possibilities and limitations of digital speckle pattern interferometry in the analysis of corrosion processes in metallic materials

1、This content has been downloaded from IOPscience. Please scroll down to see the full text.Download details:IP Address: 128.250.144.144This content was downloaded on 04/02/2015 at 02:50Please note that terms and condition

2、s apply.Possibilities and limitations of digital speckle pattern interferometry in the analysis of corrosionprocesses in metallic materialsView the table of contents for this issue, or go to the journal homepage for more

3、2013 Meas. Sci. Technol. 24 075204(http://iopscience.iop.org/0957-0233/24/7/075204)Home Search Collections Journals About Contact us My IOPscienceMeas. Sci. Technol. 24 (2013) 075204 A Andres-Arroyo et alAs corrosion sur

4、face processes are local and non- uniform, it is very useful to apply techniques that provide 2D information on the surface. In this sense, digital speckle photography has been applied to visualize the local surface chan

5、ges by recording the evolution of the correlation coefficient [8–10]. Digital speckle pattern interferometry (DSPI) has been used to obtain information on the changes at all points of the sample surface simultaneously [1

6、1–13]. DSPI has already been applied to measure the thickness corrosion layer in a metallic sample immersed in a corrosive solution [12]. In this case, measurements were performed in air after having extracted the sample

7、 from the liquid. Some attempts have been made to detect changes in metallic surfaces immersed in liquid corrosive solutions [13]. In this work, the applicability of DSPI for studying the corrosion process evolution whil

8、e the sample is kept immersed in the solution has been investigated. Corrosion has been induced by applying a given electrical current between two Cu electrodes which are immersed in a CuSO4 solution 0.1 M inside an elec

9、trolytic cell. The aim of this work is to determine what information about the corrosion process for the described conditions can be obtained with DSPI. The limitations of sensitivity and accuracy when quantitative infor

10、mation is retrieved will also be analyzed.2. ConceptsDSPI is a widely used, well-known technique, with different implementations which are well described in the literature [14]. The characteristics of DSPI, which are rel

11、evant for this work, can be described as follows.(i) A rough surface is imaged onto the sensor of a CCD camera. (ii) The laser light scattered by the surface (speckle wave) is combined with a smooth laser beam and their

12、interference is recorded on the sensor. (iii) Spatial phase shifting (SPS) is used in order to obtain the scattered light (random) phase map. The analysis is performed with the Fourier transform method [15]. (iv) Phase d

13、ifference maps are obtained by subtracting the phases for two different time instants. (v) The wrapped phase differences in the range [?π, π] are mapped into gray levels, for visualization. Low- pass filter and phase unw

14、rapping processes are used in order to obtain quantitative information, with an estimated accuracy of λ/100.Phase differences, ?φ, are directly related to optical path length changes, ?δ. ?δ are produced both by local su

15、rface displacements, L, and by refractive index changes along the beam path, ?n. For the case of only surface displacements, the phase differences can be expressed as ?φL = K · L, where K = (ko ? ki) is the sensitiv

16、ity vector and ki and ko are the illumination and observation vectors. ki and ko have the same magnitude, k = 2πna/λo, where na is the refractive index of the medium surrounding the surface and λo is the wavelength in ai

17、r.Figure 1. DSPI setup for measuring surface displacements.We will refer to the vector directions as ui and uo respectively. Thus, ?φL can be expressed as?φL = 2πλo na(uo ? ui) · L = 2πλo na|uo ? ui|LK, (1)where LK

18、is the projection of the surface displacement along the direction of the sensitivity vector. In our experiments, the rough surface is a metallic surface (sample) inside an electrolytic cell (figure 1). For a Z axis along

19、 the surface normal, when taking kiy = 0 and neglecting koy, equation (1) can be expressed as?φL(x, y) = 2πλo na[Lx(x, y)(sin θi ? sin θo)+ Lz(x, y) (cos θi + cos θo)] (2)where θi and θo are the angles between ki and ko

20、and the surface normal. When only Lz(x, y) displacements take place, the phase difference maps are Lz(x, y) maps. In the corrosion process studied in this work, Lz(x, y) is not a displacement, it is associated with the e

21、volution of the corrosion layer over the sample surface. For the case of only refractive index changes, the phase differences can be expressed as?φn = 2πλo ?δ = 2πλo? d0 ?n dl (3)where l refers to positions along the bea

22、m travelling path and d is the total length with refractive index changes. When the DSPI experiments are carried out while the surface is immersed in the corrosive solution, phase changes due to both corrosion depth and

23、solution refractive index changes can take place. In this case, the total phase difference will just be ?φ = ?φL + ?φn, when neglecting higher order terms. Since we are interested only in measuring corrosion depths, it i

24、s necessary to determine ?φn for removing it from the measured ?φ. For measuring ?φn while the corrosion is taking place, the DSPI setup shown in figure 2 has been used. In this case, the illumination beam passes through

25、 a ground glass. The produced speckle wave travels along the direction of the X axis, passing through the liquid in front of the corroding surface (sample). The phase changes of this speckle wave are produced by the refr

26、active index changes that occur in the corrosion cell in front of the sample. Therefore, d is never bigger than the corrosion cell length along X. In this case, the phase difference maps are average refractive index chan