Topic Options
Rate This Topic
#33466 - 02/13/16 02:27 PM Large/Wide Roof Overhangs
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
I see it frequently mentioned that large overhangs are essential for a long-lived, trouble free timber frame (or any structure) as well as sun/shade reasons. I was thinking a ~4ft overhang would make sense for me since there is a ~3ft walkway beneath it. Can an overhang this large be supported by just the rafter(s) of the roof frame(s) extending outward, or is a gusset brace (or even a vertical post) a guaranteed necessity for this size of cantilever? It'd be nice for the walkway under the eaves to be undisturbed by crossbraces with forehead-dents in them, and nicer to not have to make the wall taller to raise the braces up. FYI, this is to be located in central Texas, so snow loading is likely not as big a concern as wind-loading.

What type of rafter-extension is usually done for the really wide overhangs sometimes seen? I figured the usual birdsmouth jointing at the frame cap would eat up too much of your cross section to stick out very far, but those are the only type of rafter/wall joint I seem to be able to find detailed explanations of (well, roof sprockets count, but I was hoping for something that continues the same shallow roof incline as the main rafters)

TCB

Top
#33467 - 02/13/16 07:18 PM Re: Large/Wide Roof Overhangs [Re: TCB]
Jay White Cloud Offline
Member

Registered: 11/06/08
Posts: 480
Loc: Vermont
Hi TCB,

First... grin...Big Welcome to the TFG forum...!!! Glad you have joined...

I see you have two posts so, I will make this one my first to respond to...

Quote:
I was thinking a ~4ft overhang...Can an overhang this large be supported by just the rafter(s) of the roof frame(s) extending outward, or is a gusset brace (or even a vertical post) a guaranteed necessity for this size of cantilever?


Depends on the design of the architecture, the timber size/species, and load exposures for the area...Snow isn't your challenge at all, as I was just in Katy, Texas not long ago...however, wind is a challenge.

A 1.2 m (~4 feet) is not hard to achieve at all. Even 3 m (~10') overhangs are achievable without any obstructing members to block free passage, yet this size (the largest I have personally examined) is not a typical or easy design to facilitate within the context of most domestic architecture...

Quote:
What type of rafter-extension is usually done for the really wide overhangs sometimes seen?


If you mean by "type" the most common designs, I am afraid they would be voluminous to list here...Green and Green Craftsman/Prairie style is the first that comes to mind for a "homegrown" design motif, but there are many vernacular folk styles from the Middle East, Asia and beyond that also have very large overhangs.

For very local (and indigenous) design flavor the Creole/Choctaw styles that developed both pre and post Spanish/European migration to the areas of Texas have roots in local design motif...You can find Bousillage and Spanish styles both in Texas and these can have very large overhangs and veranda...

In the European modalites, large overhangs are not typical, yet the Swiss have excelled at this probably better than any other cultures of Europe. There are forms of timber architecture found in Eastern European that can have large over hangs and also the Moorish influenced styles found on the Iberian peninsula...

If "type" meant timber size...that would all depend on design...

Quote:
I figured the usual birdsmouth jointing at the frame cap would eat up too much of your cross section to stick out very far, but those are the only type of rafter/wall joint I seem to be able to find detailed explanations of (well, roof sprockets count, but I was hoping for something that continues the same shallow roof incline as the main rafters)


This is a complex and deep "rabbit hole" we are entering...It is so nebulous to approach without a "focal point" design as to be inhibitive without some form of goal or context in understanding what might be the possible target..If perhaps you have some more specific ideas or goals, perhaps then we could explore more specifics design parameters and then just begin to explore what might be possible...



Edited by Jay White Cloud (02/13/16 07:20 PM)
_________________________
http://about.me/tosatomo

Top
#33478 - 02/14/16 01:11 PM Re: Large/Wide Roof Overhangs [Re: TCB]
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
Well, it at least sounds like the concept is plausible; not having a great handle on roof design loads nor material properties at this stage in my 'training,' I wasn't sure if I was asking for the moon or a moon-pie.

"This is a complex and deep "rabbit hole" we are entering...It is so nebulous to approach without a "focal point" design as to be inhibitive without some form of goal or context in understanding what might be the possible target..If perhaps you have some more specific ideas or goals, perhaps then we could explore more specifics design parameters and then just begin to explore what might be possible..."
I kind of figured as much. I am trying to play my design a bit close to the vest at this point, since it is more theoretical than modeled in CAD at this time*, and since I'm leery of trying to focus on too many design aspects at once early on (though I do understand how they necessarily rely on each other)

Here is the summary, of the roof at least (my other thread describes the wall somewhat)
-Octagonal floorplan, roughly 40ft diameter (~16ft walls)
-Double-layer roof; exposed reciprocal truss inside, more 'normal' peaked/hipped roof outside
-Two large column posts (forget the architectural term for columns spanning multiple floors) at the center of the floorplan help support/stabilize the reciprocal ring
-Exterior roof rafters cantilever at least 4ft past walls to shade a narrow walkway
-Low roof pitch (20deg off horizontal is my starting guess) steel-cladded

The reciprocal internal roof truss is somewhat self-explanatory, but the exterior is giving me designer's block. I'm thinking something like this, only with three triangular 'facets' on the ends of the peak line for the octagonal planform; the 'broad' faces on the front/back would be rectangular, with a peaked vestibule jutting out over the front door.

