Basic Heat Treatment Term

CHAPTER -1 INTRODUCTION

1.1 Basics of Heat treatment

1.1.1 Hardening processes

The general purpose of heat treatment process is to vary the hardness of the metal

or to produce a structure or a surface that facilitates the progress of subsequent

manufacturing operations (Payson, P., 1943) or the proper fimctioning of the product.

Hardening

Hardening consists of heating of metallic materials to hardening temperature,

holding at that temperature, followed by rapid cooling such as quenching in water, oil or

salt baths. The high hardness developed by this process is due to the phase

transformation accompanying rapid cooling. The term rapid cooling simply does not

mean that cooling rate is higher than that adopted in annealing and normalizing. What it

really means is that cooling rate is equal to or more than the critical cooling rate of the

given material. Rapid cooling results in the transformation of austenite at considerably

low temperatures transformation into non-equilibrium products. The product of low

temperature transformation of austenite is martensite, which is hard microconstituent of

steel (Schimizu, N. and Tamura, I., 1978).

Brayans, R.G., et al., (1969) analyzed the mechanism of phase transformation in

crystalline solids and observed that the hardening temperature depends on chemical

composition. For plain carbon steels, it depends on carbon content alone. Hypoeutectoid

steels are heated to about 30-50°C above the upper critical temperature, whereas

eutectoid and hypereutectoid steels are heated to about 30-50°C above the lower critical

temperature. The phases, Ferrite and pearlite transform into austenite at hardening

temperature for hypoeutectoid steels. This austenite transforms to marteniste on rapid

quenching from hardening temperature. The presence of martensite accounts for high

hardness of quenched steel. If hypoeutectoid steel is heated to a hardening temperature

equivalent to that for hypereutectoid steel, the structure will consist of ferrite and

austenite. This will transform to femte and martensite on quenching. Femte, a very soft

phase, lowers the hardness of hardened steel considerably. Thus, the very purpose of

hardening, i.e. attainment of improved mechanical properties is lost. This is also known

as incomplete hardening, and is not usually employed in practice. The preferred

hardening temperature for hypereutectoid steel lies between the lower critical temperature

(A,) and the upper critical temperature (Lm). The advantage gained from hardening

temperature in this range is two-fold. The first is related to the presence of cementite in

hardened steel. The cementite in a martensitic matrix accounts for several desirable

properties. Wear resistance is one of them. The fact that both martensite and cementite

are hard constituents their presence is responsible for high wear resistance of the resulting

microstructure. Cementite is harder than martensite and so wear resistance of the two-

phase microstructure is better than what is achieved by martensite alone. The second

advantage of this hardening temperature is the attainment of fme martensite in the final

structure. In fact, heating of hypereutectoid steel above the upper critical temperature

(km) for hardening is detrimental because such a high temperature will result in

coarsening of austenitic grains and decarburization at the surface. Coarse austenite will

transform to coarse acicular marteniste (microstructure) which has poor mechanical

properties. Decarburized surface responds poorly to hardening treatment. In addition to

these factors, quenching from such a high temperature will introduce severe internal

stresses in the hardened steel.

Quenching media

When hardening tools and machine components, the heat treater generally aims at

obtaining a martensitic structure, at least in the surface layers of the steel. Hence, the rate

of cooling must be controlled so that the formation of femte, pearlite or bainite is

avoided. For cooling the hot metals, the various quenchants used are water, oil, air, salt

baths, mixtures of water and oil (emulsions) and polymer quenching media.

During the quenching of steel in liquid media, the whole operation may be split

up into three stages, viz. the vapour-blanket stage, the boiling stage, and the convection

stage, as shown in Figure 1. l(Christian, J.W., 1975).

CONVECTION I

Figure 1.1 Three stages of the cooling curve

During the vapour-blanket, stage a thin vapour film form, which is, heat

insulating and thus prevents heat abstraction. Therefore, the curve has rather flat

appearance in this range.

During the boiling stage, there is high rate of heat abstracting which in part is due

to the energy consumed by the vapour formation and in part to the vigorous agitation.

During the convection stage, the heat abstraction takes place more slowly and the

curve is flat once again.

Tempering

Jaffe, L.D. and Buffurn, D.C., (1965) investigated the Upper Temper

Embrittlement of Ni-Cr-Steel and found that, when steel is hardened to a martensitic

structure the toughness is rather low. On tempering, i.e. heating it to some temperature

between 160°C and 650°C, the toughness increases considerably.

Austempering is special heat treatment process in which austenite is transformed

into bainite. Similar to austempering, Martempering involves heating the steel to the

austenitizing temperature, followed by quenching in a constant temperature bath

maintained above M, point. The usual temperature of the bath lies between 180°C and

250°C. Steel is held in the bath till the temperature throughout the section becomes

uniform and is equal to the bath temperature. As soon as this temperature is attained,

steel is withdrawn and cooled in air. The cooling rate should be sufficiently high and

holding time considerably short to prevent transformation of austenite to pearlite or to

bainite. Martensite is formed in the second stage, namely during cooling in air.

Sub-zero treatment

The resultant microstructure of klly hardened steel should consist of martensite.

In practice, it is very difficult to have a completely martensitic structure by hardening

treatment. Some amount of austenite is generally present in the hardened steel. This

austenite existing along with the marteniste is referred to as retained austenite.

Conventional metallography does not reveal the presence of retained austenite till it

exceeds 20 percent. The presence of retained austenite greatly reduces mechanical

properties and such steels do not develop maximum hardness even after cooling at rates

higher than the critical cooling rate. The amount of retained austenite depends largely on

the chemical composition of steel. For plain carbon steels, the amount of retained

austenite increases with the rise in carbon contents (Suryanarayana, A.V.K., 1979). The

problem of retained austenite is more complex in alloy steels. Most of the alloying

elements increase the content of retained austenite.

Hardness can be improved in hardened steels containing retained austenite by a

process known as sub-zero treatment or cold treatment. Retained austenite is converted

into martensite by this treatment. This conversion of retained austenite into martensite

results in increased hardness, wear resistance and dimensional stability of steel. When

hardened steel is left with retained austenite, it is likely to get transformed into martensite

during subsequent thermal cycling. Such transformation leads to unpredictable

dimensional changes. Hence, for dimensional stability, it is necessary to eliminate

retained austenite as far as possible. This can be done through sub-zero treatment. The

process consists of cooling steel to sub-zero temperature, which should be lower than the

Mf temperature of the steel. Mf temperature for most steels lies between -30°C and -

70°C. During the process, considerable amount of internal stresses are developed in the

steel, and hence tempering is done immediately after the sub-zero treatment. This

treatment also helps to temper martensite, which is formed by decomposition of retained

austenite during sub-zero treatment. This treatment has been found to be less effective if

performed on steels, which have been kept at room temperature for several hours or

mildly heated, say up to 100 or 150°C, after the hardening treatment. The reason for this

is that the stability of austenite is increased considerably by these two processes. The

austenite so stabilized does not transform to marteniste completely. Therefore, sub-zero

treatment must be performed just after the hardening treatment. Mechanical refrigeration

units, dry ice, and some liquefied gases such as liquid nitrogen can be used for cooling

steels to sub-zero temperatures (Rajan, T.V., et al., 1998).

This treatment is employed for high carbon and high alloy steels used for making

tools, bearings, measuring gauges and components requiring high impact and fatigue

strength coupled with dimensional stability. Case hardened steels, which have a tendency

to retain austenite, are also subjected to such treatment.

1.1.2 Surface heat treatment

There are situations in which the requirement is such that the outer surface should

be hard and wear resistant and the inner core more ductile and tougher. Such a

combination of properties ensures that the component has sufficient wear resistance to

give long service life and at the same time has sufficient toughness to withstand shock

loads. Such combination can basically achieved by case hardening process (Child, H.C.,

1980).

Case hardening treatments are basically classified into two groups. The first

group is known as thennochemical treatment because the surface composition of the steel

is changed by difhsion of carbon and some times by other elements.

