Reverberatory furnaces and their roof span

I was reading a new acquisition, a book about furnace design and construction, when I noticed that it had a chapter on furnace strength and problems with furnaces, so I naturally wondered how it compared to historic reverberatory furnaces and their construction. Basically, a reverberatory furnace is where the fuel is burnt at one end, and usually by a sort of chimney effect, hot air and gases are brought over the top of the metal to be melted, the heat reverberating from the roof of the furnace. I really want to build and operate one to make copper alloy castings, but that is quite a big project.


These two photos show top and side view of Biringuccio’s reverberatory furnace design:

This type of furnace goes back to the 15th century or earlier in Europe, probably around 1500 in England and Scotland. It is built with bricks and clay and mortar and I think should end up with a lot of movement of the bricks in the roof as they expand and contract with the furnace heating up and cooling down, potentially causing various problems. As it says in Industrial Furnaces, Vol 1, by Trinks and Mawhinney, published in 1961 (5th edition, first came out in 1923), page 292, this first quote not necessarily covering reverberatory furnaces but is I think useful to note,

“… it is desirable to consider the places where failure is most likely to occur. Prominent among such places are roofs. Wherever material spans an opening, there must be tension in some part of the covering. Because hot refractories are extremely weak in tension, it is usually taken up by cold steel lying outside of the furnace. The refractory part of a roof can then be put wholly into compression.”

It carries on in page 293, more relevant to our needs – “Between room temperature and furnace temperature of 2200F, fireclay bricks expand about 0.075 in. per ft, or about 0.6%. The inside of the arch becomes longer by that amount, while the outside maintains practically it’s original length. Several things can happen under the influence of this expansion. The abutments can be pushed apart, the bricks can be compressed, or the arch can rise.”

Page 296 – “It may be stated however, that roofs of heating furnaces seldom fail from crushing of the arch bricks. Occasionally a very hot luminous flame melts an arch down; in other cases, ash is blown against the roof, sticking to it, and acting as a flux which causes the roof to melt; but as a rule, the rather rare collapsing of an arch is due to other causes, the most common of which is the yielding of the abutments (Skewbacks), and pinch spalling.”

(The spalling seems to result from rapid cooling of the bricks so different areas contract at different speeds so surfaces flake off, if it happens often enough the bricks lose strength)

This prompted me to look at the main historical sources I know of about reverberatory furnaces and more importantly, I happen to s in English translation. These are the works by Biringuccio (1540) and Cellini (1568 or so), both of whom used reverberatory furnaces in the 16th century. I don’t know of any other accessible sources for the period, and there are not many after either although I am sure they were written about. (Makes note to see if early modern sources can be found) So I read through the relevant sections of both to see what they say about the pitfalls of building a reverberatory furnace.

Cellini is not much use, declining to give a drawing of a furnace, then spending a lot of words describing how to make the very important floor of the furnace properly so it does not break under the force of the liquid metal. When it comes to the roof, he merely wrote “After the bed you build up the vault with similar bricks in the same way, … “

He is very picky about the properties of the bricks that you must use, so we can assume the ones he wants us to use do last well in the heat. Most interestingly, his design seems to let the flame and such out the far end of the furnace at the roof height. More modern designs assume you have a chimney of some sort at the far end of the furnace away from the firebox.

Biringuccio is also not much use, not describing any particular issues with roofs of furnaces.

He wrote:

This is made of baked bricks, or of crude ones if desired or of suitable fire resistant stones. … (2 pages later) I have sometimes made them with bricks and laid them with the white clay that glassmakers use for making their pots and furnaces. Peperino also servers very well and a black rock flecked with white spots of the talc inside, whoe name I cannot tell you but it is found in Valcomenica in the Brescian region, …”

The vault itself is described thus:

Thus built in this shape, I have had the vault begun not only over the firebox but also over the place for the bronze. I have taken pains to arrange that the vault of the firebox should be somewhat lower than that of the furnace, and that the aforesaid convex should be somewhat lower than that of the furnace, and that the aforesaid convex be commenced as desired from the part of the wall where it abuts a little above the level of the arches, curving in such a way that all the flamed, beating against it, press against the opening of the window that leads into the furnace. Likewise I have also had the vault made low above the taphole, so that when the flames are beaten back they may fall forcibly upon the metal.”

