S5003132

10

blouing has started a 'bole' ls produced duo Co charcoal consunption and thls ls fllled wlth ore and charcoal. Later the 'hole' ls fllled by push-Ing materiał froa the sldes lnto the 'hole'. The oxldlsed ore above the tuyere ls reaoved and placed above the 'hole' where the bloom ls forming. Ali thls reąuires an intimato knowledge of the posltlon and extent of the reducing and oxldlslng zones of the furnace. Normally, reductlon can only take place fifteen charcoal diameters beyond the end of the tuyere. Thls should be contrasted with a shaft furnace where reductlon takes place in the shaft well above the tuyere, and the tuyere level becomes hlghly oxldl-sing.

The Slav countries have madę one of the blggest contrlbutlons to Iron smelting research. Most of the work has been carrled out by Plelner (1969) in Prague, and Radwan and Bielenin in the Holy Cross Hountains of Poland.

In 1964 Plelner carrled out field experiments on the tali narrow shaft furnace - the Scharabeck type with slag pit. He failed to get this work-ing with lnduced draught. But the operation of the typically Slav (ninth century) Zelechovice furnace with its lnclined tuyere and blind alley was a great success (Fig.7).

Bielenin and Radwan in Poland concentrated on the broader Holy Cross Hountain type of slag-pit furnace. They failed to make the slag go into the pit below the furnace (as indeed did Thomsen (1963) with the Scharm-beck type) but mixed slag and metal were produced in a tuyere zonę.

Recently, however, it has been shown that by making an lnclined hole from the surface into the slag-pit, the air induced is sufficient to cause the residual charcoal in the pit to be burnt, and allow the slag to run into the empty space so produced (Fig.8).

After the slag had run into the pit, the shaft of the furnace was removed and rebuilt over a new hole beside the old. This method of operation left ninety-five enormous 'slag fields' containing over 4053 slag bottoms (Bielenin 1974).

In the 1960s BBC Televlsion encouraged some smelting experiments that were based on the Roman period furnaces at Pickworth in Lincolnshire and Ashwicken in Norfolk. These represented a common type of Roman period shaft furnace about 1.5 m high and 0.30 m internal diameter. After the field experiments recorded in the televislon programme, the furnace was noved into the laboratory and fully instrumented. The programme resulted in 32 smelts and it was found that with these furnaces very compact and homogeneous all-metal blooms could be madę weighlng up to 9 kg without the previous mixing of slag and metal. And this was done without any form of 'aanipulation', merely by continuous charging of ore and fuel in layers until the metal rosę to tuyere level (Fig.8) (Tylecote et al. 1971). At the same time many experiments were being madę by other groups on similar types of furnace in the United Kingdom (Cleere 1971, Adams 1979). Also a great deal of work has been done in Scandinavia (Thomsen 1963, Hagfeldt 1966), Austria (Straube et al. 1965), Hungary (Hackenast et al. 1968), and Germany (Ozann 1971), thus producing a great number of people familiar with bloomery techniąue.

(et Lead

Since lead is about the easiest metal to smelt it is surprising, or perbaps understandable, that little work has been done on it. It is possible to produce lead metal from galena (lead sulphide) in a wood-

fuelled. brlck-bullt brazler by charglng cobes of galena at the top, although the yield Ib not high (Tylecote 1964).

Thls ls eeeentlelly the techuląue adopted ln the 'boles' of the Pennlnee. The alags were reworked with Iron fluxee to extract the remain-lng 50% lead.

Plelner (1967) has descrlbed reconstruction experlments conducted on lead smelting. Two types of furnaces were used:    a bowl furnace and a

shaft furnace. The clay-llned bowl or shallow hearth was comparable to that used for copper smelting. The shaft furnace was also bullt from clay. It measured 70 cm ln helght and about 35 cm ln dlaneter. One tuyere was connected to a set of goatskln bel Iowa worked by hand. The dlameter of the tuyere was 4 cm. The total charge of 8.25 kg alxes of lead ore with Iron ore flux was layered with 7 kg charcoal. At the conclusion of the experi-ment a smali ąuantity of lead was tapped from the surface. Uhen the remaining contenta of the furnace were examlned, it was found that the slag had solidified round the tuyere zonę (Fig.10).

Horę recently experiments have been carrled out by Hetherlngton (1980). The furnace used was essentially a sllghtly modified version of the copper furnace Type B (Fig.3).

The ore was a galena concentrate containing 78% lead sulphlde, 3% ferric oxide, 2% silica and 2% alumina. Iron ore and sand containing 73-83% ferrlc oxide was used as a flux.

With an airflow of the order of 150 litres/mln tbere ls no dlfficulty in getting a temperaturę of 1200*C sufflcient to form and run a fayalite slag. However, at sucb temperatures lead oxide is very wolatlle (BP -1490*0, and there is considerable loss as fuming occurs. It became elear that unless one has a 'bag-house' or flue system for collecting dusts, lead smelting must suffer loss either in the slags due to Iow temperatures and poor fluxlng or, under high temperatures and good slagging practice, due to fuming. The two-stage process of Iow temperaturę smelting followed by high temperaturę recovery of lead from slag was probably the best process and was that practlsed, up to a few years ago, by the ore hearth and blast furnace combination.

As silver has a preference for the metal phase rather than the slag phase, all the silver in the ore is recovered by the first smelt and therefore the sllver/lead ratio is higher than in the ore ltself. This is an advantage that morę than outweighed the loss of the lead in the system.

It meant that recovery of lead from the slags was rarely wortbwhlle untll morę recent times.

Some work has been done on the cupellation process by R F Tylecote which was reported by McKerrell and Stevenson (1972) (see Table). It was shown that during the cupellation of silvez; tin, zinc and copper could be readily removed down to the 0.1% level while lead and bismuth tended to remain at the 0.5-1.0% level.

Since thls process is still used today a great deal should be known about it and perhaps this accounts for the lack of attentlon given to it.