Disappointing fail for Octogon table

[Embarrassed Newbie]

First time poster here.  I've been woodworking for a year and had some successes and a few failures.  Here is my greatest fail.  I did not think about wood moisture content (I've since bought a meter).  These were made in the garage when it was very humid.  Wood was Brazilian Cherry and Maple.  The project is an eight sided table with a top that would hold a marble tabletop from India.  The marble is translucent, so it I installed a strip of LED lights inside the table.  Don't worry about the stuff on the inside because that would have been inside a sealed top and bottom.  That putty was needed to make the top sealed from light leakage. 

Obviously the problem is the splitting of the wood and separation of the panels.  Each of the eight panels are strips of Brazilian Cherry, maple, and another Brazilian Cherry.  The intersection of each panel has an inlay of 1/4 inch of maple to hide the gap.  For strength and support are three rings of plywood.  The design was mine (probably the problem!) 

It was beautiful (if I say so myself) when it was completed, but then I brought it into the house for three months before taking it out to give as a gift.  By then I had multiple splits in the walls as well as the top. 

I'm looking for guidance on how to fix (doubtful) or how to improve in the future. Thanks.

[Chet]

It does look like a moisture/movement problem.  My initial guess is that if you glued it to the plywood thats not leaving any way for things to move.  Plywood is pretty stable and your hardwood had no way to move.   Fastening the outside with small screws to the rings through slotted holes in the ply might have worked more in your favor.

[Embarrassed Newbie]

The panels were glued to the plywood rings.  I understand what you are suggesting about screwing the panels to the ring.  I'll keep that in mind if I try again.  I'd have to make thicker panels to screw into it.  But if it's an octagon so the panels are all attached to each other, wouldn't the movement still be a problem?  I can see a tabletop strinking/expanding and using jointery to allow for movement, but how would you do it when the circumference of the octagon can't change?

[Chestnut]

This seems to me like an excellent veneer project. If you veneered the sides over a substrate of mdf or baltic birch ply you wouldn't have had this issue. If you wanted to accomplish the same thing with solid wood the internal support rings were the demise. This still is goign to cause major problems for the top and that connection. The top will need to be floating or sorts where it is fastened on top with figure 8 fasteners or a similar method to allow the base to move around and not pull on the top. The marble also needs to be floating in to top to allow for movement. The board you cut the shape out of wanted to contract and because there was something in the way it failed at the weakest point.

I personally love Jatoba (Brazilian cherry), it changes color to a wonderful dark red with UV exposure. So make sure that it gets even light for the first year of it's life or you'll see stark shadows. They will become less apparent with time but may never go away.

[wdwerker]

If the boards in the base went sideways movement shouldn't be a big problem. Wood moves in width and slightly in thickness but very little in length. If the top was made of 8 strips with angled butt joints movement wouldn't hurt much there either.  Dowels, dominos, biscuits or even lap joints could work. Getting all 8 joints to fit as perfect miters can be tricky, angles and lengths must all match quite accurately.

Otherwise veneer might be a better choice.

[Embarrassed Newbie]

I'll seriously consider veneer.  I have some mahogany veneer left over from a project where I veneered an old Ikea dining room table (one of my better results).  Perhaps a solid Mahogany top, although the grain pattern won't match very well.  I've only veneered flat, 90 degree surfaces.  Can you veneer over a rounded edge?  I have preglued thick veneer on hand.

[Byrdie]

I'm with Steve (wdwerker) on this one.  I'm not seeing a problem due to movement but more likely due to joint failure.  The evidence is in the closeup of the broken joint.  Typically the bonding agent (glue) in a joint, when cured, ends up being stronger than the wood its self and you see failure, splintering in the wood.  When I look at the pictured failure I see strings of the bonding agent and very little wood.  Steve also raised what would've been my other point.  This is not the direction for movement.  Movement occurs across the grain, away from the center.  The forces that would've been reactive on this joint are vertical.  

So what I'm seeing is that something to do with the edges of the pieces being joined failed.  They either weren't prepped properly or a less than ideal type of glue was used.  I'm going with prep because there were other similar joints that are still working.

As far as the top goes, I'm still going with joint failure although I believe there may have been some twist imparted by drying as well.

All that being said, the construction looks a little clunky and there were likely better methods to go about constructing this.  I think a thin ply or MDF construction with veneer would've had better results and likely been more stable.

[Ronn W]

I tend to agree that the joints failed but I also think that it is a moisture (shriknage) problem and that if the joints had not failed the wood would have.  I like that veneer idea.

 

 

[Embarrassed Newbie]

Thanks so much for all the advice guys.  I've learned a lot and really enjoyed sitting and listing to the conversation.  I love this forum and the time that you guy give to a Newbie!