- Email: [email protected]
Developing an empirical relationship to predict the corrosion rate of friction stir welded AZ61A magnesium alloy under salt fog environment
Materials and Design 32 (2011) 5066–5072
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/matdes
Technical Report
Developing an empirical relationship to predict the corrosion rate of friction stir welded AZ61A magnesium alloy under salt fog environment A. Dhanapal a,⇑, S. Rajendra Boopathy b,1, V. Balasubramanian c,2 a
Department of Mechanical Engineering, Sri Ramanujar Engineering College, Vandalur, Chennai 600 048, Tamil Nadu, India Department of Mechanical Engineering, College of Engineering, Guindy, Anna University, Chennai 600 025, Tamil Nadu, India c Center for Materials Joining & Research (CEMAJOR), Department of Manufacturing Engineering, Annamalai University, Annamalainagar 608 002, Chidambaram, Tamil Nadu, India b
a r t i c l e
i n f o
Article history: Received 26 April 2011 Accepted 20 June 2011 Available online 25 June 2011
a b s t r a c t Magnesium (Mg) alloys shows the lowest density among other engineering metallic materials. As a consequence, this light alloy has a promising future. However, these alloys have great affinity for oxygen and other chemical oxidizing agents. The limitation of low corrosion resistance restricts their practical applications. Extruded Mg alloy plates of 6 mm thick of AZ61A grade were butt welded using friction stir welding (FSW) process. Corrosion behavior of the welds was evaluated by conducting salt fog test in NaCl solution at different chloride ion concentrations, pH value and spraying time. Also an attempt was made to develop an empirical relationship to predict the corrosion rate of friction stir welded AZ61A magnesium alloy. Three factors and a central composite design were used to minimize the number of experimental conditions. Response surface method was used to develop their relationship. The developed relationship can be effectively used to predict the corrosion rate of friction stir weld AZ61A magnesium alloy at 95% confidence level. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Magnesium (Mg) alloys are one of those light weight metallic alloys currently being investigated, because of its low density (1.74 g/cm3) and high mechanical stiffness [1]. The mechanical benefits of magnesium, however, are contrasted by a high corrosion rate as compared to aluminum or steel. Because of magnesium’s electrochemical potential, as illustrated in the galvanic series, it corrodes easily in the presence of sea water. The high corrosion property of magnesium has relegated the alloy to use in areas unexposed to the atmosphere, including car seat and electronic boxes. The demand for the use of magnesium alloys as structural materials in automobile industry, electronic products, vibrating plates of vibrating test machines, automotive wheels, etc., have increased in recent years [2]. The use of magnesium alloys in automotives reduces the total weight of the component up to 20% and at the same time this will substantially reduce the fuel consumption and the CO2 emissions. Conventional fusion welding of Mg alloys resulted in much solidification related problems such as porosity, hot cracking, alloy
⇑ Corresponding author. Tel.: +91 9444859455 (Mobile), +91 44 22751393 (R); fax: +91 44 22751370. E-mail addresses: [email protected] (A. Dhanapal), boopathy@ annauniv.edu (S.R. Boopathy), [email protected] (V. Balasubramanian). 1 Fax: +91 44 22357744. 2 Fax: +91 4144 238080/238275. 0261-3069/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.matdes.2011.06.038
segregation and partial melting zone. Hence, the applications of Mg alloys in structural members were still limited. However, a recent innovation of friction stir welding (FSW) process eliminated the above said problems. FSW is a solid state, autogenously process and hence there was no melting and solidification. Though the mechanical properties and micro structural characteristics of FSW joints of Mg alloys are the topic of many researchers, the corrosion properties of these joints have not yet been fully explored. In a buffer chloride solution, the corrosion rate of magnesium and its alloys did not depend on their purity or the content of the major alloying, but solely on the pH of the solution [3]. A serious limitation for the potential use of several magnesium alloys is their susceptibility to corrosion. Magnesium alloys, especially those with high purity, have good resistance to atmospheric corrosion. However, the susceptibility to corrosion in chloride containing environments is a serious concern [4]. The thickness of the oxide film formed on the surface of the specimen is quite increased with the increase in pH. The corrosion rate of AZ91 is high in acidic solution as compared to that in neutral and highly alkaline solutions. In extruded magnesium alloy the pitting corrosion is more at lower pH [5]. Alloying magnesium with aluminum in general improves the corrosion resistance [6]. In the initial stages of the corrosion attack, the cathodic reactions could involve both reductions of oxygen and water. The oxygen reduction occurred preferentially in the periphery of the electrolyte droplets on the surface, which led to alkaline conditions and an increase in CO2 deposition in these areas. After a long exposure time, a thickened layer of corrosion products were
5067
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072 Table 1a Chemical composition (wt.%) of AZ61A Mg alloy. Al
Zn
Mn
Mg
5.45
1.26
0.17
Balance
Table 1b Mechanical properties of AZ61A Mg alloy. Yield strength (MPa)
Ultimate tensile strength (MPa)
Elongation (%)
Vickers hardness at 0.05 kg load (Hv)
177
272
8.40
57
99
NORMAL
formed on the surface. Due to localized corrosion attack brucite was formed on those areas [7]. The galvanic couples formed by the second phase particles and the matrix are the main source of localized corrosion of magnesium alloys [8]. The general and pitting corrosion behavior of parent and FSW nugget regions were nearly the same, even though they were different from the untreated condition [9]. The overall corrosion processes can be described by three stages. The first stage corresponds to the pit nucleation and growth; the second stage involves the growth of a passive oxide layer; and the third stage involves the reactivation [10]. In wet air and aqueous solutions, oxide and hydroxide layers of Mg,MgO/Mg(OH)2, were formed spontaneously on the surface. Mg(OH)2 had a Pilling–Bedworth ratio of 1.77, and could provide enough corrosion protection to the metallic surface [11]. Corrosion attack of Mg, AZ31, AZ80 and AZ91D materials under the salt fog test increased with increasing temperature and Cl concentration [12]. Individual pitting characteristics, including pit surface area and pit volume, were greater for the salt spray surfaces [13]. From the literature review, it was understood that the most of the published information on corrosion behavior of Mg alloys was focused on pitting corrosion and general corrosion of unwelded base metals. Very few investigations [14,15] had evaluated the corrosion behavior on FSW joints of Mg alloys. Hence, the present investigation was carried out to develop an empirical relationship to predict the corrosion rate on FSW joints of AZ61A Mg alloy under salt fog conditions.
95 90 80 70 50 30 20 10 5 1
2. Experimental work 2.1. Fabricating the joints and preparing the specimens The material used in this study was an AZ61A magnesium alloy in the form of extruded plates of 6 mm thickness. The chemical composition and mechanical properties of the base metal are presented in Tables 1a and 1b. The optical micrograph of friction stir welded stir zone is shown in Fig. 1. The plate was cut to the required size (300 mm 150 mm) by power hacksaw followed by milling. A square butt joint configuration was prepared to fabricate friction stir welded (FSW) joints. The initial joint configuration obtained by securing the plates in position using mechanical clamps. The direction of welding was normal to the extruded direction. Single pass welding procedure was followed to fabricate the joints. Non-consumable tool made of high carbon steel was used to fabricate the joints. An indigenously designed and developed computer numerical controlled friction stir welding (22 kW; 4000 rpm; 60 kN) was used to fabricate the joints. The FSW parameters were optimized by conducting a lot of trial runs. The welding conditions which produced defect free joints were taken as optimized welding conditions. The optimized welding parameters are presented in Table 2. From the welded joints, the corrosion test specimens were sliced to the dimensions of 50 mm 16 mm 6 mm. The specimens were grounded with 500#, 800#, 1200#, 1500# grit SiC paper. Finally, it was cleaned with acetone and washed in distilled water then dried by warm flowing air.
