J OURNAL OF MATERI ALS SCI ENCE 21 (1986) 3223- 3226
X- ray analysis of microscopic structure in
def ormed brass F. TUNG, S. ZWUI , Y. WANG Physics Department, Ji l i n University, Changchun 130023, China The microscopic structure of brass is systematically studied by X-ray diffraction profile analysis in this paper. The samples were heat treated differently and deformed in an unidirectional ten- sile state. Parameters such as the effective grain size, average dislocation density, dislocation confi gurati on parameter and stored deformation energy density are carefully deduced from X-ray diffraction data based on some recent theories and models. These parameters change wi th macroscopic stress and strain regularly and it is possible to establish a quantitative relationship between the microscopic structure and macroscopic properties. There is an unique relationship between the structure parameters and stress rather than strain for different sample heat treatments. 1. I n t r o d u c t i o n Aft er plastic def or mat i on in metals there are a lot of dislocations with various di st ri but i ons regarding different degrees and ways of def or mat i on [1]. Dislo- cations may be arranged to form subgrai n boundari es or to cause a cell structure and t hus make the coherent domai n size smaller which can be eval uat ed by an X-ray met hod. In the case of dislocation pile-ups or r andom di st ri but i on there is a large long range strain field causing a specific broadeni ng of X-ray diffraction profiles [2]. The macroscopi c mechanical properties change correspondingly with the microscopic structure which can be studied statistically by X-ray diffraction. Li ne profile analysis is able to give the average coherent domai n size, dislocation density and con- figuration. Stokes' [3] deconvol ut i on met hod allows the separat i on of the physical broadeni ng profiles from the i nst rument al ones and the War r en-Aver bach [4, 5] analysis gives the domai n size and mean square strain. Wilkens [6] has proposed a model of "restric- tedly r andom" di st ri but i on of dislocations to analyse the line profiles and has obt ai ned the dislocations density 0, confi gurat i on paramet er ~r, strain field range Ro and stored elastic energy (El V) for mono- crystal copper. Langf or d [7, 8] has suggested t hat each profile can be considered as a convol ut i on of several Gaussi an funct i ons with Cauchy ones [7-9]. Based on the above ment i oned ideas Wang et al. [10] has worked out a practical procedure and some st andar d curves for line profile analysis of f c c polycrystals. Recent l y sat i sfact ory results have been obt ai ned on h cp and b cc polycrystals and on multi-phase f c c alloys as well [11]. In this paper the results on commer- cial u-brass after uni di rect i onal tensile def or mat i on are presented. 2. Experi ment al detai l s Samples were cut from the commerci al u-brass con- taining 31.55 wt % Zn, 0.6 mm thick, al ong the rolling direction. Aft er polishing t hey were anneal ed in vacuum at two different t emperat ures for 2 h, namel y at 617 K for samples of group A and at 651 K for those of group B. As for group C, samples of filings at r oom t emperat ure wi t hout any anneal i ng were used. Tensile def or mat i on was performed at r oom t emperat ure with a speed of 0. 26mm min -1. The st r ess-st r ai n curve is shown in Fig. 1. X-ray diffraction line profiles were recorded by a rot at i ng anode X-ray di ffract omet er with CuKu radi- at i on and a graphite monochr omat or . Two pairs of reflections 1 1 1/2 2 2 and 2 0 0/4 0 0 were measured on deformed and st andard (annealed but not deformed) samples of 18 mm x 10 mm size. Aft er Stokes' [3] deconvol ut i on the physical broadened profiles and its correspondi ng Fouri er coefficients A(L) were obt ai ned and t hen the two com- ponent s, "part i cl e" coefficients AP(L) and "st r ai n" coefficients AS(L) were separated accordi ng to [10] as L AP(L) = a - - - (1) Do, AS(L) = e x p ( - 2f l 0 L - ~/~2gL2) (2) where a is the quant i t y expressing the " hook" effect, L the specific length perpendi cul ar to the reflecting planes, Dee the average "effective" coherent domai n size, and/~c and /~g the Cauchy ad Gaussi an widths of "st r ai n" broadened profile, respectively. Maki ng use of the st andard curves and procedures described in [10], the average dislocation density ~ and confi gurat i on parameter ~r have been obtained and they give directly the stored elastic energy density El V E ~r -~ = AG~b 2 In ro(~)m (3) where A = 1/4~ for screw dislocations, A = 1/4~z (1 -- v) for edge dislocations, v is Poissons ratio, G is the shear modul us, B is the dislocation strength and r0 is the dislocation nucleus radius. 0022-2461/86 $03.00 + . 12 9 1986 Chapman and Hall Ltd. 3223 o" ( 1 0 7 N m -2 ) 4 0 2O A . ~ t _ ~ _ ~ - ~ " - - ~ - ' - 4 : 3 o j . . ~ -~ 0 10 20 30 40 6(%) Figure 1 Flow st r ess-st r ai n curves for samples of group A and B. (D) Gr oup A annealed at 617K before deformation, (O) group B anneal ed at 615 K before deformation. 3. Resul t s and di scussi on The flow st r ess-st r ai n curves for samples of group A and B are shown in Fig. 1. The dat a for bot h groups cannot be put on an uni que curve, i.e. for a definite e there is big difference in a for samples anneal ed at different temperatures. The Defr-s curves are presented in Fig. 2 showing t hat D~# ~ is always bigger t han D~ ~176 especially after small deformat i on. At very severe def or mat i on as in the case of filings, a-,'effirll I 1 - - - - - a.-ef trl200. The De~-e curves for samples of group B (experimental points in figures are expressed by small circles) are located higher t han for samples of group A (experimental points in figures expressed by small squares). This is in correspondence with those in Fig. 1. On the ot her hand, Fig. 3 shows the Defr-a relation- ships which can be expressed by an uni que curve for samples of bot h groups, i.e. there is a one to one correspondence between Dee r and a in spite of the history of sample treatments. So it mi ght be possible in the fut ure to evaluate flow stress by onl y measuri ng mi crost ruct ural paramet ers such as Deer and others wi t hout dest royi ng the sample. In Fig. 4, (E/ V) is calculated by Equat i on 3 as follows. Supposing the 1 1 1 reflection is affected onl y by screw dislocations, one can obt ai n (E/V)~ ~1 . Then (El V)~ ~ , (E/V)s z22 and (El V)~ 22 can also be deduced in the same way, where the subscript e or s denotes the edge or screw type of dislocations, respectively. A crude average is made as ( E) ~' = -41 [ ( ~ ) ' ll, Deft f (nm) 20 0 0 5 100 EX~l 11 + + ~ ~ V ) s "q- ~ , , V , ] e J (4) o \ o k \ ,, \ \ \ \ \ \ ~o\111-222 \ \ - . - \ \ 200_400~o~ \ o~ 1 1 1 - 2 2 2 ~ o _ o ._~ ~- z~_~- v -_ - ~ - - - o . . . . . 2u0"~40,0 ^ i I . . . . 15 _ ] ~ 0 - - 0 . . . . ~ 0 , . ," " " ' 9 . 9 . . . . " . . : ~ I I ~ i q I I , i ~ i ] ~ 1 - - - - 25 35 45 E ( ~ - - - and similarly the (E/ V) 2~176 It is clearly shown in Fig. 4 t hat (E/ V) ll~ is always smaller t han (E/ V) 2~176 and the difference increases rapi dl y wi t h increasing defor- mat i on. Compari ng Fig. 4 wi t h Fig. 2, one can notice t hat at e < 12% the def or mat i on work is mai nl y used to dest roy the original crystal grains while after > 18% the energy stored inside the domai n increases rapidly because of the increasing dislocation density. The (El V)-e curve for group B is lower t han t hat for group A, in correspondence with those in Figs 1 and 2. One shoul d keep in mi nd t hat (E/ V) obt ai ned by us from Equat i on 4 is not a precise average. In order to improve it the weighted average of real screw and edge dislocation densities and confi gurat i on paramet ers have to be considered. Fr om Fig. 5 one can seen t hat dat a for bot h A and B group of samples again may be put on one uni que (El V)-a curve unlike the case of (El V)-e curves. It is clear in Fig. 6, dislocation density at first increases with increasing e but stops changi ng at > 10~~ -2, while )Q" rapidly reduces at large deformat i ons, especially for _M (2 0 0/4 0 0). This prob- ably means