_ 
a 
liquid nitrogen. During the final stages of the precracking 
the maximum stress intensity range was IK, — 13,7 MPa 
m (for Spec. I) and. X Ke — 15,8 MPa m (for Spec. II), 
respectively. The fracture toughness of the two specimens 
were a little bit different: in terms of the critical stress 
intensity Kyo = 25.2 MPa Vm for Spec. I and Kjc - 29.6 
MPa ^m for Spec. II. 
From the corresponding regions near the midsection of 
each specimen half stereo image pairs were produced by 
the tilting method with a scanning electron microscope 
(SEM). 
The magnification scale of the SEM-micrographs was 
about 1000 x and the tilting angle was 10°. This leads to 
stereo models, which give an acceptable accuracy and 
intersection qualitiy for stereoscopic measurement on an 
analytical plotter. 
Photogrammetric Orientation 
The photogrammetric orientation of the stereo models was 
done on a KERN-DSR analytical plotter and yielded a 3- 
dimensional local coordinate system. The main task of the 
current investigation was to find the orientation of the two 
halves of the specimen relatively one to another. The 
accuracy of this depends on the possibility of finding 
conjugate points on both surfaces of the two counterparts. 
The selection of these points was done under the influence 
of a plastic deformation of the specimen during the 
cracktest. That means, that only small undeformed 
structures can be used. The region inside of some cleavage 
facets was appropriate for that purpose and a number of 15 
- 20 corresponding points could be defined on both parts. It 
is necessary to remark, that the destortion of the 
microscope (which was determined with a calibration grid) 
was found to be smaller than the pointing accuracy. The 
tilting angle between the stereo images was taken as an 
instrument-parameter directly from the microscope. 
  
Fig.1 SEM - Micrograph of Specimen I at a 
magnification of about 1000 x, broken in liquid nitrogen: 
(a) side 1, (b) side 2.