-2.00
-1.00
-0.001
0.99
1.99
STUDENTIZED RESIDUAL Fig. 1. Normal probability plot.
1. If the pH value of the solution was less than 3 – the change in chloride ion concentration did not considerably affect the corrosion. 2. If the pH value was in between 3 and 11 – there was inhibition of the corrosion process and stabilization of the protective layer 3. If the pH value was greater than 11 – blocking of further corrosion by the active centers of protective layer. 4. If the chloride ion concentration was less than 0.2 M – the visible corrosion did not occur in the experimental period. 5. If the chloride ion concentration was in between 0.2 M and 1 Mn – there was a reasonable fluctuation in the corrosion rate. 6. If the chloride ion concentration was greater than 1 M – the rise in corrosion rate may hesitate and decrease a little. 7. If the spraying time was less than an hour – the surface was completely covered with thick and rough corrosion products. 8. If the spraying time was in between 1 and 9 h – the tracks of the corrosion can be predicted. 9. If the spraying time was greater than 9 h, the tracks of corrosion film were difficult to identify. 2.3. Developing the experimental design matrix
2.2. Finding the limits of corrosion test parameters From the literature, the predominant factors that have a greater influence on the corrosion rate of AZ61A FSW joints had been identified. They were: (i) pH value of the solution, (ii) spraying time and (iii) chloride ion concentration. Large numbers of trial experiments were conducted to identify the feasible testing conditions using friction stir welded AZ61A magnesium alloy joints under salt fog conditions. The following inferences were obtained:
Owing to a wide range of factors, the use of three factors and a central composite rotatable design matrix were chosen to minimize the number of experiments. A design matrix consisting of 20 sets of coded conditions (comprising a full replication three factorial of 8 points, six corner points and six center points) was chosen in this investigation. Table 3 represents the range of factors considered, and Table 4 shows the 20 sets of coded and actual values used to conduct the experiments.
5068
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072 Table 2 Optimized welding conditions and process parameters used to fabricating the joints. Rotational speed (rpm)
Welding speed (mm/min)
Axial force (kN)
Tool shoulder diameter, (mm)
Pin diameter, (mm)
Pin length, (mm)
Pin profile
1000
75
3
18
6
5
Left hand thread of 1 mm pitch
Table 3 Important factors and their levels. S. No
Factor
1 2 3
Unit
pH value Spraying time Cl concentration
Hours (h) Mole (M)
Notation
Levels
P T C
1.682
1
0
+1
+1.682
3 1 0.2
4.62 2.62 0.36
7 5 0.6
9.38 7.38 0.84
11 9 1
Table 4 Design matrix and Experimental results. Ex. No
Coded values
Actual values
pH (P)
Time (T)
Conc. (C)
pH
Time (T) (h)
Conc. (C) (M)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 +1 1 +1 1 +1 1 +1 1.682 +1.682 0 0 0 0 0 0 0 0 0 0
1 1 +1 +1 1 -1 +1 +1 0 0 1.682 +1.682 0 0 0 0 0 0 0 0
1 1 1 1 +1 +1 +1 +1 0 0 0 0 1.682 +1.682 0 0 0 0 0 0
4.62 9.38 4.62 9.38 4.62 9.38 4.62 9.38 3 11 7 7 7 7 7 7 7 7 7 7
2.62 2.62 7.38 7.38 2.62 2.62 7.38 7.38 5 5 1 9 5 5 5 5 5 5 5 5
0.36 0.36 0.36 0.36 0.84 0.84 0.84 0.84 0.60 0.60 0.60 0.60 0.20 1 0.60 0.60 0.60 0.60 0.60 0.60
For the convenience of recording and processing experimental data, the upper and lower levels of the factors were coded here as +1.682 and 1.682 respectively. The coded values of any intermediate value could be calculated using the following relationship.
X i ¼ 1:682½2X ðX max X min Þ=ðX max X min Þ
ð1Þ
where Xi is the required coded value of a variable X and X is any value of the variable from Xmin to Xmax, Xmin is the lower level of the variable, Xmax is the upper level of the variable. 2.4. Recording the responses Solution of NaCl with concentrations of 0.2 M, 0.361 M, 0.6 M, 0.838 M, and 1 M were prepared. The pH value of the solution was maintained as pH 3, pH 4.619, pH 7, pH 9.38, & pH 11 with concentrated HCl and NaOH respectively. The pH value was measured using a digital pH meter. The test method consists of exposing the material to be tested in a salt spray chamber as per ASTM B 117 standards and evaluating the corrosion tested specimen with the method as per ASTM G1-03 [16,17]. Basically, the salt spray test procedure involves the spraying of a salt solution onto the samples being tested. This was done inside a temperature-controlled chamber. The glass racks were contained in the salt fog chamber (30 high, 30 deep and 50 wide). The samples under test
Corrosion rate (mm/year)
14.62 10.23 11.89 8.99 15.82 11.31 12.92 10.75 17.96 10.88 11.23 8.51 11.66 15.28 9.89 8.56 9.79 9.97 8.93 9.62
were inserted into the chamber, following which the salt-containing solution was sprayed as a very fine fog mist over the samples. NaCl in tapped water was pumped from a reservoir to spray nozzles. The solution was mixed with humidified compressed air at the nozzle and this compressed air, atomized the NaCl solution into a fog at the nozzle. Heaters were maintained at 35 °C cabinet temperature. Within the chamber, the samples were rotated frequently so that all samples were exposed uniformly to the salt spray mist. The temperature within the chamber was maintained at a constant level. Since the spray was continuous, the samples were continuously wet, and therefore, uniformly subjected to corrosion. The corrosion rate of the friction stir welded AZ61A alloy specimen was estimated by weight loss measurement. The original weight (wo) of the specimen was recorded and then the specimen was sprayed with the solution of NaCl for different spraying times of 1, 2.619, 5, 7.38 and 9 h. The corrosion products were removed by immersing the specimens for one minute in a solution prepared by using 50 g chromium trioxide (CrO3), 2.5 g silver nitrate (AgNO3) and 5 g barium nitrate (Ba(NO3)2) for 250 ml distilled water. Finally, the specimens were washed with distilled water, dried and weighed again to obtain the final weight (w1). The weight loss (w) can be measured using the following relation,
w ¼ ðwo w1 Þ
ð2Þ
5069
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
where w is the weight loss in grams, wo the original weight before test in grams and w1 is the final weight after test in grams. The corrosion rate of FSW joints of AZ61A was calculated by using the following equation:
Corrosion rate ðmm=yearÞ ¼
8:76 104 w ADT
ð3Þ
where w is the weight loss in grams, A the surface area of the specimen in cm2, D the density of the material, 1.72 g/cm3, T is the spraying time in hours. Micro structural examination of the corroded specimens was carried out using a light optical microscope (VERSAMET-3) incorporated with an image analysing software (Clemex-vision). The exposed samples surface was prepared for the micro examination both in the ‘‘AS polished’’ and ‘‘AS etched’’ conditions. Picral + Acetic acid were used as etchant. The corrosion test specimens were polished in disc polishing machine for scratch fewer surfaces and the surface was observed at 200 magnification. 3. Developing an empirical relationship In the present investigation, to correlate the salt spray test parameters and the corrosion rate of welds, a second order quadratic model was developed. The response (corrosion rate) is a function of pH value (P), spraying time (T) and chloride ion concentration (C) and it could be expressed as,
Corrosion rate ðCRÞ ¼ f ðP; T; CÞ
ð4Þ
The empirical relationship must include the main and interaction effects of all factors and hence the selected polynomial is expressed as follows:
Y ¼ bo þ
X
b i xi þ
X
bii x2i þ
X
bij xi xj
ð5Þ
For three factors, the selected polynomial could be expressed as
Corrosion rate ðCRÞ ¼ b0 þ b1 ðPÞ þ b2 ðTÞ þ b3 ðCÞ þ b11 ðP2 Þ þ b22 ðT 2 Þ þ b33 ðC 2 Þ þ b12 ðPTÞ þ b13 ðPCÞ þ b23 ðTCÞ
Corrosion rate ðCRÞ ¼ 9:48 1:89P 0:88T þ 0:82C þ 1:60P 2 þ 1:27C 2
ð8Þ
The Analysis of Variance (ANOVA) technique was used to find the significant main and interaction factors. The results of second order response surface model fitting in the form of Analysis of Variance (ANOVA) are given in Table 5. The determination coefficient (r2) indicated the goodness of fit for the model. The Model F-value of 31.30 implies the model is significant. There is only a 0.01% chance that a ‘‘Model F-Value’’ this large could occur due to noise. Values of ‘‘Prob > F’’ less than 0.0500 indicates model terms are significant. In this case P, T, C, P2, C2 are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve the model. The ‘‘Lack of Fit F-value’’ of 1.69 implies the Lack of Fit is not significant relative to the pure error. There is a 28.93% chance that a ‘‘Lack of Fit F-value’’ this large could occur due to noise. Non-significant Lack of Fit is good. The ‘‘Pred R-Squared’’ of 0.8176 is in reasonable agreement with the ‘‘Adj R-Squared’’ of 0.9349. ‘‘Adeq Precision’’ measures the signal to noise ratio. P ratio greater than 4 is desirable. Our ratio of 19.440 indicates an adequate signal. The normal probability of the corrosion rate shown in Fig. 1 reveals the residuals were falling on the straight line, which meant that the errors were distributed normally. All of this indicated an excellent suitability of the regression model. Each of the observed values compared with the experimental values shown in Fig. 2.