Examples: Carburizing-Pack carburizing, Liquid Carburizing and Gas Carburizing

Vacuum carburizing, Post-Carburizing, Cyaniding and carbonirtriding, Nitriding, Plasma

Nitriding, Salt bath nitrocarburizing, Boronizing, Chromizing and Toyota Diffusion

Process

The second one is called thermo-mechanical treatment, which involves phase

transformation, by rapid heating and cooling of the outer surface.

Examples: Flame hardening, Induction hardening Electron beam hardening and laser

hardening

Thermo-chemical surface hardening

Carburizing

Carburizing is the most widely used process for surface hardening of steels. In

this process carbon is difhsed into steel by heating above the transformation temperature

and holding the steel in contact with a carbonaceous material which may be a solid

medium, liquid or a gas. Under such condition, carbon is absorbed in solid solution in

austenite. As the solubility of carbon is more in the austenitic state than in fenitic state,

fully austenitic steel is essential for carburizing (Colegate, G.T., 1951). Carburizing can

be divided into Pack carburizing, Liquid Carburizing and Gas Carburizing. Among them

Gas Carburizing is the most widely used industrial heat treatment process.

Gas carburizing is a process in which the surface of the component is saturated

with carbon in a gaseous atmosphere containing carbon. To accomplish this, first the

components are heated in a gas tight furnace in a neutral atmosphere to a predetermined

temperature in the range of 870 to 940 OC. Then the furnace is flooded with a suitable

gas such as Propane, Butane, Kerosene etc. Finally, the components are held at this

temperature to allow diffusion of carbon into the case. After the carburizing treatment is

over the components are quenched to get the required hardness, wear resistance and

fatigue resistance on the surface, supported by a tougher core. A striking feature of gas

carburizing process is that in this process the original toughness and the ductility

remains unaffected even after the heat treatment.

Cyaniding

In cyaniding, the surface layer of steel is hardened by addition of both carbon and

nitrogen. Cyaniding is carried out in a liquid bath of NaCN. In this process, parts to be

heat treated are immersed in a liquid bath (800 - 960 OC) of NaCN with the concentration

varying between 25% and 90%. Measured amount of air is passed through the molten

bath. NaCN reacts with the oxygen in the air and is oxidized. Carbon and Nitrogen so

formed in atomic form diffuse into the steel and give thin wear resistance layer. Usually

this process requires 30-gominutes for completion. Case depth ranging from 0.5 to 2mm

can be achieved. This process is not suitable for hardening those parts, which are

subjected to shock, fatigue and impact because nitrogen addition has adverse effects on

such properties of steels (Waterfall, F.D., 1949).

Carbo-nitriding

Carbo-nitriding is carried at lower temperatures (800 - 870 OC) in a gas mixture

consisting of a carburizing gas and Ammonia. A typical gas mixture contains about 15%

NH3 5% CH4, and 80% neutral carrier gas. Carbon and nitrogen are diffused at the same

time into the surface of the steel in the austenitic ferritic conditions and gives case

thickness of the order of 0.05 to 0.75mm. After Carbo-nitriding, quenching is done in oil

to a avoid cracking.

Plasma-nitriding

Plasma nitriding is also known as an Ion nitriding process. In this process the

steel components to be heat treated is kept at 450 OC in vacuum at a negative potential of

the order of 1000 ohms with respect to chamber. Then an appropriate mixture of nitrogen

and Hydrogen is passed at a pressure of 0.22 to 8m bar. As a result, plasma formation of

these gases takes place. Korotchenko, V., and Bell, T., 1978, showed that in Plasma

Nitriding the time required for achieving the case depth of 4pm is about 20 hours at

450°C.

Boronizing

Boronizing is the one of the recent methods of surface hardening. Both pack and

gaseous techniques can be applied using Boronizing. In the case of pack Boronizing, The

Components are packed in heat resistant boxes with mixtures of granules or paste of

boron carbide with addition of activators at 900 - 1000 OC. Boron diffuses inwards and

iron borides layers are formed. On the outer surface FeB phase forms, while in the

interior Fe2B phase is formed. FeB phase is more brittle and it is not desirable. Higher

temperatures, longer treatment times and high alloy steels favours the formation of FeB

Phase. The treatment time requires for a case depth of 0.15mm is 6hours at about 900

"C. The significance of this process is the treatment temperature, which is very high and

hence hardening of components before Boronizing is not required (Graf Von

Matuschka, A,, 1975).

Chromizing

Chromizing is used for surface hardening of both carbon and tool steels. The

components to be chromized are packed with fine chromium powder and additives. A

typical Chromizing mixture consists of 60% Cr, 0.2%Ammoium Iodide and 39% Kaolin

powder. Diffusion of chromium takes place at 900-1020 OC and chromium carbide is

formed at the outer surface of the steel. The treatment time needed for achieving a case

depth of 0.02 - 0.04mm is 12 hours. Hardness of chromium carbide layer is about

1500VHN (Raghavan, V., 1984).

Toyota Diffusion process

In this surface hardening process, which produces the hardest surface, the

hardness achieved is of the order of 3000 - 4000VHN. In Toyota Diffusion process, the

component is kept in salt bath of proprietary composition based on borax. The carbide

forming elements are added in the form of Ferro alloys. The process is carried out at

about 1050°C. After salt bath treatment, the components are quenched and tempered.

This process of surface hardening is very appropriate for die and tool steels (Rajan, T.V.,

et al., 1998).

Thermo mechanical surface hardening

Flame hardening

Flame hardening is the simplest form of surface hardening heat treatment. This

process consists of heating the large work piece, such as crankshaft, axles, large gear,

cam or any other complicated cross section by an oxy-acetylene or oxy-fuel blowpipe

followed by spraying of jet of water as coolant. Hardness in flame-hardened steel is due

to martenisitic and lower bianitic structure. The carbon content required for flame

hardening steels varies from 0.3% -0.6%. High carbon steels can also be hardened by

this process, but greater care is required to avoid cracking. This process can achieve

normally, a case depth of upto 3mm. There are four different methods, which are used in

general for flame hardening: i) Stationary, ii) Progressive iii) Spinning iv) Progressive-

Spinning.

There is a little scaling, decarburization or distortion in flame hardening. Since

the heating and cooling are very fast, the core remains unaffected (Davies, D.E., 1966).

Induction hardening Process

Generally this process is used to surface harden crankshaft, camshaft, gears, crank

pins and axles. In this process, heating of the components is achieved by electromagnetic

induction. A conductor (coil) carries an alternating current of high frequency, which is

then induced in the enclosed steel part within the magnetic field of the coil. As a result,

induction heating takes place. The heat so generated affects only the outer surface of the

steel components due to skin effect. The component is heated usually for a few seconds

only, and then it is quenched immediately by the jet of cold water. Due to quenching

martensitic structure is formed which makes the outer surface hard and wear resistance.

The hardening temperature ranges from 760 to 930°C. According to the Carbon content

and alloy addition, the temperature is fixed (Lakhtin, Yu.M., 1979).

Electron beam hardening

Electron beam hardening is a short surface hardening procedure for

martenisitically hardeneable ferrous materials. So this process is used for hardening

those components, which cannot be induction, hardened because of associated distortion.

Automatic transmission clutch cams (SAE 5060 Steels) are hardened by this process.

The work piece is kept in vacuum at 0.06m bar pressure. Electron beam is focused on the

work piece to heat the surface. In the beginning, energy input is kept high. With time,

power input is reduced as the component gets heated up. This is done to avoid melting.

Normally case depth upto 1.5mm can be achieved by this method (Brooks, C.R., 1996).

Laser surface hardening

Laser surface hardening of ferrous materials is an established process used to

enhance the mechanical properties of highly stressed machine parts such as gears and

bearings. Laser beams are used since these have very high intensity, but they may melt

the work piece when they are used at such a high intensity. Therefore, a lens is used to

reduce the intensity by producing a defocused spot from 1 to 25 mm wide. Ashby and

Easterling, (1984) demonstrated that a laser beam of 1kW produces a circular spot

whose diameter may vary from 0.5- to 0.25mm. Industrial lasers upto 20kW are now

available. Case depth of about 1.3mm is obtained by self quenching. The microstructure

of laser heated steel consists of bianite and ferrite at the surface of the heated spot and

pearlite and fenite in the interior.

