July 26, 2008
Influence of Al on Strength and Impact Behaviour of Hot Rolled Plain C-Mn Steels
By Mintz, B Gunawardana, W D; Su, H
The influence of Al on the strength and impact properties of hot rolled plain C-Mn low S steels has been examined at two C levels, 0.02 and 0.1% and two N levels, 0.001 and 0.004-0.005% (where percentage is to be taken as wt-%). It has been found that Al is an excellent element to add to hot rolled steels, an addition of 0.2% to a plain 0.1C-1.4Mn steel with 0.004-0.005%N resulting in an impact transition temperature of -95[degrees]C at a strength level of ~300 MPa. At this Al level, Al removes the N from solution but also importantly refines the grain boundary carbides and to a lesser degree the grain size resulting in a 40[degrees]C lowering of the impact transition temperature in a coarse grained steel (15-20 mum; d^sup -1/2^, 7-8 mm^sup -1/2^). Strength is not affected by this Al addition. The decrease in strength due to the removal of nitrogen from solution as AlN is balanced by the solid solution hardening from the Al in solution in the iron lattice, (a 1%Al addition increasing the strength by ~70 MPa), and the small grain refinement that occurs from AlN precipitation. Increasing the Al level to 1% in the 0.1%C steels, results in the formation of martensite with consequent deterioration in impact behaviour. When the N level is reduced to very low levels, 0.001%N, no grain refinement in the gamma occurs in the 0.02% and 0.1%C steels at the 0.2%Al level. The solid solution hardening effect of Al now dominates the properties with the impact transition temperature remaining constant due to Al refining the grain boundary carbides. Introduction
Recent work1,2 has shown that some very good impact properties can be obtained in hot rolled steels by adding 0-2%Al to a low S (0.002%) C-Mn steel, (all % in this paper are wt-% unless stated otherwise). For a 0.1C-0.3Si-1.4Mn steel, an impact transition temperature (ITT) of -95[degrees]C was obtained at a strength level of ~300 MPa. The steel had a coarse ferrite grain size of ~16 mum, d^sup -1/2^ of 8.0 mm^sup -1/2^, where d is the mean linear intercept. The present examination is a continuation of this work to establish the cause of this good impact behaviour and as such it will be looking closely at the impact and tensile results from the previous work1'2 as well as the results obtained from the present work.
The steels examined in the present exercise were 50 kg laboratory vacuum melts which had been hot rolled to 13 mm thick plate, finish rolling in the temperature range 1000-950[degrees]C, the average cooling rate in the temperature range 800-400[degrees]C being 40 K min^sup -1^.
All the compositions which have been studied are given in Table 1. Steels from the previous examinations2 are asterisked and were made as 20 kg melts but given otherwise similar hot rolling schedules.
It can be seen that the new steels 1-6 in Table 1 all have very low N levels (0.001%) whereas the previously examined steels,2 steels 7-14 contained about 0.003-0.005%N. In addition to examining the influence of Al on steels with 0.1%C, some steels with 0.02%C were also included in this study, steels 1-3. Since in the previous work, the 0.1%C steels with >/=1%Al resulted in the formation of martensite, Al levels were restricted in the present work to the ~0.5%A1 level in the 0.1%C steels and 1%A1 in the 0.02%C steels, levels at which martensite was not present.
The previous work2 had shown that a 0.2%Al addition is near the optimum level for the best impact behaviour and since a silicon addition improves not only the strength but also gives a small improvement in impact behaviour, all the new steels except one, steel 5, contained a 0.3%Si addition. Steel 5 was chosen to explore whether a higher Si addition, 0.56%, would give an even better combination of strength and impact behaviour.
After hot rolling, standard Charpy V notch impact samples were machined from each plate parallel to the rolling direction and duplicate tensile samples were machined transversely to the rolling direction.
The impact transition curves were obtained for all the hot rolled plates and the tensile properties determined. Tensile samples of gauge length 44 mm and diameter 6.3 mm were tested on an Instron tensile machine at a crosshead speed of 0.02 cm min^sup -1^.
