Quantitative characterization of the surface topography of rolled sheets by laser scanning microscopy and fourier transf

  • PDF / 3,370,476 Bytes
  • 9 Pages / 598 x 778 pts Page_size
  • 34 Downloads / 155 Views

DOWNLOAD

REPORT


I.

INTRODUCTION

THE surface structure of twin-roll cast aluminum sheets and its development through cold rolling are important to the quality of the final surface structure of the product. Every pass during cold rolling involves a specific change in surface topography. This is generally caused by interfacial sliding and the associated local deformation field. The specific effect on the surface structure of the sheet is affected by the initial surface structure and the topography of the rolls together with the actions of the lubricant. To understand and control the evolution of surface structure through cold rolling, it is important to characterize the surface in a relevant and tailored manner. The topography of the surface must therefore be described both statistically and microstructurally, and in particular, relevant quantitative parameters must be developed. There are several techniques for measuring surface topography, where the most established one is stylus profilometry with both single trace and area measurements with a set of traces. In the last 10 years, several different techniques for measuring surface topography without physical contact of the surface have been commercially available. Among these are interferometry,tt] optical contour display,~ nanoswing technique,t31 and the confocal technique of the laser scanning microscope (LSM). t41 These methods eliminate the problem with deformation of finer details of the surface by the profilometer stylus tip. There is also a large gain in resolution when using these new techniques compared to profilometers. The traditional roughness parameters have for decades been accepted as inadequate for detailed studies of surface topography due to the inability to yield parameters other

than mean values of the height distribution of the surface. As early as 1962, Myerst51 suggested a development of the root-mean-square by introducing the first and second derivative of a surface profile. In 1971, Spragg and Whitehouset61 introduced the average wavelength index calculated from the autocorrelation function of a profile. This was done to supplement the traditional amplitude distribution (the arithmetic mean). Thwaite[71 reported in 1979 the use of optical Fourier transformation as a direct means to achieve the Fourier (or Power) spectrum of a rough surface. The introduction of such parameters is an effort to be able to quantify the typical trends of a surface and to be able to distinguish between apparently "similar" surfaces. The present work combines the advanced confocal imaging technique of LSM and Fourier transformation of the confocal three-dimensional (3-D) measurements. The topographical images of the surface produce microstructural information, and the Fourier spectra give detailed statistical information about the same area. This gives important and new knowledge of the surface topography in addition to what is gained from traditional techniques. This is an approach highly needed in work concerning the mechanisms forming and deforming surface structure during ro