Surface Texture Measurement and Characterization

1
Surface Texture Measurement and Characterization

• Profile Measurement
• Preprocessing
• Filtering
• Numerical Evaluation
• - Roughness Ra
• - Waviness Wa
• - Profile Pa

2
Surface Texture Parameters

• Average Parameters
• Height Parameters
• Spacing Parameters
• Hybrid Parameters
• Functions

3
(No Transcript)
4
Amplitude Parameters
Ra ..Roughness Average (also AA and
CLA) Rq...Root mean square roughness (also
RMS) Rti RtRoughness Total Rymax(Ry)
(Rmax) Rtm..Roughness Peak Rp...Roughn
ess Peak Rpi Ppm.Roughness Peak Mean Rpymax
(Rpmax) (Rpy)
5
Rv..Roughness Valley Rvi. Rvm..Ro
ughness Valley Mean Rvymax (Rvmax)
(Rvy) R3z.. R3y (R3ymax) (R3max) R3i. Rz
..Ten Point Height Mean RHR..Roughness
Height Range RskSkew RkuKurtosis
6
HtpHeight between Bearing
Ratios H..Swedish Height R W
consisting of 4 mean values R
Ar W... Aw. 4 maximum
values Pt. Wt Wmax. Aw..
7
Surface Finish Definitions
8
Mean Line
The mean line is defined as the line which
bisects the profile such that the area above and
below the line equal and a minimum.
9
Terminology
Typical for an analog filter
10
Typical for a digital filter
Sampling length Cut-off Filter Cut-Off
For unfiltered measurement. Sampling Length
Assessment Length
11
Ra Ra Roughness Average Older terms
AAArithmetrc Average Ra is the
Arithmetic Average of the distance of the
filtered or unfiltered Roughness profile from
its Mean Line
12
Ra
Analog Systems
Continuous
Digital Systems
Discrete
Where L is the assessment length and N is the
number of Data points
Advantages Statistically very stable
parameter. Easy and
cheap to implement into instrumentation.
Good for random type
surfaces (Processes like grinding) Disadvantages
Not a good discriminator for different types
of surfaces Looses
its usefulness when skew (Rsk) exceeds
1.5 WARNING Ra is not the same as RMS
(Rq). Adding 11.1
to the Ra reading, or calibrating the
instrument 11.1
higher is incorrect except for sinusoidal
surfaces
13
Two different surfaces with the same Ra
14
(No Transcript)
15
Rq Rq...Root Mean Square
Roughness Average Older terms RMS ...Root
Mean Square, Still widely used in some industries
(i.e. optical)
Rq is the Root Mean Square of the distance of
the filtered or unfiltered Roughness Profile
from its mean line. Rq is the standard
Deviation of the Amplitude Density Distribution.
Because this Parameter squares Amplitudes it is
more sensitive to peaks and valleys.
16
Rq
Analog Systems
Continuous
Digital Systems
Discrete
Where L is the assessment length and N is the
number of Data points
Advantage Statistically very
stable. Disadvantage Only available in middle
to higher priced instruments. WARNING Rq. is
not 11.1 higher than Ra
17

• Rt, Rp, Rv
• Rt Total Roughness
• Rp Roughness Peak
• Rv Roughness Valley
• Rt is the highest Peak to the lowest Valley in
  the assessment length
• Rp is the highest Peak in the assessment length
  Rv is the lowest Valley in the
• assessment length
• All three Parameters can be used Filtered or
  Unfiltered In Filtered Mode Rp and
• Rv do not have to be in the same sampling
  length

Rt Rp Rv
18

• Rt, Rp, Rv
• These Parameters are essentially divergent as
  there is no guarantee that the
• Measurement Line crosses the extremes of the
  surface.
• For this reason these Parameters tend to vary
  considerably on multiple measurements.
• Large errors of measurement can also be
  introduced when using a Skidded Instrument
• in the presence of significant or isolated
  Peaks or dirt.
• Peaks are important in the study of Friction
  and Wear as the interaction between
• surfaces concentrates around them.
• Presence of Peaks can make dimensional
  measurements on Parts subjected to Wear
• unreliable, as the Wear removes the Peaks,
  which were originally included in the
• measurement.
• Valleys are significant for Lubrication
  properties.
• Propagation of Fractures and Corrosion start in
  valleys.

