If you’ve been following the build of the wine sideboard, here is the culmination of a bit more than a month of part-time effort, including a week and a half of finishing:
For those of you following @mansfinefurn on twitter, some of the following may be familiar since I did a side post about this last week. Feel free to skip to the new stuff down below.
I spoke about my finish strategy in the last post, which was to use Sherwin-Williams wood classics stain (oil) to match the client’s existing furniture, then wait ~3X longer than specified on the stain for drying time in a warm dry shop, before overcoating with Hydrocote Resisthane Plus (waterborne). I mentioned that the Resisthane can says not to use it over oil stains, and that after speaking with the people at Hood Finishes (makers of Resisthane), I was satisfied that a wipe down of the dried surface with denatured alcohol would remove any microscopic layer of residual oil that remained after drying
I also mentioned it was a MINOR concern because I had done a test piece which had absolutely NO issues of poor adhesion. I pulled a screwdriver across it to double check, no problem…
After the finish was hardened, and had been sitting for~ 24 hours in a 75°F, dry environment, I applied a thin finish coat of Briwax paste wax (which I do on many finishes as a final step, including, over Resisthane on two prior occasions). On an unseasonably sunny, warm day, I pulled the piece out into the driveway in the sunlight, and saw some ripple action (more on that in a bit…) that I hadn’t seen in the raking light in the shop.
Before I go any further, here’s a little gallery of images from BEFORE I go and undo much of the finishing I had done, call these the finished piece, pre-re-finishing:
I decided to knock it down with 600 grit paper on a sanding block, which went well, until I did the ONE LAST PASS, and a bit of the finish on the top(< mm^2) sloughed off…no big deal, plenty of time to touch it up. I proceeded to blow the sanding dust off with the compressed air gun and dropped the nozzle (not hard) on the top, and surrounding the tiny ding was from the gun, was a sizable bubble, and I could see right away that it was the finish pulling off of the surface. Long story short: a bit of sanding, peeling by hand, became peeling off the finish in semi-brittle sheets– I needed to take it all off (didn’t take any images of this, I was too busy uttering expletives).
Fortunately, where there are adhesion problems, removing the finish is pretty quick and simple with a carbide scraper (mine’s a Bahco I think, and I love it). It quickly took off the finish from the surface, with a few whisps of wood here and there, but enough that it would need to be re-stained to match the rest of the case. So I applied a quick coat of stain, and wiped it off in less than 5 minutes, really rubbing hard with a clean cloth to suck up any of the residue. I would need to leave it to dry at least 12 more hours…
After I scraped and re-stained, I came up with my plan:
- Let the stain cure at ~80-85°F, ~30%rH for at least 12h
- Wipe it down firmly with denatured alcohol to take off any surface oil
- Spray a coat of Zinsser SealCoat (dewaxed shellac) to seal the oil from the topcoat
- Dry for ?h
- Spray 3-5 coats of General Finishes Enduro-Var Satin (the can says do not apply over shellac, I’m assuming they don’t mean de-waxed under this restriction? Anyone know if it’s OK over SealCoat?) @2h between coats.
- Deliver on Xmas Eve by 4P (I got a day’s extension to deal with this, and this plan was formulated 12/22)
Needless to say, it was a sleepless night, during which I reached out to the wonderful community of woodworkers online. Unfortunately, there is little agreement to be found among woodworkers when it comes to finishing, and when presented with the proposed plan, I got two general replies: 1) sounds good, go for it, or 2) No way, just blow the deadline and let the stain dry completely.
I’m not sure how many of you have tried to read up on oil stains and how to know when they’re dry, but I found, rather quickly, that everyone says something like “make sure it’s completely dry” or “do not overcoat until completely dry”, but nobody tells you HOW to tell if its completely dry. To me, it’s dry when it feels dry. This was essentially confirmed for me in talking with Marc Spagnuolo, better known as the Wood Whisperer, who wrote a great little book called Finishing: It ain’t over till its over and spent some part of his development in woodworking working in refinishing. His description was that it is dry when it feels ‘like wood’- same temperature as ambient when at room temperature, with no tackiness, so you can rub your hand across it without ‘grabbing’. He also said that my plan sounded good, and what he might do in the same circumstance. Heartening.
