The top became loose and then came off. The top has some minor scratches in its finish (oil based gloss poly of some brand) as well as a few other blemishes. Not really scratches but more of indents in the finish from writing with a pen and it telegraphed through?
I think I should be able to just scuff sand the top and apply more poly. Good idea or no?
Not sure how to reattach the table top. It was glued (hide glue) with a wedge into the center of the post.
Any suggestions?
Thanks!
The wedge didn't hold
The whole table
Scratches in the finish. Would it be ok to just scuff sand the top and apply a coat or two of oil based poly?
First I would try a minimal approach. Lightly sand with 600 grit and then buff wit 0000 steel wool or Mirka Abralon. Might just look okay as is, unless you want it to look brand new. Poly gets very hard, and it will need a mechanical bond to adhere to the old finish. It won’t chemically bond.
To reattach the top I would remove the sub-top from the top and check the fit. I might remake that piece to get a better fit. I would look at removing the wedge or sawing a slot into the existing wedge and adding another wedge. Make sure the wedge is perpendicular to the long grain for the sub-top. That is, wedging against the end grain. I might use epoxy. I might introduce a couple of pocket hole screws from the sub-top to the pedestal.
As for the finish, I agree with Dick. Patina has to start somewhere. I doubt you can steam the dents out without ruining the finish. If you want to you can probably scuff sand the top and apply a coat of de-waxed shellac and then a coat or two of urethane. Sanding between the coats of urethane to level the surface. With care and enough coats you can fill and level the dents.
Pick and choose, some of the suggestions might be worth what you paid.
As the others have said, polyurethane does not stick to cured polyurethane. OTOH, shellac sticks to anything, and anything sticks to shellac (especially dewaxed).
I am curious why your hide glue turned white. I would only attempt to reactivate old hide glue with new hide glue if the old hide glue was still brown and transparent. So, I would remove as much of it as possible with hot water, and then rebuild the joint and reglue it with fresh hide glue. If the joint is loose, you can take a shaving, or several as needed, brush on hide glue and wrap the shavings around the tenon. Also, you may need to replace the wedge with one that is thicker, in order to get a tight fitting joint.
1. Too much glue. Was the fit loose? None of the furniture glues are gap filling with hide being the worst. If joint is loose make a new mating piece (see below).
2. I am not fond of wedges. If the joint is properly tight the wedge can't and hence is useless.
3. It failed because the lever arm of the top vastly exceeds the resistance of the too short tenon. The tenon should have been at least half again as long mated into a piece half again as thick. If there was a way of making the tenon longer what would be the best fix.
4. A solution is to replace the piece with the hole with one half again thicker. After gluing the tenon into a hole of matching length drive a robust screw (or three) through the top of the mating piece and into the tenon, thereby locking the two together with a mechanical fastener.
Clean up the top and cherish the signs that what you made got used. Next time test urethane product for hardness by- can you dent it with a finger nail test.
1) The grain in the spindle and the grain in the cleat should run the same direction, else seasonal movement can be an issue. Wedge should be 90 degrees to the grain in the cleat.
2) Maybe slightly taper the cleat's top side (like a hammer head) to give the tenon something to flare into.
3) The shoulder on the spindle is really small. Too much prying action on the tenon when a load is applied to the outer end of the top.
4) Maybe re-turn the spindle with a slight shoulder below the current tenon, turn a collar 2 to 2X that of the spindle diameter, 1-1/2" long, and glue that in place to expand the bearing area. Make sure grain orientation is followed during the glue up. Keep your spindle diameter larger next time.
I agree with Pete. I also think it’s a lot to ask if a single round tenon, there is just a lot of force on the extremes. I’d be tempted to build out a supporting batten.
This is interesting because I am setting up to make that very table based on the plan in the December 2016 Popular Woodworking.
In the plan, the tenon shoulder is 2 1/8 diameter. The tenon is 1" diameter and 7/8 long.
It seems like a single wedge in the center would not be able to expand that short/wide tenon much. The magazine shows two wedges dividing the tenon into thirds. These are more able to push the sides out.
Prepare a simulation of the joint- this dinky tenon mortised into the middle of a stick as long as the diameter of the table. When glue is set see how resistant the tenon is to force applied to the end of the stick, such as might be applied by someone leaning on the table edge. If it seems insufficient the solution is a square tenon that is larger and longer. The mating mortised board can taper toward the table edge to thin it where thickness is unimportant.
There are some good suggestions that could be incorporated in the next one you build. But this one should simply be reassembled as best you can and your daughter should treat it accordingly as a nice Shaker style table made for her by her father.
I've never seen hide glue go chalky, so I'm questioning whether you might be mistaken about that. It looks like PVA glue that was over the hill when you used it. Regardless, the wedged round tenon is quite standard construction for Shaker round tables and sewing stands, typically a 7/8"Ø by 7/8"L tenon into a 7/8"T cleat, with at least a 5/8" shoulder around. This tenon was typically wedged, cross-grain to the grain of the cleat. For larger diameter tenons, I'd suggest you use two wedges, closer to the edges of the tenon.
Here's yet another illustration of the shoulder of the joint providing the resistance to leverage that Bill discussed. The joint relies on the glue (and wedge) to keep the tenon from backing out, which holds the shoulder tight to the cleat and effectively resists pullout. Also, you want to be sure the leg is as dry or dryer than the cleat, so that it doesn't shrink away from the hole as it comes to equillibrium. (George Nakashima used to dry his chair rungs and spindles in a hot box so they would expand into their sockets when assembled.)
Eventually, depending on how the piece is treated, any compressible wood species can begin to crush or deform in a way that loosens the joint, but there are many examples of this type of furniture out there that have held up fine. I'm suggesting that fresher glue and a good, tight, wedged fit -- with a significantly wider shoulder relative to the tenon diameter -- will work fine.
Thank you and thanks to everybody who offered suggestions.
I understand this is a pretty delicate design. I've made a couple of these and this is the only that has failed. I made it the same as the Shakers did and the same as Christian Becksvoort has. I also know there are various designs too some of which offer better top attachment.
I have some ideas now, thanks to the comments, how to repair it. Nobody will ever know.
You could let in a long spline as shown below, the advantage being two additional glue surfaces with long grain to long grain. Sorry about the crude drawing.
The only example of Titebond glue "chalking" discussed in the Franklin "Trouble Shooting" technical literature results from application at too low temperature. Each product has a lower application temperature and some are as low as about 60 degrees and none less than 50 (from possibly flawed memory).
Chalking, as it is normally defined, results from the glue latex not coalescing as the water leaves the glue. The result are micron sized polymer balls loosely stuck together that have not merged into a strong film. Light scattering off these balls results in a white appearance to the inhomogeneous film.
I recently finished up a Company's sample of "yellow" glue I got at IWF 8 years ago. In my experience "yellow glue" gets thicker over time as water leaves the container but this aging only affects the ability to spread a thin film. If the coalescing solvent left the preparation chalking would result; I just have never seen this failure.
There is very little long grain to long grain mating in a round M&T joint. This table design would benefit from a rectangular M&T fastening the top to post. The male part could be easily made before turning the post and then the male part scribed to the top support to define the mortise. With proper grain orientation the joint would have a large long grain to long grain glue surface and not be subject to the seasonal stresses of round M&T joints. This is a reason that quality Period chairs were assembled with rectangular M&T joints.
