Canale L.C.F., Mesquita R.A., Totten G.E. Failure Analysis of Heat Treated Steel Components

Подождите немного. Документ загружается.

time-temperature variation at a depth of 10 mm,

it can be assumed that the specimen was heated

through the entire volume, that is, to the very

core of the specimen. Because of a different time

variation of temperature in the fourth tempera-

ture cycle, it may be concluded that the maxi-

mum temperature obtained in the depth of

10 mm is lower than the magnetic transforma-

tion temperature of the given steel. Because of

strong overheating of the cylindrical specimen

toward its center, lower temperature gradients

occur, which result in a reduction of thermal

stresses during the heating process. According to

a comparatively high temperature in the core,

temperature gradients between the surface and

the core are generally lower, which produce a

decrease in axial internal stresses generated

during quenching and also a decrease in axial

residual stresses.

Figure 39 shows the time variations of

through-thickness temperature in single-shot

induction heating with high-frequency generator

powers of 40, 60, 100, and 180 kW and a

medium-frequency current (Ref 51). Charac-

teristic temperature transformations, T

and

, for equilibrium heating are plotted (Ref 51).

Induction heating is a very fast process; there-

fore, the temperature transformations shift to

higher temperatures. To obtain homogeneous

austenite in the surface layer, it is necessary to

heat the surface layer to the hatch-marked tem-

perature range. The results shown in the four

diagrams make it possible to draw the following

conclusions:

 The heating curves differ strongly.

 Only with the power of 40 kW does the tem-

perature not reach the hatch-marked tem-

perature range in 30 s.

 With all other powers higher than 60 kW,

the hatch-marked temperature range is

reached in a shorter heating time.

 With all powers higher than 60 kW, more

intensive heating of the surface layer occurs,

which is shown in a steeper hardness curve in

the hatch-marked temperature range.

The ﬁrst question raised by an engineer would

be how to evaluate the through-thickness var-

iations of hardness and residual stresses with

reference to the time variation of temperature.

Fig. 39 Time-temperature variations in single-shot heating at various powers. Source: Ref 51

448 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 448

The second question would be how to select

induction-heating parameters to obtain optimal

properties of the surface-hardened layer with

a minimal energy input. In answer, the initial

steel microstructure must be taken into account

and an appropriate quenchant selected to en-

sure a cooling rate in the surface layer equal to

or a bit higher than the critical cooling rate

required.

Figure 40 shows the time and through-thick-

ness temperature variations of the specimen on

heating with different powers and different

velocities of workpiece movement, that is,

= 140 mm/min (Fig. 40a), v

= 220 mm/

min (Fig. 40b), v

= 370 mm/min (Fig. 40c),

and v

= 680 mm/min (Fig. 40d), in scan hard-

ening (Ref 51). In the ﬁrst case, with v

, the

hardened layer is obtained with the powers of 74

and 56 kW. In the second case, with a higher

velocity of movement, v

, the hatch-marked

temperature range is reached only in heating

with the powers of 78 and 59 kW. In the third

case, with v

, the hatch-marked temperature

range is reached only in heating with the power

of 112 kW, whereas in the fourth case, with

the highest velocity, v

, this temperature range is

not reached in spite of a very high power,

151 kW.

Quenching Systems for Induction

Hardening

Control of the quenching process of a gear

wheel from high hardening temperatures

ensures martensite microstructure. The cooling

rate should be high enough to prevent formation

of unwanted softer microstructures, such as

pearlitic or bainitic ones; therefore, it is very

important that in the development of new sys-

tems of induction hardening, quenching systems

are designed properly.

Process parameters should be accurately

controlled to ensure permanent and reproducible

results after hardening of machine parts. It is also

indispensable to deﬁne individual parameters

and determine their permissible deviation in

operation, presuming that the deviations have a

negligible inﬂuence on the results of heat treat-

ment. The quenching system for induction hard-

ening is deﬁned by eight parameters (Ref 17,

Fig. 40 Time-temperature variations in scan hardening at various scan speeds. Source: Ref 51

Induction Hardening / 449

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 449

52): heat time/scan rate, power level, power

frequency, part position/rotation, quenchant

concentration, quenchant ﬂow, quenchant tem-

perature, and quenching time.

