Effects of Quenching on Corrosion and Hardness of Aluminum Alloy 7075-T6

Abstract

Quenching affects the mechanical and corrosion properties of precipitation-hardenable alloys such as aluminum alloy 7075-T6 (AA7075-T6). In this paper, the properties of as-quenched AA7075-T6 are predicted within the framework of quench factor analysis (QFA), using cooling curves obtained from a quench test. Theoretical and computational approaches are used to predict spatial and temporal variations of temperature. The temperature variations are used to predict the quench factor and consequently the material properties. A test is carried out on a block of AA7075-T6 quenched partly in water and partly in air followed by hardness measurements and electrochemical characterizations. The results show that the hardness and the corrosion potential of the quenched block decrease as the cooling rate decreases. The results further suggest the existence of a corrosion product layer for the water-quenched part of the sample. This was not observed for the air-cooled part. A new corrosion prediction model is developed by using the QFA method, cyclic polarization, and electrochemical impedance spectroscopy test results. The present model may be used to potentially reduce the number of corrosion tests in evaluating corrosion properties of quenched AA7075-T6. Model predictions for corrosion and hardness are in good agreement with experimental results.

1. Introduction

Desired mechanical properties of heat-treatable alloys can be reached by controlling three steps of solutionizing, quenching, and aging [1]. For alloys with mechanical properties dependent on cooling rate, such as precipitation-hardenable aluminum alloys, quenching is an essential step [2,3]. Quenching of these alloys is based on the nucleation and precipitation process of their supersaturated solid solution, which is controlled by diffusion reactions [2]. At high temperatures, despite the high diffusion rate, the degree of supersaturation is low, resulting in low precipitation and supersaturation rates. Conversely, at low temperatures, the degree of supersaturation and the time for precipitation are high due to low diffusion and precipitation rates. At intermediate temperatures, the precipitation rate is the fastest because of the highest precipitation diffusion rates of solute atoms [4]. To obtain the best mechanical properties, the highest quenching rate is desirable. However, to minimize residual stresses from the quenching process, low quench rates must be employed [3]. In addition to mechanical properties, corrosion properties of quenched alloys have also been studied in the literature. Fink and Willey [5] studied the effects of quenching on the strength of AA7075-T6 and the corrosion resistance of AA2024-T4. They studied the quenching effects on maximum strength and resistance to corrosion based on average quenching rate. For an average quenching rate of between 800 F/s and 200 F/s, the maximum strength was slightly degraded, but the corrosion resistance was not affected.

Different alloys show different sensitivities to the quenching rate [6]. This can be investigated by plotting strength versus average cooling rate for a specific range of temperature [3]. Different techniques such as the end quench test (which use temperature–time–transformation (TTT) curve), temperature–time–properties (TTP) curve, and continuous–cooling–transformation (CCT) curve have been used to investigate the quench sensitivity of aluminum alloys [3]. The basis of these techniques is to plot the strength against average cooling rate for a particular range of temperatures called the critical temperature. For example, in the TTP curve method, the isothermal precipitation of a supersaturated solid solution as a result of quenching and aging is propounded [3].

The Jominy end quench test, as described in ASTM 255 [7], has been widely used to determine quench sensitivity and mechanical properties of aluminum alloys. In 1962, this method was employed by Bryant [8] to study the quench sensitivity of 7000 series aluminum alloys followed by the work of tHart et al. [9] who studied the Vickers hardness, electrical conductivity, corrosion, and microstructural properties. Microstructural investigations attempted to discover the effects of precipitation and dispersoid formation on the quench sensitivity [10]. Evancho and Staley in the early 1970s [11] developed the quench factor analysis (QFA) method followed by its theoretical justification by Staley in 1987 [3]. This method was then used to predict the post-quenched hardness and strength of aluminum alloys below their critical values [12]. The critical value is the maximum quench factor (Q-factor) that retains 85–100% of the material properties such as yield strength [12,13,14]. The QFA is a prediction technique that relates the cooling curves (i.e., temperature–time curves) to the metallurgical response of a material [15]. It requires only the cooling curve as an input [12,15]. Newkrik et al. [16] used the Jominy end quench test and the QFA method to predict the hardness of aluminum alloys and discovered significant insight into the nature of precipitates. Dolan et al. [17] used the same technique along with finite element analysis (FEA) to predict the cooling curves and the decrease in the hardness of aluminum alloys to 65% of its maximum value.

