Paul Kierstead":2yxykbom said:
pam niedermayer":2yxykbom said:
Big difference between sanding the wood and sanding the finish.
The article is -- in part -- concerned with the need to sand the wood if you are going to sand the finish as any scalloping left on the wood can play havoc since the peaks will have the finish sanded through. It is worth reading.
In theory it makes sense. In practice it depends on how the wood is finish planed.
Let's say you are planing at a 0.001" depth (very fine smoothing), with, let's say, a #4 smoother, with a 2" wide iron. Let's assume that the iron is cambered to that amount, i.e. flush with the sole at each end and sticking out 0.001" in the middle. Assuming full width shavings are taken next to each other with no overlap (which is more precise plane handling than most of us can achieve), the peaks would be 0.001" thick. I was just reading a finishing article by Chris Minick tonight, where he states that varnish coats are in the 0.003-0.004" thick range. In that case the finish would have no risk of being sanded through the scalloping.
There are lots of variables at play here. All other things being equal, if one chooses an oil finish, which leaves nearly no film on the surface, then sanding through could be an issue. In the same vein, if one uses a heavier camber and still takes full-width shavings, then the peaks will be higher, and more prone to sand through.
The way I was thought to finish plane at Rosewood Studio was to plane full width with a small camber, aiming for the 0.001-0.002" thick shaving range. The last step was to retract the iron to the point where it was taking only a little bit of "fuzz" from the middle of the iron, and take lots of overlapping strokes (it is faster than it sounds). My guess is that this took no more than half a thou off, possibly less, therefore lowering the peaks down to the un-measureable level (or close to it).
Another factor is that as we plane, with any kind of overlap in our strokes the plane sole burnishes the surface somewhat, reducing the height of the peaks further.
Although it would be interesting to make controlled experiments, playing with all the variables of influence, my take is that we don't need to worry much about sanding trough the finish (hasn't happened to me yet). If anyone has had such experiences I would like to hear about it.
My $0.02,
DC-C