Beginner looking for advice on a workshop project

[Lanyon]

Greetings from northwestern BC!

I'm brand new to these forums and I'm looking forward to learning some things about timber framing from you fine folks. I live in an area where it's not likely I'll ever be able to get professional instruction, so I have to make do with books and the internet.

To give you an idea of my experience level, I'm a complete beginner to timber framing. I have some limited experience with stick framing though.

Without further ado, here's my project I'm looking for advice on.

I have a small shed on my property that I'm going to be knocking down soon and I would like to build a 26' x 18' shop in its place. I plan to have an 8' ceiling for the ground floor shop, and I want to put a gambrel roof on it so that I can have an upper floor as well where I might put in an office or craft space for the wife, etc.

I live in a rural location and have plenty of wood on my property (primarily western hemlock, with a few scattered amabilis fir, western red cedar, pine, and spruce). I've been curious to try some basic timber frame construction for a while, so I figured why not go for putting up the shop post-and-beam style and try some timber framing? It'll be a hell of a lot cheaper than the alternative seeing as how I can procure most of my own timber.

My biggest question regards span distances though. Ideally, I'd like to keep the entire floor space open without any support posts, so we're looking at a number of posts put up only on the perimeter of my 26' x 18' floor plan. This, of course, requires me to be able to span 18 feet such that it's strong enough to act as a floor for a second floor.

A quick look at some span tables in a book I have reveals that in traditional framing, to span 18 feet with traditional dimensions of hem-fir I'd be looking at 2x12s.

Does this mean that to span my 18 feet with timber frame beams that I would need beams that are 12inch square? Or does the size of the beams in timber framing impart more strength than traditional lumber dimensions would? For example, if my span tables call for 2x12s at 16 inch spacing, with big timber frame beams can I get away with spanning the 18 feet with 10 inch square beams? Smaller? Also, what kind of spacing between the beams can I get away with?

I'm a firm believer that a picture is worth a thousand words, so rather than try to explain in great deal any more of what I have in mind, I used Sketchup to draw up a model of what I'm thinking about. Please keep in mind that I'm a beginner and there are probably some very glaring, obvious errors and reasons why my structure would fail (like, for example, I'm already fairly certain that 4"x4" braces are insufficient). That's why I'm here to ask for advice.

Thanks in advance for any information and assistance!

Edited: I didn't realize you can't upload outside links to images. I got the images and descriptions uploaded to the Forestry Forum gallery and added the link.

Edit 2: Thanks to Jeff for pointing out that links to the gallery don't work either. I think I have the image posting figured out now.



  


Basic plan
4 "bents", spaced ~ 8' 5" (from centers). The perimeter of the rectangle on the bottom is exactly 26' x 18'. Dimensions of all timbers in the image are as follows: Posts: 10" diameter logs. Horizontal beams spanning posts: 10" x 10", 18' long. Bracing: 4" x 4" timbers.



 

 Main beam tenons
From my understanding of looking up timber framing stuff, it's very important to create a "shelf" on members such that a portion of the full dimension of the timber is supported (i.e. it's not just resting on the tenon). This is a through-tenon in the log and the shelf that supports the full 10" x 10" dimension of the beam is 3" deep from the furthest outside edge of the cylinder.



 

 Ends of the shop don't require 18' span
Since I'm putting up posts in the middle of the ends of the shop, the spanning beams don't have to be 18' long. Here's a lap joint for the two 10" x 10" beams over the middle post. Again, there's a shelf left such that the full 10" x 10" dimension is fully supported. In this case the shelf is 2".



 

 Floor beams
Forgive my lack of proper timber framing terminology. I can't think of what the floor beams should actually be called right now. The floor beam dimensions are 6" x 6". For the demo drawn up here they are spaced 2' OC apart.



 

 Closer view of floor beam joints
Here's a closer view of the floor beams and their joints. I've flipped the one beam upside down to show what it looks like.



 

 Even closer view of floor beam joints.
The outer edge of the dovetail is the same dimension as the full beam, 6" wide. This tapers down to a total of 4" wide at the base of the joint, with a 1" tall shelf added in for further support at a depth of 1" from the edge of the main beam.

[Jeff]

Now all you need to do is put the embed code from the photos in your post. We don't allow links to the gallery either. They must be in your post. The gallery is simply the means to hold photos for posts. A photo in the gallery, must be in the post. It may sound silly, but there is a method to our madness. 

[Lanyon]

Now all you need to do is put the embed code from the photos in your post. We don't allow links to the gallery either. They must be in your post. The gallery is simply the means to hold photos for posts. A photo in the gallery, must be in the post. It may sound silly, but there is a method to our madness. 

