The sensitization of stainless steels may decrease their corrosion resistance in industrial applications. Traditional immersion tests exist to determine the degree of sensitization (DOS) of the stainless steels. However, electrochemical methods may be preferred because they are less expensive and faster to perform. The fast and robust double loop electrochemical potentiokinetic reactivation (DL-EPR) test has been introduced to the corrosion community some decades ago but an interlaboratory testing study was necessary to assess the repeatability and reproducibility of the DOS results. This work reports on a recent study where 11 laboratories returned results that show a high degree of confidence in the data obtained by DL-EPR.
Stainless steels (SS) are a large family of iron (Fe)-based alloys containing enough chromium (Cr) to make them passive in industrial applications. SS can generally be grouped in three families of alloys: (1) ferritic/martensitic stainless steels have the body-centered cubic (bcc) structure and are alloyed mainly with Cr but also may contain small amounts of nickel (Ni), molybdenum (Mo), copper (Cu), niobium (Nb), etc.; (2) austenitic stainless steels have a face-centered cubic (fcc) structure and contain mainly Cr, Ni, and in some cases Mo; and (3) duplex stainless steels which contain both the ferritic and austenitic phases in more or less equal volume proportions.
The most common stainless steels in the industry are the so-called 300 series in the austenitic family, which includes Types 304SS and 316SS (UNS S30400(1) and 31600) (ASTM-A240).1 Typically, Type 304 contains (in mass percent) 18%Cr and 8%Ni, and Type 316 contains 16%Cr, 10%Ni, and 2%Mo.
The corrosion behavior of austenitic stainless alloys is not only a function of the overall chemical composition of the alloying elements (i.e., mass percent) but also of the way these alloying elements are distributed in the alloy. The corrosion resistance of the stainless steel may be different if all of the alloying elements are evenly distributed as a solid solution in the alloy or if these elements are enriched in certain areas due to the formation of second phase particles. After exposure to temperatures in the order of 600°C for times as short as 1 h or less (for example during welding), some stainless steels would develop areas that are depleted in chromium (Cr) due to the formation of secondary carbides. These Cr-depleted areas (generally adjacent to grain boundaries) may have a lower corrosion resistance than their neighboring areas, leading to localized corrosion issues of the SS such as intergranular attack (IGA) and increased susceptibility to stress corrosion cracking (SCC).
The metals industry and trade organizations developed standard testing procedures to test for the metallurgical state or condition of SS to determine the degree of sensitization (DOS); i.e., how important is the Cr depletion at near grain boundaries. Common immersion tests for determining sensitization include ASTM A2622 and G28.3 Faster and less destructive electrochemical tests have also been created to assess for DOS in stainless alloys such as in the electrochemical potentiokinetic reactivation (EPR) method ASTM G108.4
The ASTM Standard G108 EPR method was developed in 1994 and it contained what it is now called the single loop (SL) test.4 In 2017, the ASTM subcommittee G01.11 Electrochemical Measurements in Corrosion Testing authorized the formation of a working group to explore the suitability of incorporating into ASTM Standard G108 the double loop (DL) method initially developed by Číhal5 and later explored by Majidi and Streicher6 and others.7-10 The double loop method is currently described in ISO 12732.11
Guidelines from the DL method have been used extensively for martensitic and duplex stainless steels12-14 and for nickel alloys containing Cr such as UNS N06600 and N06690 in narrow ranges of electrolytes.15-16
The working group within the ASTM G01.11 subcommittee developed an interlaboratory testing plan to evaluate the repeatability and reproducibility of the DL test method. The objective of the current manuscript is to report the results from the interlaboratory testing program.
INTERLABORATORY TESTING PROGRAM
Based on the literature data (mainly from Majidi and Streicher6 ), enough specimens of certified Type 304H SS ASTM A2401 (UNS S30409) were purchased to be tested in the program. Type 304H was selected because it contains higher amount of carbon than the Type 304L (UNS S30403) material. Table 1 shows the chemical composition and heat of the used material. Coupons were prepared and thermally aged side by side in one laboratory and then distributed to the participating laboratories for the DL testing. These testing laboratories included industry, government agencies, national laboratories, and universities.
