Ceramics and carbides as tool materials

Citation for published version (APA):

Gielisse, P. J., & Kals, H. J. J. (1975). Ceramics and carbides as tool materials. (TH Eindhoven. Afd.

Werktuigbouwkunde, Laboratorium voor mechanische technologie en werkplaatstechniek : WT rapporten; Vol. WT0358). Technische Hogeschool Eindhoven.

Document status and date: Published: 01/01/1975

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

AS

TOOL MATERIALS

P.J. GIELISSE

t

H.J.J. !<ALS

*

REPORT WT0358

JULY 1975

t Department of Chemical Engineering, University of Rhode Island, Kingston, R. I., 02881

U.S.A.

*

Division of Production Technology, Department of Mechanical Engineerinq, Eindhoven University of Technoiogy, Netherlands.

LOAD, WEAR AND APPLICATION OF CUTTING TOOLS

I I TECHNICAL REQUIREMENTS FOR CUTTING TOOL MATERIALS

1. High hot-hardness

2. Low chemical affinity

3.

High abrasion resistance

4. Low adhesiveness

5.

Low deforma t ion factor

6.

High toughness

7.

High fatigue resistance

8.

High resistance to therma I shock

9.

High creep resistance

10.

Inexpensive and eas i I

y

ground

I I I THE DEVELOPMENT OF CEMENTED CARBIDES

IV RECENT DEVELOPMENTS IN TOOL MATERIALS

A) Tool materials for machining high-alloy steels

1. Polycristall ine boron nitride on carbide

2. Ceramic materials

B) Tool materials for machining abrasive materials

1. Polycristall ine diamond on carbide

2. Polycristall ine diamond

C) Tool materials for machining conventional materials

1. Coated inserts

2. Surface treated inserts

3.

Ceramic cutting tools

ACKNOWLEDGEMENTS

B I BLI OGRAPHY

3

10 11 18

24

26

27

29

33

34

37

38 39

45

45

46

48

49

49

50

51

52

55

56

60

61

I. LOAD, WEAR AND APPLICATION OF CUTTING TOOLS

The requirements to be made on tool materials depend on the nature of the material to be machined, the choice of the cutting variables and the dy-namic properties of the machine tool. The phenomena wear and fracture, which determine tool life, are governed by physical, chemical and mecha-nical processes, which are caused by the mechamecha-nical and thermal load during machining. It is of importance that, prior to a further considera-tion of the various tool aspects, one is familiar with the forces, energies and temperatures coming into play in metal cutting. Below follows an out-line of the most important aspects of the metal cutting process.

During metal cutting, workpiece material is removed by shear (primary deformation zone, Fig. 1). The level of the strains and stresses generated in the process is determined by the choice of the geometric parameters (predominantly by the rake angle y ), the cutting variables (cutting rate

o

v, feed s(h) and the mechanical and physical properties of both workpiece and tool material. A so-called secondary deformation zone is found on the rake face, wherein most cases the characteristic crater wear occurs (dotted line in Fig. 1). This is represented diagrammatically in Fig. 1, which for purposes of comparison also represents the analogous situation in the grin-ding process. Note the characteristic differences between the angles y ,

o the dimensions h and the relative sizes of the plastic zones.

The value of the specific energy

_,

where E

t the total required energy per unit time, Vt = the removed volume per unit time,

for typical cutting rates between 1 and 6 m/s (60 and 360 m/min):

0.7 J/mm 3 for aluminum , 2.0 J/mm 3 for mild steel,

primary deformation v workpiece a chip

~~

I

-

-,-crater

-i~

,

plastic zone

,

workpiece \

,

_I ~

--.

--...

-b tool grinding gra in elastic zone

Fig. 1. Diagrammatic presentati~n of the

(In the case of grinding, these values are about 6 J/rom3 and 200 J/rom3 for rough grinding and high-finish grinding respectively). Of the total amount of energy, Et' more than 90 percent is carried off in the chip at high cutting rates, the rest is distributed over the tool and the workpiece. Most of the energy is transformed into thermal energy, around

two thirds in the primary deformation zone and the rest along the rake face (area of the secondary deformation).

The undeformed chip thickness may vary between 0.025 and 1.25 rom, the most frequent values being about 0.25 rom. The temperatures around the rake face, measured in average values, normally are between 500 and 9000 C. Temperatures up to 13000 C have been observed under certain conditions. Temperatures on the clearance face normally are one or two hundred degrees lower. The specific temperature and the temperature dis-tribution depend of course to a high degree on the metal cutting condi-tions chosen and on the material properties of both workpiece and tool.

A typical metal cutting process at 14 kW with miln steel (R =30) will c

readily causes a load of around 10, 000 N perpendicular to the rake face and about 5000 N parallel to it. On the active tool surface

there is a maximum normal stress of 103 N/rom2 (= 104 bar). Additionally, the maximum stress that results from $udden loading as happens during interrupted cutting, can have a magnitude of twice that resulting f rom stat1c loa 1ng: . d' 12

----

I workpiece I h _feed __ ----~-crater wear length KBo tool

Fig. 2. Cutting geometry in cross section indicating

Fig. 2 shows the cutting geometry in transverse section and puts for-ward the more critical areas such as the area of crater wear on the rake

face (with depth d) and the clearance face (with length VB)' The crater length is less than the total contact-length. The flat first part repre-sents the so-called 'sticking length'. Under certain conditions, work-piece material accumulates on the rake face near the cutting edge; this

so-called buH t-up edge (BUE) is broken down periodically, which may lead to crumbling away of the cutting edge. Fig. 3 shows the typical wear patterns. wear grooves in relief face nose radius plateau Fig. 3. cutting edge I

,

depth o~ cut I I I I crater

,

flank wear region

~7

/ crater width

~-~,L

_____ grooves corresponding with workpiece surface cu tt i ng edge

Typical wear pattern on a turning tooL

The following phenomena are clearly recognizable: crater wear on the rake face

flank wear

formation of lateral grooves on the clearance face rounding and wear of the tool nose

~vear on the auxiliary clearance face.

Fissures (microscopic and macroscopic ones), mostly parallel to or per-pendicular to the cutting edge, are a frequent phenomenon, both on the rake face and on the clearance face. It need not be emphasized that all types of wear need not occur simultaneously. The machining of certain high alloy steels (Inconel, Rene, Waspalloy) with high chemical affinity

and weldability to tool materials for instance, will normally show a very high crater wear. On the other hand, crater wear in cutting cast iron with ceramic tools is far less important, and the abrasive flank wear w'hich dominates here is of much greater influence on tool life. From this it may, however, not be concluded that crater wear does not occur at all with ceramic tools! This really does happen, as in cutting high-speed steel with ceramic tools.