Hi Jeff, thanks for the heads up! I just assumed it was working now because when I view my post myself the link to the gallery album works I'll edit my original post now.

[jueston]

someone with a lot of timber framing experiance will be here soon to offer some good advice, but just doing some rough calculations, i don't think a 10X10 will span the 18ft, i think a 10X12 will.

to figure out your load you ussually start at the top with the roof framing and snow load, and then work your way down. figuring out what load you expect on the second floor(if you use it for storage, it can accumulate  a lot of weight) and then you can figure out your posts after you have all that figured out.

[Lanyon]

to figure out your load you ussually start at the top with the roof framing and snow load, and then work your way down. figuring out what load you expect on the second floor(if you use it for storage, it can accumulate  a lot of weight) and then you can figure out your posts after you have all that figured out.

This is one of the main reasons I'm here: to figure out the basics of how to calculate load and such.

[Jim_Rogers]

Have you read this story yet?

https://forestryforum.com/board/index.php/topic,50714.0.html

Jim Rogers

[Lanyon]

Have you read this story yet?

https://forestryforum.com/board/index.php/topic,50714.0.html

Jim Rogers

Hi Jim! Thanks for your reply... I'm really looking forward to getting this all figured out and learning a ton in the process.

I've started looking at the thread you linked, and I've downloaded the fellow's Sketchup file so that I can follow through your posts and learn what you did (and how you did it). I'm not yet entirely sure how to go about calculating loads starting from purlins, as I was planning on doing a gambrel roof which doesn't really have purlins in the same sense as the structure discussed in your linked thread. But I'm sure we'll get to that later.

I do have one question right off the bat though. In the linked thread, dukndog stated that the snow load for the roof is 5 psf, and that the ground snow load is 15 psf.  Where do you get the snow load for the roof based on the ground snow load? Based on the numbers he gave, am I to just assume that the roof snow load is assumed to be one-third of the ground snow load?

I looked up the climatic conditions in the building FAQs on my community homepage and discovered that, apparently, the ground snow load here is 112.7 psf (yikes). SO if I'm starting from the top down, what roof snow load am I looking at?

As ever, thanks for the assist.

[Brian_Weekley]

Here's a roof snow load calculator:

http://www.had2know.com/househome/roof-snow-load-calculator.html

There's a number of posts that recommend using some soffit joints for the joists instead of a dovetail:

https://forestryforum.com/board/index.php/topic,59968.0/all.html

Here's one more post to check out regarding placement of the scarfs:

https://forestryforum.com/board/index.php/topic,45430.msg653619.html#msg653619

[Jim_Rogers]

Brian:
Great link on the roof snow load calculator.

I to just assume that the roof snow load is assumed to be one-third of the ground snow load?

You can't assume anything.
I don't remember how he came up with his roof snow load.

Jim Rogers

[Lanyon]

Ok, so, after the comments so far I've gone back to my model and made a load of changes. First mistake I made was trying to go from the ground up, rather than the other way around. I now have a model that represents the overall shape/plan I would like it to be. What I need now is to determine what sizes of timbers I can get away with. I'm pretty sure I've erred on the side of too large for most of the timbers. I'm hoping we can pare them down to much more conservative sizes (conservative in the sense of saving timber, not load bearing capacity).

Without further ado, here are some pics of the new model. I've also uploaded the model itself.



 


 


  

Jim, I had a look at the link you mentioned and I tried to follow along as best I could, but I have to admit I'm having difficulty figuring it out for a couple reasons. First, I think the SKP file that's linked in the post now has been modified quite a bit since the initial model that your posts were referring to. Second, since my design is for a gambrel roof I'm not sure how to go about attempting the load calculations. I based my design on a gambrel barn model that was available in the Sketchup component library, and there don't seem to be any purlins? Or should there be purlins between the rafters for this sort of design?

I was able to figure out a couple things, though, namely that the roof snow load for the top of the gambrel roof is expected to be 78.66 (round up to 80) psf. Thanks for the link to the calculator!

[Brian_Weekley]

You could probably do without that ridge beam and use common rafters with mortise and tenon joints instead.  That looks a little complicated for a building of that size.  Are you determined to use round posts?  It would probably be much easier to layout the joinery with milled timbers.

[Lanyon]

You could probably do without that ridge beam and use common rafters with mortise and tenon joints instead.  That looks a little complicated for a building of that size.  Are you determined to use round posts?  It would probably be much easier to layout the joinery with milled timbers.

Ok, so I took out the ridge beam... is this what you had in mind then?



