Deck Roof: Support on Cantilever

[midrover170]

Hi folks -

Time for my next question, but first, just wanted to say thanks for all the help given so far. Much appreciated.

I'm close to framing in my deck, but have worked myself into another situation that needs a little engineering...  . I don't need permits in this location, so I won't have an inspector looking at this detail. However, I'll be spending a lot of time on the deck and want it to be rock solid.

My site sits right on the edge of a pretty steep slope going down to USFS land. My design calls for the deck posts/girder to rest right near the edge (already installed - see picture below). The deck itself will cantilever about a foot and a half past this, kind "floating" out over the edge.



This is a 24' long deck, with 16' of it being covered by a simple shed roof. No enclosed walls. I'm planning a metal roof, 2x8 rafters and maybe some pine paneling underneath. I'd like to support this roof at the edge of the deck, like this:



There's lots of material out there saying to stay clear of this, being that the load path would not be over the foundation. There are also some uplift concerns. I get that, and really wish my design allowed for the typical code requirements. I've also read a bit from folks who have successfully braced their deck to allow what I'd like to do. I've drawn up a couple options to get the conversation going.

Option 1: Use a 4x4 at a 45 degree angle and support the roof post. I'd probably thru-bolt this to both the post and the deck floor system


Option 2: Use PT 2x6's, again at 45's, and support the deck floor system. I like this option more, as it seems stronger.


Any thoughts? Suggestions? Violent disagreements?

Thanks
~D

[OlJarhead]

The cantilever could be done strong enough to support the load, I'm sure, of a roof but with a flat roof, if you are in snow country, it better be one heck of a tough cantilever!

[midrover170]

Thanks, OlJarhead,

The cantilever could be done strong enough to support the load, I'm sure, of a roof but with a flat roof, if you are in snow country, it better be one heck of a tough cantilever!

It's a little to hard to see in the rendering, but at a minimum I'd be doubling the joists under the posts. I have a buddy that has access to a lot of steel and steel fabrication tools. I might see if he could make something for this...?

If I focus on the cantilever, should I ignore or still put in some knee bracing? Probably wouldn't hurt, right?

On the flat roof, yes, it's pretty flat (2:12). There are a number of reasons I've gone that low, but the primary is that I can't go any higher with the rafter ledger without running into the main structure's eave. I'd love to get 4 or 5:12, but if I want good head clearance at the edge of the deck, it's got to be looooow.

[Don_P]

You don't have enough particulars for an answer on the canti joists,

Violent disagreements, or more stuff to think about... you haven't braced another cotton pickin thing, why start now? At least brace the main structure, it looks wierd for a porch to be standing beside a collapsed house.
The edge of the slope you are perched on looks unstable, I'm assuming floating is the dream and sliding the more likely reality.
Are you planning on hanging the deck on the cantilevered rim of the main floor... and if so is that rim attached with joist hangers upside down to take the ledger load? Were the piers and girders sized to account for the added porch load?
When a cantilevered joist passes over a girder it should be blocked solid over that bearing point, you have that situation at the house and are setting it up again.

I would opt for options 1&2 combined, running the brace down to the pier pad. There is a vertical and horizontal component of the diagonal brace load. It corresponds to the proportions of vertical and horizontal legs of the bracing triangle. A longer, closer to vertical brace has lower horizontal loads associated with it. Connections are less stressed out and parts don't want to slide out of place as bad under load.

Draw in the rest of the building frame on that side, the wall and roof on the main structure in relation to the porch is what I'm looking for, there has to be a better way.

[midrover170]

Thanks for weighing in, Don_P

Violent disagreements, or more stuff to think about... you haven't braced another cotton pickin thing, why start now? At least brace the main structure, it looks wierd for a porch to be standing beside a collapsed house.

I'm planning to brace the main structure, as you've described. Doing the weekend warrior approach for a build doesn't always allow enough time to get everything where it should be before heading home. The walls aren't being framed for several weeks, and I'm planning to brace and block the structure before that happens. The cantilever on the main structure does not exceed the depth of the joists, for what it's worth.

The edge of the slope you are perched on looks unstable, I'm assuming floating is the dream and sliding the more likely reality.

You're right. The slope is close, and unfortunately not made of granite. Before any load is placed on the deck system, the face of the slope will be retained. That was one of the reasons the P.E. had me pour 48" deep, 12" dia piers with 36" dia., 10" thick spread footings. He was as concerned as you/me.

Are you planning on hanging the deck on the cantilevered rim of the main floor... and if so is that rim attached with joist hangers upside down to take the ledger load? Were the piers and girders sized to account for the added porch load?
When a cantilevered joist passes over a girder it should be blocked solid over that bearing point, you have that situation at the house and are setting it up again.

Yes (all the above). Everything was reviewed by the P.E. before I started. I've got a big box of joist hangers up there ready to hang out upside down. I will be blocking at the girder for the joists as well, as you pointed out. Thanks for the suggestion to support all the way down to the pier pad - can and will be done.

This is a little outdated, but this is what the wall/roof of main in relation to porch looks like. Thoughts after looking at that?

[Bob S.]

could you move the post back over the beam then cantilever the roof out to cover the deck?

[midrover170]

Thanks for bringing that up, Bob. I've been considering that as an option as well. This weekend, i'm going to calc what my overhang would be if I went that route. It might be greater than 2', in which case, I'd be in another trouble spot... Will let you know.

