Prakt. Met. Sonderband 46 (2014) 325 
4. METALLOGRAPHIC INVESTIGATIONS ON CROSS SECTIONS 
Cross sections where prepared for metallographic investigation using the Optical 
Microscope Zeiss. The applied etching solution was Nital. The samples were investigated 
in the as received condition, including the corrosion products on the fracture surface 
situation at the crack start position, see Fig. 6 left. A very thick corrosion layer of corrosion 
product, corrosion pits and intergranular separation in the coarse grain region of the HAZ 
can be observed. This observation is in full coincidence with the observations made during 
the LMI and SEM investigation; compare Fig. 3 and Fig. 4. 
) in Fig. 3 
1g”, appears 
ctly after the 
ding. 
a” The grains 
kable. When 
1S, which are 
ervice of the 
ition of the 
cks” can be ome 
emptying of Fig. 6: optical microscopy crack start area (left) and deformed area (right) in Fig. 3 
Very interesting observations were made at the cross section of the so called deformed 
area. Fig. 6 right reveals the surface area of the fatigue crack which shows traces of heavy 
plastic deformation. Furthermore deformed material separations can be observed. The 
plastic deformation of the surface area can be explained by a squeezing effect caused by 
crack closing when emptying the pipe. Here also the branching cracks, which show traces 
of micro striations in the SEM investigation on their surfaces, are visible in their deformed 
condition. 
In this connection it is to mention the fatigue crack depicted in Fig. 1 is “open”. This effect 
can be explained by the elastic relaxation of the test piece after its removal from the pipe. 
The explanation of this appearance, so called “Macro striations”, could be found in the 
presence of the textured micro-segregation occurrence in the rolled plate which will have 
an influence on the fatigue crack propagation. 
5. CONCLUSIONS 
In the pipe where leakage caused by fatigue crack propagation was observed, a hydrogen 
induced cold crack of about 1-2mm depth, has already been present in the weld from 
beginning of the service of the penstock. This crack served as crack starter of the fatigue 
crack. This could be confirmed by the observation of heavy corrosion pits and much 
thicker corrosion product appearance in this surface crack area caused by the much 
longer exposure time acting. 
propagation So it is not surprising that in just that position of the penstock where cracks already existed 
nly in a very from begin of service. the first leakage appeared after short service periode. 
crack to the 
ig. 5.