The QFA method along with the TTP and TTT approaches can be used to evaluate the quench sensitivity of aluminum alloys based on their composition. Liu et al. [1] investigated the quench sensitivity of high strength Al-Zn-Mg-Cu alloys by changing the quench rate in the Jominy test and using Q-factor to correlate the quench sensitivity with chemical composition. Their results showed that a decrease in alloying elements such as Al, Zn, Mg and Cu and an increase in Zn/Mg ratio decrease the quench sensitivity. Xie et al. [18] assessed the quench sensitivity of low Cu-containing high strength AA7097 and compared the results with the results for AA7055 and AA7085 which contain high and medium Cu, respectively. The AA7097 showed a lower quenching sensitivity than AA7055 but higher than AA7085. The authors attributed this difference to the effect of the Cu content on the microstructure evolution and precipitation formation during the phase transformation process. The phases and precipitates were found by analyzing the vertical section of the phase diagram of 7xxx aluminum alloys. Some phases have the same size and quantity at different locations of the quenched sample [19]. However, as the distance from the quenched end of the sample increases, the cooling rate decreases which in turn results in the change in microstructures of the phases [2]. Pei-yue et al. [20] analyzed the grain boundary precipitates and grain boundary spread of quenched AA7050 and observed that they became coarser and more discontinuous as the distance from the quench end increases. Furthermore, a reduction in quench sensitivity may result in more homogeneous microstructure and properties [19]. According to Lin et al. [19], quenched-induced phases of AA7475 can be modified by slow cooling rate which results in coarse η equilibrium phases and suppressed Guinier–Preston (GP) zones. A η phase, as a quenched-induced phase, is identified by many coarse particles distributed along grain boundaries. The GP zones are identified as zones with dense spherical fine precipitates distributed within grains. The density of the η phase in the grain boundaries and in the matrix impacts the mechanical and corrosion properties. For example, an increase in the η phase leads to a decrease in corrosion resistance [19].

Corrosion behavior of quenched aluminum alloys has also been studied in the literature [21,22,23,24]. Recently, Chen et al. [22] has investigated corrosion of AA7075 quenched with air and water at different temperatures. Their results showed that a slow quench rate increased intergranular corrosion as a result of the widened solute depletion in the precipitate-free zone. Liu et al. [23] evaluated the corrosion behavior of 7A46 aluminum alloy for different isothermal holding times in a salt bath. They observed that the corrosion resistance of the quenched alloy increases initially, and then decreases with increasing isothermal holding time. The corrosion properties can be modified by pre-processing, processing, or post-processing treatments. These treatments include: applying high-temperature pre-precipitation that causes an increase in Cu content [25], changing the quench time [26], using various quench rates [25,27], changing the quench conditions that leads to different intermetallic phases, and applying pre-strain that reduces the quenched induced residual stresses and accelerates the corrosion in immersion tests [28,29,30,31]. To the best of the authors knowledge, the prediction of corrosion in the quenched samples at different locations of the samples has not been studied yet.

In this work, we use the TTP diagram [1] to obtain the quench sensitivity parameters of AA7075-T6 and predict the hardness using the QFA method. To this end, a block of AA7075-T6 is heated to 500 °C, then quenched partly in water and partly in air, separated by the horizontal dashed line in Figure 1. Temperatures at three locations in the block are recorded by K-type thermocouples. These locations are shown by TC₁, TC₂, and TC₃ in Figure 1. An FEA is performed to simulate and predict the temperature at the points where the hardness is measured. By employing the QFA method and using the simulated temperature, the spatial variation of hardness is predicted and compared with the experimental measurements. Very good agreement is observed. Cyclic polarization (CP) and Electrochemical Impedance Spectroscopy (EIS) tests are carried out to measure corrosion resistance. Finally, a new corrosion prediction model based on the QFA method is introduced.

