mosaic tile rosettes split over time

[david bain]

Anyone have any ideas for preventing mosaic tiles in classical rosettes splitting as time goes by? Mine, although perfectly fitting and smooth-surfaced after finishing (french polished shellac), develop tiny cracks within a year or two and eventually begin to become uneven to the touch. I suspect this is inevitable given changes in humidity, and the fact that they're end grain, but maybe there's something I'm not doing, and should be?!

Thanks

[Michael Lewis]

I think your problem may be the tiny pieces of end grain are absorbing too much moisture from the glue when you install them. You might try super drying the rosette and sealing it on both sides with shellac or CA so it doesn't pick up moisture and expand, THEN glue it in place as normal.

[david bain]

Thanks Michael; yes I'm sure that's the basic problem! I assemble my rosettes actually on the guitar top, so I'd have to treat each tile, I suppose, in something like the way you describe. I tend to use hot hide glue (although I have used titebond; doesn't seem to make any difference), one advantage of which is that it encourages the tiles to swell and fill any tiny gaps there might otherwise be! This is the way I was taught to do it...

However, drying the tiles and then sealing them with shellac, say, before fitting sounds like a good thing to try. I've tried sealing with shellac afterwards to protect the rosette through the rest of the manufacture (I do the rosettes very early on), but obviously that's not enough; by then the damage (too much water to begin with) is done... The splits I'm talking about aren't huge, but annoying nonetheless.

[Randy Roberts]

Using polyethylene cutting board material for constructing your rosettes may solve a lot of your troubles. Here's a post from the library. The method is about 3/4 of the way down the thread:

http://www.mimf.com/library/Methods_for ... -2004.html

[david bain]

Thanks Randy, that's a nifty idea.... I have a bit of a reluctance to use CA glue for some reason (bit of a traditionalist? - although I do use it for some repairs); I note someone in the thread mentioned using PVA or hide glue, but I wonder if I did that whether I'd just be setting up the same problem for myself by the back door... Also, I've never yet tried fitting a full preformed rosette to a guitar top, except for a solid walnut one which wasn't a good fit on installation but looks great now (years later) after filling and finishing. My problem there is that my circle cutter (a lathe/drill press type which I use with a knife blade by hand) doesn't actually cut a circle, so assembling the rosette directly in the channel in the guitar top seems to give me a better chance of a good fit. Maybe I should try the router method of cutting the channel (the potential for disaster there always seemed to me to be huge!).

Lots to think about!

[Clay Schaeffer]

If you use thin CA on end grain it should wick completely through the tiles and seal them so they don't absorb or lose moisture and shrink or swell. You could then build the rosettes in place as you prefer. Shellac is not water proof and will still allow some moisture migration in and out of the tiles.
Another possibility might be to use "stabilized wood " that some knife makers use to build your rosettes.

[david bain]

Thanks for that Clay - yet more CA! I have thought of that sort of approach, but baulked at it...

But is it really the case that nobody else has had this sort of problem?! CA hasn't been around forever... I suppose I would have to look at a lot of mature handmade classicals to see if what I'm worrying about is really that much of an issue. When I had such an opportunity, I wasn't thinking about rosettes...

Thanks again for the suggestions.

[Barry Daniels]

CA hasn't been around for ever but cracking tiles have been. It is very common, but it is also exacerbated by using large tiles of ingrain wood. Most classical rosettes use very small tiles which minimizes expansion/contraction issues.