Free standing pavilion 4

[DoubleStud]

So when I have a free standing pavilion/shelter with gable roof but without walls, what are the things I need to check if it is framed with wood? There is no wall. So that means the lateral wind load has to be resisted by the columns and diagonal braces. Shall I just go ahead and specify something to add some kind of moment resistance at the bottom of the post (simpson embedded post base)? Do I just run analysis and make sure the lateral deflection is below certain number? Do I just model it with the braces and assume the base is pinned? How accurate is the deflection if I model it wood? I am afraid the thru bolts give too much play and not modeled correctly on Risa.

 

[jayrod12]

This is a way more serious question than most people realize.

Generally speaking I'm a bigger fan of discrete bracing in wood structures and would consider the bases pinned.

You'll find that the Simpson moment base doesn't have the capacity you likely need.

Please ensure you have an experienced mentor review your work. There's a situation locally here where one that was essentially just completed collapsed as it had no lateral system. The designer in this case was not qualified to do the work and failed to have his drawings reviewed by anyone. He's going to be in for a hurt real bad with our professional association.

Edit: found a news article about the one that collapsed near here pavilion collapse

 

[driftLimiter]

Could look at knee braces to get some better rigidity at the top of the column. Simpson knee brace connector gives tested load capacity. Also watch out for uplift on the columns and footings.

 

[DoubleStud]

Thanks guys. Yes I am aware of the problem it can cause. But when I add knee braces, what checks do I need to do? Just model it with risa, add pinned braces and look at the deflection? What deflection limit would you consider laterally? I am worried about the initial deflection before the bolts are engaging.

Simpson knee braces will not work because it is not visually pleasing. This is for a high end pavilion with nice heavy wood ridge beam and rafters.

 

[driftLimiter]

I would model the knee brace connection as pin pin. Of your going to through bolt it to the column you'll want to carefully look at edge distances and direction of load on both the main and side member.

Deflection limit for this is a serviceability issue as long as you include p-delta in your analysis for strength. Are there any finishes or other elements that might require limited deflection?

My gut says l/120 is a reasonable limit at the service level but that's just a judgement I'm making without all the info so take it with a grain of salt.

Simpson also has some concealed connectors you might be able to use for the knee brace connection. Other thoughts are long timberlok screws that you can flush up or countersink and plug.

 

[XR250]

Yes model in software but the devil is in the details and this is likely where you will need the most help. Figure how all those knee brace loads are being resolved into the beams and columns and then check their connections where they meet at the corner.

 

[TLHS]

Along these same lines, does anyone have reference on the through bolt connection detail for knee braces that's pretty common for these. It's a bolt (or worse, a lag screw) at a weird angle where the breakout involves a variable depth piece of material because of how it interacts with the end cut. It seems like a hell of a thing to calc out, but I also see it all over the place.

I suspect it's probably just a thing people do and it doesn't break so...

Even strong tie has the detail in one of their examples on their site.

https://blog.strongtie.com/your-diy-pergola-makes-everything-even-cooler-part-2/

 

[jayrod12]

If this is truly a heavy timber type design, I'd be expecting to see mortise and tenon joints at most locations to maintain that look.

At least that's what we've seen on these types of things.

 

[driftLimiter]

I've seen some kneebrace joint details that are kind of Hybrid heavy timber and conventional light frame. You can dap the post a few inches to let the knee brace sit inside, this gives the traditional look, but more importantly it gives you a bearing surface for the shear load that is parallel with the post. When done this way the fastener can be designed to resist the component of the knee brace that is perpendicular with the column.

What detail are you looking to use for the connection I see you say through bolt, its at a weird angle but NDS gives all the equations you need to solve that. Its quite a lot of calculation but its doable.

 

[JLNJ]

Maybe you can convince your architect to use masonry piers and you can cantilever them up from the foundations. This would simplify the structure.

 

[HouseBoy]

I agree with a liberal deflection limit for an open structure (L/120 or so)

Connections (at the knee brace) are going to be the tough thing!!
I like driftLimiters idea of some sort of ledge to bear on. Still, knee brace forces are likely to be very high and tension connection will be hard to get right. Column stresses can also be high at the brace location (unless you're using something fat).

It is sometimes hard to achieve moment capacity in a wood post base. I have spec'd Simpsons MPBZ but it seems almost hard to believe the values and requires fairly specific rebar placement. I also like Parma-Column "Sturdi-Wall" bases (used in pole buildings). Not sure if they would fit your aesthetic....

 

[AgMechEngr]

I am working on a similar design. A good reference to help you along would be the National Frame Building Association's Non-Diaphragm Post-Frame Building Design Guide. Also, NRCS released a prescriptive design guide based on it that you can find for free online. It covers pinned, fixed, and embedded base conditions.

 

[KootK]

What I do:

1) Add knee braces if they would be architecturally tolerable.

2) Per the advice of the Timber Frame Engineering Council, ignore the knee braces in tension. Consistent with that, tie the knee brace into a recess in the post which makes all manner of sense for a compression connection.

3) Per #2, use the model below and a dirt simple P-delta FEM analysis to to see if the thing flops over under any plausible load case. I consider the replacement of the knee brace with a moment connection at the beam to column interface to be a somewhat reasonable way to account for the connection slop in the knee brace connections.

4) Button up the design with the member and connection checks I which deem to be far less important than #3.

 

[TLHS]

Any chance you know where the Timber Frame Engineering Council has that recommendation. It makes sense, but I want to jam it somewhere that I can find it next time I run into this.

 

[DoubleStud]

KootK, Are you saying to just fix one corner and don't add diagonal brace member on the model?

 

[manstrom]

I sink the posts in a footing and fix the base of the column. You can wrap the post to get more longevity out of the wood. Check the posts for combined bending and compression from wind / snow load. Knee braces will only take you so far (and your column won't like the point load).

When I am working on a problem, I never think about beauty but when I have finished, if the solution is not beautiful, I know it is wrong.

-R. Buckminster Fuller

 

[KootK]

Any chance you know where the Timber Frame Engineering Council has that recommendation.

I'm afraid that I'd have to go digging for it and, as such, may never get around to it. My TFEC stuff is in the form of a bunch tech notes, draft standards etc. Most of what I have from TFEC came from this web page: Link

KootK, Are you saying to just fix one corner and don't add diagonal brace member on the model?

Exactly. At the least, that's what I do for the P-delta check which, in my opinion, is most of what matters for these kinds of structures. I usually run two frame models in the same FEM file:

1) The model without the knee brace for P-delta investigation and;

2) The model with the knee brace to streamline the checking of members and connections.

 

[XR250]

The connection of the beam to the column is usually the hardest part (given Koots suggestions). Lots of force going on there from the knee brace loads and usually congestion for the perpendicular roof beam.

 

[DoubleStud]

For the P delta check, do I basically look at the deflection from lateral load, multiply that by the reaction on the top of the column and get a moment? Then add that moment to the top of the column as live load moment?

 

[KootK]

For the P delta check, do I basically look at the deflection from lateral load, multiply that by the reaction on the top of the column and get a moment? Then add that moment to the top of the column as live load moment?

That would be pretty close if your planning to do this by hand. Normally one would convert the P-Delta moment to an equivalent, additional lateral load and then cycle through a few iterations until the thing converges (or not). Will you not be using an FEM package with P-Delta capability? That is what I normally do. Subdivide the posts and beams into at least four mini-members and run the model with geometric non-linearity enabled.