The Designers Guide to Tungsten Carbide

General Carbide
1151 Garden Street
Greensburg, PA 15601
July 2008
Introduction to
The Designers Guide to Tungsten Carbide
This publication is a reference guide for designers, engineers, fabricators and end users of
tungsten carbide material. It is a compilation of recommendations derived from practical
experience, theoretical stress analysis, proven application engineering practices and
modern manufacturing techniques. It will enable the reader to gain insight into how best
to utilize one of the most unique engineering materials available today.
It has been said that the tool materials of one generation become the engineering
materials of the next generation. This observation is certainly true of tungsten carbide. It
is a material that has been around since the early 1920s, replaced tool steel in most
cutting tool applications and developed into an engineering material used to resist the
harshest environments of corrosion, high temperature, impact, high compressive loads,
deformation, and severe abrasion. It competes with advanced ceramics in the wear parts
arena but just as tungsten carbide did not totally displace tool steel, advanced ceramics
cannot replace tungsten carbide. It exhibits superior toughness given its high hardness.
Tungsten carbide has enjoyed tremendous growth as a tooling and engineering material
and this manual should provide the designer with recommendations needed on how best
to design for success and reliability.
Thomas R. Shearer
Director of Business Development
General Carbide
1151 Garden Street
Greensburg, PA 15601
www.generalcarbide.com
The Designers Guide to Tungsten Carbide
Table of Contents:
Chapter I.... Background of Cemented Carbide
Chapter II.... Unique properties of Cemented Carbide
Chapter III.... Design Considerations
Chapter IV.... Attaching and Assembling Techniques
Chapter V.... Finishing Techniques for Cemented
Carbide
Copyright 2008 General Carbide



Chapter I - Background on Cemented Carbide

What is cemented carbide?

Tungsten carbide (WC), also referred to as cemented carbide, is a composite material
manufactured by a process called powder metallurgy. Tungsten carbide powder,
generally ranging in proportion between 70%-97% of the total weight, is mixed with a
binder metal, usually cobalt or nickel, compacted in a die and then sintered in a furnace.
The term cemented refers to the tungsten carbide particles being captured in the
metallic binder material and cemented together, forming a metallurgical bond between
the tungsten carbide particles and the binder (WC - Co), in the sintering process. The
cemented carbide industry commonly refers to this material as simply carbide, although
the terms tungsten carbide and cemented carbide are used interchangeably.

If the permanent deformation of a material at failure is quite small, the material is labeled
brittle; if the plastic deformation is very large, the material is called ductile. Carbide is
classified technically as a brittle material since it exhibits little or no plastic deformation
preceding the initiation of a crack and total failure. Without the presence of the metallic
binder phase, tungsten carbide could be considered a ceramic material much the same
as silicon carbide or aluminum oxide. The definition of a ceramic material is the
marriage of a metal to a nonmetal, for example silicon (metal) carbide (carbon, non-
metal), aluminum (metal) oxide (oxygen, non-metal), or silicon nitride. A cermet is a
composite material composed of ceramic (cer) and metallic (met) materials. It is the
addition of the metallic binder, i.e. cobalt or nickel that makes the cemented carbide (WC
- Co) a cermet and differentiates it from truly brittle materials, that is, the ceramic family
of materials.

Cemented carbide is the preferred material for parts that must withstand all forms of
wear (including sliding abrasion, erosion, corrosion/wear and metal-to-metal galling) and
exhibit a high degree of toughness. It exhibits high compressive strength, resists
deflection, and retains its hardness values at high temperatures, a physical property
especially useful in metal-cutting applications. It provides long life in applications where
other materials would not last or would fail prematurely.


The history of cemented carbide

Since the late 1800s when a French chemist, Henri Moissan, first synthesized it,
tungsten carbide has been known as one of the hardest substances in existence,
approaching diamond in this respect. In fact, he was seeking to produce man-made
diamonds, but WC was the result. Since large solid pieces could not be produced, cast
compositions containing tungsten carbide were tried, but were too brittle and porous for
use as an engineered material.

Search for a substitute to replace the diamond dies employed in drawing tungsten wire
for electric-lamp filaments led the Osram Lamp Works of Berlin to an interesting
discovery. Karl Schroeter and Heinrich Baumhauer found that hard carbide, bonded or
sintered together with a metal such as cobalt, was not only hard but acquired enough
toughness to suggest its use as a cutting tool. The substance discovered by the Osram
Company was known as Hartmetall and pointed the way to the development of the
modern sintered carbides. Friedrich Krupp, A. G. Essen, acquired the original patent
rights and undertook an extensive program culminating in the production of Widia (Wie
Diamant
1-1
, like a diamond), which consisted mainly of tungsten carbide particles
thoroughly interspersed in a cobalt matrix, which constituted from 5 to 15 percent of the
total composition.




In negotiations with Krupp, all American rights accrued to General Electric, with Krupp
retaining the right to export Widia to the United States. General Electric formed the
Carboloy Company, which sublicensed Firth-Sterling Steel Company and the Ludlum
Steel Company. At that time the American equivalents of Widia were known as
Carboloy, Dimondite, and Strass Metal. Later, the trade name Dimondite was changed
to Firthite. American patent applications were made as early as 1922, but it was not until
1926, after thorough production tests in the Essen workshops, that Krupp marketed
Widia tungsten carbide commercially for the first time in Germany. Tungsten carbide
was expensive, costing about $450 per pound, but even at that price its use could be
justified economically. The practice of making only the tool tip out of cemented carbide
was dictated as much by the cost of the material as it was by any other single
consideration. Carboloy tools were tested in General Electric plants and came to public
attention around 1928.

Following the introduction of Widia and its counterparts, Dr. Balke developed a tantalum
carbide bonded together with metallic nickel. This material, called Ramet, resisted
cratering and proved more successful in machining steel than did tungsten carbide,
which exhibited a tendency toward early cavitation near the cutting edge, where steel
chips came into intimate contact with the tool face. It was this characteristic of the first
carbides that impeded an advance in their application to machining steel comparable
with the strides made in machining cast iron, nonferrous metals, and abrasive materials.
Thus tantalum-bearing carbide opened new fields in steel cutting. Single and multiple
carbides of tantalum, titanium, and columbium were also to find use as crater
preventives or resisters. In 1935, Philip McKenna, working at the time as a metallurgist
for Vanadium Alloy Steel Company in Latrobe, PA, used a novel technique to
manufacture crater-resisting carbides with improved strength and toughness and in 1937
introduced a tungsten-titanium carbide which, when used as a tool material, proved to be
effective in steel cutting. He received a patent for this formulation and went on to form
the McKenna Metals Company, now known as Kennametal Inc. Constant progress by
the carbide manufacturers in improvement of their products has lowered the cost of
carbides from the original $450 per pound to such an extent that they are no longer
regarded as precious metals. Depending upon grade, shape, and size, the cost, with
greatly improved quality and capacity dramatically increased, is but a fraction of what it
was in 1929. Although tungsten carbide has become an extremely successful
metalcutting tool, its roots are as a wear material, a wire draw die, which is still a
successful application of this material to this day.

Although various carbide manufacturers may have different processing procedures, the
final product is obtained by compacting the powder formulation by some technique and
sintering the constituents into a solid mass in which cobalt, or a similar metal, bonds or
cements the particles of carbide together. The manufacturing process is further
described below.

Rigorous control is necessary throughout the manufacturing process since the quality of
the final product can be greatly affected by seemingly insignificant factors. Purity,
quantity, and particle size of the powdered materials must be closely watched. Mixing,
milling, pressing, pre-sintering, and sintering techniques are among the factors
influencing the characteristics of the finished material. Carbide should be uniform in
structure and grain size, free from porosity, and of maximum density, strength, and
hardness. Modifications can be achieved by using various carbides and bonding
materials, by varying the proportions of carbide to the cementing matrix, and by
regulating the carbide particle size.




1-2






How is cemented carbide manufactured?

As mentioned previously, cemented carbide is made by a powder metallurgy process.
The compaction process is performed under very high pressure in a mechanical press
as shown in Figure 1-1 or in an isostatic chamber to form a part with the consistency of
blackboard chalk. A small amount of wax (paraffin) is added to increase the green
strength and help in handling the compacted shape. In this green state, it can be
formed or shaped by conventional methods such as turning, milling, grinding, and drilling
( Figure 1-2). The formed and shaped carbide is then sintered (placed in a vacuum
furnace at a high temperature). During the sintering process, the carbide may shrink as
much as 20% linearly, or nearly 48% by volume (Figures 1-3 and 1-4)



Mechanical Press Powder Shaping Mill

Figure 1-1 Figure 1-2

Sinter-HIP furnace

Figure 1-3





1-3







Figure 1-4



For an as-sintered part, it is considered an industry standard to be able to hold a
tolerance of 0.8% of the dimension or 0.005, whichever is greater. Tighter
tolerances can be held on smaller pressed parts. After sintering, cemented carbide has
achieved its full density and hardness. It can then be fabricated by diamond wheel
grinding or electrical discharge machining (EDM) techniques, both of which are
discussed fully in Chapter V.



























