Roundwood pavilion

I'm new to timber framing, but I did build my stick framed tiny house myself, so not new to building. I am planning on building a round wood timber framed pavilion with pine logs that were felled when my property was cleared. It needs to be rectangular with a roof slope ideally at 30° since I will be putting solar panels on the south side (my house was too small for them). This is what I've drawn up so far, and I would love input about anything that might be off, things to keep in mind, things to change, etc. My biggest questions that I know I have are the lengthwise spacing between frames, and tied to that, what to use as purlins for the metal roof - I was thinking dimensional lumber because I probably don't have enough logs to do only logs, and I definitely don't have enough 40'+ ones. But I wondered what sort of lumber to use for purlins since the spans will be so large - 10' or 13'4". I also wondered whether 4 or 5 frames would be more appropriate, whether beams to brace the diagonals in the middle would be necessary, and any tips for lifting and scribing and joining the diagonals, since those would come last and be raised separately in my current setup.
I have attached the drawing I have put together of my leading plan. Thanks in advance for the help!

Nate -

What are your nominal snow loads?  Any special wind loading considerations which attend this site?

Kevin

You're not thinking of 2x6 flat purlins spanning 10', are you? Those would sag dramatically just from the metal roofing load, and collapse with serious snow load. If you use good sized beams spanning across the truss bents at ridge and eaves with intermediate rafters spaced 3 or 4 feet apart, 2x6 or 2x4 purlins would work fine.

Those purlins would be a lot of lumber to buy, and if you have a surplus of roundwood, I would use those as purlins and skip the purchased lumber. This would require quite a bit of flattening time to give a good surface for metal roofing. If you have some 20' logs for purlins, using those staggered to span multiple bents would increase the stability of the structure. You will need diagonal braces in both directions in any case unless the posts are set several feet deep in the ground, which depending on your climate and soil would likely rot in several years without effective treatment and placement. 2' deep post setting will barely keep them upright during construction, never mind live loads. If you are in a hot moist climate with wet soil and little stone, I would try different foundation methods than earthfast posts, assuming you want this structure to last.

Looking at your sketch, your trusses need to have the tie beam directly secured to the top members of the trusses, and the longitudinal beams tied into those - there are several ways that could be done. The diagonals in the trusses would be necessary at the span you show unless the top members are really strong. The truss joints need to be serious load-bearing types; again, there are numerous possibilities.

More detail on the sizes of your available logs would help us give the best advice.

Welcome to the forum!  your sketch looks good to me.

Kevin Olson wrote:Nate -

What are your nominal snow loads?  Any special wind loading considerations which attend this site?

Kevin

Kevin, I'm in Louisiana, on the south side though not deep south. Snow load is almost non existent, though we did get a freak 6" here two years ago. Max wind load is around 130 MPH, but I'm not sure how that translates to lb/sqft.

Glenn, thanks for the great and detailed response, I'll break it down and respond in detail soon.

Nate -

OK, so no snow loads to speak of until the next glaciation, but significant episodic wind loading.

James Mitchell gives scantlings for beams (i.e. rafters and purlins) in a table in Appendix 3 of his "Short Log and Timber Building Book".  I can probably scan that next week on the flatbed scanner at work (my flip phone doesn't take good enough pictures for legibility).  But, Mitchell is primarily concerned with snow loads, not wind loads.  Where I live, snow loads - up to 100 pounds per square foot, nominally, before all of the fudge factors are applied - are the primary consideration, as well.  He also has a good discussion of, among other things, foundations for permafrost locations (clearly not your principal concern either!), and computing roof loads for sod roofs.  His bog/swamp foundation discussion, though short, is worth a look and might be of interest in your neck of the woods.  This is a pretty accessible (non-specialist) reference, despite the fact that most of the book is about building modular post framed log buildings.  Used copies abound from the usual suspects and quite cheaply, since it's gone through multiple printings, so if you decide it's one for your bookshelf it shouldn't be too difficult to get one.

I also have an older edition (1974?) of the American Institute of Timber Construction's "Timber Construction Handbook" which gives some equations for wind load calculations as a function of wind velocity, roof slope, relative wind direction, etc.  It also has 50 and 100 year isotach maps, showing design wind speed for load calculations for the US.  Wind loads can either be positive (a push) or negative (a pull), and could be both a push on one side of a roof or wall combined with a pull on the other.  However, for open structures like your pavilion, the "Timber Construction Handbook" refers to a second publication, which I haven't yet tracked down.  So, I may need to dig a bit and see if I can hunt that one down.  External reference to a separate railroad engineering publication is also made for timber railroad trestles and the like.  But, I should also be able to scan the wind load calculation pages from the book I do have.

If you're in a hurry to get a reliable answer, directly consulting a structural engineer might be a good course of action.  But, my curiosity is now piqued regarding wind loads on open structures, so I'll see what I can find.  This might be generally useful information for others here on this forum, as well, for outdoor classrooms, carports, machinery sheds, hay storage, etc.  Or, even for buildings which will be constructed in stages, e.g. a timber framed structure which will subsequently have cob, adobe or cord wood infill, but will remain partially or fully open while the walls are closed in.

Please note that I am not a licensed professional engineer (PE) and none of the foregoing information should be taken as engineering services.

On the hunt...

Kevin

You don't have snow, but wind is more destructive, and you live in Louisiana, t-storms and hurricanes. Your concern is not the weight bearing down so much as keeping the building in place in high winds. Racking and lifting.
What you have drawn is neither a truss nor a timber frame bent. None of the members are joined. You don't give the sizes of material you have to work with.
The beam spanning the width of the building at each post carries about half the roof/timber weight of each bay from the ridge in the middle. Which means a deep beam.  Your span can be covered by a true king post truss and would likely need those diagonal web members to function. The design is readily available online or in a timber framing book and very easy to understand.
You want those 10ft bays and the extra work involved.
My guess is you want to frame this in pieces, in place, not having to lift a truss you assembled on the ground.
I built a pig pole barn years ago at MSU in Bozeman. We used 2x6 purlins set on edge to carry the steel roofing.

I agree with framing the trusses (as proper trusses) in place rather than trying to raise timber frame bents. I also think 10' spacing is much safer. My approach would be  to start with longitudinal eave beams the full length with all posts diagonally braced, then building the trusses on top of that base with structural tie-down from top members to posts. Maybe it would even work to extend that tie to deadmen buried in the ground, to counter hurricane uplift.

I think the extra weight of roundwood purlins spanning between trusses rather than 2x purlins would be beneficial to high wind stability.

Glenn Herbert wrote:You're not thinking of 2x6 flat purlins spanning 10', are you? Those would sag dramatically just from the metal roofing load, and collapse with serious snow load. If you use good sized beams spanning across the truss bents at ridge and eaves with intermediate rafters spaced 3 or 4 feet apart, 2x6 or 2x4 purlins would work fine.

That is what I thought would be the case. There are some tables online about 2 by rafters, but those would of course be with the long side vertical.

Glenn Herbert wrote:Those purlins would be a lot of lumber to buy, and if you have a surplus of roundwood, I would use those as purlins and skip the purchased lumber. This would require quite a bit of flattening time to give a good surface for metal roofing. If you have some 20' logs for purlins, using those staggered to span multiple bents would increase the stability of the structure.

Whether I have a surplus of roundwood is a factor of how I end up laying things out. If I did things the way I had laid out originally, that would be about 90' per bent plus 120', so 480' or 570'. I have 23 logs totaling 852', so there's definitely some wiggle room. Could I split logs in half with a froe and use both halves with the flat side up as purlins?

Glenn Herbert wrote:You will need diagonal braces in both directions in any case unless the posts are set several feet deep in the ground, which depending on your climate and soil would likely rot in several years without effective treatment and placement. 2' deep post setting will barely keep them upright during construction, never mind live loads. If you are in a hot moist climate with wet soil and little stone, I would try different foundation methods than earthfast posts, assuming you want this structure to last.

I was definitely planning on bracing in both directions. The 2' was just a newbie hazarded guess and could certainly be much longer if it needs to be. I am definitely "in a hot, moist climate with wet soil and little stone." I thought that there might be ways of protecting the in-ground posts sufficiently, but I am open to being wrong there. I would prefer to avoid concrete if possible, but I will if I have to. I know you can scribe to stones, but I wanted something better anchored to deal with the wind loads that can be up to 130 MPH.

Glenn Herbert wrote:Looking at your sketch, your trusses need to have the tie beam directly secured to the top members of the trusses, and the longitudinal beams tied into those - there are several ways that could be done. The diagonals in the trusses would be necessary at the span you show unless the top members are really strong. The truss joints need to be serious load-bearing types; again, there are numerous possibilities.

I am curious about your reasoning on this being necessary, but it certainly would be much easier to construct and raise the whole bent with trusses than to put the trusses on top of the longitudinal beams later. Would I just be able to raise each bent and then use those split-in-half purlin logs to hold everything together in that direction? If I need other longitudinal beams, then I would love to hear ideas on how to involve those without messing up the purlin structure for the roof - that's why I put the longitudinal beams on and then the trusses on top, because then I could put whatever kind of purlins on top of that without having to do any lengthwise flattening and having to have more longitudinal beams than I have.

Glenn Herbert wrote:More detail on the sizes of your available logs would help us give the best advice.

Here's a Google Sheet with the length and two end diameters of each of the 23 logs.

Thank you for such a detailed and thoughtful post!

Edit to include your next post:

Glenn Herbert wrote:I agree with framing the trusses (as proper trusses) in place rather than trying to raise timber frame bents.

Same question to you - just because these would be massive, heavy bents, or some other reason?

Glenn Herbert wrote:I also think 10' spacing is much safer. My approach would be  to start with longitudinal eave beams the full length with all posts diagonally braced, then building the trusses on top of that base with structural tie-down from top members to posts. Maybe it would even work to extend that tie to deadmen buried in the ground, to counter hurricane uplift.

So would you build and raise the two longitudinal sides like bents? I interpreted from your first post that there would not be longitudinal beams besides the purlins, but I guess you just didn't like the longitudinals being between the tie beam and the trusses? I could see how that would be less secure. Could you specify what you mean by structural tie-down? I like the deadmen idea, though I would need to figure out how to make that look decent aesthetically.

Tommy Bolin wrote:What you have drawn is neither a truss nor a timber frame bent. None of the members are joined. You don't give the sizes of material you have to work with.

This is a preliminary drawing, and I definitely plan to join the members. From what I've seen so far, especially this Permies video, It seems like most of the joints should be Lincoln Log style (or maybe the "butter pat" that I've seen elsewhere, but that seems more complicated.)
As for the sizes, you can see the previous post which links to the data on all the logs. Average of the big and small size averaged across the logs is 7".

Tommy Bolin wrote:The beam spanning the width of the building at each post carries about half the roof/timber weight of each bay from the ridge in the middle. Which means a deep beam.  Your span can be covered by a true king post truss and would likely need those diagonal web members to function. The design is readily available online or in a timber framing book and very easy to understand. You want those 10ft bays and the extra work involved.

Noted. I can definitely do 5 bents. By "which means a deep beam", are you saying I need a thick log there?

Tommy Bolin wrote:My guess is you want to frame this in pieces, in place, not having to lift a truss you assembled on the ground.

Is this just because lifting something so heavy would be a massive pain?

Tommy Bolin wrote:I built a pig pole barn years ago at MSU in Bozeman. We used 2x6 purlins set on edge to carry the steel roofing.

I hadn't thought about setting purlins on edge - seems like that might look weird, and it would definitely be much easier to miss them with the screws, but I suppose I could do that if necessary.

I also would love people's thoughts on the process of scribing, making cuts, and assembling, especially if I'm not going to be raising it as bents. Seems like any scribing and cutting significantly in the air would be a huge pain.

No. I looked to me from the way you drew it, that you intended to raise the building one piece at a time. Maybe due to (solo) labor constraints.

Deep beam across yes. Strength of the beam proportional to square of depth as I recall. Little extra goes a long ways.

I have not watched the reference video, but remember, stacking things with 'lincoln log' type joints is fine, but a true truss or bent does not allow for any displacement at any joint, it is intended to be rigid or stabilized in any direction.
Pegs are not load bearing, just keep the joint together, therefore generally neither are nails, except in shear planes.
I used to tell my apprentices without added seismic/hurricane hardware that a wood framed building is held together only by gravity/inertia. Turn it upside down and the whole thing falls apart. A building must be anchored from roof ridge to solid earth. Start your mental modeling of structure from the roof down. Figure out how much weight you intend to carry, how to distribute it, how to keep it in it's place, how to get it to the ground.
Stacking Lincoln Logs makes a structure that will stand but does absolutely nothing for wind or seismic uplift, lateral displacement.  

The Wheaton Labs videos are rough gravity-load timber framing. A true traditional timber frame has mortise and tenon joints, pegged to withstand tension as well as compression loads. The number, size and species of pegs makes a huge difference; mortise and tenon joints do not use tenons like nails, but like through bolts.

The trusses you draw are not "on top of" the tie beams, the tie beams are a critical part of the trusses, and all have to be pegged together so the joints cannot pull apart. If they are joined properly, the tie beam is largely in tension holding the bottom ends of the sloped top members from spreading. If you depend on the tie beam just supporting the gravity load of the upper members and roof, then it must be really strong to resist bending.

Here is the scan of Appendix 3 of James Mitchell's book on Short Log and Timber construction, giving dimensions for various beam loadings, with a brief tutorial on computing simple loads.

Alright, how does this look? I tried to take your advice into consideration. The only things I've seen about roof ridge joinery have been with square timbers meeting at a 90° angle, so I wasn't sure how to set that up. And if there's anything that doesn't actually need to have a mortise and tenon that I've drawn with one, let me know, because that would obviously save some work. I would still love advice on anchoring it and specifics on what tieing it to deadmen would look like. And if the trusses were joined as I showed, would they be able to be placed on the side walls Lincoln Log style, or would those also need proper joinery?

Also, is a ridge beam necessary in this setup? If so, how would it fit in, just a section between each truss with tenons that fit into mortises on the king post just below where the rafters attach, and then corner braces on each of those?

That mostly looks good. You show half-round purlins about 4' apart, but metal roofing generally wants purlins about 2' apart so the metal does not buckle when you step on it for installation. Additionally, 4' spacing would give only half the attachment points to resist wind uplift as 2' spacing would. Half-round purlins at the average size you mentioned would probably still not give adequate strength against bending in either direction, down or up.

In most places, the lincoln log connection of trusses to longitudinal beams would be fine, but for major uplift loads it would not be sufficient; you need something like through bolts tying the trusses to the posts.