19
Rti, Rtm, Rymax, Rpm, Rvi, Rvm, Rvmax
The averaging used for these Parameters makes
them more stable than pure extreme
values. Rti is the maximum peak-to-valley
height in the sampling length i. Rtm is the mean
of all the Rti values obtained within the
assessment length.
20
Rti, Rtm, Rymax, Rpm, Rvi, Rvm, Rvmax
Where n is the number of sampling length or
cut-offs.
Rymax is the largest Rti value within the
assessment length. Rymax (ISO) Rmax (DIN)
Ry (RTH)
21
Rti, Rtm, Rymax, Rpm, Rvi, Rvm, Rvmax
Rpi is the maximum peak height in the
sampling length i. Rpm is the mean of all the
Rpi values obtained within the assessment length.
Rpmax Rp is the largest Rpi value within the
assessment length. Rvi is the maximum
valley depth in the sampling length i. Rvm
is the mean of all the Rvi values obtained
within the assessment length.
Rvmax Rv Is the largest Rvi value within the
assessment length.
22
R3z, R3y, R3i, R3ymax, R3max.
R3z was originally developed in Germany, with the
intent to reduce the instability of the Peak
Parameters by ignoring Profile extremes.
R3yi The deviation from the third
highest peak to the third lowest valley in the
sample length I, where vertical and horizontal
thresholds have to be exceeded. R3ymax
Is the highest R3yi within the assessment length
.
In order to
streamline the parameter description with the
Rymax Paramater one should assume that
R3ymax (ISO) R3max
(DIN) R3y (RTH) R3z Is the mean of
all the R3yi values obtained within the
assessment of length.
23

• Rku Kurtosis
• Can detect if Profile "Spikes" are evenly
  distributed above and below the Mean Line.
• It provides a measure of the sharpness of the
  Surface Profile or ADF.
• Spiky surfaces have a high Kurtosis value
  Rkugt3, bumpy surfaces have a low
• kurtosis value Rkult3.
• A perfectly random Surface has Rku3.

24

• 
  Rsk
• Rsk..Skew
• Rsk - Skew - is the measure of the symmetry of
  the Profile about the Mean Line.
• It will distinguish between asymmetrical
  Profiles of the same Ra, Rq, Rtm, Rymax, Rt.
• Negative Skew indicates a predominance of
  valleys, while positive Skew will be seen on
• "peaky" surfaces.
• Typically
• Plateau Honing Rsk negative
• Grinding ..Rsk zero
• Turning, Bead Blasting Rsk positive

Continuous
Analog Systems
Digital Systems
Digital
25

• Rsk
• If Rsk exceeds 1.5, Ra should not be used as
  the only control Parameter.
• This condition also indicates that Phase
  Corrected Filter should be used.
• Surfaces with positive Rsk can give large
  measurement errors if Instruments with Skids
• are used.

Negative Skew
Positive Skew
Rsk is a non-dimensional parameter
26

• Rzi, Rzy, Rz, Rzm
• The original intent in creating the Rz
  Parameter was to allow the measurement of very
• short surfaces, where by choosing a cut-off
  length to suit the available surface, the
• information lost due to the filtering makes
  this prohibitive.
• Rz is also known as the ISO 10 point height
  parameter.
• It is measured on the unfiltered profile only
• It is numerically the average height difference
  between the five highest peaks and the
• five lowest valleys within the assessment
  length, where vertical and horizontal
• thresholds have to be exceeded.

27

• Presently there is an ongoing heated
  discussion whether the Rz parameter should
• actually be filtered or unfiltered.
• This led to the Rzm parameter, which now
  logically has to extend to Rzi and Rzymax
• (or Rzmax or Rzy).