However, another woodworker, and former pro finisher who replied on Google +, (no name in his profile, so +Ace HoleInOne, if you’re out there, thanks for the feedback!) suggested that the whole plan sounded rather rushed (and it was!) and that the de-waxed shellac was unnecessary if the stain was ‘completely dry’, and in fact, that the shellac (dewaxed or not) would not be a good choice under General finishes Enduro Var. (You may wish to read the full interchange here, though the link to the original post is no longer active, the contents of that link are in this post, above)
Needless to say, I wanted to test some options out, both to see if there would be any further issue of adhesion, and to see what the planned (and other) finish(es) looked like over the stain. It was also important to match the sheen and look of the original piece as closely as possible. Fortunately, I had a large test board I had finished along side the actual piece, which I scraped and restained just as I had the top of the case. On it, I tested several options:
- 1/2# cut dewaxed shellac – 1 and 2 thin coats, overcoated with General Finishes Enduro Var (satin)
- 1/2# cut dewaxed shellac – 1 and 2 thin coats, overcoated with General Finishes High Performance (semi-gloss)
- GF Enduro Var no shellac coat
- GF High Performance no shellac coat
- Minwax Spray Lacquer (solvent based) (satin)
Why reluctantly? let me count the ways…
Frankly, I’m irrationally afraid of spraying solvent lacquer, due to the extreme flammability of the solvent. My ‘spray booth’ consists of a room built of plastic sheeting inside my garage, with a furnace filter make-up air inlet, and a filtered box fan exhaust (out the window). It’s cold here, and the curing and drying of the various parts of all of these finishes require a warm environment (the stain, all of the lacquers, spraying, etc.) so opening windows and pulling in cold air with the fan for ventilation is counter-productive towards this goal. But, ceramic heaters, such as the two I was using to maintain a warm environment to get this finish cured, are not a good thing to have around solvent vapors, especially lacquer thinner. If it were summer, I would have donned the mask and taken the whole operation outdoors. Instead, I turned off all ignition sources, donned the mask, sprayed, then opened the windows, turned on the fan, and let the vapor dissipate and be exhausted for 20-30 minutes before returning and closing everything and re-heating the space.
Another reason: Solvents, particularly aerosolized, are not environmentally friendly, in general, and I try very hard to do my work with low environmental impact. For health’s sake, and for the sake of a healthy environment for my kids, I try to minimize their use whenever possible, thus the initial choice to use Hydrocote. If you want to know more about the ‘bad’ stuff about solvent based finishes, read up on VOCs (volatile organic compounds).
That said, I chose the solvent lacquer to use for this application for several reasons:
- Not incompatible with oil finish (could be used for over-coating much sooner than any others)
- Matched the look of the original piece best (which was lacquer finish), in sheen,color, and depth.
- Adhered nicely in the test piece
- Dries FAST and can be recoated sooner.
A quick side note:
I was not impressed with the quality of the Minwax spray lacquer from the can, though I didn’t have high expectations, I had hoped to be able to get adequate spray to get the job done. It was ok, but on later coats, it sputtered a little, leaving dry patches in some places, and large drips (droplets) in the finish in other spots that had to be worked out later. Bottom line: it worked in a pinch, and kept me from having to mix up my own from brushing lacquer and lacquer thinner (of which i’m even more irrationally wary), and spraying it with my HVLP, and having to clean up after the whole thing, but I won’t use it again except outdoors for very small projects.
To finish the finish tale: After 3-4 coats, the surface was looking pretty nice, and required just minimal sanding to remove the slight ‘texturing’ and some careful razor work to deal with the droplets I mentioned.
The morning of delivery, I rubbed out the finish with 600 grit wetted with mineral spirits, followed by #0000 steel wool, to bring it to uniform sheen. I wiped it down thoroughly then waited for it to dry (directly in front of my two heaters), and waxed it as earlier described. It seems that some of the mineral spirits absorbed into the deep pores of the early wood since on occasion, while it was warming up, tiny ‘volcanoes’ of liquid burst forth in droplet sized puddles from the grain. These either evaporated away or were wiped away, not affecting the finish in any way.
Oh, and that ripple I had seen? It turns out, deep under all the finish, the oak itself had a bit of figure, which I hadn’t noticed in the fluorescent lighting in the workshop. I attempted to get an image of it, but struggled. If you look carefully just around the reflected light you’ll see the waves in this image:
And here’s a shot of the finished top by itself:
Since my engineering senses tingle when anyone claims a root cause without having sufficiently tested to determine causation, and since I haven’t done said testing, I’ll just discuss some of the potential issues that I speculate may have been at the root of the problem, and will allow for amending or adding to the list as comments from readers/further data come to light:
Time: Undoubtedly, deep down in the grain of this oak the liquids in the stain were not fully dry, though the surface felt as completely dry as any might describe. It seems like the instruction on the can of 12-24 hours of drying time might be sufficient for closed grain or dense woods, but even with proper ventilation and temperatures, 3 days did not suffice in this case.
Grain: The early-growth is soft and porous, and sucks in anything in liquid form, which means issues with drying, and difficulties with evaporation. Not only did the stain get in and, it seems, not dry, but the mineral spirits used in wet sanding the finish seeped in an ‘boiled’ out when the temperature started to rise.
Denatured Alcohol: I’m not certain about the chemistry on this, but +Ace HoleInOne suggested that wiping down with denatured alcohol before applying the first coat of finish sounded wrong, and just may have ‘re-wet’ the stain in the grain, where it could have actually been dry after the 3 days. It made enough sense to me that I chose not to do it after re-staining (at least on the test piece, since it was not an issue with the solvent lacquer).
Incompatible finishes: As much as any manufacturer claims their finishes will work with otherwise incompatible stains, some of the basic chemistries of, or rather, between the finishes may be more or less compatible. some may work, just not that well, some are completely incompatible, and some, of course, are completely compatible. On this, all I can suggest is read the can, and start your work by trying some finishes, before you start joinery, before final milling, surface a few boards, and test some ways you might choose to finish, before you start. Then, don’t put 100% stock in the results, since, for me, the test piece showed no signs of failure. Leave alot of time for finishing.
There are probably more, and I’m happy to amend the list if others have good suggestions.
Lessons-Learned (or re-learned, as it were):
On the Finish:
Let it dry, Let it dry, let it dry – I find this among the least satisfying thing to say, and the most irritating thing to try and research online. Since few can say when is the oil deep down dry enough for topcoats, let it dry according to the instructions for use, then double that, then wait until it ‘feels’ dry, then wait twice again as long. then, wait some more…Think weeks, not days.
I don’t care much for oil stains – at least not with waterborne finishes. I haven’t used water borne dyes/stains, and have been meaning to try them. Now I have greater motivation to do so. I’m not sure it would have helped with this project, the color was precisely specified. I would have had to done alot of experimentation and mixing (not to mention buying) to get the color just right.
I don’t care much for red oak – I actually enjoyed most of the aspects of ‘working’ the wood on this piece, at least, much more so than past oak projects, with the possible exceptions of the large ‘cathedrals’ of early-growth (the dark grain) that was so porous and soft relative to the denser, harder late-growth, that it sanded and scraped poorly, wearing much faster and tearing-out much more severely, and led to difficulties in flattening without ‘valleys’ of early growth surrounded by hills of late-growth. This wasn’t an issue in quartersawn and riftsawn areas as much as in the plain-sawn areas, so wouldn’t prevent me from considering red oak again for projects where quartersawn will be used for it’s stunning medullary ray patterns.
When using waterborne finishes (the same is likely true for other spray finishes) it’s tricky to get just the right amount on the piece; enough to get the droplets to flow together and flatten, but not so much that deep grain pockets are bridged (leaving gas pockets underneath) or drips form on vertical surfaces. Too little and it dries too fast and forms orange-peel, a surface that literally looks like the texture of an orange peel, and needs to be flattened by hand, and often re-coated.
On the Woodworking:
Make many like parts at once -It would have been more efficient to cut all of the scalloped supports from one wide piece, with the tenon shoulders established before separating them into individual spans. This would reduce the handwork required to fit each one individually, and assure proper fit in a more productive manner.
Get a stock of air dried woods- Air-dried works so much more nicely than kiln-dried. Even in hand selecting boards, I found checks, honeycombing, case-hardening, and all kinds of internal stresses that made stock preparation longer than it should have been, and waste higher than necessary. I have a sizable amount of ash, mulberry and maple air drying in my yard. I have a good source of air-dried quartersawn red oak (just not plain) and I’m working on getting a few logs of cherry and walnut (my preferred woods) to mill and stack for future projects so I know the quality and provenance of the woods I use.
Tannins are a PITA -I’m not accustomed to working with red oak, but I was surprised every time I scraped, planed, or chiseled, my hands would turn black, and the black would rub off and stain the light oak any time I touched it. Nothing, and I mean Nothing that I tried, could get it off of my hands, but time (days). This oxide is the product of tannic acid in the oak, oxygen, iron from the steel, and moisture from my hands.
On the Project as a whole:
It’s good to have clients who want it done well – in this case, my client preferred to delay delivery until the finish was redone to my satisfaction, even if it meant delivery after Christmas. I was able to get it done to my satisfaction before Christmas, as it happens, but not without extra effort, which I’m happy to deliver for a client who wants good quality work. I appreciate a client who expects good work, and is willing to wait if it means the difference between ok work, and great work.
There are few good guides for dimensioning for wine bottle storage, so it took a bit of research. If anyone is building wine storage and wants to know the differences in spatial needs between a pinot bottle and a standard bottle or a magnum, let me know, I’m happy to share what I’ve learned. I prototyped it, and drank my share (and a bit more) of wine to get enough empty bottles to test it. This piece holds almost 4 cases, including 10 positions for magnums.
The woodworking community is a phenomenal resource- knowledge, feedback, tips, inspiration, motivation, sources, supplies, leads, collaborations, and more. If you’re a woodworker, get online and get networked. Here are some places you should be:
- Twitter, specifically #Woodchat (using tweetchat) on Wed. Nights, 9PM EST. If your new to Twitter, start by getting in touch with me (@mansfinefurn) and see the group of woodworkers I follow.
- Google +, a growing and big community of woodworkers. Get in touch with me on G+ here, I can help get you started by linking you to my woodworking circle.
- Modern Woodworkers Association, a new group of connected woodworkers from around the nation
- WoodTalkOnline, forums, links, all kinds of resources…
- The Wood Whisperer, a great free resource. Or, join the The Wood Whisperer Guild (you get the finishing e-book I mentioned free with membership, as well as a wealth of other resources)
- The Hand Tool School, a great resource for hand tool users, and loads of other good woodworking info.
Thanks to all who’ve been following along with progress on this build. As always comments, questions, and suggestions are welcomed!
A few parting shots (though not the best photography) of the finished piece in it’s new home: