Wood stringer in Torsion due to railing loads 1

[VA-Struct-Engr]

Hello everyone.
I am designing a wood stair stringer. The member size is 3 1/2"x14" Glulam (exterior use). The treads (3x12 SYP) are attached to the stringers with Simpson TA9 angles. See attached PDF for a couple of details we are using.
The railing newel posts are at 4'-0" o.c. (1 1/2" pipe) and is attached to the top of the stringer. In the event of a lateral live load of 50plf at the top of the newel post (per ASCE 7), torsion will be applied to the stringer.
First of, does anyone typically look at torsion in a stringer design (be it wood or steel)?
Second, how do I go about looking at it from an analysis point of view? The treads will provide some support, but not sure how to quantify it. The stringer in the far side is against a 2x- wall, but not attached to the wall, so it will help somewhat but not sure how much.
Thank you for your thoughts in advance.

[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1582292072/tips/stair_stringer_details_yyajgd.pdf[/url]

 

[DaveAtkins]

This is one of those times when I would not try to design by the numbers.

There is a lot going on. All of the treads and risers are attached to the stringers, so when a person pushes on the guardrail, the load is distributed to more than one tread/riser. The torsion is ultimately resolved as a couple between the two stringers.

DaveAtkins

 

[VA-Struct-Engr]

DaveAtkins,
I would agree with you. However, I have to satisfy the permit office. They have specifically requested calculations to prove that the stringer has the capacity to support the railing loads.

 

[azcats]

I'd be more concerned about the cross grain bending where the post attaches than the overall torsion in the member.

Any way to turn the torsion into a couple going into the treads (kinda like a little diaphragm) that is then resisted by another couple where the treads attach to each stringer where the final couple works in the strong direction?

This is clear in my head but I prolly need to make a sketch...

 

[XR250]

How do you plan on attaching the post to the top of the glulam? The tension component of the force couple will be quite large.

 

[JAE]

 

[AK4S]

I agree with JAE. Need to check if the clips connecting the treads can take the pull away tension. It helps that there are multiple treads/clips in action per guardrail post.

Not to highjack this post, in a similar situation: wood decks with guardrail posts, the peripheral beams have deck planks nailed to them, however I see connectors (

http://embed.widencdn.net/pdf/plus/ssttoolbox/zqzekua0ld/F-DECKCODE20-p029-032.pdf)

specified to transfer the tension inward, rather than relying only on the peripheral beam (and torsion resistance offered by the connected deck planks).

These connectors avoid torsion in the beam and concerns with cross grain bending where the post attaches (as pointed out by azcats).

 

[KootK]

I'd be more concerned about the cross grain bending where the post attaches than the overall torsion in the member.

This is something that I've always struggled with. Consider:

1) Cross grain bending is bad because it produces cross grain tension. Check.

2) You can, I believe, install a lag bolt in the top of a glulam and use it in tension which would induce cross grain tension in the member.

3) In imagining a stanchion mounted to the top a glulam, the only difference between this and #2 is the addition of the compression force of the moment couple.

So are we really saying that the #2 case, which is okay, is made not okay by the addition of the compression part of the moment couple? I'd actually think that the compression would improve matters.

KootK no understando.

 

[phamENG]

For what it's worth, the APA Engineered Wood Design Guide calls for an 18% reduction in allowable shear strength for radial tension and torsion. Don't have time to dig deeper into that, but the reasoning behind it may alleviate some of the concerns. See Page 21

On a slightly different topic, I'm not sure I'm excited about your connection to the landing. I'm pretty sure this is right out of the "Timber Connections to Avoid" article I read the other day:

 

[azcats]

I share my hypocrisy KootK.

I wouldn't have even considered the cross grain bending had the post been attached to the side of the stringer even though technically it is (could be anyway) the same mechanism.

Maybe it's because bending stresses aren't uniform?

Or just something I've blindly followed forever b/c someone told me 6 months into my first job.

I haven't done much wood lately - does NDS offer design values for cross grain bending? Tension perpendicular to grain? How do they compare?

 

[KootK]

I share my hypocrisy KootK.

You misunderstand me. I wasn't questioning your logic; I was voicing my identical concern (and personal confusion). Like you, my very first thought was "cross grain bending problem". Viscerally, glulam bothers me even more than rough sawn because of obvious laminations. I'm sure that's just in my head though; the glue's probably stronger than the wood.

 

[human909]

It seems a reasonable request to show appropriate torsional stiffness and strength. I've seen steel walkway construction where torsion is a problem. It certainly can catch a designer off guard when they are used to just throwing railings onto the edges of beams and not worrying about it.

 

[KootK]

One issue that I would have with considering the beam in torsion is that you'll collect a fair bit of that torsion over the length of the beam and need to resist it at the end connections. Torsional resisting timber support connections details that respect shrinkage and cross grain bending concerns can be difficult to develop.

 

[azcats]

You misunderstand me. I wasn't questioning your logic; I was voicing my identical concern (and personal confusion).

Nah - Pretty sure we're on the same page - I was really just agreeing w/ you and talking thru my thinking. Although I can see that I came off defensive.