- —e—8 
20 — : ' 
mes —%—8 
a z 
a cen 8 Los 
— Le z2 
= Ta ıl_e—C 
10 ka- - u az cm 
% 5 a = a Com 
? | oo | - 
? 612 5 : YS 
5 | - a U C 
3 4 ' gs 
= ELLE .. vk 
AT ooo] 
aA EEE S 
? re AS, 
Te S, 
i =! ey Sa 
0,0 0,1 0,2 0,3 0,4 0,5 
NER true plastic strain 
ter J Th, . on 
ain and te Fig. 3: Roughness parameters as specified by Topograf 
un meld | evaluation of data by Hommel-tester versus plastic strain 
vatlable usi . . oo . 
- be A linear relationship is found between these variables which holds even when the range of uniform 
LCL . . . . . . . . 
i ot . elongation (true plastic strain of 0.4) is exceeded. Considering the experimental scatter by drawing a 
iol i scatter band including the data points by two parallel lines, the experimental error of strain derived 
\ . N in from roughness measurements with an investigated area of 1 x 1 mm? is estimated to +/-0.05 as 
A shown in Fig. 3 by the construction outlined in bold lines. 
RLS cada For methods 2 and 3, the experimental error of the mean diameter of the RSSE and of the linear 
ad intercept of the grains depends on the evaluated area. As both the number of RSSE and of intercepts 
were do . . . . - 
tele dd os increase linearly with area, a square root dependence is expected and actually observed as shown in 
tan 0 . . . 
Sa of Fig. 4. Methods 4 and 5 both have a comparatively very small experimental error. As the 
EE magnification of the image can be adjusted to needs, the resolution is only dependent on the size of 
iron features which can be identified in the images. 
2 anton For methods 2 and 3, Fig. 4 allows to estimate the area necessary to obtain a given relative standard 
Se 2F 4 adoro error or to define the area needed to be investigated for achieving a preset accuracy. Though the size 
of RSSE is nearly uniform while the size of grains has a large distribution in itself, the area 
0,10 — 
0,09 | : Method 
=» 3D roughness 
0,08 | ” m RSSE diameter 
5 0.07 3 grain size 
5 0.06 ” .a comp. RSSE 
3 ’ 5 A comp. microstructure 
£2 0,05 
5 
= 0,04 
3 0,03 - 
X 
© 0,02 
0,01 {A —_ 
0,00 — - nn DL 
0 . 3 ‘ 5 
area [mm’] 
Fig. 4: Experimental error of strains for the different methods 
67