Professional Documents
Culture Documents
Hall 1985
Hall 1985
Samples of INCONEL* 600 were examined in the mill-annealed and solution-annealed states, and
after isothermal annealing at 400 ~ and 650 ~ The corrosion behavior of the samples was ex-
amined, analytical electron microscopy was used to determine the microstructures present and the
chemistry of grain boundaries, and Auger electron spectroscopy was used to measure grain boundary
segregation. Samples of different alloys in the mill-annealed state were found to have quite different
microstructures, with Cr-rich M7C3 carbides occurring either along grain boundaries or in intragranular
sheets. The corrosion behavior of the samples correlated well with the occurrence of grain boundary
chromium depletion. Solution annealing at 1190 ~ caused dissolution of all carbides, whereas at
1100 ~ the carbides either dissolved or the grain boundaries moved away from the carbides, de-
pending upon alloy carbon content. Low-temperature annealing at 400 ~ had little effect on mill-
annealed or fully solutionized samples, but in samples with intragranular carbides present, the grain
boundaries moved until intersecting or adjacent to the carbides. Isothermal annealing at 650 ~ caused
carbide nucleation and growth at grain boundaries in fully solutionized samples. Chromium depletion
at grain boundaries accompanied carbide precipitation, with a minimum chromium level of 6 wt pct
achieved after 5 hours. Healing was found to occur after 100 hours. Solution-annealed samples with
intragranular carbides present had more rapid corrosion kinetics since the grain boundaries moved back
to the existing carbides. Thermodynamic analysis of the chromium-depletion process showed good
agreement with experimental measurements. The Auger results found only boron present at grain
boundaries in the mill-annealed state. Aged samples had boron, nitrogen, and phosphorus present,
with phosphorus and nitrogen segregating to the greatest extent. The kinetics of phosphorus segre-
gation are much slower at 400 ~ compared with 650 ~
1226--VOLUME16A,JULY1985 METALLURGICALTRANSACTIONSA
0.5 ' ' ' ' I ' ' ' ' I . . . . I 0.5 , ' ' ' I . . . . I . . . . I
"~0.3 ~0.3 -
(fJ
(,t3
._1
~=0.2 ~ 0z
hl
01 0.1
o
, - , ,~ , o ,o, ,I , , , ,r Oc , ,o, ,~ , I I II I , , ,6
- 10 10 100 1.0 t0 100
TIME (hrs) TIME (hrs)
(a) (b)
Fig. 1 --Corrosion behavior (weight loss during test) v s annealing time at 650 ~ for samples of alloy A (a) solution annealed for 1 h at 1t90 ~ (b) solution
annealed for 0,5 h at 1100 ~ Closed circle: no attack; open circle: broke upon bending after test; closed circle with arrow: disintegrated during test.
Table II. Corrosion Response of Mill-Annealed ary chromium concentration very near to the carbides was
Alloys Used in This Investigation measured in this sample, values from 10.3 to 13.1 wt pct
were found, with a mean of 11.9 wt pct. However, the low
Corrosion Rate density of carbides at grain boundaries in this sample re-
in Nitric Acid Stress Corrosion suited in very little chromium depletion for most areas of the
Alloy (mg/cm ~) Test Response
grain boundaries.
A 0.134 intergranular SCC observed The mill-annealed microstructure of sample A was found
B 0.164 intergranular SCC observed to be very complex and heterogeneous. The majority of the
C 0.014 no SCC observed sample had moderate M7C3 precipitation along the grain
boundaries (Figure 2(c)) and grain boundary chromium lev-
times, with samples breaking upon bending after the test for els between carbides of - 1 2 wt pct. This is quite similar to
annealing times of 0.5 and 1 hour. Some weight loss was the previous sample. However, some regions of this sample,
seen after 3 hours, and the 5, 15, and 23 hour samples all occurring in bands, had a very different microstructure
showed large amounts of weight loss. Once again, after which is difficult to display using transmission electron im-
100 hours no attack or weight loss occurred. ages. Figure 2(d) is a secondary electron image of one of
these regions, in which the M7C3 carbides appear bright.