Although this outline is mainly directed at the developments in the field of ceramics and cemented carbides, it is of importance to know that high-speed steel and stellite are still widely used as materials for

. *

cutting tools. The total consumption of tool materials 1n the U.S.A. 1n 1965, spread proportionally over the various kinds of material was:

Highspeed steel 65 %, carbide (and stellite) 32.3

%,

tool steel 1.8 % and ceramics (and diamond) 0.9

%.

The 10,000 tons of cemented carbides produced in 1968-1969 were consumed approximately as follows: cutting tools 25

%,

spikes for winter tires

(2 x 106 tires) 20 %, tools for mining raw materials 50 % and dies and other wear resistant applications 5

h.

In the mean time the usage of carbide has, particularly in the area of milling, been on the incrase. The application of ceramics has also grown considerably. An estimate of current usage indicates:

high-speed steel 40-45 %, carbides 50

%

and ceramics 5-10 %.

A general impression of the fields of application of the various types of tool material is given in Fig. 4. We should observe that the cutting speed of a particular tool material can be enhanced proportionally to the higher heat resistance - consider for instance the hot-hardness. the values of the cutting speed in the case of ceramic materials usually are twice as high as those for cemented carbides. In section IV-C-3 it will become evident that notwithstanding the good properties a ceramic will not always be the proper tool material. It should however. be pointed out that modern trends in the sector of ceramic tool materials contain indications that these will find ever wider applications. \.Je have in mind not only the

improvement of the generally applied aluminum oxide but also their 'alloys' as well as nitrides, borides and non-metallic two-phase

...-.. ~ ... '00 UJ ex:: :::>

·

·

...J

·

·

c:( • 1..1.. 75 •

.

UJ ex:: 0 1..1.. UJ co I-50 :::> u 1..1.. 0 :J: I-<.!:l :z 25 UJ ...J

\

\

.

_•

•

_•

·

•

_•

• •

·

_·

• _•

.

•

\

•

•

\

\ '. HSS

.

•

\

carbide " stell ite

"

...

'-....

_...

_.

... ceramIc

...

... alloy steel, Rc

=

21 1.5 mm feed o~---~~---~---~---~---~ Fig. 4.

a

2 3 t. 5

CUTTING SPEED

(m/s)

Fields of application of various tool materialf in maching an alloy steel. After H. Field and A.U. Daniels.1

or multiphase compounds. The figures 5 and 6 illustrate once more the differences between cemented carbides and ceramics and also put forward the necessity of reducing the cutting speed with an increase in hard-ness of the material to be removed. With increasing hardhard-ness, the possi-bility of varying the cutting speed decreases more rapidly in the case of cemented carbides than in that of ceramics.

40

-

III Q) ... :l c: 30 E ... UJ I..L. ....J 20 ....J 0 0 l-10 Fig. 5. 50 ..-.. 1.0

-III Q) ... ::J c: E 30 ' - ' UJ I..L. ....J 20 ....I a a I-Fig. 6. carbide 2 4340 steel. Rc=50 0.125 m m frev 1.25 mm feed without coolant 3

CUTTING SPEED

(m/s)

Fields of application of carbide and ceramic tool materials for 4340 steel with R

=

50.

c After M. Field and A.U. Daniels.1

/ ? ' ~ ?" ~ . / ~. 3

CUTTING SPEED

{m/s}

4340 steel Rc=55 , 0.125 mm /rev 1.25 mm feed without coolant ceramic

Fields of application of carbide and ceramic tool materials for 4340 steel with R

=

56.

c After H. Field and A. U. Daniels.1

II. TECHNICAL REQUIREMENTS FOR CUTTING TOOL HATERIALS

At the moment it is not (yet) possible to indicate what exactly are the precise requirements, expressed in the classical definitions of proper-ties of materials, to be fulfilled by a tool material under certain

~vork conditions. This is particularly due to the fact that one has not yet succeeded in laying down the quantitative contributions of various material parameters in a possible 'quality function'. This is to say

that at the moment, we are not yet able to establish unequivocally the material properties that are of importance, and then to determine pre-cisely their influence on friction, wear, fracture etc. The solution to the problem outlined is hampered in no small measure by the use of values of material properties at room temperature and at atmospheric pressure

in-stead of at conditions prevailing in metal cutting. In many cases the values at such conditions are simply not known. The solution is also interfered by the widely varying cutting conditions, examplified by the occurrence of continuous chips or intermittent removal, the changing nature of the load of the tool brought about by differences ~n properties of the workmaterial, wether or not a coolant or a lubricant of varied and usually unknown composition is being used which creates further

un-known effects and the use of machine tools with great differences in their dynamic characteristics (stiffness, natural frequencies, etc.). The fact that failure of a tool may be occasioned either primarily by fracture or principally by processes of wear (or, as mostly is the case,by a combina-tion of both)and the many possibilities that arise regarding nature of wear

(abrasive or 'chemical') and rate.of wear for different combinations of tool/workpiece material are further problems. It is foreseeable that in the future there will be a better chance to discover a 'quality function'

for the so-called primary fracture than for (secondary) fracture intro-duced by wear. Both phenomena are subject of very active research in the field of metal cutting.

With respect to the proper choice of materials, a summary in the form of a 'Ten Commandments for Tool Materials~ and principally based on quali-tative considerations of the most important properties is given below.

1. High Hot-hardness

Hardness at elevated temperatures is essential for cutting tool materials in order to withstand, under the prevailing process conditions, plastic deformation caused by normal (F ) and tangential (F ) load on the rake

yn y

face of the tool. See Fig. 1. High-hot hardness is one of the most im-portant properties of good tool materials. The influence of temperature on hardness is demonstrated for a number of materials in Fig. 7.

30 .-... N E E ₂₀ ... Z .::t. "-" (/') (/') w :z 0 a::: c:t :::c

t

10 0 0 Fig. 7. \

\ TiC

\ \ \

.

\ " .

,

...

~

,

..

_,

-

..

,

.

-

..

"

".

"

_"

'. WC-Co 6%

'.

,

_,

.

_.

_.

_.

'.

W '

..

.

. .

,

2

.

"

....