1-4

Note the difference in size as this part
progresses through the manufacturing process
from iso-pressed billet to partially machined in
the green state to fully machined to final
sintered size.
Chapter II - Unique Properties of Cemented Carbide
MECHANICAL PROPERTIES
Hardness
This is one of the most important properties of cemented carbide. It is the one physical
property that is thought to be the most important when it comes to abrasion resistance,
although this property alone does not dictate the success of a carbide grade in a wear
application. Hardness is determined by indenting a sample with a diamond penetrator
per ASTM standard B-294. Hardness values for cemented carbide are usually
expressed in terms of Rockwell A or Vickers values. Steels are measured in a similar
fashion and are expressed in terms of Rockwell C. Figure 2-1 depicts the approximate
conversion of Rockwell A to C. It can be seen that a D2 tool steel heat treated and
hardened to a Rockwell C value of 62 is still quite soft when compared to a 6% cobalt
binder grade of carbide with a value of 92 Rockwell A.
Hardness Conversion Chart (HR
A
-HR
C
)
Rockwell A Rockwell C
91.8-92.8 79.5-81.5
91.5-92.5 79.0-81.0
90.5-91.5 77.0-79.0
90.2-91.2 76.5-79.5
89.8-90.8 75.6-77.6
89.0-90.0 74.0-76.0
88.5-89.5 73.0-75.0
88.0-89.0 72.0-74.0
87.5-88.5 71.0-73.0
87.0-88.0 71.0-72.0
86.0-87.0 69.0-71.0
83.0-84.5 63.0-66.0
81.5-83.0 61.0-63.0
Figure 2-1
Density
Density is determined according to the ASTM standard B311. Since cemented carbide
is a composite material, and its constituent ingredients have varying individual densities,
the density of cemented carbide varies with composition. Combining these materials in
various proportions creates variation in the density of the cemented carbides in line with
their composition. A density of 14.5 g/cc is typical for a 10% cobalt binder material. This
value is twice the density of a 1040 carbon steel, which is an important consideration
when weight is a factor in design.
2-1
Transverse Rupture Strength
The mechanical strength of cemented carbide is usually determined by this method
rather than a tensile test commonly used for steel. The reason for this is that brittle
materials are extremely sensitive to misalignment of the tensile test apparatus and
sensitive to any notch or surface defect, which would cause a stress concentration and
lead to incorrect test results. The transverse rupture strength (TRS) is determined by
placing a standard sample (per ASTM B-406, ISO 3327) between two supports and
loading it until fracture occurs, as shown in Figure 2-2. The value obtained is called the
transverse rupture strength or bending strength and is measured as the load that is
needed to fracture the sample. This is shown as load per unit area, expressed in psi or
N/mm
2
. Several tests are conducted and the value is taken as the average of all
observed tests because cemented carbide exhibits a range of fracture values caused by
the existence of micro-voids, inherent in all brittle materials.
Figure 2-2
The values for transverse rupture strength of cemented carbide grades that appear in
various suppliers properties charts are based on the above standard test and thus,
reflect the mechanical strength for this sample size only. Many designers, even those in
the carbide industry, often regard the TRS values as the design strength value for a
grade and use it in designating which grade should work in a particular application,
expecting that value to hold true. In reality, these values decrease as the size of the part
increases and the design strength value for larger pieces must be calculated due to the
size effect. More on this in Chapter III. The TRS values increase with increasing binder
content as shown in Figure 2-3.
2-2
Effect of Binder Content
Figure 2-3
Another factor that affects the mechanical properties of cemented carbide, specifically
the transverse rupture strength, is the grain size. The effect of grain size is shown in the
chart below. As grain size increases, TRS values increase and wear resistance
decreases.
Effect of Grain Size
Figure 2-4
2-3
< 4%
4% -10%
10%-16%
>16%
Mechanical Strength
W
e
a
r

R
e
s
i
s
t
a
n
c
e
ultrafine 0.5 m
submicron 0.8 m
medium 1-2 m
coarse >= 3 m
Mechanical Strength
W
e
a
r

R
e
s
i
s
t
a
n
c
e
Compressive Strength
This property is also one of the most important attributes of cemented carbide. Ductile
materials under compression loading merely bulge or swell without fracture but a brittle
material will fail, due mostly to a shear fracture than true compression. Cemented
carbide exhibits very high compressive strength when compared to most any other
material and the value increases with decreasing binder content and decreasing grain
size. (Figure 2-5) Depending on the grain size and binder content, values between
400K-900K psi 7 kN/mm
2
) are typical for cemented carbide.
Figure 2-5
Modulus of Elasticity of Cemented Carbide
The modulus of elasticity or Youngs Modulus (E) is a measure of the stiffness of a
material and is measured as the rate of change of tensile or compressive stress with
respect to unit strain and is expressed as:
E =

The Youngs Modulus for cemented carbide is as high as 94,000,000 psi (>650 kN/mm
2
)
and is 2 to 3 times higher than steel. It increases linearly with decreasing binder
content. This property of carbide is used to resist deflection and is extremely useful
when employing this material in a cantilever beam, such as a boring bar application.
2-4
Compressive Strength
400,000
500,000
600,000
700,000
800,000
900,000
1,000,000
0 5 10 15 20 25
% Cobalt (wt)
C
o
m
p
r
e
s
s
i
v
e

S
t
r
e
n
g
t
h


p
s
i
Impact Strength
Finding exact solutions to impact problems can be extremely complex. Formulae for
impact stresses show that stress varies directly with the modulus of elasticity hence,
cemented carbide, having a high modulus of elasticity, is not suitable for all impact
applications. However, given the hardness of cemented carbide, especially a higher
binder grade containing 25% cobalt binder with a coarser grain structure, a surprisingly
high degree of impact strength is exhibited. Transverse rupture is often mistakenly used
as a measure of impact resistance when, in fact, fracture toughness is a better indicator
of the ability of cemented carbide to withstand mechanical shock or impact. Fracture
toughness varies according to grain size and binder content as shown in Figure 2-6.
Figure 2-6
With the skillful use of design techniques, the impact energy imposed on the carbide can
be transmitted in many instances to a more ductile mating part, thereby allowing the use
of a higher wear resistant grade of carbide. This utilizes the best characteristics of both
materials.
2-5
Fracture Toughness vs Grain Size
3
5
7
9
11
13
15
17
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
% Binder
F
r
a
c
t
u
r
e

T
o
u
g
h
n
e
s
s

M
P
a
m
0
.
5
4 micron
2 micron
0.8 micron
0.5 micron
Fatigue Strength
When a material is subjected to repeated cycles of stress reversal or fluctuation, failure
may occur even though the maximum stress is less than what would have caused a
failure had the loading been constant. Fatigue properties are determined by subjecting
test specimens to stress cycles and counting the number of cycles to failure. This
usually results in a plot of values shown on an S-N diagram where failure stress is the Y
axis and the number of cycles at which failure occurred is shown on the X axis. Several
large companies in the carbide industry have conducted such tests and have reported
their findings. Sandvik reports that the fatigue strength of cemented carbide under
pulsating compression loading can be taken to be 65 85% of the static compressive
strength at 2 x 10
6
cycles. No definite fatigue strength limit, which corresponds to an
infinite life, has been found as in the case of steel and other metals. The fatigue
strength increases with decreasing tungsten carbide grain size and decreasing binder
content.
THERMAL PROPERTIES
Coefficient of Thermal Expansion
Cemented tungsten carbide has a very low coefficient of thermal expansion (CTE). This
property is quite useful when shrink fitting a carbide die into a steel casing, which is
described thoroughly in Chapter IV. Compared to steel, WC-Co cemented carbides
have values of approximately half that of ferritic and martensitic steels while the ratio is
closer to 1:3 for austenitic steels. The value for CTE increases with increasing binder
content as shown in Figure 2-7 and can be increased with the addition of alloying
agents.
Figure 2-7
2-6
Thermal Expansion Coefficient for Cemented Carbide
20 - 800C
3.5
4
4.5
5
5.5
6
6.5
7
0 5 10 15 20 25 30
% binder
T
h
e
r
m
a
l

E
x
p
a
n
s
i
o
n

x

1
0
-
6
/

K
Thermal Conductivity
Pure WC is an excellent conductor of heat but WC-Co cemented carbides have a
thermal conductivity factor of about one third that of copper. Grain size has no
noticeable effect on this property, however, the presence of titanium carbide or tantalum
carbide additives will decrease the thermal conductivity factor significantly.
ELECTRICAL AND MAGNETIC PROPERTIES
Electrical Resistance and Conductivity
WC-Co cemented carbides have low electrical resistance with a typical value of 20
cm. Consequently, cemented carbide is a good electrical conductor having a value of
about 10% of the copper standard. This property is useful because it allows the use of
EDM (electrical discharge machining) as a fabrication tool for cemented carbide. See
more on this technique in Chapter V.
Magnetic Properties
Cemented carbides show ferromagnetic properties at room temperature due to the
presence of the metallic binder phase, cobalt or nickel. This property is useful in non-
destructive testing of a piece of carbide to determine magnetic saturation and coercivity.
Permeability
Low magnetic permeability is a characteristic of WC-Co cemented carbides that contain
a ferromagnetic binder phase. It increases with the cobalt content and the typical range
of values is 1.01 to about 12 when the standard is considered a vacuum with a value of
1. This property can be useful in abrasion resistant applications involving computer or
disk drive media where magnetism would have a deleterious effect. A low magnetic
permeability is important in a compacting die for pressing magnetic powders.
CORROSION PROPERTIES
More detailed information on corrosion resistance is presented in Chapter III but some
fundamental information is presented here.
Tungsten carbide particles themselves are resistant to most corrosive media. It is the
binder material that is susceptible to leaching in the presence of a strong acid or alkali
solution. The binder material will leach from the surface of cemented carbide, leaving a
skeletal structure, which is unsupported. The carbide particles will then abrade away
quite readily, exposing new surface area to be attacked. When binder content is low, the
carbide skeleton is denser. Low binder grades show a slightly higher combined wear
and corrosion resistance than those grades with a higher binder content.
Straight WC-Co grades are corrosion resistant at neutral pH, which is a value of pH7.
This is also true for WC-Co grades that contain additives like titanium carbide (TiC),
tantalum carbide (TaC) or niobium carbide (NbC). Certain alloyed titanium
carbide/nickel binder based grades possess the highest corrosion resistance down to
about pH1. When compared to straight WC-Co grades these grades are brittle and have
inferior thermal conductivity.
2-7
They are also hard to grind and braze, and are only used in specific applications where
corrosion and wear resistance are a must and mechanical strength and thermal shock
resistance are not as important.
When corrosion/wear is a prime design requirement, specially alloyed WC-Ni grades are
the best choice. They are resistant down to pH2-3. In certain solutions, where pH value
is less than 2, they have proven to be resistant to corrosion. Because they have WC as
the hard principle, and nickel and cobalt are similar metals in most respects, their
mechanical and thermal properties are similar to those of straight WC-Co grades.
The pH factor is one of the most important parameters when determining how corrosive
a medium will be. Other major influencing factors include temperature and the electric
conductivity of the medium. The latter is dependent on the ion concentration, i.e. the
amount of dissolved salts in the solution. Therefore, it is hard to simply determine how
corrosive a certain medium will be. No general rules apply to all situations, however, it is
generally accepted that WC-Co cemented carbides should not be exposed to pH7 or
below or leaching will occur. For a particular grade, it is recommended that tests be
conducted with the intended medium.
2-8
Chapter III - Design Considerations
Stress Analysis
There are several considerations when designing an industrial component such as:
Geometry
Stress
Wear
Corrosion
Impact/Shock resistance
Friction
Fatigue
Thermal conditions/properties
Stress risers/concentrations
Safety factor
All these factors play a role in determining the proper shape, size and grade of a
component but the amount of stress involved in an application is typically the most
elusive. Approximate calculations are not usually considered adequate for general
design problems, but often the carbide designer does not have sufficient information with
which to perform basic engineering calculations, such as in many metal-forming
applications. Hence the carbide designers first priority is to ensure that the component
must first withstand the operating load and not fracture. If the part wears out, the grade
formulation can then be adjusted to impart a higher degree of wear resistance and still
withstand the stresses involved. Failure can come in several forms ranging from micro-
cracks to premature wear to total fracture. Contributing factors can be corrosion, fatigue
or material flaws but avoiding total fracture is the first priority.
Designing by todays engineering standards, initially involves determination of the
shape of a part in order to arrive at the proper decision regarding the final size and
tolerance of the part followed by attempting to understand the stresses that the part will
see under operating conditions. An understanding of these factors then leads to a
material recommendation. The amount of stress that the part experiences and the
amount of deflection that occurs are key factors in any general design situation.
All materials deform when subjected to load. For most materials a change in load
results in a corresponding, but not necessarily linear, change in deformation. If, upon
removal of the load, a body returns to its original size and shape, the body has
undergone elastic deformation. If the body does not completely recover to its original
shape, it is said to be partially elastic; it is perfectly elastic when full recovery takes
place. In steel, if a part is stressed beyond its elastic limit deformation will occur but not
necessarily failure. For cemented carbide, classified as a brittle material, very little
plastic deformation will occur preceding the initiation of a crack and ultimate failure,
hence the yield strength and rupture strength are essentially identical. This is very
different than the yield strength and ultimate strength of low carbon and structural steels
where the ultimate strength is 1 to 1 times higher than the yield strength.
3-1
Generalized in its simplest form for tensile stresses and compressive stresses, Hookes
Law says that stress is proportional to strain and Youngs Modulus of Elasticity is a
constant of this proportionality expressed in equation form as:
E =