Salt bath hardening

Salt bath heat treating methods are uniform, quick, efficient, economical and

environmentally responsible. Salt bath are used in a wide variety of commercial heat

treating operations. Parts that are heated in molten salt bath are heated by conduction.

The molten salt bath provides a ready source of heat as required. Although materials

being heated come in contact with heat through their surfaces, the core of a part rises in

temperature at approximately the same rate as its surface. Heat is quickly drawn to the

core fiom the surface and salt bath provide heat at an equal rate over the total part. Salt

bath hardening consists of four stages namely Preheating, Austenitizing, Quenching and

Drawing (Grossmann, A,, and Bain, E.C., 1964).

1.1.3 Gas carburizing

During the last few decades, carburizing by means of gas has become the most

popular method of case hardening. The furnaces are very dependable and relatively

simple to operate. Since hydrocarbons enter into the process, several reactions take place

simultaneously. In addition to the reaction

2com C+ C02 ...l.l

there is also the methane reaction

-2%

CH4 - C+ 2H2 ... 1.2

and the water-gas reaction a

CO+H2 '- C+H20 ... 1.3

The composition of the furnace gas can be determined by analysis and hence the

carbon potential may be determined. The moisture content of the gas has a great

influence on the carbon potential as indicated by reaction (eq. 1.3) and may be measured

by simply determining the dew point of the gas. The dew point is that temperature at

which the water droplets (dew) are deposited from the gas (saturation).

Harris, F.E., (1943) described a method of calculating the carbon potential from

the composition of the gas. There are a number of methods in existence for the

continuous measurement and control of the carbon potential. One such method is based

on infrared rays, which are used to register the concentration of C02 in the gas. This

value gives a measure of the CO concentration, which, in turn, relates to the carbon

potential. By means of a special measuring unit, the resulting impulse can be converted

direct to a numerical value of the carbon potential. The control may also be carried out

my measuring the resistance of a steel wire, which is calibrated for different carbon

concentrations. The resistance measured is proportional to the carbon concentration.

Indirect measurement of the carbon content can take place by means of an oxygen probe.

When gas carburizing is being employed, a primary requirement is the availability

of furnaces sufficiently gas tight to prevent air ingress, since this would interfere with the

process.

In principle, there are four different ways of producing the gas, viz.:

Dripping carburizing liquid into the furnace. This liquid generally contains ethyl

alcohol and white spirit or only isopropyl alcohol. It drops on to a plate inside the

hrnace where it is dispersed and vaporized. By this method, it is possible to

employ a liquid that produces a carrier gas having a relatively low carbon

potential. This potential can then be adjusted by dripping a more active liquid

into the furnace. In practice, however, only one liquid is used and the carbon

potential is adjusted by varying the volume of liquid added.

Admission of gas from a gas-producing unit. The gas is produced by endothermic

combustion. The raw gas, usually propane, is mixed with air in carefully

balanced proportions and burnt in the gas-producing unit to produce a carrier gas

with a carbon potential of 0.35% - 0.50%. Before it enters the furnace, the carrier

gas is given an extra addition of propane, thereby adjusting the carbon potential to

the required level, which is usually around 0.8%.

Direct admission of gas and air into the furnace. The gas actually doing the

carburization is produced in the furnace chamber by the reaction between the gas

and air, which are admitted in balanced amounts.

Carburizing in a vacuum furnace. Only hydrocarbons are admitted into the

furnace fi-om which air is evacuated. Since no oxygen takes part into the reaction,

there is no internal oxidation. The process can take place at a temperature higher

than that normally used, thus allowing the carburization to proceed more quickly

(Western, H.W., 1972 and Grosch, et al., 1981).

Depth of hardnesspenetration (Case depth)

The depth of hardness penetration depends on the carbon content of the

carburized layer. Provided only martensite is formed in the case on quenching, the depth

of case hardening would be equivalent to a depth of carbon penetration down to 0.40% C.

This would agree well for small parts but as the section dimensions increase, the rate of

cooling decreases and hence the conditions necessary for the formation of martensite are

changed. This implies that steel components of different sizes but which have the same

depth of carbon penetration obtain a casehardening depth that is dependent on the

dimensions of the piece (Beurnelburg, W., 1964).

Surface hardness

The carbon dissolved in the austenite is the decisive factor determining surface

hardness. When the carbon concentration at the surface of conventional alloy case-

hardening steels exceeds 0.70%, the M, temperature falls steeply, the amount of retained

austenite after quenching increases and the hardness decreases. If the surface layer of the

steel contains the appropriate concentration of carbon for maximum hardness, the

quenching temperature is of minor importance to the hardness provided that the grain size

is not altered. By varying the quenching temperature of an 'over-carburized' steel it is

possible to control the amount of carbon going into solution and hence the amount of

retained austenite which, in turn, affects the hardness (Robinson, G.H., 1957).

Fatigue strength of case-hardened steel

By and large, the increase in fatigue strength brought about by case hardening is

proportional to the hardness increase in the surface zone. It is also dependent on the

compressive stresses in this zone. The maximum fatigue strength value is obtained when

the carbon is about 0.6%. Most fatigue strength investigations of case-hardening steels

show that optimum properties are obtained when the surface carbon content is 0.6-0.8%

and the surface hardness at least 700HV (Diesburg, D.E., et a1.,1978).

ASM (Metals Handbook: Heat Treating, 1981) suggested the following general

heat-treatment recommendations for casehardening steels. The treatment is designed to

give a surface hardness of atleast 60 HRC or 7 10 HV after quenching.

Carburizing in solid compound (pack carburizing) - Temperature: 900-925OC

When direct quenching is employed, a carburizing temperature of 900°C is

recommended. If a casehardening depth of more than 1.25 mm is aimed at, it might be a

good practice to perform the traditional practice of a double quench, i.e. a first quench

from about 880°C and then a final quench. During the first heating for quenching, the

carbon concentration is reduced by diffusion. In addition, it breaks up and dissolves the

carbide network that usually forms when pack carburizing is used to give a deep carbon

penetration.

Depth of Case hardening in mm < 0.50 0.50 - 1.25 < 1.25

Method of hardening Direct Quench Single quench from 800 to 820°C Double quench

Carburidcing in a salt bath (liquid bath carburizing)-Temperature: 850-900°C

Depth of Case hardening in mrn Method of hardening I Direct - Quench I

For direct quenching, a carburizing temperature of 8 50-870°C is recommended; for

0.50 - 1.00

double quenching, 900°C may be used.

Requench fAm 800 to 820°C (Double quench)

Carburizing in gas (Gas carb urizing)-Temperature: 900-940°C

The carburization is carried out to the required depth of case hardening and to a

surface carbon concentration of 0.70-0.80%. When the process is completed, the

temperature in the h a c e is lowered to 83Q°C and the part is quenched in a suitable

medium.

Case-hardened steels are tempered at temperatures generally around 1 60-220°C.

Temperatures below 160°C should not be used, particularly if a grinding operation is to

follow, since grinding cracks develop very easily. The hardness falls quite rapidly when

the steel is tempered between 160°C and 200°C. If a hardness of 60HRC is required, the

tempering temperature should not be higher than 1 80°C (TheIning, 1984).

1.1.4 Induction hardening

When electric current passes through a conductor, a magnetic field is created

round it. If the conductor consists of a coil, a magnetic field is established inside the coil.

This field persists even if a metal ba is inserted into the coil, as shown in Figure 1.2. If

the magnetic flux is created by a high-frequency alternating current, it gives rise to eddy

currents in the surface of the metal bar, which consequently becomes heated (Figurel.3).

In iron, hysteresis losses also contribute to some extent to the temperature rise upto to the

Curie point (768'C), above which iron is non-magnetic (Semiatin S.L., and Stutz, D .E.,

1986)

The depth of penetration of the heat is governed mainly by the power and

frequency employed. The normal power input is 0.1 - 2kw/cm2 of the heated surface.