B. Analytical Electron Microscopy Results These regions consist of large grains with rod-shaped intra-
granular carbides present, and smaller grains with massive
1. Mill-annealed samples
grain boundary precipitates. It was not possible to find suit-
The microstructures present in the mill-annealed samples
able grain boundary regions between carbides in these re-
are shown in Figure 2. Figure 2(a) shows a typical grain
gions for the measurement of chromium concentration due
boundary in sample B. Extensive precipitation is seen along
to the extensive precipitation. However, the carbide size and
the boundary, and some precipitation has also occurred
density is larger than that seen in either of the other mill-
along dislocation lines near the boundary. All precipitates
were identified by electron diffraction as M7C 3. EDX spec- annealed samples.
troscopy showed that the composition of the M portion of 2. Mill-annealed plus heat treated samples
this carbide was 94 wt pct Cr, 3 wt pct Ni, and 3 wt pct Fe. The effects of three different heat treatments on the
The composition of grain boundary regions between car- previously-described mill-annealed microstructures were
bides in this sample was also measured using EDX spec- examined. These heat treatments consisted of 400 ~ for
troscopy. Chromium contents in the range 9.2 to 10.3 wt pct 200 and 2000 hours and 650 ~ for 100 hours. The results
were measured, with a mean of 9.8 wt pct. Several com- showed that neither of the 400 ~ heat treatments had a
position profiles across grain boundaries in the vicinity of significant effect on the microstructures. A slight increase in
carbides were also obtained, and one result is shown in carbide size and density may have occurred.
Figure 3. The width of the chromium-depleted region is The effect of heat treating the mill-annealed samples for
approximately 800 nm (0.8/xm). 100 hours at 650 ~ can be illustrated by the results for
A typical area in the mill-annealed sample of alloy C is alloy A, shown in Figure 4. Comparing this microstructure
shown in Figure 2(b). Only occasional particles, again to Figure 2(c), it can be seen that significant carbide nucle-
M7C3, were found along grain boundaries. Strings of M7C3 ation and growth has occurred. The M7C3 grain boundary
carbides were found in the matrix apparently delineating the precipitates in Figure 2(c) have grown, often along planes
prior positions of grain boundaries. When the grain bound- not corresponding to the grain boundary plane. Additional
(a)
(d)
(c)
Fig. 2--Microstructures in mill-annealed samples: (a) transmission electron image (TEl) of alloy B, (b) TEl of alloy C, (c) TEl of alloy A, and
(d) secondary electron image of alloy A.
smaller carbides have nucleated and grown both along grain The results showed that the higher solution-anneal tem-
boundary regions between large carbides and also along the perature (SA2) caused all precipitates to dissolve, so that the
interfaces between large carbides and the matrix. Exten- microstructures of all samples were free of precipitates after
sive new precipitation along dislocations in the vicinity of that treatment. The SA1 treatment produced different re-
the grain boundary is seen. Unlike the grain boundary pre- suits. For sample C, which had the lowest carbon content
cipitates, the dislocation precipitates are Cr-rich M23C 6 and had moderate precipitation at the grain boundaries in the
which have a cube-cube orientation relationship with the mill-annealed state, SA1 caused all precipitates to dissolve.
Ni matrix. After the solution-anneal, the only precipitates present were
3. Solution-annealed samples some very small (0.1 /xm) Cr-rich carbides at some grain
The mill-annealed samples described previously were boundaries. However, in sample A, which had large, nu-
solution annealed using one of two different treatments: merous precipitates at the grain boundaries in the mill-
1100 ~ for 0.5 hour (SA1) or 1190 ~ for 1 hour (SA2). annealed condition and an intermediate carbon content, SA 1
(a) (b)
(c) (d)
Fig. 7--Transmission electron images of grain boundaries in alloy A after solution annealing at 1190 ~ for 1 h and then aging at 650 ~ for: (a) 0.5 h,
(b) 5 h, (c) 23 h, and (d) 100 h.
