Bhadeshia H.K.D.H., Honeycombe R. Steels: Microstructure and Properties

Подождите немного. Документ загружается.

168 CHAPTER 8 THE HEAT TREATMENT OF STEELS: HARDENABILITY

8.2 USE OF TTT AND CONTINUOUS COOLING DIAGRAMS

TTT diagrams provide a good starting point for an examination of hardenability,

but as they are statements of the kinetics of transformation of austenite carried

out isothermally, they can only be a rough guide.To take one example,the effect

of increasing molybdenum, Fig. 8.1 shows the TTT diagrams for a 0.4 wt% C–

0.2 wt% Mo steel and a steel with 0.3 wt% C–2 wt% Mo. The 0.2 wt% Mo steel

begins to transform in about1sat550

◦

C,but on increasing the molybdenum to

2 wt% the whole C-shaped curve is raised and the reaction substantially slowed

so that the nose is above 700

◦

C,the reaction starting after 4 min. The latter steel

will clearly have a greatly enhanced hardenability over that of the 0.2 wt% Mo

steel.

(a)

Fig. 8.1 (a) TTT diagram of a molybdenum steel 0.4C, 0.2Mo (Thelning, Steel and its Heat

Treatment, Bofors Handbook, Butterworth, 1975).

8.2 USE OF TTT AND CONTINUOUS COOLING DIAGRAMS 169

(b)

Fig. 8.1 (b) TTT diagram of a molybdenum steel: 0.3C, 2.0Mo (Thelning, Steel and its Heat

Treatment, Bofors Handbook, Butterworth, 1975).

The obvious limitations of using isothermal diagrams for situations involv-

ing a range of cooling rates through the transformation temperature range have

led to efforts to develop more representative diagrams, i.e. continuous cool-

ing transformation (CCT) diagrams. These diagrams record the progress of

the transformation with falling temperature for a series of cooling rates. They

are determined using cylindrical rods which are subjected to different rates of

cooling, and the onset of transformation is detected by dilatometry, magnetic

permeability or some other physical technique. The products of the transform-

ation, whether ferrite, pearlite or bainite, are partly determined from isothermal

diagrams, and can be conﬁrmed by metallographic examination. The results are

then plotted on a temperature/cooling time diagram, which records, e.g. the

time to reach the beginning of the pearlite reaction over a range of cooling

rates. This series of results will give rise to an austenite–pearlite boundary on the

170 CHAPTER 8 THE HEAT TREATMENT OF STEELS: HARDENABILITY

Fig. 8.2 Relation between cooling curves for the surface and core of an oil-quenched 95-mm

diameter bar and the microstructure. The surface is fully martensitic (Thelning, Steel and its

Heat Treatment, Bofors Handbook, Butterworth, 1975).

diagram and, likewise, lines showing the onset of the bainite transformation can

be constructed. A schematic diagram is shown in Fig. 8.2 in which the bound-

aries for ferrite, pearlite, bainite and martensite are shown for a hypothetical

steel. The diagram is best used by superimposing a transparent overlay sheet

with the same scales and having lines representing various cooling rates drawn

on it. The phases produced at a chosen cooling rate are then those which the

superimposed line intersects on the CCT diagram. In Fig. 8.2 two typical cool-

ing curves are superimposed for the surface and the centre of an oil-quenched

95-mm diameter bar. In this example, it should be noted that the centre cool-

ing curve intersects the bainite region and consequently some bainite would be

expected at the core of the bar after quenching in oil.

8.3 HARDENABILITY TESTING

The rate at which austenite decomposes to form ferrite, pearlite and bainite is

dependent on the composition of the steel, as well as on other factors such as

the austenite grain size, and the degree of homogeneity in the distribution of

the alloying elements. It is now possible to estimate hardenability using phase

8.3 HARDENABILITY TESTING 171

transformation theory, but there is also a reliance on one of several practical

tests, which allow the hardenability of any steel to be readily determined.

8.3.1 The Grossman test

Much of the earlier systematic work on hardenability was done by Grossman

and co-workers who developed a test involving the quenching, in a particular

cooling medium, of several cylindrical bars of different diameter of the steel

under consideration.Transverse sections of the different bars on which hardness

measurements have been made will show directly the effect of hardenability. In

Fig. 8.3, which plots this hardness data for an SAE 3140 steel (1.1–1.4Ni, 0.55–

0.75Cr,0.40C wt%) oil-quenched from815

◦

C,it is shown thatthe fullmartensitic

hardness is only obtained in the smaller sections, while for larger diameter bars

the hardness drops off markedly towards the centre of the bar. The softer and

harder regions of the section can also be clearly resolved by etching.