2. Theoretical Background

The quench factor, Q, is the amount of transformation kinetics during continuous cooling and is the sum of the incremental quench factor, q, over the entire range of temperature for a particular alloy. The Q is calculated by [3,11]:

$$Q={\displaystyle \sum }q={\displaystyle \sum }_{i=1}^n\frac{\Delta t_i}{C_i\left(T\right)}={{\displaystyle \int }}_{t_0}^{t_f}\frac{1}{C\left(T\right)}dt$$

(1)

where q is calculated for each time step $\Delta t_i$ or dt in the cooling process based on critical time, C(T). The critical time is the time to achieve a constant amount of precipitation. t₀ and t_f are the start and end times of quenching, respectively. The time step is chosen as Δt = 0.1 s to account for less than 25 °C drop in the critical temperature [32]. The critical quenching rate for AA7075-T6 is about 300 K/s [1].

The C(T) for steels is extracted from the TTT curves rationalized by kinetics of isothermal homogeneous and heterogeneous precipitation formation [33]. For aluminum alloys, the TTP curves for isothermal precipitation and rapid quenching are similar to the TTT curves for steels. Mechanical properties such as hardness and yield strength along with corrosion behavior can be determined by using the TTP curves [3]. The TTP curves are called C-curves because of their characteristic shape [3,11]. The C-curves are described by:

$$C\left(T\right)=-k_1{k}_2\exp \left(\frac{k_3{k}_4^2}{RT{\left(k_4-T\right)}^2}\right)\exp \left(\frac{k_5}{RT}\right)$$

(2)

where k₁ is the fraction of untransformed strength in the natural logarithm scale assumed to be ln(0.955) [3]. k₂ is the reciprocal of the number of potential nucleation sites. k₃ corresponds to the free energy required for nucleus formation. k₄ relates to the solvus temperature. k₅ corresponds to the diffusion activation energy. R is the universal gas constant, and T is temperature in Kelvin. The parameters k₂ to k₅ are constant coefficients determined by nonlinear least-square fitting to the hardness [34]. The values of the parameters for AA7075-T6 are listed in

Table 1. k₂–k₅ parameters for TTP diagram of AA7075-T6 [1] and the gas constant. Adapted with permission from Elsevier, 2010.

$k_2$(s)	$k_3$(J/mol)	$k_4$(K)	$k_5$(J/mol)	R (J/Kmol)
$4.10\times {10}^{-13}$	1050	780	$1.40\times {10}^5$	8.3143

.

The yield strength of precipitation-hardenable aluminum alloys, such as AA7075-T6, is proportional to hardness [12], and the hardness depends on the amount of the unprecipitated fraction of hardening solutes. Therefore, the following equation can be used to predict either the yield strength, σ, or the hardness, H [12]:

$$\frac{\sigma -{\sigma }_{min}}{{\sigma }_{max}-{\sigma }_{min}}=\frac{H-H_{min}}{H_{max}-H_{min}}=\exp \left(k_{1}Q\right)$$

(3)

where ${\sigma }_{min}$ and $H_{min}$ are the minimum values of the yield strength and hardness obtained by very slow cooling rate, respectively. For simplicity, the minimum values of predicted properties, i.e., ${\sigma }_{min}$ and $H_{min}$, can be set to zero following the work of Evancho and Staley [3,11]. Thus, Equation (3) yields:

$$\frac{\sigma }{{\sigma }_{max}}=\frac{H}{H_{max}}=\exp {\left(k_{1}Q\right)}^{n/2}$$

(4)

where n = 0.92 best fits the data for AA7075-T6 in the present work. The exponent $n/2$ in Equation (4) is considered to account for nonlinearity that exists for the yield strength variation with solute concentration [11]. By having the cooling curves during the quench process, the Q can be obtained from Equation (1) and σ or H can be calculated using Equation (4).