ð6Þ
where b0 is the average of responses (corrosion rate) and b1, b2, b3, . . . b11, b12, b13, . . . b22, b23, b33, are the coefficients that depend on their respective main and interaction factors, which were calculated using the expression given below,
Bi ¼
number of combination considered. All the coefficients were obtained applying central composite face centered design using the Design Expert statistical software package. After determining the significant coefficients (at 95% confidence level), the final relationship was developed using only these coefficients. The final empirical relationship obtained by the above procedure to estimate the corrosion rate of friction stir welds of AZ61A magnesium alloy is given below,
X ðX i ; Y i Þ=n
ð7Þ
where ‘i’ varies from 1 to n, in which Xi is the corresponding coded value of a factor and Yi is the corresponding response output value (corrosion rate) obtained from the experiment and ‘n’ is the total
4. Discussion Fig. 3 shows the micrograph of stir zone of friction stir welded AZ61A magnesium alloy before corrosion tests. Fig. 4 shows the scanned images of the corrosion test specimen after salt fog tests at different pH value, chloride ion concentration and spraying time. It was seen that, at lower concentration, the surface of the specimen was relatively slightly corroded in neutral and alkaline solutions while severely corroded at all pH values at higher chloride ion concentrations [18].
Table 5 ANOVA test results. Source
Sum of squares
df
Mean square
F value
p-value Prob > F
Model P T C PT PC TC P2 T2 C2 Residual Lack of Fit Pure error Cor total
126.60 49.03 10.55 9.12 1.83 0.047 0.033 37.07 3.404E-004 23.17 4.49 2.82 1.67 131.09
9 1 1 1 1 1 1 1 1 1 10 5 5 19
14.07 49.03 10.55 9.12 1.83 0.047 0.033 37.07 3.404E-004 23.17 0.45 0.53 0.33
31.30 109.11 23.48 20.29 4.08 0.10 0.072 82.48 7.575E-004 51.55
<0.0001 <0.0001 0.0007 0.0011 0.0710 0.7543 0.7934 <0.0001 0.9786 <0.0001
Significant
1.69
0.2893
Not significant
5070
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
(a) Experiment No. 1
(b) Experiment No. 2
(c) Experiment No. 3
(d) Experiment No. 4
(e) Experiment No. 5
(f) Experiment No. 6
(g) Experiment No. 7
(h) Experiment No. 8
Fig. 2. Correlation graph for response (corrosion rate).
Fig. 3. Microstructure of friction stir welded stir zone before corrosion test.
From Table 4, it shows the corrosion rate obtained from salt fog tests at different pH value, chloride ion concentration and spraying time. At every chloride ion concentration and spraying time, the joints usually exhibited a decrease in corrosion rate with the increase in pH. The highest corrosion rate was observed at pH 3 and at neutral pH, the corrosion rate was remained constant approximately and comparatively low corrosion rate was observed in alkaline solution. It was seen that the influence of pH was more at higher concentration as compared to lower concentration in neutral and alkaline solutions. Fig. 5 shows the microstructure of FSW joints after salt fog tests at different pH value, chloride ion concentration and spraying time respectively, it was observed that the matrix shows the pitting marks and the pitting corrosion that has taken place at the friction stir welded microstructure. The particles are Mn–Al compound and fragmented Mg17Al12. The pits were marked in the microstructure with arrows. The numbers of pits were more in the joints when it is sprayed with the solution of low pH. Hence the corrosion rate increases with the decrease in pH value. Since the increase of grain and grain boundary in the joints, the grain boundary acts cathodic to grain causing micro galvanic effect. Corrosion tends to be concentrated in the area adjacent to the grain boundary until eventually the grain may be undercut and fall out [19]. It meant that the pH value was one of the major factors of corrosion rate. At lower pH values, the specimen exhibited a rise in corrosion rate with an increase in chloride ion concentration. But the quantity of this rise was different in such a way that, the change in chloride ion concentration at lower concentrations affected the corrosion rate much more as compared to that of higher concentra-
(i) Experiment No. 9
(j) Experiment No. 10
(k) Experiment No.11
(l) Experiment No. 12
(m) Experiment No. 13
(n) Experiment No. 14
(o) Experiment No. 15
(p) Experiment No. 16
Fig. 4. Scanned image of the corrosion test specimens.
tion. It showed that with the increase in chloride ion concentration, the rising rate at corrosion rate decreased that is, the influence of
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
(a) Experiment No. 1
(b) Experiment No. 2
(c) Experiment No. 3
(d) Experiment No. 4
(e) Experiment No. 5
(f) Experiment No. 6
(g) Experiment No. 7
(h) Experiment No. 8
(i) Experiment No. 9
(j) Experiment No. 10
(k) Experiment No. 11
(l) Experiment No. 12
5071
From the microstructure of FSW joints after salt fog tests at different pH value, chloride ion concentration and spraying time, it showed that the alloy exhibited a rise in corrosion rate with the increase in Cl concentration and thus the change of Cl concentration affected the corrosion rate much more in higher concentration solutions than that in lower concentration solutions. When more Cl in NaCl solution promoted the corrosion, the corrosive intermediate (Cl) would be rapidly transferred through the outer layer and reached the substrate of the alloy surface [19]. Hence, the corrosion rate was increased. This corrosion behavior is consistent with the current understanding that the corrosion behavior of magnesium alloys is governed by a partially protective surface film with the corrosion reaction occurring predominantly at the breaks or imperfections of the partially protective film. This is consistent with the known tendency of chloride ions to cause film break down, and the known instability of Mg(OH)2 in solutions with pH less than 7. However, it was observed that, with the increase in chloride ion concentration, the rising rate at corrosion rate decreased. The increase in corrosion rate with the increasing chloride ion concentration may be attributed to the participation of chloride ions in the dissolution reaction. Chloride ions were aggressive for magnesium. The adsorption of chloride ions to oxide covered magnesium surface transformed Mg(OH)2 to easily soluble MgCl2 [18]. Table 4 shows the corrosion rates obtained from salt fog tests at different pH value, chloride ion concentration and spraying time respectively. The corrosion rate decreased with the increase in spraying time. It resulted from the decrease in hydrogen evolution with an increase in spraying time; this is attributed to the corrosion occurring over an increasing fraction of the surface, which is the insoluble corrosion product. The insoluble corrosion product on the surface of the alloy could slow down the corrosion rate [19]. During the experiment, some black areas appeared initially, these areas become larger and additional similar areas appear with the increase in time. It was characterized by the observation of localized attack and many upheavals with pitting occurrence. Also the grain is refined, and quite a lot of b particles were distributed continually along the grain boundary. In this case, b phase particles cannot be easily destroyed and, with the increase of corrosion time, the quantity of b phases in the exposed surface would increase and finally play the role of corrosion barrier. Although, there are some grains of a phase still being corroded, most of the remaining a phase grains are protected under the b phase barrier, so the corrosion rate is decreased with the increase of spraying time [20].
5. Conclusions
(m) Experiment No. 13
(n) Experiment No.14
(o) Experiment No. 15
(p) Experiment No. 16
Fig. 5. Optical micrographs of corroded test specimens.
chloride ion concentration was much lower at higher concentrations [18].
1. An empirical relationship was developed to predict the corrosion rate of friction stir welded AZ61A magnesium alloy with a 95% confidence level. The relationship developed by incorporating the effect of pH value, spraying time and chloride ion concentration. 2. In AZ61A magnesium alloy friction stir welds, the highest corrosion rate was observed at pH3. The corrosion rate was higher in the acidic media than the alkaline and neutral media with the same concentration and spraying time period. 3. AZ61A magnesium alloy friction stir welds corroded more seriously with the increase in Cl concentrations. Since more Cl in NaCl solution promoted the corrosion, the corrosive intermediate (Cl) would be rapidly transferred through the outer layer and reach the substrate of the alloy surface. 4. The corrosion rate would decrease with increase in spraying time. It resulted in the decrease of hydrogen evolution with the increase in spraying time, which tended to increase the
5072
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
concentration of OH- ions thus strengthening the surface from further corrosion. The insoluble corrosion products on the surface of the alloy could slow down the corrosion rate.
Acknowledgements The authors would like to thank sincerely to the Department of Manufacturing Engineering, Annamalai University, Annamalai Nagar, for extending the facilities of Material Testing laboratory to carry out this investigation. The authors are ever grateful to CEMAJOR (Centre for Materials Joining & research), Department of Manufacturing Engineering, Annamalai University, Annamalai Nagar, for extending the facilities of corrosion testing laboratory to carry out investigation. References [1] Nagasawa T, Otsuka M, Yokota T, Ueki T. Structure and mechanical properties of friction stir weld joints of magnesium alloy AZ31. In: Kaplan HI, Hryn J, Clow B, editors. Magnesium technology 2000. Warrendale: TMS; 2000. p. 383–7. [2] Zettler R, Blanco AC, dos Santos JF, Marya S. The effect of process parameters and tool geometry on thermal field development and weld formation in friction stir welding of the alloy AZ31 and AZ61. In: Neelameggham NR, Daplan HI, Powell BR, editors. Magnesium technology. San Francisco: TMS; 2005. p. 409–23. [3] Inoue H, Sugahara K, Yamamoto A, Tsubakino H. Corrosion rate of magnesium and its alloys in buffered chloride solution. Corros Sci 2002;44:603–10. [4] Murray RW, Hillis J. Magnesium finishing: chemical treatment and coating practices. SAE Technical Paper Series #900791, Detroid; 1990. [5] Zeng Rong-Chang, Zhang Jin, Huang Wei-Jiu, Dietzel W, Kainer KU, Blawert C, et al. Review of studies on corrosion of magnesium alloys. Trans Non Ferrous Metals Soc China 2006;16:S763–71. [6] Makar GL, Kruger J. J Electrochem Soc 1990;137(2):414–21.
[7] Martin Johnsson, Dan Persson, Dominique Theirry. Corrosion product formation during Nacl induced atmospheric corrosion Mg alloy AZ91D. Corros Sci 2007;49:1540–58. [8] Chengying-Liang, Ting Weit Qin, Wang Hui-Min, Zhang Zhoa. Comparison of corrosion behavior of AZ31, AZ91, AM60 and ZK60 magnesium alloy. Trans Non Ferrous Metal Soc China 2008:514–7. [9] Balasubraminan P, Zetler R, Blawert C, Dietzel W. A study on the effect of plasma electrolytic oxidation on the stress corrosion cracking behavior of wrought AZ61 Mg alloy and its friction stir welding. Mater Charact 2009;60:289–389. [10] Zhang Li-Jun, Zhuxiu-Bei, Zhang Zhao, Zhang Jian-Quing. Electrochemical noise characteristic in corrosion process of AZ91D magnesium alloy in neutral chloride solution. Trans Non Ferrous Metal Soc China 2009;19:496–503. [11] Abady Ghada M, Hillal Nadia H, El-Rabieee Mohammed, Badary Nahzed A. Effect of Al content on the corrosion behavior of Mg–Al alloys in aqueous solutions of different pH. Electrochem Acta 2010;55:6651–8. [12] Merins MC, Paradr A, Arrabal R, Morine S. Influence of chloride ion concentration and temperature on the corrosion of Mg–Al alloy in salt fog. Corros Sci 2010;52:1696–764. [13] Holly Martin J, Horste Meyer MF, Paul Wang T. Structure property quantification of corrosion pitting anoles immersion and salt spray environment on an extruded AZ61 magnesium alloy. Corros Sci 2011. [14] Zeng Rong-Chang, Chen Jun, Dietzel Wolfgang, Zettler Rudolf, Jorge Dos Santos F, Lucia Nascimento M, et al. Corrosion of friction stir welded magnesium alloy AM50. Corros Sci 2009;51:1738–46. [15] Song G, Atrens A. Understanding magnesium corrosion – a framework for improved alloy performance. Adv Eng Mater 2003;5:837–58. [16] Standard Practice for Operating Salt spray (Fog) apparatus, ASTM B 117, American Society for Testing of Materials; 2003. [17] Standard Practice for Preparing, Cleaning and evaluating Corrosion Test Specimens, ASTM G1, American Society for Testing of Materials; 2003. [18] Hikmet Altun, Sadrisen. Studies on the influence of chloride ion concentration and pH on the corrosion of AZ63 magnesium alloy. Mater Des 2004;25:637–41. [19] Gao Lili, Zhang Chunhong, Zhang Milin, Huang Xiaomel, Sheng Nan. The corrosion of a novel Mg–11Li–3Al–0.5RE alloy in alkaline NaCl solution. J Alloy Compd 2007:285–9. [20] Zeng Rong-Chang, Chen Juro, Dietzel Woltang, Zettler Rualolt, Jorge Dos Santos F. Corrosion of friction stir welded magnesium alloy AM50. Corros Sci 2009;51:1738–46.
