Rigid Frame 1

[dberry42]

I'm looking at APAW485.pdf (Rigid frame roof structures). The pitch of these is 5/12. Does anyone have a link to these types of structures with a 20/12 pitch? I'm trying to span 24'. Essentially it's an A-frame with 4' kneewalls and I'm trying to find some information on how large the plywood gussets need to be to resist the outward thrust.

Thanks!
Dan

 

[youngstructural]

Good Day;

I'm going to do a bit of looking for you tonight, but if I understand your situation correctly, this is not exactly a solution you want to pursue. Kneewalls are a notoriously ubiquitous design atrocity in most of the Canada and US residential construction. Put simply, they don't work, aren't safe, have serious maintenance issues and are a seismic horror. Your rigid A-frame would have to be very, very rigid to achieve a thrust onto your kneewalls that can be considered negligible.

I'm basing my answer on some broad assumptions: The knee walls @4' only give you half of your desired ceiling height, and you want to try and extend a roof to 8' over the majority of your space. The kneewalls are true knee walls, not cantilevered projection walls (which would solve your problem and make an A-frame a viable solution). I'm also assuming that you want to spec the A-frame roof for the structure in wood.

Like I said, I am going to do more looking, but here is my two cents on the matter: Use a proper scissor truss (see

http://www.cwc.ca/products/trusses/shapes.php

for examples). Have it made by a reputable truss manufacturer, with quality in-house engineering - These guys are experts in creating trusses for all sorts of problems, and you can rely on their expertise (and insurance!) in an area that really does demand experience. Specify that the truss is to have minimal trust load (often termed a "support loading without horizontal thrust"). And you DO have to specify that because even scissor trusses can have kick (it will be a factor of their relative stiffness). Bear in mind that you will still have to do something to your kneewall connections to try and give it some seismic capacity, and your roof will always be able to impart a thrust when subject to an unbalanced snow load, and particularly under storm conditions. Kneewalls can also have problems under severe wind loads if their connections aren't given special consideration.

I'll take a close look at this tonight, since I have a very good truss design text in the home office. Let me know if I've helped, and please confirm or rebuff any (or all if need be) of my assumptions.

Cheers,

YS
B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[dberry42]

A picture is worth...

http://users.adelphia.net/~dberry42/cabin.pdf

that's the cross section of what I'm trying to do. Your assumptions are right on the money. It's 24' x 36' with studs, joists and rafters all on 24" centers.

From bottom to top is workshop, living and then 2 bedrooms in the tiny loft area.

My options are to balloon frame so that the kneewall isn't acting like a hinge, but I really don't want to do that if I can help it.

2nd option is to use 2x8 for vertical kneewall and bolt it to the floor joists and the 2x12 rafters as indicated on the drawing.

3rd is using plywood gussets similar to the APA handbook.

What I'm thinking is that since the roof is so steep and has a 2x10 collar tie the outward thrust on the kneewalls will be managable. I know folks build kneewalls all the time and get away with it out of sheer luck, but I don't want to risk winging it and my best alternative is to over build the heck out of it.

I tried to hire a structural engineer, but after working on it for a while, he wanted $500 and said it needed a lot more work before he could figure it out. This doesn't seem like rocket science, but I realize it's a somewhat unusual design these days of factory built trusses.

I appreciate you looking at this and am willing to pay a reasonable fee for a workable solution but $500 to start struck me as kinda high.

Oh I almost forgot. Since this design has been giving me fits, I made an alternative design with a structural ridge beam supported every 12 feet from foundation to ridge. That's kind of a klutzy solution, but it is an option.

Thanks a bunch!
Dan

 

[EG]

After seeing the picture you attached, why not simply balloon frame the walls and design the floor to resist the outward thrust from the roof? Seems simple. Also, $500 sounds like a bargain...Good luck!

 

[youngstructural]

You've hit the nail on the head with the structural ridge beam. All jabbing about prefab trusses aside, no one will do something atypical for you on the cheap, because there are just too many unknowns. I don't blame your engineer in the least for insisting upon enough of a fee to be able to properly tackle the situation. With this in mind, let's see if we can't simplify the situation...

I absolutely would not rely on a 2x10 collar tie in a situation like this: it would work, but for how long? One of the greatest strengths we have when we work with wood is how well it takes momentary overloads (wind gusts, seismic racking, etc), but the inverse is just as true. You would be flexing this joint continually for a number of years, and I would think it would eventually fail, at a minimum in the deflection serviceability limit state through rotation at the joint. Basically, you'd get a crazy bouncy floor, or perhaps even a dangerous structure.

You're certainly right in saying your design is atypical (always a Bad Thing, capital B, capital T), but you can easily improve it by bringing in as much typical framing as possible. Balloon framing is not the enemy it is made out to be: 18th & 19th century balloon framing is, but that involved notching members to provide bearing area. Don't notch your members and you fix the problem. Instead of notching, bring jack studs up both sides of your continuous member, and support your floor system on the jacks. Make the continuous member a deeper section, and calculate it's bending etc based on pin connections at base and at 2nd storey floor. You'll find that there is still more thrust than you think and that your connection between the column and 2nd storey framing is going to be a critical design concern. And make sure you do use a ridge beam (be careful in your detailing to ensure that it is a true ridge beam, not a ridge board: Load path, load path, load path.... Properly sized column to a properly founded foundation). That will remove a good deal of load from your roof, and will acheive a significant reduction in thrust. Although I'm still betting on fair size thrust under unballanced loading conditions. You're also going to have to be quite careful with your deflection limits, both on the ridge beam and on the vertical columns.

Take a look at getting diaphram action to carry your load from rafters which come down between columns accross to the columns at 4, 6 or perhaps 8 feet. I think you're going to have to keep the columns quite close in order to support the thrust in a reasonable manner, but 8 feet should be acheivable if you're willing to have a small bulkhead around the columns. Keep in mind that corners = cost when drywalling. Careful with diaphrams mixed with balloon framing: You need continuous top plate between area acting as a diaphram! Your engineer should be able to spec this quite quickly if he/she's familiar with wood diaphrams.

Collar ties in the traditional sence are hard to size and can be very problematic in the field. A tensile load transfer through a column to beam joint, with a ridge beam above to reduce overall tensile load as we've discussed may just be a realistic long term solution. Be careful to check your continuous third of your column in shear! Double it up if you need to, and DON'T allow services to pass through if they will be far from the point of inflection on your shear diagram. Wood does not like shear, and you won't like what shear does to wood.

Please don't take offence here, but I would encourage you to have this designed by a competent Structural Eng. in your area. Make the changes I've suggested (to simplify his/her life and save you money) and then get a stamp. You'll also find that all of your hard work could be for nothing if the local building authority refuses your plans without a seal...

EG: I think Dan's issue was that $500 still had not yielded a solution...

Dan: Good luck! Hope I haven't discouraged you too much, but I think you should get a SE to take a look for you. Please do not think that this is sufficient advise to procede on your own. I wish I could help you more, and assure you of what you should do; but this is just impossible without being there and seeing the plans. Also, given the abnormal construction, I would encourage you to find a SE who's willing to come take a look at the finished product. Do not rely on this advice for your design: Take advantage of it to re-work your design and make the SE's life easier. You've got an interesting idea, and I think it'll be a great place.... You do need advice you can rely on in order to acheive it. I really hope I've been helpful, and not too stressful!

Regards,

YS


B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[dberry42]

YS, Thank you for your considerate reply.

In all likelyhood I'll end up building a traditional 2 story with a loft since this idea is just a little too much outside the norm. The traditional way will be easier and cheaper as well as not require any funky analysis.

What I hoped and still think is possible is an SE might see this thread and say "hey! I just did a project just like that. What you need to do is...." Or someone might have had a link to a little known APA or Agriculture extention document that had something close to what I needed.

Thank you again for taking the time to reply
Dan

 

[youngstructural]

One last comment for you: If you extend your roof line to the ground, and give it foundation, that is a very common form of construction in several regions of Europe. At that point you just design your roof "rafters" as axial & bending members.... Works very nicely, but you'd probably have to increase your roof slope more, and give thought to the thrust of any snow build-up on the sides of the structure (I don't know where you are in the States, but I'll not make any assumptions).

Take care,
Good luck,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[dberry42]

A true A-frame just isn't as attractive to my eye as a more traditional looking cabin. I've seen a few around here though. (Colorado in the rocky mountains)

I still have 120 days to finish the design, and I may yet try and hire someone locally if I'm not happy with the regular 2 story.

If I do, I'll do as you and EG suggested. Balloon frame and use a real strucural beam that is properly supported to the foundation to simplify the design. That and the appropriate connections (steel plate and bolts) may be just enough to get this out of the hairbrained into the possible realm.

Thanks again for the tips!
Dan

 

[youngstructural]

Hey there Dan;

Don't want to beat a dead horse here, but one last suggestion: If you are going to go ahead and take the step of using true steel plates (not something I would think necessary, but only an analysis will let you know for sure), you should use true timber rivets, and leave your members exposed. They really look amazing! And they can be very cost effective, because it will allow you to use VERY large peices of lumber (probably best to use GLULAM, but I've even heard of local large size (18"+) dressed lumber working well).

Take a look here:

http://www.cwc.ca/products/connections/heavy_connections/rivets.php

And a bad picture (couldn't find a good one) of what can be acheived can be found here:

http://www.timsys.com/images/rivet_mockup_thumb.jpg

Let me know what you decide to do... I have all the respect in the world for your taking on a challenging project!

Regards,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...