Lag Screw Capacity w/ Shear and Bending 3

[UTvoler]

Hi All,

See attached; what do we do with case where a lag screw is loaded in shear and bending? Thinking of something like a kneewall clip or handrail bracket; does the load have to resolve to shear and tension only in the lag such that there has to be some component of compression on the supporting substrate from the bracket due to a resisting couple? What about a gate hinge lag that has no bracket whatsoever?

I don't find any provisions in NDS that would allow the determination of a lag's capacity when subjected to bending directly. What are your thoughts; all input is appreciated!

 

[haynewp]

I never put bending through the threads or shank of any screw or bolt. I guess you could calculate the section modulus of the solid part of the thread, but it’s not a good idea in my opinion. Use screws for tension and shear with no standoff in bending.

 

[Eng16080]

For cases where there is no gap/space between the applied shear force and the resisting material, the Yield Limit Equations of NDS, Section 12.3.1 do account for bending of the fastener. Appendix I and Figure I1 in particular provide a decent explanation. In cases where there is a gap/space, AWC TR12 accounts for this, I believe.

Beyond that, for a fastener with combined lateral and withdrawal loads, refer to NDS Section 12.4.

In the case where the end of a fastener sticks out from a piece of wood and the load is applied at the end, I'm not sure if AWC TR12 has a means to account for that. If not, in that situation, I might design per the equations of Section 12.3.1 as well as provide an additional bending check of the fastener using the root diameter for the area and checking it as if it was a cantilever beam (with a circular cross section).

 

[phamENG]

The moment clip and the guardrail thing are just tension and shear in the threaded portion. In both instances, you have a force couple consisting of bearing pressure under the plate and tension in the screw that resolves the moment and friction and/or shear in the screw to resist shear.

For the gate hinge...I wouldn't look too closely at that. If you're hanging a standard tube gate from Tractor Supply, they work (if opening and closing the gate at the end of my driveway everyday counts as extensive testing, I have tested them extensively). If you need more, then design a bracket that you can be comfortable with. Those hinges are likely a 'tested assembly'.

 

[haynewp]

If it’s soemthing like a gate hinge lag that has small loads and isn’t very important then yeah. But I wouldn’t advise someone to actually design lag screws to transmit bending to wood as a general design consideration.

 

[Stress_Eng]

You may find looking at strength of socket joints useful. A key aspect to these types of joints is that the maximum bending moment and shear force are inside the socket. Admittedly the loading is bearing only, and threads are not included.

 

[UTvoler]

Thanks for the feedback all.

@Eng16080 I have not had enough time to really study NDS/AWC TR12 on this particular topic, but it seems that the fastener bending/bearing is generated due to the lateral load in pure shear and the bending/bearing occurs due to the side/main member interaction. To @phamENG's point, I believe the applied moment has to be resolved to lateral/tension components and compared to the allowable lateral/tension NDS gives us. I don't believe NDS or TR12 directly provides an allowable bending check; TR12 "plugs in" the dowel beading/bearing capacity to arrive at an allowable lateral load. How do we arrive at an allowable moment via NDS/TR12?

@phamENG, the force couple route is my normal approach. But, as the plate gets narrow the moment arm gets small and the lag tension and bearing pressure gets astronomical. Seems like there should be some solution that combines the plate bearing, bolt bearing/couple in the main member, and dowel bending. Probably requires testing...

We'll leave the gate hinge alone.....I'll probably just punt on this one and require a thru-bolted detail; I'm not seeing a solution with a lag connection.

 

[phamENG]

Seems like there should be some solution that combines the plate bearing, bolt bearing/couple in the main member, and dowel bending.

I think the issue will lie in the longevity of such a connection. You could test something like that and prove it works...but what about how long it will work? Though I don't think it's stated explicitly anywhere, the implied service life of structures designed to building codes in the US (IBC, ASCE 7, etc.) is 50 years. I get that primarily from the return periods for live loads and from the Beta values in chapter 1 of ASCE 7.

So, would a connection like what you're talking about still be reliable after 50 years? My guess is no. I don't expect the gate hinge to last 50 years...but it's just a gate.

 

[Eng16080]

it seems that the fastener bending/bearing is generated due to the lateral load in pure shear and the bending/bearing occurs due to the side/main member interaction

I agree with this description.

I believe the applied moment has to be resolved to lateral/tension components and compared to the allowable lateral/tension NDS gives us

Correct. In my comment above, I was referring to NDS Section 12.4 which is used to check combined lateral and withdrawal (tension) loads. So, per your recent sketch, you have an applied moment acting on a base plate. The left end of the base plate will bear on the wood below in compression (as I think you're showing in the sketch) and the lag screw will be in tension (withdrawal). If you combine that tension force with your lateral force, there will be a resultant force which is acting almost vertically (since the tension force is significantly greater than the lateral). The angle of that force from the horizontal is equal to the angle alpha used in the equation of Section 12.4. In your case, you could just check withdrawal alone and get almost the same result. With the loads per your sketch, this will essentially have no chance of working.

Seems like there should be some solution that combines the plate bearing, bolt bearing/couple in the main member, and dowel bending. Probably requires testing...

I think the procedure outlined above is probably reasonably accurate, or at least, I don't see anything in this particular connection which would add to the strength in a significant way which isn't already being accounted for. In your particular case, I see the failure essentially being the plate pulling up on the fastener head which is then being resisted by the engagement of the screw threads embedded in the wood below. The lateral force also contributes, but, again, is almost insignificant in comparison.

If anything, the biggest unknown here is probably where the resultant compression force occurs between the plate and wood surface and thus the moment arm length. We were actually debating this in a recent thread with there being no clear conclusion.