----

-

--

-4 6 8 10 12 14

-

TEMPERATURE

{ °c

X 102}

Influence of temperature on the hardness of tungsten, carbides and a carbide cutting tool material. Compiled from T. Tabor, A.G. Atkins and T.N. Loladze

Although TiC behaves somewhat more favourably than WC, both are subject to a very rapid drop in hardness with increasing temperature. The com-posite WC-Co, which by definition must show a different behaviour, pro-vides a somewhat more favourable picture in its usual compositions, and makes metal removal possible with these materials. Nevertheless, temperature continues to play an important role in the resistance to plastic deformation overall. The addition of TaC and NbC (up to about

]0 percent) to we-co and ve to TiC-Ni-Ho causes the hot-hardness to rise considerably. The additional carbides are soluble

in

the main carbi-des. For the rest, hardness and hot hardness of composites and polycrys-talline materials are greatly dependent on composition, porosity, grain size, and the properties of the constituent phases.

As regards the properties of the grains the following may serve as an example. The hardness of TiN, and hence also that of the individual grain, greatly depends on the composition (stoichiometry) as shown by the hard-ness values (at 50-gramme load).

Compound TiNO•59 TiNO•63 TiNO•B5 TiNO•92 TiNO•97 2 Hardness (N/mm ) 12 x 103

]4

x 103 16.3 x 103

17.B

x 103 19 x 103

In TiC,too, hardness is considerably influenced by its carbon content. TiC_O• BO becomes 'soft' at a temperature 'tvhich is 200 to 3000 Clower

than the corresponding temperature for TiC

O•96' Referring to the influence of porosity on hardnessJit has been proven repeatedly for various kinds of materials that the following relation may be used for expressing

. . 8.9,10

hardness as a function of gra1n S1ze.

H = H

o + k

l-~

where H a constant (intercept ordinate at 1

=

0)

o

k = a constant,

This expression appears to have the same form as the Hall-Petch relation for the strength of a material:

_1

a = a + kl 2

I

Although not established experimentally, it may be expected that the influence of porosity on the hardness of a material is not expressed very inaccurately if one writes

where H = H e-AP o H = hardness at P = 0, o A a constant ~ 7,

P volume per cent pores.

This behaviour 1S assumed on the ground of an empirical relation of

the same form, originally derived for the tensile strength and accounting for the corresponding behaviour between a en H mentioned above. As an example may serve the experience that the bending strength of polycrys-tal line A1

203 drops by around 50 per cent with each 10 per cent rise in porosity. The hardness of the individual crystals may greatly depend on the crystallographic direction (anisotropic behaviour). This behaviour manifests itself very strongly in the hardest material we know of

- diamond - in which the Knoop hardness for the various faces takes the

3 2 3 2

following values: (100) 54 x 10 N/mm, (116) 77 x

to

N/mm, and on the hardest face (Ill) 95 x 103 N/mm2• For TiC we have the values (110) 27.5 x 103 N/mm2 and (010) 22 x 103 N/mmZ. Only mean values can be given for strongly anisotropic crystals, unless the face or the direction is accu-rately defined. Also, hardness greatly depends on the method of prepara-tion. Differences in hardness between Al

Z03 as a pure monocrystal (sapphire) and the 'electro corundum' used in grinding wheels, both artificially pre-pared aluminum oxides, may be more than 30 per cent.

Hardness is determined in the first place by the binding energy between the atoms (molecules, ions) of the material. Fig. 8 shows the relation existing between the 'physical' hardness (with energy as unit) as deter-mined by one of the authors, and the technical hardness deterdeter-mined by the

Fig. 8. Relation between Wooddell hardness and physical hardness for a number of cutting materials

11 From J.N. Plendl and P.J. Gielisse.

Wooddell method; the latter is an arbitrary method used to measure the relative resistance to abrasive wear. The results of other measuring me-thods, such as the Knoop method, yield practically the same picture. This graph Fig. 8 is not only of importance because it shows the re-lation between physical and technical hardness, but it also illustrates the enormous differences in hardness between the various materials. One notices that the hardness of 98 per cent of all materials is below that of corundum (A1

203). Except for diamond, only synthetically pre-pared materials (i.e. non-natural ones) have a hardness exceeding that of corundum. l1oreover, in the region between cubic boron nitride (with a structure analogous to that of diamond) and diamond ~ a region covering half of the whole hardness area - no other materials are known. The

hard-ness values of i\fC and TiC average about 22 x 103 and 25 x 103 N/mm2 respectively; the composite WC-CO is appreciably less hard, viz.

about 17 x 103 N/mm2•

For a number of tool materials the decrease in hardness with increase of temperature is apparent from Figs. 9 and 10. Fig. 9 mainly refers to polycrystalline materials in use for cutting tools. Fig. 10 is con-cerned with monocrystals of materials which are used in abrasive (grin-ding) processes. From Fig.

9

we can see that at mean temperatures of

o

for instance 800 C,the hardness values of various materials show the same relative relationship as indicated for the mean cutting r~tes

of Fig. 4. The ceramics which at 8000 C show a hot-hardness twice as high as that of cemented carbide, appear to be applicable at cutting rates that are also twice as high. Perfect quantitative correspondence can, of course, not be expected. Fig. 10 shows that as regards hot-hardness, diamond is not surpassed. There are still enormous differences between diamond and the next lower cubic boron nitride. For that matter, the latter material exhibits quite a different relation between hardness and temperature. The figure does refer to the data for monocrystals of BN, SiC and Al

203 on their hardest faces (measured in vacuum). These abra-sives thus show a considerable decrease in hardness with rising tempe-rature. The various curves do not, however, intersect and hence the ad-vantage of one material over another remains at temperatures exceeding

13000 C.

13

The experimental data for Al

203 and BN may be expressed in the formula:

where H = H = 0 t = k = H = H e -kt/lOOO o hardness (N/mm ), 2

hardness at room temperature, temperature (oC),

.-, N E E ... z .:::t. ... (/) (/) I.J.J Z 0 0:: <t: :r: (/) 0:: I.J.J ::.:: L> >

t

F • 9.

r

20 _ceramic !A1₂0₃

l

carbide 10 ste

11

i te tool

o~--~----~----~--~----~---a

2 L. 6 8 10 12

--Hardness as a function of temperature for a number of polycrystalline tool materials.

100 75

1

-N E E '-_z ~ Vl V> W Z

£::) cubic boron nitride

a:.

