News has filtered through that we are going to have close the current experimental facility at St Fagans in the very near future. A stay of execution has been granted until after the National Science Week activities.

The furnace has survived the winter well, despite the deteriorating condition of the shelter. The problems experienced with the low blow rate during smelt 25 left the furnace with relatively little damage, so repairs today to get prepared for smelting later in the week were very light.

The first of two smelts planned for National Science Week 2004, smelt 26, took place. The day was intended to be used to solve the low blowing rate problems experienced during smelt 25. A minor problem with the attachment of the tube to the blower was identified as a possible problem - and rectified. A further problem with smelt 25 may well have been the somewhat damp fuel; but new supplies of charcoal from Gower ensure good fuel this time. The plan was therefore to progress to a full smelt only if the starting phase went well, otherwise the day would be spent in troubleshooting.

In the event the smelt started well (despite the maintenance of a wood fire for longer than usual whilst we awaited the key to the electricity supply). The furnace was clearly burning well, so we progressed to charging 1h:04m after starting the blow. We used 22.4kg of ore over 1h:23m, with a burn rate of about 14kg/hr. The first tap was made at 37m after the last ore charge (4h:00m). The slag flowed very freely and was ponded above a "hearth bottom" with a free space of about 20-40mm below the base of the bloom. Apart from a sample retained for demonstration purposes, the slag from this and the subsequent taps was returned to the furnace top. The subsequent taps at 4h:25m and 4:56m also produced voluminous (probably 3-4kg each time) freely flowing slag. After this the slag became much viscous and the bloom-furnace bottom gap started to become occluded. A tap at 5h:42m yielded viscous slag and it was decided that should be the last tap. However, by 6h:56m the slag puddle on top of the bloom was threatening to block the blowhole, so it was decided to make one more tap to try to drain this a little. This seems to have had some effect and the blowhole was kept open for the remainder of the burn-down. The bloom was rather well stuck, and had cooled rather by the time it was removed. The furnace had suffered quite a large failure of the blowing-side wall - which will need repairing before the next smelt, but otherwise survived well.

After the rather balmy day on Wednesday, today marked a return to stormy conditions. The plan was to do some smithing, and the first task was to cut the bloom from smelt 26. Aaron took the bloom away to cut and returned with an unexpectedly good-looking section. The bloom appears to be around 4.5kg, with the whole mass about 6.8kg. Most of the difference was made up by brittle, dominantly glassy slags, on the upper surface of the bloom. Near the iron the slags take on a somewhat khaki colour, with some very bright green glass immediately adjoining the iron.

One half of the cut bloom was crudely cleaned (weighing now 2kg) and bloomsmithing was started. Aaron decided that we could proceed with using a steel hammer from the start, rather than the usual wooden mallet. This decision was certainly supported by the integrity of the bloom when being worked, although there was rapid loss to the hearth initially. Over 11 reheats the bloom was reduced to a circular billet weighing a little over 1kg. This was divided for sampling, and a piece weighing 620g was worked on. Over a further 16 reheats this was worked to a crude elongate billet of about 440g. The metal worked well at high temperatures, but started cracking. One major crack broke the bar and the two fragments were successfully welded back together. The forging to bar was not complete when work was stopped.

The second of the smelts for National Science Week 2004, smelt 27 was undertaken with yet another day starting off with high winds and rain. The plan was to follow the design of smelt 26, with the exception of not recharging the tapped slag. The fire was lit by using embers from the smithing hearth rather than kindling a new fire and this may possibly have affected subsequent progress. However, for whatever reason, despite and initial burn rate of close to 20kg/hr the rate dropped to around 10kg/hr after about an hour. This change seems to have occurred just before the top gasses were lit and charging started at 1h:14m. The most likely explanation is that the third bag of charcoal used was either damp, or had a high fines content, or both.

Charging was rather slow, because of the reduced burn rate, with 19.8kg being charged over 1h:30m (i.e. burn rate of about 12kg/hr). The burn rate was apparently helped slightly by more careful selection of the charcoal going in. The first opening of the tap arch was at 2h:32m, during charging, just to see what showed - and nothing did. The first proper tap was undertaken at 3h:15m (30m after end of charging) and a good fluid flow of around 4.3kg was obtained. A third opening at 3h:44m found a further 4.7kg of cooled flow behind the tap arch blocking, but no fresh fluid slag. The burn-down started at 3h:54m and was complete by 5h:28m - i.e. 1h:34m, which is probably the fastest and easiest burn-down yet. The bloom was still very hot and removal was relatively straightforward - with only a few small bar holes needed from above, then manipulation through the tap arch to put it on end, so it could be grabbed from above with the tongs. The slags around the bloom were still quite fluid at this point.

The bloom was rather irregular in appearance and the total mass weighed 6.79kg. Of this, 1.65kg of brittle slag from above the bloom separated very easily and a further 1.5kg of slag was removed from the bloom with a hammer, leaving a crude bloom of 4.61kg.

Today saw the removal into storage of the contents of the smelting shelter. The furnace has been left standing, with a view to monitoring its survival once in the open. The removal of the tarpaulin and the scaffolding in likely to be next week.