In the Grossman test, the transverse sections are metallographically exam-

ined to determine the particular bar which has 50% martensite at its centre.

The diameter of this bar is then designated the critical diameter D

. However,

this dimension is of no absolute value in expressing the hardenability as it will

obviously vary if the quenching medium is changed, e.g. from water to oil. It

is therefore necessary to assess quantitatively the effectiveness of the different

quenching media. This is done by determining coefﬁcients for the severity of the

Fig. 8.3 1.1 Ni–0.75Cr–0.4C steel. Hardness data from transverse sections through

water-quenched bars of increasing diameter (Grossman et al., in Alloying Elements in Steel (eds

Bain and Paxton),ASM, Ohio, USA, 1961).

172 CHAPTER 8 THE HEAT TREATMENT OF STEELS: HARDENABILITY

Table 8.1 H-coefficients of quenching media.

Agitation Cooling medium

Oil Water Brine

None 0.25–0.30 1.0 2.0

Moderate 0.35–0.40 1.2–1.3

Violent 0.8–1.1 4.0 5.0

quench usually referred to as H-coefﬁcients. Typical values for three common

quenching media and several conditions of agitation are shown in Table 8.1.

The value for quenching in still water is set at 1, as a standard against which to

compare other modes of quenching.

Using the H-coefﬁcients, it is possible to determine in place of D

, an ideal

critical diameter D

which has 50% martensite at the centre of the bar when the

surface is cooled at an inﬁnitely rapid rate, i.e. when H =∞. Obviously, in these

circumstances D

, thus providing the upper reference line in a series of

graphs for different values of H (Fig. 8.4). In practice, H varies between about

0.2 and 5.0 (Table 8.1), so that if a quenching experiment is carried out at an

H-value of, say, 0.4, and D

is measured, then the graph of Fig. 8.4 can be used

Fig. 8.4 Chart for determining ideal diameter (D

) from the critical diameter (D

) and the

severity of quench (H) for carbon and medium alloy steels (Grossman and Bain, Principles of

Heat Treatment,ASM, Ohio, USA, 1964).

8.3 HARDENABILITY TESTING 173

to determine D

. This value will be a measure of the hardenability of a given

steel, which is independent of the quenching medium used.

8.3.2 The Jominy end quench test

While the Grossman approach to hardenability is very reliable, other less elab-

orate tests have been devised to provide hardenability data. Foremost amongst

these is the Jominy test, in which a standardized round bar (25.4 mm diameter,

102 mm long) is heated to the austenitizing temperature, then placed on a rig in

which one end of the rod is quenched by a standard jet of water (Figs. 8.5a, b).

This results in a progressive decrease in the rate of cooling along the bar from

the quenched end, the effects of which are determined by hardness measure-

ments on ﬂats ground 4 mm deep and parallel to the bar axis (Fig. 8.5c).A typical

hardness plot for a steel containing 0.4C–1Cr–0.25Mo wt% (En 19B) is shown

in Fig. 8.6, where the upper curve represents the hardness obtained with the

upper limit of composition for the steel, while the lower curve is that for the

composition at the lower limit. The area between the lines is referred to as a

hardenability or Jominy band. Additional data, which are useful in conjunction

with these results, is the hardness of quenched steels as a function both of car-

bon content and of the proportion of martensite in the structure. These data are

given in Fig. 8.7 for as-quenched steels with 50–90% martensite. Therefore, the

hardness for 50% martensite can be easily determined for a particular carbon

content and, by inspection of the Jominy test results, the depth at which 50%

martensite is achieved can be determined.

The Jominy test is now widely used to determine hardenabilities in the range

=1–6 cm; beyond this range the test is of limited use. The results can be

readily converted to determine the largest diameter round bar which can be

fully hardened. Figure 8.8 plots bar diameter against the Jominy positions at

which the same cooling rates as those in the centres of the bars are obtained

for a series of different quenches. Taking the ideal quench (H =∞), the highest

curve, it can be seen that 12.5 mm along the Jominy bar gives a cooling rate

equivalent to that at the centre of a 75-mm diameter bar. This diameter reduces

to just over 50 mm for a quench in still water (H =1). With, e.g., a steel which

gives 50% martensite at 19 mm from the quenched end after still oil quenching

(H =0.3), the critical diameter D

for a round rod will be 51 mm.

The diagram in Fig. 8.8 can also be used to determine the hardness at the

centre of a round bar of a particular steel, provided a Jominy end quench test has

been carried out. For example, if the hardness at the centre of a 50-mm diameter

bar, quenched in still water, is required, Fig. 8.8 shows that this hardness will

be achieved at about 12 mm along the Jominy test specimen from the quenched

end. Reference to the Jominy hardness distance plot then gives the required

hardness value. If hardness values are required for other points in round bars,

e.g. surface or at half-radius, suitable diagrams are available for use.

174 CHAPTER 8 THE HEAT TREATMENT OF STEELS: HARDENABILITY

Fig. 8.5 The Jominy end quench test: (a) specimen size; (b) quenching rig; (c) Jominy hard-

ness–distance curves for a shallow and a deep hardening steel (Wilson, Metallurgy and Heat

Treatment of Tool Steels, McGraw-Hill, 1975).

8.3 HARDENABILITY TESTING 175

Fig. 8.6 Jominy curves for upper and lower limits of a steel, En 19B, giving a hardenability band

(Thelning, Steel and its HeatTreatment, Bofors Handbook, Butterworth, 1975).

Fig. 8.7 The effect of percentage of martensite and carbon content on as-quenched hardness

(Thelning, Steel and its HeatTreatment, Bofors Handbook, Butterworth, 1975).

176 CHAPTER 8 THE HEAT TREATMENT OF STEELS: HARDENABILITY

Fig. 8.8 Equivalent Jominy positions and bar diameter, where the cooling rate for the bar

centre is the same as that for the point in the Jominy specimen. Curves are plotted for a range

of cooling rates (Grossman and Bain, Principles of Heat Treatment,ASM, Ohio, USA, 1964).

8.4 EFFECT OF GRAIN SIZE AND CHEMICAL COMPOSITION

ON HARDENABILITY

The two most important variables which inﬂuence hardenability are austenite

grain size and composition. The hardenability increases with increasing austen-

ite grain size, because the grain boundary area per unit volume decreases. The

sites for the nucleation of ferrite and pearlite are reduced in number, with

the result that these transformations are slowed down, and the hardenability

therefore increases. Alloying elements which slow down the ferrite and pearlite

reactions increase hardenability. However, quantitative assessment of these

effects is needed.

The ﬁrst step is to determine the effect of grain size and of carbon content.

Data are available, in so far as D

has been determined for steels with carbon

in the range 0.2–1 wt%, and for a range of grain sizes (ASTM 4–8), as shown in

Fig. 8.9. Use of this diagram for any steel provides a base hardenability ﬁgure,

8.5 HARDENABILITY AND HEAT TREATMENT 177

Fig. 8.9 Effect of carbon content and grain size on base hardenability (Moser and Legat,

HärtereiTechnische Mitteilungen 24, 100, 1969).

, which must be modiﬁed by taking into account the effect of additional

alloying elements. This is done by use of multiplying factors which have been

experimentally determined for the familiar alloying elements (Fig. 8.10). The

ideal critical diameter D

is then found from the empirical relationship:

= D

× 2.21 (%Mn) × 1.40 (%Si) × 2.13 (%Cr)

× 3.275 (%Mo) × 1.47 (%Ni) (weight percentages). (8.1)

This relationship, due to Moser and Legat, appears to be more accurate in

practice than a much earlier one put forward by Grossman. Further corrections

have to be made for different austenitizing temperatures when dealing with high

carbon steels, but, on the whole, the relationship is quite effective in predicting

actual hardening behaviour.

8.5 HARDENABILITY AND HEAT TREATMENT

While alloying elements are used for various reasons, the most important is the

achievement of higher strength in required shapes and sizes and often in very

large sectionswhich may be up toa metre or more in diameter in thecase of large

shafts and rotors. Hardenability is, therefore, of the greatest importance, and

one must aim for the appropriate concentrations of alloying element needed to