where stress is the force per unit area and strain is the deformation per unit length.
Cemented carbide has an E value of 93,000,000 psi ( ~ 650 kN/mm
2
) for its Modulus of
Elasticity, meaning it is approximately 2-3 times stiffer than steel which has a Youngs
Modulus of 30,000,00 psi.
For a shear stress and strain, the proportionality constant is G (sometimes called the
modulus of rigidity, or torsional modulus) and it is represented in equation form as:
G =

where is the shear load per unit area and is the shear deformation per unit length.
G is anywhere from one third to one half the magnitude of E for most materials and this
holds true with experimental findings for cemented carbide. Thus, the resistance to
torsional loading and shear stress for cemented carbide is still 2 to 3 times that of steel,
a very useful property for rotating tooling.
In many cases, simple tensile or compressive loads are not the only stresses put on a
machine component or tooling component. Radial loads may be combined with axial
loads as well thermal stresses to confound the designer. Stresses, deflections, strains
and loads may be determined by the application of strain gages to the surface of the
part, then applying loads simulating those encountered in operation. Many times this is
not feasible or practical for a carbide application and elegant finite element analysis
(FEA) models have been created to try to determine the stress levels in parts where
strain cannot be measured and stress cannot be calculated.
The scope of this chapter will be confined to special considerations governing the design
of a cemented carbide component. Consult General Carbide application engineers for
specific recommendations on grade selection based on various operating conditions
where experience plays an important role in successful design.
3-2
Weibulls Statistical Strength theory
All materials contain some amount of defects in the form of voids, pores or micro-cracks.
These defects lead to reduced material strength. For ductile materials, defect frequency
and size are important but in the case of cemented carbide, defect frequency and size
are limiting factors. In fact, the mechanical strength of cemented carbide is volume
dependent because the probability of finding large defects increases with the size of the
part.
According to Weibulls statistical strength theory, the size effect can be expressed as:

1
=
V
1
1/m

2
V
2
where:

1
= fracture stress of No. 1 size specimen

2
= fracture stress of No. 2 size specimen
V
1
= volume of No. 1 specimen
V
2
= volume of No. 2 specimen
m = factor derived from the spread in fracture stress of the material, known as the
material constant or Weibull modulus.
The material constant m can be shown to be an index of the relative number of voids in
the material, or of its homogeneity. High quality cemented carbides made today should
have an m value of 9 or higher. High m values correspond with small variations in
fracture stress and less volume-dependent material.
For larger volume parts, size effect should be taken into consideration when evaluating
the strength properties of a grade of carbide for a particular application.
Working Stress
Design stress or working stress can be defined as:

w
=

N
........where is the strength or mean fracture stress of the material, and N is the
safety factor. The reliability of the design depends primarily upon the accurate
determination of both and N, which is discussed in the following paragraphs.
3-3
Safety Factor
Design engineers have to take into consideration many factors when selecting a
reasonable safety factor. Estimating the load, for example, placed on a metalforming
punch, the slight misalignment of the press causing a bending moment or the pressure
experienced by an out of tolerance die can affect the performance of the punch and may
endanger the operator. Operating and environmental conditions can change and
become more severe. Thus, the selection of a reasonable safety factor requires a good
knowledge of design, a thorough understanding of the strength of the material and
application engineering experience.
Determination of Failure Stress
As mentioned in Chapter II, cemented carbide exhibits a range of fracture values caused
by the existence of micro-voids, inherent in all brittle materials. This characteristic
requires an entirely different approach to the evaluation of the failure stress. The value
of the stress at fracture can also vary widely with size, stress state (tensile, bending,
torsion), shape, and type of loading.
Transverse Rupture Strength (TRS) or bending strength is the most common way of
determining the mechanical strength of cemented carbide. TRS is determined by ASTM
or ISO standard methods whereby a specimen of rectangular cross-section is placed
across two supports and loaded in the middle until fracture occurs (Figure 2-2). The
TRS value for a particular grade formulation is the average of several observed values.
These values are usually provided by the carbide supplier in grade specification data
sheets and are shown as the transverse rupture strength (TRS) values. However, the
mean ultimate strength values obtained from standard TRS specimen tests, used as the
basis for determining failure stress, are not directly applicable to design, hence there is a
need for an approach based on the probabilities of failure.
Weibulls Statistical Strength Theory is based on the premise that a brittle material is
subject to a flaw of random size and random distribution locating itself in the area of
highest stress thus creating a stress concentration and weakening the material, causing
it to fail at below than expected stress levels or published TRS values.
This statistical strength theory or probability analysis requires the use of a material
safety factor. Figure 3-1 shows the curves for the material safety factor for two levels of
reliability, 99% and 99.99%, for two material constants or m values. The safety factor
can be determined by interpreting the point where the m value intersects the survival
probability curve desired. These values should be applied to the mean fracture stresses
determined by the TRS value and also corrected for size effect. The material safety
factor and the design safety factor previously discussed are not the same and should not
be confused.
3-4
Figure 3-1
Common practice in the carbide industry today is the use of hot isostatic pressing (HIP)
in the manufacture of cemented carbide. This process of sintering cemented carbide
under simultaneous application of heat and pressure raises the m value shown above
by eliminating many of the micro-voids in the material and lowers the material safety
factor. This ensures a more reliable probability of survival in any given application.
Stress Concentrations
Critical to the success of any cemented carbide component, especially in an impact
situation, is the elimination of all stress risers or stress concentrations. These are points
in the component where the cross sectional area changes abruptly thereby increasing
the strength or intensity of applied stress. If a micro-void would happen to be located at
this point of highest stress, it would be the initiation site for a crack. It would propagate
rapidly under continued stress and lead to premature failure. Ductile materials are not
as susceptible to stress risers since they can yield or plastically or deform at the points
of localized stress and not exhibit immediate failure.
Avoiding stress concentrations should be a major consideration when designing with
cemented carbide. A minor modification in the shape of the part can reduce the stress
concentration considerably. For example, utilizing the largest possible radius when
transitioning from one diameter to another will minimize the stress concentration in round
tool parts. Never allow internal sharp corners to exist and, if possible, spell out the
expected radius on the print. On external surfaces, add the note on the technical
drawing to Break all sharp edges.
3-5
Considerations to guide the designer can be confusing. In some instances, operating
conditions are not well understood, hence application engineering experience is
invaluable. The value of the expertise of the designers at General Carbide, who have
applied carbide in similar applications, cannot be overemphasized.
Determining the Relationship between TRS, tensile strength and torsional
strength
As we learned in Chapter II, conventional tensile tests used for steel are not suitable for
brittle materials, such as cemented carbide, due to the erroneous results that occur from
misalignment and improper clamping. These conditions impose additional stresses on
the material, rather than capturing the true scatter of values, thereby causing large
deviations in the test data. If it is required to estimate the tensile strength of a cemented
carbide grade, a calculated result can be arrived at using the following equation:
(bending)
=
V
T
1/m
(tension) V
B
if the volumes are equal:
(bending)
=
2 (m+1)
1/m
(tension)
using m = 7.5 (a conservative value for the material constant, where 9 is more indicative
of cemented carbide today):
(bending)
= 1.46
(tension)
Thus, a designer is safe in using 45% to 50% of the transverse rupture strength as the
mean tensile strength of a cemented carbide specimen that has the same volume as the
transverse rupture test specimen. The effect of size must again be considered in
estimating the tensile strength of a component.
Although tensile strength is the weakest mechanical property of all brittle materials,
including cemented carbide, using proper design techniques to take advantage of
cemented carbides high compressive strength can overcome this weakness. These
techniques are reviewed in Chapter IV.
3-6
2 (m+1)
For reference, it is estimated based on experimental test results and Weibulls theory,
that the torsion or shear strength of a specimen of cemented carbide is, in general, 50%
to 58% of the bending strength or transverse rupture strength.
Elevated Temperatures
Cemented carbide will retain most of its strength at elevated temperatures. It has
exceptional high hot hardness, a property readily taken advantage of in metal cutting
applications. However, as temperatures approach 1000
0
F, oxidation will occur. This
appears as a powder layer or flakes on the surface of the carbide, which are easily
abraded away. Above 1000
0
F, oxidation is too severe for cemented carbide to be
used. Fortunately, most industrial applications do not reach this temperature extreme.
The grade formulations that are most suitable for elevated temperatures are the higher
binder grades, which can withstand higher impact stresses and thermal shock.
The difference in coefficient of thermal expansion (CTE) between cemented carbide and
steel is significant. Carbide has a CTE value 1/3 that of steel. This low thermal
expansion rate value is readily used in designing cemented carbide for shrink-fit
assemblies using steel die cases. This subject is covered in more detail in Chapter IV.
Corrosive Environments
Using cemented carbide for wear resistant components such as seal rings, flow control
devices, nozzles, and bearings, has become common practice in todays industry. In
some processing operations, the environment may include severe corrosion or extremes
of temperature.
Corrosion of cemented carbide is usually referred to as leaching which is the removal of
the binder phase and, thus, the surface region will remain only as a carbide skeleton.
The bonds between adjacent carbide grains are rather weak so the skeletal structure will
result in higher abrasion rates and exposure of more surface area to be affected by
leaching. Stress concentrations caused by surface pitting will affect the strength of the
carbide. In lower binder content grades, the carbide skeleton is more developed and,
accordingly, such grades exhibit a somewhat higher combined wear and corrosion
resistance than corresponding grades with higher binder phase contents. The
photomicrographs below show how the surface appears when leaching occurs and a
side view of the removal of the binder phase, leaving the grains uncemented.
3-7
Figure 3-2 Figure 3-3
The limited corrosion resistance of straight tungsten carbide and cobalt (WC-Co) grades
often makes them unsuitable in applications where the corrosive conditions are severe.
For these applications, most carbide manufacturers have formulated a series of highly
corrosion resistant grades, which substitute nickel for cobalt or contain mixed binder
phases of chrome, nickel, cobalt and molybdenum. Some manufacturers also produce
grades with combinations of titanium carbide and nickel to add an extra measure of
corrosion resistance.
In most corrosion-wear situations, the better choice is specially alloyed WC-Ni grades,
as shown in Figure 3-4, which are resistant down to pH 2-3. Even in certain solutions
with pH values less than 2, they have proven to be resistant to corrosion. Since they
have tungsten carbide as the hard principle ingredient, and Ni and Co are similar metals
in most respects, the mechanical and thermal properties of WC-Ni grades are
comparable to those of straight WC-Co grades.
The pH value is one of the most important parameters when determining the corrosivity
of a medium, but other factors also have a major influence, such as the temperature and
the electrical conductivity of the medium. The latter is dependent on the ion
concentration, i.e., the amount of dissolved salts in the solution. Thus, one cannot
define the corrosivity of a certain medium in a simple way and, accordingly, no general
rules are valid in all situations.
3-8
Figure 3-4
The amount and rate of corrosion may change considerably with factor changes such as
concentration and temperature of the corrosive fluid plus exposure time to the carbide.
The most accurate way to select a carbide grade is to test the grade under the actual
corrosive conditions in which the carbide will be used. Contact General Carbide for
assistance in selecting appropriate grades for testing.
With the introduction of advanced ceramics into the engineered materials market, the
use of cemented carbides for their corrosion resistance alone is difficult to justify.
However, when abrasion resistance and toughness requirements are also involved, the
combined corrosion and wear resistance of cemented carbide proves its usefulness.
3-9
Chapter IV - Attaching and Assembling Cemented Carbide Parts
In order to take advantage of the unique properties of cemented carbide, such as its
outstanding wear-resistance, high level of compressive strength, hardness and rigidity,
and at the same time minimizing the consequences of its inherit brittleness, it is
advisable and economically more practical to join the carbide to a tougher material.
Steel as well as other non-ferrous alloys have been widely utilized, thereby forming a
composite assembly. By definition, a composite product /assembly should be
considered to be a macroscopic blend of two or more distinct materials which have a
discernible interface between them.
Indeed, by applying intelligent joining techniques, it is possible to get the optimum
combination of functional properties at a significantly lower cost for the final product.
In order to achieve compatibility of these opposite materials, a good understanding of
the overall physical properties of each material is necessary. A description of the
properties of cemented carbide was provided in Chapter II, along with comments and/or
a direct comparison to steel.
Attachment of carbide to another mating member of a multifunctional and multilayer
assembly may be accomplished by various joining techniques including brazing,
industrial adhesives, interference/shrink fit assembly and mechanical fastening.
Conventional welding with oxy-acetylene gas or arc welding is a popular joining
technique but not in joining carbide to steel. In general, traditional welding cant be
considered to be a reliable technological solution because of oxidation issues and the
high potential for chemical imbalance (e.g. carbon deficiency). However, recent
achievements in vacuum technologies including electron-beam welding have been
applied successfully to weld carbide to steel or two carbide components to each other.
In addition, friction welding has been successfully used in small cross sectional joints,
such as carbide tip being attached onto a steel band saw for a metalcutting band saw
application.
Both brazing and epoxy adhesive techniques are very common and are used extensively
but each has its own inherent limitations, mostly related to their sensitivity to operating
temperatures and the presence of corrosive substances in the immediate environment,
like sulfur, phosphorus, halides, etc. Interference/shrink fits are very commonly used for
