In this article we will discuss about:- Introduction to Hardenability of Steels 2. Hardness and Hardenability of Steels 3. Measurement 4. Quantitisation 5. Calculation of Hardenability of Steel from Composition and Grain Size.
Contents:
- Introduction to Hardenability of Steels
- Hardness and Hardenability of Steels
- Measurement of Hardenability of Steels
- Quantitisation of Hardenability of Steels
- Calculation of Hardenability of Steel from Composition and Grain Size
1. Introduction to Hardenability of Steels:
A steel transforms from austenite to martensite, if its cooling rate is more than, or equal to the critical cooling rate, the surface hardness then depends principally on the carbon content of the steel, provided,
ADVERTISEMENTS:
(i) All the carbon was in solution in austenite, resulting in martensite of same carbon content.
(ii) Critical cooling rate is attained.
(iii) No austenite is retained.
(iv) There is no auto-tempering of martensite (during cooling).
ADVERTISEMENTS:
If cooling rate is lower than critical cooling rate, the amount of martensite formed is reduced, or it may not form at all, the overall hardness of the steel is reduced. Fig. 4.1 illustrates the relationship between hardness, carbon content of the steel, and the amount of martensite formed. Even the hardness of 50% martensitic structure mainly depends on the carbon content.
A small sample of steel may be quenched in water to obtain 100% martensite. However, when the steel of appreciable size is quenched, the cooling rate at the surface may be fast but decreases towards its centre. This difference in cooling rates increases with the severity of quench (i.e., with increased rate of heat removal) and section thickness. Thus, ≈ 100% martensite may form only up to a certain depth from the surface (up to which critical cooling rate is attained), and softer products may form inside.
With the increase of the cooling rate, austenite undergoes transformation to various products- ferrite, pearlite, upper bainite, lower bainite or martensite. Fig. 4.2 illustrates schematically, CCT curves of three different steels cooled with same cooling rates at the surface and at the centre of same size of part of each steel, quenched in the same coolant. The microstructures formed (and thus, the hardness) in the surface and at the centre are different.
For example, the centre has 100% pearlite in steel in (a), bainite and martensite in steel in (b), and 100% martensite in steel in (c). This illustrates that the amount of martensite (thus the resulting hardness) may vary considerably throughout a given cross-section, or between identical cross-sections made from different steels (as illustrated in Fig. 4.3). Thus, it is obvious that the depth of hardness-penetration (on hardening) is a characteristic property of the steel, and is called hardenability.
Hardenability—the susceptibility to hardening by rapid cooling is defined as the property of the steel which determines the depth and distribution of hardness produced by quenching, or as the capacity of steel to develop a desired degree of hardness usually measured in terms of depth of penetration.
It is also defined on the basis of the microstructure produced by quenching, such as by the depth to which austenite transforms to 50% martensite + 50% other transformation products, or to 90% martensite, depending on the criterion used for measuring hardenability.
ADVERTISEMENTS:
Carbon steels, except in small sections, normally harden to a depth slightly below its surface, and are called shallow hardening (D1 < 1″)- Alloy steels (like steel A, or even B in Fig. 4.3) can harden uniformly through its cross-section, or to greater depths, and are called deep-hardening (D1 > 6″).
2. Hardness and Hardenability of Steels:
Hardness is the resistance to indentation. The hardness of a quenched steel, generally refers to hardness of 100% martensitic structure, and is basically a function of the carbon content even in most alloy steels. Hardenability refers to the depth of hardening on quenching.
A steel quenched to 100% martensite up to its centre may have a lower hardness (right from the surface to the centre), but still has higher hardenability as compared to a steel having higher surface hardness due to quenching to 100% martensite there, but lower hardenability as illustrated in Fig. 4.4.
Hardenability is dependent on:
(i) Amount of alloying elements
(ii) Amount of carbon content
(iii) Grain size
(iv) Geometry and size of the sample
(v) Nature of coolant
(vi) Criterion of hardenability
(vii) Austenitising temperature and time.
For quantitative evaluation of hardenability, factors (iv) and (v) have to be standardised, so that hardenability of the steels could be compared and specified. Thus, factor (iv) is taken cared by evaluating hardenability in terms of diameter (of hardening), and factor (v) is taken cared by evaluating this diameter with reference to an hypothetical coolant called ‘ideal quench’.
3. Measurement of Hardenability of Steels:
Universally agreed criterion of hardenability is the depth which contains 50% martensite and 50% other products. Any other criterion (such as 90% martensite), if used, has to be normally mentioned in the specifications. The 50% criterion shall be used here too to demarcate between hardened from the unhardened core of the bar. This 50% criterion has been universally adopted because it is easily delineated in practice and is reproducible.
The reasons for this selection are:
1. A cylindrical hardenable steel sample of about 5 cm diameter on water quenching and fracturing at half its length (steel should not harden thoroughly), the fractured surface has two distinct appearances as schematically illustrated in Fig. 4.5 (a). The outer case having predominantly martensitic structure has brittle fracture (very smooth or faceted, intergranular fracture).
The inner core having softer non-martensitic products shows ductile fracture (rough and transgranular). The 50% martensitic structure forms the transition boundary of these two types of fractures, easily distinguishable by visual examination.
2. If the fractured surface is polished and etched with nital, and then macroscopically examined, the outer surface has white colour as martensite etches white, while the inner case shows black colour as pearlite etches black. The 50% martensitic structure is the transition of this change in colour, which is easily detectable. Detection of martensite at levels above 50% in microstructure would be much more difficult.
3. If hardness is measured across the cross-section along a diameter and results plotted to get a hardness contour such as illustrated in Fig. 4.5 (b). The hardness is high and is almost constant initially (meaning almost 100% martensite), and then hardness decreases at an increasing rate, passes an inflection point, and then decreases slowly. This steep fall or point of inflection corresponds to the same 50% martensitic structure.
Thus, at 50% martensite, the change in etching colour, nature of fracture as well as abrupt change of hardness takes place, and can be easily detected, and thus, is used as a criterion for- hardening, or depth of hardening for measuring hardenability (though 50% martensitic structure is not necessarily of greatest metallurgical and engineering interest).
Fundamental Considerations:
Basically, the amount of martensite, or the hardness at a particular depth below the surface of a particular quenched steel is a function of the cooling rate at that point. The cooling rate at any point during quenching is generally not constant, but varies with the temperature as illustrated in Fig. 4.6.
Grossman suggested then that time required to cool from the austenitising temperature to an arbitrary chosen 410°C (as this temperature is approximately half-way between the austenitising and room temperature, is thus designated as half time) may be used as a suitable measure of cooling rate, that is, in a steel section, points with similar half cooling time have similar microstructure, and thus, the similar resultant hardness.
The hardness at various depths can be plotted as a function of the cooling time as illustrated in Fig. 4.7. This is a fundamental curve, and is not related to the size of the bar, or the severity of quench. The hardenability in this case may be defined in terms of critical cooling time at which the inflection in the hardness occurs, i.e., 50% martensite is formed. But, it is a time consuming method.
4. Quantitisation of Hardenability in Steels:
i. Critical Diameter:
It is more convenient and useful to express hardenability as the depth of penetration of hardness. One common method is to find out the hardness distribution across the section on quenching round sections ranging from 12.7 mm to 127mm in diameter as illustrated in Fig. 4.8.
It illustrates the importance and effect of the mass of the steel, and illustrates:
(i) Maximum hardness is obtained on the surface of bar of small diameters.
(ii) Even in small diameter 12.7 mm, hardness in the centre drops significantly.
(iii) With increasing diameter, the surface hardness of the steel drops significantly.
(iv) As the diameter of bar increases, the centre hardness continues to drop.
If a plot is made of centre hardness vs specimen diameter, there is most rapid fall in centre hardness, when it has 50% martensitic structure. The position of the steepest fall in hardness, or point of inflection is the critical diameter, and is 1.25″ in oil, and 1.83″ in water in Fig. 4.9. Critical diameter, Dc, of a steel under a given quenching condition is defined as the diameter of the cylindrical bar which hardens up to centre, i.e., has 50% martensitic structure at the centre.
It can also be determined as the diameter of cylindrical bar at the centre of which nature of fracture just changes from brittle to ductile under given quenching conditions. Critical diameter of a steel depends on the efficiency of quenching medium, and increases as the rate of heat removal from the specimen increases, i.e., with the severity of quench.
Fig. 4.9 illustrates this fact as, Dc in water is 1.83″, but in oil is 1.25″:
Thus, when hardenability is expressed in terms of critical diameter, it is necessary to specify the severity of quench. Ideally, the hardenability of a steel should be expressed in terms of a common ideal quench. It is now possible, as a method exists, that Dc obtained under a standard quenching condition may be transferred to a different quenching condition, which is being used in a given heat treatment operation, or to an ideal quench.
ii. Severity of Quench:
The effectiveness of a given cooling medium is specified by the value of a parameter called its ‘severity of quench’, i.e., its quenching power, and is designated by the symbol ‘H’, which is also called coefficient of surface heat transfer. ‘H’ is defined as the rate of heat removal by the quenching medium from an object.
The cooling rate of heated (austenitised) steel, when quenched in a coolant depends on:
(i) Ability of the heat to diffuse from the interior to the surface of the steel, i.e. on thermal conductivity, volume specific heat of the steel, and thus, its thermal diffusivity.
Thermal diffusivity, (a) is defined as:
and it increases as does thermal conductivity of the steels with the decrease of temperature.
(ii) Ability of the quenching medium to remove heat from the surface of the steel.
Newton’s law of cooling states that the rate of heat transfer is proportional to the difference in temperature of the surface of steel (Ts) and the temperature of the coolant (Tc), and thus-
Q = S(TC – Ts)
where,
Q is the rate of heat transfer per unit surface area,
S is the proportionality constant, and is the heat transfer coefficient of the metal-coolant interface.
‘H’, the severity of quench is employed to include the properties of steel as well as the rate of heat transfer, and is given by-
H = S/K
where,
H is the severity of quench and has units of (length)-1, and is not constant.
S is the proportionality constant, and
K is the thermal conductivity is J/m.s.°C.
Grossmann assumed (a), K and H to be constant over a range of temperatures. The microstructure and hardness at any point are function of the half cooling time only. He made calculations based on unsteady state of heat transfer to come out with very satisfactorily workable solution to the problem of comparing the hardening responses under different quenching conditions. By knowing H, the cooling rates in the interior of round of steel of diameter D, can be calculated.
The results of such calculations are presented as plots of Du/D against log (H x D), and is summarised in Fig. 4.10. Here, Du represents the diameter of the circular area in the centre of the round bar, where the cooling rates are less than an arbitrary specified cooling rate, or such as with 50% or less martensite as illustrated in Fig. 4.11.
The curves in Fig. 4.10 correspond to many possible values of ‘H’ and the specified cooling rates:
This fig. can be used for:
(a) Experimental Evaluation of ‘H’ of a Coolant:
Round bars of a hardenable steel, for example, SAE 3140 steel, of various diameters are quenched in the coolant after austenitisation. For example, Fig. 4.12 (a) illustrates results in two coolants-one in oil and the other water. The rounds normally have a soft core and a hard case. The soft core containing less than 50% martensite can be clearly seen on an etched cross-section, and is measured as Du for each D.
The experimental values of Du/D are plotted against log D on a transparent paper in the same scale as the Fig. 4.10 as illustrated in Fig. 4.12 (b). The steeper curve is for oil-quenching due to reduced ability of oil quenching to produce hardening in larger sections. This also explains the fact that large diameter bars, quenched in milder quenching medium do not even form hardened rim at the surface Fig. 4.12 (a).
It also proves that greater is the critical diameter of a steel of given hardenability if greater is the severity of quench (Fig. 4.12 b), or greater is the critical diameter in same severity of quench, if hardenability of the steel is higher. Thus, two different steels of different hardenabilities may have the same critical diameter in two different quenching media, but their unhardened core at all diameters higher than the critical diameter would be always different.
The experimental curves (Fig. 4.12 b) are compared one by one with the theoretical curves of Fig. 4.10, and the best matching theoretical curve is determined. The coordinate of the point on the theoretical curve on the x-axis gives (H x D). As D is known, ‘H’ can be calculated.
For example, point x in Fig. 4.12 (b) corresponds to a cylindrical diameter of 1.83 inch and results in a value of 2.6 inch as (H x D), when the best curve has been matched. Thus, of this water bath usually a large number of points are considered on the curve and then, an average value of ‘H’ is obtained. Table 4.1 gives ‘H’ values of a number of coolants in different state of agitation.
(b) Evaluation of Critical Diameter:
The severity of quench, H of a coolant is known. First, find Du experimentally for a particular D of the given steel. Calculate Du/D and H x D for this diameter, and project these points from y and v-axes in Fig. 4.10 to get the points of intersection P at some curve in that Fig. and the curve is followed to Du/D = 0, that is, H x D critical is obtained for Du/D – 0 as illustrated in Fig. 4.13, and thus.
is the critical diameter.
iii. Ideal Critical Diameter:
The critical diameter, D, of a given steel, changes considerably depending on the quenchant (H value) as illustrated in Fig. 4.12 (b). Due to this fact, one cannot compare the hardenabilities of two steels on the basis of their critical diameters if quenchants are different. To avoid this limitation, it is necessary to obtain a measure of hardenability independent of the quenching conditions.
This was achieved by Grossman in the concept of ‘ideal critical diameter’. The ideal critical diameter Dl, sometimes also called as ideal diameter, is defined as the diameter for which the unhardened core just disappears if the bar is subjected to an ideal quench, or is the diameter of bar of steel which when quenched in an ideal quench would have 50% martensite in its centre. In an ideal quench, the severity of quench ‘H’ is infinite. Dl gives value of inherent hardenability of the steel.
It may appear that the ideal critical diameter is infinite with infinite severity of quenching. This notion is false, since all that the ideal quench does is to cool the surface of the bar instantly to the temperature of the quenching medium and holds the surface temperature constant, heat being extracted from the surface as rapidly as it is conducted to the surface from the interior i.e., the heat transfer coefficient of metal/coolant interface, S, is infinite.
The interior of the bar still cools at a finite rate since the thermal diffusivity is finite. This ideal quench would produce the maximum possible depth of hardening, and the critical diameter would be the largest attainable, called the ideal critical diameter, Dl.
We can write:
H = ∝, then, Dc = Dl
H < ∝ then, Dc < Dl
Ideal quench is a hypothetical quench and cannot be attained in practice. However, quenching in highly agitated water, or brine is very close to ideal quench (see Fig. 4.14 and 4.15).
Fig. 4.14 and 4.15 have also been obtained by heat flow analysis, and that equivalent microstructure is obtained from the equivalent half cooling time in a steel. These figures can be used to obtain Dl if Dc of a steel is known, or experimentally found in a known ‘H’ value of a coolant. Dc can also be found out for any other severity of quench than for which it was obtained.
It may be necessary to obtain the case with greater percentage of martensite than 50%, such as 90% or more, i.e. the criterion of hardenability may be 90% or more martensite. Dl corresponding to higher percentage of martensite can still be read from Fig. 4.14 and 4.15, if ‘H’ is known and the experimental value of Dc corresponding to the higher percentage of martensite is found. Fig. 4.16 gives average relationship between Dl and Dl (for higher percentage of martensite) for medium carbon and low alloy steels. This figure could be used to get Dl for higher percentage of martensite from normal Dl (i.e. with 50% martensite).
5. Calculation of Hardenability of Steel from Composition and Grain Size:
Hardenability, DI, can be calculated from the composition and grain size of low-alloy medium carbon steels. Originally, Grossman proposed the multiplying factors for the alloying elements, and had been used to obtain DI. Moser and Legat have given a more precise relationship between grain size and the carbon content of the steels, and have also, revised the multiplying factors and is being used here to calculate, DI.
Steps in calculation of DI are:
1. First step is to get base hardenability, DIC based on carbon content and grain size of the steel with the help of diagram 4.23. This diagram in useful for steels with carbon in the range 0.2-1.0% and grain size in range of ASTM 4 to 8.
2. The base hardenability, DIC is modified by taking into account the effect of additional alloying elements. This is done by the use of multiplying factors, which have been experimentally determined for common alloying elements as illustrate in Fig. 4.24.
Thus:
or, Hardenability, DI can be calculated from the following empirical relationship based on the results of Moser and Legat.
DI = DIC x 2.21(%Mn) x 1.40 (% Si) x 2.13(%Cr) x 3.275 (% Mo) x 1.47 (% Ni)
This relationship appears to be more accurate in practice than the one put forward by Grossman. Fig. 4.24 gives values of average multiplying factors. No account has been taken of all the possible alloy interactions that exist. Thus, the calculated value of hardenability represents only a first approximation and should be used as such. Table 4.5 gives Grossman hardenability multiplying factors. Multiplying factor in unity for element having no effect. Sulphur and phosphorous are impurities and have unity factor.