The relationship between depth of penetration and frequency can be calculated

approximately by using the following simplified expressions, which are valid for the

temperature rise in steel upto the hardening temperature:

= 20/ f cold state (20°C)

dh= 5001 fhot state (800°C)

Where, d, and dh = depth of penetration in mm, suffices, c and h represents the

cold and hot state conditions.

f- frequency in cycles per sec (Hz)

Owing to heat conduction in the material during heating, the overall depth of

penetration is larger. It is possible to calculate the additional penetration due to heat

conduction from the following expression.

dl = 0.2 t

where dl = depth of penetration in mm

t = time in seconds

Figure 1.2 Path of magnetic flux through a metal bar inserted in a coil

through which an electric current is flowing

1 0

Figure 1.3 Path of the electrical current in the coil and metal bar during

Induction hardening

The total depth of penetration is obviously c& ordh+ dl, It should be stressed that

these expressions give only a rough estimate of the depth of penetration and they have

been included here only to show the fbndamental effects of frequency and time

(Osborn, H.B., 1955).

Working coils and fixtures

The coils, also known as inductors, also got heated by the electric current and by

thermal radiation fiom the heated steel. The coils are often made from copper tubing and

are cooled during operation by internal water flow. Figure 1.4 shows three basic types of

working coils designed for external and internal heating, and for heating of a flat surface.

The internal diameter of the copper tube is matched to suit generator power

output. For the power output normally required, around 50kW, tubing with an internal

diameter of around 5mm is used. The coils are usually wound with a distance of 2-5mm

between turns and with roughly the same spacing between the coil and the workpiece.

By varying this distance, it is possible to influence the rate of heating to a very large

extent.

Parts that are to be hardened must be held in a fixture of some sort so that they are

correctly located in the inductor. Small components must be f m l y held so that they are

not displaced by the magnetic filed. A metallic fixture that comes near the coil may

become heated by the induced current. To avoid this fixture and centre washers are

normally made from an insulating and heat-resistant material such as 'sindango' or a

heat-resistant plastic (KatsuyuM Matsui, et al., 1998).

External Internal Surface

Figure 1.4 Types of coil used for Induction hardening

The influence of various factors on hardness and depth of hardening

The carbon content has a decisive influence on hardness after hardening. As the

carbon content increases, upto to about 0.80% the hardenability also increases,

particularly when in combination with other alloying elements. Both the hardening

temperature and holding time are decisive factors that govern the hardness and depth of

hardening. For induction hardening with conditions otherwise remaining, the same the

depth of hardening i s governed by the fkquency of the inductive current. Since the

hardening temperature, heating time and holding time can be controlled relatively easily

dwing induction hardening, it is possible with this method to determine in advance the

required hardness depth with a relatively high degree of certainty by using calculated and

empirical values (Rudnev.V., and Demidovich, V., 1997).

The hardening temperatures employed for induction hardening are normally some

50°C higher than those used for conventional hardening. The holding times are

extremely short. In many cases, there is no holding time, i.e. cooling starts directly when

the desired temperature has been reached. Temperatures are normally controlled by

means of optical pyrometers (Siedel, W., and Netz, W., 1982).

The pressure of the cooling medium and the angle of incidence of the jet as well

as its width are o f paramount importance to the intensity o f cooling. Hence, these factors

influence the depth of hardening and the hardening itself. Kegel and Pennekamp (1974)

arrived at the following recommendations based on experiments and theoretical

calculations of the coefficient of heat transfer through the vapour film:

Pressure 1-3 bas

Angle o f incidence 20-30'

Width of slit 0.8 -1.2 mm, preferably 0.9 -1.0 mm

Usually the spray nozzle contains a number of holes, the diameters of which are

taken as equivalent to the slit width of the above recommendations.

1.1.5 Material Testing and Inspection

Testing of material is one o f the important and essential steps for judging

suitability for engineering applications after heat treatment. Most of the properties of

interest are mechanical properties since heat treatment essentially alters mechanical

properties. This section i s confined to the testing o f mechanical properties; non-

destructive methods of testing materials have also been included, as they constitute an

important aspect of testing heat treated components (Testing of Materials ASTM, 1970:

Raval, B.M., 1995). Many tests and testing methods are in use now-a-days.

Vickers Hardness Test

Vickers hardness test is also referred to as Vickers Diamond Pyramid test. The

indenter used in this test is a square base diamond pyramid. The included angle between

opposite faces of the pyramid is 136". The loads generally employed in this test vary

fiom ikg to 120kg, depending on the hardness of the material under test. Therefore, at

varying loads, the same indenter can be used for the measurement of hardness of a

number of metals and alloys. Due to the accuracy associated with this test, it is generally

used for research and high precision work.

Microhardness testing can be defined as indentation hardness testing it involves

forcing a diamond indenter of specific geometry into the surface o f the test material at

loads ranging from 1 to 1000kgf (Andrew, et al., 1995).

In Vickers microhardness testing, the hardness value is determined by measuring

the size of the resulting indentation by using a microscope and established formula.

The microhardness test uses preciously shaped diamond indenter and various load

ranges to determine the hardness variety of the material. The specific application for

microhardness includes.

1 Measuring hardness of precision work-pieces that are too small to be

measured by conventional macroscopic hardness testing methods.

2 Monitoring of carburizing or nitriding operations, this i s usually accomplished

by hardness surveys taken on cross sections of test pieces that accompanied

the work-pieces through production operations.

3 Measuring hardness of individual micro constituents and hardness close to

edges, thus detecting undesirable surface conditions such as grinding bum and

decarburization.

Figure 1.5 Diamond pyramid indenter used for the Vickers Hardness

test and resulting indentation in the work- piece

The Vickers indenter is a highly polished, pointed, square-b ased pyramidal

diamond with face angles of 136' C (Figure 1.5). With the Vickers indenter, the depth of

indentation is about one seventh of the diagonal length. The Vickers hardness number

(VHN) is the ratio of the load applied to the indenter to the surface area of the

indentation;

VHN=2Psin ( 0 / 2 ) l ~ *

Where,

P - Applied load, (kg9

D - Mean diagonal o f the indentation, (mm) and

0 - Angle between opposite faces of the diamond, 136'

Rockwell hardness test

This test differs from both the Brine11 and Vickers tests in the sense that here the

depth of penetration, and not the surface area, is used as the parameter for arriving at the

hardness value. It works on the principle that the depth of penetration varies with the

hardness o f material. The higher the hardness, smaller will be depth of penetration and

vice versa. In this depth of penetration is not measured. Instead of that, the hardness

values can be read directly on a dial attached to the tester. Accuracy that can be

achieved by this test is not comparable to either Brine11 or Vickers hardness test.

However, the test is very popular in day-today industrial practice. There are two basic

reasons for this: Firstly, it is a fast process and, secondly, very small indentation is made

on the surface. Therefore, even finished parts can be subjected to this test.

Non-destructive Testing methods

A number of situations are encountered in day-to-day industrial practice where it

becomes necessary to study the flaws in metallic components without destroying them.

Under such circumstances, recourse is taken to non-destructive testing (NDT) technique.

This technique makes the use of the component more reliable, safe and economical,

Non-destructive testing has two important roles to play: (i) From these investigations, the

presence o f defects is established; (ii) it is also useful in estimating the nature and extent

of defects and amving at a conclusion about accepting or rejecting the product for service

(Dieter, G.E., 1986). Non-destructive tests also help in measuring the defect size and

distribution, which plays a crucial role in taking the final decision on accepting or

rejecting the component. A number of non-destructive tests are available now a days to

detect the flaws in metallic components. Some of these tests are Magnetic Particle

Inspection, Liquid Penetration Inspection, Radiographic Inspection, Ultrasonic Inspection

and Eddy Current Inspection. Magnetic Particle inspection and Liquid Penetration

Inspection are used to detect only surface defects. Radiographic and Ultrasonic

Inspection Techniques are useful for studying both the surface and internal defects. Eddy

current Inspection is not as commonly used as the other Non-Destructive Testes.