cylindrical members since this technique takes advantage of the difference in Coefficient
of Thermal Expansion (CTE) of the two materials. Mechanical attachment techniques
overcome the limitations of both brazing and epoxy adhesives and tend to make the task
easier for replacement of new parts. Interference fits require that the mated sleeves be
separated by pushing out the inner carbide insert. Advantages and disadvantages exist
with all four techniques and each technique must be judged to be most appropriate for a
given application. General Carbide application engineers can recommend which
technique is appropriate for any given application.
All four techniques are discussed in detail below.
4-1
1. Brazing
Brazing is a technique to provide a solid joint between two (or sometimes several)
dissimilar materials, mainly metal-based materials where the interface is produced by
heating of an assembly while utilizing a filler metal. This filler metal compound has a
melting point (liquidus temperature) which is substantially below the solidus temperature
of the base materials (metals). For metals, brazing is usually accomplished by using a
filler metal with a melting temperature above 450
o
C (840
o
F) while another similar
process, known as a soldering, employs a filler metal with a melting temperature
(liquidus temperature) below 450
o
C (840
o
F).
During both brazing and soldering, the molten filler metal is distributed by capillary action
between the closely fitted surfaces of the two materials to be joined, which then forms a
solid interface upon cooling. Due to the physical nature of the brazing process, it is
imperative that the filler material has excellent wetting ability (i.e. dihedral angle to be
close to zero), meaning that the braze material must adhere evenly and uniformly over
the entire surface via capillary action.
Presently, brazing is quite commonly used for cemented carbide in a wide variety of
applications and in particular, it has been very successful for small surface areas and
shorter length joints. Common industrial practices for brazing of cemented carbide are
furnace brazing, including vacuum, brazing and induction brazing, as well as the torch
brazing and resistance brazing. Through the use of certain design principles, brazing
can also be accomplished satisfactorily on larger joints and long blades. Cemented
carbides can be easily wetted with brazing alloys ranging from silver solder to pure
copper, which is discussed below.
Bi-metal thermal expansion difference
The Figure 4-1 shows the bi-metal strip effect involved when two dissimilar materials are
bonded by fusion of a braze filler material. As mentioned in Chapter III, the Coefficient
of Thermal Expansion (CTE) for each material is significantly different, with steel having
having a CTE two to three times higher than carbide. At the solidification point of the
braze material, the two strips remain straight and parallel. However, as the bonded
strips cool, the steel contracts about twice as much as the carbide, causing the bimetal
strip to bend as shown. (Note: curvature may not be representative. When the carbide
passes its elastic stress limit, it will crack).
Figure 4-1
4-2
Braze Strain
As the steel base above cools and shrinks, a strain is set up in the joint, the severity of
which is dependent on the type of steel and its cooling curve, the grade of carbide and
its cooling curve and the type of braze compound. Since the amount of curvature of
brazed parts is usually limited by the size of each assembly member as well as the
stiffness of the materials within the joint, the strain must be absorbed within the
assembly or the strain induced by contraction will be relieved by the carbide
experiencing a crack.
Relieving Braze Strains
A simple braze of well-proportioned parts, not excessively long or of large surface area,
presents no serious issues when brazing carbide to steel. As a rule, no abnormal strains
are observed in the assembly because generated stresses are easily absorbed or
relieved by the braze material. However, on longer joints the strain becomes
proportionately greater and, therefore, other techniques must be considered.
Our extensive industrial experience suggests that braze strains can be relieved by three
proven methods:
a) Sandwich Braze a sandwich braze consists of a copper shim between the carbide
and steel parts to be assembled. The copper is not melted because only a low or
medium temperature brazing (filler) metal is used. The copper shim is malleable enough
to deform under the brazing strain without losing its bond to the steel or carbide parts.
Figure 4-2
The copper shim, as illustrated in Figure 4-2, can also be sandwiched between two
shims of silver filler material. 3-ply shims are commercially available with a layer of
silver filler material on either side of the copper.
The copper sandwich braze is useful only for light or medium duty since it will tend to
mush and be squeezed out if the application involves heavy loading or high impact.
It will not provide the uniform support required to prevent breakage of the carbide.
A potential alternative is a nickel shim, which will withstand more impact. However,
nickel does not have the malleability of copper and will not relieve the braze strains
as effectively as copper.
When brazing long joints is unavoidable, counter-straining, peening or soaking
process steps are recommended as methods for reducing stress and strain.
4-3
b) Counterstraining This technique utilizes the principle of force and counterforce.
By pre-stressing the assembly beyond normal, it is often possible to have the part come
out nearly straight when the clamp is removed and the part cools. This pre-stress
reduces the tensile stress in the outer edge of a carbide strip by forcibly overcoming the
curvature of the assembly. One way of doing this is to clamp the assembly in a jig
during cooling. Most of the curvature can be eliminated this way. Fig. 4-3 shows how
this principle has been applied to a wear insert.
Figure 4-3 Figure 4-4
Peening is another method of counterstraining and relieving braze stress. This
technique is accomplished by using a ball peen hammer peen the steel surface on the
opposite side of the braze joint. This will expand it and set up a counterstrain to restore
straightness of the assembly (Fig. 4-4).
c) Opposing braze - Another solution to relieve strain is to braze a carbide blade to
opposite faces of the steel component. This sets up balanced opposing stresses and an
added benefit is that two working surfaces are now available for use. (Fig. 4-5).
Figure 4-5
With all of these strain-relieving methods, the assembly should be allowed to stand for a
few hours (or even days) to allow any residual stresses to relieve themselves. Since the
strains may still be locked in, some creep could take place and that must be taken into
account before any final grinding of the assembly should occur.
Heat soaking is the process of bringing the assembly to temperatures in the 400F range
and is useful for minor stress relief only. It is not as effective for major stresses that
occur in larger joints of 1 or longer because the temperature needed to remove such
stresses is 1000F or above. Since this is just below the temperature at which silver
based braze compounds solidify, the cooled assembly would again be strained to about
the same temperature after cooling.
4-4
Design Recommendations to avoid Braze Strains
Since braze strains result from the difference in the thermal expansion rates of the
materials being brazed, it follows that the use of materials with similar thermal expansion
coefficients would result in less braze strains being introduced into the assembly.
Heavy tungsten alloy has a coefficient of thermal expansion approximately equal to that
of cemented carbide. The material wets well with silver filler material, producing very
sound brazed joints with minimum braze strain. Heavy tungsten alloy has physical
properties similar to those of mild steel and is easily machined.
Figure 4-6
If the application being considered can employ heavy tungsten alloy, as depicted above
for a grinder spindle housing shown in Figure 4-6, the probability for success is
enhanced. The two materials, having similar expansion and contraction curves, will
minimize the braze strains in the assembly.
Using a series of short pieces of carbide instead of one larger piece will reduce the
cumulative effect of braze strain. An assembly that is designed with joints at non-critical
points will avoid strain at highly stressed areas and reduce the likelihood of brazing
failure or cracking.
Strains can often be relieved on a miss-proportioned assembly by brazing only on one
surface. A special mask-like paint can be applied to the surface on which no attachment
is needed so that the braze alloy will not wet those surfaces. A relief gap, machined into
the body, will also help to prevent braze material from bonding all surfaces.
A combination of the above principles showing how to reduce braze strains is shown in
Fig. 4-7.
Figure 4-7
4-5
Thickness of Braze Joints
The thickness of a braze joint is critical. A thick braze layer can more readily absorb
strain but it can fail due to wash out by abrasive action or fail due to peening out by
impact loads. It can mush due to a lack of strength. For this reason, a thinner braze
layer is better because when a braze filler metal compound is stressed, it tends to
elongate before failure. This elongation requires a reduction in area. This, in turn, is
restrained in the joint by the proximity of the steel and carbide to which the brazing
compound is bonded. Thus, the braze compound resists elongation, not only by the
shear strength or tensile strength, but by resistance to flow in several directions at once.
The result is that the tensile strength of the thin braze can have several times the tensile
strength of the braze filler metal used.
A word of caution is needed regarding a braze joint that is too thin. Various tests have
shown that going too thin can bond the two materials too tightly and cause cracks in
otherwise well-proportioned assemblies. An ideal braze thickness is between .003 and
.005.
To assist operators that are unfamiliar with brazing work, it is often helpful to use a prick
punch that can be employed to raise an embossed area on the bonding surface of the
steel part to approximately .004 height. Three such punch marks will prevent the joint
from being squeezed too tightly during brazing and will frequently eliminate braze strain.
Shape of the Braze Joint
For maximum reduction of braze strains, a single brazed surface is best. Some
assemblies may not permit this so when more than one surface is to be brazed
simultaneously, the design should be such that the carbide will slide along one surface
until it engages a second surface. For the situation where three surfaces are involved,
the carbide should slide along one surface until it contacts against a second, then move
along the two surfaces until it rests against the third surface.
The shape of a brazed joint should be designed so that it will best resist the expected
loading when the part is put into operation. For example, a carbide punch in a metal-
forming application would tend only to encounter axial forces and compressive loading in
operation. In this case, a simple butt braze would be sufficient. However, that same
carbide tip brazed to steel and used as core rod in a powder compacting application
would be subjected to longitudinal forces and lateral thrust, so a conical braze joint
would be better in that instance.
Note that conical brazed joints should incorporate a vent-hole to permit exit of gas
pockets, flux, or excess braze material which might otherwise become trapped and
prevent proper control of braze thickness.
With any of these designs, as the diameters or surface areas increase, so does the
braze strain. Keeping the areas to be brazed as small as possible is essential.
4-6
Types of Brazing Compounds
The most common brazing compounds for bonding cemented carbide to steel are the
silver-copper series, commonly referred to as silver solder alloys. Various other brazing
compounds are available for bonding cemented carbide to ferrous alloys and they range
from low temperature tin/lead/zinc alloy solders to a high temperature copper brazing
filler composition.
Consideration must be given to the temperature range of the application and the bonding
temperature of the material when specifying a braze compound for a given job.
Low Temperature Soldering The lowest temperature solders have the advantage of
less thermal strain after bonding but they have low mechanical strength and certainly will
not stand up to operating conditions much above room temperature. Bond temperature
is only 700F or lower. This low temperature permits brazing of large areas or long wear
strips of cemented carbide and the thermal strains are held to a minimum. Low temp
solders of this type contain approximately 40% tin, 35% lead, and 25% zinc.
Medium Temperature Brazing The silver solder braze compounds are the most
common brazing alloys for bonding cemented carbide to steel, as mentioned previously.
Silver solder alloys are available in rod form as well as pre-cut shims or strips and
typically contain other elements to aid in wetting of the carbide surface. A typical
combination of ingredients would be 50% silver with copper, zinc and nickel additives.
They are high enough in melting point to be a good choice for the vast majority of wear-
resistant applications. This material begins to melt at 1170F and is completely fluid at
1270F.
High Temperature Brazing Straight copper is used as a brazing material because it
retains practically all of its strength up to a temperature of 1000F. Beyond this
temperature, most cemented carbides begin to oxidize anyway so this alloy will perform
at the highest practical operating temperature for most cemented carbides. However,
copper brazing requires a hydrogen atmosphere furnace for best results and thus is
usually limited to high production operations. It is usually accomplished with pre-cut
strips or shim material because the parts are inaccessible during brazing.
Copper makes a good braze as far as bonding is concerned. However, at the 2100F
bonding temperature required to use copper, most common steels will suffer excessive
grain growth and, as a result, will be brittle and weak. Some high-speed steels, as well
as air-hardening steels or silicon-manganese steels can take these high temperatures
without detrimental effects. Other high temperature brazing alloys are composed
primarily of nickel and flow at temperatures from approximately 1820F to 1925F.
To ensure proper tinning of the surfaces and freedom from inclusions or voids, it is best
to heat the assembly well above the fusion point of the filler material. If the designer is
unsure which of the above brazing compounds is best for a certain application, please
consult the application engineers at General Carbide for recommendations.
4-7
Brazing to Hardened Steel
Brazing to hardened steel presents it own unique issues since cemented carbide will not
withstand a liquid quench, typically required for hardened steel. Thus, a brazed
assembly cannot be subjected to rapid cooling without causing severe cracking. If a
very high-hardness steel is desired, an air-hardened steel can be used. It should be
brought to recommended hardening temperature and soaked long enough for proper
hardening, then air-cooled to obtain hardness prior to any brazing operation. With the
steel microstructure properly established, it can be reheated to hardening temperature
for a short time, brazed with a suitable brazing alloy for the particular temperature
involved and then air cooled. The assembly can then be drawn to the required hardness
of the steel without damage.
It is suggested that a nickel base air hardening steel, rather than a chromium base steel,
be used because the oxide of chromium that forms on the steel is difficult to flux away.
Moreover, nickel steel has higher toughness and, thereby, can better relax stresses
generated during the brazing operation. If an intermediate hardness of the steel part is
satisfactory, an oil hardening steel can be used, but it must be air quenched to obtain a
moderate hardness reading.
High-speed steel can be used for components if hardened, then brazed with a low
temperature silver solder at a maximum temperature of 1200F. The high-speed steel
member will retain most of the hardness at this temperature. The resulting combination
is suitable for many uses.
Brazing Procedures For Cemented Carbide
Brazing of cemented carbide can be accomplished in any one of several ways, high
frequency induction brazing, furnace heating, or torch brazing. Torch brazing of INVAR
plugs is described below but in all cases cleanliness of the mating parts is of utmost
importance.
Procedure for brazing INVAR plugs in Cemented Carbide for tapped holes:
Turning Plugs:
1. Inspect the hole diameter and the depth/length of the hole.
2. Turn an INVAR plug to .014 - .020 under the hole diameter, or .007 - .010 per
side.
3. Cut the plug length .010 over size for counterbored holes, and .025 over for
through holes. (You may want to leave more on the length depending on the job
and the experience of the operator).
4. If the preformed hole has a radius in the bottom of the counterbore, you will need
to radius the base of the plug. You can usually use a file to do this.
5. If the part has through holes, turn the plugs to .020 - .025 under the diameter of
the hole.
4-8
Preparing the Parts:
1. Grit blast parts using silicon carbide ceramic grit. CAUTION: Aluminum oxide
abrasive should not be used since it will inhibit proper wetting of the surface and
decrease the bonding strength, especially when brazing flat surfaces together.
Silicon carbide is a more friable abrasive and will not penetrate into the
micropores of the surface.
2. Wearing rubber gloves, clean all pieces with Denatured Alcohol or Toluene.
3. Be sure the parts are dry.
4. Flux the hole in the part and the plug with Stay-Sylv high-temperature flux.
(Fluxing will alleviate air and help capillary action of the braze material to take
place).
5. Set parts on a ceramic tray.
6. If the plug is to be inserted into a blind hole, drop small pieces of braze material
into the hole and insert the plug. (You can also partially wrap braze material
around the plug).
7. In some cases you may have to use a graphite plug to prevent the flow of braze
into unwanted areas. For example, a counterbored hole with a through hole
being plugged.
8. When brazing flat bars with through holes use a .007 shim on each end of the
bar so the plug protrudes.
You are now ready to braze.
Brazing the Parts:
1. Turn on the vapor hood.
2. Cut pieces of braze material to desired length.
3. Heat the part with an oxygen/acetylene torch making sure to heat the part evenly
all over. (Take care not to get the part super hot. The torch can reach
temperatures up to 1600F. If heat is not even, the part could crack).
4. Feed the braze material into the gap between the carbide part and the INVAR
plug. Also, work the INVAR plug up and down and around to ensure good braze
material flow.
5. When brazing carbide parts with counterbored through holes, turn the part over,
if possible, and feed the braze material from the bottom.
6. Large mass carbide parts may need to be pre-heated in a furnace to about
1150F and held there until the part is red/orange in color (approximately one
hour per inch of wall thickness).
7. When the carbide part is red/orange in color, remove the parts from the furnace
and place them on a pre-heated tray to braze.
8. To prevent the brazed assembly from sticking to the tray when brazing, simply
tap itafter the braze sets up but while the assembly is still hot.
9. Finer grain grades of cemented carbides need to be staged and preheated to
400F for about one hour. When brazing is complete, put them back into the
preheated furnace at 900- 1000F.
10. Turn furnace off and let the assembly cool in the furnace.
Cooling:
1. Let the brazed assembly air cool.
2. Do not set the assembly on a cold surface or expose it to cold air or anything that
would cool it rapidly.
Clean-Up and Inspection:
1. Grind off excessive braze material.
2. Grit blast and inspect the assembly to ensure a good braze.
4-9
Additional Notes:
1. When brazing round parts it is extremely important to heat the entire part evenly
to prevent cracking.
2. Put a piece of steel or graphite inside the I.D. of the part to help hold the heat.
Be sure this part has sufficient clearance from the I.D.
2. Industrial Adhesives
Structural adhesives fall into four broad polymer families: epoxies, cyanoacrylates,
silicones and acrylics. With the exception of silicones, these polymers have bond
strengths on the order of 2,500 to 7,500 psi in tensile-shear mode. They also tolerate
temperature swings as wide as 350
0
F and endure impact loads of 10 ft-lb/in
2
or greater,
which demonstrate the advancements that this industry has made in the last 20 years.
Industrial adhesives have several major benefits over other joining methods like brazing
and mechanical fastenings. Joint stress is reduced by evenly distributing the load over a
broad area. Adhesives are invisible because they are applied inside the joint, and they
resist flex and vibration stresses by forming a seal that can protect the joint from
corrosion.
Adhesives are also a perfect choice for joining irregularly shaped surfaces, which may
prove problematic for brazing. Minimal weight is added to an assembly and there is
virtually no change in part dimensions or shape.
Some adhesive limitations include a potential need to disassemble the joint, curing time,
and surface prep requirements.
Adhesives and mechanical fasteners, when used together, form a stronger bond than
when used separately. For example, a bolt that is tightened to the correct torque setting
and has a thread-locking adhesive applied to the threads will improve the strength of the
assembly. The thread-locking adhesive ensures the assembly will not loosen and
corrosion will also be minimized.
Some adhesives require the addition of a hardening agent, often referred to as a catalyst
or activator. Others require only heat to obtain the bond. When a catalyst is required, a
variable is introduced into the completeness of the mixing. There is a limited pot life, or
time during which the epoxy can be applied. In some cases, this can be extended by
refrigeration. Most adhesives requiring hardeners have a specified shelf life and the
manufacturer should be consulted on their recommendations.
Heat and humidity usually have the most damaging effects on bonded joints, although
exposure to solvents and ultraviolet light can also take a toll. Operating temperature is
the most important variable that qualifies an adhesive for a particular application. While
a device mounted outside is exposed to cold, wet, sunlight, and other conditions, the
maximum temperature is not likely to exceed 60C (140F). Therefore, an outdoor
environment does not eliminate any of the potential adhesive chemistries described
above.
4-10
Thermal Cycling:
When devices operate in environments that cycle between extremes of heat and/or
humidity, they experience thermal cycling or thermal shock. All materials expand when
heated and shrink when cooled. This rate of dimensional change is called the coefficient
of thermal expansion (CTE) as discussed in Chapter I. Differences in CTE produce
stress on the bond joint.
Resistance to thermal cycling is generally achieved in two basic ways:
A very high strength, rigid adhesive may resist the applied stress. Classic rigid
chemistries include acrylics and epoxies, but many urethane modified or elastomer-
modified formulations are available.
A softer, more flexible adhesive can absorb the applied stress by flexing or
moving rather than cracking. Silicones and urethanes are typical of these softer and
more flexible chemistries.
Surface Preparation:
Proper surface preparation is key to ensuring a good bond. It can be as simple as
cleaning the surfaces with a solvent to remove oils, greases, and other potential
contaminants that could hinder bond strength. Other applications may require surface
abrasion or grit blasting to enable proper adhesion. It is our experience that grit blasting
alone does not ensure a good bond and chemical cleaning with an approved solvent is
recommended.
Adhesive Recommendations:
The most frequent causes of adhesive joint failures do not involve adhesive strength.
Typically, adhesive joint failure may be attributed to poor design, inadequate surface
preparation, or improper adhesive selection for the substrate and the operational
environment. A competent carbide application engineer, familiar with successful
assembly techniques will be able to provide the optimum bonding technique for a
specific application. Testing under load may be necessary to ensure success of an
adhesive assembly.
The joint should be from .003 to .006 in thickness to assure maximum strength in the
bond. The facing surfaces must be clean and free from dirt, grease, and scale.
Dozens of adhesives are available for use today. General Carbide has had experience
with the products listed below which provide excellent bonding to cemented carbide:
1) 3M - DP460, two-part epoxy.
2) LOCTITE 320 acrylic adhesive with LOCTITE 7075 Activator.
4-11
3. Interference/Shrink Fit Assembly
A widely used and highly reliable method of mounting round sections of carbide into
steel is to employ an interference fit. The high compressive strength of carbide makes it
ideal for the compressive loading encountered with shrink fits and the tensile strength of
steel is ideally suited to withstand the hoop stresses encountered with this method.
In a shrink fit design, the amount of interference designed into the assembly depends
entirely on the requirements of the application. In heading die applications, the
maximum amount of interference and compressive loading is required on the carbide
insert in order to overcome the stress reversals that occur when the carbide is subjected
to the cyclical loading and pulsating internal pressures of the application. On the other
hand, a smaller interference may be adequate in the design of powder metal dies (see
guidelines below).
This method of attachment takes advantage of the difference in Coefficient of Thermal
Expansion (CTE) of steel and cemented carbide, as discussed previously under Brazing.
The difference is in excess of 2 to 1 between steel and carbide and this fact allows for
convenient assembly and disassembly of the joint. In applications involving operation at
elevated temperatures, the difference in the coefficient of thermal expansion will cause a
decrease in the amount of interference and must be considered in the design of the joint.
The amount of interference of the shrink fit can be calculated from Lams equations.
See Figure 4-8.
where...
diametral interference
P Pressure between cylinders
E
s
modulus of elasticity of steel
E
c
modulus of elasticity of carbide