The cooling curves are required to predict the hardness and the corrosion properties of the quenched samples. To this end, it is necessary to calculate the heat transfer coefficient for both air and water-quenched parts of the block. Analytical and empirical models are used to predict the heat transfer coefficients. Detailed heat transfer analyses are presented in the Supplementary File and all parameters used in the mathematical modeling are listed in Table S1. Figure 2 shows the predicted heat transfer coefficients for air-cooled part at TC₂ and TC₃. Their average is also shown in this figure. The heat transfer coefficients are in the range of 4 to 12 W/m²K and increase as the surface temperature increases. Figure 3a shows the predicted temperature for the water-quenched part at TC₁. The experimental temperature is also shown for comparison. A good agreement is observed. Figure 3b shows the heat transfer coefficient for the water-quenched part. This curve is obtained through analytical heat transfer analysis detailed in the Supplementary File. The heat transfer coefficient is maximum at a temperature of about 320 °C. The predicted heat transfer coefficients will later be used in the FEA to predict the full field temperature profile (see Section 4).

3. Experimental Work

3.1. Quench Test

The sample is an AA7075-T6 block with the dimensions illustrated in Figure 1. The chemical composition of the alloy is given in

Table 2. Nominal chemical compositions of AA7075-T6 (wt.%).

Al	Zn	Mg	Cu	Fe	Si	Mn	Cr	Ti
Bal.	5.6	2.5	1.6	0.5	0.4	0.3	0.23	0.2

. The sample is heated in a furnace until a steady-state temperature of 500 °C is obtained by thermocouples embedded in the block. We should note that the furnace is not supplied with inert gas; therefore, high temperature oxidation of the block is not ruled out during the heating process. The block is then removed from the furnace and is immersed partially in a large bucket of water. The initial temperature of the water is 22.5 °C. The immersion process is captured by a video camera. The average speed of immersion is calculated by using the time duration of the video from the moment the sample touches the water until the water level reaches the horizontal dashed line (“Water level”), shown in Figure 1. The average immersion speed is calculated to be 0.06 m/s. The part of the sample below the “Water level” is water-quenched and the part above this line is air-cooled. One of the thermocouples, TC₁, is in the bottom part of the sample, where the sample is quenched with water, and the other two (TC₂ and TC₃) are in the air-cooled part.

3.2. Hardness Test

The post-quenched hardness of the block is measured at 30 points, equally distanced in the y-direction, as shown in Figure 1. A Struers Rockwell hardness tester is used to measure the Rockwell B hardness (HRB). Indentation tests are performed according to ASTM E18–20 test procedure [35]. All tests are performed with a 1.588 mm ball indenter and total test force of 100 kgf. The indentation depth varies, on average, from 105 µm to 167 µm for the maximum and minimum hardness sites on the black, respectively. The tests are repeated three times on parallel lines in the y-direction and average values are calculated. The indentation sites for hardness measurements can be seen in Figure 4. We should note that at nanoscale the hardness is expected to highly depend on the indentation depth, load, and local microstructure [36,37]. However, our indentation tests are performed at mesoscale for which these dependencies can be captured.

3.3. Corrosion Test

Corrosion tests are carried out at six locations on the surface of the block. Figure 1 shows these locations, labeled as “Water 1” and “Water 2” and “Air 1” through to “Air 4”. The samples for electrochemical measurements were prepared by cutting the block into a stripe so that it can be mounted in a corrosion cell. The corrosion cell is a Gamry ParaCell which can accommodate bulky and large samples. The reference electrode is an Ag/AgCl filled with saturated KCl and the counter electrode is a graphite. The stripe is polished using silicon carbide grit papers #320, #400, and #600, and cleaned with deionized water ultrasonically for five minutes. The measurement locations are covered with mask and only an area of 1 cm² is exposed for corrosion tests (see Figure 4). A Gamry potentiostat 600+ is used to conduct the electrochemical impedance spectroscopy (EIS) and cyclic polarization (CP) tests in a 3.5% NaCl solution. Each measurement starts with an open circuit potential (OCP) test to obtain voltage stability. The EIS test is first performed followed by the CP test. In the EIS test, a small sinusoidal potential perturbation of 10 mV is applied over a frequency range of 100 kHz to 0.1 Hz to measure the current density. The impedance and the phase shift of the potential and the current density are measured [38]. Gamry EIS300 Electrochemical Impedance Software is used for the EIS data fitting. In the CP test, the sample is polarized by swiping the voltage across a voltage range with a specific scan rate. The voltage swipes from −0.5 V to 1.5 V versus open circuit potential with scan rate of 0.167 mV/s. Various corrosion properties, such as corrosion potential, E_corr, corrosion current, I_corr, and pitting potential, E_pit, can be extracted from the EIS and CP tests. The results are presented in the Results and Discussion section.