Inspection

Inspection can be defined in several ways. Simply, it can be defined as a tool to

judge the quality of a product. According to Lansburgh, R.H., and Spriegal, W.R.,

(1940), cGInspection is the process of measuring the quality of a product or service in

terns of established standards". Kimball, D.S., (1933) defined the term inspection by

stating, "Inspection is the art of comparing materials, products or performances with

established standards". Alford, L.P., and Beatty, H.R., (1951), defined it as the art of

applying tests, preferably by the aid of measuring appliances to observe whether a given

item of a product is within the specified limits of variability.

Inspection is an important step in any production unit as it decides whether the

material or the product under consideration is suitable for use. It separates accepted fiom

rejected items. Therefore, inspection checks entry of defective items from further

processing, resulting in the saving of material, time and money. Also, it protects he

interests of the producer and customer: It helps the producer by improving reputation and

competitive strength, and assists the customer by ensunhg supply of quality items. It

should be clearly understood that inspection does not include any corrective action.

Inspection in Heat treatment

Inspection in heat treatment can be divided into three classes, namely, preheat

treatment inspection, post-heat treatment inspection, and inspection during heat

treatment. Various characteristics to be inspected in each class are as follows:

(i) Pre-heat Treatment inspection

Size

Chemical composition

Macro and micro structure

Q Austeniticgrainsize

Hardenability

Tensile strength, percentage elongation and reduction in area

* Non-metallic inclusions

Surface flaws

(ii) Inspection during heat treatment

* Heat treatment temperature

Q Heat treatment time (heating time, soaking time and sometimes

cooling time)

r Properties of gas in case of controlled atmosphere process

Properties of solid, liquid or gaseous materials used in case of

hardening

(Cafburizing, nitriding and Carbonitriding) processes. Properties of

quenchants, if in use.

Post-heat treatment Inspection (iii)

The characteristics to be inspected are almost the ~ame as the ones

included in pre-heat treatment inspection.

Residual stresses in materials

Introduction

All the heat treatment processes introduce residual stresses due to non-uniform

changes induced by the process within the body. These stresses influence the mechanical

properties like fatigue strength, depending on their nature, magnitude and distribution

across the body. There is basically no material or situation free of these stresses. Hence,

the general interest today is the recognition and measurement of these residual stresses.

The X-ray diffraction method is used to determine the residual strains present in

the surface of the material. This method is non-destructive in nature. This is based on

the measurement of changes in lattice spacing. Such relative changes of the distance

between lattice planes are called lattice strains. In the X-ray method, an X-ray beam,

which leaves the material unscathed, is irradiated on the specimen's surface and the

lattice strain is calculated from the difiiaction peak shift.

Principle of X-ray diffraction method

X-ray diffraction method employs Bragg's law to estimate residual strains present

in the atomic planes. In this method, a monochromatic X-ray beam of sufficient intensity

is made incident on the atomic planes. The reflected beam from successive planes of

atoms is observed Bragg's law defines the condition for diffraction through the following

equation.

A =2d sin 0 . . (1.5)

Where,

A - Wavelength of incident X - Rays

8 - Angle between incident or reflected beam and reflecting planes

d - Interplanar spacing and

n - Orderofreflection(n=l,2,3 ...)

The Equation (1.5) shows that, if the wavelength of X-rays i s known, the

interplanar spacing'd' can be determined by measuring the angle 9 . In presence of

residual stresses, the d-spacing changes and this in turn results in shift in X-ray

diffraction peaks. Therefore, this shift in diffraction peaks is a measure of residual

stresses. Figure 1.6 Shows the configuration generally followed for residual stresses

measurements PI, P2 and P3 refer to three orthogonal directions relative to the sample

under investigation and 4, L2 and L3 describe the laboratory or measurement same of

reference.

The angles yl and Q define the relationship between Pi and Li axes describes the

angle between the specimen surface normal (P3) and the direction of strain being

measured (L3); 6 denotes the angle between one of the principal stress axes (PI) and the

projection of the measured strain direction (L3) on to the specimen surface. In the widely

used "sin2 y Method, diffraction measurements are made at several tilt angles y . For

the general case of two measurements at y =0° and y =(II i (Figure 1.7), Noyan and

Cohen, (1986) have given the following equation for the surface residual stress a + .

Figure 1.6 Axial system for residual Figure 1.7 sampling of different grains in stress measurements the specimen by incident

X-ray beam at (a) y =0° and (b) =W i

1 + v sin" d # ~ * Where,

d - Interplanar spacing in the described by the angles 4 and y

d 4 = 0 - Stress free interplanar spacing value

E -- Young's modulus of the material and

v -- Poisson's ratio.

The tern E / (sin2 y ) ( I+ v ) is constant and is defined as K. Using the linear

relationship given in Equationl.6 for surface residual stress, the lattice strain A d/d

plotted against sin2 y would produce a straight line whose gradient is a function of a

#, v and E.

Then, m* = 6 ~ 4 ~ / 6 sin2 y

a 4 = m*/{(l+v)/E}

Where,

m* - Gradient of a least squares straight line fit through the data points and

&q5y=Adld

Depending on the incident beam energy and the material under study, the

information on surface residual stresses with in a depth of 10 - 30 pm is possible. For

example, in the case of Cr, Ka , and X-rays the depth o f penetration in steel is of the

order of 15 to 20 pm.

1.2 Literature Review

The growing demand for materials having low specific weight, high strength,

stifhess, toughness, wear resistance and ability to operate at elevated temperatures has

led to the development o f newer materials. These newer high performance materials are

used in various engineering applications like automobiles, aeroplanes, space vehicles etc.

Modern production methods are available to process these materials more efficiently.

However, most of the engineering properties of metals and alloys are related to their

metallurgical structure.

Mechanical properties of the metals can be changed by varying the relative

proportions of micro constituents. In practice, change in micro constituents is achieved

by subjecting the material to heat treatment process. T h i s process consists of heating a

metal or alloy to a specific predetermined temperature, holding at this temperature for

required time, and finally cooling from this temperature. All these operations are carried

out in solid state. Sometimes, it becomes necessary to repeat these operations to impart

some characteristics. Therefore, heat treatment may be defmed as heating md cooling

operation(s) upplied to metals and alloys in solid state so as to obtain the desired

properties.

Steel is the most important metallic material used in Industry. This is because of

the versatility of its engineering properties suitable for different working conditions or

applications. In one application it can be more ductile and suitable for any forming

operation. In another application the same steel can be very hard and wear resistant. In

yet another application, it may be having harder outer skin and tougher inner core. This

versatility is possible because of its amenability as far as its mechanical and physical

properties are concerned through different heat treatment processes that the steel can be

subjected to (Thelning, 1984).

Heat treatment is an integral part of the manufacturing cycles of ferrous and Non-

ferrous products. It is a highly energy intensive process having strong bearing on the

final product quality and it influences the productivity of a plant. The Primary objectives

of Industrial heating and heat treatment processes are to enhance the formability of

materials for shaping operations or to provide desired mechanical properties to the end

product. Enhancement of formability is achieved by heating the components to higher

temperature where material can be easily shaped. Whereas, the tailoring o f mechanical

properties involves phase transformations, microstructural changes or surface

modifications. Since, these heat treatment operations impart desired mechanical

properties to materials they have a strong influence on the final product quality

(Malhotra and Satyam Sahay, 2002).

Satyam Sahay and Arun Kumar (2002) reported that the high temperature

requirement and the sluggish kinetics o f solid state transformation render heat treatment

operations highly energy intensive and they become productivity bottlenecks. Due to

their effect on the key plant performance and quality parameters, it is essential to

optimize heat treatment operation for maximum efficiency and minimum cost.

Suitable thermal or mechanical treatments will produce extensive rearrangement

of atoms in metals and alloys and corresponding marked variations in Physical, Chemical

and Mechanical Properties.

Other processes where phase transformations occur are casting, welding and

machining etc. Phase transformation can be classified in many different ways.

Christiane (1975) classifies them into homogenous and heterogeneous transformations.

The former involves rearrangement in the structure of the materials taking place

simultaneously in all parts of the solid, while the latter involves structural changes which

are more localized. Alternatively, they could be classified into isothermal and non-

isothermal transformations.