1 Influence of Al on strength and impact behaviour of 0.1%C steel with low Si and Si containing C-Mn steels having N levels in range 0.004-0.005%
2 Impact transition curves at 0.02%, 0.48% and 0.94%Al levels for 0.02%C steels with low N levels
Grain size measurements were made using the mean linear intercept method (1000 intercepts being taken) and pearlite volume fraction obtained by point counting (a total of 700 counts being made). The grain boundary carbide thickness was measured using the method of Mintz et al.3
The results from the previous work2 for low Si and Si containing steels are summarised in Fig. 1, where it can be seen that adding 0.2%Al causes the ITT to decrease by almost 40[degrees]C for the Si killed steel. Yield strength however, is hardly influenced by this Al addition. It can also be seen from this figure that the addition of Si at the ~0.3% level is beneficial, since as well as marginally improving the impact behaviour it also significantly increases the strength by ~25 MPa.
Table 1 Compositions of Al containing steels, wt-%
Table 2 Mechanical properties of steels examined
3 Impact transition curves at 0.02%, 0.21% and 0.46%Al levels for 0.1%C steels with low N levels
The impact transition curves for the new steels are given in Figs. 2 and 3 for the 0.02% and 0.1%C steels respectively. The tensile and impact results are summarised for all the steels in Table 2 and their metallographic measurements together with their accuracies4 in Table 3. In Table 2, 54 J ITT is the test temperature which corresponds to an energy reading of 54 J in a Charpy test.
4 Influence of Al on grain boundary thickness for all steels examined
Relative error of measurement are:4 ferrite grain size, 2%; pearlite volume fraction, 2-2.7%; grain boundary carbide thickness, 7%.
It can be seen from Table 3, that in the previous examination,2 steels 7-10, adding ~0.2%Al to the 0.1%C steel with 0.004-0.005%N leads to a small grain refinement of ~0.5 mm^sup -1/2^, (steels 7 and 8 having 0.02%Al and steels 9 and 10 with ~0.2%A1 ).
In contrast for the presently examined 0.1%C, low N steels, adding Al results in a coarsening of the grain size (steels 4, 5 and 6 having d^sup -1/2^ of 7 mm^sup -1/2^ (20 mum), 6.8 mm^sup -1/2^ (22 mum) and 6.5 mm^sup -1/2^ (24 mum) respectively).
For the low C steel, which was examined only at a low N content of 0.001%, adding 0.48%Al led to no change in grain size (steels 1 and 2) while at the 1%Al level some grain refinement took place (Table 3).
The grain boundary carbide thickness measurements are plotted against Al content for all the steels 1-14 in Fig 4. The behaviour is similar for all the investigations, the carbide thickness markedly refining on adding 0.2%A1 with little further change with increasing Al content. The change from a coarser to a finer carbide distribution is shown in Fig. 5a and b for the ~0.02%C at the 0.02 and 0.94%Al steel respectively and Fig. 6a and b for the 0.1%C steels at the 0.02 and 0.19%Al levels respectively.
Table 3 Microstructural measurements for steels examined
5 Grain boundary carbides in a steel 1, 0.028%C with 0.02%Al and b steel 3, 0.022%C with 0.94%Al
Pearlite volume fraction also varied in these steels from ~3% for the low C steels, (steels 1-3) to ~12.5% for the 0.1%C steels in the previous examination, (steels 7-10) to ~19% (steels 4-6) for the presently examined steels (Table 3).
6 Grain boundary carbides in a steel 8, 0.098%C with 0.02%Al and b steel 10, 0.095%C with 0.19%Al
Interpretation of results for previously examined 0.1%C, 0.004- 0.005%N steels (steels 7-10)
When discussing the results it is always necessary for the impact behaviour to be considered in relation to the strength.
The factors that give rise to improved impact behaviour in these higher N steels are:
(i) removal of N from solution by Al
(ii) refinement of grain boundary carbides
(iii) refinement of the ferrite grain size.