Important simultaneous changes of one or

more of these parameters can produce unwanted

effects on the workpiece, which, in extreme

cases, result in an unsuitable microstructure, de-

viations in the depth of the hardened proﬁle,

unsuitable hardness variation (too low hardness,

soft spots), and exceeding the distortion of the

machine element.

In practice, there are machine parts of differ-

ent shapes and sizes, requiring different depth of

the hardened layer. In these cases, the type of

material chosen and its through hardenability

should be considered.

Thus, with alloyed steels having good through

hardenability and/or machine parts with a com-

paratively thin hardened surface layer, the mar-

tensite microstructure can be obtained without

the application of a quenchant. In such cases,

heat sinks from the surface into the cold work-

piece core, so that the critical self-cooling rate

obtained at the surface is higher than the critical

cooling rate (Ref 24, 53).

Figure 41(a) shows the time variation of the

temperature measured at the individual mea-

suring points of the gear-wheel tooth during

quenching with a pressurized water jet (Ref 52).

Data on the temperature variation show that

the cooling temperatures were very similar at

measuring points 1 and 2, somewhat lower at

measuring point 3, and the lowest at measuring

point 4, where, after 40 s of quenching, it still

equaled approximately 350



Figure 41(b) shows the calculated time var-

iation of temperature at the same measuring

points of the gear-wheel tooth during quenching,

taking into account selected heat-transfer coef-

ﬁcients, a

, of 5000 and 10,000 kcal/m



(Ref 52). The calculated time variations of

temperature differ from the measured ones only

at those measuring points with higher cooling

rates. The time variation of temperature refers

only to the selected measuring points; therefore,

assessment of the stress state during the quen-

ching process and the distortion of gear-wheel

teeth or a tooth is very difﬁcult.

The authors focused on an analysis of the con-

ditions during the quenching process by means

of a calculated distribution of temperature,

shown as isotherms (a

= 5000 kcal/m



Ch)

at the half-cut of the gear-wheel tooth. Figure 42

shows the variation of isotherms during the

quenching process after 0.1, 1.0, and ﬁnally 10 s

of cooling (Ref 52).

In quenching, it is very important that the

temperature gradients are high enough to pre-

vent plastiﬁcation, that is, distortion, of the gear

wheel. Such quenching conditions, including

high-temperature gradients, result in a decrease

Fig. 41

Temperature proﬁles of tooth gear at given measur-

ing points during cooling process. (a) Measured. (b)

Calculated. Source: Ref 52

Fig. 42

Temperature distributions during the cooling

process. Source: Ref 52

450 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 450

in favorable compressive residual stresses after

hardening. It follows that a sufﬁciently high

cooling rate must be ensured at the tooth surface

as well as at the required depth of hardening. That

means that temperature gradients in the gear-

wheel tooth must be attained so that gear-wheel

teeth are hardened with the minimum required

cooling rate, ensuring the required martensite

transformation. A sequence of images showing

isotherms in very short time intervals from the

beginning of quenching allows assessment of the

size of distortion of the tooth and residual stresses

to be expected after quenching.

Figure 43(a) shows the measured volume-

fraction distribution of martensite and Fig. 43(b)

the calculated volume-fraction distribution of

martensite when different values of the heat-

transfer coefﬁcient, a, that is, 5000, 7500, and

10,000 kcal/m



Ch, are taken into account (Ref

52). The main difﬁculty encountered in the cal-

culation is how to determine the heat-transfer

coefﬁcient to obtain a description of real quen-

ching conditions. Quenching is very intensive,

since it is carried out under a pressurized water-

jet; the gear wheel has been induction heated,

whereas the core is cold.

Two common quenching methods that use a

quenchant are spray quenching and immersion

quenching. Particularly popular spray quenching

techniques that offer different possibilities are:

 Spray with progressive scan heating (scan

hardening)

 Spray after heating in position (single-shot

hardening)

 Spray quenching out of location after

heating

Spray quenching is carried out immediately

after heating with a short pause. It is used with

machine parts made of steels with good hard-

enability and with a comparatively thin hard-

ened layer. Immersion quenching is carried out

after the inductor has been disconnected from

the high-frequency generator. Gradual hard-

ening is very efﬁcient and prevents the inﬂuence

of inverse heat from the core toward the surface,

so that no reverse heat ﬂow and no heating of the

already cooled surface occur. This problem is

characteristic of large parts, where the work-

piece surfaces should be hardened selectively.

