa)	fy (MPa)	fu (MPa)	fye (MPa)	fue (MPa)	εy	εu	fu/fy	εu/εy
S0-1	0	216,347	694.21	770.08	694.21	770.08	0.0052	0.1075	1.11	20.67
S0-2	0	216,162	675.61	747.59	675.61	747.59	0.0052	0.0979	1.11	18.98
S0-3	0	216,188	686.13	761.31	686.13	761.31	0.0052	0.1016	1.11	19.61
S5-1	5.07	202,088	646.86	721.23	681.41	759.74	0.0052	0.1079	1.11	20.83
S5-2	4.97	200,110	651.09	727.56	685.14	765.61	0.0052	0.0991	1.12	18.95
S5-3	5.21	202,952	654.12	726.94	690.07	766.89	0.0052	0.0993	1.11	19.17
S10-1	10.15	192,322	605.42	676.10	673.82	752.47	0.0052	0.0681	1.12	13.12
S10-2	10.01	192,381	618.72	683.55	687.54	759.59	0.0052	0.0843	1.10	16.09
S10-3	10.16	193,571	622.60	693.71	693.01	772.16	0.0052	0.0999	1.11	19.13
S15-1	15.21	184,316	596.52	662.22	703.53	781.01	0.0053	0.0857	1.11	16.29
S15-2	15.13	183,518	562.66	627.36	662.96	739.20	0.0051	0.1055	1.11	20.80
S15-3	14.79	183,089	598.13	666.81	701.94	782.55	0.0052	0.0963	1.11	18.44
S20-1	20.06	173,076	535.76	572.74	670.20	716.46	0.0051	0.0497	1.07	9.77
S20-2	19.34	172,918	565.80	624.87	701.47	774.69	0.0053	0.0612	1.10	11.61
S20-3	17.96	173,131	563.81	583.28	687.24	710.97	0.0052	0.0417	1.03	7.97
S25-1	25.18	162,940	530.91	591.78	709.59	790.94	0.0053	0.0714	1.11	13.54
S25-2	25.26	162,613	538.74	591.74	720.83	791.73	0.0053	0.0882	1.10	16.58
S25-3	22.52	162,154	536.65	598.36	692.64	772.27	0.0053	0.0763	1.11	14.37
S30-1	29.35	151,442	486.58	544.20	688.71	770.27	0.0052	0.0707	1.12	13.61
S30-2	30.47	151,352	509.12	543.39	732.23	781.52	0.0054	0.0701	1.07	13.05
S30-3	30.56	151,526	-	-	-	-	-	-	-	-
S40-1	40.01	129,415	405.05	451.14	675.20	752.02	0.0051	0.0221	1.11	4.31
S40-2	41.37	129,422	444.45	483.14	758.05	824.06	0.0054	0.0289	1.09	5.35
S40-3	41.21	129,562	385.68	421.26	655.92	716.42	0.0051	0.0205	1.09	4.01
S50-1	50.34	103,777	296.91	312.04	597.89	628.35	0.0049	0.0122	1.05	2.50
S50-2	50.08	105,616	323.58	351.53	648.20	704.19	0.0051	0.0146	1.09	2.88
S50-3	51.06	91,625	299.03	315.49	611.01	644.65	0.0053	0.0153	1.06	2.90
S60-1	64.21	73,025	234.82	240.99	656.11	673.34	0.0053	0.0071	1.03	1.34
S60-2	58.88	74,779	249.62	261.20	607.05	635.21	0.0053	0.0101	1.05	1.91
S60-3	61.12	76,374	241.92	254.79	622.22	655.32	0.0052	0.0094	1.05	1.80

Note: S30-2 denotes that the predetermined corrosion degree was 30% and series number was 2.

). There were some errors between the real corrosion degree and the predetermined corrosion degree. The above errors might be caused by uncertainties of the electrochemical accelerated corrosion test. The errors were comparatively small compared with the predetermined corrosion degree. Therefore, the electrochemical accelerated corrosion test conducted in this study could be used to obtain corroded Q620 HSS specimens effectively. The corrosion performances of test specimens were recorded (see Figure 6). When the corrosion degree was comparatively low (5~20%), the reductions caused by corrosion kept almost the same along the length, which indicated that the corrosion was relatively uniform. No clear weak area was observed in the middle portion of Q620 HSS specimens. However, with increases in corrosion degree, the reductions caused by corrosion became more and more non-uniform, correspondingly.