«

::J: 0 a:. u ::E:

t

carbide O~--~----~---L----~----~----~--~

a

2 4 6 8 10 1 2 1 4

Fig. J O. _{Hardness as a function of temperature for a} number of typical grinding and cutting materials (single crystals). From T.N. Loladze~3

The value of k varies between 1.4 and 1.6. The 'ordinary' electro corun-dum usually shows a k-value of 1.4. If strengthened by means of Cr, Zr or Ti, k rises to 1.6, which is also the appropriate value for cubic BN.

In cutting, special care should be excercized to prevent the occurrence of plastic deformation as well as that of spontaneous fracture. As re-gards material properties, the occurrence of spontaneous fracture is closely associated with the value of rupture strength or the ultimate uniaxial strain of the tool material. Fracture phenomena that are not directly catastrophic, such as chipping, are also particularly detri-mental to tool life. They enhance to a great extent the chance of pre-mature failure and accelerate the normal wearing process considerably. Plastic deformation manifests itself in the form of loss of shape sta-bility of the cutting edge with consequent accelerated wear and increa-sed chance of fracture. As a material property, hardness at operating temperature constitutes a most important and easily determined quantity in connection with the occurrence of shear at the cutting edge.

In general one uses the relation:

where H

C

H CY

hardness,

conditional factor (often 2.8.-3) or constraint factor, Y = the 0.2 per cent yield point or, in the case of

elonga-tions below 0.2 per cent, the resistance to fracture. flore generally the micr6yield or flow stress.

On empirical grounds a plastic safety factor NT has been developed to approximately predict whether or not shearing of the cutting edge

or since y ~ 2T

where H

u

H u

the hardness of the relevant tool material at cutting temperature,

He('e)

=

the hardness (max. shear stress) of the workpiece material in the deformation zone,

H (, )

=

the hardness (max. shear stress) of the workpiece c c

material in the chip-tool contact zone.

At NT > I, no shearing of the cutting edge need be expected. At NT < 1, on the contrary, there is a great chance of shear at the cutting edge. Thus it appears that when grinding high-temperature resistant nickel alloys by means of corundum, where temperatures of the order of 11000 C may occur, NT becomes smaller than 1. Such alloys indeed are very dif-ficult to machine with corundum. The value of NT is characteristic of the process behaviour for different combinations of tool/workpiece ma-terials, and thus the inequality NT > 1 appears to be an important ma-chining criterion. For mama-chining the alloys mentioned above, alterna-tives are being looked for in the application of cubic BN (NT

=

7.5 at 11000 C) and SiC (NT

=

3.5 at 1100° C). Extensive application of the criterion NT > 1 is hampered considerably by the absence of reliable

values of the requisite properties under process conditions. This applies in particular to cutting tool materials.

2. Low Chemical Affinity

A low chemical affinity will counteract the following detrimental processes: mutual reaction of tool and workpiece material,

mass transport of certain components by means of diffusion.

Both will result in a mutation of the structure and the properties of the cutting tool material. The correct choice of material will generally go far to prevent chemical and diffusion wear.

material by means of the simplified notation

M + B +MB

the reaction will proceed to the right if the free energy decreases owing to the reaction. The ultimate free (reaction) energy will then be:

where

8G(T,P)

=

free energy change,

G

_x

=

the free energy; X

=

B, II, MB.

The more the free energy G(T.P) becomes negative, the tendency

to form -r-ffi will increase • Now if we have a comparatively large nega-tive energy for the tool material (this can also be the case for the workpiece or for both), then the tendency to form MB will be small. We

should bear in mind that even if this information is known to us, it does not yet tell us anything regarding the possibility of other reac-tions taking place such as for example :

and the quantity of reaction product that is formed. All this has to be established experimentally, and properties such as chemical potentials and activities have to be measured. One is aware that this kind of de-tailed information for metal cutting systems is hardly available. Using the information regarding the value of the free energy of the various tool materials, a good indication of possible reaction can, however, be obtained. Fig. 11 gives an outline of this. We observe that

we

does not exactly occupy a favourable position. Hence the generally serious crater wear in WC-Co cutting tools. TiC behaves far more favourably, and retains

this property without much change also at elevated temperatures. For this reason TiC is applied to cemented carbides in very thin layers (the so-called coated carbides, of which more later) to prevent crater wear. It

a

_{Fe 3C} we vc ..-.. _{- 25} NbC +-' TaC ro

.

Zre TiC I-Hfe 01 "- HfN

ro -

5 0 AI203 u .:::t. T102 >- Zr02 C!l 0::: I..LJ -75 :z I..LJ TIO I..LJ Hf02 I..LJ 0::: LL.

t

100 -125

o

0,5 1.0 1,5 2.0 -+- TEMPERATURE (oC xl 0

3)

Fig. 11. Free energy of a number of compounds that find

l ' • • • D f • 15

app LcatLon Ln tool materLals. ata rom H. SchLck

16 17

and R. Keiffer, after N.P, Suh •

has already been proven that TiN, with an even lower free energy, offers resistance to this form of wear still better. The fact that

the oxides of Zr and Hf play a great role in present research in the field of tool materials will require no further explanation if refe-rence LS made to the information given in Fig. II.

The weakening of tool materials by processes governed by diffusion is still more difficult to discuss quantitatively. First of all one may have to account for different processes : surface, grain boundary, and bulk diffusion. Diffusion speed decreases in the order indicated. Further, atoms or ions which do not occupy a specific place in the crys-tal lattice but move interstitially, diffuse approximately one order of

magnitude more rapidly than the so-called substitutional atoms (ions). The incorporation of substitutional atoms is more or less determined by the Hume-Rothery rule, which says that atoms or ions with equal charge and with radii not deviating more than 15 per cent from those of the atoms in the parent lattice, can easily be incorporated in it. This rule provides insight into the measure of occurrence of substitu-tional diffusion. The dimension of cations of Hf, Zr,

AI,

Cr and Ti is about equal to or falls within the 15 per cent limit of the dimen-sion of the carbon atom in

we.

Diffusion processes are subsequently influenced by the values of the diffusion constant D , which in turn

0,

depends on the atomic distance and the step frequency of the atoms in the lattice, and the activating energy for diffusion

Q ;

all this

ac-o cording to: where diffusion coefficient 3 D = the (mm Is), T the temperature,

R = the gas constant.

For

metaZs

Q

~ 20 RT ,T being the melting temperature. With respect

o m m

to these materials one might state that:

From this it can be concluded that it is preferable to choose tool ma-terials with a very high melting point, or that cutting should take place at process temperatures as low as possible. Carbides possess the highest melting temperatures, immediately followed by the borides, the ni-trides, the simple oxides, and subsequently the silicides, the multiple oxides, and the sulphides. Below is given a list of melting temperatures of raw materials that are for several reasons important to the manufacturer

18

Helting points (oK) Compounds

3875 - 3700 HfC, TaC

3700 - 3300 NbC, ZrC

3500 - 3050 HfB

2, ZrB2, TaB2, TiC, HfN, Th02

3050 - 2750 _{TiB2, VC, TaN, ZrN, TiN, Hf02' Zr02,}

2750 - 2500 Al

4C3, M02C, BeO, CaO

2500 - 2300 LaB₆, B₄C, SiC, AIN, VN, Cr

203

2300 - 1950 _{A1 203•}

Non-binary compo~nds with melting points of over 27500 C are not known. Compounds of more than two elements generally show melting points below

WC

o

2200 C. Some of these compounds, mainly those arising from the oxides of Ca, Ba, Sr, Ce, Hf, Zr, Th and Cr, show melting points in the range of 2750 to 23000 C. These materials are of importance in the development of future ceramic tool materials and special applications of carbides. with cemented carbides.

Coatings can form diffusion barriers which, immediatel,y belmv, resul t in substantial concentration differences such as between Co and HC.

The carbide acts in a very brittle manner and often provides the locus for crack initiation.

In order to prevent chemical reactions and diffusion processes, which in most cases are undesirable, the operating temperature must be kept as low as possible. The value of the formation energy depends on tem-perature and the situation becomes less favourable when the tempe-rature rises (Fig. 11.). The value of the diffusion coefficient also increases considerably with temperature. Often the point is overlooked that physical and chemical mutations may also have positive consequences. In certain cases the cutting or grinding process is promoted, if not made possible, by diffusion or chemical reactions. Incorporation of Cr in Al₂0

3 strengthens the lattice and reduces crater wear in tools of this material. In-diffusion of Mg prevents accelerated grain growth in Al

203 at high temperatures. Vanadium in TiC prevents rapid fall of the hot hardness of this material. Diffusion of Zr, Hf, Ti, Al and other ele-ments in WC-Co acts very favourably in preventing crater wear. On the other hand, there are lots of unfavourable processes. Diamond reacts with steel and forms Fe

3C, rendering grinding or cutting steel by the most 'ideal' material impossible. A1

203, as also cubic BN, reacts with titanium; grinding is impossible with this combination. WC dissolves in Fe at approx. 1000 to t1000 C, which results in heavy crater wear at such temperatures. Certain Ca and Si containing steels form a liquid

layer of glass of low viscosity on A1

crater wear.

In conclusion it may be mentioned that not only chemical processes between tool and chip but possibly also those between tool and atmos-phere may cause problems.

Fig. 12.

I

: 1t

_{o '} I-:z ~ 10 V) 0::: W > :z o u iN> 5

t

boron carbide

r

5 i 1 icon ca rb i de n , )~~--~---~~~---O---o 550 600 650 ~

TEMPERATURE

(OC)

Comparison between oxidation rates of B4 C and sic. Data from H.F.G. Ueltz1~

The graph of Fig. 12 serves as an example. Despite very high hardness, reasonable resistance to thermal stresses and other good qualities, boron carbide can not be used for, e.g., grinding. The reaction:

proceeds very rapidly, in particular at higher temperatures, and causes B

4C not to be applicable as a tool material. sic is satisfactory in certain cases. The difference in resistance to oxidation between the two materials is remarkable.

3. High Abrasion Resistance

In general, by abrasion resistance is understood the resistance to material removal due to purely mechanical processes occurring in the plane of contact of two bodies moving with respect to each other. Mate-rial removal is brought about by plastic deformation (formation of chips) and/or brittle fraction (granulation), caused by the interaction of loose particles and bound asperities. In general, the hardness of one material must at least be 2.5 times higher than that of the other to make possi-ble effective penetration. Therefore, in principle no purely abrasive wear can occur between two materials of the same hardness. In the case of composites it is of importance to realize that one of the phases is sometimes appreciably less hard that the other and therefore may wear off much more rapidly. The soft phase sets the degree of resistance to abrasive wear of the material as a whole. Actually both bodies will lose material ~n the case of mutual friction. Although a large difference between the wear rates in both directions will occur, wear of the hardest body may sometimes be significant. This point is often overlooked.

The degree of abrasive wear is thus determined by properties of materials - in particular the relative hardness - and system parameters such as contact pressure, relative speed, and surface roughness. This form of wear contributes greatly to tool wear taking place when cutting with ce-mented carbides at low speeds (flank wear), and generally plays a very

important part when high-speed steel tools are used. Abrasive wear is decisive in cutting non-metals with ceramic, carbide or diamond tools. Diamond tools are often used in grinding, cutting and drilling such ma-terials as rocks, minerals and synthetic inorganic mama-terials.

It is not easy to establish a measure of abrasive wear quantitatively. The values of the material parameters as well as of the system parameters may differ enormously in the various cases. An adequate quantitative assessment seems only possible empirically and after profound study of the system involved. Literature references are not usually available. In Fig. 13 an attempt has been made to give some insight into the relative

l.LJ 60 u z <x::

.--

_tf) tf) SO l.LJ 0::: l.LJ > 40 t/') <x:: 0::: a:l <x:: UJ > 30

.--<x:: . J l.LJ 20 0:::

t

10 0 0 Fig. 13. \ 6% (895l ( WC-Col (883 J 6 °/0

It

(i.1.. A)

,

\ [SSAl 2 3 4 5 6

-

() 2/2 E, (N/mm 2)

Relative resistance to abrasive wear of a number of WC-Co grades. Numbers in brackets

General Electric grades.

resistance to abrasive wear for several kinds of abrasion

re-sistant cobalt bonded tungsten carbides. The modulus of resilience, 2

cr /2E, has been chosen as independent variable and the wear unit has been defined as

abrasion resistance

=

l/volumeloss.