4. Finite Element Analysis (FEA)

Temperature measurements are taken at a few locations. FEA is then used to predict temporal and spatial temperature variations during the quench process. COMSOL Multiphysics 5.5 is used with material properties for AA7075-T6 [39]. A quadrilateral mapped mesh with a minimum element size of 0.369 mm and a total of 20,196 elements are used. The temperature-dependent heat transfer coefficient for both water-quenched and air-cooled parts of the block are obtained from the analyses in the Supplementary File. An element activation method is used for the water-quenched part of the block to mimic the immersion process. In this method, the water-quenched part is divided into several elements each of which being activated in sequence according to the immersion velocity of 0.06 m/s. Each element becomes activated by changing the heat transfer coefficient from air natural convection to water-quenched heat transfer. These heat transfer coefficients are functions of block surface temperature as explained in the Supplementary File. Material properties such as thermal conductivity, heat capacity, and density are assumed to be temperature dependent. A transient heat transfer analysis with implicit solver is used. The spatial temperature distribution at 5 s and 700 s after the start of the quench process are shown in Figure 5a,b, respectively. The FEA is performed for the entire duration of the quench process. The temporal variations of the temperature at locations of TC₁, TC₂, and TC₃ are shown in Figure 6. Comparisons between the recorded temperature by thermocouples and the simulated temperatures at the same locations are shown. The simulation results are in good agreement with the experimental data. A slight difference between the measured and simulated results can be attributed to the change in temperature of the water during the quenching process. The water temperature slightly increases during the test while it is assumed to be constant in the simulations.

5. Results and Discussion

5.1. Hardness

The simulation results are used to calculate the Q-factor along the length of the block in y-direction (see Figure 1). The QFA method is then used to calculate the spatial variation of the hardness along the length of the block. The hardness is experimentally measured at the exact locations for comparison. Figure 7 shows the measured hardness with error bars and the predicted hardness as a function of the distance from the leading edge. A fairly good agreement is observed. Within about 40 mm of the water-quenched part, the hardness remains high and does not change much. The hardness of the air-cooled part drops notably until 100 mm from the leading edge and then remains constant at about 49 HRB. It is to be noted that lower hardness for slower cooling rates may be related to the different microstructures that are formed as a result of slow cooling rate [40]. According to Liang and Mudawar [40], slow cooling causes coarse CuAl₂ precipitate formation along the grain boundaries. This will produce a structure which possesses low strength and low hardness [40]. Furthermore, it is known that Cu-rich particles (such as CuAl₂) promote micro-galvanic corrosion between the particle and the aluminum matrix in AA7075-T6 [41] which, in turn, results in an increase in corrosion susceptibility. Accordingly, we expect that the water-quenched part of the block exhibits better corrosion properties than the air-cooled part. This is demonstrated later by the results of the corrosion analysis.

5.2. Corrosion

The results of corrosion characterization tests are presented below followed by the introduction of a corrosion prediction model based on the QFA method.

5.2.1. EIS Results

The Nyquist and Bode plots for water-quenched and air-cooled parts are shown in Figure 8a–d, respectively. As illustrated in Figure 8b,d, a large semicircle is followed with a small semicircle or a line which indicates a mixed kinetic and charge transfer control. The small semicircles which follow the large semicircles in Water 1 and Water 2 vanish for Air 1 to Air 4.

The equivalent circuit method is used to analyze the EIS results. In this method, the corrosion cell is modeled as an electrical circuit with various arrangements of the electrical elements such as the resistor and capacitor. Each of these elements must have a physical interpretation in the corrosion cell of the EIS test.