3.2. Stress–Strain Property

The stress–strain properties of Q620 HSS specimens with different corrosion degrees were discussed in this section. After the above tensile test, the load L and displacement D results of the Q620 HSS specimens with different corrosion degrees were obtained. The cross area of the Q620 HSS specimen without corrosion, A₀, was selected to calculate the nominal stress σ through σ = L/A₀. Considering that the gauge distance of the extensometer is 100 mm, the nominal strain ε was calculated through ε = D/100. Therefore, stress–strain curves of corroded Q620 HSS specimens were introduced in Figure 7. Based on test results of specimens without corrosion, there was no yield plateau in the stress–strain curve. After the elastic segment, the stress–strain curve entered the strain-hardening segment directly. Considering the comparison shown in Figure 8, it is clear to see that with the increase in ρ, strength and ductile properties of Q620 HSS were weakened, gradually. To investigate the strength property of corroded Q620 HSS accurately, the yield and ultimate strength were selected. Considering the feature of the stress–strain curve of Q690 HSS, it should be noted that stress corresponding to the 0.2% residual strain was considered to be the yield strength f_y and the corresponding strain was considered to be the yield strain ε_y. The above definition of mechanical properties of materials without a yield plateau was accepted by many researchers [51,52,53]. The effective cross area A_e was determined by A_e = A₀/(1 − ρ). The effective yield strength f_ye and effective ultimate strength f_ue were calculated, correspondingly. The test results of nominal yield strength f_y, nominal ultimate strength f_u, effective yield strength f_ye, and effective ultimate strength f_ue are shown in Table 2 and Figure 9. It should be noted that the tensile coupon test of the S30-3 Q690 HSS specimen failed, because the non-uniform corrosion was concentrated in the area outside the middle portion of the specimen. When the corrosion degree was comparatively low (0~40%), the f_ye and f_ue remained almost unchanged. When the corrosion degree was larger than 40%, clear decreases in the f_ye and f_ue were observed, which might be caused by the inhomogeneity of the corrosion. The f_y and f_u were negatively related to corrosion degree, as shown in Figure 8 and Figure 9. There were clear differences between effective and nominal strength properties. The elastic modulus E was negatively related to the corrosion degree. The ultimate strain, ε_u, decreased with the increase in corrosion degree. Therefore, the corrosion had an adverse effect on the ductile property of Q620 HSS specimens. Furthermore, effects of corrosion on dimensionless stress and strain properties are shown in Figure 10, where the dimensionless property was the ratio between the stress (strain) property with corrosion to the mean value of stress (strain) property without corrosion. Results showed that with an increment in corrosion, the dimensionless ratios of f_y, f_u and ε_u decreased significantly. The effects of corrosion on dimensionless ratio of ε_y could be ignored.

3.3. Failure Performance

After the above tensile test, failure performances of specimens with different corrosion degrees were obtained, as shown in Figure 11. For the Q620 HSS specimen without corrosion, the obvious necking occurred in the fracture section, which indicated a ductile failure. With development in corrosion degree, the necking was weakened gradually. The reason might be that with increases in the corrosion degree, the weakening effect caused by corrosion got more and more non-uniform. The plastic deformation was confined to the weak area in the middle portion. The stress concentration resulted in the earlier fracture, which was caused by the non-uniform corrosion. Therefore, the corrosion had an adverse effect on the ductile properties of Q620 HSS specimens, which agreed with test results introduced in Section 3.2. Based on the aforementioned discussion, the similar phenomenon might be observed in the steel component, where plastic deformation was confined to the weak area caused by the non-uniform corrosion, which resulted in a significant decrease in the ductile property. This research was focused on the mechanical properties of Q620 HSS with corrosion. It is helpful to clarify the microstructures of the fractures, which was desirable for study in future research.

3.4. Simplified Constitutive Model

3.4.1. Relationship between Corrosion Degree and Characteristic Point

Based on the test results introduced in Section 3.2, features of stress–strain curves of corroded Q620 HSS specimens were obtained. Considering the above features of stress–strain curves, the simplified constitutive model was selected to quantify mechanical properties of corroded Q620 HSS specimens (see Figure 12). The simplified constitutive model consisted of two characteristic points, which were the nominal yield point (ε_y, f_y) and the ultimate point (ε_u, f_u). The mathematical expression of the above simplified constitutive model was shown as Equation (3). Considering the investigation conducted in Section 3.2, the aforementioned characteristic points were influenced by the corrosion degree. The numerical fitting was conducted to clarify relations between corrosion degree and mechanical properties. Based on the test results introduced in Table 2, the linear model was selected to describe the effect of corrosion on mechanical properties. After the numerical fitting, Equations (4)–(7) were proposed to quantify the aforementioned relationships. The f_yc, f_uc, ε_uc, E_c denoted the nominal yield strength, nominal ultimate strength, ultimate strain and elastic modulus of Q620 HSS specimens with different corrosion degrees, respectively. The f_y0, f_uo, ε_u0, E₀ denote the average value of the nominal yield strength, nominal ultimate strength, ultimate strain and elastic modulus of Q620 HSS specimens without corrosion, respectively. The comparisons between the results from Equations (4)–(7) and tensile tests were conducted, as shown in Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17. The Equations Equations (4)–(7) obtained from the numerical fitting built into the f_yc, f_uc, ε_uc and E_c well with R-square = 0.997, 0.980, 0.846 and 0.993, respectively. The test results of ε_u showed a greater discreteness than those of the f_y and f_u. With the increase in corrosion degree, the f_yc, f_uc, ε_uc, E_c decreased, correspondingly. The degradation coefficients of f_yc and f_uc were 1.028 and 1.059, respectively. The degradation coefficient of the idealized uniform corrosion was 1. Therefore, the adverse effect of the non-uniform corrosion in this study was greater than that of the idealized uniform corrosion. Considering that degradation coefficients 1.566 > 1.059 > 1.028, the adverse effect of corrosion on ε_uc was more obvious than that on strength properties. Based on the test results, the average value of ε_yc was 0.0052 and standard deviation was 0.000104, where ε_yc denoted the yield strain of the Q620 HSS specimens with different corrosion degrees. The effect of corrosion on ε_yc of Q620 HSS specimens with different corrosion degrees could be ignored, as shown in Figure 17 Considering the physical significance, the ε_uc should not be less than ε_yc. Therefore, the Equation (6) was only suitable for Q620 HSS where the corrosion degree was less than 60.61%.

$$\sigma =\left\{\begin{array}{l}E\epsilon ,0\le \epsilon \le \left({\epsilon }_y-0.002\right)\\ f_y,\left({\epsilon }_y-0.002\right)<\epsilon \le {\epsilon }_y\\ f_y+\left(f_u-f_y\right)\left(\frac{\epsilon -{\epsilon }_y}{{\epsilon }_u-{\epsilon }_y}\right),{\epsilon }_y<\epsilon \le {\epsilon }_u\end{array}\right.$$

(3)

$$f_{yc}=\left(1-1.028\rho \right)f_{y0}$$

(4)

$$f_{uc}=\left(1-1.059\rho \right)f_{u0}$$

(5)

$${\epsilon }_{uc}=\left(1-1.566\rho \right){\epsilon }_{u0}$$

(6)

$$E_c=\left(1-1.041\rho \right)E_0$$

(7)

The comparison between results from the Equations (3)–(7) and test results was conducted to clarify the accuracy of the simplified constitutive model proposed in this study (see Figure 18). When the corrosion degree was comparatively low, the simplified constitutive model proposed in this study could be used to predict the mechanical properties of corroded Q620 HSS accurately, especially for the elastic segment. With increases in the corrosion degree, the corrosion became more and more non-uniform. The accuracy of the simplified constitutive model reduced, correspondingly. It is worth noting that the degree of corrosion occuring in steel structures usually increased with time. Most of corrosion issues were detected when the corrosion degree was comparatively low. Therefore, the simplified constitutive model proposed in this study was useful for the evaluation of existing steel structures with corrosion issues.

According to GB/T 1591-2018 [54], the f_y and f_u of the Q620 HSS should not be less than 620 and 710 MPa, respectively. Based on the Equations (4) and (5) proposed in this section, the corresponding corrosion degrees were determined for the Q620 HSS investigated in this study. The corrosion degree corresponding to f_yc = 620 MPa and f_uc = 710 MPa were 9.27% and 6.17%, respectively, as shown in Figure 13 and Figure 14. Therefore, to address requirements in GB/T 1591-2018 [54], the corrosion degrees of the Q620 HSS investigated in the research should be controlled within 6.17%.