 

I have to admit it just looks foreign to me without a ridge beam Even the stick framing I've done on small structures all had ridge beams, so I never really conceived of this structure not having one!

As for round posts, I'm not set on them for any aesthetic reason or anything. I'm just trying to cut down on the milling required, where I can, since I'll be doing it all myself with a Stihl 070 chainsaw and an Alaskan mill.

[BCsaw]

Just thought that I would throw this in. It sounds like your timber framing skills are limited. Consider building a couple sets of timber saw horses. This will give you some practice with cutting and layout! There are lots of designs on here and the net!

Good luck with your project.

[Lanyon]

Just thought that I would throw this in. It sounds like your timber framing skills are limited. Consider building a couple sets of timber saw horses. This will give you some practice with cutting and layout! There are lots of designs on here and the net!

Good luck with your project.

Hey, thanks for your comment My TF skills are definitely limited. I'm actually in the process now of putting up a shelter with some logs I had lying around for a new pig pen. I'm putting it up to get some hands on mortise/tenon experience before I try my workshop. However, I like to plan in advance so that I know what I'm going to have to deal with with regard to size and numbers of timbers required to put up the shop. That way I can cut and mill a few at a time until they're all ready to go!

[Brian_Weekley]

I don't know anything about Gambrel roofs-never built one.  But I was thinking Queen posts on top of the tie beams with common rafters (held together with open mortise and tenon joints).  Here's some examples:





Open mortise and tenon common rafters (not the same pitch, but you'll get the idea):



You're going to cut all the timbers with an Alaskan saw mill?  Wow, I don't envy you.  I know people have done it, but that's an incredible amount of work!  You'd probably do much better if you felled and neatly stacked all your trees, then hire a guy with a portable sawmill to come and mill them for you.

[Dave Shepard]

I would lean more towards Brian's drawing for a gambrel roof.

[Jim_Rogers]

Ok, so, after the comments so far I've gone back to my model and made a load of changes. First mistake I made was trying to go from the ground up, rather than the other way around. I now have a model that represents the overall shape/plan I would like it to be. What I need now is to determine what sizes of timbers I can get away with. I'm pretty sure I've erred on the side of too large for most of the timbers. I'm hoping we can pare them down to much more conservative sizes (conservative in the sense of saving timber, not load bearing capacity).

Without further ado, here are some pics of the new model. I've also uploaded the model itself.



 


 


  

Jim, I had a look at the link you mentioned and I tried to follow along as best I could, but I have to admit I'm having difficulty figuring it out for a couple reasons. First, I think the SKP file that's linked in the post now has been modified quite a bit since the initial model that your posts were referring to. Second, since my design is for a gambrel roof I'm not sure how to go about attempting the load calculations. I based my design on a gambrel barn model that was available in the Sketchup component library, and there don't seem to be any purlins? Or should there be purlins between the rafters for this sort of design?

I was able to figure out a couple things, though, namely that the roof snow load for the top of the gambrel roof is expected to be 78.66 (round up to 80) psf. Thanks for the link to the calculator!

So, I've been busy with everything going on even on Sundays.
I haven't had a chance to study your frame design, but some previous advice is good.

What you have in a gambrel roof design is a "principal purlin" holding up the roof at the location where the upper rafters meet the lower rafters.

To do the math on the design you'd start with the rafters. Size them for the load based on type of wood and spacing.
Then size the principal purlin to support the upper load. The beam going between them, as in Brian drawing is to prevent the principal purlins from being pushed together by the lower rafter's load. It is called a "straining beam".
Next after that you'll need to size your tie beams to hold up all the load coming down the queen's posts from the upper roof. Not an easy job for sure.

Jim Rogers

[Lanyon]

Ok, I'm back! Bear with me gents, as this is going to be a long one....

After spending some time thinking and deliberating, I've decided to abandon the gambrel roof design. I didn't originally choose the gambrel for any aesthetic reasons, I just thought at the time that that roof design gave me the most useable floor space on the second floor. However, after looking over other designs, including the one of DnDs that Jim pointed towards in a link, I found that I can still get the same amount of useable floor space on the second floor simply by raising the posts up to give the second floor some vertical rise before starting a normal gable roof.

So, here are some pictures of the latest edition of my Sketchup model:



 


 


 

Now for some math!

A huge thanks to Jim again, before I get started, on his awesome replies to DnDs post about his project. It's pretty awesome being able to crunch some numbers myself, and I hope I have this figured out properly. Let's take a look.

Roof Load

Looking up my municipal building FAQs, I discovered that the ground snow load is 112.7 psf. That seems like a heck of a lot to me, but it must be based on historical records as a just-in-case measure. There certainly have been some years where we've gotten a dump of a couple feet or more overnight, but that's fairly rare.

Using the calculator generously linked to by Brian_Weekley, I figured my exposure factor as 1, thermal factor as 1.2 (unheated), roof angle of 36.9° (9:12), and a roof type of "C" (cold roof, slippery unobstructed surface). Plugging all this in I get a return of a FLAT roof snow load of 94.67 psf, and a SLOPED roof snow load of 56.97 psf. For my calculations, to err on the side of caution, we'll just use 60 psf. Add in a dead load of 10 psf, and we get a total roof load of 70 psf.

Purlin calculations

For the new gable roof on my model, I used purlins of 4x6 dimension. Let's see if they can handle the load.

The longest spacing between rafters is 8' 6" over the "middle" bay of the structure (the spacing between rafters for the two end bays is 8' 4" each). I'll just use 8' 6" (8.5 ft) for all my calculations as it'll be the most conservative. I measured this spacing on center between the rafters, like so:


 

Based on Jim's post, I next measured out the horizontal run of the purlins. Again, a picture is worth a thousand words:


 
The measured run of 2' 2 3/4" (2.2292 ft) is the same between all purlins, as well as between the purlins and the ridge/edge of the roof.

For the surface area, we have 8.5' x 2.2292' = 18.95 (call it 19) sqft. For the load on the purlins, we now have 19 sqft x 70 psf total load = 1330 lbs on the purlins.

The span of the purlin is 8' 6" (102"), the beam width is 4", and the beam depth is 6". Plugging all this into the uniformly loaded beam calculator, I find that the purlins pass with flying colours using #1 Western Hemlock P&T values. The required section modulus was 16, and we got 24, with a deflection of 0.16".

Rafters

Next up we have the rafters. For the model I used rafters of 6x8 dimension.

As stated previously, the spacing between the rafters is 8' 6" (8.5). I wasn't quite sure how to measure out the rafter run, so I assumed it is similar to the purlins in where you actually only measure the horizontal distance. Here's how I got my measurement of 8' 11' (8.9167):


 

If I measured that right, that gives us a surface area of 8.5' x 8.9167' = 75.792 (call it 76) sqft.

Following Jim's post on DnDs structure, at this point Jim mentioned something about adding additional load to the dead load based on the roof pitch. I did not understand what was going on here, and Jim referred to outside spreadsheets to determine the extra value, so I was stuck here. Given that my roof pitch is very similar to DnDs, and Jim added on an extra 2.5 psf, I went with adding on an extra 3.5 psf, hoping that that'll be more conservative than it should be. This gives me a new load for the rafters of 73.5 psf.

For the load on the rafters, we now have 76 sqft x 73.5 psf total load = 5586 lbs on the rafters.
For the span of the rafter I used the horizontal distance already measured at 8' 11" (107"). (Is this correct? Or should I have used the actual length of the rafter along the slope?)

The rafter width is 6", and the depth is 8". Plugging all this into the uniformly loaded beam calculator, I find that the rafters fail using #1 Western Hemlock P&T values. The required section modulus was 71, and we got 64.

Ok, so what about 6x10? Everything else in the calculator remains the same except now I have to switch to #1 Western hemlock B&S (instead of P&T).

This time around the rafters pass with a required section modulus of 65, and we're getting 100.

Floor Joists

Moving along, now I checked out the floor joists. They weren't in the original photos above, but here's what they look like with some dimension measuring thrown in:


 

The floor joist dimensions are 4x6. Note that in the image the joists in the two end bays appear to be the full depth of the tie beams. This was an error on my part while building the model, which I have rectified. All joists now have the 4x6 dimensions just like the ones in the middle bay of the photo.

So, for the longest span between tie beams I used the 8' 6" (8.5) on center measurement. The on center spacing between the joists is exactly 2' (2.0). Thus, the area is 8.5' x 2' = 17 sqft.

Considering a 60 psf live load and a 10 psf dead load, we end up with a total load of 70 psf. 17 sqft x 70 psf = 1190 lbs load on the joists.

The longest span of the joists is 8' 6" (102") with a width of 4" and a depth of 6". Plugging this into the calculator using #1 Western Hemlock P&T values we find that the joists pass, with a required section modulus of 14.5, and we are getting 24. Also, the deflection is 0.14, so less than 1/4".

Tie Beams

Finally, we're at the point where we can calculate what my initial question was all about... what size tie beams do I need to be able to clear an 18' span with no supporting posts (other than the ends) while also being able to support a second floor?

