THREE-DIMENSIONAL INVESTIGATION OF CORRODED AND ENGINEERED SURFACES BY SEM IMAGES STEREO PAIR TECHNIQUE

KYUNG WON KANG; PABLO BILMES; CARLOS LLORENTE; M. E. CANAFOGLIA; R D BONETTO

Ariston Hotel Convention Center, ROSARIO, ARGENTINA

Congreso; 10th Inter-American Congress of Electron Microscopy (CIASEM 2009); 1st Congress of the Argentine Society of Microscopy (SAMIC 2009).; 2009

CIASEM, the Inter-American Committee of Societies for Electron Microscopy

<!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0cm; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman"; mso-ansi-language:ES-AR;} @page Section1 {size:612.0pt 792.0pt; margin:70.85pt 3.0cm 70.85pt 3.0cm; mso-header-margin:36.0pt; mso-footer-margin:36.0pt; mso-paper-source:0;} div.Section1 {page:Section1;} --> <!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0cm; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman"; mso-ansi-language:ES-AR;} @page Section1 {size:612.0pt 792.0pt; margin:70.85pt 3.0cm 70.85pt 3.0cm; mso-header-margin:36.0pt; mso-footer-margin:36.0pt; mso-paper-source:0;} div.Section1 {page:Section1;} --> <!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0cm; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman"; mso-ansi-language:ES-AR;} @page Section1 {size:612.0pt 792.0pt; margin:70.85pt 3.0cm 70.85pt 3.0cm; mso-header-margin:36.0pt; mso-footer-margin:36.0pt; mso-paper-source:0;} div.Section1 {page:Section1;} --> The large depth of field of the Scanning Electron Microscope (SEM) allows the three-dimensional study of surfaces from SEM images stereo pairs, obtained by scanning the same area under two different perspectives achieved by eucentric tilt of the sample. Features of different heights on the specimen surface are different in their lateral displacements in both images. These disparities in both images are measured to obtain quantitative surface topography data and to generate a dense height map. The main parameters to be considered in the Z component error calculation are the tilt-generated parallax precision and the tilt angle accuracy. Accuracy of Philips SEM 505 microscope used and of data of the dense height maps obtained by EZEImage program [1] was evaluated in the study of corrosion pits on corroded surfaces of boiler tubes and surfaces engineered for orthopedic implants.  The exact orientation of the tilt axis on the micrographs varies according to the working distance due to beam electron spiraling. It was found to be 12° with horizontal axis at a working distance of 12 mm for this microscope. Since acquiring stereo pairs in eucentric tilt is crucial to obtain precise measurements, studying repeatibility of the correct eucentric position was necessary. Thus, the specimen stage was placed at a specific tilt angle and one image was obtained, then it was removed to another angle and it was again placed at the first angle obtaining another image. This sequence was carried out for ten times showing a maximum change of one pixel with the corresponding change in the dense height map data. Nevertheless, when the eucentric position was corrected before image processing, identical dense height maps were obtained. In pitting corrosion, attack localized points (pits) appear on metal surfaces spreading then to inner regions. The depth and morphology of the deepest pit are usually the most meaningful parameters in order to define the geometry of critical defect and to be used to predict catastrophic failure conditions or values for life-assessment in industrial equipments, by means of statistical methods. Particularly, the pit aspect ratio (depth/diameter) is the main parameter affecting stress concentration factor in steel and other alloys [2]. Reliable three-dimensional data are therefore necessary. The accuracy of z calculations in a corrosion pit by EZEImage program (Fig.1a and 1b) was tested by means of many cross section measurements of the same pit (Fig. 1c). In Table 1a, some of these measurements differ much less than the estimated maximum error. Measurements at two different magnifications were also obtained and show a good agreement (Table 1b). Although EZEImage program allows obtaining a wide set of parameters to characterize the three dimensional surface topography, fewer parameters are necessary in each specific case. Particularly, in the 3D quantitative study of engineered surfaces for orthopedic implants carried out in this work, it was found (Fig. 2a), that the Sci functional parameter was the only that allowed characterizing the differences between two surfaces with different treatments (Ti2 and Ti5 samples). These results are in agreement with those reported by other researchers in the study of three-dimensional surface topography in dental implants [3]. <!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0cm; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman"; mso-ansi-language:ES-AR;} @page Section1 {size:612.0pt 792.0pt; margin:70.85pt 3.0cm 70.85pt 3.0cm; mso-header-margin:36.0pt; mso-footer-margin:36.0pt; mso-paper-source:0;} div.Section1 {page:Section1;} --> References [1] E. Ponz, J.L. Ladaga, R.D. Bonetto, Microscopy and Microanalysis 12 (2) (2006) 170-177. [2] M. Solari, P. Bilmes, Failure Analysis of HT Steel Components, 2008, pp7-34, ASM Int. Ed. [3] A. Arvidsson, B.A. Sater, A. Wennerberg, Clinical Implant Dentistry and Related Research 8 (2) (2006) 70-76The Extended Abstract (EA) will be evaluated by the Reviewers Committee to be accepted for publication in the CD-ROM Proceeding. This Proceeding will be published by Acta Microscopica (ISSN: 0798-4545), wich is the official Journal of the CIASEM. Thus the EA may need corrections, modifications or it may be considered not appropriated for publication in the CD. The final version should be ready before October 1st, to be included in the CD-ROM Proceedings.