Making a domed Late Roman shield

[M. Demetrius]

I plan to make two oval Late Roman shields. Robert V, you'll be pleased to know that we have finally decided to put together a Late Roman impression, Mike Boyd, Doug Mather, and myself will be the first members of the Cornuti Seniori. 4th/5th Century. We'll be the 5th Cohort of that Legion. Interestingly, a loose translation of the Legion title could be "horny old men", which, well, you know.

So here's my plan. We are not ready to use the splits of wood and lay them up in a true historical fasion. I think I'll try this method:

1. cut the oval (THANKS FOR THE FORMULA, ROBERT.) from 1/4" plywood.
2. drill the holes for the edging.
3. put the ovals in a soak tank for three or four days at least
4. on a 3/4" (19mm) piece of plywood, larger than the shield, I'll fasten a circle of wood 1 1/2" thick, about the same size as the flange of the boss, right in the center.
4. When saturated with water, using wood screws through the edging holes, bend the wood over the block.
5. Let it dry for however long it takes, I expect a week or more, in the shade

The hope is that it will dry and retain some of that dome shape when the screws are removed. Does that sound crazy?

[jvrjenivs]

Well, I would also suggest you to look in the greek section were a couple of methods of making a reproduction Aspis are discussed.

Don't know if that method would work good enough. Well, just try I would say. You can also make the wood more flexible by heating it just before you place it into the mould. Get it soaked very well, heat the wood with a blow-torch (big one used for roofing), as the wood will get more and more flexible. (make sure you add enough water during heating so you will only heat the wood, not burn it!) After the process (the longer you heat, the more flexible the wood should become) you can bent it for about 10 minutes. Just enough to place it into the mould with a couple of people.

Let it dry for some time and the wood will stay bended and you can start finishing the shield. (Note that you should start with a set of wood slightly more thick as you want it to be, as the heating will blacken the outer layer of wood, which you have to grind away after the shield has its shape

Anyway, good luck with your project. It's good to see there are some more Late Romans comming in the US

[M. Demetrius]

Thanks for the encouragement, Jurjen.

I have a "weed burner" that might do the same thing as the roof tar heater...but I don't plan to build a dished oval mold, and am hoping to avoid all the sanding and shaping that most of the aspis construction methods use. Using the soak/bend/dry method, I've had fair success with the traditional scuta, so it just seemed worth a try. 8mm Plywood is prone to warp on its own, so I just thought, "Hey, what if "

I'll keep a record and show how it works.

[Epictetus]

Would heating the water help? I know some Finnish craftsmen use steam to bend wood. Is it the heat or the moisture content that helps bending?

[M. Demetrius]

Yes, it would. My plan is to put the soak tank out in the sun. On a bright day, using the black plastic as a collector, the water gets pretty hot to the touch. BTW, if anyone tries this, make sure it's EXterior plywood. Interior will simply delaminate into a big mess of slats.

[Matthew Amt]

There are a few folks I know of who have tried to make dished medieval shields in this sort of way, and from what I recall they only get mixed success, at best. The problem is that you are trying to not just bend wood, but dish it, either stretching or compressing it in a way it was designed to prevent! But I've heard of a method that might help:

Instead of supporting the center, set the wet wood on a raised ring (4 or 5 inches high should do). Then you can either pull the center downward using a large threaded bold of some sort through a hole in the center of the blank (with a circular plate or block of wood to serve as a washer, maybe 6" in diameter); or force the center downwards with weights or a jack braced against a handy ceiling. (Don't wreck the shed roof...)

If I were trying this, I'd use 2 layers of thinner wood, soaking both and leaving in the press to dry, then I'd do it again. After drying the second time, THEN I'd put glue between the dry layers and put them back in the press until the glue is dry, giving a total of 3 sessions in the press.

But however you do it, don't be surprised if you only have an inch or two of depth when it's done... If it works better than that, though, we want to know exactly how you did it!! Cuz I know a number of guys who would like a dished shield, including myself. And all I can think of at the moment is laying a number of thick slabs together and chiseling of a heck of a lot of wood... Authentic as heck, and a heck of a chore!

Good luck and Vale,

Matthew

[M. Demetrius]

Thanks, Matt. Good observations and advice, as always.

How much dome is supposed to be in a Late Roman oval, anyhow?