It turns out it's really hard to make an octagonal roof look attractive, without a lantern or cupola at the middle, and not have it be a cone. Or have tons of up/down peaks, valleys, or dormers, and their associated leak liability. A real shame, since the faceted exterior walls always look attractive. I get the impression it must have been a constant vexation for Thomas Jefferson, & probably Da Vinci too, who apparently were quite taken with the concept themselves.

TCB

*I model aircraft structures daily, but houses are 'different,' so setting up the assembly structure, constraining design geometry, and the general order of operations one uses to efficiently approach such a big project aren't quite online yet


Edited by TCB (02/14/16 01:14 PM)

Top
#33510 - 02/15/16 08:00 PM Re: Large/Wide Roof Overhangs [Re: TCB]
Jay White Cloud Offline
Member

Registered: 11/06/08
Posts: 480
Loc: Vermont
Looks like your concepts are developing nicely thus far...
_________________________
http://about.me/tosatomo

Top
#33531 - 02/16/16 11:10 PM Re: Large/Wide Roof Overhangs [Re: TCB]
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
I appreciate the encouragement. Hopefully this roof concept isn't too ripe for dis-couragement;



The outer (red) roof layer isn't all that crazy, although I know more internal support structure between it and the inner (grey) layer will be needed, but the longest members as modeled are about 25ft, which I believe means a splice will be required somewhere. Other than that, a fairly basic peaked/hipped roof, only octagonal in planform. The outer edge of the cover is the same elevation, but the pitches vary per facet, so horizontal eave paneling on the underside of the overhangs will hide the varying height of the roof profile on the wall (and make burning-out wasp nests from the blue patio walkway easier). However the roof over the front entry will be handled, will likely be to cut the red rafters short at the wall plate & scab the vestibule rafters onto them. These exterior roof rafters are supported entirely by the outer layer of wall frame posts/plates, and only indirectly load the interior wall frame members.

Now, for the odd part, the interior (grey) reciprocal frame. These don't actually extend past the outer wall-frame post, and at present are supported entirely by the interior layer. Technically the black posts in the center supporting the ring formed by the eight intersecting timbers don't need to be there, but this roof layer is quite flat (maybe a 10deg pitch) so any flexing or settling over time will be greatly magnified. Also gives some more margin for imperfections in the joinery & greatly reduces stresses in the timbers. Plus, the two pillars actually frame the interior nicely (between them is a large opening in the upper floor that connects to a living room. Also technically, this roof layer isn't carrying much more besides the ceiling, some roof insulation, and its own weight. The plan is wooden slats or paneling will fill in the eight triangular regions, leaving the frames themselves exposed to the interior. I think a fancy light fixture will block the closeout in the reciprocal ring atop the pillars (or I could leave it open to the attic to make tossing stuff up there to be forgotten easier :))

I hadn't planned on the black pillars extending all the way up to the ridge beam, but it looks awfully convenient. They are also largely decorative as modeled, being far fatter than I suspect is really required. As far as the bracing between the roof layers, my first instinct was to insert braces wherever the two rafters cross vertically, but adding a second ring of timbers (perlins, sort of?) halfway down their length & tying to that also seems reasonable & much more predictable.

TCB

Top
#33675 - 05/18/16 10:29 PM Re: Large/Wide Roof Overhangs [Re: TCB]
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
I have a mostly architectural vs. timberframe conundrum, but heck, we're all architects here, right? wink

As I think I've mentioned before, the bane of octagonal plans is roof design (well, one of the banes, anyway). I wanted to avoid domes, cones, and other "circular symmetry" patterns since they draw the eye to the 'nub' at the peak too strongly (makes them great for temples and domed structures, but I'm after neither)

My thought was to place a ridge beam roughly the length of one of the side facets at the centroid, and run facets from it to the side walls. Very convenient in my case, since I have two large structural pillars in the interior that can support it.

Less convenient is that, while the walls of the floorplan are octagonal, the walkways and balcony ringing them that will be covered by the roof overhang are not. Knowing that the edge of the roofline must be the same height all around (not building a Stetson hat, here), this means the horizontal projection of opposite facets --and their slope angle-- will not be the same. This also means the hip rafters (I think this is the right term for the beams forming the facet corners) will not pass directly over the octagon corners --no biggie, since I don't have corner posts and will have to support them through the plates.

