) and
tool
off-
t the
t cost
agrees
n line,
and
d be
of
r the
uring
acing
and
aking
taken
area
the
nworn
d).
n of
the
light
ghted
level
the
1:- in
level
the
as:
image
aking
that
picture processed 43 shown as the Fy.3 of the
appendix.
3.Bulding gauging coordinate system
In the theory and practice, the tool orthogonal
plane coordinate system is often used to gaug, design
and check the angle of cutting tool according to the
Internation Oarganization for Standardizalion.
Because this research only uses two dimensional
coordinate for measurement we build the gauging
coordinate system as follows Fig3.1
0 >
Cutting Tool
y Major: Flank
Fig 3
In the coordinate, the coordinate point is on the
tool corner, the axis of abscissa is parallel to
tool reference plane, and the axis of ardinate is
vertvcal to tool reference plane. These axes are
called relative axes. While the axes, which are
definited by imagion system, are called absolute axes,
In absolute system, the axis oj refers to vertical
direction and axis oi refers to line scanning
direction in the computerimage system, These two
coordinace systems shoued be posited correctly
follows as Fig.3.2
o. +
/ a
De
Fig 3.2
But, the xoy system often takes correct pasition
to ioj system showed as Fig3. 3 when taking
micrograph. In this way, gauging erroes have been
produced.
It is nessary to adjust the relative position
between ioj and xo’y. The o’xy system must be build
at the practical position of cutting tool in taking
picture, and then rotated or moved until the axes
0’xy take correct position to oij system, Taking
lath cutter with 0° tool cutting edge inclination as
an example, The processing of building and
transforming relative coordinate system is given as
follows.
Oo".
Fi4 3.3
Fig3. 3 shows the position of cutting centre in
microscope visual field. To constructe o’xy
coordinate, the Lines, li and 12, which are the
projctions of the face and the major flank of cutting
tool in o’xy plane, should be produced, then the
intersection point of l|] and l2 could be determind.
This point could be refered to as cutting corner
after wearing,
After cutting procession, the parts of the face
and the major flank around the corner have been worn
partly. So the projectors of the two planes in o’xy
are irregular curve, We take regression method to
gain l| and 12.
Cutting corner o’ appears a little curve in
micrograph. Basing on caculating every slope of the
point on curve, the point which has the largest
slope could be selected, Then, the whole picture of
the wear being taken by camera is rotated around the
picture’s weight centre until the point with largest
slope is placed on the highest position of the
picture, From the highest point to the lowest point
of the curve, we can make regression respectively in
the two direction along the right and left of the
highest point.
It be supposed that the points which take part in
regression are on the edge of the worn area. We can
take several x with unequal values(x],x2,:-,xn) to
test independently, then we could gain a series of
the samples: (x1,y1), (x2, y2), =, (xn, yn). IF there are
2
y~NCa+bx, ¢ ) for every x, or y=atbxit& i, 8 i-NCo,
2
oc), i=l, 2,-,n for every point(xi, yi) , the
probability density function is:
1
2
exp [- —(yi-a-bxi )]
2
oon 20
Because yl,y2,:-, and yn are independent for each
other their coalition's probability density is:
n 1 1
2
L-II —————exp[- —«yi-a-bxi) ]
2
i=l oVèT 20
1 1 n
n 2
=( ) exp [- — Z2 (yi-a-bxi>] 3-1)
2
om 2g i-1
In above formula, a and b can be estimated by proble
method. Obviously, when L has the largest value the
n
function QCa, b)= 27 (yica-ixiX has the least value.
i=1
The following coupled equations can be obtained by
deriving Q to a and b respectively:
20 n
—=-2 T(yi-a-bxi)=10 (3-2)
2a i=l