Принципы нанометрологии (1027506), страница 30
Текст из файла (страница 30)
A fundamental problem is that a 2D profile does not necessarilyindicate functional aspects of the surface. For example, consider the mostcommonly used parameter for 2D surface characterisation, Ra (see section8.2.7.1). Figure 6.4 shows the profiles of two surfaces, both of which returnthe same Ra value when filtered under the same conditions. It can be seenthat the two surfaces have very different features and consequently verydifferent functional properties.With profile measurement and characterization it is often difficult todetermine the exact nature of a topographic feature. Figure 6.5 shows a 2Dprofile and a 3D surface map of the same component covering the samemeasurement area.
With the 2D profile alone a discrete pit is measured on121122C H A P T ER 6 : Surface topography measurement instrumentationFIGURE 6.4 Profiles showing the same Ra with differing height distributions.FIGURE 6.5A profile taken froma 3D measurementshows the possibleambiguity of 2Dmeasurement andcharacterization.the surface. However, when the 3D surface map is examined, it can be seenthat the assumed pit is actually a valley and may have far more bearing on thefunction of the surface than a discrete pit.The measurement of areal surface texture has a number of benefits overprofile measurement.
Areal measurements give a more realistic representationof the whole surface and have more statistical significance. Also, there is lesschance that significant features will be missed by an areal method and themanufacturer gains a better visual record of the overall structure of the surface.6.6 Surface topography measuring instrumentationOver the past one hundred years, and especially in the last thirty years, therehas been an explosion in the number of instruments that are available tomeasure surface texture.
The instruments can be divided into three broadSurface topography measuring instrumentationclasses: line profiling, areal topography measuring and area-integratingmethods [25]. Line profiling methods produce a topographic profile, z(x).Areal topography methods produce topographic images, z(x, y). Often, z(x, y)is developed by juxtaposing a set of parallel profiles. Area-integratingmethods measure a representative area of a surface and produce numericalresults that depend on area-integrating properties of the surface.
This chapterwill highlight the most popular instruments available at the time of writingand more instruments are discussed in [7–10]. Scanning probe and electronbeam instruments are described in chapter 7.6.6.1 Stylus instrumentsStylus instruments are by far the most common instruments for measuringsurface texture today, although optical instruments and scanning probemicroscopes are becoming more common in MNT manufacturing facilities.A typical stylus instrument consists of a stylus that physically contacts thesurface being measured and a transducer to convert its vertical movementinto an electrical signal. Other components can be seen in Figure 6.6 andinclude: a pickup, driven by a motor and gearbox, which draws the stylus overthe surface at a constant speed; an electronic amplifier to boost the signalFIGURE 6.6 Schema of a typical stylus instrument.123124C H A P T ER 6 : Surface topography measurement instrumentationfrom the stylus transducer to a useful level; and a device, also driven ata constant speed, for recording the amplified signal [1,26,27].The part of the stylus in contact with the surface is usually a diamond tipwith a carefully manufactured shape.
Commercial styli usually have tip radiiof curvature ranging from 2 mm to 10 mm, but smaller or larger styli areavailable for specialist applications and form measurement respectively.Owing to their finite shape, some styli on some surfaces will not penetrateinto valleys and will give a distorted or filtered measure of the surface texture.Consequently, certain parameters will be more affected by the stylus shapethan others. The effect of the stylus shape has been extensively coveredelsewhere (see for example [7,28–30]). The effect of the stylus force can havea significant influence on the measurement results and too high a force cancause damage to the surface being measured (see Figure 6.7).
ISO 3274 [26]states that the stylus force should be 0.75 mN but this is rarely checked andcan vary significantly from the value given by the instrument manufacturer.The value of 0.75 mN was chosen so as not to cause scratches in metals witha 2 mm radius stylus, but it does cause scratches in aluminium. Smaller forceslimit the measurement speed due to the risk of ‘stylus flight’. Someresearchers ([31,32] and, more recently [33]) have developed constant-forcestylus instruments to improve the fidelity between the surface and the stylustip plus reduce surface damage and dynamic errors.To enable a true cross-section of the surface to be measured, the stylus, asit is traversed across the surface, must follow an accurate reference path thathas the general profile of, and is parallel to, the nominal surface.
SuchFIGURE 6.7 Damage to a brass surface due to a high stylus force.Surface topography measuring instrumentationa datum may be developed by a mechanical slideway; for examples see [34]and [35]. The need for accurate alignment of the object being measured iseliminated by the surface datum device in which the surface acts as its owndatum by supporting a large radius of curvature spherical (or sometimes withdifferent radii of curvature in two orthogonal directions) skid fixed to the endof the hinged pickup. At the front end of the pickup body the skid rests on thespecimen surface (note that skids are rarely seen on modern instruments andnot covered by ISO specification standards).All the aspects of stylus instruments are discussed in great detail elsewhere [7]. The main sources of error associated with a stylus instrument aresimply listed below:-surface deformation;-amplifier distortion;-finite stylus dimensions;-lateral deflection;-effect of skid or other datum;-relocation upon repeated measurements;-effect of filters – electrical or mechanical;-quantization and sampling effects;-dynamic effects;-environmental effects;-effect of incorrect data-processing algorithms.The lateral resolution of a stylus instrument, or the shortest wavelength,l, of a sinusoidal signal where the probe can reach the bottom of the surface,is given bypffiffiffiffiffil ¼ 2p ar(6.1)where a is the amplitude of the surface and r is the radius of the stylus tip.Note that equation (6.1) only applies for a sinusoidal profile.
Quantizationeffects and the noise floor of the instrument will determine the axial, orheight, resolution.Modern stylus instruments regularly obtain measurements of surfacetexture with sub-nanometre resolution but struggle to obtain true traceabilityof these measurements in each of their axes. It is worth pointing out here thatmany of the pitfalls of mechanical stylus techniques are often highly125126C H A P T ER 6 : Surface topography measurement instrumentationexaggerated [36]. For example, the wear on the surface caused by a stylus isoften stated as its fundamental limit, but even if a stylus does cause somedamage, this may not affect the functionality of the surface.
There havebeen some proposals to speed up the performance of a stylus by vibrating itaxially [37].One drawback of a stylus instrument when operated in an areal scanningmode is the time to take a measurement. It is perfectly acceptable to takeseveral minutes to make a profile measurement, but if the same number ofpoints are required in the y direction (orthogonal to the scan direction) as aremeasured in the x direction, then measurement times can be up to severalhours.
For example, if the drive mechanism can scan at 0.1 mm$s1 and1000 points are required for a profile of 1 mm, then the measurement willtake 10 s. If a square grid of points is required for an areal measurement, thenthe measurement time will increase to 105 s or approximately 2.7 hours.This sometimes precludes the use of a stylus instrument in a production orin-line application.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