In the past, straight oils and water-soluble oils

were used for quenching after surface induction

heating. Straight oils and water-soluble oils

produce mild quenching effects, which reduce

difﬁculties due to distortion and/or crack initia-

tion.

Straight oils require immersion quenching

to minimize the risk of oil ignition, whereas

water-soluble oils are more suitable for spray

quenching. Other quenchants used are water,

polymeric water solutions of different con-

centration, water-soluble oils, saltwater, and

so on.

Polymeric water solutions are inﬂammable

quenchants. They are prepared in various con-

centrations to be suitably adapted to various

cooling rates. For spray quenching, various ﬂow

rates of quenchants can be selected. They

depend on the size of inlet openings and the

Fig. 43 Martensite distribution of the hardened tooth gear. Source: Ref 52

Induction Hardening / 451

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 451

number of outlet nozzle boreholes in one, two, or

more rows at the quenching ring. The ﬂuid ﬂow

rate can also be regulated by pressure. Technical

literature provides data on shaping of quenching

nozzles as well as on the selection of the quen-

chant ﬂow rate required. It is very important that

a sufﬁcient ﬂow rate of the quenchant is chosen

to ensure sufﬁcient heat removal from the work-

piece surface layer during quenching.

The ﬂow rate is also controlled by the number

of boreholes for the ﬂow and spraying of the

quenchant, respectively. An appropriate arrange-

ment of the boreholes in the ring permits a

uniform heat removal after spray and proper

and uniform cooling. The borehole cross sec-

tions should be only 15% of the total available

ring surface and of the inductor in the single-

shot-type application, respectively. Rotation

velocities of the workpieces in surface induction

heating are relatively high and can be selected

between 800 and 1000 rotations per minute,

whereas the rotation velocity of the workpieces

during quenching is considerably lower, that is,

40 to 60 rotations per minute. Very high rotation

velocities of the workpieces are required when a

uniform and reproducible depth of hardened

layer is to be ensured, and particularly with

relatively short heating times.

Figure 44 shows the dependence of cooling

rate changes on the momentary temperature at

the workpiece surface for four different con-

centrations of polymeric water solutions (Ref

24). The selection of an appropriate concentra-

tion of the polymeric water solution for the

selected steel and the required depth of hardened

layer should ensure minimum distortion of the

workpiece, which is the ﬁnal purpose of any heat

treatment. Thus, the required heat removal from

the heated surface layer of the workpiece, con-

sidering the thickness of the heated layer, should

be ensured by an appropriate quenching system.

Time Variation of Stresses and

Residual Stresses

With a certain heat treatment performed, the

required microstructural changes and an appro-

priate magnitude and variation of hardness are

obtained. It is also required that distortion of a

machine part be as small as possible so that the

ﬁnal size of a machine part can be obtained with

minimum precision machining. Consequently, it

is very important that distortion be kept under

control. An appropriate machining technology

and appropriate heat treatment processes should

be selected, and internal stresses lower than the

yield stress should be ensured at any moment

and any location of a machine part during heat-

ing and/or cooling. Analytical methods provide

an insight into heat treatment conditions if the

time variation of internal stresses is monitored

and the dependence between the cooling time

and the specimen temperature at each point is

known. With regard to the specimen tempera-

ture determined in this way, the specimen yield

stress at each point can be determined as well.

Consequently, with different heat treatment con-

ditions, different values of physical quantities

can be selected. They are reﬂected in the chan-

ged conditions in the material, which make it

possible to study distortion of the machine part

during cooling and to determine the magnitude

of the residual stresses.

Figure 45 shows the time variation of axial

stresses at the surface and in individual depths,

that is, from 2.0 to 4.0 mm under the surface

and in the middle of a cylindrical specimen

(Ref 50).

In surface heating, tensile stresses occurred

at the surface due to thermal extension of the

Fig. 44

Polymer additive ratio effects on the cooling rate.

