Rotella

Technical information

Blank: Darkwood round shield, 20 in
Fabric: Medium weight linen
Gesso: Golden acrylic gesso
Colors (All Golden Heavy Body acrylics): Iridiscent Silver, Mars Black, Permanent Violet Dark, Light Purple, Zinc White
Gel Medium: Liquitex gel midium, matte
Varnish:

For a long time I've wanted a decorated shield that I could use in combat. I just love how they look and how much they add to any figther in any kind of field.

Given that my persona is based in 16th century Venice, and I already have a schiavona sword, I decided to go with a round shield called Rotella that was common in that time and place. And decorate it with a fancy version of my heraldic badge.

I got a rotella blank from Darkwood Armory and with some help I attached straps made of sole weight leather. To attach them I used binding post screws, also called chicago screws because that way if I ever need to replace the leather straps I do not need to rip the decorated face of the shield.

For those that do not know, chicago screws are made of two parts, a post that has an inner thread and the screw itself. They are similar to the inner threaded piercings that many people use. They are used a lot on leather crafting and bookbinding so they come in a range of sizes. The ones I ended using are M5 x 6 mm which means that the posts fit in a 5mm hole and once closed are 6 mm tall.

The top strap is for the arm and goes around to give more room or armor, while the bottom one is for the hand and is a tighter strap to keep my hand and wrist secure. As is it fits perfectly with a rapier glove, but If I want to use it with a padded glove or a gauntlet I will need to change it for a longer strap

The next step is to create a surface that we can paint on. While we could paint directly on the metal, it would not have enough texture for the paint to grip and will probably peel and crack under the effect and time and combat stress, because remember that the plan is to use this shield in C&T combat

I took a piece of medium weight linen that I had lying around and glued it to the face of the shield using 3M 77 spray adhesive, let it rest around 30 minutes and then cut around the shield, leaving an in or so of extra fabric so I could fold it over the edge of the shield and glue it to the back using more adhesive so I would not have to worry about the fabric eventually unravelling.

Any fabric can be used for this. Some people use muslin, some use canvas. In my case this linen is not a fabric that I use for garb, but I have a lot of it so it worked just fine for this project.

Then is was time to get the design. Given the round shape of the shield, and that my badge it also a radial design, I decided to use some Islamic tiling designs instead of the traditional diapering. This was not only because the shapes fitted, but also because one of my rabbit holes is the mathematics behind the Girih tiles and its use across the medieval Islamic architecture. So being able to use them for my shield because as symbolic as the use of my badge.

The starting design concept is on the left, and the center and right images are the base imaged for the background (slightly simplified) and my badge.

The general shape of the background design is one of the many inner designs that have been used for the decagon Girih tile, as it can be seen in this picture from the Imam Reza shrine in Iran, built around the year 800.

The inner design of the background comes from using the same Girih tiles to fill the general shape, this design was taken from the paper Modularity and Hierarchy in Persian Geometric Ornament which is one of the few papers freely available I've been able to find in the subject.

Because of the complexity of the design, I was not willing to draw it free handed, so I decided to go for a design transfer. For that I needed first to have a 1:1 print which I obtained from Office Depot with their blueprint printing service.

Doing research on print transfer, I found that while there were multiple alternatives for laser prints (modpodge, gel medium, acetone) that was not the case for inkjet prints. All the sources I read pointed to gel medium as the only alternative.

However all of them were for making decoupage or scrapbooking, which meant that the process was meant to preserve the transfer, something that I was not going to do, because I was going to fully paint over it. What that caveat counted on, I decided to do a test transfer using plain Elmer's glue. I used some print scraps that I had to trim away from the design, covering the printed face with a tick layer of glue and pressing it against some unprimed rough linen scrap.

Note: Looking back I should have used the same surface I had on the shield for testing (same linen, primed), but at the moment my logic was that if it worked on a more porous, rougher surface it should work on the better one. I was wrong.

I left the test to dry overnight and the next morning took some cold water and started rubbing away the paper very gently.

It was a very slow and delicate process (the sample piece took me like 5 minutes to do), but I managed to rub away enough paper to leave only the ink and the glue on the fabric. While most of the sites I read recommended using an old toothbrush for this, I'm happy that I used my fingers, because that way I was able to feel the texture of the paper rubbing away and felt when I was actaully rubbing away glue and ink.