But that is about it. Nothing useful, no personal experiences or anything. At the moment, it looks like there wasn’t much of a problem with such furnaces and their roofs.

There is of course a clear assessment that you need furnace bricks which will survive the fire, and it is implied that if you do not, then things will go very wrong. But then if you use the correct bricks then things will go well and your furnace will work. Moreover if you look at the thickness of the walls in the drawings near the top of the post you will see the furnace is intended to be very strong and therefore should resist the walls being pushed apart by the expanding bricks in the vault, even if Biringuccio doesn’t make much of it.

Of course one difference is that nowadays we build to last and to minimise maintenance. Back then I suspect they rebuilt parts of the furnace after every few firings but finding any evidence of this is difficult. Also, if you did carefully choose the right stone or bricks to make it from, and built it well, there would be no problem with the roof pushing the side wall out, or of spalling. A roof to keep the rain off would definitely be a good idea though, but it seems that such a weak spot in a furnace was not a major issue or it was assumed you would add a roof of tiles or temporary one of wood.

Indeed I have read of one in Italy I think, which was in the basement of a large building, and therefore completely protected from the weather.

What I actually did on the foundry at Kentwell in 2017

This year, we had four of us, all having been to Kentwell before, and two with well over a decade each under their belt.

I was head of station, and often referred to the others on the foundry as Master Alexander. Which was especially nice when I had just corrected something or made a useful comment about things. My many years experience has given me a good foundation of knowledge and confidence about my own capabilities and knowledge.

The great thing about having experienced people there was that all I had to do was mention a few facts and answer their questions and they immediately incorporated it into their spiel to the public. I was initially a bit stressed on Sunday because it was the first day and I didn’t quite know what I had to work with, but the three of them swung into action and by Monday were talking convincingly to school children about various things.

The week did start a bit badly, because I found on Saturday night when dropping my stuff off (After driving 430 miles, so a bit tired) that someone had adapted the furnace to look like this:

Kentwell 17 furnace front top


Continue reading

What I did on my holiday this year – bronze casting

Once again, I was on the foundry at Kentwell hall in Suffolk.

This time, I was the master of the foundry. Unfortunately there was a slight lack of assistants, so there were 2 of us most days. Still, I managed to get a few things done and refine my understanding of the processes and equipment.

Last year, which I meant to write up and publish but never really did, involved a group of us trying to make cannon. We failed, due to lack of time and lack of appreciation of the difficulties. Replicating the techniques of 16th and 17th century foundry work, even with at least 3 of us having a great deal of expertise, was something that would have taken more than the 8 days that we had, and so we failed. Just two more days would have been enough though. Which isn’t bad, all things considered.

For those who don’t know, this is the foundry at Kentwell, first the bellows end and then the furnace end:

bellows end of kentwell foundry 2016 furnace of kentwell foundry 2016

Continue reading

Maybe a sighting of double action bellows?

In Lazarus Erckers book about the analysis and production of various minerals and salts, there is great use made of furnaces. But most are natural draft powered. However there is one to be found later in the book, I stumbled upon it by accident. The accompanying text merely calls them bellows, for the purpose of assaying copper ore.

And the picture is:

POssible double action bellows

From page 219 of the hardback of Sisco and Smith’s translation of the Treatise on Ores and Minerals.

The key point here is that it shows a man reaching into the furnace with tongs whilst manipulating the bellows with his other hand. Moreover the end of the pole seems to be attached to a rod which pulls up the bottom of the bellows and there is a valve on top. Of course the small issue is the presence of the valve, which would suggest that air goes in the top, when the bottom is falling down.

Modern double action bellows work by having a valve inside, splitting the top from the bottom, so that when you pull the bottom part up it pushes air into the top part, which then expells it out the front due to the force of gravity on the top piece of wood, whilst the bottom section is opening out again.

So in this case I don’t think I can say it definitely is double action bellows because of the apparent valve in the top. This image has pretty much everything you would expect for double bellows such assingle handed operation and only one set of bellows, but I can’t really be sure. I shall keep my eyes open anyhow.

Finishing the casting off so it is nice and shiny

A quick foreward to this post – I originally did some “lets see what happens when I try some methods of finishing castings off that I think they might have used” last year, about 5 minutes worth. That convinced me that using files and stones was possible. Then of course I realised it would be a good idea to do some research as well, using my well stocked library. Hence this post, which follows on nicely from the previous one.

So, now you have a nice cast object. It is still discoloured and a bit rough from the casting process. What do you do?
To answer this I turn to the usual two printed sources that bracket my period of interest, Theophilus and Biringuccio.

The 12th century monk Theophilus describes the use of files to finish a cast silver chalice off before engraving it.
More interestingly, he writes, page 175 (Of the Dover paperback edition), after casting a bell and bracing it on the lathe, “…the bell can be turned and smoothed all over with a sandstone.” Which is surely good evidence for the use of a natural and often freely available precursor to sandpaper! You can get different types of sandstone too, some fine, some coarse, which would give you great ability to smooth it out properly.

Earlier, he writes of the making of burnishers, which are for finishing the work, they are like scrapoers but more rounded. They are still in use today for smoothing out the surface of a metal object by plastic deformation.

You should also, (page 93) have files made out of pure steel, hardened with a form of carburisation. Large, medium sized, four cornered, three cornered and round. The descriptions are not so clear, but he does refer to making finer files “.. work whic h has previously been filed with other coarser files should be smoothed.”

Obviously I need to expand my toolkit, get some more files and burnishers as well as different grades of sandstone.

Biringuccio writes on page 377, regarding the finishing of flasks, saltcellars etc, “With rasps and scrapers and other cutting tools they are smoothed, polished and made beautiful.” I get the impression that the finishing of metal items is so well known about and widespread that he doesn’t see any need to go into details, whereas the art of casting in bronze and such is tricky and difficult and he wants you to know all about it (And that he knows all about it).

I have an interesting book on English pewter, called “A history of British Pewter” by John Hatcher and T. C Barker, published in 1974. It has an outline of the manufacturing process, and it is clear that you used the lathe for it as told by the period sources, with several types of iron tools similar to those used by wood carvers for burnishing the surface. Interestingly though you also hammered some types of wares, like plates.
But it obviously only covers pewter.

Another source is the treatises of Benvenuto Cellini, of the mid or later 16th century. Page 57 of the edition I have talks about finishing a work of art, some sort of cast and soldered together object, and he wrote “I took some four or five hard pointed stones, which are sharp at the ends and thicken upwards in the manner and I used them with some well powedered pumice stone. The object of using these stones is to take out the markes of the steel tools, the punches, chisels, files and suchlike and to give it a fine uniform surface; and last, but not least, a brilliancy of colour which would not be so easy if the marks of the steel tools (And the skin they make) were not obliterated.”

Page 94 tells us about the cleaning up of a large silver statue. He used pumice to clean and smooth it, then heated it to red hot and put it in a blanching solution of water with tartar and salt. Then it was scrubbed and when the silver started going white it was put in another vessel with clean water, to wash the blanching solution off. After that it was ready and clean for gilding.

Between them all that gives me enough information to improve my displays and production processes. But as usual, practise is required.

For contrast, here’s a photo of a later medieval bronze cooking cauldron from Dublin. Note that it isn’t really well finished off and clearly shows the flashing line where the outer mould was split in half to enable it to be taken off the mould it was shaped on.

Dublin late medieval cauldron

The origin and use of bellows, especially in medieval Europe

I’m sure most of you are familiar with bellows, through seeing small, often ornate and decorative, sets of them by fireplaces. However the shape that they are today originated comparatively recently, becoming widespread in Europe in the medieval period.