Two Steps in the Interlaboratory Testing Program
The interlaboratory testing program consisted of two steps. The objective of Step 1 was to narrow down the appropriate heat treatment or sensitizing conditions (time and temperature) for the thermal aging of the Type 304H SS (Table 1) that will be used in Step 2. Three laboratories participated in Step 1. In Step 2, the most relevant aging condition for the Type 304H SS selected after analyzing the results from Step 1 was tested by 12 participating laboratories. For both steps, coupons and specimens of Type 304H SS (Table 1) were thermally aged side by side in the same location and then distributed to the participating laboratories.
Step 1: Selecting the Most Relevant Time and Temperature for Sensitizing Type 304H SS
Table 2 shows the thermal aging matrix used to prepare the test specimens for the Step 1 study. Specimens of as-received (AR) Type 304H SS were solution annealed (SA) and sensitized (Sen) at four times and temperatures following the matrix used by Majidi and Streicher.6 The time in Table 2 refers to the time at temperature in the furnace in a natural air environment. The temperature was controlled with a precision of ±3°C.
Three laboratories participated in the Step 1 study, performing electrochemical testing for all of the thermally-treated conditions listed in Table 2.
Testing Instructions for Step 1 Study Laboratories
The specimen preparation and electrochemical testing instructions for the three laboratories in Step 1 were as following.
Specimen preparation: Grind the surfaces of the specimens with silicon carbide paper 320 grit to expose a fresh surface.
Solution Preparation: To perform the electrochemical testing, use the electrolyte 0.5 M H2SO4 + 0.01 M KSCN, naturally aerated, at ambient temperature. To prepare 1 L of this solution, dissolve 28 cm3 of concentrated sulfuric (∼96%) acid and 0.97 g of KSCN salt with deionized (DI) water17 to a final volume of 1 L.
Procedure: Immerse the Type 304H SS test specimen (working electrode) into the test solution and monitor the corrosion potential (using for example a saturated calomel electrode [SCE]) for 30 min. Record the value of Ecorr. The corrosion potential value should be in the vicinity of –400 mVSCE. Perform a cyclic potentiodynamic test (such as described in ASTM G61) starting at the corrosion potential (e.g., –400 mVSCE) at a forward anodic scan rate of 1.67 mV/s. When the potential reaches +300 mVSCE, reverse the direction of the potential scan also at 1.67 mV/s and end the scan at the original Ecorr. Note: the peak current ia in the forward scan could be in the order of 35 mA/cm2, therefore the potentiostat should be properly set up (e.g., to be able to read a maximum current density of 100 mA/cm2).
Results: Record the values of ia and ir and calculate the ratio ir/ia.
Testing Results from Step 1 Interlaboratory Testing
Figures 1 and 2 show the corrosion potential evolution as a function of the immersion time for up to 2 h for Type 304H SS in the six metallurgical conditions in Table 2. Figure 1 includes the potential for a reference platinum electrode and Figure 2 shows only the values for the Type 304H SS working electrodes. Figure 1 shows that the Ecorr of Pt was approximately +400 mVSCE, while the Ecorr of the six Type 304H SS electrodes were approximately –400 mVSCE. Figure 2 shows that the Ecorr for the six Type 304H SS electrodes gradually increased for immersion time of approximately 2 h. The maximum difference in final Ecorr among all six electrodes was approximately 15 mV at the end of the exposure time.
Figure 3 shows the DL-EPR tests to Type 304H SS in the AR condition (blue line) and in the thermally-aged condition of 677°C for 2 h (Sen3 in Table 2) (red line). Both electrodes had practically the same initial corrosion potential (Ecorr). As the polarization increased in the anodic direction, both electrodes showed a characteristic anodic current peak (ia) and then the electrodes passivated in the vicinity of −0.1 VSCE and reached the preset passive potential of +0.3 VSCE. In the reverse scan, the AR electrode reached the corrosion potential in the passivated condition but the Sen3 electrode showed a reactivation anodic current peak (ir). These tests show that the AR electrode was not sensitized (no peak in the reverse scan), while the Sen3 electrode was sensitized. The value of the ratio ir/ia is the DOS. The higher the DOS, the more sensitized the material. In cases where the tested specimen does not show an ir peak, the ir peak is taken as the current value in the reverse scan at the potential of the ia peak.
The three laboratories that participated in the Step 1 study reported the Ecorr and DOS ratios for the six types of electrodes shown in Table 2. Table 3 shows the results from the Step 1 interlaboratory study.
Table 3 shows highly reproducible results from the three laboratories involved in the Step 1 study. For example, the average corrosion potential (Ecorr) reported by the three laboratories for the six metallurgical conditions of Type 304H SS (Table 2) was –0.418 VSCE with a standard deviation (SD) of 14 mV. Figure 4 shows in graphical form the values of DOS in Table 3 from the three laboratories for the six metallurgical conditions. It is apparent that the tested Type 304H SS in the AR, SA, and Sen1 (550°C for 8 h) did not show sensitization because the value of DOS < 0.05.6,11 However, for the Type 304H SS thermally treated at 677°C and 732°C, the value of DOS was higher than 0.05 and therefore sensitization was expected.
Figure 5 shows the appearance of the Type 304H SS specimens after the DL-EPR tests for the six metallographic conditions in Table 2. It is apparent from Figure 5 that the AR, SA, and Sen1 specimens did not show grain boundary attack, a sign of the absence of sensitization (see Figure 4). The other three specimens showed clear grain boundary attack, suggesting the presence of sensitization confirmed by the DOS values in Figure 4.
Based on the results from the Step 1 study, the Sen3 sensitization condition (677°C 2 h) was used to prepare specimens to distribute to the 12 laboratories that agreed to participate in the Step 2 study.
Step 2 Interlaboratory Testing Study
Initially 12 laboratories agreed to participate in the interlaboratory study to determine repeatability and reproducibility of the DL-EPR test. For the Step 2 study, three Type 304H SS specimens were distributed to each laboratory. One specimen was in the AR condition and two specimens were in the Sen3 (677°C for 2 h) condition. The same testing guidelines used in the Step 1 study were given to the Step 2 participating laboratories. It was requested for the 12 laboratories to report the values of Ecorr and the DOS (ir/ia) ratio. At the end of the study 11 laboratories returned results.
Table 4 shows the results returned by 11 laboratories for the AR (control) specimens.
The data for the AR specimens are provided to show that these specimens were not sensitized, that is, the ir values were significantly lower than the ia values; therefore, the coefficient of sensitization ir/ia was also low, or below 0.05, showing no sensitization.11 The values of Ecorr show that the specimens were actively corroding during the open-circuit potential period, and they are included here for information purposes, to have another control parameter that would inform that the 11 laboratories had a similar Ecorr for their AR specimens. The average and standard deviation values of Ecorr show high reproducibility among the 11 laboratories (–429±31 mVSCE).
Table 5 shows the results from the 11 laboratories for the Sen3 condition. Material called HT1 is for the first Sen3 specimen and HT2 for the second Sen3 specimen for each laboratory. If the letter R appears, it means that after one test, the specimen was resurfaced to conduct another test.
The values of Ecorr are for information purposes, to have a control parameter that supports the similar starting condition for the DL-EPR for the 11 laboratories. The average Ecorr and standard deviation showed agreement among the 11 laboratories (Ecorr = –427±32 mVSCE). The values of ir/ia reported in Table 5 are higher than 0.05 showing sensitization of the tested specimens.11
Statistical Data Summary for Step 2 Interlaboratory Testing Study
The results from the Step 2 study have been analyzed according to the procedure shown in ASTM E691.18 The results of these analyses are shown in Tables 6 and 7 for the HT1 and HT2 specimens of sensitized (Sen3) Type 304H stainless steel.
In Tables 6 and 7 the scell values are the standard deviations of the replicate values obtained at the laboratories in question for the same material by the same operators using the same equipment. The “d” values are the differences between the Avg. values and the average of all of the Avg. values in the table. The “h” values are the ratios of the “d” values to the standard deviation of the Avg. values in the table. These “h” values are also known as the “between-laboratory consistency statistics.” The “k” values are the ratios of the scell values to the standard deviation of the Avg. values, i.e., savg, for the table. These values are also known as the “within-laboratory consistency statistics.” The sr value is calculated by pooling all of the scell values, and the sR value is calculated by pooling the sr value with the savg value.
