Ferrite structure and mechanical properties of low alloy duplex steels

Scripta METALLURGICA

Vol. IS, pp. 867-872, 1981 Printed in the U.S.A.

Pergamon Press Ltd. All rights reserved

FERRITE STRUCTUREAND MECHANICAL PROPERTIES OF LOW ALLOY DUPLEX STEELS R.H. Hoel and G. Thomas Materials and Molecular Research Division Lawrence Berkeley Laboratory and Department of Materials Science and Mineral Engineering University of C a l i f o r n i a , Berkeley, C a l i f o r n i a 94720

(Received April 20, 1981) (Revised May 26, 1981)

Introduction Dual phase steels consisting of strong load carrying martensite in a d u c t i l e f e r r i t e matrix form a r e l a t i v e l y new class of strong d u c t i l e steels for applications requiring t e n s i l e strengths of the order of 700 MPa. The mechanical properties are optimized by c o n t r o l l i n g the composite morphology, the volume f r a c t i o n of martensite, the carbon content of martensite and the addition of d i f f e r e n t ternary and quaternary a l l o y i n g elements, as summarized in recent symposia (1,2): I t is generally accepted that the ultimate t e n s i l e strength of the dual phase aggregate w i l l increase with the martensitic volume f r a c t i o n , Vm. Experimentally i t has been shown that the t e n s i l e strength increases l i n e a r l y with the martensitic volume (4-9) and is given e m p i r i c a l l y by the law of mixtures = %nVm + of(1 - Vm)

(1)

where ~ and ~ are the t e n s i l e strengths of respectively the martensite and f e r r i t e phases. NormallY, the ~ i e l d strength is also found to increase with Vm. However, recent studies (3,1) have shown that some dual phase steels show an unusual decrease in y i e l d strength as the martensitic volume f r a c t i o n increases. Figure l shows such a decrease in dual phase steels of compositions Fell.5SiIO.15ClO.O3Nb (base + Nb) and FeIl.5SiIO.15ClO.38~:o (base + Mo). These are a l l weight percent, unless otherwise is specified. Thomas and Koo (1) suggested that the reason fo r the unexpected decrease in y i e l d strength is as follows. The y i e l d strength of the dual phase aggregate is p r i m a r i l y dependent on the y i e l d strength of the f e r r i t e phase when this phase forms a continuous matrix with a f i n e dispersion of martensite islands. This assumption depends on the martensitic volume f r a c t i o n not being too high. Futhermore, since the carbide forming elements Nb and i!o are present, the f e r r i t e phase is subject to p r e c i p i t a t i o n strengthening. The level of this p r e c i p i t a t i o n strengthenin~ in the f e r r i t e phase w i l l decrease as the martensitic volume f r a c t i o n increases, leading to an o v e ra l l decrease in the y i e l d strength of the dual phase aggregate. The purpose of this communication is threefold. I) To confirnl the presence of and to characterize the p r e c i p i t a t e s in the f e r r i t e phase of the base + Nb and base + Mo steels, 2) to study any possible v a r i a t i o n in p r e c i p i t a t e density as the martensitic volume f r a c t i o n is changed and 3) to determine the level of p r e c i p i t a t i o n strengthening. Experimental The dual phase steels described in this paper were a l l vacuum melted. The dual phase heat treatment consisted of a u s t e n i t i z a t i o n at llO0°C, followed by intermediate quenching into an ice brine, annealing in the two-phase f i e l d , a f i n a l ice brine quench, and aging at room temperature fo r a few weeks. In order to study the possible v a r i a t i o n s in p r e c i p i t a t e density with martensitic volume f r a c t i o n s , it was necessary to measure f o i l thicknesses with an accuracy of a few percent. The most s u i t a b l e method was found to be that of convergent beam m i c r o d i f f r a c t i o n (14-16). A b r i e f description and an example of such a measurement is presented in the appendix.

867 0036-9748/81/080867-06502.00/0 Copyright (c) 1981 Pergamon Press Ltd.

868

PROPERTIES OF DUPLEX STEELS

Vol.

15,

No.