28
Spacing Parameters
S Mean Local Peak Spacing Sm..Mea
n Peak Spacing HSCHigh Spot
Count PC...Peak Count K-Factor.Rando
mness
29
S S...... is the mean spacing between adjacent
local peaks, measured over the assessment length.
(A local peak is the highest part of the profile
measured between two adjacent minima, and is only
included if the distance between the peak and its
preceding minima is at least 1 of the Rt of the
profile.
Where n is the number of peak spacings.
30
Sm Sm....... is the mean spacing between profile
peaks at the mean line, measured over the
assessment length. (A profile peak is the highest
point of the profile between an upwards and a
downwards crossing of the profile of the mean
line.)
Where n is the number of peak spacings.
31
HSC HSC...... The high spot count is the number
of complete profile peaks (within the assessment
length) projecting above the mean line, or a line
parallel with the mean line. This line can be
set at a selected depth below the highest peak or
a selected distance above or below the mean
line. HSC is always expressed in number of
counts per assessment length.
This parameter has been found useful in
applications where the penetration of asperities
through a threshold can be critical. A typical
example would be peaks exceeding a flight height
level for memory disks. Peaks that are to high
can cause a crash of the reading head.
32
Pc Pc...... The peak count is the number of local
peaks which project through a selectable band
centered about the mean line. The count is
determined only over the assessment length though
the results are given in peaks per cm (or per
inch). The peak count obtained from assessment
lengths of less than one cm (or 1 inch) is
obtained by using a multiplication factor. The
parameter should therefore be measured over the
greatest assessment length possible.
This parameter is mostly used in the USA. It has
been found extremely useful for painting, forming
and plating applications.
33
HYBRID PARAMETERS
tp Bearing Ratio, Material
Ratio Mr.DIN Symbol for tp Rk For
describing the Material Portion in the roughness
Profile divides into the following Parameters\
RkCore Roughness Depth
Rpk..Reduced Peak Height Rvk
.Reduced Valley Depth Mr1 Material
Component relative to Peaks Mr2.Material
Component relative to Valleys
A1...Material filled Profile Peak Area
A2Lubricant filled Profile Valley
Area laAverage Wavelength lq.RMS
Average Wavelength Da..Average Slope Dq.RMS
Average Slope
34
Material Ratio Curve The Material ratio curve,
formerly called Abbot-Firestone curve or Bearing
ratio curve, is a graphical representation of the
tp parameter in relation to the profile level.
This curve contains all the amplitude information
of a profile. It is a close relative to the
amplitude density function (Integration of the
ADF yields the material ratio curve. As the shape
of this function varies significantly relative to
different processes the description of it's shape
can provide very detailed information about a
surface profile. See Rsk, Rku and RK parameters.
The abscissa can be displayed in two ways As a
linear axis (see above picture). As a cumulative
Gaussian probability axis. This is particularly
useful for random surfaces as the plot becomes a
straight line. For the assessment of multiple
processed parts, with the individual processes
being Gaussian each process is shown as a
straight line.
35

• la lq
• Average wavelength, is a measure of the spacings
  between local peaks and valleys,
• taking into account their relative amplitudes
  and individual spacial frequencies.
• Being a hybrid parameter, determined from both
  amplitude and spacing information,
• it is,, for some applications, more useful than
  a parameter based solely on amplitude or
• spacing.
• For a profile that is not sinusoidal this
  parameter must not be confused with true spacing
• parameters like S or Sn.
• Average wavelength takes every point of the
  profile into account.

36

• la lq (contd)
• Average wavelength can be particularly useful in
  applications where the presence of certain
  harmonics on parts changes with time of usage.
• Closely spaced irregularities of a surface are
  normally of a relatively small amplitude, but
  wear rapidly with the part being used like in
  roller or ball bearings or rotational or
  reciprocational friction applications.
• This leaves the lower frequency components of the
  surface more dominant.
• In process control average wavelength is useful
  to monitor the condition of grinding wheels, as
  it relates to the average grit size.
• In turning the condition of the tool edge and
  machine feeds can be monitored.
• Average wavelength is related to the power
  spectrum and represents an estimate of the
  weighted mean of the Fourier spectrum.

37
(No Transcript)
38
Now we can finally put it all together into one
single picture
39
n The shaded area is computed (Area3) o Rvk is
calculated as the height of the triangle with an
area equal to the shaded Area3. The base of
this triangle is DF.... which is 100 - Mr2.
o Shows Rvk exceeding below the lowest valley of
the profile. This is correct, it can, but it
does not happen all the time. It is also
possible that Rpk goes higher than the highest
profile peak. Adding up Rpk RK Rvk does
NOT equal Rt.
40

• This parameter attempts to numerically evaluate
  the material ratio curve of surfaces manufactured
  
• with a process resulting in a negative skew.
• The processes that this parameter relates to in
  particular are Plateau Honing, Lapping and all
  kinds of
• multiple machining operations which intend to
  remove peaks but leave larger valleys from a
  previous
• process.
• The parameter originated in Germany and has been
  implemented in the Automotive Industry.
• A strong argument as to the validity of RK and
  it's associated parameters is continuing, but if
  the
• Industry asks for it - We got it.
• The way Rk is derived is essentially twofold
• 1 ... A special filtering technique is applied
  which is intended to minimize the residual
  distortions of the filtered profile still
  existing after phase corrected filters have been
  applied.

