Some Bloom to Bar

 My iron smelting partner, Neil, came up Monday for a scheduled one day 'bloom to bar' session. What with COVID, and honestly, Neil's limited experience blacksmithing, this turned out to be a combination of observation and trying out. I worked up two pieces, then later Neil did one himself. Although all this was using smaller pieces of bloom, there were good results all around. 

Partially sectioned raw bloom, plus finished 'currency bars' - Norway, Viking Age.

The results should at best be considered as only approximations of the overall conversion of raw ore into final working bars. Much attention has been directed to the first element of this process - that of bloomery smelting itself (ore to bloom). In truth, an understanding of the full cycle also needs to assess the important next step, bloom to bar. The desired finished product was never blooms, but the working bars sold into the hands of blacksmiths.  

There are a couple of major limits that need to be remembered when comparing the experiments here at Wareham, and how them might be applied to a larger understanding of historic iron making.

First is the relatively small size of the blooms produced. Intentionally, the normal ore amount used is in the range of 25 - 30 kg. The standard bog ore analog has a rough iron content in the range of 55 - 60 %. It certainly has been found that smaller ore weight smelts tend to have lower yield results than those using larger amounts. In effect, it takes a certain amount of ore to create an effective 'working base' in the furnace, and later additions of ore simple pack more and more iron on to the developing bloom.

The objective has been to aim for blooms in the 3 - 5 kg weight range. This is large enough to show effective production, yet also leaving individual blooms small enough to make later forging to bar possible at the hands of a single worker (admittedly  assisted by powered equipment!) Yields have been found to vary considerably however, the general impression is as the effect of air supply systems. The expectation is yields around the 20 % mark for these small ore smelts. (1)

A clear second exception is the use of modern forges and forging equipment. The Wareham Forge is a well equipped, if small scale, artisan blacksmith's shop. A two burner propane forge is commonly used for an initial 'heat soak' of the bloom pieces (to ensure full heating of the central core). Then pieces are transferred to a coal forge to bring up to full welding temperatures. As well as hand hammering, compression sequences are undertaken under a 30 ton hydraulic press. Fuller compaction and drawing makes use of a 75 lb air hammer. (2)

Bloom pieces from the 'Bones' smelt : June 2020. 

The piece seen to the left was one of those used (# 3).

A number of roughly fist sized pieces of earlier blooms were considered, most in the 500 gm range. These admittedly small pieces were chosen primarily for the ease of heating them, and also to reduce the amount of raw force that would be needed while hammering.

1) The first piece attempted was a smaller fragment from the 2018 'Gromps' experiment. The texture of the piece certainly suggested a high carbon content. At the first light hammer stroke, the piece completely disintegrated - a sure sign that in fact it was composed of unforgeable *cast* iron!

2) The second piece was part of the bloom from June 2008, one of the early tests of the 'Dark Dirt' bog ore analog. This was a specifically small ore smelt, with only 20 kg used. This seriously impacted the initial yield, although a considered a 'nice compact bloom' the total produced was only 1.9 kg (so 10 % yield). This section weighed 727 gm at the start - and had already had some compacting undertaken. The material worked down into a bar fairly easily, using a single overall sequence of compressing and compacting, with no major flaws developing. The end result was a bar 26 x 2.2 x 1 cm at 435 gm, considered ready to forge into some object. This then represents a roughly 60 % return from bloom to bar. 

Finished bars # 2 / # 3. Inset is the starting piece used for bar #2

3) The third piece selected was from the 'Bones' experiment - June 2020. This smelt had lower yield results than expected, a 25 kg smelt with 2.6 kg / 11 % return (3). The individual bloom pieces had been compressed then cut right after extraction via the hydraulic press. The starting weight was 577 gms. Through a full compression and compaction series, the resulting block still had some cracking and voids on the surface. It was prepared for a second weld series, being cut into two roughly equal pieces. At this point the weight is 382 gm, at roughly 15 x 3 x 1.8 cm, tapering in width to the two ends. This represents a bloom to bar return at 66 %, but it is important to note that some loss can certainly be expected over the remaining welding series. 

4) The last working sequence was carried out by Neil, who has limited forging experience. The piece selected was a fairly dense segment of the October 2005 smelt (very early!). This used a combination of industrial taconite and Virginia limonite, a roughly 19 kg smelt resulting in a 4.3 kg bloom / yield at 22 %. The bloom segment was 492 gm at the start.

The end result was a small bar, down to 209 gm. Measurements at this stage are roughly 8.5 x 2.7 x 1.2 cm (later addition from Neil). There were still some pits remaining on the centre surface, and a major flaw along one edge. The combination would require cutting and another welding series (as seen with bar #3). Neil was justifiably pleased with his 'first time' result, and decided to retain the bar as a lecture sample. The difference in the bloom to bar yield, here 42 % (at this stage) was put down to the relative lack of experience in the worker. 

It is clear that not too much in way of conclusion can be drawn from this limited experience. The starting quality of the blooms will obviously have a major impact of the bloom to bar sequence results. Because the of the intentionally small ore smelts certainly significantly lowers the yields at the ore to bloom stage, it can be expected that any overall ore to bar numbers are clearly distorted. 

Other tests rendering the DARC team blooms down into bars have also resulted in roughly 60% returns at this phase. Although not conclusive, this at least is suggestive of the kind of results than may be expected. 

A big thank you to Neil Peterson - for suggesting undertaking this working day. Overcoming personal inertia is a huge problem for me of late (if not for us all in the world of COVID).

1) This is certainly bore out by comparing smelts with different air systems and ore amounts.

When human powered bellows of various types have been used, yields typically drop dramatically, with roughly 15 % ore to bloom being a rough average.

When ore amounts are increased to 40 - 45 kg, so the yields also dramatically improve, into the range of 35 - 40 % ore to bloom (occasionally as high as 45%!)

2) Even still, all of these equipments also impose certain size limits. The die surface on the press is about 5 x 6 inches. The die on the air hammer is a still smaller 2 x 4 inches. The existing propane forge is limited to about 4 x 8 inches, The coal forge has an effective heat zone at best roughly 6 inches in diameter. The combined result is that any piece much larger than a half grapefruit simply is to large to effectively heat or work. (Another reason to limit smelting production to 3 - 5 kg.)

3) The best guess here is that the additional moisture from the several additions of bones with meat still attached might be the reason ( ?? )