276 Prakt. Met. Sonderband 38 (2006) 
It is worth to mention, that during preparation of TEM replica preferential etching of the Line 
various phases and agglomeration of particles after the etching process can occur and der 
modify the particle distribution. Therefore, it is important to compare the microstructure diff 
observed at replica with the one obtained at metal foils. Fig. 3 shows as an example a 
montage of four low magnification TEM micrographs of an extraction replica of the FBS 
material in the initial state along with the SEM image of a TEM metal foil. As can be seen 
similar microstructures are found. 
Fig. 2: OM images of the initial state of FB8 (left) and TAF (right) 
— ST — ET N Sum The 
Fig. 3: SEM (left) and TEM (right) images of the initial state of FB8 san 
28% 
The subgrain size and dislocation density of the steels was studied at metal foils in the aw: 
TEM, while the analytical characterization of the precipitates was carried out at extraction pre 
replica. The extraction replica were chosen for these studies, since no contribution from tec 
the surrounding matrix exists in the EDS data as compared to metal foils, yielding a more Spe 
precise determination of the chemical composition. Electron diffraction experiments can be on 
performed on both types of samples. However, the use of extraction replica prevents that inv 
the electron beam is deflected by the magnetic matrix, which makes diffraction est 
experiments at metal foils more tedious. Fig. 4 shows images revealing the subgrain size are 
and the dislocation density of the TAF material before and after creep. As can be seen the det 
subgrain-size increases during the creep-test, while the dislocation density within the ma 
subgrains decreases. For a quantitative evaluation of these parameters, at least 9 wit! 
micrographs of each type of sample were taken and analyzed. For the subgrain-size 2.2 
determination a magnification of 30k was used for each image corresponding to an area of the 
4.8 um? The value for the magnification was chosen to achieve a sound statistical 10