In my model, the tie beams are 10x10. The longest spacing between tie beams is the now-familiar 8' 6" (8.5). For the length, I'm just going to use 18', even though the real value (I think) would be less due to the amount sitting on the posts. However, using 18' will introduce some conservatism to the calculations.

The area is therefore 18' x 8.5' = 153 sqft. A 60 psf live load plus the 10 psf dead load gives us a total load of 70 psf.

153 sqft x 70 psf total load = 10,710 lbs on the tie beams... yikes!

Span of tie beam is 18' (216"), with a width of 10" and a depth of 10". Plugging the #1 Western Hemlock P&T values into the calculator gives us an immense FAIL. Required section modulus 275, and we are getting 167.

Ok, let's try a 10x12 beam. Still fails with required section modulus of 275 and we're getting 240, but we're on the right track.

Last, let's try a 10x14 beam. Now we have to switch out to #1 Hemlock B&S instead of P&T. This time we PASS with required section modulus of 251, and the 10x14 beams gives us 326 (room to spare). Also, we get a deflection of 0.38, but for this span a 1/360 allowable deflection is 0.6, so we're good there too.

Next Steps

I'm feeling pretty good about this project now. I'm happy with the new layout with the gable roof... seems to me to be much easier and simpler than the gambrel I was going with originally.

At this point my biggest concern is the outwards pressure of the roof/rafters on the wall sides. How to you check to see if the model is acceptable in this respect?

I'm looking forward to any comments on where we're at right now, and on whether or not I cam close to correct with my calculations

Thanks for putting up with the huge post!

EDIT: I forgot to add a link for the new Sketchup file

[Jim_Rogers]

I haven't had a chance to review everything or comment. Sorry busy here today.

Jim Rogers

[Lanyon]

I haven't had a chance to review everything or comment. Sorry busy here today.

Jim Rogers

Jim! You don't have to apologize or explain yourself for not replying immediately!  :D

I forgot to add the new Sketchup file anyway... it's now linked to the new post about changing tack to a gable roof (not the original post).

[Brian_Weekley]

A few thoughts, but I defer to the timber frame experts...

Increasing the distance between the tie beam to the top of the post does give you more headroom upstairs.  However, the longer that distance, the greater the chance of the tie beam joint failing from the outward rafter thrust.

The lower the roof pitch, the more outward thrust where the rafter meets the top of the post.  Increasing the pitch to say 12:12 would also give you more headroom upstairs and also reduce the outward thrust.

You might want to add lower collar ties on the outside rafter sets to help counter-act rafter spreading (at least at those locations).

When you cut the multiple purlin mortises into the principal rafters, they will weaken them.  Make sure the calculations account for the material removed.  I don't think the building is so wide you still couldn't use common rafters instead of the principal rafter-purlin design.

I don't know if it would be better to use a continuous plate on top of the posts?

[witterbound]

I don't think collar ties would help much.
http://www.timberframeengineeringcouncil.org/images/pdf/TFG%20Article%20December%202007.pdf

[Jim_Rogers]

I have read most of this thread but I have not "run the numbers" myself.

The first thing that "jumps out" to me is your choice of grade for your timbers. Do you understand what "grade #1" will mean? This is a very high quality timber. We don't usually ever select this grade of timber for our tests.

If you haven't had someone like a professional forester view your trees you may need to consider doing that. Grade #1 can be a challenge to acquire. If your trees are that good, that's great. But you may need to "brush up" on what grade #1 means. Mostly it means very good if not great timbers. No defects or very small defects if any. By defects I mean very small knots, no slope of grain at all, no splits, no wane, nearly perfect timbers on all sides and both ends.

Next, on your tie beam sizing, if you have that much extra, then consider a timber not as wide. 10" wide is a lot of wood, drop back to an 8 and see if it passes. And if you're milling these yourself you can also consider off sizes, such as 9" or like that.

Now if you do change your tie width to a smaller width, also consider what that will do to your floor joist lengths. It will make them a bit longer. Making a timber deeper is usually the best way to make it stronger.

Also, you may need to read up on the placement of scarfs. We don't normally put a scarf over a post. When timbers are over a post, you need them to be as strong as they can be. If you cut them to create a half lap scarf joint then you're asking this timber to be as strong as it can be and then you're taking away half of it. Not the best solution.

Rafters pushing outward on the post is a real challenge to understand. This force is called "thrust".

It is why we put through tenons on the tie beams through the posts. To tie the two long walls together and to overcome the thrust of the rafters.

This aspect of frame design is what I hire engineers to review and calculate.

Jim Rogers