Prakt. Met. Sonderband 38 (2006) 211 
rograph 
cipitates 
with the 
fraction, 
tions of 
esigned 
nethods 
> matrix. 
3S were 
2d to be 
d a size 
he heat 
nt cycle. Fig. 1: TEM/EFTEM micrographs of sample A7000. Zero loss image (ZL), Fe- and Cr jump 
ELNES) ratio images and V elemental map (V Lz; map) indicate different precipitate types. EELS 
Though and SAED analyses confirmed M23Cs (Cr jr), VN (V map) and Laves Phase. 
alread 
Y 3.2 QUANTITATIVE EVALUATION OF PRECIPITATE PARAMETERS 
;o0mmon 
ined by To obtain statistics for the precipitate populations of M23Cs and VN, Cr jump ratio images 
articles and V elemental maps were used. The numbers of Laves phase, NbC and mod. Z-phase 
were too low for a proper statistic. The number of M33Cs detected per sample was between 
/ in the 49 and 154, the number of VN between 18 and 101. respectively 
longest 
ne in Nearly all size distributions turn out to be left shifted with one single peak. Analysing the 
Z-phase M23Cs particles, one clearly recognises the shift of the peak position from 60nm in the 
ye so by sample “heat treated” to 120nm after 16000h of tempering. In contrary to this behaviour, 
ky et al VN precipitates only seem to grow until 4000h of tempering. After this sample condition no 
clear tendency is obvious. For further analysis it is inevitable to determine mean 
ansity of diameters, the volume fraction and. most important, the number density in the course of 
number the experiments. 
ne time, 
ı time of Very definite trends can be found when analysing the number of precipitates per unit 
ason for volume. For thermally aging and creep loaded samples, the number density of the 
precipitates is decreasing for both precipitate types (Figure 2 and 4). 
tatistical Although the statistic errors are larger when analysing the fewer VN precipitates 
ly MasCo (compared to M23Cs), the drop of the number density is eye-catching here: for the longest 
running aged and creep exposed samples the number goes down by a factor of 20 (creep 
exposed) and 10 (aged), respectively (Figure 3). This striking decline is probably one of 
the main reasons for the deteriorating long-term creep resistance