The volumeloss of the various grades was measured by means of a stan-dard wear test with aluminum oxide in water as the abrasive medium. In accordance with the above it appears that the resistance to abrasive wear of the materials involved decreases rapidly with increasing cobalt

content. A significant part is played by the grain size. This appears

*

from a comparison between the types 895, 883, and 55A of G.E., all of which contain 6 per cent cobalt, and of which the mean estimated grain diameters are 3, 5 and 7 vm respectively. The resistance to abrasive wear of carbides having a high percentage of TiC (and also the Ni-Ho bonded titanium carbides) is mostly one order smaller than the best we-co

grades such as the type 999 or 895. It is known that the so-called 'micro-grain' carbides (grain size belo~ 1 ~m) possess a higher resistance to wear than a standard grade with the same cobalt content. Abrasive wear of cemented carbides in aluminum oxide spraying processes (sandblasting) exhibits the same tendency as the one outlined in Fig. 13. For the rest, the value of the parameter a2/2E appears to be a fair measure of the re-sistance to ,.;rear on aluminum oxides. The higher the values of a2/2E the less the material will wear.

4. Low Adhesiveness

During metal removal, attempts are made to prevent as much as possible adhesion between workpiece material and tool material. Adhesion often gives rise to so-called adhesion wear. Strong adhesion prevents wear in the adhesion plane, but stimulates fracture in the immediate neighbour-hood. This form of wear manifests itself by relatively large particles of the tool material being broken off. The occurrence of the built-up edge (BUE) is also promoted by a high degree of adhesion and may lead to fracture along the whole cutting edge. Chatter mostly contributes to high welding strength. Further, adhesion wear is particularly detrimental to tool materials having a strongly heterogeneous structure.

Generally attempts are being made to prevent adhesion by making an ade-quate choice of tool material and also by using liquids which should prevent direct contact between chip and tool material (lubrication), therefore rendering impossible the reaction between the two materials. In certain cases the effect of liquids is clearly perceptible, but expla-nations for this are still very controversial. It is possible to make certain workpiece materials better machinable by the addition of certain elements e.g. lead in the case of steel. Addition of other elements such as silicon makes the material more brittle and prevents adhesion. Silicon

cau~es a silicate type of protective coating on carbides having a high titanium or tantalum content. The assumption that a low tendency for work-hardening of the workpiece material helps to prevent adhesion seems justified. The degree of adhesion must be established experimentally. Determining

at a comparison of the behaviour for different cases. Very little pioneer-ing has been done in this field. It is of importance to distpioneer-inguish between adhesiveness and chemical affinity

(11-2.).

Low chemical affinity does not a priori mean low adhesiveness and vice versa. In metal cutting.one tries to restrict the occurrence of both phenomena as much as possible.

S. Low Deformation factor

Due to the fact tl1at all cutting tools are wedge-shaped and nearly always have the same geometry, the relative resistance against elastic defor-mation can be expressed by the value of Young's modulus (E). In this re-spect a high E-value is always favourable. Regarding the clamping pro-blems which arise in the case of indexable tools, however, a relatively

low E-value is advantageous. Particularly in the case of inserts,the re-sistance to plastic deformation is much more important. This can be ex-pressed by the ratio between the E-modulus and the yield stress (Y). We shall call this ratio the deformation factor S:

S

= Ely

The value of the deformation factor 1S 300 - 1000 for metals, about

100 for most ceramic materials, and around 2S for glass and polymers. Calculations for a series of cemented carbides yield values between 1]0 and 160. The low values for glass are due to the low E-value, while crystalline ceramics have a relatively low value because of the very high value of Y. A high value of the deformation factor is undesirable

if deformation is to be prevented. The stability of the shape of the cutting edge is low in that case. Generally, this is dictated by a low Y-value. All this can be extended by the following considerations. As has been mentioned above, (II - 1, High hot-hardness), the hardness of a material can be expressed by :

from which follows the (compressive) yield strength:

Y := H/3

This ratio applies to metals. The yield stress of certain ceramic materials (KBr, NaCl, MgO and TiC) seems to be considerably lower.

In these cases it is observed that often:

Y '" H/35

This is due to the fact that these materials are very anisotropic

and/or 'Vlill very much strainharden, or that they show irregularities (im-purities, porosity) which decreases the compressive strength but not the hardness. Another factor is the measuring accuracy, which with these hard materials is rather problematic. Moreover, owing to a com-pressive load, phase changes often occur which are not accounted for. A familiar example is the melting of ice under a skate, which for that matter makes skating possible. More accurate investigations carried out recently, however, show that most ceramics follow the expression Y :::: H/3 when the value of Y is derived from compression strength. ~ve

then arrive at:

or, since Y :::: 2T (T

3E S '"

_H

In the case of carbide cutting tools it appears that the compression strength is very nearly twice the yield strength. This means that

2

S ~

€

C

where € is the ulimate uniaxial strain in compression.

c

The deformation factor concerns the resistance to plastic deformation of the different materials, while the plasticity factor (see item 1, NT) concerns the material behaviour of the tool in cutting, i.e. in its relation to a1_{particular workpiece material. In conclusion the}

temperature dependence of Young's modulus is mentioned. As an example, for A1

203 the value of E decreases by around

1/3

in the temperature range between 250 C and 16000 C. The influence of porosity can

gene-23

rally be expressed by:

where:

E

=

E (1 - 1.9 P) a

E the value of the modulus at P

=

0, o

P

=

the porosity in volume per cent.

6. High Toughness

While the deformation factor gives a good impression of the possibili-ty of plastic deformation occurring, toughness is rather a measure of the total deformation that is possible before fracture ensues. The parameters hardness, stiffness and toughness are interdependent. They are actually different definitions in which we try to represent the behavioural characteristics of a material. Since it is not yet possible to describe the mechanical behaviour univocally with the aid of inde-pendent variables. We shall have to rely on these separate definitions.

In principle, toughness is the capacity of a material to deform until coherence is broken (fracture). This may include elastic as ';'lell as plastic deformation. A measure of toughness is the ultimate uniaxial strain. A material behaves brittly if the elongation curve is linear till the point of fracture. The ultimate uniaxial strain in brittle materials is sometimes so small that it is difficult to measure. With completely 'brittle' materials we will observe spontaneous fracture,

while tough materials show appreciable plastic deformation before fracture. Another definition of toughness is found in the amount of energy that the material is capable of taking up till the point of fracture. It is clear that materials showing a high ultimate uniaxial strain are also capable of taking up a large amount of energy per volume unit. A dif-ficulty is formed by the determination of experimentally comparable values of this energy. It should also be remarked that toughness can be measured by different tests such as the tensile, bending and notched-bar impact tests. A short exposition of definitions of toughness is given below.

