Common Carbide
Defects
Internal
The most common and
most serious problems are internal.
Why This is
Important
The customer
reported some
chipping and breaking during grinding. It wasn't
all the time and some times it was a lot worse than
others. The carbide supplier had examined the parts
and said that there was nothing wrong with them. We
received the parts and sent them to an outside
source. They polished the surface off and took
pictures of the interior. The most visible features
were internal cracks. The crack in the middle
picture runs though a pinhole. There were also
pinholes without cracks.
The tips also had
internal voids or holes. Finally the carbide grains
were not evenly distributed
in the cobalt matrix so there is a puddle of cobalt
with no carbide in it.
Explanations
The first series of
photos supplied by Valenite Die and Wear Walmet
Copyright © 1999-2002 and used with permission. The
remainder are from our labs.
Clusters
(Cls)- Clusters are defined as groups of three or
more WC grains that are significantly larger than
the average grain size. A WC cluster can be a weak
spot in the carbide microstructure. Clusters are
thought to form during the cooling cycle through WC
crystallization. It is not completely clear what
controls this crystallization but small impurities
in the powder and non-uniform carbon distribution
have been implicated.
It is difficult, if not impossible, to avoid
clusters completely. Low levels are not considered
harmful to the integrity of cemented
carbide parts. Large numbers of these clusters can
adversely affect performance, especially where shock
is involved.
Binder Lakes (Blk) -
Binder lakes are pools of cobalt binder in the
microstructure. They are formed when melted
cobalt or nickel binder flows into open pores during
sintering (1345°C+). Sinter HIP'ing eliminates
binder lakes by sintering at high temperatures under
an inert gas pressure of approximately 700psi. The
pressure forces carbide grains, along with the
binder, into the open pores. Low levels of binder
lakes are not considered harmful to performance, but
a large number of lakes may structurally weaken a
cemented
carbide part.
Eta
Phase (Eta-1, Eta-2, Eta-3) - Valenite's internal
rating system for eta phase. Eta phase is a carbon
deficient form of tungsten carbide that results in a
harder, more brittle cemented
carbide part. Insufficient carbon levels are
generally the result of improper formulation of the
carbide powder, long term exposure of unsintered
parts to the atmosphere, or poor control of
sintering conditions. We rate eta phase on a scale
of 0 to 3. Zero indicates that no eta phase is
present, eta -3 indicates the most severe level. Eta
phase is generally considered to be harmful to the
performance of cemented
carbide parts. With our typical 5 minute etch, the
eta phase is rapidly etched leaving a void with
characteristic geometric patterns.
Grade
Contamination: An area of a distinctly different
grade in the microstructure, round to oval in shape,
whose longest axis exceeds 25 microns. Cross Grade
Contamination is generally the result of ineffective
cleaning of powder processing equipment. Low levels
of grade contamination are not considered harmful to
performance, but a large number of these areas may
adversely alter the physical properties of the cemented
carbide part.
Porosity
A
Porosity B
Porosity C Porosity
A Porosity: Pores in
the microstructure less than 10 microns in diameter.
Rated from
A01 to A08.
B Porosity: Pores in
the microstructure 10-25 microns in diameter. Rated
from B00 to B08.
C Porosity: Not
true porosity. Rather, carbon porosity consists of
discrete areas of graphite in the microstructure
resulting from an overabundance of carbon. Rated
from C00 to C08.
Free Carbon: A term
used to describe C Porosity in excess of C00
Microphotographs
of Bad Material
500 x magnification
Top right shows a
very large porosity. This material looked good
until it was ground. These pores were hidden under
the surface. During grinding pores like this opened
up. The bottom two photos show cracks in the
material. These cracks were in the material as
supplied. In this case about 30% of the material
broke during the initial sharpening
This
is at 1500x magnification This is a photo of the
surface and the white is binder. The small black
specks in the top right photo are areas of
porosity. The bottom right photo shows a very large
A porosity about ten microns in diameter. The
bottom left photo shows uneven distribution and
contamination by foreign materials.
