278 Prakt. Met. Sonderband 38 (2006) 
am 
To obtain quantitative results about the precipitate state (size, shape, and distribution of the 
each precipitate type) a digital image processing system was used to analyze the EFTEM cre 
images and/or TEM bright field micrographs. To get statistical relevant numbers specimen cre 
areas of up to 40 pm? were quantitatively analyzed, however, this value should be adopted pre 
accordingly to the material under study. In principle, TEM bright-field images, where the in | 
different precipitate types are marked by their chemical composition (as determined by the Z 
EDS measurements), can be used. This method needs more time for the data processing OT 
compared to the analysis of EFTEM images since for the latter a higher degree of nu 
automation can be achieved. The results of the precipitation analysis is shown in Fig. 5. 
478 nm 183 nm 
58 nm suas 
129 nm 
7 167 nm 
60. N Laves 
00 
S = LIV (MX) 
M,C. 
04 oe _ ps — 
TAF TAF FBS FBS x“ 
initial 100 MPa initial 70 MPa 
state 26913 h state 19436 1 . 
Fig. 5: Precipitations smaller than 1 um present in the investigated steels. jit 
The fraction of the different precipitate types observed are given in relation to the total Z-| 
precipitation number. In all samples, the most prominent precipitates are from the M23Ce the 
type. Their fraction ranges from 76 % in case of the creep tested TAF steel to about 99 % CO 
for the initial state of the TAF material. Remarkable is the low initial size of these Co 
precipitates in the TAF steel of only about 77 nm. But of even greater importance for the ter 
creep stability of the TAF material is the slow coarsening rate of these particles, which pr 
reached 138 nm after 26 931 hours. This slow growth prevents the degradation of the 
creep properties. A further positive effect on the creep strength is the dynamic precipitation 
of fine V(C,N) particles during creep. The analysis of the less creep resistant FB8 steel Ac 
yielded different results. The initial size of the M23Cg was relatively high with 98 nm, and Th 
they coarsened quickly. More important is the appearance of the Z-phase in this steel. This SP 
relatively large phase consumed all fine MX particles during creep, which is very fin; 
detrimental for the creep stability. In both steels, Laves-phase has been found after creep. ny 
Fig. 6 gives an overview of the growth rate of M23Cg at 650 °C in different steels developed tec 
within the COST projects along with data for the P92 steel [18]. This graph demonstrates 
that both, the absolute size and the growth rate, are very low for the TAF test material. 
This is usually attributed to the relatively high boron content of the material. The boron is Li 
built-in in the M23Cg particles and reduces their rate of ripening (e.g. [5, 7]). Additionally to . 
the slowly coarsening M23Ce precipitates, MX-type vanadium carbonitrides precipitate [1 
during the heat treatment and even more during creep of the TAF steels. The initial size 
and the growth of these particles is summarized in Fig. 6 for the two test steels and other 
COST steels. The MX particles in the TAF steel exhibit the smallest initial size of less then ' 
20 nm in the initial state and still only 41 nm after creep testing (Fig. 5). Additionally, the