Plan B!

Wondering why a previously-successful process didn't work very well at all twice in a row, I decided that I should just do an experiment.  Since I have a number of different kinds of varnishes in my stash, I got out some unvarnished paint-chip squares that were left over, and tried every one of the varnishes, partly to figure out why the artists medium didn't work and partly to audition a different varnish.

The results were quite interesting, actually.  I stuck my samples on a piece of card stock, so that I can have a record of the various finishes -- at the top in the photo above is one of the leftover tiles from my very first batch, as charming as it was the day it was made!  Below it, on the far left, are a cluster of three of the faux tiles varnished with the Liquitex gloss medium I used originally, then Mod Podge satin finish, then DuraClear ultra-matte, CeramCoat gloss varnish, and DuraClear satin.  They were all smoothly covered when the varnish was newly-applied, so the dips appeared during the drying process for some reason I don't know yet -- I applied it intentionally quite thickly, by the way, so these flaws might not appear if applied thinly.  The DuraClear ultra-matte is ultra-matte, indeed! but too matte for my purpose, though beautifully smooth and even.  I wasn't willing to risk large irregularities like the CeramCoat sample, so that one was probably out, too, at least for now.  This left the Mod Podge satin finish and the DuraClear satin, both of which probably would be successful, but the Mod Podge was still a bit tacky the next day, and indeed the instructions say it might be, and recommend a varnish top-coat, so I decided to use the DuraClear even though the Mod Podge had just a bit more depth to it, which to my eye looked more like a real glazed tile.  But by the second day, the Mod Podge sample was not tacky at all, and so I changed my mind and used the Mod Podge after all.  (I figured that if the faux tiles were still tacky after a few days, I could varnish them with the DuraClear satin ...!)

My conclusion from this experiment is that the Liquitex gloss medium is just getting old.  I bought it and first used it in 2014!  It has worked fine only a month or so ago for "varnishing" faux paintings (printed on paper with the computer), so I'm guessing that enough of the water in it has evaporated over time that when laid thickly like this, it just doesn't set as well as it did when first opened.  (The first batch of mini tiles I glazed a month ago are still tacky, in fact.)  It's also possible that the water content helps it to self-level while it is drying, and therefore the bubbles and irregularities "stick" instead of smoothing themselves out.

And so I made a third set of tiles -- but this time, I had been scrolling, rather disheartened, through photos of "tiled shop fronts" on Google (as one does), and seen one with a sort of self-border of half-tiles, and thought it looked very handsome.  And ... if I did rows of half-tiles along the top and bottom, I could make them bigger and still have an odd number of rows!  The only remaining problem was that of getting each square the same height and depth as its neighbors.  I cut a piece of watercolor paper a good bit bigger than the area to be tiled, painted it with FolkArt milk paint in "Pirate Black" (!), and tacked to the back of it a piece of 1/2" grid paper.  Instead of cutting the painted area into strips and then cutting each of the strips into squares -- which is what I did before, resulting in ever-so-slightly different "squares" -- I numbered the squares on the unpainted side before cutting, enough squares to fit the space to be tiled, then used my big paper-cutter to cut along the horizontal lines but only almost to the edge, leaving them still attached to each other at one end.  Then I turned the whole thing clock-wise 90 degrees and cut along the other set of lines, this time all of the way (because I could put them back in order using the numbers).  This was more than a bit awkward, obviously -- I had to make sure that all of the partly-cut strips were flat and butted up closely to each other, so that the second cuts would be in the right place, and the grid paper didn't stick very well so that I had to carefully peel it off of each cut square and write the correct number onto the back of the square! -- but it wasn't as onerous as it sounds, and it achieved the result I was hoping for, a nice straight set of squares that line up evenly, and in both directions.

The Mod Podge, just like the artists medium in my very first set of faux tiles, took on some of the texture of the watercolor paper, which is fine with me.  I'm very pleased with the satin finish, too -- it has enough shininess to say "glazed tiles" but not so much that it's distracting or reflects irregularities.  Now that I have a better idea of the predicted-success-rate, in future I might try a smooth paper -- the paint chips are worth using again, for instance -- or something more Arts & Crafts/"hand-painted" like a watercolor wash.

So I'm happier with the finish, and I'm happier with the new size!

(The façade developed a crack running from the upper left corner of the door frame through the cornice molding, so I've pulled it apart and reglued it, which is why it's clamped now.  The tiles are not attached to the bay window yet ...)

Comments

[Sarina]

Sometimes research pays off!

Hope life with the new normal is all right for you all. A lot of changes for everybody. We are well and I'm happy being home for now.

[juliet brown]

Wow forensic attention to detail, do you have a particular town in mind for your little shop to inhabit? I'm thinking York or maybe around Thirsk, when the tea shop is ready you must invite us for a warming - I can bring virtual pots of jam, as long as I follow more or less the National Trust recipe book for such things I am quite proficient. Now what virtual jam might you like I wonder???