SurfaceFinish Report

Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 8

Ryan Bourget 1

Final Manufacturing Assignment

Surface Finishing
From Ra to Rz

April 2003
Ryan Bourget
Ryan Bourget 2

Surface Finish

Surface finish, by definition, is the allowable deviation from a


perfectly flat surface that is made by some manufacturing process.
Whenever any process is used to manufacture a part, there will be some
roughness on the surface. This roughness can be caused by a cutting tool
making tiny grooves on the surface or by the individual grains of the
grinding wheel each cutting it’s own groove on the surface. It is affected by
the choice of tool, speed of the tool, environmental conditions, and definitely
by what material you are working with. Even when there is no machining
involved, as in casting/injection molding, the surface of the mold will have
surface deviation, which in turn will be transferred onto the part. Even if
you could create a mold which was perfectly flat, the cooling process and
thermal properties of the material would cause surface imperfections. So,
like everything else in the manufacturing world, we have had to make a
compromise between function and cost of manufacturing. If you don’t need
a mirror finish all over the brake drum, then you just cast it and worry about
machining the surfaces that need to be (relatively) flat for the function of the
part. I have included at the end of the report a chart that details the surface
finishes that can be obtained by many different processing methods.

The science of metrology – the and this will create (either by


study of surface finish/texture/etc. analog or electronic means) a
goes into such depth with the profile of the surface. A profile
statistical analysis and complex might look like this:
calculus (I’ll explain how this
enters later) that I’m only going to
go over the basics of two-
dimensional surface profile
analysis. For most types of surface
finish measurements, the testing
equipment works as follows: a
stylus is dragged across a surface,
Ryan Bourget 3

Surface texture is generally broken up into three components upon


analysis: roughness, waviness, and form. Roughness is generally the marks
made on a surface by the machining tool. eg: grooves from the tool or from
each grinding granule on a grinding wheel. Waviness is the result of the
distance between the cutting tool and the workpiece changing. This is caused
by vibration in the tool, a class of students walking by the machine, etc.
Vibration can also be inside the tool; from the power motor, or from a worn
spindle. If you assume that the part is straight and flat, form errors are
because the machine tool's ways are not straight or are worn. This will create
surface irregularities, but will do so in a consistant manner, because the
machine will always follow the same path along it’s ways.

All three surface finish components exist simultaneously. They just


overlap one another. We often will want to look at each (roughness,
waviness, and form) separately, so we make the assumption (a correct one,
in most cases) that roughness has a shorter wavelength than waviness, which
in turn has a shorter wavelength than form.

When analyzing surface finish, there are more than 100 different
parameters in existence (in recognized standards) and many more that have
been developed for special products/circumstances. Many of these are either
redundant or just plain unnecessary. I am going to examine in some depth
Ra and the different forms of Rz in use today, and touch on some others, as
they have their own usefulness.

The parameter most used in North America for general surface


roughness is Ra. It measures average roughness by comparing all the peaks
and valleys to the mean line, and then averaging them all over the entire
cutoff length. (Cutoff length is the length that the stylus is dragged across
the surface; a longer cutoff length will give a more average value, and a
shorter cutoff length might give a less accurate result over a shorter stretch
of surface) It’s also referred to as CLA (center line average) and AA (area
average). The old name for the specification is RHR. In the picture below,
the Ra can fall between 10 and 20μin. Benefits to using the Ra method are
its simplicity and its widespread use.
Ryan Bourget 4

However, Ra is a simple method of analyzing the profile, and it


doesn’t tell you everything about a surface. The three surfaces below will
all display the same Ra value, but as you can see, they are quite different
surfaces.

So if Ra doesn’t give a detailed enough picture of the surface finish of


a part, we need to use a different parameter. Another parameter that can be
useful is Rymax. This is an ISO standard that measures the distance between
the highest peak and the lowest valley over a cutoff length. This is a
sensitive method of interpreting the data, and if over the measurement length
a scratch or imperfection was encountered, you would get an extremely high
reading. It was formerly called just Rmax.