The solution resistance, R_s, corresponds to high frequencies at the start of the Nyquist plot. The impedance at low frequencies corresponds to the charge transfer resistance, R_ct. In the range of 10 Hz to 1000 Hz, the slope of the impedance module in the Bode plot is −1 and the phase angle is about −80° which indicate capacitive behavior [42]. The diameter of the large semicircle, in Figure 8b,d, is proportional to the R_ct [42]. The large semicircle represents superposition of R_ct with at least one capacitor. The capacitor can be a double layer, C_dl, or an imperfect constant phase element, CPE, with an impedance constant and exponent denoted by Y_dl and a_dl, respectively [43]. For Water 1 and Water 2 (Figure 8a), a circle appears at low frequencies which confirms the existence of a corrosion product layer capacitor, Y_f, and a_f, and corrosion product layer resistor, R_f. Therefore, for the water-quenched part, two imperfect capacitors and resistors and for the air-cooled part only one CPE and one resistor are considered. The equivalent circuits used for the water-quenched and air-cooled parts are shown in Figure 9a,b, respectively. At high frequencies, the diffusion impedance is controlled by reaction kinetics because the diffusion of the electroactive species cannot follow the alternating potential. In this case, the electron flow favors the double layer formation pathway. At low frequencies, the mass transfer controls the diffusion associated with the charge transfer. The electrical elements of all six corrosion measurements are determined by fitting the equivalent circuits to experimental data. The values of the elements along with the Chi-squared are listed in

Table 3. Corrosion parameters of the quenched AA7075-T6 block.

Element	Water 1	Water 2	Air 1	Air 2	Air 3	Air 4
Rs (Ω)	10.12	12.29	11.30	11.76	11.88	11.72
Yf ($\times {10}^{-6}$F)	20.23	11.03	-	-	-	-
af	0.85	1	-	-	-	-
Rf (Ω)	19.13	47.48	-	-	-	-
Ydl ($\times {10}^{-6}$F)	11.1	0.33	8	2.18	5.13	1.98
adl	0.97	0.84	0.62	0.77	0.67	0.80
Rct (Ω)	1534	3030	4995	7339	11,891	12,617
${\chi }^2\left(\times {10}^{-4}\right)$	1.72	4.69	3.14	1.33	1.14	1.01

. The charge transfer resistor, R_ct, increases as the distance from the leading-edge increases.

5.2.2. CP Results

The results of CP tests for water-quenched part and air-cooled part are presented in Figure 10a,b, respectively. The potential with respect to the reference electrode is plotted versus the current density. The corrosion potential is the potential at which the current density reverses from cathodic to anodic current. The corrosion potentials of Water 1 and Water 2 are almost similar. The corrosion potential of the air-cooled part is lower than that of water-quenched part. It decreases as the distance from the leading-edge increases. We did not observe a distinct pitting potential. It coincides with the corrosion potential for all measurements. This may be due to the formation of an oxide film on the surface [44] and the pitting susceptibility of the AA7075-T6 in chloride solution [45]. The point of scan reversal or maximum potential attainment is referred to the vortex potential, E_v [44]. At E_v, the current densities reach a threshold value of 0.01 A/cm² and decrease afterward. The values of corrosion potential, the corrosion current, and the vortex potential for all measurements along with their standard deviations are presented in

Table 4. Corrosion potential and corrosion current along with their standard deviations at different locations of the quenched AA7075-T6 in 3.5% NaCl.

Elements	Water 1	Water 2	Air 1	Air 2	Air 3	Air 4
Ecorr (V) and Epit (V)	−0.722 ± 0.003	−0.728 ± 0.005	−0.768 ± 0.012	−0.783 ± 0.009	−0.793 ± 0.005	−0.792 ± 0.003
Icorr (µA)	1.911 ± 0.199	1.612 ± 0.205	2.411 ± 0.594	7.633 ± 2.013	3.620 ± 0.987	0.953 ± 0.184
Ev (V)	−0.583 ± 0.005	−0.591 ± 0.006	−0.638 ± 0.037	−0.697 ± 0.042	−0.652 ± 0.037	−0.655 ± 0.009