The deformation energy per volume unit of material is given by the area under the stress-strain curve:

o where Ed = deformation energy, V

=

original volume, a 0'

=

nominal stress, E

=

strain.

v

=

o

and we observe that more energy can be taken up according as the material has a smaller E-modulus and a higher rupture strength. The fact that this definition of toughness is applicable to cemented carbide and ceramic tool materials is associated with the fact that no or only little plas-tic deformation occurs before failure. Toughness values that include plastic deformation energies (so-called integrated values) are not known.

The critical value of the stress intensity factor, K, derived from the 'weakest link! mode of fracture m.echanics and based on Griffith formalism

_,

where: ! K=o(rrc)2 and thus K

o

(m;)~

or where c c K =

~E

• 2y c

(critical) stress intensity factor, (tensile, bending) strength,

E

=

elasticity modulus,

y

=

surface energy,

c half the crack length.

~s also taken as a measure of toughness. The factor K is called the c

fracture toughness. For materials which show a substantial amount of plastic deformation, K is expressed as:

c

where G ::: 2(y + y ) is the critical fracture energy and y the plastic

c p p

deformation energy. Application of these quantities to cemented carbides is limited owing to the inhomogeneous nature of these materials. }1easure-ments of bending toughness where E has to be in the formulas by

E = EI

b (I

=

moment of inertia)

and experimentally determined values of the (notched or unnotched) impact toughness offer further possibilities. It appears that a linear relation exists between the ultimate uniaxial strain and the unnotched impact toughness of cemented carbides~5

Therefore, to realize some comparison and estimation of the deformation energy, one actually only needs the values of E and I, considering the limitations above. The behaviour of E has been discussed before. Frac-ture strength of brittle materials depends chiefly on grain size, poro-sity and quantity as ,yell as properties of the (sinter) admixtures • It has been possible to determine. empirically the following expressions as re-gards the quantities mentioned above.

[

k) and a: constants k d- a

-cr = I d grain size -bP

[

k2 and b: constants cr

=

k 2e

-P : porosity in volume percent'

[

v admixtures in volume per cent

G = _k3

-

r counteracting grain growth .

v

-r average radius of the grains

of admixture

G grain size

For many materials the value of a has been established fairly accurately

20,24

at approx. 1/3 and that of kl at around 60. In general we have, therefore:

- 1/3 - bP v 1

cr

=

Kd

e (_)2

r

From a strength point of view it will be desirable to compose materials with a grain size and porosity volume as small as possible (the effect of ad-mixtures is disregarded). One of the authors has succeeded in expressing

the toughness in terms of deformation energy for polycrystalline aluminum "d 21

ox]. e as :

E

from which it may be concluded that for instance aluminum oxide having a gra~n s~ze of 2 pm is capable of taking up seven times as much energy as aluminum oxide having a grain size of 40 pm. This is seen as the rea-son for the success of the so-called micrograin carbides (d ~ 1 pm).

Fig. 14.

A=

grain size

A1>A2 >A3

· .. i ncreas i ng

~Al

TOUGHNESS

Influence of composition and grain size on hardness and toughness (qualitative)2~

As regards cemented carbides, figure 14 shows roughly the relation be-tween toughness and hardness. Since modulus and strength depend consi-derably on temperature, the possibility of taking up deformation energy is al so determined to a large exten t by temperature. As an example the value of the rupture energy of A1

203 at 1600

0

C is only one fourth of that at 250 C.

7. High Fatigue Resistance

Even if the stresses in the tool during cutting are below the rupture strength, allowance must be made for the chance of fracture as a result

of periodic changes in the load (fatigue fracture). Low fatigue strength may affect tool life very unfavourably. In general, little attention is

paid to fatigue phenomena in tools; among the reasons is the fact that only little quantitative information on fatigue strength of, for instance, cemented carbides is known. In metal cutting the nature of the load is mostly variable. The dynamic component of the cutting force contains se-veral frequencies, and the composition depends on the material to be ma-chined, the tool, the machine, the cutting rate and the feed. The frequen-cies normally lie bettveen 0 and 20,000 Hz. Ordinarily it is not feasible and economically certainly not justifiable to give attention to fatigue influences during continuous metal cutting. An exception is perhaps made by the machining of such materials as titaniL® where as a result of the formation of crumbling chips highly varying cutting forces occur. For this phenomenon as also for processes with interrupted cut the following as-pects may be considered:

A high ratio between the hardnesses of the tool material and the material to be machined diminishes the chance of fatigue frac.ture. Fatigue strength for high load frequencies is generally greater than for low frequencies.

Fatigue strength for cutting at fixed conditions may differ from that at varying load conditions. Differences in tool life can only be determined empirically for specific conditions.

Surface conditions play an important part. fatigue strength in the presence of coarse surfaces is smaller than that when 'smooth' sur-faces are involved. It would be commendable to polish tool sursur-faces

i f only this were economically justified. The chance fox fatigue

phe-nomena and/or initiation of cracking at the surface is influenced by residual stresses which are introduced during grinding of the surface. Corrosion (before and during the use of the tool) may influence fa-tigue strength adversely. Keeping tools clean, especially before use, is important. Even finger perspiration (acid) way have injurious con-sequences.

The higher frequencies are the more damaging. They are often generated as a result of the specific clamping mode of tool or insert.

8. High Resistance to Thermal Shock

Large temperature differences between rake face or clearance face on the one hand and the bulk of the tool, such as occur at the beginning and

the strength of the material. This influence may lead to chipping and local crack formation, which will result in accelerated -';vear and even catastrophic fracture.

The thermal stresses may be calculated from the following formula:

where E

=

11T = v

=

Il.

=

A

:::::

::::: A Young's modulus aE

-

\)

the temperature gradient Poisson's ratio

linear coefficient of expansion function of the Biot number

The value of A lies between 0 and 1, depending on the size of the test piece, the value of the heat transfer coefficient and the coefficient of thermal conductivity. In the case of infinitesiMally rapid cooling,

A

reaches its maximum value of 1. The temperature difference is a result of a heat flux ~

which analogous to Ohm's law can be written as:

=E

k

tp L

where: LIT = the temperature gradient per unit length, k = the coefficient of

It now follows from the above that:

thermal conductivity.

cr _ Il. E

th - A 1 - \)

<P L

k

There are two definitions of the resistance to thermal shock:

where a

b is the rupture strength.

