Effect of Initial Grain Size and Shape on Textures in Sheet Steel

8/12/2019 Effect of Initial Grain Size and Shape on Textures in Sheet Steel

1/6

Textures and Microstructures, 1991, Vols 14-18, pp. 691-696Reprints available from the publisherPhotocopying permitted by license onlyC 1991 Gordon and Breach Science Publishers SAPrinted in the United Kingdom

EFFECT OF INITIAL GRAIN SIZE AND SHAPE ON TEXTURES IN SHEET STEEL

J.R. HIRSCH*, W.B. HUTCHINSON** and K. LOCKE**** ALCOA Laboratories, ** Swedish Institute for Metals Research, RWTH AachenABSTRACT

Steel samples have been prepared with a range of different equi-axed grainstr.uctures and with columnar grains having an aspect ratio of 20 1. Thesehave been col d rol Ied to different reductions and the textures examined usingODFs. Computer simulations have been carried out using various models of con-straint. Starting textures were quite weak but these had nevertheless a majorinfluence on resulting deformation textures. Grain size and shape appear toplay subordinate rol es.INTRODUCTION

It is now well established (e.g. that the rolling texture of iron or lowcarbon steel comprises two main fibres. One of these is the partial -fibrewith //RD extending from {100}


2/6


3/6

GR IN SIZE IN SHEET STEEL 69370 reduction which is in agreement with a state of homogeneous plane straincompression. The central M) zone of this specimen produced the usual flattenedand elongated structure.Initial textures were generally weak but not random as shown by the 2=45sections in Fig. 2 for materials A, C, M and Q. They may be described as

A 13 m equi-axed) random slight cubeC 39 m equi-axed) strong a-fibre {100} to {211}, 5 x RM 50 m equi-axed) strong cube//ND, {001}, 4 x RQ 50 m x 1000 m columnar random, slight Cube//ND and T-fibre, 1.5xR

.

A

2

CFig. 2 p2 45 sections of ODFs of the starting textures.

After cold rolling, the expected a- and T-fibres were developed but somedifferences existed between the different materials. Figure 3 shows 45sections after 80 cold rolling. Steels D irrespective of grain size,produced a major peak in the vicinity of {233} which was not evident ineither M or Q. This discrepancy, as well as variations within the strength ofth e a-fibre can be traced back to starting textures. For example, materials Cand M had very similiar initial microstructures but developed rather differentrolling textures due to the presence of an initial a-fibre in C Orientationdensities along the a and T fibres 80 reduction are shown in Fig. 4 a) andb) together with the position of the T-fibres in Fig. 4 c). The most perfect-ly developed //ND T-fibre texture is found in the columnar grained steel Qwhere the {111} end of the fibre is most symmetrical. In all cases theorientation density was maximum within 2 of {111} and was not located atthe Taylor orientation {11,11,8} which is 8 away from this.

Fig. 3 f2 45 sections of ODFs after 80 cold rolling reduction.


4/6

694 J.R. HIRSCH ETL20 20 70

15

zw 10-

5-

(x-fibre

< 011>//RD I < 111 >//ND Fig. 4 Orientation densities and T-fibre position after 80 rollingreduction

ANALYSISSimulations of the rolling textures have been carried out using the measuredinitial textures as starting points, assuming {110} glide and a variety ofdifferent relaxed constraints conditions as described previously. Model F isthe fully constrained Taylor condition. C is the lath model where onlye is relaxed whereas the transverse shear e is relaxed in S. Model SC isthe pancake model where both these shears are relaxed. Examples of some ofthese simulations are given for steel Q in Fig. 5 (compare Fig. 3 (d)), whilepositions of the predicted skeleton lines are shown in Fig. 6 for cases wherethe initial textures were assumed to be random. The significance of initialtextures was clearly evident when comparing the simulated textures. This is animportant point since it emphasises the difficulty of drawing conclusions aboutmechanisms of texture evolution. Even weak starting textures have a distortingeffect. Present observations supported by theoretical simulations show that the

Fig. 5 Simulated rolling textures p= 45 according to different modelsbased on the initial texture present in steel Q.