8

As the basis for the determination of the precipitation strengthening level we used the classical Orowan mechanism. Orowan (lO) described the mechanism by which dislocations encounter and bypass point obstacles. The flow stress, ~, required to bypass such an obstacle is aiven by

T

= ~o + ~ ~b , where T O = flow stress of the f e r r i t e in the absence ~n

of point obstacles, m = a geometrical factor depending on particle shape and distribution, G = shear modulus, b = burgers vector of the glide dislocations, and ~n = average spacing betweenobstacles in the glide planes. I f the obstacles are spaced at Bistances less than the mean free dislocation slip length, an enhancedstrengthening of the matrix occurs. In a m e t a l l i c matrix the d i s l o c a t i o n s l i p length is t y p i c a l l y 1 - 50~im. Results and Discussion The presence of f i n e homogeneously nucleated p r e c i p i t a t e s in the f e r r i t e matrix is evidenced in Figure 2. The matrix is oriented along the [001] zone axis. Upon t i l t i n g away from this o r i e n t a t i o n , some p r e c i p i t a t e s give rise to d i s l o c a t i o n loop contrast, depending on t h e i r o r i e n t a t i o n r e l a t i v e to the d i r e c t i o n of the electron beam (Figure 3). This is t y p i c a l of coherent or p a r t i a l l y coherent plate p r e c i p i t a t e s (17). In Figure 2 these plates are seen edge-on and they are oriented on or within 12° of the {I00} planes. The plates oriented face-on are i n v i s i b l e as ~ . b = 0 and no s t r a i n contrast is observed in the surrounding matrix. Weak beam imaging ( I I , 12) reveals plate thicknesses of less than 2.5nm and plate diameter ranging front 4nm to 25nm. The mean plate diameter is 13.5 (~2)nm. Figure 3 shows examples of loops changing from outside contrast to inside contrast as the sign of (~.b)s is reversed. The inside-outside contrast analysis (13) of these p r e c i p i t a t e s shows that they give rise to vacancy type loops which has been predicted f o r the ear ly stages of p r e c i p i t a t e formation (18). These c h a r a c t e r i s t i c plate p r e c i p i t a t e s are common to both the base + Mo and base + Nb a l l o y systems. Thin f o i l s of the base + Nb a l l o y prepared w i t h i n a few hours of heat treatment showed the presence of the same c h a r a c t e r i s t i c p r e c i p i t a t e s when examined in the electron microscope. Hence the p r e c i p i t a t e s are not formed as a r e s u l t of room temperature aging. Upon tempering at IO0°C f o r I / 2 hour no p r e c i p i t a t e growth was observed. The size and morphology also remained unchanged as the tempering temperature was raised to 190°C and the time increased to 50 minutes. A s i m i l a r sequence of tempering of an FeI2SilO.IC a l l o y , a composition s i m i l a r to the base of the a l l o y s described above, showed that upon tempering at IO0°C for I / 2 hour some carbides s t a r t to coarsen. Tempering at 170°C f or 1 hour (19) and 200°C f or 1 hour (20) results in a much more pronounced coarsening, and the presence of cementite precipitates '~,17nm wide and 150nm long, has been reported (20). The increase in tempering resistance of the base + Nb alloy must be attributed to the presence of the niobium. For precipitate density determinations, thicknesses weremeasuredtypically within an area of 100nm to 200nln in diameter (although better spatial resolution is possible) with an estimated accuracy of a few percent. The number of precipitates was counted from photographic prints within a typical true cross-sectional area of a few lOnm~. The effect of particle overlap on the projected image could be ignored without introducing serious errors in the final result, and the effect of particles cut by the f o i l surfaces was corrected for by the equation (21). NA = Nv (x + 2r)

(2)

where NA is the number of p a r t i c l e s per u n i t projected area in a f o i l of thickness, x, NV is the p a r t i c l e density and 2r is the mean p a r t i c l e diameter. 2r is taken to be the diameter of an imaginary spherical p r e c i p i t a t e whose volume is equal to that of an average sized plate p r e c i p i t a t e of diameter D and thickness t. The value obtained f o r 2r is 7.3nm. The l i n e a r p a r t i c l e density, NL, is calculated as the cube root of NV and the mean distance between the centers of neighborinq p r e c i p i t a t e s , ~, is taken as the reciprocal of NL. Such distances were determined t y p i c a l l y with an accuracy of ,4%. T a b l e l shows that as the martensitic volume f r a c t i o n of the base + Nb a l l o y is increased from 20% to 40% there is an unmistakable decrease in p r e c i p i t a t e density. The respective mean i n t e r p a r t i c l e distance increases from 49.8 (±3)nm to 64.1 (±3.5)nm. The uncertainties quoted represent the limits of reproducibility from six independent measurements. Such limits are

Vol.