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/matdes
Technical Report
Developing an empirical relationship to predict the corrosion rate of friction stir welded AZ61A magnesium alloy under salt fog environment A. Dhanapal a,⇑, S. Rajendra Boopathy b,1, V. Balasubramanian c,2 a
Department of Mechanical Engineering, Sri Ramanujar Engineering College, Vandalur, Chennai 600 048, Tamil Nadu, India Department of Mechanical Engineering, College of Engineering, Guindy, Anna University, Chennai 600 025, Tamil Nadu, India c Center for Materials Joining & Research (CEMAJOR), Department of Manufacturing Engineering, Annamalai University, Annamalainagar 608 002, Chidambaram, Tamil Nadu, India b
a r t i c l e
i n f o
Article history: Received 26 April 2011 Accepted 20 June 2011 Available online 25 June 2011
a b s t r a c t Magnesium (Mg) alloys shows the lowest density among other engineering metallic materials. As a consequence, this light alloy has a promising future. However, these alloys have great affinity for oxygen and other chemical oxidizing agents. The limitation of low corrosion resistance restricts their practical applications. Extruded Mg alloy plates of 6 mm thick of AZ61A grade were butt welded using friction stir welding (FSW) process. Corrosion behavior of the welds was evaluated by conducting salt fog test in NaCl solution at different chloride ion concentrations, pH value and spraying time. Also an attempt was made to develop an empirical relationship to predict the corrosion rate of friction stir welded AZ61A magnesium alloy. Three factors and a central composite design were used to minimize the number of experimental conditions. Response surface method was used to develop their relationship. The developed relationship can be effectively used to predict the corrosion rate of friction stir weld AZ61A magnesium alloy at 95% confidence level. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Magnesium (Mg) alloys are one of those light weight metallic alloys currently being investigated, because of its low density (1.74 g/cm3) and high mechanical stiffness [1]. The mechanical benefits of magnesium, however, are contrasted by a high corrosion rate as compared to aluminum or steel. Because of magnesium’s electrochemical potential, as illustrated in the galvanic series, it corrodes easily in the presence of sea water. The high corrosion property of magnesium has relegated the alloy to use in areas unexposed to the atmosphere, including car seat and electronic boxes. The demand for the use of magnesium alloys as structural materials in automobile industry, electronic products, vibrating plates of vibrating test machines, automotive wheels, etc., have increased in recent years [2]. The use of magnesium alloys in automotives reduces the total weight of the component up to 20% and at the same time this will substantially reduce the fuel consumption and the CO2 emissions. Conventional fusion welding of Mg alloys resulted in much solidification related problems such as porosity, hot cracking, alloy
⇑ Corresponding author. Tel.: +91 9444859455 (Mobile), +91 44 22751393 (R); fax: +91 44 22751370. E-mail addresses: [email protected] (A. Dhanapal), boopathy@ annauniv.edu (S.R. Boopathy), [email protected] (V. Balasubramanian). 1 Fax: +91 44 22357744. 2 Fax: +91 4144 238080/238275. 0261-3069/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.matdes.2011.06.038
segregation and partial melting zone. Hence, the applications of Mg alloys in structural members were still limited. However, a recent innovation of friction stir welding (FSW) process eliminated the above said problems. FSW is a solid state, autogenously process and hence there was no melting and solidification. Though the mechanical properties and micro structural characteristics of FSW joints of Mg alloys are the topic of many researchers, the corrosion properties of these joints have not yet been fully explored. In a buffer chloride solution, the corrosion rate of magnesium and its alloys did not depend on their purity or the content of the major alloying, but solely on the pH of the solution [3]. A serious limitation for the potential use of several magnesium alloys is their susceptibility to corrosion. Magnesium alloys, especially those with high purity, have good resistance to atmospheric corrosion. However, the susceptibility to corrosion in chloride containing environments is a serious concern [4]. The thickness of the oxide film formed on the surface of the specimen is quite increased with the increase in pH. The corrosion rate of AZ91 is high in acidic solution as compared to that in neutral and highly alkaline solutions. In extruded magnesium alloy the pitting corrosion is more at lower pH [5]. Alloying magnesium with aluminum in general improves the corrosion resistance [6]. In the initial stages of the corrosion attack, the cathodic reactions could involve both reductions of oxygen and water. The oxygen reduction occurred preferentially in the periphery of the electrolyte droplets on the surface, which led to alkaline conditions and an increase in CO2 deposition in these areas. After a long exposure time, a thickened layer of corrosion products were
5067
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072 Table 1a Chemical composition (wt.%) of AZ61A Mg alloy. Al
Zn
Mn
Mg
5.45
1.26
0.17
Balance
Table 1b Mechanical properties of AZ61A Mg alloy. Yield strength (MPa)
Ultimate tensile strength (MPa)
Elongation (%)
Vickers hardness at 0.05 kg load (Hv)
177
272
8.40
57
99
NORMAL
formed on the surface. Due to localized corrosion attack brucite was formed on those areas [7]. The galvanic couples formed by the second phase particles and the matrix are the main source of localized corrosion of magnesium alloys [8]. The general and pitting corrosion behavior of parent and FSW nugget regions were nearly the same, even though they were different from the untreated condition [9]. The overall corrosion processes can be described by three stages. The first stage corresponds to the pit nucleation and growth; the second stage involves the growth of a passive oxide layer; and the third stage involves the reactivation [10]. In wet air and aqueous solutions, oxide and hydroxide layers of Mg,MgO/Mg(OH)2, were formed spontaneously on the surface. Mg(OH)2 had a Pilling–Bedworth ratio of 1.77, and could provide enough corrosion protection to the metallic surface [11]. Corrosion attack of Mg, AZ31, AZ80 and AZ91D materials under the salt fog test increased with increasing temperature and Cl concentration [12]. Individual pitting characteristics, including pit surface area and pit volume, were greater for the salt spray surfaces [13]. From the literature review, it was understood that the most of the published information on corrosion behavior of Mg alloys was focused on pitting corrosion and general corrosion of unwelded base metals. Very few investigations [14,15] had evaluated the corrosion behavior on FSW joints of Mg alloys. Hence, the present investigation was carried out to develop an empirical relationship to predict the corrosion rate on FSW joints of AZ61A Mg alloy under salt fog conditions.
95 90 80 70 50 30 20 10 5 1
2. Experimental work 2.1. Fabricating the joints and preparing the specimens The material used in this study was an AZ61A magnesium alloy in the form of extruded plates of 6 mm thickness. The chemical composition and mechanical properties of the base metal are presented in Tables 1a and 1b. The optical micrograph of friction stir welded stir zone is shown in Fig. 1. The plate was cut to the required size (300 mm 150 mm) by power hacksaw followed by milling. A square butt joint configuration was prepared to fabricate friction stir welded (FSW) joints. The initial joint configuration obtained by securing the plates in position using mechanical clamps. The direction of welding was normal to the extruded direction. Single pass welding procedure was followed to fabricate the joints. Non-consumable tool made of high carbon steel was used to fabricate the joints. An indigenously designed and developed computer numerical controlled friction stir welding (22 kW; 4000 rpm; 60 kN) was used to fabricate the joints. The FSW parameters were optimized by conducting a lot of trial runs. The welding conditions which produced defect free joints were taken as optimized welding conditions. The optimized welding parameters are presented in Table 2. From the welded joints, the corrosion test specimens were sliced to the dimensions of 50 mm 16 mm 6 mm. The specimens were grounded with 500#, 800#, 1200#, 1500# grit SiC paper. Finally, it was cleaned with acetone and washed in distilled water then dried by warm flowing air.