3.4.2. Comparison with Other Test Results

Based on the investigation conducted in this study, corrosion effects on the mechanical performance of Q620 HSS were clarified through corrosion factors ν_y, ν_u and ν_E, as shown in Equations (8)–(10). A comparison of Q235 mild steel and Q460, Q550 and Q620 HSSs was conducted, as shown in

Table 3. Comparison among mild steels and HSSs.

Study	Type	Accelerated Corrosion Test	νy	νu	νE
Present study	Q620 HSS	electrochemically	1.028	1.059	1.041
Zhang [58]	Q460 HSS	electrochemically	0.925	1.050	1.216
Zhang [58]	Q550 HSS	electrochemically	0.832	0.812	1.196
Shi et al. [55]	Q235 mild steel	neutral salt spray	0.9852	0.9732	-
Cheng [56]	Q235 mild steel	acid salt spray	0.802	0.955	0.836
Zheng et al. [57]	Q235 mild steel	neutral salt spray	0.9684	0.9438	0.8913

. For studies conducted in references [55,56,57], all experiments were focused on Q235 mild steel, and the test results indicated that corrosion factors were influenced by the corrosion method. Q235 mild steel specimens corroded in neutral and acid salt spray produced different corrosion factors. For tests in reference [58] and the present study, Q460, Q550 and Q620 HSSs were treated through the electrochemical accelerated corrosion test. The ν_y and ν_u of Q460 HSS was larger than that of Q550 HSS and smaller than that of Q620 HSS. The ν_E of Q550 HSS was larger than that of Q620 HSS and smaller than that of Q460 HSS. Based on the above comparison, the effects of corrosion on the mechanical properties of Q235 mild steel, Q460, Q550 and Q620 HSSs were different from each other. The effect of corrosion on the mechanical properties of mild steels and HSSs produced a noteworthy uncertainty. The effect of steel strength on corrosion factors was not clear. To determine the corrosion effect on the performance of structural steels and predict the resistance of steel structures during the service life accurately, more corrosion tests should be encouraged.

$$f_{yc}=\left(1-{\nu }_y\rho \right)f_{y0}$$

(8)

$$f_{uc}=\left(1-{\nu }_u\rho \right)f_{u0}$$

(9)

$$E_c=\left(1-{\nu }_E\rho \right)E_0$$

(10)

3.5. f_u/f_y and ε_u/ε_y

The indexes f_y/f_u and ε_u/ε_y were selected in this study to quantify the ductile property of the corroded Q620 HSS. Based on the test results, the average value of the index f_y/f_u was 1.09 and standard deviation was 0.0281, as shown in Figure 19. Therefore, the effect of corrosion on the index f_y/f_u was weak and could be ignored. With increases in corrosion degree, the ξ = ε_u/ε_y decreased, correspondingly, as shown in Figure 20. After a numerical fitting, Equation (11) was suggested to determine the corrosion effect on the ξ_c, where ξ_c denoted the ε_uc/ε_yc and ξ₀ denoted ε_u0/ε_y0. Based on the comparison between the results of Equation (11) and the test, the Equation (11) fitted into the ξ well with R-square = 0.851. The index ξ_c was related to a clear physical significance. Therefore, the index ξ_c should not be less than 1. Hence, the proposed Equation (11) was suitable for the Q620 HSS, whose corrosion degree was less than 60.24%.

$${\xi }_c=\left(1-1.576\rho \right){\xi }_0$$

(11)

3.6. Energy Indexes

For the indexes f_y/f_u and ε_u/ε_y, they focused on either strength or strain properties. To clarify the corrosion effect on both the strength and strain properties, the energy absorption κ and the degree between total energy density and elastic energy density ψ were selected in this section [59]. The energy absorption κ denoted the sum of the area under the nominal stress–strain curve. Considering the test results introduced in Section 3.2, the κ was calculated through the Equation (12). The ψ was obtained through Equation (13). After the calculation, the test results indicated that with the increment in corrosion degree, the κ and ψ decreased, correspondingly, as shown in Figure 21 and Figure 22. Therefore, the corrosion had an adverse effect on the ductile property of Q620 HSS. Based on the distribution of the scatter, the polynomial was selected to conduct the numerical fitting. After the numerical fitting, Equations (14) and (15) were suggested to determine the corrosion effect on the indexes κ_c and ψ_c, where κ_c and ψ_c denoted the energy absorption and the degree between total energy density and elastic energy density of the Q620 HSS specimens with different corrosion degrees, respectively. It is clear that Equations (14) and (15) fitted into the κ_c and ψ_c very well with R-square = 0.901 and 0.852, respectively. Considering that indexes κ_c and ψ_c were related to the clear physical significances, they should not be less than 0 and 1, respectively. Hence, the proposed Equations (14) and (15) were suitable for the Q620 HSS, whose corrosion degree was less than 60.25% and 59.88%, respectively.

