Riving Knife Question

dusty · Post by **dusty** » Mon Jan 03, 2011 7:20 am

Same title as another active thread but a different subject.

There has been some discussion about building riving knives (splitters) for use on the Shopsmith without an upper saw guard. This started my quest for for information because I was going to build still another (I have three).

Just how thick should a riving knife be?

Easy question. Can't be thicker than the kerf! Just a tad bit thinner than the blade seems to be the obvious answer.

Oh but not so obvious.

How thick are your saw blades? That is the question.

Mine are .093" (Shopsmith Thin Kerf Combination), .126" (Freud L72 General Purpose), .129" (Shopsmith Combination).

I also found several thin kerf blades that were a little thinner than my .093" from Shopsmith.

My Shopsmith splitter (part of the Upper Saw Guard) is .075". That is thinner than I would have expected and thinner than the last splitter that I made.

That brings up the question "Why do I use a splitter anyhow"? I think I use a splitter to prevent the stock being ripped from closing on the saw blade and thus creating a bind that can easily lead to a kickback.

Therein is the confusion. Should the splitter not then be as thick as the kerf of the saw blade being used at the time?

Hmmmm. Maybe it should be a wedge from about .075" to .125".

JPG · Post by **JPG** » Mon Jan 03, 2011 8:17 am

How about flat on the fence side and wedge shaped on the 'away' side. Ahh but then the fence would only 'work' one way, and alignment again rears its head!

So; I would at first think your suggestion is plausible, but how should it be 'aligned' to the blade?

Assuming a true wedge shape, the alignment would include the following goals. At the rear(widest, furthest away from the blade) the edge away from the fence needs to be no closer to the fence than the widest blade tooth set on that side. The edge closest to the fence needs to be no further from the fence than the widest blade tooth on that side.(the knive must NOT be closer to the fence than the workpiece width after being cut-otherwise a bind will occur).

As blades are changed, those locations vary. The fence-knife separation needs to be 'wider' to allow for the thinner blades, but that creates a more excessive(dragging) position on the 'away side'.

Conclusion: Anything wider than the thinnest blade will either pinch the workpiece to the fence or drag on the cutoff side. Anything thinner than the saw kerf will allow the back side to close in on the knife and thus also the blade. One would at first think the dragging is the lesser of the evils, but that would tend to pull the workpiece away from the fence(into the blade). Thinner IS the lesser of the evils, but some closure will be allowed.

So; there is no pat answer. There is no 'universal' answer. The only size that 'works' is one that matches the kerf and is positioned to align with the kerf relative to the fence. Anything else is a compromise.

dusty · Post by **dusty** » Mon Jan 03, 2011 8:29 am

JPG40504 wrote:...........

The only size that 'works' is one that matches the kerf and is positioned to align with the kerf relative to the fence. Anything else is a compromise.

That would have been my thought too until I started analyzing about this and realized that my splitter (from Shopsmith) is thinner than my thin kerfs and I have never had a problem that I would attribute to this relationship.

JPG · Post by **JPG** » Mon Jan 03, 2011 8:45 am

dusty wrote:That would have been my thought too until I started analyzing about this and realized that my splitter (from Shopsmith) is thinner than my thin kerfs and I have never had a problem that I would attribute to this relationship.

As I said thinner is the lesser the evils. Tis still a compromise. It will allow a slight closing in of the cutoff but I think the distance from the blade to the back edge of the knife minimizes the effect to an acceptable amount.

The important thing is that it prevents the workpiece pinching between the blade and the fence. Again thinner is better and again a compromise and the distance again minimizes the effect.

I think we can conclude they knew what they were doing by making it 0.075"!;)

anmius · Post by **anmius** » Mon Jan 03, 2011 11:46 am

I think that the riving knife/splitter is also keeping the wood (including the cut-off) parallel to the kerf. If the wood "cocks" a little, then the edge of the blade (either front or back) could catch the wood and throw it back at you. Keeping it parallel would help prevent that particular action.

Ed in Tampa · Post by **Ed in Tampa** » Mon Jan 03, 2011 11:48 am

Dusty
I think your thoughts are valid.

A splitter is to prevent the wood from closing on the blade and to keep the wood going straight after the blade.

For wood to pinch the blade it has to have some real force and this comes from the kerf completely closing. But if the splitter keeps it from completely closing there isn't enough pinch in the wood to cause kickback. A difference between the blade and splitter thickness isn't enough for the wood to pinch tight to the back of the blade and kickback.

The other thing a splitter does is keeps the cut wood side riding true after it leaves the blade. Again the really small thichness difference between the blade and splitter isn't enough to cause the wood to make a drastic direction change and cause a problem.

I try to match my blade and splitter thickness but as long as the splitter is thinner I usually don't worry about it.