/I
B 0.054 o 0.07
I.--
C 0.160 r
~O.OE
ACKNOWLEDGMENTS
V. CONCLUSIONS The authors would like to acknowledge the assistance
1. Samples of INCONEL 600 which had nominally re- of the following: Ms. C.S. O'Toole, General Electric
ceived the same mill-anneal treatment were found to have Company, San Jose, CA, who provided samples and some
very different microstructures and corrosion properties. corrosion data; Mr. N. Lewis, General Electric Corporate
In some cases the grain boundaries were found to be Research and Development, who assisted with the AEM
covered with Cr-rich M7C3 carbides; in other cases the experiments; and Dr. R . L . Fullman, General Electric
grain boundaries had moved away from the carbides, Corporate Research and Development, who provided assis-
leaving sheets of intragranular carbides. tance with the thermodynamic data.
2. A solution anneal at 1190 ~ was found to dissolve all
carbides in all mill-annealed samples. A solution an-
neal at 1100 ~ either dissolved all carbides or caused
the grain boundaries to break away from the carbides, REFERENCES
depending upon the alloy carbon content and the micro- 1. E. Serra: EPRI Special Report NP-2114-SR, Electric Power Research
structure in the mill-annealed state. Institute, Palo Alto, CA, 1981.
2. J. Blanchet, H. Coriou, L. Grail, C. Mahieu, C. Otter, and G. Turluer:
3. Low-temperature aging at 400 ~ of fully solution- Stress Corrosion Cracking and Hydrogen Embrittlement of lron Base
ized samples caused no microstructural changes. Low- Alloys, NACE, Houston, TX, 1979, p. 1149.
temperature aging of samples with rows of intragranular 3. J.R. Cels: J. Electrochemical Society, 1976, vol. 123, pp. 1152-56.
carbides present caused grain boundaries which were 4. J.R. Cels: Corrosion, 1978, vol. 34, pp. 198-209.
free of carbides to move until they intersected or were 5. G.J. Theus: Corrosion, 1977, vol. 33, pp. 20-26.
6. N. Pessal, G.P. Airey, and B.P. Lingenfelter: Corrosion, 1979,
adjacent to the carbides. vol. 35, pp. 100-07.
4. Isothermal annealing at 650 ~ of fully solutionized sam- 7. N. Pessal: Corrosion Science, 1980, vol. 20, pp. 225-42.
ples caused carbide precipitation at grain boundaries to 8. Ph. Berge, J. R. Donati, B. Prieux, and D. Villard: Corrosion, 1977,
occur, Carbides were observed after 0.5 hour, increasing vol. 33, pp. 425-35.
9. G.P. Airey: Corrosion, 1979, vol. 35, pp. 129-36.
in size and density with increasing time. After 5 hours, 10. P.K. De and S.K. Ghosal: Corrosion, 1981, vol. 37, pp. 341-49.
nearly continuous grain boundary coverage was seen, 11. R.C. Scarberry, S.C. Pearman, and J.R. Crum: Corrosion, 1976,
and precipitation had begun along dislocation lines. Both vol. 32, pp. 401-06.
MTC 3 and M23C6 carbides were observed, with both car- 12. G.P. Airey, A.R. Vaia, N. Pessal, and R.G. Aspden: Journal of
bides rich in Cr. Metals, Nov. 1981, vol. 33, pp. 28-35.
13. C.S. Tedmon, Jr. and D.A. Vermilyea: Corrosion, 1971, vol. 27,
5. Composition profiles across grain boundaries in samples pp. 376-81.
isothermally annealed at 650 ~ showed significant chro- 14. R.L. Cowan, II and C.S. Tedmon, Jr.: Advances in Corrosion
mium depletion at the boundary in the vicinity of car- Science, Plenum Press, New York, NY, 1973, vol. IIl, p. 293.
bides for annealing times of 3, 5, and 23 hours. The 15. C. Stawstrom and M. Hillert: J. Iron Steel Inst., 1969, vol. 207,
pp. 77-85.
width of the chromium-depleted region increases with 16. G.P. Airey: Metallography, 1980, vol. 13, pp. 21-41.
increasing time. The minimum chromium level was mea- 17. G.S. Was, H. H. Tischner, and R. M. Latanision: Metall. Trans. A,
sured to be 6 wt pct and occurs after 5 and 23 hours of 1981, vol. 12A, pp. 1397-408.