You are not Logged In! -- Please consider creating an account. (It's free!)
This is the BETA version of the Articles Library -- Expect occasional bugs -- Report them to Daemon_Lotos => [Here]
[ Articles Home | Newest Articles | Submit an Article ]
[ Random Article | Search Articles ]

Making Roman-Like Riveted Maille
Article © MAIL User: Neuralmancer

Making Roman-Like Riveted Maille

Lee Holeva

December 27, 2005


One of my first desires within the hobby of roman reenactment was the obtainment of a shirt of roman lorica hamata, a long maille shirt. Reconstructions of these usually consist of butted rings in the usual 4-in-1 pattern. While there may of been some use of butted maille during roman times, a more common construction method was the use of alternating rings of riveted and solid rings. The riveted rings were fastened together using a round rivet. Solid rings were either stamped from sheet or welded together from wire. Compared to most medieval maille, roman maille was small. Typical ring thickness ranged from 0.8mm to 1.5mm and outside diameters ranged from a tiny 3mm to at least 12mm.

My next step was a search of the internet for instructional information on how to make suitable roman riveted maille. Most of the instructional material I found concerned the fabrication of medieval maille. Starting out I knew nothing at all about maille and its construction so my first step was to tailor the techniques that I found to the making of something closer to what I wanted: roman maille. Though I was willing to sacrifice a degree of historical accuracy, I wanted to go beyond the use of butted maille. Ninety-nine percent of the fabrication literature I found describes the uses of wedge, not round rivets. I concluded that this was a minor detail. I discovered that once peened, taking into account the burr resulting from the punching operation, the wedge rivet forms a round semi-spherical head on the outer surface of the ring, appearing much as a peened round rivet would be expected to look. Because of my use of wedge rather than round rivets I call my maille roman-like to distinguish it from more accurately made roman maille.

In this document I'm not going to say much about the construction of a roman shirt of hamata. This subject is already covered elsewhere in great detail. See the twin how-to manuals by Robert Norton.1,2

What I achieved looks like the following:

Image: image1.jpg
Fig. 1 Roman like maille

The riveted rings are made from 16 gauge mild steel wire. The solid rings are steel washers. Each ring has an inside diameter of 1/4 inch. The solid rings have a square cross section of 0.06 inches by 0.06 inches. Notice the slight flattening of the riveted rings. As is typical roman fashion only the rivet tab is severely flattened. The remainder of the riveted ring more or less retains its spherical shape. This piece of maille illustrates several key characteristics of roman maille:

1. Only the tab is flattened, not the entire ring.

2. Solid rings have a square cross-section.

3. The overlap of the tab of the riveted rings is left over right.

4. The maille appears dense.

The picture also shows how the maille should hang while a piece of the garment, with alternating rows of rings, solid rings oriented to the right and riveted rings oriented to the left.

Though small, other ring sizes are possible. The use of 18 gauge wire certainly works. This would then require solid rings of about 0.04 inch cross section. See the suppliers appendix for a source of non-galvanized steel wire and solid rings.

Making Riveted Rings

The first step in the fabrication of rings for maille is to wind a coil. To do this you need a coil winder. Mine is made from a long piece of 1/4 inch steel rod, bent to form a crank:

Image: image14.jpg
Fig. 2 The coil winder

A hole of diameter equal to that of the wire to be wound is drilled into the rod after the handle. By happy accident I set the winder to exclusively produce left over right rings. I feed the wire from the left table edge, holding the wire down with my gloved left thumb and turn the crank with my right hand.

Rings are cut from the coil using a modified pair of Knipex mini bolt cutters. Take a dremel tool and cut a notch in both the lower and upper cutting blades. The objective is to pass the cutters over the first ring of wire so that the rings are cut with an overlap. Experience will decide how much overlap is right. Too little overlap and you have to overly flatten the tab and the ring may split upon riveting. Too much overlap makes for ugly rings. I've gotten to the point where I can visually judge the degree of overlap.

The next step is to anneal the rings. It is because of the need to anneal that galvanized wire may not be used. The zinc coating will vaporize into a lethal gas. To anneal the rings slide the rings onto a piece of wire and heat till they glow orange:

Image: image20.jpg
Fig. 3 Annealing rings

After heating I usually let the rings cool overnight. To anneal ferrous metals, heat to high temperature quickly and slowly cool.

A riveted ring is then made in the following sequence of steps:

1. The tabs are pre-flattened by placing the annealed ring into a flattening jig. One quick pounding with a 3lb hammer, turn the ring over, and then one more whack does the trick. The effect is to partially flatten just the tabs while only minimally flattening the rest of the ring. Here's the flattening jig:

Image: image15.jpg
Fig. 4 The flattening jig

The flattening jig is made from a piece of 3/4 inch steel bar with a 1/2 inch hole drilled into it. The piston is a piece of 1/2 inch grade 8 bolt. Grade 8 bolts are harder than ordinary bolts.

2. Flatten the tab using an 8oz ball peen hammer. Recently I changed my tab flattening procedure by performing most of the flattening with a modified sheet metal punch. The punch part of a ΒΌ inch punch is ground flat and a grade 8 bolt with a flattened end substitutes for the die. See Figure five. A few taps with the ball peen hammer finishes the flattening job.

