correlation of steel hardness and thermal conductivity 1

[dalo]

I need information about the correlation of hardness and the thermal conductivity of steel. I heard that there exist some correlations, which are used in non-destructive hardness measurement, but I couldn't find more detailed information on that topic. If someone knows more about this, maybe even an empirical equation, please let me know.

 

[Goahead]

No correlation known between Thermal Conductivity - a Physical Property, and Hardness - a Mechanical Property dependent on microstructure.

http://www.welding-advisers.com/

 

[kenvlach]

Correlations do indeed exist between hardness and thermal conductivity. However, the only practical ones with which I am familiar are for copper, brass and aluminum alloys. For these metals, the electrical conductivity (directly related to thermal conductivity by the Wiedemann-Franz law) can be used to follow annealing and age-hardening processes.

In principle, a correlation could be developed for steels. However, it would need to be done on a case-by-case basis per specific steel alloy, since conductivity varies with alloying, heat treatment and cold working. A composite model would be needed, as conductivity differs for the austenite, ferrite, martensite and pearlite microstructures. A plot of thermal conductivity vs. temperature for each of these structures is given on p. 13.7 in Physical Metallurgy Handbook (2003), reproduced from Quenching Flow of Heat in Metals, J. B. Austin, Amer. Soc. Metals (1942). The data are for an 18-8 austenite, a ‘low-metalloid’ &[ignore]alpha[/ignore];-Fe ferrite, annealed 0.83%C steel for pearlite, and quenched 0.83% C steel for martensite. Further testing could reveal the ferrite and cementite components of the pearlite conductivity, with corrections for interfaces.

 

[Goahead]

Thank you kenvlach.
My previous answer above should be reworded as follows:
For any practical purpose there is no correlation available for steels between Thermal Conductivity and Hardness.

http://www.welding-advisers.com/

 

[EnglishMuffin]

An old British reference that I have - "Physical Constants of some Commercial Steels at Elevated Temperatures" - British Iron & Steel Research Association 1953 - lists the following data :

Steel Carbon Content Thermal Conductivity
(@ 100 degC)
(%) g.cal/cm/sec/degC

.06 .144
.08 .138
.23 .122
.4 .121
.8 .115
1.2 .107

This indicates that the correlation may not necessarily lie purely with hardness per se, but may be at least partially related to alloy content. The reference also lists thermal conductivities for low and high alloy steels, and for these the thermal conductivity varies widely with alloy type and content. For example, a 13% manganese steel at 100 degC has a thermal conductivity of only .035 g.cal/cm/sec/degC.

 

[kenvlach]

The data given by EnglishMuffin are included in ‘Table 14.11 PHYSICAL PROPERTIES OF STEELS,’ PP. 14-27 to 14-41 in Smithells Metals Reference Book, 7th Edn. (1992). Thermal conductivity units are W/m/^oK ( = 418.4 x cal/cm/sec/deg^oC).

The carbon steel data listed above (%C as stated, Mn contents range from 0.31 to 0.64%) are for the annealed state. For each datum for other steels, in addition to composition, the heat treat state is specified:
Oil quenched from 850^oC + Tempered 1 hr at 600^oC + oil quenched, sometimes simply Q + T, annealed, normalized, normalized + tempered, normalized + stress-relieved at 600^oC, solution treated [for PH SS], solution treated + aged, etc.

Thermal conductivity can be greatly decreased by alloying (especially, with a change in structure to austenite) and slightly by hardening. The lack of systematic data suggests that few studies have been conducted of thermal conductivity per se. The lack of studies suggests that funding authorities did not expect worthwhile correlations to be obtainable. However, the existing data could be studied by matrix statistics with modern computers. With further input of the volume % of each microstructure for each datum, the thermal conductivity could be predicted as a function of composition, heat treatment and temperature. Sufficient hardness data exist such that some further work would yield the desired thermal conductivity-hardness correlation. As much hardness data exist, and as hardness is easily measured, this would perhaps be more useful for calculation of thermal conductivity.

 

[dalo]

First of all I want to thank all of you for your detailed answers. EnglishMuffin and kenvlach, thanks for the references. I will compare them with my available data.

But at least I think I can now specify my question a little bit more. I have a C-Mn steel (22MnB5), and I have temperaturedependent data for the thermal conductivity for the ferritic and the austenitic phase. Since I want to calculate the temperature distributin in a quenching process I need the data for the martensitic phase too. I was told that due to the increasing hardness in quenching the thermal conductivity will decrease.
So my first guess was to use the data for the ferritic phase and decrease it by a factor. But I do not know how much I should decrease it. I thought about 30%, but without any guarantee.
As an additional information: the hardness is increased from approximately 165 HV10 to 475.
Hope that one of you can help me.

 

[kenvlach]

Some thermal conductivity data at 200^oF read from the figure mentioned in my Dec. 9 post. Units are cal/cm/s/^oC:
Annealed 0.83%C steel (pearlite): 0.11
Martensite 0.83%C: 0.10
Ferrite: 0.16
Austenite (18-8 SS) 0.037

The conductivities of the ferrite and pearlite phases decrease linearly with increasing T, that of martensite is nearly constant (slight decrease with increasing T), and that of austenite increases linearly with T. Also, the ferrite, pearlite and austenite (albeit 18-8 SS) plots converge to about 0.063 at 1600^oF. The high-temperature &[ignore]gamma[/ignore];-Fe plot is nearly an extension of that for 18-8 austenite, so maybe structure is more important than alloying for the fcc phase.

Your estimate of a 30% decrease going from the ferrite to martensite appears OK at 400^oF. The decrease is greater at lower T, less at higher T.

What are the composition & equivalent alloy designations for 22MnB5? It’s not in my 1992 Smithells, which uses designations En 8 and 060A42 for a 0.64%Mn-0.42% wrought steel, BS 1617A for a 0.35%Mn-0.11%C cast steel, etc.

 

[dalo]

Hi kenvlach,

thanks for your helpfulness. Here a little bit more information about the 22MnB5. The Materialc Number is 1.5528
Here the alloying composition from the steelmaker (ARCELOR):

min[%mass] max[%mass]
C 0.20 0.25
Mn 1.10 1.40
Si 0.15 0.35
P 0.025
S 0.008
Cr 0.15 0.35
B 0.002 0.005
Ti 0.02 0.05

The strength of the ferrit is ~600MPa and that of the martensit is ~1500MPa.