Rajan, et al., (1994) described heat txeatrnent of metals as an important operation

in the final fabrication process of many engineering components. During heat treatment,

metals can be subjected to thermal cycling. The effects of thermal cycling differ from

one metal to another. Heat treatment could result in,

i) Improved Ductility

ii) Decreased internal stresses

iii) Refined grain size

iv) Increased hardness or Changes in Chemical composition of metal surface as in the

case hardening.

Other beneficial effects of heat treatment include improvement in the

machinability, alteration in the magnetic properties, modification of electrical

conductivity, improvement in toughness and development o f recrystallised structure in

cold worked metal.

Avner (1982), described that there are number of factors of paramount

importance which are to be considered when heat treating a metal or alloy. Some of them

are the temperature upto which the metal i s heated, the length of time that the metal is

held at the elevated temperature, rate of cooling, and the atmosphere surrounding the

metal when it is heated. Heat treatment temperature, holding time, rate of heating and

cooling are some of the parameters which affect the heat treatment processes and are

commonly referred to as heat treatment process variables. The required magnitudes of

these variables depend on the chemical composition, size and shape of the object and the

final properties desired in the metal.

Numerous industrial applications require a hard wear resistance surface called the

case and a relatively soft and tough inside called the core. Five principal methods of case

hardening exist. They are Carburizing, Nitriding, Cyaniding or Carbonitriding, Flame

hardening and Induction hardening.

Among the more important of these treatments are heat treatment processes such

as Induction hardening and Case carburizing which rely on phase transformation to

produce desired changes in mechanical properties (Wang, et al., 1993).

Induction surface hardened low alloyed medium carbon steels are widely used

for critical automotive and machining applications such as propulsion shafts, crankshafts

and steering knuckles, which require high fatigue resistance. Kristoffersen and

Vomacaka (2001) studied the influence of process variables on Induction hardening and

concluded that fatigue behaviour of induction hardened components depends to a great

extent on the correct combination of hardening depth and the magnitude and distribution

of residual compressive stresses in the surface layer.

In order to reduce the manufacturing costs interns of material consumption and

elimination of the number of processing steps, the effect o f induction hardening

parameters on the fatigue behaviour should also be considered at the design stage itself.

In present competitive industrial scenario, there is a tremendous need for saving

time, energy and cost without sacrificing the quality of the product processed. That

underlines the importance of optimizing the induction hardening process parameters to

have designed properties of materials for various applications.

Gas carburizing is the oldest and one of the cheapest methods of case hardening in

which the surface of the component is saturated with carbon in a gaseous atmosphere

containing carbon.

Anand M Deshpande et a1.,(2003) analyzed the optimization of carburization

profile for minimizing the process cost and reported that Gas carburizing is a complex

process in itself as a number of variables affect the success of the process and quality of

the components. An industrial survey indicates that there is a rejection of 10-12% of case

hardened components due to various defects like crack formation, over hardening, change

in size and shape etc, and the extent of rejection can be kept at a minimum by closely

monitoring the process and installing proper quality control measures. In the case of

surface treatment processes this is possible by controlling the process variables. If the

process variables are optimized fkom the point of view of the obtainable material

characteristics, it will be a good measure. Thus, determination ofprocess variables lies

in the proper selection and introduction of suitable limit design concept at the earliest

stage of the process and product development cycle that will result in the quality and

improved productivity.

As reported by Woodward, et al., (1999) Quenching is an integral part of

industrial heat treatment for steels and it provides a means by which the mechanical

properties of a steel part can be controlled. During quenching, the steel is typically

cooled from above the austenizing temperature by liquid spray (Induction hardening) or

immersion in a liquid bath (Gas carburizing).

Grossmann & Bain (1964) described that depending on the temperature field

development within the material, different types of phase transformations occur, which

result in a variety of microstructures, residual deformations and hardnesses and residual

stress distribution. This involves phase transformation by rapid heating and cooling of

the outer surface.

The demerits involved in induction hardening are as follows:

*3 Dimensional changes and localized volume changes which may require

reworking.

*:* Quench crack.

9 Excessive grain growth in the region just below the hardened surface produces

stresses induced by phase transfonnation and thennal stresses. These stresses

cause shape and size distortion in the components.

The distortions in the components are troublesome to the manufacturers. If the

distortion is controlled within the design tolerance limit, the post hardening processes and

rejections can be eliminated. This will result in cost saving and increase in productivity.

Modeling of induction hardening to predict the volume fraction of martensite in

the treated component is a complex issue which involves numerically solving the

Maxwell's equations for electromagnetic field around component, coupled with the

modeling of the thermal process for the component using heat transfer principles. The

system is inevitably highly nonlinear because of temperature dependent material

properties of the component (Nemkov, V.S., and Goldstein, R., 1999). In the past, more

attention of numerical simulation had been paid to modeling, taking into account only

certain aspects of the induction hardening problem. Little effort has been taken to model

the integrated process from heating to final hardening. Numerical simulation of the

magnetic field provided valuable analysis for the magnetic vector potential which was a

very important parameter for eddy current and Joule heat determination. Simulation of

induction heating had been carried out and provided models for solving coupled

electromagnetic/thermal analysis.

Koistinen and Marburger law (1959) and Avrami (1936 and 1940)

developed mathematical models for quenching and phase transformation processes and

determined the volume fraction of micro-constituents formed in the quench cooling

process. However, few studies have been found in predicting the hardness distribution

using either mathematical modeling or simulation.

Jiankun Yuan et al., (2003) has developed an integrated FEA based modeling

system with the capacity to simulate the electromagnetic field induced heating process,

austenite state holding processes and severe quenching processes. It is also possible,

from the above model, to find out the volume fraction of micro constituents such as

martensite formed in the quenching process, and final hardness distribution pattern in

workpiece. However, this model cannot be used in obtaining continuous cooling curves

with temperature time transformation diagrams in induction hardening applications.

Hence, in this thesis as an extension of Jiankun Yuan et aL, model; an equation has been

developed to predict the volume fraction of martensite. A Regression model has also been

developed to estimate the Hardness in HR4. Experimental validations have been done toprove

the suitabiliq of these models in predicting the characteristics of the heat treatmentprocess.

Metallic materials must have different mechanical properties in accordance with

the area of application. Surface hardness is the most important mechanical property for

working conditions like frictional and twist moment transfer. The working conditions

determine the grade of hardness and the depth of hardness required on the work

materials. Most important characteristics for the material hardening are hardness value,

which depends on chemical composition and variation of hardness, which depends on

hardening conditions (Bodart and Touzani, 1998 and 2001; Kayacan, 1991).

Wear resistivity of material, fatigue life, impact resistivity and compressive

strength, and resistance of twist force are increased by the increasing surface hardness

value. Martensitic surface hardening method is an important method for increasing these

properties (Kayacan, 1991).

Martensitic surface hardening method can be classified as Induction hardening,

flame hardening, electron beam and laser beam hardening. Upto now dominant variables

in the martensitic surface hardening methods have been identified by various studies

(Chen and Shen, 1999; Sphepeyakovskii, 1987; Shary and Osborn, 1974; Iozinskii,

1969).

Hardness depth control is very complex in the Induction hardening process, for

many variables affect hardness depth obtainable. Because of the non linearties involved

in Induction Surface Hardening process, it is very difficult to obtain the hardness depth to

the exact values (Mehmet Cengiz Kayacan and Oguz Colak 2003).

In spite of the extensive studies that have been carried out so far, controlling the

correct depth of hardness in induction hardening process remains difficult. Several

methods are being applied to obtain the correct value of required hardness thickness.

Relationship among the influential variables can be obtained in optimum value using

intelligent techniques.

Unlike induction hardening process, which is used for treating medium carbon

steels, Gas carburizing is applied to low carbon steels. The process of Gas carburizing

and post carburizing of steel constitutes three separate steps as given below:

0 Austenitizing (over a temperature of 900°C).

* Quenching in oil to obtain the martensitic structure of steel.

Tempering for restoring the good mechanical characteristics to the work piece.

The final quality of a material depends on the correct combination of case depth

and level of hardness without any thermal damage. Mitra (2004), demonstrated a cost

model based optimization of carburizing operation and used Furnace Temperature,

Carbon potential, Quenching time, tempering temperature, Preheating and tempering time

as the influential variables.