s
Poissons ratio of steel

c
Poissons ratio of carbide
=
bP b
2
+ c
2
+
s
+ bP a
2
+ b
2

c
E
s
c
2
b
2
E
c
b
2
a
2
4-12
Figure 4-8
If a steel ring is to be shrunk on a solid carbide cylinder instead of a liner, the
diametral interference can be calculated by considering a to equal zero in the
above formula. In Figure 4-8 above, the tangential stress at the inner surface of the
steel ring due to shrink is:

t
=
P (b
2
+ c
2
)
c
2
b
2
The maximum compressive pre-stress at the inner surface of the carbide, due to
shrink, is:

t
=
-2 Pb
2
b
2
a
2
The shrink allowances listed below are general guidelines derived from practical
experience. Two levels of interference are shown with each appropriate for different
circumstances. Actual calculations with Lams formulas above are preferred over
the guidelines. Individual calculations should be performed when any new or
complex design is employed (i.e., high internal pressure on a die, thin walled
cylinder, unusual geometry, or elevated temperature). Consult General Carbide
application engineers for guidance.
Shrink Allowance General Guidelines-
Carbide Cylinder Mounted Inside a Steel Ring:
OD of
Carbide
Low (1)
Diametral
Interference
Medium (2)
Diametral
Interference
- 1 .0003 .0022
1 - 1 .0005 .0030
1 - 1 .0007 .0037
1 - 2 .0009 .0045
2 - 2 .0012 .0060
2 - 3 .0017 .0075
3 - 3 .0022 .0090
3 - 4 .0027 .0105
4 - 5 .0035 .0125
5 - 6 .0047 .0155
6 - 7 .0055 .0185
Low diametral interferences are intended for applications involving low to medium
torque and low internal pressures.
Medium diametral interferences are used in applications where the carbide cylinder
is subjected to internal pressure. The compressive pre-stress applied to the carbide
by the interference fit must be large enough to keep the carbide in compression
throughout the working cycle and not allow it to be exposed to tensile stress.
4-13
The temperature needed to achieve the assembly fits listed in column (1) of the table
above is 450F. Heat both the carbide and steel parts together. The difference in
CTE of the two materials will allow ample time to assemble the joint and to position
the two components before the joint sets.
When assembling a heated steel part over a carbide part at room temperature, the
steel part loses heat rapidly since the carbide acts as a heat sink. The assembly will
set very fast. As mentioned above, it is best to take advantage of the low CTE by
sizing the joint for assembly with both parts heated to the shrink temperature. This
will allow sufficient time to orient the parts relative to each other before the joint sets.
For medium diametral interferences as shown in column (2), it is best to heat only
the steel part to achieve maximum clearance between the two rings.
Shrink Allowance Guidelines for PM dies:
OD of Carbide Shrink Fit for Powder Metal
Dies
"- 1" .0006 - .0015
1" - 2" .0015 - .0025
2" - 3" .0025 - .004
3" - 4" .004 - .0055
4" - 5" .0055 - .007
5" - 6" .007 - .008
6" - 7" .008 - .0095
7" - 8" .0095 - .0105
Generally speaking, caution must be taken not to impose maximum interferences in
powder metal dies that have various geometries or case materials that range from
softer heat-treated steel to harder more brittle materials. Maximum interferences will
often split harder cases, thus the guideline above recommends interference fits that
fall between the low and medium recommendations in the general guidelines table.
4-14
For carbide liners with oval or complex geometry, it is a good idea to contact your
supplier for the thermal expansion rate of your specific grade of carbide.
In addition, another important consideration is the scope of the properties of the steel
case to be used in the shrink fit assembly. The steel case should have a pre-heat
treatment that will retain the mechanical properties of the material, despite any self-
tempering and/or self-stress relief that occurs during the shrink assembly procedure.
Shrink Guidelines - Carbide Cylinder Mounted Outside of a Steel Ring:
Normally, tungsten carbide should not be subjected to tensile stress, thus a carbide
sleeve shrunk over a steel shaft seems doomed to failure. However, sometimes it is
necessary to do so and it can be successful. A complete set of stress calculations
must be performed in these cases. Lams equations may be used if the subscripts
c and s are reversed. Additional stresses must also be considered. Any
operation above room temperature will rapidly increase the tensile stress at the
carbide I.D. leading to premature failure. When the carbide is shrunk on the outside
of the steel, the designer should use extreme caution and use only the minimum
interference necessary.
In summary, if it is necessary to shrink-fit a carbide sleeve onto the outside of a steel
shaft, all operating conditions and stresses must be considered. General Carbides
Engineering Department should be consulted when this type of design is desired.
Press Fits This mounting technique is an interference fit but no temperature is
involved. The carbide part is merely forced into the mating steel part by mechanical
pressure or force. Light interferences are generally used to permit assembly without
biting into the steel wall. Steel requirements are much the same as far as elongation
and yield strength are concerned, but the draw temperature is no longer critical.
Most tapered dies or pins can be assembled at room temperature. Pressing a pin
into the steel recess (Fig. 4-9A and 4-9B) requires a chamfer to properly lead the pin
into the hole
Figure 4-9
In cases where maximum interference is desired and the high temperature required
for shrink fitting would adversely affect the hardness of the steel, there is no
alternative than to use a heavy interference press fit. In this method, a taper in the
range of 1to 2included angle, is ground on mating parts so that they can be forced
together to give the required radial interference. This technique is used in the
multiple ring approach of assembling ultra high-pressure anvils or dies. The
tangential stress in each steel ring of the multiple ring anvil design, illustrated in Fig.
4-10, is calculated to approach the yield point of the steel in order to provide
maximum support to the anvil. This technique provides greater tensile strength than
one large steel ring. The assembly of the part should be accomplished by lubricating
all surfaces with a lubricious coating such as molybdenum sulfide or a suitable
lubricant.
4-15
Figure 4-10
Mounting External Rings and Sleeves In the assembly of some large diameter
carbide applications such as rolls or slitters, it is desirable to use external sleeves or
rings mounted on steel or a low expansion alloy. In order to avoid hoop stress in the
ring or sleeve, the carbide should be mounted by axial clamping as shown in Figures
4-11 through 4-13. In the case of slitter knives, the stresses are principally radial and
no positive drive is needed so axial clamping is sufficient. However, in high torsional
applications such as drive rolls, a positive drive key is needed. Fig. 4-11 through Fig.
4-13 illustrate proven methods where positive drive is used. Note that a lug should
be put on the carbide instead of a keyway put into the carbide. As previously noted,
a keyway provides a stress concentration and the drive lug is more dependable. A
large radius should be used for the inside corners in Figure 4-11.
Figure 4-11 Figure 4-12
Figure 4-13
4-16
4. Mechanical Fastening
Mechanical fastening represents a proven method for fastening cemented carbide to
steel. It cannot be overemphasized that when utilizing mechanical joining, no
allowances need to be made for the temperature difference between the thermal
expansion rate of carbide and that of the steel holding part or other material used. Even
when the temperature range is somewhat wide, and there would be a bimetal strip effect
as depicted in Figure 4-1, design modifications can be made to compensate for this
effect. Another advantage to be considered is that the steel-based member of the
assembly can be hardened and receive the required final surface finish that can be
easily matched later on by grinding or polishing of the carbide member of the assembly.
Design Rules For Mechanical Fastening
Certain principles of design should be followed when designing a mechanical mounting
for a carbide wear part or component:
a) To maximize support, allow the carbide to rest against a solid surface or shoulder so
that the clamping mechanism only has to hold the carbide in place and not resist the
direction of the operating thrust of the application. Secure the carbide in a pocket or
up against a stop for maximum rigidity and support.
b) A highly polished cemented carbide surface, such as a polished shank on a rotary
tool, looks beautiful good but is more difficult to grip than the as-sintered surface
due to its lower coefficient of friction. The same is true of a working knife-edge
where the body of the blade need not be highly polished to allow a clamping device
to better secure the blade.
c) Whenever possible, an assembly should be designed to take advantage of the
extremely high compressive strength of carbide. Thus, the designer should focus on
generating a compressive load on the carbide wherever possible but keeping in mind
that notches and keyways should be avoided because they represent (mostly tensile
and shearing) stress risers as discussed in Chapter III.
Methods of Mechanical Fastening
A variety of methods can be used to fasten cemented carbide mechanically. Some of
the most common are the following:
(a) Clamping The carbide insert is fitted into a pocket or supporting recess, then held
in place by one or more clamps. Fig. 4-14 shows a typical clamp design used for
wear parts and machine components. Many other similar variations or designs are
used. The key here is to exert uniform pressure over all contact surfaces. All sharp
edges nested in the recess should be broken as noted in Chapter III.
Figure 4-14
4-17
(b) Wedge This method is sometimes more practical than clamping if higher pressures
are needed to secure the component or if longitudinal movement of the carbide in a
slot or recess must be prevented by more than just friction against a clamp. The
coefficient of friction of cemented carbide against steel is quite low, therefore, the
ability to hold the piece may not be enough to a firm grip by means of friction alone.
Wedges can be designed to hold much tighter than clamps. .
Fig. 4-15 illustrates a common type of wedge mounting used today. The strip-type
wedge holds a carbide scraper blade by a hex head cap screw inserted into a drilled
and tapped hole, placing pressure on the wedge and capturing the carbide in
compression mode.
Figure 4-15 Figure 4-16
(c) Dovetail Dovetail mounting is closely related to wedge mounting and is depicted in
Figure 4-16. This technique utilizes the strength of the supporting surfaces to absorb
the working stresses while providing a uniform clamping pressure over the entire
carbide component. As in all types of mechanical mounting, supplementing rather
than opposing the major operating forces is important to the success of the
assembly.
(d) Screw Mounting The most practical and economical mechanical attachment
technique in use today is screw mounting. This method uses screws to attach
carbide to a steel supporting structure and the following figures illustrate the use of
countersunk holes with flat head screws. If holes are put in the carbide before
sintering, the spacing will vary somewhat but can be held fairly closely per the
guidelines noted in Chapter I. In all cases, the hole through the carbide should be
oversize, as shown.
Figure 4-17A Figure 4-17B Figure 4-17C
Use the carbide part as a drill jig in order to assure proper alignment of the
countersunk holes with the tapped holes in the steel body. An alternate solution to
align the holes is to use a nut in an oversize counterbore (such as is shown in Fig. 4-
17B). This method permits the screw and nut to float or shift to accommodate
variations in the center-to-center distance in the carbide part and is practical on
relatively thin carbide strips. It will not weaken the holding power under the screw
head.
4-18
Figure 4-18
Use a counterbored hole to accommodate a heavier screw head when a thicker
section of carbide is required. This technique is displayed in Fig. 4-18. All but one of
the counterbored holes can be elongated to permit variations in center-to-center
distance of the holes. Elongated holes in the mounting body may also be used.
Whenever counterbored holes are used, certain precautions are required to eliminate
stress concentrations as discussed in Chapter II. (see Figure 4-19).
Figure 4-19
Please note the large fillets and broken sharp edges of the carbide. A minor
modification in the shape of the part can reduce the stress concentration
considerably.
(e) Draw Rod Mounting Assemblies utilizing steel and carbide in which carbide is the
integral material of the assembly, such as a pump plunger used in an abrasive
pumping application, can be assembled with a draw-rod mounting technique, again
employing the carbide in a compression mode. This technique is depicted
conceptually in Fig. 4-20.
4-19
Figure 4-20
The high compressive strength of carbide makes it well suited for the compression
loading and the high tensile strength of the steel complements this technique. The
steel can be pre-heated and then all parts assembled. Upon cooling, the steel draw-
rod shrinks and the pre-load is applied axially to the carbide compression member.
Please note that the pre-load must be calculated so as not to exceed the yield
strength of the steel.
(f) Tapped Holes Tapped holes can be achieved in different ways, either by threading
of inserts or plugs or the use of external studs brazed to the non-working surface of
the carbide.
A common technique is to use INVAR plugs that can be brazed (see page 4-8) into a
hole previously green machined into the carbide workpiece. Because INVAR has a
CTE closer to that of cemented carbide, the resulting assembly is secure through the
temperature range of the braze filler metal. The insert is then drilled and threaded to
required position and thread size.
If the assembly allows for an external boss, a tapped stud can be brazed to the
workpiece. Studs are usually cross-slotted to reduce braze strains.
With the advancements in CNC machining, it is possible to put internal threads into
cemented carbide in the green state but care must be taken not to over-torque the
bolt and strip the carbide threads. Also, calculating shrink factors to ensure the
proper hole location is critical when employing this method.
Another technique, developed by General Carbide engineers, is to use a
combination of internal carbide threads and epoxy adhesive. In this method,
depicted in Figure 4-21, a hole is drilled and internally CNC threaded in the green
carbide to an oversize condition of .010 - .014 in diameter from the desired tap size.
After sintering, an A2 threaded steel plug is screwed into the hole with high temp