The critical values for the consistency parameters determined at the 0.5% significance level are 2.34 for the h parameter for 11 laboratories’ participation, and 2.30 for the k parameter in Table 6 (with 7 laboratories), and 1.95 for Table 7 (with 4 laboratories). These critical values are available in ASTM E69118 for a number of participants and replicates.
The process of determining precision information for ASTM test methods is required because these tests are called out in specification for products to meet customer demands for product quality. The precision values are necessary to show whether a test result meets the specified value. Product rejection can result when the test result is outside of the range determined by the reproducibility precision around the specified value. Tests such as the EPR test are known as performance tests, and this type of test is preferred by many customers because it measures a parameter that determines product performance in a situation that simulates real world environments.
The ASTM E69118 practice provides an unbiased evaluation of the test method precision involving a number of independent laboratories. The repeatability value is an important measure of the variability of the method while holding operator, specimen, and equipment variables constant. It represents the best result that can be obtained. It is particularly valuable to determine if an operator is able to successfully perform the procedure. The reproducibility value is valuable in determining how much variation can be expected when a number of laboratories perform the same test. The ASTM E69118 practice uses normal (Gaussian) probability statistics.
ASTM also requires a bias statement for test methods. Bias is a systematic error that is related to the actual procedure used in the method. In this case, bias would relate to the comparison of this method with other methods used to determine sensitization of Type 304H SS. Bias is not covered in ASTM E691.18
The consistency parameters determined by the ASTM E691 practice are intended to determine if the variations found in the interlaboratory program are within ranges expected for a normal distribution. When laboratories develop results that exceed the critical levels of these parameters, it may indicate that there is problem either with the laboratory’s method or a problem with the write-up that leads to variations in the results. It is considered good practice to investigate situations where consistency outlier results are found.
The results of these analyses show that the average ir/ia ratios of the two specimens were very close, and the standard deviations were also very close. A Student’s t-test of the difference between the two averages showed a value of 0.59, which is not significant. Similarly, F-tests of the variances used to calculate the standard deviations were 1.11 for the repeatability data and 1.02 for the reproducibility data. These are not significantly different. Therefore, the pooled variances may be used to calculate standard deviations. This procedure yields a better estimate of the standard deviations.
The consistency analysis showed that three laboratories, 3, 4, and 10, had results that were out of line with the others in the program. Laboratories 3 and 10 obtained ratios that were significantly below the others for the HT1 material, and laboratory 10 also obtained a significantly low ratio for the HT2 material. Laboratory 4 obtained a ratio for the HT2 material that was significantly higher than the others. None of the laboratories that performed duplicate runs obtained results that were out of line in terms of repeatability statistics.
Earlier results using the SL method demonstrated that the method was not applicable to more corrosion-resistant stainless steel alloys, such as Types 316L or 317L (UNS S31703). As a result, these alloys were not included in the evaluation of the DL method. The DL method produces a quantitative value showing the DOS ranging from not sensitized to completely sensitized. This method is easily performed and is relatively rapid. The solution does not have to be deaerated nor is it necessary to perform careful preparation of the specimen surface before the test is done. Also, the surface area of the specimen does not have to be measured accurately. These are advantages over the SL method. However, it is necessary to perform the test using electrochemical glassware and components including a reference electrode and an auxiliary electrode.
The results of the interlaboratory test study have been incorporated in the ASTM Standard Test Method G108 and as such this method may now be included in specifications for products made from Type 304SS.
SUMMARY AND CONCLUSIONS
A comprehensive and lengthy interlaboratory testing study was conducted to determine reproducibility and repeatability of the double loop electrochemical potentiokinetic reactivation (DL-EPR) test to determine the degree of sensitization (DOS) in Type 304H SS.
Eleven laboratories returned results that showed that the DL-EPR test method is rather simple and robust, showing highly reproducible and repeatable DOS results.
UNS numbers are listed in Metals and Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International.