15, No. 8

PROPERTIES OF DUPLEX STEELS

869

t y p i c a l l y ±6%. The absence of precipitate growth as this alloy is tempered at IO0°C is Confirmed by measurement of the particle density which, within experimental error, is the same as that of the untempered material. Table l shows that the precipitation density of the base + Mo alloy with a martensitic volume fraction of 20% is the same as that of the base + Nb alloy. TABLE l Precipitate density, Nv, and mean interparticle distance, ~, as a function of composition and martensitic volume fraction, Vm Composition

|V~,

t. . . . . . . . . . . . . . . . . . . . ~. . . . . . . . . . . . . . . . . . . . X l O l h c m -11 | ~,, nm

~ I Nv, .

Base + 0.03% Nb Base + 0.03% Nb ~ Base + 0.03% Nb I Tempered, IO0°C, I /2 hr. Base + 0.38% ~io

40 20 20

.

I I L

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

81 ' 15 38, 6

49.8 ' 3 64.1 , 3.5

85; 15 77, 16

49.0~ 2.7 50.6, 3.5

..............................................

The v a l i d i t y o f applying the Orowan model in c a l c u l a t i n g the level of p r e c i p i t a t i o n strengthening is j u s t i f i e d since fo r these a l l o y carbides i t has been calculated that they w i l l act as impenetrable barriers to d i s l o c a t i o n motion (22). The dispersion strengthening term, ~ b___, in the Orowan equation (eqn. I) has been modified by various authors (see (23) for a review) to take into account the f i n i t e particle size, more refined estimates of the dislocation line tension, the mutual interaction of the two bowed out arms of the dislocation on either side of the particle, and f i n a l l y the particle shape. For plate (or rod) shaded particles of diameter D and thickness (or rod length) t the following expression has been derived for the dispersion strengthening, [D' 083

Gb

,

LD = ~3TICiT-~/2

x

ln( D)

(5)

where v is Poisson's r a t i o , L is the e f f e c t i v e i n t e r p a ~ t i c l e distance (see below) and r o is the d i s l o c a t i o n core radius. Calcu-~a~i-on-sbased upon an equation of this type have shown that thin plates can exert a strengthening influence un to 3 times greater than the same number of spherical p a r t i c l e s of equivalent volume. The equation derived by Fullman (24) for the e f f e c t i v e p a r t i c l e spacing is L = ( t D l f ) ~/~, where f is the p a r t i c l e volume f r a c t i o n . By d e f i n i t i o n f = r~/(C/2) z, where r is the mean p a r t i c l e radius. Equation 5 was used to estimate the magnitude of the p r e c i p i t a t i o n strengtheninq levels of the base + Nb a l l o y , assuming the fo l l o w i n g values: G = 79.8 MPa, b = 0.248 nm, v = 0.29, t = 1.4 ( ! ] . l ) nm, D = 13.5 (~2) nm and r o = l . l (!0.3) nm. The values f o r ;. are l i s t e d in Table 1. Assuming these values the level o f p r e c i p i t a t i o n strengthening with a martensitic volume f r a c t i o n of 20% is T2o = 274 ~!Pa. As Vm is increased to 40% the strengthening level drops to t , o = 211MPa. I t is the change in t~e p r e c i p i t a t i o n strengthening l e v e l , At = l ~ 0 x~o , that is of p a r t i c u l a r i n t e r e s t here. Inserting the uncertainties quoted above into equation 5 shows that the decrease AT = 71.8 (~27) MPa. With reference to Figure 1 and reference (3) i t is seen that the observed decrease in the y i e l d strength o f the dual phase f e r r i t e - m a r t e n s i t e aggregate is 5T~.~ MPa as the martensitic volume f r a c t i o n increases from 20% to 40%. Let us consider an imaginary base + Nb a l l o y which in every way is id e n t ic al to the experimental one describe~ above, except that the f e r r i t e phase is p r e c i p i t a t e free and i t shows the usual increase in y i e l d strength as Vm increases. The above calculations show that the unexpected drop in the y i e l d strength can be explained as a r e s u l t of a reduced level of p r e c i p i t a t e strengthening in the f e r r i t e phase.

