Gluing Wide Panels and Warping/Cupping

Mike,

Experts (former spurts) like Jeff Jewitt say it makes no difference which way you turn the boards. They wil cup towards the moisture. Old tables always cupped up because they were washed down after each meal. Water on top, cup up. Look at your deck. The boards are laid randomly; some rings up, some down, yet they ALWAYS CUP UP—TOWARDS THE WET SIDE.

Bob Flexner also has stated the same position and offered up a lot of evidence in support. A lot of these ideas seem logical and make it to cult status causing us to worry unnecessarily. Most experts today say compose your top so it looks good to your eye and don’t worry about what the wood is going to do.

Steve

-- Steve in KY. 44 years so far with my lovely bride. Think I'll keep her.

Sounds like finishing both sides is best to control moisture balance then, even if one side, in this case, will never be seen. Thanks for the feedback Steve.

-- HorizontalMike -- "Woodpeckers understand..."

Actually Mike, Flexner and the other guy say finishing both sides is not necessary. Peter Gedrys agrees. I have never finished both sides if they won’t be seen and have never had problem. I built a trestle desk of yellow pine over forty years ago. At the time, my wife gave me hell for leaving the underside and back unfinished. For over forty years those parts have been unseen and uncupped/warped/ split, etc. More urban legends and old wives tales. Keep at it old son. You’re doing good.

Steve

-- Steve in KY. 44 years so far with my lovely bride. Think I'll keep her.

Resaw your glue-ups and do it again. Do you clamp the boards top and bottom when you glue them up?

-- Failure does not stop me, it makes me try harder..... because I'm crazy.

Yes Russ, I alternated clamps and used 6 clamps over the 22” span. Resawing is not in the cards at this point.

I just checked again this morning and the warp is ~3/32” toward the bark side. Looks like “bark side” will be my UP side. Since the four attachment screws are near the edges of the span (~1 1/2” from edge), I am sure that they will pull the cupping out of the piece. If the butt joints of these 7” boards ever fail, then I will indeed resaw and re-glue. In a way, with an “all-crown” panel the cupping is probably more even/consistent than it would have been had I alternated C-V-C, as that may/could have created an “S” shaped warping.

It seems that the soft nature of this maple(~950-1000) is more prone to the warping/cupping, as my last two projects with wide wide board panels, have stayed straight as an arrow once assembled. But then again one was White Ash (1320), much harder and the other maple panel, just two 7” boards wide C-V, was out of this same maple. No wash-boarding at all on the Ash, even though they were in a C-V-C-V-C sequence with narrower boards.

It looks like the (outer edge only) attachment points will make all the difference as long as I put “valley” up.

-- HorizontalMike -- "Woodpeckers understand..."

Mike,
Depending on who you listen to and what you read, we all do it differently. I was trained to close the wood on all sides to stop moisture finding its way in but it’s not as easy as it sounds, also cup up cup down is okay but the way it was cut before milling makes a big difference to – think stave or engineered material, even block tops which have proven that the orientation of grain should vary often and it helps to be in shorter joined pieces – another reason why good quality VC or ply with veneer is chosen… This is all well a good until you start building furniture and table tops where we want wide long grain, with solid material but unfortunately we will experience movement in most all directions and once it moves it’ll always demand the right to spring back!!!
In the meantime you can run it through my widebelt if you want;

-- Steve, The Hill Country, TX; www.themillworkguys.com

Steve,
I think I will take you up on that offer. Like I mentioned above, I managed to get it all joined with 13/16” in thickness, so I do have some to spare. Are you in the shop this morning, will be. Just let me know and I’ll swing by.

-- HorizontalMike -- "Woodpeckers understand..."

I know this is like closing the stable door after the horse has bolted, but I find it helps to put a pencil mark on the dry fit before gluing, that way the boards will go back together the right way when you get to your glue up.
With regard to fixing the top, I’d fix the front and leave some way for the back edge to move in or out for seasonal movement, either a slotted bracket or button. And definitely apply the same finish to the underside.

-- Do or do not, there is no try

Yes, I have slots in back and holes in front for screw anchors in order to take care of seasonal movement. And YEP, need to clearly mark beforehand. Lesson learned on that one for sure.

I THINK I GOT TO THE BOTTOM OF THIS ISSUE - It appears to NOT be traditional “cupping.” Upon close inspection, the 3-boards are each flat but are joined at an angle. Checked my TS blade = 90*, but noticed the slightest “rise” in my TS insert on the fence side of the blade. This was hardly noticeable, but when making 4-cuts (for 2-joints) with this slight error, it multiplied the size of the error. If I had actually gone with C-V-C, then the error would have been cancelled out since each pair of cuts would have opposite errors. Lesson learned. Looks like my leaving the board at 13/16” was a good thing, with Steve letting me run this thru his wide-belt sander. Looks like it should end up very close to flat at 3/4”

Looks like I need to finally toss that OEM metal TS insert and make a real ZCI to avoid this issue in the future.

-- HorizontalMike -- "Woodpeckers understand..."

Mike, its good to see you went after the root cause rather than jumping to conclusions. My boss always tells me to go to the problem wheel, because like most of us I tend jump to step 4:

1. Define the problem
2. Collect data
3. Form a hypothesis
4. Create a solution
5. Monitor performance (go back to 3 if it doesn’t work)

-- Brian

Mike, I enjoy reading your posts.

-- Bert

Mike,
OTOH I’m thinking if you would have glued as intended that slight angle would have met its complimentary slightly off angle. You would have had a flat glue up rather than cup.

-- Brian

Brian,
That was what I was trying to say above. It is definitely something to pa y attention to in the future. It may even be a “best practices” to guard against the potential of the TS being off slightly. I was thinking with the board ~1/16” cupped divided by 4-cuts, than each cut would have contributed to 1/64” “cup/warp”. Hard to know that until you do the glue up. “Murphy” will always win given the chance, after all it’s his “Law.” ;-)

-- HorizontalMike -- "Woodpeckers understand..."

Glad you found it, Mike.

For those interested, the reason why guys like Jewitt and Flexner say that the need for finishing both sides of a board is a myth is because finishes are not vapor barriers. They fight off moisture, but not water vapor (or humidity). Some finishes ward off vapor more than others, but over time, humidity still affects the work, finished or not.

-- jay, www.allaboutastro.com

Horz…give yourself a break…soft maple will want to move…and you did wide boards (I stopped doing that a long time ago regardless of species unless I know it’s old growth stuff that has been laying around for some time and is still flat).

on an earlier post you mentioned that the bottom will never get seen…a few hardwood cleats screwed to the bottom will work (pull apart any old oak table leaf and you’ll know what I’m talking about).