Source: Ref 24

452 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 452

surface layer, but when the transformation

temperature of the pearlitic- to austenitic micro-

structure was exceeded, additional compressive

stresses occurred at the surface.

The compressive axial stresses transform

into tensile stresses in the zone of undercooled

austenite. In the transformation of undercooled

austenite into martensite, the compressive

stresses increase with the increase in martensite

fraction. In the core, the opposite sign of the

stress was obtained, that is, the tensile stress. At

the end of quenching, the compressive residual

stresses obtained in the surface-hardened layer

were approximately 1600 N/mm

, and the

tensile stresses in the core were approximately

+870 N/mm

Internal stresses are induced in heat treatment

by temperature and microstructural changes.

Residual stresses in the induction surface-

hardened layer are always of a compressive

nature, are relatively high, and have a good

effect on dynamically loaded components. The

existence of residual stresses in the radial

direction, that is, into the depth of the hardened

layer, is very important, as is the absolute value

of residual stress on the surface and the stress

proﬁle in the transition from compressive into

tensile stresses (Fig. 46) (Ref 15).

In the case of induction hardening, a maxi-

mum compressive residual stress in the surface

layer is achieved, which is very desirable for

dynamically loaded components. The transition

from compressive into tensile stresses should be

as gentle as possible, lessening the effect of

stress concentration in loaded components. This

contributes to the fact that a machine component

is less susceptible to overloads in operation. It

has been shown that residual stresses are closely

linked to hardness variation and microstructure

in the transition zone of the hardened layer, that

is, in the narrow range between the hardened and

nonhardened microstructure.

Some examples of the dependence between

microhardness and residual stresses are shown in

Fig. 45 Axial stres s distribution at various depths below the surface during single-shot induction surface hardening. Source: Ref 50

Fig. 46

Residual stress proﬁle below the surface after

induction surface hardening. Source: Ref 15

Induction Hardening / 453

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 453

Fig. 47 (Ref 15). Figure 47(a) shows a very steep

microhardness proﬁle in the transition zone and

the highest compressive stress in the surface,

and, related to this, a very steep transition of

residual stresses into the tensile range. The

change from compressive into residual stresses

happens at the transition between the hardened

and nonhardened areas. Figures 47(b and c)

show graphs of broader transition zones, which

produce a change in the microhardness and also

the residual stresses (Ref 15). Thus, Fig. 47(c)

shows a very slight drop in hardness in the

transition zone for the same microhardness on

the surface.

This is reﬂected in lower compressive stresses

in the surface, accompanied by a slight change

of stresses in the transition zone into lower

tensile residual stresses. In induction surface

hardening, the engineer should choose the kind

of heat treatment conditions that will result in

the microhardness and residual-stress proﬁles

shown in the last two examples in Fig. 47.

Investigations of residual stresses after induc-

tion surface hardening have conﬁrmed that when

the hardened layer is 2 mm thick, the change of

compressive into tensile stresses happens in

compliance with the transition zone, that is,

the achieved depth of the hardened layer. When

the thickness of the hardened layer is greater

than 2 mm, the transition from compressive

into tensile stresses happens in the hardened

zone, that is, the martensite microstructure. This

means that induction surface hardening is much

more difﬁcult if the thickness of the layer is

above 2 mm. Due to intensive cooling, the

internal stresses in the surface layer may become

so high that they cause failure of the component.

This failure is very typical, because the sur-

face layer separates from the core due to high

radial stresses. Important features on the curve

are:

 Maximum value of compression residual

stress in the hardened surface layer

 Maximum value of tension residual stress in

the hardened surface layer

 Transition width of compression to tension

of residual stress in the hardened surface

layer

 Transition steepness of compression to ten-

sion of residual stress proﬁle

 Layer depth with transition microstructure

The examples shown in Fig. 46 and 47 were

based on a presumption that the maximum

hardness and the maximum value of the residual

stresses obtained were independent from the

heating power density and the heating time

(Ref 15). In the case of induction surface hard-

ening, a maximum compressive residual stress

in the surface layer is achieved. The transition

from compressive into tensile residual stresses

should be as gentle as possible, lessening the

effect of stress concentration in loaded compo-

nents. This contributes to the fact that a mach-

ined component is less susceptible to overloads

in operation. It has been shown that residual

Fig. 47 Various residual-stress and hardness proﬁles below the surface. Source: Ref 15

454 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 454

stresses are closely linked to hardness variation

and microstructure in the transition zone of the

hardened layer and base material, that is, in the

narrow range between the hardened and non-

hardened microstructure.