Aluminium provided there is little nitrogen present has been shown to be a potent solid solution hardener5-7 in much the same way as Si. For pure Fe-Al alloys an increase of 1%Al results in an increase of ~45 MPa, but recent work has established that for steel, a greater increase of 70 MPa occurs.8
In Fig. 1 it can be seen that the yield strength does not change with Al content but the ITT improves as the Al addition increases to ~0.2% and then deteriorates at the 1%Al level.
The influence of free nitrogen on yield strength has been reviewed by Morrison et al and 0.001%N will raise the yield strength by ~5 MPa.9 Removing 0.004%N from solution as AlN by the addition of 0.2%Al should therefore lead to a 20 MPa decrease in strength. However, the solid solution hardening effect of a 0.2
In the case of the impact behaviour for 54 J ITT, the multiplying factor in [degrees]C for removing N from solution in hot rolled Si killed plain C-Mn steels has been found to be 2750 per 1%N so that removal of ~0.004%N by the addition of 0.2%Al would give rise to a 10[degrees]C decrease in the ITT.10 The refinement in the grain boundary carbide thickness (0.31 to 0.2 mum) that also occurs would account for a further 20[degrees]C decrease in the ITT.3 This with the small grain refinement, which would lead to ~5[degrees]C fall in the ITT,3 would account for observed decrease of ~40[degrees]C in ITT (Fig. 1).
Thus, the behaviour shown in Fig. 1 with regards to the impact and strength changes with Al content up to 0.2% is to a large degree explained.
Al levels >/=1% lead to martensite being formed, Fig. 7 and the interpretation of changes in the structure/ property relationships then becomes very difficult, as it has been shown that martensite causes preyielding so that the yield strengths are reduced to what normally would be expected for ferrite/pearlite structures.8
Presently examined steels
Low C steels (steels 1-3)
It can be seen from Fig. 2 that for the 0.02%C steel, Al additions have little influence on the impact behaviour. However, a marked strengthening of 50 MPa was noted on increasing the Al from 0.02 to 0.94% as can be seen from Fig. 8, which summarises the changes in the 54 J ITT and the yield strength with Al addition. In the presently examined steels (steels 1-6) only a small amount of N is present so the benefit to impact behaviour from its removal is very limited and hence the solid solution hardening effect of Al dominates the changes in properties. Also, because of only a very small amount, 0.001%N, AlN precipitation would be restricted preventing any grain refinement of the gamma. Solid solution hardening by Al would therefore largely account for the 50 MPa increase in strength that is observed as the Al content increases.
The problem then is explaining why the impact behaviour does not change with Al content. An increase in strength of 10 MPa from precipitation or solid solution hardening has been shown to correspond to an increase in ITT of 5[degrees]C.3 Such an increase in strength as 50 MPa without any grain refinement would normally therefore give rise to a ~25[degrees]C rise in ITT.3
8 Ifluence of Al on yield strength and 54 J ITT ([degrees]C) of 0.02%C steels with low N levels (steels 1-3)
9 Yield strength standardised to 0.3%Si and 54 J impact transition temperature against Al content for the 0.1%C low N steels (steels 4-6)
Examination of Table 3 and Figs. 4 and 5, shows that adding Al again refines the grain boundary carbides. The grain boundary carbide thickness refines from 0.32 to 0.19-0.17 mum on increasing the Al level from 0.02% to 0.94%Al additions (the data points are also plotted in Fig. 4 for the 0.1%C steels and behaviour can be seen to be similar). This refinement of the grain boundary carbides would correspond from previous work3 to a decrease in the ITT of 20- 30[degrees]C. Thus, the detrimental influence of solid solution hardening by Al on the ITT is balanced by the refinement in the grain boundary carbides that occurs on adding Al. Although no grain refinement occurred on increasing the Al from 0.02 to 0.48%, there was a small refinement in grain size (~0.8 mm^sup -1/2^) on increasing the Al level from 0.48 to 0.94% which may account for a part of this increase in strength. This grain refinement suggests that the transformation temperature has been raised so much by Al that finish rolling is taking place in the ferrite rather than austenite.11
Steels containing 0.1%C (steels 4-6)
Examination of Fig. 3 indicates that Al appears to have little or only a small influence on the impact behaviour of the 0.1%C steels and this time there is no significant influence of Al on strength (Table 3). However, it has to be noted that as well as Al varying in these steels, steel 5 also has a higher Si level. If one standardises the strength for all the steels to the same Si level, 0.3%Si, using a solid solution hardening multiplying factor of 1%Si increases the strength by 80 MPa,12 then again it is apparent that increasing the Al level increases the strength without influencing impact behaviour in a similar way to that noted for the low C steels (Fig. 9). In the case of the impact behaviour, refinement in the grain boundary carbides again occurs on adding Al (Figs. 4 and 6). This grain refinement would lead to a 20-30[degrees]C decrease in ITT but is balanced by the solid solution hardening influence of Al and the grain coarsening that occurs, d^sup -1/2^, 7 to 6.5 mm^sup - 1/2^ (20-24 mum) as the Al increases from 0.02 to 0.46%. In these steels because of their low N content there is no refinement of the gamma grain size and as Al raises the transformation temperatures the ferrite comes out at higher temperatures and so will be coarser the greater the Al addition.