-2.00
-1.00
-0.001
0.99
1.99
STUDENTIZED RESIDUAL Fig. 1. Normal probability plot.
1. If the pH value of the solution was less than 3 – the change in chloride ion concentration did not considerably affect the corrosion. 2. If the pH value was in between 3 and 11 – there was inhibition of the corrosion process and stabilization of the protective layer 3. If the pH value was greater than 11 – blocking of further corrosion by the active centers of protective layer. 4. If the chloride ion concentration was less than 0.2 M – the visible corrosion did not occur in the experimental period. 5. If the chloride ion concentration was in between 0.2 M and 1 Mn – there was a reasonable fluctuation in the corrosion rate. 6. If the chloride ion concentration was greater than 1 M – the rise in corrosion rate may hesitate and decrease a little. 7. If the spraying time was less than an hour – the surface was completely covered with thick and rough corrosion products. 8. If the spraying time was in between 1 and 9 h – the tracks of the corrosion can be predicted. 9. If the spraying time was greater than 9 h, the tracks of corrosion film were difficult to identify. 2.3. Developing the experimental design matrix
2.2. Finding the limits of corrosion test parameters From the literature, the predominant factors that have a greater influence on the corrosion rate of AZ61A FSW joints had been identified. They were: (i) pH value of the solution, (ii) spraying time and (iii) chloride ion concentration. Large numbers of trial experiments were conducted to identify the feasible testing conditions using friction stir welded AZ61A magnesium alloy joints under salt fog conditions. The following inferences were obtained:
Owing to a wide range of factors, the use of three factors and a central composite rotatable design matrix were chosen to minimize the number of experiments. A design matrix consisting of 20 sets of coded conditions (comprising a full replication three factorial of 8 points, six corner points and six center points) was chosen in this investigation. Table 3 represents the range of factors considered, and Table 4 shows the 20 sets of coded and actual values used to conduct the experiments.
5068
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072 Table 2 Optimized welding conditions and process parameters used to fabricating the joints. Rotational speed (rpm)
Welding speed (mm/min)
Axial force (kN)
Tool shoulder diameter, (mm)
Pin diameter, (mm)
Pin length, (mm)
Pin profile
1000
75
3
18
6
5
Left hand thread of 1 mm pitch
Table 3 Important factors and their levels. S. No
Factor
1 2 3
Unit
pH value Spraying time Cl concentration
Hours (h) Mole (M)
Notation
Levels
P T C
1.682
1
0
+1
+1.682
3 1 0.2
4.62 2.62 0.36
7 5 0.6
9.38 7.38 0.84
11 9 1
Table 4 Design matrix and Experimental results. Ex. No
Coded values
Actual values
pH (P)
Time (T)
Conc. (C)
pH
Time (T) (h)
Conc. (C) (M)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 +1 1 +1 1 +1 1 +1 1.682 +1.682 0 0 0 0 0 0 0 0 0 0
1 1 +1 +1 1 -1 +1 +1 0 0 1.682 +1.682 0 0 0 0 0 0 0 0
1 1 1 1 +1 +1 +1 +1 0 0 0 0 1.682 +1.682 0 0 0 0 0 0
4.62 9.38 4.62 9.38 4.62 9.38 4.62 9.38 3 11 7 7 7 7 7 7 7 7 7 7
2.62 2.62 7.38 7.38 2.62 2.62 7.38 7.38 5 5 1 9 5 5 5 5 5 5 5 5
0.36 0.36 0.36 0.36 0.84 0.84 0.84 0.84 0.60 0.60 0.60 0.60 0.20 1 0.60 0.60 0.60 0.60 0.60 0.60
For the convenience of recording and processing experimental data, the upper and lower levels of the factors were coded here as +1.682 and 1.682 respectively. The coded values of any intermediate value could be calculated using the following relationship.
X i ¼ 1:682½2X ðX max X min Þ=ðX max X min Þ
ð1Þ
where Xi is the required coded value of a variable X and X is any value of the variable from Xmin to Xmax, Xmin is the lower level of the variable, Xmax is the upper level of the variable. 2.4. Recording the responses Solution of NaCl with concentrations of 0.2 M, 0.361 M, 0.6 M, 0.838 M, and 1 M were prepared. The pH value of the solution was maintained as pH 3, pH 4.619, pH 7, pH 9.38, & pH 11 with concentrated HCl and NaOH respectively. The pH value was measured using a digital pH meter. The test method consists of exposing the material to be tested in a salt spray chamber as per ASTM B 117 standards and evaluating the corrosion tested specimen with the method as per ASTM G1-03 [16,17]. Basically, the salt spray test procedure involves the spraying of a salt solution onto the samples being tested. This was done inside a temperature-controlled chamber. The glass racks were contained in the salt fog chamber (30 high, 30 deep and 50 wide). The samples under test
Corrosion rate (mm/year)
14.62 10.23 11.89 8.99 15.82 11.31 12.92 10.75 17.96 10.88 11.23 8.51 11.66 15.28 9.89 8.56 9.79 9.97 8.93 9.62
were inserted into the chamber, following which the salt-containing solution was sprayed as a very fine fog mist over the samples. NaCl in tapped water was pumped from a reservoir to spray nozzles. The solution was mixed with humidified compressed air at the nozzle and this compressed air, atomized the NaCl solution into a fog at the nozzle. Heaters were maintained at 35 °C cabinet temperature. Within the chamber, the samples were rotated frequently so that all samples were exposed uniformly to the salt spray mist. The temperature within the chamber was maintained at a constant level. Since the spray was continuous, the samples were continuously wet, and therefore, uniformly subjected to corrosion. The corrosion rate of the friction stir welded AZ61A alloy specimen was estimated by weight loss measurement. The original weight (wo) of the specimen was recorded and then the specimen was sprayed with the solution of NaCl for different spraying times of 1, 2.619, 5, 7.38 and 9 h. The corrosion products were removed by immersing the specimens for one minute in a solution prepared by using 50 g chromium trioxide (CrO3), 2.5 g silver nitrate (AgNO3) and 5 g barium nitrate (Ba(NO3)2) for 250 ml distilled water. Finally, the specimens were washed with distilled water, dried and weighed again to obtain the final weight (w1). The weight loss (w) can be measured using the following relation,
w ¼ ðwo w1 Þ
ð2Þ
5069
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
where w is the weight loss in grams, wo the original weight before test in grams and w1 is the final weight after test in grams. The corrosion rate of FSW joints of AZ61A was calculated by using the following equation:
Corrosion rate ðmm=yearÞ ¼
8:76 104 w ADT
ð3Þ
where w is the weight loss in grams, A the surface area of the specimen in cm2, D the density of the material, 1.72 g/cm3, T is the spraying time in hours. Micro structural examination of the corroded specimens was carried out using a light optical microscope (VERSAMET-3) incorporated with an image analysing software (Clemex-vision). The exposed samples surface was prepared for the micro examination both in the ‘‘AS polished’’ and ‘‘AS etched’’ conditions. Picral + Acetic acid were used as etchant. The corrosion test specimens were polished in disc polishing machine for scratch fewer surfaces and the surface was observed at 200 magnification. 3. Developing an empirical relationship In the present investigation, to correlate the salt spray test parameters and the corrosion rate of welds, a second order quadratic model was developed. The response (corrosion rate) is a function of pH value (P), spraying time (T) and chloride ion concentration (C) and it could be expressed as,
Corrosion rate ðCRÞ ¼ f ðP; T; CÞ
ð4Þ
The empirical relationship must include the main and interaction effects of all factors and hence the selected polynomial is expressed as follows:
Y ¼ bo þ
X
b i xi þ
X
bii x2i þ
X
bij xi xj
ð5Þ
For three factors, the selected polynomial could be expressed as
Corrosion rate ðCRÞ ¼ b0 þ b1 ðPÞ þ b2 ðTÞ þ b3 ðCÞ þ b11 ðP2 Þ þ b22 ðT 2 Þ þ b33 ðC 2 Þ þ b12 ðPTÞ þ b13 ðPCÞ þ b23 ðTCÞ
Corrosion rate ðCRÞ ¼ 9:48 1:89P 0:88T þ 0:82C þ 1:60P 2 þ 1:27C 2
ð8Þ
The Analysis of Variance (ANOVA) technique was used to find the significant main and interaction factors. The results of second order response surface model fitting in the form of Analysis of Variance (ANOVA) are given in Table 5. The determination coefficient (r2) indicated the goodness of fit for the model. The Model F-value of 31.30 implies the model is significant. There is only a 0.01% chance that a ‘‘Model F-Value’’ this large could occur due to noise. Values of ‘‘Prob > F’’ less than 0.0500 indicates model terms are significant. In this case P, T, C, P2, C2 are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve the model. The ‘‘Lack of Fit F-value’’ of 1.69 implies the Lack of Fit is not significant relative to the pure error. There is a 28.93% chance that a ‘‘Lack of Fit F-value’’ this large could occur due to noise. Non-significant Lack of Fit is good. The ‘‘Pred R-Squared’’ of 0.8176 is in reasonable agreement with the ‘‘Adj R-Squared’’ of 0.9349. ‘‘Adeq Precision’’ measures the signal to noise ratio. P ratio greater than 4 is desirable. Our ratio of 19.440 indicates an adequate signal. The normal probability of the corrosion rate shown in Fig. 1 reveals the residuals were falling on the straight line, which meant that the errors were distributed normally. All of this indicated an excellent suitability of the regression model. Each of the observed values compared with the experimental values shown in Fig. 2.