epoxy cement. The .005 - .007 gap per side allows room for the epoxy cement to
securely fasten the steel insert into the carbide. The carbide is then ready for screw
mounting.
4-20
Figure 4-21
This technique offers some advantages over the brazing of INVAR plugs:
1. no thermal stress induced by the brazing operation
2. plug can never pull out in service
3. plug can be removed if necessary
Please review the type of internal threads desired with General Carbide application
engineers before specifying on a drawing.
(g) External Threads In many assemblies, there is a need for external threads. While
threading carbide is possible by hard grinding, it reduces the cross sectional strength
due to the notch effect of the threads acting as stress risers. An alternate approach
is to braze a thin-walled steel or INVAR bushing in place and thread the bushing.
4-21
Chapter V - Finishing Techniques For Cemented Carbide
Cemented carbide parts can be finished to the desired shape, size, flatness, and surface
finish by diamond wheel grinding or by diamond lapping and polishing. In addition, EDM
(electrical discharge machining) has risen in prominence and popularity among carbide
fabricators. Both techniques are covered here in detail.
Fundamentals of Grinding
Diamond wheel grinding, in an overly simplified way, can be described as removing
undesirable portions of material from a part by subjecting it to repeated overlapping
contact with a rotating diamond wheel (Figure 5-1). During the grinding process, the
rotating diamond wheel is brought down on the work piece so that the tips of the
exposed diamond particles barely touch the surface to be ground (Figure 5-2). At this
starting point, the work piece is subjected to either a reciprocating or a rotating motion,
and the wheel is dropped further by an amount equal to the depth of cut (D
c
). The
process is repeated n times until the desired amount of material equal to n x D
c
is
removed.
5-1
Figure 5-1. The grinding process. The depth of
cut is exaggerated for illustration purposes.
Grinding Factors
It is apparent from the above figures that if the wheel is fed into the work piece deeper
than the exposed diamond length (i.e., if D
C
> D
L
), damage to either the grinding wheel
or the work piece will result. In cases where D
C
= D
L
, a considerable amount of heat is
going to be generated due to the rubbing that occurs between the work piece and the
wheel bonding material. Coolant used for removing excessive heat will also not be very
effective due to the collapse of annular space between the wheel and work piece.
Therefore, the ideal situation is when D
C
= 1/2 D
L
and is maintained throughout the
grinding range. In almost all types of grinding (reciprocating, cylindrical, centerless,
etc.), the feed rate is maintained at the depth of cut per pass. The greater the D
L
, the
greater the D
C
can be, resulting in a higher rate of material removal. The limiting factor,
of course, comes from the fact that to increase the D
L
, coarser grit diamond particles
must be used, which influences the surface finish of the part.
Surface Finishes
An as-sintered piece of cemented carbide has a surface finish of approximately 50 to
60 microinches and may have surface irregularities several thousandths of an inch deep,
as a result of the sintering operation. This as-sintered surface can be improved and
the part brought to correct size by proper grinding techniques and the appropriate
diamond wheel.
5-2
Figure 5-2. A typical diamond wheel
contains uniform but randomly
distributed diamond particles.
The following table indicates the finishes that can be obtained with good grinding
practices and the appropriate grit size of the diamond wheel. For lower microinch
readings, diamond lapping or polishing of the part will produce the desired result.
Grit Size Particle Size Expected Surface Finish
Micron Inch (R
A
)
80 267 0.0105 24-36
150 122 0.0048 14-16
180 86 0.0034 12-14
220 66 0.0026 10-12
320 32 0.0012 8
400 23 0.0009 7-8
600 14 0.0006 2-4
1200 3 0.0001 1-2
Grinding cautions
It is essential to avoid thermal shock caused by sudden changes in temperature, such as
loss of coolant flow. Thermal shock will produce grinding cracks, which can appear as
crazing and large cracks. Crazing is a pattern of hairline checks that appear on the
surface of the part caused by a rapid rise in the surface temperature of the carbide
relative to the cooler interior. These thin cracks can be difficult to detect and may only
be visible under magnification or by dye penetrant crack detection methods.
Thermal shock can also produce large, visible cracks. These will also result from an
extreme temperature gradient from the interior to the surface, particularly in a localized
area. Avoid excess pressure on a small contact area even when using the proper type
of grinding wheel to prevent this condition.
Grinding recommendations
Wet grinding is strongly recommended to minimize overheating and cracks. Grinding
fluids specially developed for carbide grinding, are commercially available and should be
used. These fluids do not contain leaching chemicals that will leach or degrade the
carbide structure (as mentioned in Chapter 3) yet they provide adequate heat
dissipation. The coolant should flow liberally onto the wheel and cover the entire carbide
surface. Interrupted or insufficient coolant flow will produce alternate heating and
quenching of the carbide surface and cause the carbide to crack because of these
temperature changes. If a machine is not suited for a continuous flow of coolant, a mist
spray should be used to keep the wheel wet and clean.
5-3
The following guidelines are presented to help avoid grinding cracks and crazing:
1. Never let coolant flow be disrupted
2. Avoid dry grinding unless the operator is extremely skillful and can quickly detect
heat buildup.
3. Generally, maintain a feed rate equal to the depth of cut per pass. Avoid an in-feed
per pass over 0.0003 on low binder, hard carbide grades, and over 0.0010 for
higher binder carbide grades.
4. Avoid using hard-bond diamond grinding wheels, such as vitrified and metal-bonded
diamond wheels. Resinoid-bonded diamond grinding wheels are recommended
when dimensional accuracy, flatness, and surface finish are important.
5. Dress the grinding wheel frequently to avoid a loaded wheel, particularly when
grinding steel and carbide in the same pass. If the steel is a blue color from the heat
of grinding, a crack will probably be found in the carbide. However, the crack may
not reveal itself until the assembly is in use.
Electrical Discharge Machining of Cemented Carbide
What Is It?
EDM or, electrical discharge machining, is the process by which a part is machined
using the erosive properties of electrical discharges. Most people think of EDMing as a
relatively recent discovery, when in fact it dates back to 1770 when an English scientist
named Priestly first discovered the corrosive effect of electrical discharges. However, it
was not until 1943 that two Russian scientist/brothers named Lazarenko developed the
idea of exploiting the destructive effect of electrical discharges and developed a
controlled process to machine electrically conductive materials.
The Lazarenko brothers perfected the EDM process. Simply put, it is a succession of
electrical discharges that take place between two conductors that are separated by a
non-conducting liquid called a dielectric. Today, this process still bears their name, the
Lazarenko Circuit.
EDM today has two basic types: wire and probe (die sinker). Wire EDM is used primarily
for shapes that are cut through a selected part or assembly. A hole must first be drilled
into the workpiece and then a wire is fed through the hole to complete the machining.
Probe EDMs are used for complex geometries where the EDM machine uses a
machined graphite, copper tungsten or copper graphite electrode to erode the desired
shape into the part or assembly. Probe EDM does not require a pre-drilled hole in the
part.
How It Works
During the EDM process, a series of non-stationary, timed electrical pulses remove
material from the workpiece. The electrode, workpiece, and the dielectric are all held by
the machine tool. A power supply controls the timing and intensity of the electrical
discharges and the movement of the electrode in relation to the workpiece.
5-4
An electrical discharge is initiated at the spot where the electric field is the strongest.
Under the effect of this field, electrons and positive free ions are accelerated to high
velocities and rapidly form a channel that conducts electricity. At this point, current will
flow and a spark will form between the electrode and the workpiece. This causes a
great number of collisions between the particles. During this process a bubble of gas
develops and its pressure rises quickly and steadily until a plasma zone is formed. This
plasma zone can reach temperatures in the range of 8,000 to 12,000
0
C. due to the large
number of particle collisions. This in turn causes instantaneous local melting of a certain
amount of the material at the surface of the two conductors. When the current is turned
off, the sudden reduction in temperature causes the bubble to implode, which projects
the melted material away from the workpiece. This leaves a tiny crater in the eroded
material. The dislodged particle then resolidifies into small spheres and is removed by
the dielectric.
Designing Carbide Parts With EDM In Mind
EDM, in past years, has been used to produce parts that were difficult to make by other
machining methods. Today, more parts are being designed to take advantage of the
EDM process. More and more design/manufacturing teams are using it as their first
choice to manufacture parts.
EDMing should be considered when the part being designed has very thin walls, has
small internal radii, has high depth to diameter ratios, or are very small and hard to hold
while machining. It also should be considered if the workpiece material is hard, tough,
burrs easily or needs to be heat-treated. This makes it especially well suited for
machining cemented carbide parts.
As noted in Chapter 1, cobalt is used as a binder in tungsten carbide to hold the particles
together when sintered. The amount of cobalt added determines the hardness and
toughness of the carbide. The electrical conductivity of cobalt exceeds that of tungsten,
so EDM erodes the cobalt binder in tungsten carbide. The carbide granules fall out of
the compound during cutting, so the energy applied during the cutting determines the
depth of binder that is removed.
When cutting carbide on certain wire EDM machines, the initial cut can cause surface
micro-cracks. To eliminate them, skim cuts are used. Skim cutting produces finer
finishes because less energy is applied to the wire, thereby creating smaller sparks and
thus smaller cavities. However, it is advisable to alert the carbide supplier that this part
is intended to be cut by wire EDM. This is done because some suppliers adjust the
formulation of the grade to provide maximum resistance to cracking during the EDM
process.
Some older wire EDM machines used capacitors. Since these machines applied more
energy into the cut, there was a greater danger for surface micro-cracking. Later, DC
power supply machines without capacitors were introduced. This helped to produce less
surface damage when cutting carbide.
5-5
Today, many machines come equipped with AC power supplies. These machines are
especially beneficial when cutting carbides because they produce smaller heat-affected
zones and cause less cobalt depletion than other machines.
To eliminate any danger from micro-cracking, and to produce the best surface edge, it is
a good practice to use sufficient skim cuts when EDMing carbide parts. Because EDM
does not involve contact with the workpiece, it is possible to design shapes in carbide
that would break during production when using conventional machining practices, such
as grinding. Parts that cannot take this type of stress can be machined effectively with
EDM.
Sometimes it is difficult to machine a part that has thin walls. EDM is ideal for this sort of
part because the process does not involve force, contact or deformation. A wire EDM
can be used on parts with wall thicknesses as thin as 0.005 inch. A probe EDM can
produce walls as thin as 0.002 inches. This makes EDM a good choice when designing
small surgical tools from tungsten carbide.
Another consideration for use of EDM is when the part being designed has a high ratio
of cavity depth to width, like slots or ribs. Since there is no force between the tool and
the workpiece, long electrodes can be used to make very intricate ribs. Wire and probe
EDMs are excellent for jet engine blade designs.
Use probe EDM when your design calls for difficult recessed cuts. Conventional cutting
tools cannot reach these cutting areas to apply the necessary force.
Hardness of the carbide grade is of no concern when considering EDM. Because the
EDM process vaporizes material, instead of cutting it, the hardness need not be a factor
for consideration.
Cutting Speed
Cutting speed is rated by the square inches of material that is cut in one hour. It can
vary according to the conductivity and the melting properties of materials. For example,
aluminum, a good electrical conductor with a low melting temperature, cuts much faster
than steel.
On the other hand, cemented carbide, a poorer conductor, cuts much slower than steel.
It is the binder, usually cobalt, that is melted away. Different carbide grades EDM at
different speeds because of carbide grain size and the binder amount and type.
EDM Limitations
Maximum workpiece dimensions for wire EDM are approximately as follows:
Y axis = 59 inches
Z axis = 24 inches
X axis = no limit
For probe EDM:
Y axis = 59 inches
Z axis = 17 inches
X axis = 98 inches
5-6
Maximum taper angle for wire EDM is + or 45 degrees
Maximum angle/height combination is 30 degrees at 16 inches high
Accuracy of wire EDM is about 0.00002 inches
Probe EDM is + or 0.0001 inch
Surface finish is about 4 microinch for wire EDM on a finish pass and 24 microinch finish
with a roughing pass.
5-7
References
Acknowledgement is given to the following publications and articles that were used as
reference documents in the compilation of this information:
1. Publication A80-184 (10) FO, Designing with Kennametal, Kennametal Inc,
Latrobe, PA.
2. Publication H-9100B - ENG, Understanding Cemented Carbide, Sandvik Hard
Materials, Sweden.
3. Harry C. Moser, When Should You EDM? Charmilles Technologies Corporation
and Mikron, Lincolnshire, IL.
4. Carl Sommer, Understanding the Wire EDM Process, taken from the Non-
traditional Machining Handbook, Tooling & Production Magazine, November
1999.
5. N. Langer, Where Adhesives beat Mechanical Fasteners, Dymax Corporation,
MACHINE DESIGN magazine, August 2005.
6. D. Small and P. Courtney, Fundamentals of Industrial Adhesives, Henkel
Corporation, Advanced Materials and Processes magazine, May 2005.
TRS
7/25/08