The influence of the die geometry was studied by e.g. Heil [7, 8], Vater and Heil [9], Haverkamp, Heil and Pill— kahn [10), Kopp and Tuke [11], Korneev and Petrosyan [12], Ambaum [13], Erman and Shah [14], Ono et at. [15] and Dudra and Im [16]. Forging of round sections in V- shaped dies may at low reductions be favourable compared to flat dies. The combination of a V-die and a flat die may cause cracking at half-radius. Other special methods are also reported, but forging of square sections with flat dies is a flexible method and the experience is good.

Of outmost importance at draughting is the bite ratio, i.e. the ratio between length of contact over the height of the bar. This parameter will determine the distribution of the deformation and the state of stress during the deformation. This will give the possibility of eliminating voids. Heil [7- 9, 17] studied the distribution of deformation over the cross section with flat dies and square specimens of 120 mm. His experiments showed that under the centre of the die and in the centre of the bar the local deformation was less than the average if the bite ratio was less than 0.28. A bite ratio over

0.45 gave a negligible increase of the local deformation in the centre. At a bite ratio in excess of 0.6, the elongation in the centre decreased due to increased spread. All bite ratios gave a local deformation under the tool edge less than aver— age. Similar results were repeated by Kopp and co-workers [18, 19]. Juretzek [20], made experiments with 30 tons in— gots. He recommended a bite ratio not less than 0.5 and more preferably 0.6-0.8. Kroneis and Skamletz [21) used 80 mm square specimens, with a 25 mm axial hole in which Metal Forming — Forging

20 mm long steel cylinders were inserted. They used 307o reduction and varied the bite ratio. At 0.25 all the seams be- tween the cylinders were expanded. At a bite ratio of 0.35 the seams were not expanded and some seams were welded together. At a bite ratio of 0.66 all seams were welded to- gether. Kopp, and Ambaum [22] made similar experiments with a round specimen and obtained similar results.

Ambaum [13] studied the elongation of a longitudinal hole at a height reduction of l5'7o. After forging the height of the hole under the centre and the edge of the tool was the same for bite ratios up to 0.35, where the height of the hole was reduced by 509'o. At higher bite ratios the height of the hole was reduced more under the centre of the tool, but was independent of the bite ratio at the edge of the tool. The width of the hole was increased at bite ratios below 0.4 but decreased at higher bite ratios, especially under the centre of the tool. Ambaum found that a split of the reduction into two passes improved the reduction of the hole under the tool edge. The reason was variations in the location of the tool edge. The width of the hole decreased because the spread of the bar decreased. Ambaum also found that in- creased forging temperature improved the closure of the hole, especially under the tool edge. The reason was the in- crease in temperature difference between centre and surface in the hot bar. Similarly the closure of the hole was better for a 129'o Cr steel compared to a 0.45 'c carbon  steel be- cause of the thinner scale and higher cooling rate of the al- toyed steel. These temperature gradient effects are impor- tant for small specimen size but may be negligible for forg- ing of heavy ingots.

Kopp, Ambaum and Schultes [23] studied the influence of direction of artificial defects. Defects parallel to the forg- ing force were hardest to close. Kopp and co-workers [18] also studied the variation in local deformation between dif- ferent cross sections. Their results show that even after nine passes the local deformation in the centre may show con- siderable fluctuations.

Many investigations show that the bite ratio is of outmost importance for the elimination of voids in the forging of heavy parts. But the definition of bite ratio is ambiguous. In order to avoid surface defects, the forging tool edge must have a radius. This will give a shorter contact length and a reduced active bite ratio. The simplest estimation of the ef- fective bite ratio is to reduce the length of contact with the radius [24]. The bite ratio will also change with the move- ment of the press. Normally the bite ratio is defined from the ratio at the initial height of the hot bar. This ratio will in- crease during the press movement. A higher reduction will thus give a higher average value of the bite ratio. Najz- abekov et at. [25] studied shear strains for elimination of voids in model material. They found that it was possible to eliminate voids with small shear strains. Their conclusion was that shear strains are the main cause of closing internal defects. Stdhlberg et. al [26] and Keife and Stdhlberg [27] used an upper-bound solution to study elimination of voids. They verified the results with plane strain forging of slip- line field wax. The conclusion was that the height reduction before turning of the ingot should not be too small. An ear- ly attempt to use FEM for forging was done by Shah et al. [28]. They  used  a program  named  PAPFOR  to study the elimination of voids. They found that a bite ratio of 0.6-1.0 gave the best elimination of voids, but the improvement when increasing the bite ratio beyond 0.6 was negligible.