ð6Þ
where b0 is the average of responses (corrosion rate) and b1, b2, b3, . . . b11, b12, b13, . . . b22, b23, b33, are the coefficients that depend on their respective main and interaction factors, which were calculated using the expression given below,
Bi ¼
number of combination considered. All the coefficients were obtained applying central composite face centered design using the Design Expert statistical software package. After determining the significant coefficients (at 95% confidence level), the final relationship was developed using only these coefficients. The final empirical relationship obtained by the above procedure to estimate the corrosion rate of friction stir welds of AZ61A magnesium alloy is given below,
X ðX i ; Y i Þ=n
ð7Þ
where ‘i’ varies from 1 to n, in which Xi is the corresponding coded value of a factor and Yi is the corresponding response output value (corrosion rate) obtained from the experiment and ‘n’ is the total
4. Discussion Fig. 3 shows the micrograph of stir zone of friction stir welded AZ61A magnesium alloy before corrosion tests. Fig. 4 shows the scanned images of the corrosion test specimen after salt fog tests at different pH value, chloride ion concentration and spraying time. It was seen that, at lower concentration, the surface of the specimen was relatively slightly corroded in neutral and alkaline solutions while severely corroded at all pH values at higher chloride ion concentrations [18].
Table 5 ANOVA test results. Source
Sum of squares
df
Mean square
F value
p-value Prob > F
Model P T C PT PC TC P2 T2 C2 Residual Lack of Fit Pure error Cor total
126.60 49.03 10.55 9.12 1.83 0.047 0.033 37.07 3.404E-004 23.17 4.49 2.82 1.67 131.09
9 1 1 1 1 1 1 1 1 1 10 5 5 19
14.07 49.03 10.55 9.12 1.83 0.047 0.033 37.07 3.404E-004 23.17 0.45 0.53 0.33
31.30 109.11 23.48 20.29 4.08 0.10 0.072 82.48 7.575E-004 51.55
<0.0001 <0.0001 0.0007 0.0011 0.0710 0.7543 0.7934 <0.0001 0.9786 <0.0001
Significant
1.69
0.2893
Not significant
5070
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
(a) Experiment No. 1
(b) Experiment No. 2
(c) Experiment No. 3
(d) Experiment No. 4
(e) Experiment No. 5
(f) Experiment No. 6
(g) Experiment No. 7
(h) Experiment No. 8
Fig. 2. Correlation graph for response (corrosion rate).
Fig. 3. Microstructure of friction stir welded stir zone before corrosion test.
From Table 4, it shows the corrosion rate obtained from salt fog tests at different pH value, chloride ion concentration and spraying time. At every chloride ion concentration and spraying time, the joints usually exhibited a decrease in corrosion rate with the increase in pH. The highest corrosion rate was observed at pH 3 and at neutral pH, the corrosion rate was remained constant approximately and comparatively low corrosion rate was observed in alkaline solution. It was seen that the influence of pH was more at higher concentration as compared to lower concentration in neutral and alkaline solutions. Fig. 5 shows the microstructure of FSW joints after salt fog tests at different pH value, chloride ion concentration and spraying time respectively, it was observed that the matrix shows the pitting marks and the pitting corrosion that has taken place at the friction stir welded microstructure. The particles are Mn–Al compound and fragmented Mg17Al12. The pits were marked in the microstructure with arrows. The numbers of pits were more in the joints when it is sprayed with the solution of low pH. Hence the corrosion rate increases with the decrease in pH value. Since the increase of grain and grain boundary in the joints, the grain boundary acts cathodic to grain causing micro galvanic effect. Corrosion tends to be concentrated in the area adjacent to the grain boundary until eventually the grain may be undercut and fall out [19]. It meant that the pH value was one of the major factors of corrosion rate. At lower pH values, the specimen exhibited a rise in corrosion rate with an increase in chloride ion concentration. But the quantity of this rise was different in such a way that, the change in chloride ion concentration at lower concentrations affected the corrosion rate much more as compared to that of higher concentra-
(i) Experiment No. 9
(j) Experiment No. 10
(k) Experiment No.11
(l) Experiment No. 12
(m) Experiment No. 13
(n) Experiment No. 14
(o) Experiment No. 15
(p) Experiment No. 16
Fig. 4. Scanned image of the corrosion test specimens.