Fabrizio Dughiero and Mara Battistetti (1997) did the optimization procedures

in the Design of continuous Induction hardening and tempering installation for magnetic

steel bars.

Power Potential, Scan speed, Quench flow rate, and Frequency are the critical

variables which affect the Hardness, Case-depth and desirable mechanical and

metallurgical properties of the material of the component. At this point, it has to be noted

that there are two important questions in the Gas carburising and Induction hardening of

steels.

What are the microstructural, hardness, deformation and case depth produced

under a given set of operating conditions? and

What are the operating conditions required to obtain a specified micro structure,

maximum hardness and maximum case depth?

The answers to these questions are closely related, which necessitate the study of

micro structures, microhardness and macro hardness and case depth.

In the present Research work, the Rack and pinion assembly used in the Power

steering of automobiles has been taken up as samples for experimentation. Rack is a

medium carbon alloy steel. Selective surface hardening is done only on the teeth portion

by Induction hardening process. Pinion is a low carbon alloy steel and it is case hardened

by Gas carburising process.

The dimensional changes in the components which have been either induction

hardened or Gas Carburized proved costly and troublesome to the manufactures as they

adversely affect the performance of the mentioned components in terms of life, trouble

free operation and noise of operation. As reported by Dong-hui, Xu and Zhen -Bang

Kuang (1996), there exists a definite link between distortion and the initiation of fatigue

failure.

In Induction hardening, the term 'distortion' usually describes the dimensional

changes brought about by releasing of internal stresses. This occurs in a component after

heat treatment. Several investigators have found it convenient to divide the total

distortion into two classes of dimensional change. The first of these is usually called

'volume change' or 'inherent distortion' and is said to be the result of the dilation due to

transformation. The second type of dimensional change is usually called warpage or

change in shape and is said to be associated with the thermal stresses produced by non-

uniformity of heating or cooling.

The main problems in the Gas carburized components (pinion) are shape and size

distortion. Shape distortion can be reduced by proper stress relieving between machining

as they are mostly due to residual stresses, whereas, size distortion is due to structural

transformations in steel. When austenite transforms into martensite there is an expansion

in volume. While tempering, there is a contraction due to formation of carbides. The

presence of retained austenite and its change during tempering introduces complex

overall changes in size. (Shen - Chih Lee and Weio-Youe Ho, 1989).

Many investigations indicate that the following are some of the reasons for

distortion.

i) Rapid heating.

ii) Methods of stacking or fixcuring of parts while heatreating.

iii) Increase in grain growth with increase in case depth.

iv) Severity of quenching.

Only a few isolated attempts have so far been reported on experimental analysis

of the thermal distortion developed in the components like Rack and Pinion. At this

juncture, it appears that the practical optimization and strategy and reliable

mathematical models are still needed to analyze and control thermal distortion, which

happens in Gas carburising and Induction hardeningprocesses.

The present work addresses the gap in the studies that exist in obtaining the

quality heat treated components.

Cause and effect analysis for Induction hardening and Gas carburising have been

carried out and Shewerts' diagram has been constructed to identify the influential process

variables that may affect the quality of the components.

Taguchi's mixed level series design of experiments approach (Gas Carburising)

and 3 -Level Factorial Design of experiment approach (Induction hardening) have been

used to accomplish the objective. The statistical tools such as Response Graph approach,

S/N ratio approach and Regression analysis used in the analysis of Control of distortion.

Residual stress has considerable effect on the properties of the materials such as

stress corrosion resistance and fatigue strength of steel after surface hardening (Liu et al.,

1988). Prediction of residual stress has become significant in optimizing heat treatment

processes. Components which are shaped by machining and other mechanical operations

do contain residual stresses. It is well known that the presence of residual stresses in a

component can lead to various problems such as distortion, cracking, reduction in fatigue

life, and increased susceptibility to stress corrosion. In processing operations such as

machining and cold working, the forces acting on the work piece cause non-uniform

plastic deformation leading to the formation of residual stresses. Processes such as

welding, flame heating and heat treatment involve non uniform heating and cooling.

They can lead to residual stresses on account of thermal expansion and contraction and

additionally in some cases, on account of the volume changes arising from

microstructural transformations (Horger, 1950).

There is very little information available in the literature pertaining to the analysis

of residual stresses during heat treatment processes. Weiner and Huddleston (1959)

developed equations for the computation of stress states in solid and hollow cylinders

subjected to transient temperature distribution. The equations, however, involved

constants of integration and the axial stress rate, the determination of which in general,

seemed difficult. A solution was provided for a particular formal problem, namely, a

solid cylinder in which the stresses were due solely to a phase transformation occurring at

a fixed temperature.

The analytical approach by Tanaka (1975) for residual stress in quenched

cylinders was based on a finite element programme and the stress distribution was

obtained by assuming that Young's Modulus and Poison's ratio were independent of

temperature. The results of his experiment showed good agreement with the predictions

by which the theoretical compressive residual stresses were indicated near the surface.

Rathbun and Coffin (1975) presented some experimental evidences in which

compressive stresses of the order of 2 6 0 ~ ~ l m ~ occurred on the surface of a stainless

steel specimens of the shape of a flat circular disc.

Gas carburising and induction hardening can generate compressive residual stress

in the surface layer of a component. Having compressive residual stresses is an effective

method for extending the fatigue life of the work piece (Dong -Hui Xu and Zhen-Bang

Kuang, 1996). The improvement in fatigue behaviour depends on the magnitude and

distribution of the residual compressive stress in the hardened layer. The attainment of

the correct combination of hardening thickness and the residual stress requires the

proper control of hardened thickness and initial yield stress. X-ray diffkaction

technique, which is using Bragg's Law, has been used to estimate the residual stresses in

the present study. To explore the relation between process variables and the residual

stress developed, phase transformation study has been carried out in SEM and residual

stresses have been measured using residual stress analyzer.

The main aim of the investigation has been the prediction of desired operating

conditions to avoid thermal and metallurgical defects in Gas Carburising and Induction

hardening. The process variables are the limiting factors. The above said problem

encouraged to investigate the following to obtain distortion free engineering

components, like Rack and pinion, afer they are subjected to surface hardening

process.

i) Gas Carburising process variables optimization using Taguchi's Mixed Level ,

series of Experiment.

ii) Induction hardening process variables optimization using 3-Level Full

Factorial Design of Experiment.

iii) Phase transformation study in Induction hardening using mathematical

modeling and validation through experiments.

iv) Thermal and metallurgical effects associated with Case Hardened

components.

v) The distribution and pattern of Residual stresses in Case hardened

components.

1.3 Limit design concept

Introduction

Since the late, 1940's 6. Taguchi has introduced several new SQC concepts,

which have proven to be valuable tools in the subject of quality improvement. Taguchi

has differentiated the quality into three stages as System Design, Parameter Design and

Tolerance Design. The Parameter Design stage is also called Robust Design. Its main

aim is to reduce costs and improve quality. The quality of a product normally depends on

the parameters that govern the behavior of the process for manufacture it. This is

achieved through deriving optimum parameters setting using statistical techniques and

experiments. Taguchi has suggested a new approach for the design of experiments,

which identifies the nature of parameters, by conducting minimum number of

experiments, which is extensively applicable in Research and Development sectors and

manufacturing industries (Genichi Taguchi, 1987).In this study an effort has been made

to optimize the gas carburizing process variables using Taguchi's approach.

Definition

The study of most important variables affecting quality characteristics and a plan

for conducting such experiments is called the Design of Experiments.

Need for planned experimentation

In a highly competitive market, most enlightened companies recognized the need

for continuous improvement to their products and services as a key success factor to

maintain market leadership. The challenge therefore for any organizations is to find out

the methodology to achieve design optimization for quality, cost and delivery.

The basis for the engineering design activity is the Knowledge of scientific

phenomena and past engineering experiences with similar product design and

manufacturing processes. However, when a new product has to be developed a lot of

new decisions have to be made with regard to product profile, critical parameters of the

product design, to various manufacturing processes to be adopted etc. There may be so

many interactive forces that may impact the decision, that it seems to be an

overwhelming task to figure out a simple, economic safe course of action.