~D

[Pine Cone]

Looking at your photos you appear to be in the western US in a ponderosa pine ecosystem that is designed to burn with periodic wildfires.  If your deck roof joins with either the existing roof or a wall of the building you are building a building that will burn nicely when the wildfire races up the slope below.   Hot air will rise upslope as the fire burns and then get trapped under your deck and deck roof.   You might want to look into a more fire-safe design.  See the publication below for more info on wildfire safe building guidelines...

http://anrcatalog.ucdavis.edu/Details.aspx?itemNo=8393


Fires in that ecosystem aren't a matter of if, just a matter of when.  If you are up against the USFS then they won't be doing much land management to reduce fire risk since cutting costs is their major operating mode these days.  After 40 years in the forestry biz I don't have much hope that the US Forest Service will change their ways anytime soon.

You might consider having an air gap between the porch roof and the side of the building or other roof.  That would allow you to have a higher slope to your roof.  I also agree that you should move the roof posts back to be over the footing posts.  It would be stronger and you don't really use much useable space. 

[midrover170]

Hey Pine Cone. Appreciate the insight, especially given your experience in forestry.

You are correct. I'm in western Idaho (Boise National Forest). A healthy mix of ponderosa (although not so much in my area), lodgepole and Doug fir. Even though USFS is 20ft. from the building envelope, I'm fortunate (?) to be close enough to the Treasure Valley and it's very arid and not heavily forested around me - my southern exposure probably helps with that as well. There's actually a good 80ft-100ft from the porch to the nearest tree down slope. About half of these need to be removed due to beetle kill. The slope surface is primarily exposed fill and some weeds. The retaining wall that I'm planning (two runs of 4ft. high concrete block) will create some additional separation during a burn.

In fact, the only inspection the county conducts at my site is for the Wildland Urban Interface (WUI) to confirm adequate separation from combustibles. A lot of it follows Firewise (http://www.firewise.org/?sso=0) protocols, which is probably similar to what you linked (the link won't work). That looks for 30ft buffer around the build and I'm good on all sides there.

For the roof, I'm going to get the measurements so I can consider moving the posts back. That seems like the safest bet for supporting the roof. We'll see what the overhang calc comes back at...

You might consider having an air gap between the porch roof and the side of the building or other roof.  That would allow you to have a higher slope to your roof. 

So by an air gap, do you mean fully disconnecting the two? Or some other means?

[Pine Cone]

Maybe the revised link will work.

And yes, I do mean disconnecting the two pieces to get an air gap.  I did a similar thing for my build. 


My cabin is on flat ground near the Puget Sound (ocean) so I'm not too worried about fires here.

Lots more info on my build here....
http://countryplans.com/smf/index.php?topic=8030.0

[midrover170]

Hi all -

I've been spending some time with engineers on the deck roof quandary, and a solution is close. I'll most likely pull the roof posts back to sit on top of the beam, shorten the deck canti by approx. 6 inches, and have a 2ft. roof overhang (or slightly more). Designing it so that I have two 4x6 posts on either end of the roof, and have the space between (approx. 15' span) open and clear of posts. I'll be using a 5.5"x12" glulam for this span.

Here's my question (missed asking during talks with engineer): Is there much benefit to notching the 4x6 to sit on AND around the beam with thru bolts? Or, is it roughly as strong just sitting on top of the beam and then through bolted to the deck framing? Drawing it out in Sketchup makes the notched approach look much better... Probably adds some buckling resistance?

[John Raabe]

The notched option is stiffer and makes the two parts work together as a unit. Adds rigidity against unusual forces such as high wind or earthquake.

[Don_P]

I typically don't use smaller than a 6x6 for something like this, at this beam span check column buckling on those posts, I doubt a 4x will pass. This also concentrates pier loads, check their footing size.

I like notching in, that said a notch concentrates stress and if there is movement the split will tend to run out of that re-entrant corner of the notch. A professor got me to nail together a notched 4x4 porch railing post in front of a group of engineers in the university lab. He asked us where it would fail when we loaded the top of the post as in a railing being pushed on. We all pointed to my nails. The notch split out of the corner. He and the tech had a good laugh then they pointed out the grain in that area... they had picked a ringer but with a point in mind, you want good straight, clear wood around joinery. But, it wood gets into tension perpendicular to grain, that is the splitting direction, there isn't anything reliable there.

I have taken to embedding a 1/4x4" steel plate in between plies of the girder, routing a "dap" for the plate to rest in and bolting through the beam and plate. Then I deck the floor using a drill and jigsaw to let about 8-10" of plate extend up through the porch floor. Then with a long 1/4" drill bit, drill a series of holes into the bottom of the post in a line, connect them with the bit a angles and whatever, then run a sawzall in to clean out that blind mortise for the plate. Drop a post base over if you are using a standoff, then the post onto the plate. Drill through everything stopping just shy of the far side of the post and pin with steel rod, double shear connection (awc.org connections calc) small wood plugs cover the rod end. Check your uplift, it will be substantial on just 2 posts, whatever method you choose check the connections against up as well as down.

[midrover170]

Thanks John and Don.

I have taken to embedding a 1/4x4" steel plate in between plies of the girder, routing a "dap" for the plate to rest in and bolting through the beam and plate.

Don, that sounds like one tough connection. I wonder if there's a way to retrofit my beam to accept a 1/4 x 4" plate once they've already been built. Maybe if I loosened the carriage bolts, I could drive something like that through the liquid nails in there...

Your approach sounds kind of like a Simpson CPT if I'm following you correctly, but with some serious modifications of course.

If I go a notched route, could I attach some steel plating that ties the post to the beam (especially to the face where the notch seat rests on the beam)? I could be wrong, but this would help with the tendency to crack at the notch that you explained.