The origin of bellows is well back in the prehistoric period, but they were made differently then. The first metal smelting, around the 6th millenium BC, that is, the 5,000’s BC, was done in draft and wind furnaces, that is like they fired pottery, and at that time or shortly after they realised they could smelt using blowpipes to introduce more oxygen. (

Of course the problem is that this required lots of blowers to keep up the air rate, and the air ends up being damper and lower in oxygen because it has been through someone’s lungs. So someone invented bellows, or maybe several people invented them in different places at different times. Archaeologists have spent many backbreaking days operating bag bellows, made out of the skins of animals sewn together with a hole which can be opened and closed by movement of the fingers.

At the bottom of the bags are pipes called tuyères sticking out of the bottom, leading into the heart of the furnace. These are usually made of clay and sand or muck of some sort or another, so as to withstand the temperature of the fire, although the nozzles do tend to get fused and glassy. Finding them in an excavation is usually a good indication of a metal furnace, although a better one is finding the remains of the hearth itself, contaminated with traces of metal. People back then though did like to tidy up every little droplet of precious copper that they had.

It is a little hard to see what type of bellows are in this Old Egyptian wall painting:

They look rather pot or more concertina like.

These are more efficient and easier to use. Still backbreaking work though.

Or perhaps they were more like this sort of iron age bellows:

Scroll down to find photos of them in operation – essentially the leather sits as a lid on a pot, and has a hole in it which can be opened or closed, and the leather is pushed up and down.

As for the Greeks and Romans, it has proven remarkably hard to locate images of Hephaestus or Vulcan, the Roman equivalent. Those I can see tend to show him with a hammer and anvil, no sign of bellows.

The same fellow, Dave Budd, reckons that, pot bellows were in use at this time, and also at some point kite bellows, although it is not clear from his photos exactly what they are. The internet turns up this:

Which is a kits shaped bag bellow, which fits later descriptions apparently.

After the fall of Rome, obviously less evidence is available in Europe, but to judge by Theophilus’s work On Divers arts, the type of bellows in use in northern Europe in circa 1122 was still the comparatively primitive sort based on a bag, no different to those used in previous centuries. They were to be made this way:

“When the rams are killed, the skins should not be slit under the belly but opened from the rear and turned inside out so that they can be stripped off whole. (He tells you how to dry and cure them) Then tye should be greased and stretched againl. After this make a wooden head for the bellows to go through the neck and be tied there and make a hole in the head through which an iron pipe can go. Now, along the width of the bellows at the back, place four pieces of wood, which should be fitted to each other in pairs and tied together in the middle. Each pair should be sewn onto the bellows so that the places where they are joined are at the top and bottom of the middle [ of the opening of the skin]. At these points also two loops of the same skin should be sewn on, a smaller one at the top to hold the thumb and a larger one at the bottom for the four fingers.” (Page 83-84 of the Dover reprint)

This setup is for a small scale worker in metal. I’ve found online a 9th century AD illustration from the Stuttgart Psalter, via mention of it on a forum (f 121r):

Note that the two smiths are beavering away whilst Moses is coming down from the mountain and the others are worshipping a golden calf. It looks as if one advantage of this design of bellows is that you can operate the set with one hand whilst moving things about in the charcoal with the other.

Naturally there might be differences in technology between the UK and Germany, but really, given what was available and the communications of the period I think it unlikely that the UK was any more advanced in terms of bellows than Germany.

The earliest illustration of what we know of as bellows that I have found mention of is the Hylestad church carvings:

The church was estimated to have been built in the late 12th to early 13th century, and the illustration clearly shows a pair of single action bellows being used to run a smith’s forge. Note two tuyeres, one from each bellows, running into it.

So we have the modern sort in the late 12th/ early 13th century, compared to bag bellows circa 1122, which narrows it down nicely to about a century in which it seems the modern bellows were invented and spread about Europe. Perhaps rather rapidly as well, since illustrations abound from mid to late 13th century onwards:

from 1250, courtesy of the Larsdatter links page. The problem there of course is that these are just the small set of handheld bellows you use at home. But at least next time you are blowing on a fire you can think of how old the design is, over 700 years.

This circa 1326 illustration is the oldest one Larsdatter has of the use of bellows on a forge:

which makes it clear that the same design as for domestic use was in industrial use by this time.

Later illustrations are usually a bit clearer, and show better working too. For instance, the Holkham bible, English, circa 1327-1335, showing the same sort of bellows as we expect, so they were in use across Germany, France and England:

You can clearly see that the illustrator has tried to draw a tiller to raise the bellows up and down alternately, but hasn’t succeeded. At least he drew the rope connecting the sticky out bits of the bellows to the pivoting piece of wood, and shows what are two rocks on top of the bellows, presumably to press them down and get a good blast of air out of them.

(As an aside, an ex-member of my re-enactment group had some bellows made, but didn’t have such handle like extensions on the end, claiming there was no evidence for them. Oddly enough most illustrations I have ever seen do show them, and without them it is much harder to operate the bellows, so I don’t know where he got his ideas from)

Also a couple of decent pictures here:

A 15th century German illustration from the Mendel Hausbuch showing only one set, but that is often how they are badly drawn:

The advantage of using two sets of bellows is obvious. Arranged the way they are, one deflates whilst the other inflates, leading to a steady stream of air through the fire.

Several modern re-enactors have made such systems, such as this seen at Tatton:

Tatton 2007 pair of bellows and frame

And this at Kentwell, copying Bringucio’s “De Pyrotechnia”:

better rear view of Kentwell bellows

My own bronze casting setup at events is like this:

melting furnace Lanark 2010

I still need to make the proper rocking setup to make using them easier, and make a good strong stand for the bellows.

Meanwhile, domestic bellows continued to vary in shape, e.g. this 1520’s set:;jsessionid=F5E05BD40C4638C7904E80D6DBF78848?trs=3&mi=0&qvq=q%3Abucket%3Blc%3AODLodl~14~14%2CODLodl~1~1%2CODLodl~23~23%2CODLodl~24~24%2CODLodl~6~6%2CODLodl~7~7%2CODLodl~8~8

is distinctly rounder in shape than some earlier ones appear to be. But that was probably a matter of personal preference.

Another mystery to me at the moment is when double action action bellows were invented. They are the ones you can still see in historic blacksmiths. They have two sections, a top and bottom, with a hole between and a valve in it. The rope pulls up the bottom, which pushes air into the top section, which is pressed down upon by a weight. The advantage of these is the constant flow of air and the fact you can pump them up more than the older type, leaving them to deliver air for long enough that you can guddle about in the fire.

The earliest illustration I have seen so far is perhaps this:

Which dates from 1689.

Unfortunately you see them all the time at re-enactment events, even in the medieval period. But I can find no information on them before the 17th century, the problem being compounded by some potential confusion on wikipedia between double action piston and accordion bellows. I have certainly seen piston bellows in a late medieval picture:

And vertical accordion bellows seem to be a continental thing of the late medieval period, as you can see on the far left of this engraving:

But no sign of double action bellows, not in Biringuccio, (1540), Ercker (1580) mentions a pair of bellows, so obviously not single one, and I don’t have any more sources for later than that.

You can get instructions online to build them, and if you wanted a hand blown furnace I would make them.

On the other hand it is much easier to use this:

melting furnace with two airbed pumps

At the right you can see two small black air pumps powered by two 12v batteries, which deliver enough air to melt two or three hundred grams of metal in ten minutes. The photograph shows the furnace actually at 1100C in the hot part, but you can’t see the white hot area because the charcoal is in the way. It acts as insulation, helping keep the heat in.

Another option is a hairdryer, which would move much greater volumes of air, and they are frequently recomended in books about home foundry work, the problem being they need a 240v supply and you have to be wanting to melt a lot of metal, since if you put too much air through the charcoal you won’t get the best reactions and will blow some of the heat out.