$$\kappa ={\displaystyle {\int }_0^{{\epsilon }_u}\sigma d\epsilon }$$

(12)

$$\psi =1+\left(1+f_u/f_y\right)\left({\epsilon }_u/{\epsilon }_y-1\right)$$

(13)

$${\kappa }_c=117.7{\rho }^2-198.9\rho +77.11$$

(14)

$${\psi }_c=10.81{\rho }^2-75.71\rho +42.46$$

(15)

3.7. Forecast of Mechanical Properties

Based on the above investigation, the relations between corrosion degree and mechanical performance of Q620 HSS were established. To perform the reliable resistance analysis of HSS structures during the service life, the relation between corrosion degree and rate was meaningful and necessary. The corrosion rate of structural steel was influenced by many factors, such as the service environment, material type and stress condition. In this study, the experimental investigations conducted in references [35,60,61] were selected. In reference [60], test results of low-alloy steels on 1st, 2nd, 4th, 8th and 16th year were introduced (see Figure 23). The predictive Equation (16) was proposed to fit the above test results, where R denoted the average corrosion rate (μm/a), the t denoted the exposure time (year), and the A and B were parameters obtained from test results in [60]. For the JY235(RE) in reference [60], the parameters A and B were calculated to be 44.12 and −0.7841, respectively. The comparisons between test results and results from Equation (16) were shown in Figure 23, which indicated that the proposed equation could be used to predict the test results of the corrosion rate accurately. The structural steel was homogeneous. The density of structural steel was believed to be independent of the dimensions of structural steel. Therefore, the average thickness loss ratio was equal to the mass loss ratio. When considering the relationship between erosion factors and the surface of steel plate, there were single and double side corrosion forms, as shown in Figure 24. Based on Equations (4), (5) and (16), the relationships between exposure time and strength properties of Q620 HSS with 5 mm thickness were calculated, as shown in Figure 25. In the early stage of service time, the corrosion rate was comparatively high, which meant that the slope of the remaining strength curves decreased, with increases in service time. The reductions in strength properties of Q620 HSS with double side corrosion were more severe than those of Q620 HSS with single side corrosion. In Beijing, for Q620 HSS with 5 mm thickness and single side corrosion, the f_y and f_u could meet the requirements in GB/T 1591-2018 [54], when the service time was less than 50 years. For Q620 HSS with 5 mm thickness and double side corrosion, considering the requirements of f_y and f_u in GB/T 1591-2018, the limitations of the service time were 25 and 9 years, respectively. When the service time was 50 years, the reduction ratios of f_y and f_u of the Q620 HSS with 5 mm thickness and double side corrosion were 0.94 and 0.87, respectively.

To investigate the effects of the service environment, the relationship between corrosion penetration and exposure time in references [35,61] were selected. The corrosion penetration was determined through Equation (17), where the C denoted the average corrosion penetration (μm), the t denoted the exposure time (year), and the φ and ζ were parameters obtained from experimental results. Different environments were included, such as rural, urban and marine environments. In reference [35], the φ of carbon steel in the rural, urban and marine environments were 34.0, 80.2 and 70.6, respectively. The ζ of carbon steel in the rural, urban and marine environments were 0.650, 0.593 and 0.789, respectively. The above values of the φ and ζ were also used in this study. The relations between corrosion penetration and exposure time were obtained, as shown in Figure 26. Based on Equations (4), (5) and (17), relationships between the remaining strength properties and the service time of Q620 HSS with 5 mm thickness were obtained, as shown in Figure 27 and Figure 28. The adverse effects of double side corrosion on remaining strength properties of Q620 HSS were more obvious than those of single side corrosion. The decreases in strength properties of Q620 HSS serving in a marine environment were much more severe than those of Q620 HSSs serving in rural and urban environments. Without corrosion prevention, neither f_y and f_u of Q620 HSS with 5 mm thickness and double side corrosion could meet the requirements in GB/T 1591-2018 during the service time of 50 years, as shown in Figure 27 and Figure 28. For Q620 HSS with 5 mm thickness and single (double) side corrosion, when the service time was 50 years, the remaining ratios of f_y in rural, urban and marine environments were 0.91 (0.82), 0.83 (0.67) and 0.68 (0.36), respectively. For Q620 HSS with 5 mm thickness and single (double) side corrosion, when the service time was 50 years, the remaining ratios of f_u in rural, urban and marine environments were 0.91 (0.82), 0.83 (0.65) and 0.67 (0.34), respectively.