An old form of measuring is by using a parameter called Rq. For Rq,


you take the profile and turn it into a pure sine wave. Integrating the area
under the curve can get the mean deviation from the mean line, and calculate
the surface roughness. But because the profile of a surface will NEVER
look like a sine wave, this approximation has been almost completely
replaced by more accurate means of analyzing and reporting the data.
Ryan Bourget 5

The parameter most widely used in Europe is Rz, or mean roughness


depth. The Rz ISO standard is also called “Ten Point Average Roughness”.
It averages the height of the five highest peaks and the depth of the five
lowest valleys over the measuring length, using an unfiltered profile. The
Rz DIN standard (a German standard) averages the highest point and lowest
point over five cutoffs. The newer Japanese standard (JIS) measures the
same points, but filters (slightly smooths) the raw data before creating a
profile. There is also a different method of interpreting the same data, in a
parameter called R3zi. It takes the third highest peak and the third highest
valley over one sample length. This is much more forgiving of things like
scratches and uncommon imperfections in the surface, but you wouldn’t
want to evaluate the surface that you are mounting a seal to if it will
disregard a gouge in the surface. There is also R3z, which averages a
number of R3zi values.

There are many many more parameters and evaluation methods for
surface finish. Some are quite specialized for a small number of
applications. One company manufactures tubing for process (neutral) gasses
for the manufacture of microchips and microprocessors. The smallest
imperfections on the inside of the tube can cause a buildup of contaminants,
and these in turn can be picked up by the inert gas. They have developed
new parameters – RaCH : continually shifts the mean point and identifies the
largest Ra over the entire length. RaCL: the same, but identifies the smallest
Ra over the length. RaCA: continually averages the Ra values over the
cutoff length. These parameters haven’t been adopted by any governing
standards body, but have been accepted as a “standard” within that industry.

Other parameters, such as skewness, kurtosis, core roughness depth,


Swedish height, etc. go into much more depth in analyzing the profile.
Some of these are used in optical applications where the angle of the wave
entering the surface will be refracted/reflected differently based on the angle
of the profile “bumps”.
Ryan Bourget 6

The surface finish symbol, as denoted on drawings, can specify many


more things than general surface roughness (Ra, usually in μin or μm). See
drawing below
f – roughness value other than Ra

a – roughness value (Ra)


b – production method, treatment,
coating, etc.
c – sampling length (cutoff)
d – direction of lay
e – minimum material removal

preceded by its symbol (eg Rz 1.5)

Surface finish may be denoted by a roughness grade number. Here is a


table that specifies the Ra values for roughness grade numbers. (ISO standard –
1992)

Roughness values Ra Roughness Grade


µ m µ in Numbers
50 2000 N12
25 1000 N11
12.5 500 N10
6.3 250 N9
3.2 125 N8
1.6 63 N7
0.8 32 N6
0.4 16 N5
0.2 8 N4
0.1 4 N3
0.05 2 N2
0.025 1 N1
Ryan Bourget 7

There is hundreds of different styles of 2-D surface testers, but they all
essentially do the same thing. Stylus tips are the same as on most precision
measuring equipment, being a very hard stone usually. The stylus on the
Mitutoyo Surftest 211 (the tester in the shop) is diamond tipped. This is to
prevent deflection of the tip when it encounters the tiny bumps on the surface.
The stylus tip can have a different radius depending on the application and how
you plan to analyze the data. The radius on the diamond tip on the Mitutoyo
tester is 5μm. The downward force of the stylus when measuring is 4 mN.
While something of this nature may be good for most circumstances, if you are
testing the surface of a soft material like gold, you might be scratching the
surface with your surface tester. The customer will probably fail to see the
irony in that, so different methods of testing must be used. There are optical
and other non-contact methods of measuring surface quality, but are generally
very expensive (starting above $10,000). Surface testers generally are
calibrated before being used (and periodically thereafter) because the tester
amplifies the signal. It’s reading is compared to a known value, and adjusted
until the tester displays the same reading as the reference specimen. The surface
tester in the lab gets calibrated by a specimen with an Ra value of 116μin
(2.95μm). The cutoff lengths you can choose are .01in (.25mm) .03 in (0.8mm)
and 0.1in (2.5mm). I found that if you use the shortest cutoff length, the
reading is inaccurate. The longest cutoff length returns an error more often than
not, and the middle length returns a reasonable result. (There is no other tester
to compare accuracy for…)

This report has barely “scratched the surface” of metrology but gives a
quick insight into this interesting and indepth field.
Ryan Bourget 8

You might also like