My question is what to do about the junction between the roof and the walls. Because the slopes vary, the floor to shingle height at the walls will change as you go around the structure. While architecturally ugly, this can be covered with eaves (and will be, if only to mitigate wasp colonization, unless I can parlay that into an ag-exemption). With a low-rise roof like this, the eaves don't steal too much headroom, either.

I'm more concerned about the structural aspects of a roof assembly squatting on a wall of unequal height; is this largely just a matter of putting in support posts of height necessary to connect the rafters to the top plates (they'd be only a few inches from what I can tell so far), or is the better practice to increase the height of the top plate as needed, and lap-jointing it to the adjacent walls' top plates at their different elevations?

The 'filler' post method seems simpler, but the best solutions often aren't.

Attached image is a plan view of the roof concept, about 12deg slope on the broad faces. Interior rafter rough-scheming for a 'plain' flat interior ceiling also shown. As you can see, the overhang is larger on the SW and SE faces, which have larger southerly patio/walkways than the two faces on either side of the North entrance (8ft vs 6ft, for a 6ft patio vs. 4ft walkways otherwise)



Gonna try shrinking the floorplan a bit further (40ft diameter, seems a bit barn-like even for a three-bed, 2.5 bath), but the roof puzzle should remain (that's where the other bane of octagonal floorplans --room shape/size-- comes into play)

TCB

Top
#33676 - 05/19/16 09:11 AM Re: Large/Wide Roof Overhangs [Re: TCB]
timberwrestler Offline
Member

Registered: 11/07/05
Posts: 269
Loc: Becket, MA
Cool concept. My 2 cents is that the ridge and those commons supporting jack rafters are going to have to be hefty. Unless you can support the rafters somewhere else (with purlins, posts, struts, etc).

On your wall/roof intersections, do you mean that the plate height varies? It can, but I don't think that it needs to.

How about a shot of the reciprocal frame and upper roof?
_________________________
www.uncarvedblockinc.com

Top
#33677 - 05/19/16 09:59 AM Re: Large/Wide Roof Overhangs [Re: timberwrestler]
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
I ditched the recip roof; I concluded that the interior wall/ceiling mating would cover much of it, plus that a 'focus' point would look better offset toward the entrance/foyer (there is a vaulted area here, now). Lastly that it would require really long rafter elements, worse than they are now.

I figured the supporting rafters need to be heavy, I'm considering running the jacks over them to also tie into the next parallel rafter, so the load immediately dissipates a bit wider. Also planning to tie the outer roof rafters to the interiors, but more to support the interior than to carry tension loads between the walls.

If the plate height is constant (ideal for constructing bents, I would think) how would the varying height of the rafters interface?

TCB

Top
#33678 - 05/19/16 10:10 AM Re: Large/Wide Roof Overhangs [Re: TCB]
timberwrestler Offline
Member

Registered: 11/07/05
Posts: 269
Loc: Becket, MA
Roof pitch could change, and/or the HAP (height above the plate). There are times though that the plate heights do need to be different. You could make the timber plates different depths so that the bottoms are level?
_________________________
www.uncarvedblockinc.com

Top
#33679 - 05/19/16 07:39 PM Re: Large/Wide Roof Overhangs [Re: TCB]
TCB Offline
Member

Registered: 02/10/16
Posts: 26
Loc: Texas Hill Country
I think the roof pitch is more or less fixed; the perimeter/overhang and the peak location fully define it, apart from the ridge beam width I suppose (I prefer it to be slightly narrower than the wall faces since the taper is more visually appealing, not sure if that will be noticeable in person, though)

HAP; height above plate. That's a great term that describes my question perfectly; the height of the plate and rafter gap is not consistent wall to wall. On the diagonal walls (NE, NW, SE, SW) it will actually slope at some shallow angle.

When you say the plates can be different depths, are you referring to how deeply the rafters seat down into the plate recesses? I think that could work, and would be much easier to lay out & accomplish before assembly. My only concern would be the 'tall' rafters either won't have as much support, or the 'low' rafters will be beefier than needed. The HAP is 6" higher than the N/S walls at the diagonals, and 3" higher at the E/W walls.

On last idea that occurred to me; though the hip rafters do not cross the corners, they do nearly cross vertical post members (close enough to be adjusted to match them). The Jack rafters could also be placed to align with the vertical posts (2ft centers). Would running the posts up past the common plate elevation to carry the taller rafters directly be advantageous? I plan on using purlins to support metal roofing, so the rafter/post connection wouldn't be an unsupported 'L' sticking up past the last horizontal support. I'd wanted to make all the wall truss elements identical for logisitics reasons, but maybe varying this one feature at the end of otherwise-identical posts isn't so bad?

I'll start up a proper design thread to get all these features in one place for discussion, since I now have a floor plan and rough truss layout.

TCB

Top

Moderator:  Jim Rogers, mdfinc, Paul Freeman 
Newest Members
Maisie_Holmes, Snowhunter, SaintAnselm, bhlaban, pinhead
4744 Registered Users