On the basis of the heating temperature cycle,

microstructural changes in the surface layer

after quenching can be predicted. Consequently,

the temperature cycles at the surface, that is, a

through-thickness heated layer, have a decisive

inﬂuence on the variation of residual stresses

after quenching. The variation of the tempera-

ture cycles in heating can be adjusted by the

selection of adequate energy input with refer-

ence to the steel microstructure and its thermal

properties.

Figure 48 shows three examples of induction

surface hardening, each with different energy

input during heating, whereas the quenching

process was the same in all cases (Ref 15, 54).

The variation of residual stresses after induction

surface hardening is affected by the energy input

through the workpiece surface and thermal

conductivity of the workpiece material. The

energy input in the workpiece depends on the

power chosen, the duration of high-frequency

current, the shape of an induction loop, the size

of a gap between the induction loop and the

workpiece, and the area exposed to the energy

input.

In progressive hardening, the velocity of

progression of the high-frequency loop along the

heated area must be considered.

The transition from tensile to compressive

residual stresses presents a serious danger for

catastrophic failure. From the aforementioned,

it can be understood that the workpiece must

be slowly heated and also quenched with the

correct cooling rate. The actual cooling rate is

very important throughout the martensitic trans-

formation and must be as close as possible to

the critical cooling rate. The quenching process

must be carried out very carefully in the mar-

tensitic transformation temperature range to

ensure internal stresses lower than the yield

point. The only barely-seen difference is in the

transition area, when the compressive residual

stress starts to fall rapidly. In the transition area,

the residual-stress curve is steeper and less

desirable as far as notch effects under dynamic

loads are concerned. In general, efforts should

be made to achieve low or moderate residual

stresses in the transition area after induction

surface hardening. Manufacturing engineers are

often confronted with the question of how to

obtain this kind of residual-stress proﬁle. It

should be emphasized that because of their

lower heat conductivity, alloyed steels for sur-

face hardening are very difﬁcult to treat, and it is

difﬁcult to ensure desirable residual stresses in

the transition area. This is why special alloyed

steels are available that display a more favorable

behavior in heating and quenching. Because of

these difﬁculties, delamination of the surface to

the very core may occur. Figure 48(a) shows that

that the energy input was too weak, since only a

partial transformation of the steel matrix into

austenite occurred (Ref 15, 54). After quench-

ing, this will produce only a small portion of

martensitic microstructure, which is followed by

a measured variation of residual stresses. The

second example, shown in Fig. 48(b), indicates

that heating with a high-power density was very

intensive. The result of a too-high power density

is a distinct inﬂuence of heat conduction from

Fig. 48

Residual-stress proﬁles after induction surface hard-

ening at various input energies. Source: Ref 15, 54

Induction Hardening / 455

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 455

the surface to the subsurface, which, in turn,

resulted in a higher temperature measured in the

subsurface than at the surface after the delayed

interruption of heating at the beginning of the

quenching process. Because of such a heating

regime and energy input at the lower limit of

the required energy, considerably lower com-

pressive residual stresses were achieved at the

surface. In this case, the highest value of com-

pressive residual stress was obtained at a depth

of 1.0 mm. The variation of the stress was not

the most adequate, since induction surface hard-

ening can provide considerably higher com-

pressive residual stresses in the surface layer.

The anticipated variation of the residual stresses,

that is, very high compressive residual stresses

ranging between 800 and 1000 N/mm

,is

shown in Fig. 48(c) (Ref 15, 54).

A very favorable variation of residual stresses

can be found next to the transition zone between

the quenched and unquenched surface layers,

which ensures favorable behavior of a machine

part under dynamic loads.