5/6

GR IN SIZE IN SH T ST L 695final texture is a strong a-fibre in cases where i) cube is present due to NDand subsequent R rotations e.g. A) or ii) cube//ND is present M) or i/i)a-fibre orientations are already present C). Excepting these cases, mostorientations rotate into the T-fibre e.g. Q). However, even when the knownstarting textures were taken into account here none of the models produced rea-listic simulations of the complete experimental textures. Closest approach wasfound for the pancake model SC or BSC where also the strain de21 is relaxed.

70

_o 60-

Fig. 6 Calculated positionsof the T-fibre accordingto different models ofconstraint relaxationref. 10).

DISCUSSIONWhile the present observations do not rule out some effect of grain size,examination of observed and simulated texture evolutions suggest that thedifferences observed for different initial grain sizes are probably mostattributable to variations in their starting textures. A previous report of acoarse grain size effect was for material processed in a similar way as forthe present steels C and The close agreement observed between all thesetextures suggests that starting texture may also have played a significant rolein the earl er work.More interesting are the observations relating to the initial columnargrain structure in material Q since these permit some interpretation of theconcepts underlying relaxed constraint theories. These rolling textures are infact closely similar to the corresponding equi-axed material M and show verywell developed //ND T-fibres, extending fully to the orientation{111} and with virtual ly no deviation away from the {111} orien-tation. They are thus typical of bcc rolling textures and show no tendencytowards the fully constrained Taylor condition This situation contrasts withobservations on aluminium where an initial columnar structure of similar naturewas used and the observed texture was of the Taylor type. However, thealuminium had an intense fibre starting texture which may have been adominant factor in those experiments. For intermediate rolling reductions- 70 ) the grain dimensions d) in steel Q can described as dN dR > nsDn a geometrical basis it is not therefore justifiable to relax consral dof the type de3 or dez3 although some freedom in delz might be possible.


6/6

696 J.R. HIRSCH ET ALThus, even though the pancake model (SC) of texture formation does bear somesimilarity to observation it is not apparently as a result of grain shape.Constraints associated with grain boundaries have been suggested as playing asignificant role in formation of the T-fibre but this is not supported by therlative insensitivity to grain size and shape observed here. Other patterns ofconstraint relaxation must be considered possibly involving cooperativeeffects within clusters of grains as has recently been discussed in connectionwith rolling textures in fcc metalst .CONCLUSIONS

Rolling textures in iron are sensitive to the starting texture of themetal even when this is quite weak. Initial grain size appears to have arelatively minor effect on development of the deformation texture. An initialcolumnar grain structure does not affect rolling texture ev ol ut io n to any majorextent which indicates that grain shape is not the principal source ofconstraint relaxation.REFERENCES1. U von Schlippenbach F Emren and K Lticke, Acta Met., 7_ 1929 (19B6)2 N Hansen, B Bay, D Juul-Jensen and T Leffers, Strength of Metals andAlloys (Ed. H McQueen et al, Pergamon, 1985) p.317.3. H.J. Bunge, Kristall und Technik 5 145 (1970).4. I.L. Dillamore and H Katoh, Met. Sci. J.,2I, 21 (1974)5 P van Houtte, ICOTOM 7 (Ed. C.M. Brakman et al, Holland, 1984) p.7.6. P van Houtte, ICOTOM 5 (Ed. G. Gottstein et al, Aachen, 1978) p. 347.7 A.D. Rollett U.F. Kochs, ICOTOM 8 (Ed. J.S. Kallend, Santa F 1987 p.3758 W.B. Hutchinson, K-I. Nilsson and J Hirsch, Metallurgy of Vacuum Degassed

Steel Producs (Ed. R Pradhan, TMS-AIME, Indianapolis, 1989).9. J Hirsch G. Burmeister L Hoenen and K Liicke Experimental Techniques oTexture Analysis (Ed. H-J. Bunge, DGM, 1986) p.63.10. J Hi rsch and K L{]cke, Acta Met., _ 6 2863 (1988).11. K Liicke and J Hirsch Aluminium Technology 86 (IoM, London, 1986).12. H Honneff H Mecking ICOTOM 5 (Ed. G. Gottstein Aachen, 1978) p. 265.13. H Inagaki, Z Metal lkde, 79, 164 (1988).14. J.R. Hirsch, Materials Science and Technology, in press.

Documents

Effect of Initial Grain Size and Shape on Textures in Sheet Steel