4. With shelves, it is not the strength of the screw that holds up the weight, but the friction of the joint. Make sure it is flush and tight.

[Sean]

[back to the list]

As I am typing this, I am not sure that I agree with it. It is counter-intuitive to me. I invite comments. The intent of this item is to provide insight on the failure mode of shelves and also the importance of square, flush joints.

It is my understanding and experience that it is not the strength of the screw which bears the load of a full shelf to the sides, at least not as the primary mechanism. Rather, the screw holds the shelf snug to the side, and it is the wood on wood friction which bears the weight.

The impact of this item is that if you do not have square cut wood that forms a flush joint, this friction will not be present and all of the load will be supported by the screw, much by the threads and wood around the threads. The angle of the pocket screw helps - this places a lot of the load is shear across the screw.

As the joint flexes, say with books added/removed, the screw can work out of the joint, weakening it to the point of failure.


Sean


[back to the list]

[Gene Maiorano]

You mostly have shear working for you; consider if the joint is sloppy and a gap occurs near the screw then you have leverage working against you, that is the intersection of each point of force, the edge of the wood and the length of the screw into the wood, minus the gap, creates unwanted leverage, toggling the shear and exerting a lateral (removal) force or moment.

[Sean]

Like I said, the more I think about it the more it bothers me too. The original concept, as explained to me, was that the coeff of friction for wood to wood joints was supposed to be greater than one. That would mean for every pound of force exerted by the screw in forcing the shelf to the side, that many more pounds of force resists loading on the shelf (ie, coeff = 10, screw tension is 20 lb, friction can support 200 lb).

But that is not correct - I looked in multiple sources and the coeff of friction for wood to wood is less than 1, more like 0.5 or even 0.25. So, friction is only 50% or 25% of the tension from the screw.

I was trying to get to this concept - if the joint is not tight or there is a gap, the joint can flex, both loosening the screw and weakening the wood around the threads to the point of failure. The screw will not fail in a shear mode, but it will back out of the material.

I would like to have some rule here that reflects failure modes from poor joints (wood not square, screw not tightened enough, other operator induced errors, etc.) Please comment on ideas and I will use them to revise this rule. Imagine KregRep has some resources on this one.

[Bob Sanders]

I strongly disagree with this. Image for example that you take a bar clamp and squeeze one shelf between two sides of a cabinet. Imagine a second shelf where you just simply screw through the face grain (no pocket hole even) into the shelf.... not tight enough to benefit from any friction.

Are you telling me that if you start bashing both shelves with your fists or a hammer that the shelf held in place with the bar clamp (though simple friction) would stay up longer? I don't think so. Friction certainly adds some strength, although probably not as much as glue, but no where near the strength of adding screws, especially with pocket holes.