In practice, the progress of heating and

quenching can be monitored by physical mod-

elling, in which case the temperature cycle at the

workpiece surface as well as the temperature

cycles in the individual depths are obtained. The

variations of the temperature cycles in heating

and the known temperatures, that is, temperature

ranges, make it possible to determine changes

of a pearlite-ferrite microstructure into an

austenitic one. The latter makes it possible to

predict the microstructure after surface hard-

ening.

The progress of induction surface heating can

be supervised by measuring the temperature at

the workpiece surface using a pyrometer. The

temperature measurement is a determination of

temperature cycles in individual depths during

induction heating and subsequent cooling. There

are three basic and important metallurgical

conditions that will affect a successful induction

surface-hardening operation:

 Lower critical temperature, where metal-

lurgical phase transformation begins

 Upper critical temperature, where the for-

mation of austenitic grain microstructure is

complete

 Surface temperature, which affects the

resultant grain size of the heat treated micro-

structure

For a typical AISI 140 material, the steel

reference books will list T

=750



=795



C, and T

smax

=900



C. However,

because of the shorter heating times with

induction heating, the actual values required for

a satisfactory microstructure and process may be

30 to 60



C higher than for conventional furnace

heat treatment processes.

Figure 49 shows the temperature cycles in the

surface and in the individual depths during in-

duction heating and quenching (Ref 15, 54).

The temperature range of the pearlite-ferrite

transformation into austenite, which is a con-

dition for the formation of the martensitic micro-

structure after quenching, is plotted also.

On the basis of the heating temperature cycle,

microstructural changes in the surface layer after

Fig. 49 Temperature cycles during single-shot induction surface hardening, at various depths. Source: Ref 15, 54

456 / Failure Analysis of Heat Treated Steel Components

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 456

quenching can be predicted. Consequently, the

temperature cycles at the surface, that is, a

through-thickness heated layer, have a decisive

inﬂuence on the variation of residual stresses

after quenching. The variation of the tempera-

ture cycles in heating can be adjusted by the

selection of adequate power density with refer-

ence to the steel microstructure and its thermal

properties.

Figure 50 shows graphic representations of

temperature cycles at the surface at various

power densities in single-shot induction hard-

ening (Ref 15, 54). Each temperature cycle

includes induction heating, which is followed by

quenching a short while after the interruption

of heating. Four temperature cycles of induction

heating and quenching measured at different

heating power densities using high-frequency

current were plotted. With the same inductor

having the same gap width between the inductor

and the workpiece surface, but with different

power densities, different temperature cycles of

heating were obtained. With a higher power

density, the heating curve was steeper and the

heating period shorter (Ref 54).

In all cases of heating, however, the same

depth of the hardened layer was ensured. With

the given power density, however, it is more

difﬁcult to ensure the same depth of the hard-

ened layer, since the heating times are con-

siderably reduced. With very short heating

times, it is much more difﬁcult to ensure the

same depth of the hardened layer, which indi-

cates a worse repeatability of the results. The

shorter heating times with the same cooling

times contribute to a shorter hardening cycle and

result in higher productivity. This higher pro-

ductivity, however, entails a high consumption

of electric energy for heating and less favorable

through-thickness microhardness and residual-

stress variations in the hardened layer. Figure 51

shows microhardness and residual-stress varia-

tions with the highest power density (Q

) and

different heating times, t

... t

(Ref 15, 54).

The inﬂuence of heating time affects the

microhardness and residual-stress variations.

With a correctly chosen condition of induction

heating of the thin surface layer, adequate values

of the maximum compressive residual stresses

are obtained at the surface. The variation of

residual stresses across the hardened surface

layer, that is, in the transition zone between

the hardened and the unhardened material, is

more important. The variation should not be

steep, which means that the gradient of residual

stresses in the transition zone between the hard-

ened and unhardened material should be very

gently sloping. The microhardness variation in

the transition zone, with a gradient of residual

stresses as gently sloping as possible, can be

accomplished by selecting the given power

density and an appropriate heating time, since

Fig. 50 Temperature cycles at the surface in induction surface heating and quenching at various power densities. Source: Ref 15, 54

Induction Hardening / 457

Name ///sr-nova/Dclabs_wip/Failure_Analysis/5113_417-501.pdf/Chap_13/ 18/8/2008 4:01PM Plate # 0 pg 457