~ h e s d situations are common in industry; they affect all departments across the

organization and at all levels. In these cases, it is necessary to experiment to make a

planned change, determine the effect of the change, and use this information to make a

decision about accepting or rejecting the new alternative considered. It is the quality of

this decision, which can be improved upon when proper test strategies are utilized.

In general, planned experiment is necessary to distinguish between critical factors

and non-critical factors as well as to identify the optimum level of the critical factors so

as to pave the way for significantly improved performance. It also enables to predict the

extent of improvements possible over the existing performance.

Terminologies used in Design of Experiments

Response

It is the output of interest to be optimized i.e., Maximized, Minimized, Targeted,

etc.

Factors

A factor is one of the things (Variable) being studied in the experiment. A factor

may be Quantitative or Qualitative.

Level

Levels of a factor are values of the factor being examined in the experiments.

Interaction

It is defined as the joint effect of two or more factors. We consider two factor

interactions only in Industrial experiments.

Treatment combination

A Treatment combination is one set of levels for the factors in a given

Experimental run.

Experimental Design

The analysis of any data is dictated by the manner in which data are collected.

Design of experiment is then a plan for collection of data on response(s) when the chosen

factors are varied in a prescribed manner. The three basic principles of experimental

design are:

i) Replication - Repetition of the basic experiment.

ii) Randomization - The allocation of the experimental material and the order in

which the individual runs are to be performed are randomly determined.

iii) Blocking - When known sources of extraneous and unwanted variation can be

identified, blocking technique is used in such ways that eliminate their influence and

provide a more sensitive test of significance for the variables under study.

Types of experiments

One factorial a t a time These are experiments when in each experiment only one factor is changed fiom

one level to another level, keeping all the other factors unchanged.

Full Factorial experiments This is an experiment method where all factors are tried for all combinations of

their levels.

Fractional factorial experiments

As the name indicates, instead of doing fill factorial, partial factorial is done.

This essentially means a reduction in the number of experiments.

Experimental Design Procedure

Experiments are carried out by researchers or engineers in all fields of study to

compare the effects of several conditions or to discover something new. If an experiment

is to be performed most efficiently, then a scientific approach to planning it must be

considered. The statistical design of experiments is the process of planning experiments

so that appropriate data will be collected, the minimum number of experiments will be

performed to acquire the necessary technical information, and suitable statistical methods

will be used to analyze the collected data (Figure 1.8).

Taguchi's Method and steps in designing experimental layout

G . Taguchi (1959) of Japan, by developing the associated concept of linear

graph, was able to device numerous variants based on the Orthogonal Array (OA) design,

which can easily be applied by an engineer or a scientist without acquiring advanced

statistical knowledge for working out the design and analysis of even complicated

experiments (Ross J. Philip, 1989).

These methods have the advantage of being highly flexible and readily enable

allocation of different levels of factors, even when these levels are not the same in

number for all the factors studied. The beauty of these methods lies in cutting to the bare

minimum the size of experimentation, at the same time, yielding results with high

precision, thus, by a mere 27 experiments, we may be able to evaluate all the main effects

along with one technologically relevant first order interaction through the OA DESIGN,

as against 59,049 experiments needed by a full factorial design for 10 factors each at

three levels (Rout, B.K., 2003).

Figure 1.8 Outline of experimental design procedure

-- Selection of Performing the experimental design experiments

Statement of the problem

1

(i) List down the Response, Factors & levels along with the desired interactions.

-A,

I I

(ii) Find the Degrees of Freedom for each factor and for each interaction.

Recommendation and follow-up

(iii) Compute the Total Degrees of Freedom (TDF).

V

Choice of response

il

(iv) The minimum number of Trials (MNE) is equal to the Total Degrees of

Freedom Plus one (TDF + 1).

Planning of

Understanding of present situation

subsequent experiments

Choice of factors and levels

Confirmation test /- * Data analysis Analysis of results and conclusions

/-

(v) Choose the nearest orthogonal array series like L4, L8, L16 or L9, L27, etc.

(vi) Draw the required Linear Graph (LG).

(vii) Number the linear Graph by starting with the Number 1 for Factor A and

Number 2 for Factor B. Then check whether any interaction exists. If not,

proceed with the Number 3 for Factor C. If there is an interaction, check with

the Lnteraction Table, which Column is to be allotted to the interaction? Then

proceed with the next number for the next factor.

(viii) Complete the numbering as described till the following is achieved.

All the factors and interactions are numbered.

There is no repetition of numbers.

The interaction numbers are as per the Interaction table.

The numbers used do not exceed the number of columns permitted for the

Orthogonal Array Table.

a For Example, only four numbers (Numbers 1 through 4) are permitted for

the L27 OA (Tablel.1).

(ix) Write the column numbers against each factor. That is the Design

Assignment. Rewrite the OA Table with only those columns represented by

factors and all the rows as per the OA Table. Replace the 1, 2 and 3 in the

Table with the Physical value of the level from the Factors and Levels

identified.

This completes the Design Layout.

One need not conduct the Experiment in the same order as in the OA Table. We

can randomize the order by any method of Random Number generation. In the present

research, Taguchi's Design of Experiment (L27 Orthogonal array, Table 1.1) is employed

in the optimization of Gas carburizing process variables.

Table 1.1 Orthogonal array of L 27 showing 5 variables with 3 levels

Degrees of Freedom

It is the number of independent comparisons. In general, if there are n results,

then the number of Degrees of Freedom is n-1.

Orthogonal array

This also called as Design Matrix it means a Balanced Table.

Linear graphs

Enables scientists and engineers to design and analyze complicated experiments

without requiring the basic knowledge of the construction of designs. Associated with

orthogonal mays and pictorially presents the information on main effects and

interactions. Consists of nodes joined with lines - node denotes variables and line denotes

interactions.

Analysis of Variance - ANOVA

ANOVA is a technique for determining equality of two or more Averages based

on data from samples. It is mainly used to isolate the dominate factors or interactions

from a list of suspects and to estimate the proper level for each important variables in

order to yield optimum end results.

F-ratio

It is the ratio of two variances. This ratio follows a distribution known as F-

Distribution. F-Distribution is defined through the degrees of freedom. It is defined by

the numerator and denominator degrees of freedom (Richard A. Johnson, 2000).

Signal to Noise ratio

It was developed as a proactive equivalent to the reactive loss function. Signal

factors ( Y ) is set by the designer or operator to obtain the intended value of the response

variable. Noise factors s2 are not controlled or very expensive or difficult to control. In

elementary form, S/N is Y / s2.

Factorial Method of Design of Experiment

An experiment involving the study of more than one-variable. Such an

experiment is called a full factorial if all possible variable-level combinations are being

experimented upon otherwise it is called a fractional-factorial or a fractional-replicate

(Douglas, C., and Montgomery, 2001).

Factorial experiments are very efficient ways to study multiple variables in the

same experiment. They allow one to assess any possible interaction, and if no

interactions are present, then the effect of each factor can be evaluated with the same

efficiency as though the whole experiment has been devoted to entirely that variable.

However, factorial experiments even with moderate number of variables can lead

to large number of experimental trials, industrial experiments are always restricted trials,

and industrial experiments are always restricted as to the number of experiments that can

be run due to time and expense. Therefore, two extremely usehl factorial design are

those with all the variables at only two levels or all variables at only three levels. These

are referred to as 2k or 3' fractional design, respectively.

In the present work of Induction Hardening process variables study 33 Factorial

design of experiment is followed and Table 1.2 shows the details.

Table 1.2 33 Factorial method of Design Matrix

The word sources intended to include variables and levels as well as interaction effects.

The variables P, S, and Q were assigned the first three columns.

(i) All possible combinations of levels (33 = 27) do appear in the first three

columns.

(ii) A11 possible interactions (three two ways and one 3-way interaction) are

assigned to a column, which simply is the product of the columns

corresponding to the factors in the interactions.

(iii) The design of Table 1.2 is an orthogonal and balanced design.