The quantity R1 is used in the case of very rapid phenomena (heat conduc-tivity normally plays no longer a part then), while R2 is applicable for conditions in which the temperature gradient is limited by heat flow. If

the value of Poisson's ratio is held constant (it is

approxima-tely the same for cemented carbides (= 0,25), the relative resistance to thermal stresses can be expressed a.s follows:

This is the formulation of the resistance to thermal shock as generally This is the formulation of the resistance to thermo shock as generally

applied. A high resistance may be due to a high value of the mechanical

crb k

part ~ and/or of the thermal part a of the above expression.

. Ea h . . .

The quant~ty ~ represents the t ermal stress sens~t~v~ty.

R can also be expressed as:

t

k a with E

f as the ultimate uniaxial strain in tension (or bending).

This way of formulating implies that RI and R2 refer actually only to a purely elastic behaviour preceding the fracture, and that these

defini-tions are only meaningful in the case of materials that exhibit an essential-ly brittle behaviour.

Materials having great resistance to thermal shock are therefore those that, owing to their composition and structure, have high ultimate uni-axial strain, conduct heat very well and have a low coefficient of ex-pansion. In the case of ceramics the coefficient of expansion appears to be the most influential parameter. SiC is actually the only material among the monocarbides (B

4C, ZrC, VC, HfC, NbC, WC, W2C) that offers a satisfactory resistance to thermal shock. This is a result of the thermal conductivity coefficient being approx. three times that of the other car-bides, added to the fact that the expansion coefficient is equally small.

In the case of cemented carbides the coefficient of expansion varies little. The grades having a high TiC-TaC content show the highest values, lying up to 50 per cent above those of WC-Co. On the contrary, the heat conductivity is diminishing as the TiC-TaC content grows larger. Of par-ticular importance is the greater heat conductivity of the WC-Co compo-sites. Its value increases according as the WC content increases. A high TiC content results in a high thermal stress sensitivity. Figures 15 and 16 illustrate the behaviour of the characteristic quantities for a number of qualities arranged as to ISO classification. It will be clear

that for operations such as milling the K-grades are indicated, if only for reasons of thermal shock resistance.

Sensitivity to thermalshock is particularly important in the case of ceramic tools. For polycrystalline Al

203 this sensitivity is about ten times that for the average carbide grade and a hundred times that of high speed steel (measured at room temperature). A typical phenomenon in the behaviour of white ceramics is that their strength decrease

ab-N E ,t,41 I I'? ' 0

-3 )(

I

-....

Fig. 15. : - : . : : : : -.. 5t - ; -K01 1(10 K20 M20 M40

Ultimate uniaxial strain in bending (EfT)' thermal stress sensi-tivity (St) and resistance to thermal shock (R

t) of a number of (ISO) K and M grades.

R_t

a particular temperature (about 200 to 2500 C for A1

203 cooled in water). Black ceramics on the contrary exhibit a gradual decrease in strength after ther-mal shock. ?-1oreover, the

tempe-rature at which this gradual decrease occurs is a few hun-dreds of degrees. higher than for white ceramics. Like strength and Young's modulus, the coeffi-cient of heat conductivity is strongly influenced by porosi-ty. The behavioral trend

is represented in the following relation:

k

=

k (1 - P)

o

4 _fX105 _{where k}

₌

_{thermal conductivity}

o

..,

's? at P

=

0, with P

=

the volume

3 x

l- .15 per cent of the pores.

....

...,

"

2

t

110 ... S? )(

,

" .50

...--....

~I""

.25 .~

ur

+

POl Pl0 P20 P30 P40

Fig. 16. Ultimate uniaxial strain in bending (sfT)' thermal stress sensitivity (St) and resistance to thermal shock (R

t) of a number of (ISO) P grades.

9 .JHgh Creep Resistance

to a degree not to be ignored. Commonly, however, the influence of creep on tool life will be insignificant. Very little specific informa-tion on creep behaviour of cemented carbides and the influencing fac-tors has been published. Therefore, we shall have to restrict oursel-ves to.a few general guide-lines. As a rule of thumb it may be assumed

that cold creep occurs at values of TIT (T = melting temperature) of

m m

over 0.25. High-temperature creep occurs starting from TIT

=

0.4 to m

0.5. It is quite possible that at the common cutting temperatures both types of creep, separatelyin the various phases, occur in the compo-site. The principal deformation mechanisms which cause creep are: local slip, grain boundary slip and material transport from areas under com-pression to areas where tensile stresses occur(diffusion).Hence, creep can be counteracted by

the presence of a very finely divided carbide phase in the binder (decreased local slip),

choosing a coarse-grained grade (fewer grain boundaries), the use of grades having the highest melting or decompo-sition temperature (less diffusion).

The user of the tool will mostly have to accept the product offered and moreover, will be given little quantitative information on the

resis-tance to creep.

10. Inexpensive and Easily Ground

As regards the buying of tools, inexpensive does not mean advantageous. In general one can say that production costs depend more on the use than on the price of the tool. This implies that the choice of the quality can not be made to depend on price alone.

The grindability of a prospective tool material is important with respect to the possibility of the economical manufacturing of tools. This aspect also comes to bear when tools have to be reground (e.g. in the case of milling), All this makes grindability an important criterion in the

It must now be evident that not all desired properties can be combined in one material. Most materials which show the so important high hot hardness, appear to be very sensitive to thermalshock. Materials which do combine high abrasive resistance with low chemical affinity and low adhesiveness (such as BN and diamond) are very expensive. Other com-pounds are very hard, have a high modulus of elasticity and high fati-gue strength (as B

4C and SiC) but exhibit an insufficient chemical sta-bility (B₄C) or a poor toughness (SiC). A better choice would be Si3N4 mentioned above, or one of the Sialons·. Other materials will satisfy

nearly all conditions except the so important low chemical affinity. For instance, during cutting A1₂0

3 and BN react disastrously with ti-tanium, diamond with steel and

we

appears to be soluble in iron above certain cutting temperatures. An ideal tool material does not exist. A compromise will always have to be made in which the properties of

the tool material are adjusted to those of the workpiece material. The machining conditions and external influence must also be considered.

The recent developments as described in the following chapters, will be found to center on three important concepts:

Correct choice of tool and workpiece material.

Improvement of the existing materials or the development of new materials.

The use of different materials for differently loaded parts of the tool.

III. THE DEVELOPMENT OF CEMENTED CARBIDES

The very brittle tungsten carbide (WC) cannot just be used as a tool material. As early as 1927, Schroeter succeeded in combining the brittle WC-grains into a Co binder in a composite (WC-Co). The good qualities of the composite are the high (hot-)hardness and, with respect to WC it-self, reasonable toughness. Moreover, it appeared that the two materials