I t has been shown that previously unreported fine plate precipitates may be present in the f e r r i t e phase of dual phase steels. The plates are oriented on the ~]00} planes and in the Fell.5 SilO.15 CI0.03 Nb and Fell.5 SilO.15 CI0.38 Mo alloys they s i g n i f i c a n t l y strengthen the f e r r i t e phase. Based upon existing theories for precipitation strengthening,

870

PROPERTIES

OF DUPLEX

STEELS

Vol.

IS, No.

i t has also been shown that there is a significant drop in the level of precipitation strengthening as the martensitic volume fraction, Vm, increases from 20% to 40%. The correspondinq drop in the observed yield strength of the dual phase aggregate supports the assumption that for a martensitic volume fraction up to 40% the yield strength of this aggregate is dominated by the yield strength of the f e r r i t e phase. Koo and Thomas (1,2) have pointed out that the carbon content of the martensite decreases as Vm increases. I t is well-established that the strength of this phase is proportional to its carbon content (22). Hence the strength of the martensite w i l l also decrease as Vnl increases. I t is therefore important to take the variations of '~r1' and ()f, with martensite volume fraction, into account when applying the law of mixtures leqn. l ) to the ultimate tensile strength of dual phase steels. Acknowledgement_s" The authors wish to thank Drs. U. Dahmen, R. Fisher and J. Y. Koo for helpful discussions and advice during this work. The work was supported by the National Science Foundation (R.H. Hoel), D,~.IR77-19358, Office of Energy Research, Office of Basic Energy.Sciences, ~laterials Science Division of the U.S. Department of Energy under Contract No. W-7405-ENG-48. References I. 2. 3. 4. 5. 6. 7. 8. 9. lO. II. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24.

G. Thomas and J. Y. Koo, Proceedings of the Symposium on Structure and Properties of Dual Phase Steels, R. A. Kot and J, W. Morris, eds., p. 183, The Het. Society of AII4E, ~darrendale, PA (1979). J. Y. Koo and G. Thomas, Proceedings of the Symposium on Formable IISLA and nual Phase Steels, A. T. Davenport, ed., p. 40, The Met. Society of .A.IUE, !rarrendale, P~. (1979). P. K. Costello, Lawrence Eerkeley Laboratory Report No. 8628 and ~I.S. Thesis, University of California, CA (1978). B. Karlsson and G. Linden, r~aterial Science Engineering 17, 209 (1795). J. Y. Koo and G. Thomas, ~let. Trans. A 8A, 525 (1977). P~. C-.. Davies, ,Pier. Trans. A 9A, 41 (1973--~. S. Hayalai and T. Furukawa, Pr-oceedings of the Symposium on HSLA Steels, J. Crane, ed., D. 311, Union Carbide Corporation, New York (1975). L. F. Ramos, D. K. Matlock and G. Krauss, ~let. Trans. A IO~,, 259 (1979). J. Y. Koo, H. J. Young and C.. Thomas, Met. Trans. A !I_A,-]~'5-2 (1980). E. Orowan, f!ature 149, 643 (1942). F). J. H. Cockayne,--Na-ture. 27A, 452 (1972). C. Thomas and M. J, Goringe, Transmission Electron rlicroscoD.Y of {]aterials, Ist. ed., D. 305, John 14iley and Sons, NewYork (1979). H. F611 and M. Wilkins, Phys. Stat. Sol. 31A, 519 (1975). P. f4. Kelly, A. Jostsons, R. G. Blake and J]--G. Napier, Phys. Star. Sol. 3~_A, 771 {1975). T. Tan, k'!. L. Bell and G. Thomas, Phil. Nag. 24, 417 (1971). S. !,I. Allen, Phil. Mag., in press. G. Thomas and 1.I.J. Goringe, Transmission Electron Microscopy in l:aterials, Ist. ed., p. 159, John Wiley and Sons, New York (1979). U. Dahmen, K. H. Westmacott and G. Thomas, Acta. l'4et., in press. A. Pelton, Department of Materials Science and .~]ineral Engineering, Universit.v of California, Berkeley, CA, private con~nunication. J. Y. Koo and G. Thomas, Proceedings of the Symposium on Formable HSLA and Dual Phase Steels, A. T. Davenport, ed., p. 51, The f.let. Soc. of AIIIE, VJarrendale, PA (1979). P. B. Hirsch, A. llowie, R. B. Nicholson, D. t.l. Pashley and {:. j . i.!helan, Electron ,~!icroscopy of Thin Crystals, 2nd ed., D. 424, Robert E. Kriegger Publishing Conpanv, New York (1977). G. Thomas, D. Schmatz and W. W. Gerberich, Proceedings of the Second Berkeley International Conference on High Strength Materials, V. F. Zackay, ed., p. 291, John Wiley and Sons, New York (1965). P. M. Kelly, Int. Met. Rev. 18, 31 (1973). P.. L. Fullman, Trans. Ainer. Tnst. ~lin. Het. ing. 19_~7.,447 (1953).