Whereas, grain size is influenced as to the presence or absence of N this is not so for the grain boundary carbide thickness. Throughout this work it has been found that Al refines the grain boundary carbides. Aluminium additions to steel have been shown to decrease the activity coefficient of C in ferrite as well as increase the activity coefficient of C in cementite.13 This results in a higher C solubility in ferrite and a lower C solubility in cementite both effects retarding the cementite precipitation and this may be the reason for the refinement in carbide thickness that has been noted.
Influence of higher Si level
Unfortunately, because of the difference in N levels between the past and previous work it is difficult to assess the influence of the higher Si level on properties for steels having the same Al addition, i.e. steel 5, having 0.56%Si, 0.001%N can not be compared directly with steel 10, 0.29%Si, 0.004%N. However, if one compares steels 5 and 6, having the same N content, it would seem that the additional amount of Si is increasing strength without influencing impact behaviour. However, further work is required to confirm this before recommending a higher Si level.
Comparison of present work with previous examination
It is clear that at the 0.1%C level there are optical metallurgical differences between the present and past work (steels 1-6 and steels 7-14 respectively).
The previously examined steels had less pearlite (~7%) and finer grain sizes (Table 3). For steels with
Irvine et al.14 has shown that a multiplying factor of 2.2 per percentage pearlite is needed to account for the impact transition temperature in as rolled steels and such a multiplying factor would explain 15[degrees]C of the observed change in impact behaviour. It is therefore not surprising that the impact behaviour was better than in the previously examined steels, a 54 J ITT of - 95[degrees]C, steel 10, compared to about -70[degrees]C for steels 5 and 6.
Mintz et al. have also shown that increasing the pearlite volume fraction leads to an increase in yield strength.15 However, in these higher N steels, steels 9 and 10, the reduction in yield strength, as a result of a reduced pearlite content compared to that present in the 0.001%N steels, is compensated for by the additional grain refinement (Table 3). The most likely reason for this refinement in grain size and reduced pearlite content leading to better impact behaviour is their higher N level allowing AlN precipitation to occur in the austenite thus refining the austenite grain size. This grain refinement in hot rolled C-Mn-Al steels has also been observed by Smith et al.16 at the 0.1%Al level which they also related to AlN precipitating out during the hot rolling.
Commercial implications and future work
The present work has shown that it is possible to obtain a strength level of 295 MPa with a 54 J ITT of -95[degrees]C in a hot rolled, simple C-Mn steel with a 0.2%Al addition.
To achieve these properties it is necessary to have ~0.005%N present so that AlN can refine the y grain size and hence to a small degree refine the ferrite grain size.
It would be expected that higher N contents will encourage more AlN precipitation and give a more substantial refinement of the ferrite grain size leading to higher strengths and even better impact behaviour. Increasing the Si content to 0.56% may also give some further benefit to strength without influencing impact behaviour although further work is required before such a recommendation could be made.