$$R=A{t}^B$$

(16)

$$C=\phi t^{\zeta }.$$

(17)

4. Conclusions

In this study, the mechanical performance of corroded Q620 HSS was investigated experimentally, which was useful to determine the resistance of HSS structures during the service time. The main contributions of this research include the following:

• The electrochemical accelerated corrosion test was conducted in the study to obtain Q620 HSS specimens with different corrosion degrees (ρ = 0%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50% and 60%). When the corrosion degree was comparatively low (5~20%), reductions in the width and thickness kept almost the same along the length, which indicated that the corrosion was approximately uniform. When the corrosion degree was comparatively large (25~60%), the corrosion became more and more non-uniform. The weak area occurred in the middle portion of the specimen, which had an adverse effect on ductility of Q620 HSS;
• The tensile coupon test was conducted to study the mechanical performance of corroded Q620 HSS specimens. After considering the test results, there was no clear yield plateau in the stress–strain curves of corroded Q620 HSS specimens. With the increment in corrosion degree, the f_y, f_u, ε_u and E decreased, correspondingly. The effect of corrosion on the ε_y could be ignored;
• After the tensile coupon test, failure performances of corroded Q620 HSS specimens were obtained. For the Q620 HSS specimen without corrosion, the obvious necking occurred in the fracture area, which indicated a ductile failure. With the increment in corrosion degree, the necking was weakened. With the non-uniform corrosion, plastic deformation was confined to the weak area in the middle portion, which reduced the ductile property of Q620 HSS specimens with different corrosion degrees. The stress concentration caused by the non-uniform corrosion resulted in the earlier fracture;
• The simplified constitutive model consisting of the nominal yield point (ε_y, f_y) and the ultimate point (ε_u, f_u) was proposed to quantify mechanical performance of corroded Q620 HSS. After numerical fitting, the relationships between corrosion degree and mechanical properties were clarified through Equations (4)–(7). The degradation coefficients of f_y and f_u are 1.028 and 1.059, respectively. Considering that the degradation coefficient of the idealized uniform corrosion was 1, the adverse effect of the non-uniform corrosion in this study was larger than that of the idealized uniform corrosion. Because the degradation coefficients 1.566 > 1.059 > 1.028, the corrosion effect on ε_uc was more severe than that on f_yc and f_uc. A comparison in mechanical properties among mild steels and HSSs with different corrosion methods was conducted, where it transpired that the effects of corrosion on the mechanical properties of Q235 mild steel, Q460, Q550 and Q620 HSSs were different from each other;
• Indexes f_y/f_u, ε_u/ε_y, κ and ψ were selected to clarify the effects of corrosion on the ductile property of Q620 HSS with different corrosion degrees. Based on the test results, the effect of corrosion on the index f_y/f_u was weak and could be ignored. With increments in corrosion degree, the ε_u/ε_y, κ and ψ decreased, correspondingly. The corrosion had an adverse effect on the ductile property of Q620 HSS. After numerical fitting, Equations (11), (14) and (15) were suggested to clarify the corrosion effect on the indexes ε_u/ε_y, κ_c and ψ_c;
• Combining with investigations on corrosion rates, the relationship between service time, service environment, corrosion form and strength properties of Q620 HSS with five thickness were discussed. Based on the test results of corrosion rates for structural steel in Beijing, the service life of Q620 HSS was discussed. The decreases in strength properties of Q620 HSS serving in a marine environment were much more severe than those of Q620 HSSs serving in rural and urban environments.