. Using the polarization curve, current density is determined by the Tafel extrapolation method [46,47,48]. Extrapolation is generally conducted using both cathodic and anodic branches of the polarization curve. The values of E_v are higher for the water-quenched parts than the air-cooled parts. This observation may be due to the fact that there exists a corrosion product layer for the water-quenched part when compared to the air-cooled part for which the corrosion product layer resistor was not realized from the EIS data. A negative hysteresis loop is observed for Water 1 at the intersection of the forward and reversed scans. While a positive hysteresis loop indicates the localized corrosion sites increase in size, the negative one indicates no change in size [49]. At Air 1, a step in the reversed direction is detected. This step in potential, referred to as the pit-transition potential, E_ptp, represents a complete repassivation for small pits but an initial stage for deeper pits in the reversed scan direction [50]. The E_ptp is measured to be −0.8 V at this location.

5.3. Corrosion Modeling

We developed a simple empirical model for prediction of corrosion properties (i.e., E_corr and R_ct) based on the Q-factor. Figure 11 shows the variations of E_corr and R_ct as functions of the distance from the leading-edge. Data in this figure are presented in Figure 12 in terms of the quench factor, Q. The abscissa represents $-\ln \left(-k_{1}Q\right)$ and the ordinate represents $\ln \left(-\left(1/k_1\right)\ln \left(P/P_{max}\right)\right)$, where $P$ stands for either E_corr or R_ct. The results in this figure show linear trends on the log–log scale for both E_corr and R_ct. The corrosion properties are modeled by:

$$\ln \left(-\frac{1}{k_1}\ln \frac{P}{P_{max}}\right)={\eta }_i\ln \left(-k_{1}Q\right)+{\gamma }_i$$

(5)

The form of Equation (5) is inspired by the model presented with Equation (4). This is a two-parameter empirical model with ${\eta }_i$ and ${\gamma }_i$ being the model parameters calibrated against experimental data for E_corr and R_ct.

Table 5. Values of the model parameters for E_corr and R_ct in Equation (5).

	${\mathit{\eta }}_{\mathit{i}}$	${\mathit{\gamma }}_{\mathit{i}}$
For Ecorr	−0.4939	3.1833
For Rct	−0.5435	6.1828

provides the values of the model parameters. $P_{max}$ in Equation (5) is the value of either E_corr or R_ct at the location where the Q-factor is minimum. This is the leading-edge of the block. However, the closest point of corrosion measurement on the block is 13 mm away from the leading-edge. Therefore, the values of E_corr and R_ct for the loading-edge are approximated to be −0.71 V and 1300 Ω for E_corr and R_ct, respectively. It is to be noted that the approximation is based on the trends shown in Figure 11 and the best fit of the curves in Figure 12. The strength of Equation (5) is in prediction of E_corr and R_ct based only on the Q-factor. Thus, once the history of quench process in terms of Q-factor is known, corrosion properties of a quenched part can be predicted. The present model may be used to potentially reduce the number of corrosion tests in evaluating corrosion properties of quenched AA7075-T6 parts. Further work will involve investigating the applicability of Equation (5) by expanding the number of alloys and tempers, including microstructural examinations, and quench mediums other than water.

6. Conclusions

An experimental, theoretical, and computational investigation is carried out to study the effects of quenching on the hardness and corrosion of an AA7075-T6 block. The results of this investigation are as follows:

• The hardness and the corrosion potential of the quenched block decrease as the distance from the leading-edge increases. This trend is opposite for the charge transfer resistance.
• The EIS results suggest the existence of a corrosion product layer resistor, R_f, for the water-quenched part of the sample. This was not observed for the air-cooled part.
• The values of vortex potential are higher for the water-quenched parts than the air-cooled parts. This may be due to the fact that the corrosion product layer grows thicker at higher potentials and becomes protective. This observation is in agreement with the existence of a corrosion product layer resistor for the water-quenched part when compared to the air-cooled part for which the corrosion product layer resistor was not realized from the EIS data.
• A new empirical model is developed that predicts the charge transfer resistance and corrosion potential of a quenched sample fairly well. It is based on the quench factor analysis method. The model only requires cooling curves as inputs.