t hat dislocations exist mai nl y in pile-ups at small deformat i on to f or m a long range strain field while wi t h increasing def or mat i on the dislocations will gradual l y form the regular lattice of dipoles or rows normal to the glide plane and the l ong range strains cancel each ot her out. There is significant difference in the above st at ed st ruct ural paramet ers between those deduced from 1 1 1/2 2 2 and from 2 0 0/4 0 0. For our plate sample, any h k l reflection is given by those crystal grains whose {h k l } planes are parallel to the sample surface. Different reflections are at t ri but ed to grains of dif- ferent ori ent at i ons which certainly have different Figure 2 Effective domai n size against strain curves for samples of group A, B and C (D) group Ao (O) group B, (A) group C (filings). 3224 D ef t ( n m) 2 0 0 1 0 0 20 o \ o \ \\ \ \\ \ \ \ ~ o 1 1 1 - 2 2 2 k \ ,, \ k 2 0 0 - 4 0 0 \ \ - - - - o- o- o_ _ _ - - - ~ . . . . . . . 30 4 0 5 0 60 cr (107N m -2) Figure 3 Effective doma i n s i ze agai ns t f l ow stress curves. (O) 111 222, (El) 20 0 - 4 0 0 . Figure 4 Av e ra g e s tored energy de ns i t y a g a i ns t strain curves. (IO G J ~-3,~ 20 1 0 2 0 0 - 4 0 0 . ~ O ' - - - - - ~ - - - - O ...o-- I I I - 2 2 2 / 4 b l 0 . . . . . . 0 o ~ / ~ ~ o . . . . / O I I I O ~ ~ ~ ~ / / q I I I I I I I I I I I I I I ] I I { I 15 25 35 4 5 E ( % ) ( { 4 (1 0 ~ n~3) I t 1 0 2 0 0 - 4 0 0 . " rq.C ~ . . . . El" / / ~ d / . O ' ~ o _ o ~ 1 1 1 -222 . . . . . . . . / O ~ ~ _ _ O / ~ "O ~ / O - - - - - - 0 - - - - - - 0 ~ .ZX 20 30 4 0 5 0 6 0 0" (107N m "2) Figure 5 Av e ra g e s tored energy de ns i t y a g a i ns t f l ow stress curves. Figure 6 Di s l oc a t i on parame t e rs agai ns t f l ow stress c u r v e s . (ldOcm-2) 30 2 0 1 0 , ~ 1 1 1 -222 \ \ \ \ \ p 2 0 0 - 4 0 0 \ \ \ \ ~ [ 3 \ rn~ n ~ \ B / / / N / - ' / " \ 9 M 2 0 0 - 4 0 0 . . o / / \ \ o - - / o ~ i 0 1 1 1 - 222 - I : ] . . - o ~ 2 ~ ' 2 ' ' I , , , 3O 30 2O 10 9 . . . . . , 4 ~ c ] , . " " , I , , , , I , , , ' 6 ~ 0 ' ' Q - - N O " 0 [ 4 0 5 0 O- m - 2 ) 3 2 2 5 stress and strain states hence different structural parameters. The data obtained from the filings are put together on the curves in Figs 2 to 5 with those o f tensile samples to s how the reasonable tendency o f curves. B ut our data for filings are a little different from others [12, 13]. 4 . Concl usi ons The X-ray method for studying a deformed structure is a non-destructive one which can give the statistically averaged microscopic parameters associated uniquely with the macroscopic fl ow stress regardless o f the heat treatment history o f the sample. This may al l ow a prediction of the macroscopic properties from the microscopic structure. The quantitative relationship between macroscopic mechanical properties and the microscopic deformed structure coul d probably improve the work-hardening theory for metals. R e f e r e nc e s 1. P. HAASEN, in "Physi cal Met al l ur gy" (Cambr i dge Uni - versi t y Press, Cambr i dge, 1978) p. 243. 2. H. P. KLUG, i n " X-r ay Di ffract i on Pr ocedur es " (Wiley, New Yor k, 1974) p. 618. 3. A. R. STOKES, Proe. Phys. Soc. 61 (t' 948) 382. 4. B. E. WARREN, in "Progress in Metal Physics" (Pergamon Press, London, 1959) p. 147. 5. Idem, " X-r ay Di f f r act i on" (Addi son-Wesl ey, Readi ng, Massachusetts, 1969) p. 25[. 6. M. WI LKENS, Phys. Status Solidi A 2t (1970) 359. 7. J. I. LANGFORD, J. Appl. Cryst. 11 (1978) 19. 8. J. I. LANGFORD and A. J. C. WI LSON, in "Cr ys t al - l ogr aphy and Crystal Perfections" (Academic Press, New Yor k, 1963) p. 207. 9. J. M. COWLEY, in "Diffraction Physics" (Nor t h-Hol l and, Amsterdam, 1975) p. 25. 10. WANG YUMI NG, LEE SHANSAN and LEE YEN- CHI N, J. Appl. Cryst. 15 (1982) 35. I 1. S. ZWUI , J. Mater. Sci. Lett. 4 (1985) 1434. 12. J. B. COHEN a nd J. E. HI LLI ARD, in "Local Atomic Arrangement Studied by X-r ay Diffraction" (Gordon B reach, New Yor k, 1965) p. 289. 13. E. MAGEE and W. J. KI TCHI NGMAN, J. Phys. D: Appl. Phys. 9 (1976) 939. Received 19 September and accepted 30 October 1985 3226