Image: image5.jpg
Fig. 5 The Tab flattening Tool

I use a penny to hold the ring in place while I hammer the tab. Believe it or not, the round piece of copper lying on the left end of the anvil was once an USA penny. Canadian pennies don't seem to work as well. Here's another view of my anvil:

Image: image6.jpg
Fig. 6 The ring flattening anvil

An anvil as large as this one, 110lb, isn't required. All that is needed is a sufficiently stable block of flat metal to bang on.

3. Having flattened the tab, punch a hole for the rivet. See the appendix for details on how to make the punch. I initially purchased mine, but learn how to make one anyways. I guarantee that the punch will eventually break. Besides, your purchased punch may not have the best shape. I've settled upon a squarest screw driver shape (actually, since taking these pictures I think that the punch should be just a bit more square):

Image: image7.jpg
Fig. 7 The punch

Here's another view:

Image: image8.jpg
Fig. 8 The punch, side view

The punch is made from a piece of 5/32 inch drill.

As is shown below, the rings go from a hard loop of wire to the punched riveted ring:

Image: image9.jpg
Fig. 9 The riveted ring fabrication steps

In summary: the right most ring is after cutting from the coil. It is then annealed to soften the steel for flattening. After a couple of whacks using the 3lb hammer in the flattening jig the tabs are forced together while barely touching the remainder of the ring. A half dozen or so well placed blows using the 8oz ball peen hammer fully flattens the tab. The punch leaves a rectangular opening for the rivet.

To rivet the ring into place first set the rivet:

Image: image10.jpg
Fig 10 Setting the rivet

Here I'm using a pair of riveting tongs from the Forth Armory. The riveting tongs have three sets of indentations in them. On the left and right sides is a deep hole. This allows the rivet to pass through the ring, distorting its shape. The key to successful riveting is to get the rivet all the way through the hole. Having the right size and shape hole has a big impact. The burr from the hole punching process is left in place. A good setting operation has the rivet noticeably extending in-between the burrs.

Next peen the rivet. As is shown below I do not use a hammer. I destroyed too many rings by hammering. You'll quickly build up your wrist muscles.

Image: image11.jpg
Fig. 11 Peening the rivet

Here I'm using the center indentation of the peening tongs. In the center of the tongs a shallow concavity has been formed by a Dremel. The concavity serves as a die for the peened rivet head.

Appendix 1: Tools

The following photo shows the primary tools needed to make riveted maille:

Image: image18.jpg
Fig. 12 hand tools

From left to right, is a mapp gas torch used to anneal the rings, a pair of smooth jaw piers, tweezers, an 8oz ball peen hammer, the ring punching tongs, a couple pairs of Knipex mini bolt cutters, and a couple of riveting tongs. The right most riveting tong has been made from a pair of cheap piers.

I like to have the tools lined up ready to use. Bins keep rings and rivets ready for assembly.

Image: image19.jpg
Fig 13 Tools Ready for Use

Also useful is a bench grinder used to shape the rivet hole punches and a drill press needed to make the flattening jig.

Appendix 2: Making the punch

First shape the end of the drill bit into the narrow screw driver shape as shown in the preceding photos. Next heat treat the punch. It took me awhile to figure this out, but the process is actually quite easy:

1. Using the torch, heat the punch till it glows orange followed by quick quenching. To quench I use brine made from about a half cup of tap water into which is dissolved a generous amount of table salt.

2. Don't attempt to use the punch yet. Part one has left the punch brittle and if used now it will surely break! Use a piece of emery cloth to shine up the punch. Now temper the punch. Reheat the punch, but this time aim the torch flame onto the drill bit down aways from the punch itself. Allow the heat to flow into the punch. A different color range results. When the punch turns a solid blue color quench again.

3. Now cut the punch to the proper length.

You can now use the punch as intended and if you did this correctly you should be able to punch many rings with it without the punch either breaking or bending.

Appendix 3: Suppliers

1. Forth Armory:

This is where I obtained my larger pair of riveting tongs and wedge rivets.

I deviate from his instructions as I do not hammer the tongs to peen the rivet.

2. The Ring Lord

You'll find great deals on wire here. I've also obtained my ring punching tongs from the Ring Lord. Put the punch that comes with the tool aside and make your own.

3. Seastrom Manufacturing

Any kind of flat washer of any size and made from any material is available from Seastrom. You might have to wait a month or two for a quantity order, but it beats what you would pay at the local hardware store.

My large anvil I purchased on Ebay.


1. Norton, Robert, vest.doc, This document leads you through the steps to make the basic shirt of lorica hamata. Keep in mind that his "basic unit" is based upon a much larger ring than I advise to use here. His patch dimensions need to be scaled.

2. Norton, Robert, double.doc, This document leads you through the steps to make the doubler for the lorica hamata of the 1st and 2nd centuries. For later periods the doubler disappeared.

3. Bishop, M.C.,Coulston, J.C.N., Roman Military Equipment, B. T. Batsford,1993.

4. Feugere, Michel, Weapons of the Romans, Tempus, 1993.

5. James, Simon, Excavations at Dura-Europos, Final Report VII, The Bristish Museum Press, 2004.

6. Van der Sanden, W.A.B, Fragments of a lorica hamata from a barrow at Fluitenberg, Netherlands, JRMES Vol. 4 (1993), pp. 1-8.

7. Sim, David, Roman Chain-Mail: Experiments to Reproduce the Techniques of Manufacture, Britannia, Vol. XXVIII, 1997, pp.359-371
Original URL: