Herringbone Flooring

We decided to try our hand at herringbone floors, and it turned out to be extremely time consuming. Not counting sanding or finishing, it took myself and my wife 30 hours to nail down the flooring in this 14 x 14 foot room. Although you might think it would make sense to start in the middle of the room, the only way I could figure to lay this flooring (and still use a flooring nailer to nail into the tongues) was to start at one side of the room and work back and forth, from one end of the room to the other.

The pattern got so out of whack on the first two rows of flooring that I decided to rip it all up and start over. That was a waste of about 3 hours. With a false start behind us, we started cruising. But just when I thought I had things under control and had about a third of the room finished, the pattern started getting out of whack again. What I mean is that it became impossible to draw the pieces of wood tightly together and simultaneously keep the pattern running straight. I called a time-out and started thinking about the problem.

Our solution turned out to be simple but tedious. We sorted through hundreds of pieces of 3"x12" flooring, measuring the width of each piece with digital calipers. Although they were supposed to be 3 inches wide, manufacturing tolerances and shrinkage (or grow-age!) differences had caused the all of the pieces to be a little over or under 3 inches. We labeled each piece according to how many thousandths of an inch over or under 3" it was. Some were as wide as 3.050" and some were as narrow as 2.980" with the average size around 3.010". It doesn't seem like much of a difference, but randomly grab three pieces in a row that were 3.033" and you would create a 1/10" of an inch gap that could not be drawn tight. Throw into the equation the fact that the lengths were not exactly 12.00", and you can see how the pattern could diverge quickly.

With the flooring sorted into piles (the piles looked like the textbook version of a bell-curve), we set out to get our pattern back on track. Before nailing, I would test fit each piece and either accept it or ask for one that was 0.010" larger or smaller. The solution worked, and we were back on the job. We were determined and able to get the joints extremely tight. Who knows, maybe there will be cracks everywhere as the flooring seasons with the house, but it should be fairly stable because it is quarter-sawn white oak.

In these pictures, I have 3 coats of oil based polyurethane on the flooring (and one heavy coat of general "house-under-construction" dust). I applied two coats of gloss and one coat of satin, sanding lightly between each coat. I'm not entirely satisfied with what-was-to-be the last coat, so I plan to sand and apply at least one more coat of polyurethane. Tedious, tedious, tedious, but probably worth it. When my older son moves out of this room, I'm going to claim it for my office, if my wife doesn't beat me to it.