tion. It showed that with the increase in chloride ion concentration, the rising rate at corrosion rate decreased that is, the influence of
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
(a) Experiment No. 1
(b) Experiment No. 2
(c) Experiment No. 3
(d) Experiment No. 4
(e) Experiment No. 5
(f) Experiment No. 6
(g) Experiment No. 7
(h) Experiment No. 8
(i) Experiment No. 9
(j) Experiment No. 10
(k) Experiment No. 11
(l) Experiment No. 12
5071
From the microstructure of FSW joints after salt fog tests at different pH value, chloride ion concentration and spraying time, it showed that the alloy exhibited a rise in corrosion rate with the increase in Cl concentration and thus the change of Cl concentration affected the corrosion rate much more in higher concentration solutions than that in lower concentration solutions. When more Cl in NaCl solution promoted the corrosion, the corrosive intermediate (Cl) would be rapidly transferred through the outer layer and reached the substrate of the alloy surface [19]. Hence, the corrosion rate was increased. This corrosion behavior is consistent with the current understanding that the corrosion behavior of magnesium alloys is governed by a partially protective surface film with the corrosion reaction occurring predominantly at the breaks or imperfections of the partially protective film. This is consistent with the known tendency of chloride ions to cause film break down, and the known instability of Mg(OH)2 in solutions with pH less than 7. However, it was observed that, with the increase in chloride ion concentration, the rising rate at corrosion rate decreased. The increase in corrosion rate with the increasing chloride ion concentration may be attributed to the participation of chloride ions in the dissolution reaction. Chloride ions were aggressive for magnesium. The adsorption of chloride ions to oxide covered magnesium surface transformed Mg(OH)2 to easily soluble MgCl2 [18]. Table 4 shows the corrosion rates obtained from salt fog tests at different pH value, chloride ion concentration and spraying time respectively. The corrosion rate decreased with the increase in spraying time. It resulted from the decrease in hydrogen evolution with an increase in spraying time; this is attributed to the corrosion occurring over an increasing fraction of the surface, which is the insoluble corrosion product. The insoluble corrosion product on the surface of the alloy could slow down the corrosion rate [19]. During the experiment, some black areas appeared initially, these areas become larger and additional similar areas appear with the increase in time. It was characterized by the observation of localized attack and many upheavals with pitting occurrence. Also the grain is refined, and quite a lot of b particles were distributed continually along the grain boundary. In this case, b phase particles cannot be easily destroyed and, with the increase of corrosion time, the quantity of b phases in the exposed surface would increase and finally play the role of corrosion barrier. Although, there are some grains of a phase still being corroded, most of the remaining a phase grains are protected under the b phase barrier, so the corrosion rate is decreased with the increase of spraying time [20].
5. Conclusions
(m) Experiment No. 13
(n) Experiment No.14
(o) Experiment No. 15
(p) Experiment No. 16
Fig. 5. Optical micrographs of corroded test specimens.
chloride ion concentration was much lower at higher concentrations [18].
1. An empirical relationship was developed to predict the corrosion rate of friction stir welded AZ61A magnesium alloy with a 95% confidence level. The relationship developed by incorporating the effect of pH value, spraying time and chloride ion concentration. 2. In AZ61A magnesium alloy friction stir welds, the highest corrosion rate was observed at pH3. The corrosion rate was higher in the acidic media than the alkaline and neutral media with the same concentration and spraying time period. 3. AZ61A magnesium alloy friction stir welds corroded more seriously with the increase in Cl concentrations. Since more Cl in NaCl solution promoted the corrosion, the corrosive intermediate (Cl) would be rapidly transferred through the outer layer and reach the substrate of the alloy surface. 4. The corrosion rate would decrease with increase in spraying time. It resulted in the decrease of hydrogen evolution with the increase in spraying time, which tended to increase the
5072
A. Dhanapal et al. / Materials and Design 32 (2011) 5066–5072
concentration of OH- ions thus strengthening the surface from further corrosion. The insoluble corrosion products on the surface of the alloy could slow down the corrosion rate.
Acknowledgements The authors would like to thank sincerely to the Department of Manufacturing Engineering, Annamalai University, Annamalai Nagar, for extending the facilities of Material Testing laboratory to carry out this investigation. The authors are ever grateful to CEMAJOR (Centre for Materials Joining & research), Department of Manufacturing Engineering, Annamalai University, Annamalai Nagar, for extending the facilities of corrosion testing laboratory to carry out investigation. References [1] Nagasawa T, Otsuka M, Yokota T, Ueki T. Structure and mechanical properties of friction stir weld joints of magnesium alloy AZ31. In: Kaplan HI, Hryn J, Clow B, editors. Magnesium technology 2000. Warrendale: TMS; 2000. p. 383–7. [2] Zettler R, Blanco AC, dos Santos JF, Marya S. The effect of process parameters and tool geometry on thermal field development and weld formation in friction stir welding of the alloy AZ31 and AZ61. In: Neelameggham NR, Daplan HI, Powell BR, editors. Magnesium technology. San Francisco: TMS; 2005. p. 409–23. [3] Inoue H, Sugahara K, Yamamoto A, Tsubakino H. Corrosion rate of magnesium and its alloys in buffered chloride solution. Corros Sci 2002;44:603–10. [4] Murray RW, Hillis J. Magnesium finishing: chemical treatment and coating practices. SAE Technical Paper Series #900791, Detroid; 1990. [5] Zeng Rong-Chang, Zhang Jin, Huang Wei-Jiu, Dietzel W, Kainer KU, Blawert C, et al. Review of studies on corrosion of magnesium alloys. Trans Non Ferrous Metals Soc China 2006;16:S763–71. [6] Makar GL, Kruger J. J Electrochem Soc 1990;137(2):414–21.
[7] Martin Johnsson, Dan Persson, Dominique Theirry. Corrosion product formation during Nacl induced atmospheric corrosion Mg alloy AZ91D. Corros Sci 2007;49:1540–58. [8] Chengying-Liang, Ting Weit Qin, Wang Hui-Min, Zhang Zhoa. Comparison of corrosion behavior of AZ31, AZ91, AM60 and ZK60 magnesium alloy. Trans Non Ferrous Metal Soc China 2008:514–7. [9] Balasubraminan P, Zetler R, Blawert C, Dietzel W. A study on the effect of plasma electrolytic oxidation on the stress corrosion cracking behavior of wrought AZ61 Mg alloy and its friction stir welding. Mater Charact 2009;60:289–389. [10] Zhang Li-Jun, Zhuxiu-Bei, Zhang Zhao, Zhang Jian-Quing. Electrochemical noise characteristic in corrosion process of AZ91D magnesium alloy in neutral chloride solution. Trans Non Ferrous Metal Soc China 2009;19:496–503. [11] Abady Ghada M, Hillal Nadia H, El-Rabieee Mohammed, Badary Nahzed A. Effect of Al content on the corrosion behavior of Mg–Al alloys in aqueous solutions of different pH. Electrochem Acta 2010;55:6651–8. [12] Merins MC, Paradr A, Arrabal R, Morine S. Influence of chloride ion concentration and temperature on the corrosion of Mg–Al alloy in salt fog. Corros Sci 2010;52:1696–764. [13] Holly Martin J, Horste Meyer MF, Paul Wang T. Structure property quantification of corrosion pitting anoles immersion and salt spray environment on an extruded AZ61 magnesium alloy. Corros Sci 2011. [14] Zeng Rong-Chang, Chen Jun, Dietzel Wolfgang, Zettler Rudolf, Jorge Dos Santos F, Lucia Nascimento M, et al. Corrosion of friction stir welded magnesium alloy AM50. Corros Sci 2009;51:1738–46. [15] Song G, Atrens A. Understanding magnesium corrosion – a framework for improved alloy performance. Adv Eng Mater 2003;5:837–58. [16] Standard Practice for Operating Salt spray (Fog) apparatus, ASTM B 117, American Society for Testing of Materials; 2003. [17] Standard Practice for Preparing, Cleaning and evaluating Corrosion Test Specimens, ASTM G1, American Society for Testing of Materials; 2003. [18] Hikmet Altun, Sadrisen. Studies on the influence of chloride ion concentration and pH on the corrosion of AZ63 magnesium alloy. Mater Des 2004;25:637–41. [19] Gao Lili, Zhang Chunhong, Zhang Milin, Huang Xiaomel, Sheng Nan. The corrosion of a novel Mg–11Li–3Al–0.5RE alloy in alkaline NaCl solution. J Alloy Compd 2007:285–9. [20] Zeng Rong-Chang, Chen Juro, Dietzel Woltang, Zettler Rualolt, Jorge Dos Santos F. Corrosion of friction stir welded magnesium alloy AM50. Corros Sci 2009;51:1738–46.