I considered the test a success, after I only needed to keep the transfer just long enough to paint over it. And the acrylic paint and sealant would take care of any possible peeling that could occur, and decided to go for the real deal.

That's when I found my first error on the design. The pattern and the print were done on a flat shape, and the shield was domed. Because I had not counted on the curvature of the shield when making the design, I would need to make a radial cut and give it a very flat conical shape so it could fit the shield's surface, which would distort the design.

This is the same that happens on cartography. When you see flat cartography projections of the earth's is not uncommon to see something the make shaped like orange wedges. This is done so when the map is "assembled" it actually fits correctly.

I decided to press on and see how far I could go. The trimmed design was very flimsy which is one hand allowed for a lot of adjusting, but also made putting the glue very tricky, so I decided to add a thick layer of glue to the surface of the shield and then just adjust the pattern as needed. Which ended not being a good idea, because I had a lot of wrinkles that I could not smooth away, and once wet with glue the paper was a lot more delicate than I expected

You can see on the picture to the right where I did the radial cut to adjust the pattern to the surface (close to the bottom, 6 o clock), and how different does that section look. Now, at that moment I was OK with it, because with the paint, the badge on top and from 6 feet away it would be barely noticeable, so I set it outside to dry.

At the time we were in the middle of summer in Texas (100+ F), so I thought that just a couple of hours would be enough time for the glue to try, even though I had left the test to dry overnight indoors. And THAT was the big mistake. I took the shield in and again with cold water started rubbing the paper away. Immediately I noticed I had made the wrong decision when the glue on the uncovered surfaces started to dilute (thankfully not enough to run, but enough to get my fingers sticky), the rubbed away paper stayed stuck on the uncovered surfaces, and the ink was smudging because the glue was not dry enough. I even noticed that the gesso was rubbing away in some spots because I was starting to see the pink of the fabric.

That in addition to the distortion of the design helped me decide to scrap this transfer and try again. So I went to the sink and with a heavy duty sponge rubbed the whole thing away under running water. It was a good way to vent the frustration of my own mistakes.

So now I had learned a few things: It was a bad idea to put glue on the whole shield, I needed it to dry overnight not matter how hot it was outside and had quite the dilemma... how to distort the pattern so when I fit it into the shield I had the design I wanted, without having to make the pattern again from scratch.

Looking again at the pattern, I decided to try something: The pattern has the shape of a decagon, which is a regular figure. This mean that I could take a slice of it and because it was a triangular shape (like a slice of pie) and small enough, it would not distort when I laid it over the curved surface. Then I could just repeat it all around until I got the full design.

The single slice was small enough that I would fit into a letter page, so I did that, cut it off and got to work. Marking some guide points every time I laid it I covered the full shield and saw that I still had an extra portion, which did not surprised me because I had not modified the design itself. But now I could measure how much I had to trim away.

I was a triangle measuring 1 3/8 inches on the outer side, which seemed a lot because the side of the decagon measured 6 3/8 inches. But I did not have to take all that at once. If I split that almost 1.5 inches among the 10 sides then I only needed to trim 0.15 inch form each side, or 0.075 inch from each end of the side. The second was a better decision, because that way I could keep the design balanced and minimize even more the visual distortion. I made a template of a single side of the decagon and printed 10 copies, measured the tiny slivers I needed to trim on each and cut them off with an exacto knife.

Having a plan of action, I went to execute try number 2. To make sure I would not have any glue around, I added a new layer of gesso and decided to buy the gel medium that was recommended to transfer inkjet prints. (https://www.liquitex.com/us/product/liquitex-professional-mediums-matte-gel/). With a thick brush, I started laying a generous amount of medium over the printed design on each of the sides, laying them face down over the shield and smoothing them out as much as possible before starting with the next one and left it to dry overnight.

The next morning, I took a cup of cold water and dunking my fingers, I started rubbing the paper out, being as careful as I could when going over the design to not rub the gel and ink off. I definitely got better results than the previous time, but still in some parts I rubbed too much or too hard and ended taking everything off. But I was happy enough with the design to keep going.

After leaving the rubbed shield dry again overnight, it was time to add my badge. I repeated the same steps I did with the decagon sides, putting the rose as centered as possible which was more difficult tan I expected, because the medium made the rays curl and stick to themselves. Once the rose was on the shield it would not be moved on the risk of tearing it apart, so I just again let it dry overnight.

When it was time to rub the paper of the rose, I noticed that while the paper used to print the rose was thicker than the regular printer paper, it also absorbed water easier, which made the process a lot more delicate because the paper was more prone to tear, which you can see it happened in one of the rays.

Let everything dry again, and then take a good look at it, to see if it was worth to keep going or if should just start over again. Because I did not want to rub off the medium, most of the shield still had paper layers half rubbed off, which gave the shield a texture similar to parchment. The design was broken in some parts, not visible enough in others... But I did not need it to be perfect, I just needed it good enough to paint over it so I decided to keep going and start painting.

My palette was going to be black, silver and 4 shades of purple, which I wanted to make in a spectrum. I decided to start with the "heraldic purple" for the rose and one of the shapes. Got my purple to mix, and I noticed that it was too light, almost a lavender. Still I tried to mix a color and painted the first shape, and I decided it looked too dull. Also the dirty background was making me start hating the project. In a quick impulse, I grabbed the black and painted all the voids in the background and the seeds and leaves of the rose, thinking (correctly) that it would minimize the raw paper visible and help me not focus on that while I was working.

I was frustrated by my bad choice of purple, so looking around in internet I went to the brand's website and checked their full palette. There I also found that they have a virtual color mixer so I spent a while playing with the shades of purple until I got something I liked. A couple of clicks in Amazon and a few days later I had my new purple to keep working.

My "heraldic" purple was done mixing 1:1 light and dark purple and not only I was happy with the results, I could already see that the new colors made that dull lavender a little less dull. Because there was so much heraldic purple on my badge, for this color I picked up a shape that was not common in the center of the background so the badge would not look distorted at a distance.

Once dry, I was happy to notice that the final purple was pretty similar to the once I got in the virtual color mixer. Not an exact match but pretty close, considering the differences due to screen, light and all that. So to avoid repeating the experience with my first purple, I went back to it and decided to create the full palette, so I knew what to work with or if I needed to buy more colors.

My final palette was (All in the order of White, light purple, dark purple): 0/1/1, 1/2/0, 3/1/0, and 10/1/1. After that, there was only left to paint all the colors

The lightest purple gave me a lot of issues. I had to paint this color in parts, so I had to mix the color multiple times. I never managed to get quite the same, and i had problems with the saturation. The colors looked transparent even when using them undiluted, and the paint would often start to dry before I hwas finished so a lot of parts are grainy and semi transparent.

As mentioned before, the last color was going to be silver, but at last minute I decided to change it to white because I felt if I used silver the rays of the rose would get distorted in the background. Keeping it white gives the rose enough contrast to show.

The last step was to spray the painting with a non yellowing, matte varnish that would protect the painting from the sun and from chipping and peeling due to use and combat.

Conclusions and Lessons Learned

This project was fun, difficult and a challenge altogether. But there a few things that I would change for the next time:

Have a colored master: When preparing the design and the colors, take the time and print a copy of the full design to color. It does not have to be the same colors you use in the final version, but having a guide on what color was which piece would have made things quite easier when working on the background.
Transfer, then glue: The main challenge and frustration point was that I could not fit the original design on the shield because I had not considered its curvature. That led me to a whole detour on how to distort the design to fit and made things unnecessarily complicated. A better option would have been to lay the fabric over the shield, trace the perimeter. Then measure the circle flat and use that measurement to print the design. Prepare the fabric, transfer the printed design and glue it to shield as a last step before painting.
Multiple layers of transfer medium: To transfer the design to the fabric I used a single (relatively thick) layer of gel medium painted over the print. This caused that when removing the paper I had very leeway before removing the medium itself. I should have painted multiple layers of the gel medium, letting them dry completely in between.
Thinner fabric: I used medium weight linen because it was what I had on hand. I think that a thinner fabric would have given a smoother surface (even with the use of the gesso) and would have allowed to paint sharper edges.

Page updated

Google Sites

Report abuse