@~_nd~x Figure 4 shows an example of a f o i l thickness measurement by twn-beam converqent beam k A~l . microdiffraction. The method involves plotting si21ni 2 vs. ~Ini 2 where s i = ~ ~ anon i is an '

8

Vol.

15,

No.

8

PROPERTIES OF DUPLEX STEELS

871

integer. ~ is the electron wavelength, d is the int er planar spacing of the observed r e f l e c t i o n , 20d is the distance on the photographic p r i n t between the d i f f r a c t e d and undiffracted discs, and AOi is the distance from the center of the d i f f r a c t e d disc to the i t h dark f r i n g e .

Such a p l o t w i l l s a t i s f y the equation ( s i l n i ) Z + II&g~ni~ = t -z, for the

correct choice of n i . A s i m i l a r equation holds f or the bright fringes (16), and i t is important to include these fringes in the measurement as they w i l l compensate f or any possible systematic error in the f i n a l r e s ult . As i l l u s t r a t e d in Figure 4, a t r i a l p l o t with d i f f e r e n t n i values is made and a s t r a i g h t l i n e is only obtained for the correct value. The s t r a i q h t l i n e intersects the ordinate in t -2. The slope is given b y e- 2 and provides a good check for the v a l i d i t y of the technique used. Linear regression-and trend analysis is used to f i t the best values to t and ~; . For the Nb containino steel described in the text and the r e f l e c t i o n used in Figure 4,gthe experimental e x t i n c { i o n distance is within 2" of the t h e o r e t i c a l value f o r pure iron.

r

I

~

I

i

"

~000

t40~

[

ULTIMATE

900

-

~2d -

800

-

I I'--

~voo4

u r00-

(12 F-

I

Y,ELo

- 600

8oi500

L t

o

t

i I 1 20 40 % MARTENSITE_

l 60

Figure l Fig. 1 - Shows the v a r i a t i o n in ultimate t e n s i l e strengths and y i e l d strengths of the base + 0.03 wt%Nb, A and base + 0.38 wt%Mo, | , a l l o y s with martensitic volume f r a c t i o n . Data from (3).

Figure 2 Fig. 2 - Shows the presence of f i n e homogeneously nucleated p r e c i p i t a t e s in the f e r r i t e phase of the base + Nb a l l o y . This is also typical of the base + I~o a l l o y .

872

PROPERTIES OF DUPLEX STEELS

Vol.

15,

No.

Figure 3 Fig. 3 - In certain o r i e n t a t i o n s the p r e c i p i t a t e s show loop contrast. The encircled p r e c i p i t a t e s in figures a and b are examples of how the contrast changes from inside to outside as the sign of (~.b)s is reversed. The micrographs show p r e c i p i t a t e s in the base + rlb s t e e l , but t h e i r appearance is i d e n t i c a l to that of the Mo containing steels.

Fig. 4 - An example of thickness determination by convergent beam m i c r o d i f f r a c t i o n . The s t r a i g h t l i n e intersects the ordinate in t -2 (t = thickness) and the slope equals C~-~ = extinction distance~. (Cg

Figure 4

8