Prakt. Met. Sonderband 38 (2006) 225 
as the flow curves and the thermal properties of the material were used as inputs in the 
simulation. The values of friction and heat transfer between the anvils and the specimens 
‚ 0.011N, were adjusted to obtain the shape of the sample and the gradient of temperature during 
r for this compression and thus, the distribution of effective strain and effective strain rate. Fig 2 
shows the distribution of the effective strain along the specimen simulated with 
with the DEFORM™ 
‚E 
c system 4 
of 768°C Sirala « EMactivg 
Ip to the 8: 8178 1.7 
minutes Pe in g 12 
nocouple F- 9000 £ 10 
used to =F 2 0,87 £ 0.7 
re of the 3 0.6. ie 
e friction EY ) 
long the 034 
sured by 0.2- 
hermore, 00+ EE 
768 and o TORE U ? 2 ; Z 5 6 
osphere. a) a 18 b) Radius [mm] 
Fig 2. Simulation of a deformed sample at 788°C and 1s°' showing a) the distribution of the 
effective strain within the specimen and b) the effective strain as a function of the distance 
along the radius perpendicular to the load in the centreline of the deformed specimen. 
3. RESULTS AND DISCUSSION 
3.1 MICROSTRUCTURE AS RECEIVED 
The as received material presented a bimodal a + 3 microstructure as a result of the pre- 
forging process (cogging) consisting of four steps in the ß and a + f fields. Fig. 3 shows 
the distribution of the primary a phase and the fine acicular a phase formed inside the B- 
sub-grains. The content of primary a. was around 42% 
ed. They 
scattered 
tures for 
1e mayor 
s40 v4.4. 
such as 
ording to 
from the . 
s a, um wv. 2pm _ 
2 values Fig 3. Microstructure of the as received material a) primary o. phase distribution, where a is 
sion test dark, and Bp sub-grains bright. b) Details of the fine acicular a. phase and of the different 
s as well contrasts of primary alpha due to the variable alpha orientations (IBE etching).