However, the hot rolled steel will not reach the high strength levels that can be achieved on control rolling unless some precipitation hardening can be introduced into the steel.
As noted a 10 MPa increase in strength from precipitation hardening results usually in a 50C rise in ITT.3 Hence, for example, to increase the strength to 400 MPa would result for the presently examined hot rolled steel in raising the ITT to -40[degrees]C.
The most likely element to use to give this precipitation hardening and increase strength will be to add Nb and this will form the next part of the programme.
1. Al refines the grain boundary carbides thus improving the impact behaviour.
2. In a 0.003-0.005%N steel, an Al addition of 0.2%Al improves the impact behaviour of hot rolled steels without it materially influencing strength.
3. The greater volume fraction of pearlite and lack of grain refinement by AlN during hot rolling in the present examination on low N steels gave rise to worse impact behaviour than was found in the previous hot rolled examination.
4. In the low N steels the N level is too low to have a significant influence on strength or impact behaviour. The strength therefore increases with Al addition. Since Al refines the grain boundary carbides the impact behaviour is unaffected. 5. Previous work on 0.1% plain C-Mn steels2 has shown that additions of >/=1%Al introduce martensite into the structure and so should be avoided for hot rolled plates. Reducing the C content to 0.02%, leads to higher levels of Al being accommodated without martensite being formed.
6. N is required to ensure that there is sufficient Al to give some refinement to the y grain size. Higher N levels may lead to even better properties and this needs exploring.
7. To raise the strength level to a more useful level, it is suggested that Nb be added to these steels to give precipitation hardening and perhaps some further grain refinement so that hot rolled HSLA steels can be made with properties similar to those obtained by control rolling.
The authors would like to thank EPSRC grant EP/ C512839/1 for financial support.
1. B. Mintz, A. Williamson, H. Su and W. B. Morrison: Mater. Sci. Technol., 2007, 23, 63-71.
2. B. Mintz: Mater. Sci. Forum, 2003, 426-432, 1219-1224.
3. B. Mintz, W. B. Morrison and A. Jones: Met. Technol., 1979, 6, 252-260.
4. F. B. Pickering: 'The basis of quantitative metallography'; 1976, London, The Institute of Metals.
5. W. Muschenborn, D. Grzesik and W. Kupper: in 'Steel a handbook for materials research and engineering'. Vol. 1; 1992, Stahleisen/ Berlin/Dusseldorf, Springer-Verlag.
6. G. Frommeyer, E. J. Drewes and B. Engl: Revue Metall., Oct. 2000, 1245-1252.
7. J. Herrmann, G. Inden and G. Sauthoff: Acta Mater., 2003, 51, 2847-2857.
8. B. Mintz, W. D. Gunawardana and H. Su: Mater. Sci. Technol., 2008, 24, (5), 596-600.
9. W. B. Morrison, B. Mintz and R. C. Cochrane: Proc. British Steel Corp. Product Technology Conf. on 'Controlled processing of HSLA steels', September 1974, York University, 1-39.
10. B. Mintz: JISI, 1973, 211, 433-439.
11. R. C. Cochrane: Private communication, 2007.
12. B. Mintz and P. J. Turner: Met. Trans. A, 1978, 9A, 1611- 161.
13. W. C. Leslie and G. C. Rauch: Metall. Trans. A, 1978, 9A, 343.
14. K. J. Irvine, F. B. Pickering and T. Gladman: J. Iron Steel Inst., 1967, 205, 161.
15. B. Mintz, G. Peterson and A. Nassar: Ironmaking Steelmaking, 1994, 21, 215-222.
16. C. I. Smith, R. R. Preston and N. L. Richards: Met. Technol., Oct. 1978, 341.
B. Mintz*, W. D. Gunawardana and H. Su
School of Engineering and Mathematical Sciences, City University, London, UK
* Corresponding author, [email protected]
Copyright Institute of Materials May 2008
(c) 2008 Materials Science and Technology; MST. Provided by ProQuest Information and Learning. All rights Reserved.