In scutum making using the same methodology, a blank pressed over an oil drum 24" diameter, (~60mm) relaxes to around a 30" (~90mm)diameter. So if I'm able to get 2" (50mm) of bend on the frame, I would expect around 1 1/2" (~38mm). Is that enough? The side axis will be no problem because of the direction the grain runs. But the top to bottom may pose more of a problem.

Here's a thought, too. Cutting a series of horizontal slits or chiseling shallow grooves in the backside (maybe a dozen or so above and below the boss area) might reduce the resistance of the long grain on the compression side. If that works, and the shield will hold its curve better, then bracing with slats will be unnecessary, and a layer of cloth on the back should help hold the shape. Hmm.

Still brainstorming. Won't start that project until next month, so there's time to decide. I might take a strip of similar plywood and try to warp it. About a foot wide ought to be sufficient, I'd think. Put cuts/grooves in one end and nothing in the other. Might be worth a try...will post photos.

More ideas, folks?

[Robert Vermaat]

Hi David,

I plan to make two oval Late Roman shields. Robert V, you'll be pleased to know that we have finally decided to put together a Late Roman impression, Mike Boyd, Doug Mather, and myself will be the first members of the Cornuti Seniori. 4th/5th Century. We'll be the 5th Cohort of that Legion. Interestingly, a loose translation of the Legion title could be "horny old men", which, well, you know.

Great idea! Always good to see some more allies across the pond...
That should be Cornuti seniores gentlemen, btw...

I think I'll try this method:

have a look here, the site of the producer of the Fevtio dished scuta:
http://fabricae.free.fr/
It's in French but with an English summary top right - you'll get the idea.

Good luck!!!

[M. Demetrius]

Thanks for the tip and the correction, Robert. From what I can discern on that website, it looks like the dish is around 50mm deep. Is that about right?

[M. Demetrius]

Truly, that's a serious question. How deep of a dish is deep enough? It's very hard to tell from the photos I've seen, and I've never handled a dished shield that wasn't made from metal.
2" enough?
3" better?
4" too much?

[Medicus matt]

Dug Simon James' Dura report out of the book room to see what he had to say about the oval shields found there.

Average is 1.05m high by 0.9m wide. Board thickness at centre under the boss was 9mm, tapering down to 3-5mm at the edges. Covexity is impossible to judge accurately due to post depositional distortion (some of them show no sign at all and may well have just been flat ovals) but he estimates that the high point at the centre would be 10 cm above the height of the rim (as evidenced by the remains of shield 619, where some planks were still curved)

Looking at the drawings of the surviving boards, most of the concavity must have been achieved by shaping/carving/planing after assembly, as the planks themselves don't seem to show much evidence of the tapering needed at the top and bottom to allow a barrel stave construction. You can get a good curve in one plane by butting parallel edged boards but to get a real lenticular shape you have to shape the boards either before or after assembly.
The boards do show signs of having been shaped with a plane or draw knife.

Holes for the attachment of the leather/hide rims are about 8-10mm in from the edge and 6-9 mm apart.

[M. Demetrius]

Thanks, Matt. Good info. A 1" (25mm) dome should certainly be achievable. Test in progress.

EDIT: Sorry, Matt, I misread. Should have said 4". I don't know if the wood will bend that far without cracking. We'll see what we see, though. Test still in progress.

[M. Demetrius]

I have had some success, and can achieve a not-entirely mathematical curve with a dome depth of 2 3/4" (about 7cm). Because the wood is not uniformly strong/weak, there are a few slightly high or slightly low spots. The dark grain appears to be harder, and bends less.

One I was able to get a horrid shape, but 4" (10cm) depth. The wood broke in two or three places. Can't bend it that far, period. To do that deep of a dish, it would be necessary to build a form, and lay up strips of plywood the way the originals did. I don't think it could be done from plywood without having a steamer that could hold the sheet, and a press that can work with wood that hot...basically impossible for the normal shop. A laid up strip shield is possible.

I'll see what our other guys think of the shallower dome. It looks better to the eye than just flat plywood, I think. Pictures someday.

I have bosses to make, too, so it might be a while.

[Medicus matt]

Good going! 2 3/4" is enough to bring the weight of most of the shield to behind the grip and deep enough to get your body into. Very interested in seeing some pics and description of your method when you've got the time.

[M. Demetrius]

OK. Sorry to have been remiss in posting about this, but I have been working on it.

Here's where I am.