1500x
magnification The top left photo shows the average
quality of this material. There are at least two
oversize grains of material. The top right photo
shows a very large grain of the material. The bottom
two photos show B porosity.
Inside of a good tip Inside
of a bad tip Close up of bad
tip
Small
regular grains, Lots of
porosity Grain size is
irregular
nice
tight structure these
holes mean which means a
weaker
no big
voids a much
weaker tip tip and poor
wear
GREEN FRACTURE: A
fracture that developed before the part had been
fully sintered. Green fractured surfaces are coarse
when compared to hard fracture surfaces.
HARD FRACTURE: A
fracture that developed after the part had been
fully sintered. Hard fractured surfaces have a
smooth texture and usually contain ripples or
How Bad Carbide
Breaks & Wears Out Faster
Good
Carbide small, regular grains locked tightly in a
matrix.
Porosity & binder
lakes lots of holes create weak areas susceptible
to cracking
Grade contamination,
reground powder & other odd large pieces
Eta Phase Carbon
depleted
which means there are areas where the grain
structure is weak
Bad carbide wears
out much faster because the holes and weak spots
direct and concentrate the forces into the carbide
23. External
Problems
Green State
Problems
As part of the
manufacturing process carbide has a stage where it
is very soft. This is where chipping, bending,
cracking and similar damage occurs most often.
Crack Recognition
This is what a half
inch saw tip and various size cracks look like blown
up to 20 times actual size.
Cracking
Any piece of carbide
showing a crack should not be used. This is also
true if there is a corner knocked off. There are
two arguments. One is that the part is acceptable
if the material will be ground down anyway. The
other argument is that the force that caused the
crack or loss of a corner may have created
smaller cracks that are not visible and the part
should not be used.
Chipping
and Edge Radius
This is about the
only thing suspicious and is probably best
characterized
as a surface blemish rather than a
crack.
Chips, edge
radius and parallelism
These all looked
good to naked eye inspection. We used 30x
magnification for inspection with standard light
and black and white high contrast. These are
picture of 5 separate tips fixtured between the jaws
of dial calipers.
The gap on the right
side of this tip is about 0.005 with the largest
chip about 0.002. On the left side the gap is
about 0.002 The gap is a comibnation of the
flatness of the side, the parallelism of the sides
and the radius of the edeges.
Chipping, Corner
and Edge Radius Standards
Saw Tips Need To Be
Flat And Square. A radius of 0.001 to 0.002 is
common on competitively priced production tips from
good suppliers.
Tip 1
These were taken at
30X with Proscope. They are tips held between
tweezers.
There is one edge
deliberately rounded. As near as we can tell the
rest of the edges have a radius of 0.001 and maybe
up to 0.002
Tip 2
This
time we set a dial caliper to 0.005 when we took
the pictures. The caliper is resting against a
deliberately rounded edge.
The two edges
facing you are maybe 0.001 to 0.002
Establishing the
Standards
We used the figures
of Top grind 0.015, Face grind 0.005and
Side
grind = 0.005 assuming grinding in that order.
Face view - tip is
0.150 wide
Grind top & sides
Finding the maximum
allowable radius ends up being a matter of adding
the two grinding tolerances to get 0.020.
Chips On Stump
Grinder Tips
These routinely come
in with some chipping. Those you sent back were
worse than usual
I
would suggest you specify these when you order them
and send us a copy of the specification. We will be
happy to inspect them for you.
I would suggest that
you specify then as:
No chip larger than
0.010 wide x 0.005 deep.
No more than 2 chip
per tip
No more than 5
chipped tips per hundred
One chip appears to
be about .180 by 0.060
and there are chips
all along the edges.
We have chips on the
current tips. The biggest chip we found was .030 x
.015
Measuring Cracks and Chips
Crack Detection
and Inspection Techniques
Visual
Naked eye, assisted
by magnifying glass, low-power microscope, lamps,
mirrors.
Only at places
easily accessible. Detection of small cracks
requires much experience.
Saw tip naked eye
The cracks you saw at 30x The cracks you
didnt see at 30x
(Note: these cracks
were caused , at least in part, by too much heat
from grinding.)
Penetrant
Colored liquid
penetrant is brushed on material and allowed to
penetrate into cracks. Penetrant is washed off and a
developer is applied. Remnants of penetrant in crack
are extracted
by developer and give colored line. Only at places
easily accessible. Sensitivity on the same order as
visual inspection.
Magnetic particles
Part to be inspected
is covered with a layer of a fluorescent liquid
containing iron powder. Part is placed in a strong
magnetic field and is observed under ultraviolet
light. At cracks, the magnetic field lines are
disturbed. Applicable only to magnetic materials.
Parts have to be dismounted
and inspected
in a special cabin. Notches and other irregularities
give indications. Sensitive method.
X-ray
X-rays emitted
by portable X-ray tube pass through structure and
are caught on film. Cracks, absorbing less X-rays
than surrounding material, are delineated
by black line on film. Has great versatility and
sensitivity. Interpretation problems arise if cracks
appear in fillets or at edge of reinforcement. Small
surface flaws in thick plates are difficult to
detect.
Ultrasonic
The probe
(piezoelectric crystal) transmits a high-frequency
wave into material. The wave is reflected
at the ends and also at a crack. The input-pulse and
the reflections are displayed on an oscilloscope.
Distance between first pulse and reflection
indicates position of crack. Interpretation:
Reflections of crack disappear upon change of
direction of wave. Universal method since a variety
of probes and input pulses can be selected.
Information about the size and nature of the defect
(which may not be a crack) is difficult to acquire.
Eddy current
Coil induces eddy
current in the metal. In turn this induces a current
in the coil. In the presence of a crack, the
induction changes; the current in the coil is a
measure of the surface condition. Cheap method (no
expensive equipment) and easy to apply. Coils can be
made small enough to fit into holes. Sensitive
method when applied by skilled person. Provides
little or no information about nature and size of
defect.
Holes in the
Carbide
Left - Triangle at
bottom leads into a
line Right line
of holes
which turns into a
line of holes
In
this case the tips were sintered improperly. It
appears that the coating on the graphite sintering
trays was done improperly or not done at all. Parts
of the tips have pulled out where they stuck to the
graphite boat. This is only on the side of the tips
that stuck to the boat so it is not all sides of all
tips.
After sintering the
tips were sandblasted. Then the manufacturer y
oxidized them in an attempt to make them braze
well. The surface as supplied did not braze well.
When we cleaned the bad surface to prepare them for
pretinning this defect became readily apparent.
Bent Corners
# 4
Railback corners are badly bent - #8 look much
better.
Size of parts
There is still a lot
of art in making carbide parts however a good
carbide company should be able to deliver parts
within specifications 100% of the time. This is
tough but possible because some companies do
actually do this. Typically the specifications are
given to cover a range so that a dimension of .200
may be specified as +.003" to +.007". It is ideal
if all the parts are identical but a difference of
.001 so that some parts measure .003 and some
measure .004 is generally not significant
Example: WFC
7170
These are good tips
except for being a bit undersize.
Count is 250 which
is what it is labeled. 251 252 is preferred but
250 is acceptable.
WFC 7170
Thick length
0.171
nominal .125 nominal .500
0.169
0.130 0.510
0.169
0.130 0.508
0.170
0.130 0.509
0.170
0.130 0.508
0.169
0.130 0.509
0.169
0.171
0.169
0.170
Mixed parts
Are all the parts in
the box what you ordered? We occasionally find a
drill bit in with saw tips, for example. I dont
think Id worry about an occasional part but if
there are too many mixed parts it might be
indicative of a generally sloppy operation.
Box count vs.
Actual count
It is very seldom
that the box count exactly matches the actual count
but its probably not that big of a deal. These
parts are weigh-counted
on scales that are generally + or - 1/2 of 1% (1/2%)
so the scales are not dead-on. We can do counting
that is dead-on but we usually have to hand count
three times. We did a test for a customer that ran
six months and covered more than 500,000 parts.
After six months we showed that the carbide
manufacturer had shipped 115 extra parts. The
problem is
