: Although I think your solution will work just fine, my problem isn't the
: strip closing the kerf, so far I have that under control. My problem is
: that the plank itself is closing the kerf on the last few feet of the cut.
: I cobbled up a stout steel splitter and I'll see if that helps ... no
: reason it shouldn't.

: Ken

It's the same solution for this problem.

Waht you are encountering is because there is no support for the board after it passes the blade. Let's assume you have an 8 foot long board. When you start the cut you could have the whole length guided by the infeed fence. once you get past the midpoint though, you have progressively less and less boar supported by the infeed fence, and more and more board coming out behind the saw blade with nothing to support it, or prevent it from moving into the blade. If you could fill the kerf completely with small spacers that were the exact same width as the kerf, then you could hold the board against the outfeed side of the fence, and the strip and the kerf-filler combined would serve to keep the boar in line. Dropping in a bunch of spacers every foot or so will work, but it slows you down, and there are some safety concerns.

On the otherhand, when you have that "island" located behind the blade, and in line with the outside edge fo the blade, you have two control surfaces. (on from the infeed and this one on the outfeed. Unless the fence or that "island" slips there is no way the board can get past them to pinch the blade.

Even if the board is under great internal stresses so that it bends (warps) as it is being ripped, if the "island" is about a foot behind the blade, then the most pressure you can get from the relaxaton of those internal stresses will be from the amount of distortion present in just the exposed foot-long section of the board. It won't be very much.

The geometry is going to be of four inline points. Three of them will be in line with the outside edge of the sawblade. These are the "island", the trailing edge of the blade, and the leading edge of the blade. Any "set" in the teeth of the sawblade will provide some clearance alongside the blade itself. The fourth control point is the area in front of the blade where the board rests against the fence extension.

If the board is not straight in the first place then the high and low spots will be shaved off with the first few passes, much like a jointer straightens the edge on an irregular board. After that, you have a straight edged boar riding against a straight fence, as it goes into the saw. The leading edge fo the blade cuts the wood, leaving a kerf that is slightly wider than the thickness of the metal body of the blade. If the wood should bend into the blade at this point, the trailing edge of the blade will hit it (lifting it slightly from the table -- here is where you want fingerboards and hold downs and anti-kickback pawls) and effectively widen the cut kerf. A foot further the wood encounters that "island" and has a new bearing surface which prevents it from easily twisting back into the blade.

The idea is very similar to the common split-fence arrangement used mostly on shaper tables. There the fence after the cutter is adjusted to support the work. The only reason you can't use a standard shaper fence on your table saw is that there would be no place for the strip to go. It would hit the outfeed fence. So, you make a simple tool from a "C" clamp, and maybe some scraps of wood, that serves the same purpose, but is thin enough that the strip can pass around on one side, while the other side supports the board being cut.

Try it. You should certainly have an extra clamp to experiment with. If you don't use it here, you'll use it elsewhere.

PGJ