*I'm using standard (but chosen to be the most even grain and least cracked and spliced) 1/4" (6mm?) 3 ply, BC sanded exterior plywood. Costs around 18 dollars a sheet at the Home Depot. The test shield, I domed, as said before, but it was unsuccessful in some ways, and very irregular. You can't simply start at one spot and go around the shield with a series of screws. The sides are no problem. The top and bottom are a little more difficult, but they can be bent. The problem areas are at 10, 2, 4 and 8 o'clock. These have to bend in two directions at once, and are not interested in cooperating. It was about 2 3/4" in depth.

*I resoaked it and domed it again, a little deeper, and still a little irregularly, but by then around 3 1/4". I was encouraged enough to continue.

*The third soak (still on the 1st board) I was determined to get the full 4". I did, but the plywood broke in two or three places, one of them pretty badly, so I put the screws in that area through a scrap piece of plywood laid on the surface, like an ad hoc splint, spreading the pressure over a wider area. That was a great discovery. I used that method from then on to help press down any high spots on the surface from. Looks like a bunch of bandaids all over the place, which I guess it is. In the end, that test shield was a lumpy mess anyway, but I did get the 4" depth. The edge is so full of screw holes it's not stable any more. I set it aside, and looked at it many times over the next several days before I proceeded, thinking about how to prevent the problems that I'd had with it.

* *

I have determined that it's not reasonable without some kind of a steamer unit to get the plywood to bend that far, and it would take a 3 dimensional 2 piece press to make the wood bend and stay and three men and a strong boy to press it down into place. That's all beyond my equipment/ability/pocketbook at the present time. The big boys are able to take a flat sheet of plywood and bend it into spiral rocking chair struts and runners, office chair arms and so forth, from triple the thickness I'm using, so I know it's possible, just not by me. I'm sure they don't use the cheap plywood like the rest of us do, either.

So by working out a set of "concentric" ("epicentric?") ellipses stacked up to press against the shield from the under side, I was able to settle on about 2 1/2 to 2 3/4 inch dome, which is how they'll be. These elliptical braces under the shield help to prevent places where the shield dips instead of domes, and by backing off the deeper dish, the surface irregularities are reduced.

It is domed enough that it looks clearly "not flat", and by using a cross-piece handle, the more flexible side-to-side arch can be held to the desired depth and stabilized. I'll use a 1 x 4 cut down to the right arch and glued and nailed to the face to maintain that lateral curve. I think that will solve the problem, and make a handle that's built into the shield, plus giving the added advantage that it will have the handle removed a bit from the plane of the shield, hopefully around the center of gravity, so the back of the carrier's hand won't bash against the wood. The vertical seems to be stiff enough without any help.

When daylight comes, I'll measure out the curve and band saw some cross brace/handle pieces. I now will have 3 elliptical domed scutum blanks!

There are quiet a few screw holes in the shields, of course, most of them around the edges. But I will run the sander over both sides to knock down the splintery places, then fill the holes with wood putty and sand again before putting on the fabric.

I cut two sheets of plywood into shield sized ellipses way back at the start, believing I could make the project work, or if it didn't work, at least I'd have 3 more flat shields. By using a system of putting screws widely spaced around, but not all the way screwed in, then tightening one a little, then another, then adding a scrap wood splint here and there, then tightening more screws a little at a time, adding another splint, adding a screw where the wood began to bow, I achieved a domed shield by the third one of the "production blanks" that is nearly regularly arched all around. I'm pretty proud of it. If that sounds tedious and aggravating, well, that's because it is. Now, if I can figure out how to get the cloth to go on smoothly on the front face without wrinkling, I'll be in the happy place.

The first blank, the test copy? I decided it had too much experience and history involved in its life to toss away, so I got a nail and a string and cut a 32" circle from the middle part. It was sort of domed already, but I put it back in the water for a while, then bent it over a circular piece of 3/4" plywood fastened to a piece of 2 x 6 for height. That makes about a 2 1/4" center height. I screwed it down to another board, and will use that for a parma. Were parmae domed? Hard to say, but we'll be the first kids on the block to have one that is. I'll bet others will be jealous.

When all is dry, and the handle blanks are cut, I'll put up some pictures. Whew! It's a relief to find that my mental gymnastics were not all for nothing. Robert V: Thanks for that formula for determining the nail distance and string length. That was a great help, especially when I had to make those other ellipses, and make them the same proportions as the shield blank.