Hip Roofs & Large Dormers 2

[jerseyshore]

Normally when I design roof framing for a house it's all structural ridge/hip/valleys. If it was simple, they wouldn't be hiring an engineer (generally).

Example below of this house with a tradition 6/12 sloped main roof, 46 ft deep front to back, and a matching 6/12 46 ft wide hip roof off the back, flanked by a couple of dormers on either side. Attic floor joists run front to back as well.

Around here most architects do their own structural work and majority of the time it's typical gable-type framing with ridge & valley nailers, not beams, for a roof just like in this example. They wouldn't even think twice to do it that way, even if it doesn't work on paper. Then they try to make me feel like the crazy one for suggesting structural ridges.

So what rules of thumb do you guys use for determining when hips or dormers become too big to work without structural beams?

 

[jayrod12]

Where I am, pre manufactured trusses dominate the market, I'd never stick frame that roof

 

[XR250]

Usually over 10 ft wide and I start adding ridges etc. Each is case dependant though.

 

[jerseyshore]

Do you usually use trusses when the attic is finished as a half-story?

Unless it's bulk framed apartment buildings or a new development, trusses are rare around here for single family homes.

 

[XR250]

I try to "stick" with stick framing - unless the builder wants trusses - which is rare.

 

[kissymoose]

I use rafter-tie action only where I have a simple gable frame and dormers/perpendicular roof sections can be overframed. I'm not even a fan of rafter tie action for hips where you turn the ceiling joist direction for the last X feet (though I don't often see structural roof problems even in those stupid 100 year old 2x4 hip and valley roofs with no bracing). If I'm designing new, we're doing it right and doing it simple in a way that doesn't make my brain hurt. Anything more complicated gets structural hips, valleys, and ridges with dimensioned bracing locations drawn on the plan. And a lot of houses these days have varying ceiling heights or vaults in the master bedroom which instantly invalidates using rafter-tie action as the ceiling joists don't lap and thus can't be the "tie."

With the house you posted, it looks easy enough to do with rafter-tie action for the main roof portion and overframing for the front facing gables, assuming the main front facing gable with windows is just for street aesthetic and they don't to see those windows from the inside, otherwise that'd require structural ridge and valleys (big ones) with bracing.

 

[TheDW]

Around here custom houses are mostly stick framed, tract builders all use trusses.

For me, valleys are always beams, hips/ridges are nailers unless the ceiling is funky/nonexistent

 

[phamENG]

For me, it depends on the use of the space. If it's a decorative dormer, I'll try to stick with over-framing it, and just increase the rafter size if necessary. If the space is usable, beams. I had one recently that was just a mess - only way to make it work with anything but steel beams was to have a bunch of point loads, but none of them aligned with bearing walls (because none of the second floor walls aligned with any of the first floor walls). So I pushed the architect to reduce the floor area in the 'attic' and went with attic trusses.

 

[SE2607]

When it doesn't meet the conventional framing "rules" in the IBC.

No, using 2x ridges and hips won't calc, but the history of lack of failure keeps this stuff in the code.

 

[jayrod12]

Yes to the question about half storey attic space being trusses still. Or at least the bulk of the roof is still an engineered product. They can make you attic style trusses no problem. On long span situations, often they build parallel chord trusses into the bottom chord space. I'm not sure if I've got one handy, but I know that's usually how it's done around here. In fact, if we were to try and propose stick framing a roof around here, we'd get laughed out of town.

 

[jerseyshore]

That half-story/ attic is livable space so not just decorative windows.

Like I said above, most house roofs I do involve structural ridges and hips, but I know a lot of architects who would just design this house roof A-frame style regardless of the large hips and gables. They'll ignore thrust and ceiling/ floor joist direction and it'll probably hold up for a long time.

 

[jayrod12]

They'll ignore thrust and ceiling/ floor joist direction and it'll probably hold up for a long time.

Yeah, until unfortunately it doesn't anymore. We've had to break that news to homeowners recently. But I digress.

I almost want to stick frame a roof, just to see what it's like again.

 

[jerseyshore]

No you don't, you want to write "Roof trusses by others" forever. That's the dream.

For this house they want to use floor trusses instead of TJI's for some reason. So maybe I'll convince them to use roof trusses too.

One step closer to my drawings just saying "Framing by others".

 

[phamENG]

Simple hips are fine as long as you get longitudinal rafter ties in. The hip is essentially a sloping ridge board at that point. The thrust is transferred across the diaphragm and the rafter ties are often 2x4's on the flat across the ceiling joists and run the length of the house at 2-4ft on center. I've seen lots of them around here built in the 70s and earlier.

I'm okay with it as long as a) rafters are single span with no purlins, knee walls, or other breaks and b) roof geometry is regular. In other words, I never get to design them.

 

[jayrod12]

roof geometry is regular

I haven't seen a regular roof in too darn long. I almost got a square box project once recently. But then there was magically an influx of available funding and the architect went all architecty on it and it was no longer regular.

 

[jerseyshore]

I don't think I've ever seen the 2x's on the flat along the ceiling joists that you are describing. Have any photos?

I don't mind simple hips either, usually when I see issues it's rafters separating from the hip, just like the do at the ridges, and most common on low-sloped roofs.

But it's things like this house (46' wide gable and matching 46' wide hip) and other more "unbalanced" roof layouts where I say we need structural members and the architects and builders go but why we've built X number of houses and they're all fine.

 

[phamENG]

I'll have to look and see. Having stopped doing assessments they'd be from a while ago.

but why we've built X number of houses and they're all fine.

My trick for this is to look up design level storms/events. What is the load case that's the problem? Snow? Wind? I flip the script on them and point out that, in our area, a design level wind event hasn't happened since 1953, and even then it covered about a 5 square mile area. And it's worth noting there are 0 houses in that area that predate 1954. So you and your daddy and maybe even your grandaddy have been doing it this way for 65 years. Awesome. None of those buildings have ever really been tested.

Most roll their eyes at me, but I've had a few contractors come up short after finding that out.

 

[jerseyshore]

Sandy was only 10 years ago and the impacts from that are still very fresh in everyone's' minds down here so I do reference that storm a ton. It does help because everyone around here either got water/ house got destroyed and/or know 10 people who did have that happen to them. It works.

 

[kissymoose]

I'm okay with it as long as a) rafters are single span with no purlins, knee walls, or other breaks

Out of curiosity, why does an internal support negate the use of ties? As long as it accounts for the total amount of thrust, I don't see why it wouldn't work.

 

[jerseyshore]

TheDW mentioned this above and I see architects do this alot too, but do other people use nailers for ridges and hips, but beams for valleys? I've never mixed and matched before and I always wonder how the ridge is supposed to hold up a valley beam if it's just a 2x nailer and not a beam itself.