41
There are 5 primary parameters Rk Core
Roughness Depth The depth of the Roughness Core
Profile. Sometimes called "Kernel". This is the
working part of the surface. It will - after the
initial running in period - carry the load
and influence life and performance. Rpk Reduced
Peak Height The top portion of the surface which
will be worn away in the running in
period. Rvk Reduced Valley Depth The lowest part
of the surface which has the function of
retaining the lubricant. Mrl Material Component
relative to Peaks The Material Ratio at which Rpk
and RK meet. This is the upper limit of the Core
Roughness Profile. Mr2 Material Component
relative to Valleys The Material Ratio at which
Rvk and RK meet. This is the lower limit of the
Core Roughness Profile.
42
A final word about the dimensions of these
parameters Rk in mm (micrometer) or min
(microinches) Rpk in mm (micrometer) or min
(microinches) RVK in mm (micrometer) or min
(microinches) M,l in M,2 in For Al and A2
the following applies The areas computed in j,k
and n,o are over the assessment length of the
filtered Roughness Profile. To unify this number
it is expressed in A I.......... in mm2/mm or
min2/in Al........... in m m2/ mm or min2
/in Meaning that the Al and A2 parameter were
computed as if the measurement had been taken
over one millimeter (or inch). min stands for
milliinches (0.00x inches or thousands of an
inch). Because numbers tend to get rather high
m in2/ in is rarely used There are 1000000 min
2 in I min 2
43
(No Transcript)
44
Gaussian Plotting of Material Ratio Graph
The process transition point helps in putting a
numerical surface finish value to different
functional parts of the surface. The profile
height between say 0.05 material ratio (to
ignore stray extremes) and the process transition
point is approximately four times Rq.
Reference Whitehouse D.J. Assessment of surface
finish profiles produced by multi-process
manufacture. Proc Instn Mech Engrs 1985 Vol 199
No B4 p.263-270
45
ISO 4287Geometrical Product Specification (GPS)
Surface Texture Profile method Terms,
definitions and surface texture parameters

• Scope
• The standard specifies terms, definitions and
  parameters for the determination of surface
  texture by profiling methods.
• Summary
• Terms and Definitions
• General terms
• Geometrical parameter terms
• Surface profile parameter definitions
• Amplitude parameters (peak and valley)
• Amplitude parameters (average of ordinates)
• Spacing parameters
• Hybrid parameters
• Curves and related parameters

46

• Amplitude - Rp, Rv, Rz, Rc (mean value of profile
  points), Rt, Ra, Rq, Rsk, Rku
• Spacing RSm
• Hybrid - R?q
• Curves Abbot firestone curve, amplitude curve
  etc

47
Stratified Surfaces

• ISO 13565- 1, 2 and 3 for plateau-honed surfaces
• ISO 13565 1 two stage Gaussian filtering (Rk
  filter) to obtain a more robust mean line

48
Stratified Surfaces

• ISO 13565-2 Parameters based on the Linear
  material ratio curve
• Rk, Mr1, Mr2, Rpk and Rvk

49
Stratified Surfaces

• ISO 13565-3 Parameters based on Probability Curve
  
• Rpq, Rvq and Rmq

50
Surface Texture Symbols
Max Waviness Height
Max Waviness Width
0.002.-2
Max Ra
63 32
cut-off Lay
0.030 0.015

Min Ra
Max Roughness Spacing
51
Surface may be produced by any method
Material removal by machining is
required Material must be provided for that
purpose.
Material removal allowance in inches (or
millimeters)
0.001
Material removal prohibited
52
Calibration

• Calibration Specimens
• NIST Procedure
• PTB Procedure

53
Roughness Specimens
54
(No Transcript)
55

• Calibration with the aid of Roughness Standards
• Carry out 12 measurements on each roughness
  standard, distributed over the measurement
• surface as shown in the measuring points
  pattern (Figure 2).
• Calculate the arithmetical mean for each
  roughness parameter.
• Give the deviations from the mean values stated
  in the PTB calibration certificate in percent and
  
• microns.

56

• NIST Calibration Procedure
• 4000 Data points
• 2RC Filter
• Measured at 9 locations
• Each measurement repeated twice

57
PTB Calibration Procedure
58

• Measurement Practice
• Parameter Variation
• Vibration
• Gage R R

59
(No Transcript)
60
(No Transcript)
61
(No Transcript)
62
Repeatability and within-sample variation
63
Da Dq Average slope is the mean (arithmetic or
rms) of the slopes at all the points of a profile
within the assessment length. The slope of a
profile is the angle (expressed as a gradient)
between that profile and the mean
line. Being the slope of the profile at a
given point. These parameters have been found
useful in assessing contact, optical and
frictional properties as there is a relation to
hardness, elasticity, electrical and thermal
conductivity, plastic and elastic deformation,
reflectivity, friction, adhesion and
others. Average slope is also used to compute
true profile length (See appropriate
sheet). These parameters are sensitive to the
numerical model used in computing slopes. Large
differences in computed values can show between
different instruments.