Woodshed plans - trying again

[everythingwood]

My last post was a formatting disaster so I'll try this again.  Thanks again Don P for the instructions.



 

So this is my timber frame woodshed plan.  Not Sketchup but, believe it or not, for me this is excessively  detailed!  Standard would be a freehand sketch on the back of a piece of junk mail.

Anyhow, I would appreciate some feedback.  If you have trouble reading my chicken scratching here are the basics:

• 12 x 36 four-bent timber frame with a conventionally framed lean-to on each side.
• The long sides will be open for drying firewood and lumber.  The gable end and roof will be covered in metal siding/roofing.
• All frame members are green ash.  7x7 posts, 7x9 tie beam, 7x7 plate, 4x6 principal rafters and 3-4" ash pole purlins flattened on the top.
• Building on a slab with 1' tall concrete piers to raise posts above the floor.
• 4x6 joists 30" o.c. in the end bays with a 1" thick plank floor.
• 3x5 braces put in after raising with half lap dovetails.
• Principal rafters 6' o.c. with purlins 2' o.c. 
• Treated posts and dimensional lumber too frame lean-to's.

Here is the side view without the lean-to attached.


  

Any thoughts or comments would be greatly appreciated.  I need to start cutting trees soon!

If the moderators can delete my earlier messed up post that would be appreciated.

[Don P]

I was repeating a rumor, luckily I've not had to rotate a pic. You done good.
If the moderators hold off for a little bit to let you play with the buttons on your old post before merging or deleting it.

As an experiment, go back to your original thread and click on the modify button, you should be able to modify posts you make. When the posting screen comes up delete your old sideways picture, hit the "add a photo to post", when that happens at the top of the screen that opens at the top go to "my gallery" and click on the thumbnail of the correct view of the pic and then "add to post". It should fix the pic orientation in that thread. That saves having to post again in the future if you can make that work. Half the time I'll read a post right after putting it up and something is wrong, I'll modify till I get it right, I think  :D


Not trying to check anything just general thoughts. I usually look at a building from top down, gathering loads and looking at resistance.

A rafter couple, no ridgebeam, look for what is forming the triangle to resist the rafter feet trying to slide outward, hmm, no tie. There is a plate on top of a post, it will have a desire to roll. The plate has a central point load of a common rafter at midspan, size that.

 Carry the load and thrust reactions to the adjoining posts. Assuming it is connected stiffly enough to resist that the outward push is thrust against the extended post above the tie in bending. Rotate that view in your mind 90 degrees, it is basically not much different than a beam overhanging a post with concentrated load at the end. Start there but the true situation is the post has an axial, gravity load of the roof plus the bending load against it of the rafter thrust. A "Beam-column" or in the NDS, go to 3.9 "Combined bending and axial loading" and there is an interaction description and equation.

The thrust then is restrained by the tie beam shear and tension joinery. Is it adequate.

The thrust could also be viewed as being resisted by the platelike semi rigid diaphragm of each side shed.

This is small, just showing how I look at it. My mind is looking at load and resistance as pinned connections, free to rotate as they transmit loads. There is some rigidity that comes into play especially at small scale. 

The tie beams are notched for drop in joists. The conservative way to size them is to look at the imaginary rectangular beam that fits vertically behind ,or between the pockets if you have drop ins from both sides. That really reduces the available section. If space allows I prefer to not notch beams, rather putting joists or purlins on top of beams and blocking in between, again heavy structure vs lighter load resistance. I've packed lumber to the rafters before  :D

Size the joists.

Then look at the wind resistance, the metal ends are probably the stiffest element to resist that, detail to take advantage of. In the long axis it is your let in bracing. If you can stay on temporaries until the wood shrinks then your let in bracing later will be tighter and more capable. Horizontal sheathing across the brace to column area would help to gusset that area and make it more rigid. but that is your weaker side from what I'm seeing.

Just rambling while looking at it.

[everythingwood]

Don, 

From what I have built, read, and seen in plans I'm comfortable with the H-frames for each bent.  I'm sure the sizes and joinery will be more than sufficient.  It is the rafter and purlin roof that raise questions.  I used the beam calculator to calculate the midspan point load of one principal rafter.  I used a load of 2950 lbs (9.86'x 6'x 50 lbs/sq ft) and the numbers for mixed oak #1.  If I'm doing this right I got a deflection of 0.91 inches, which is excessive.  If I bump the plate up to a 9" wide x 8" deep deflection drops to 0.47.

With a 50 lb load I get 900 lbs of horizontal thrust... I don't know what to do with this.  Is that too much for the plate to resist?  I assume if it can handle the vertical load it can resist the thrust.  But is there a danger of rotational failure with the force trying to overturn the plate?  I planned on using simple M&T joints with two pins to attach the plates to the posts.  Where does a person find resources to see if this is an acceptable level of thrust?

[Don P]

Well, I had a long winded really enlightening and wonderful post almost ready, and the power blinked, so you get this one  :D.

There is a running series of articles in the Guild's Red book that touches on this but gives no concrete solutions. It is good to see how different people look at it. Ed Levin wrote several articles "Frame Engineering","Joint Engineering II","Joinery Decisions".
At which point Jack Sobon responded with "Joinery".
You're seeing engineering and empirical design thinking about the problem.

The articles on tension joinery might be helpful as well.

I have seen large barn plates slide, shear 1-1/4" pegs and show a desire to roll. In log construction it is called plate log roll and I've worked on repair of it a couple of times also. These were larger but it is a force to be reckoned with.

At 6' oc spacing x 6' horizontal rafter span x 50psf (did you add dead load?)=1800 lbs vertical for the thrust calc= 900 lbs horizontal

Axial force coming down the rafter at the plate is going to be a² + b² = c²
Square root of (1800² + 900²) = 2012 lbs axial

From Levin the force is acting through the plate at an angle double the pitch, so 24/12.  

And then they go silent. One way I can see is to figure the section properties of the plate rotated to that angle. For this arctan(24/12)=88.4 degrees, most of the force is darn near vertical, I'd say checking the vertical at that section is close enough, as the pitch flattens I'd think more.

Shear at the post tenon is just 900 lbs/ (allowable shear x tenon cross sectional area.)

The plate mortise there is taking the 900 lb load in tension perp to grain, basically frowned on in the NDS so you'r left to your own there. But is the force getting ever closer to vertical here?

More rambling  

[everythingwood]

Thanks again Don.  lots to think about!

I was reviewing the plan and realized I can add a tie at the ends of the plate but I wanted to keep the center of the building free of obstructions at the level of the top plates.  Of course adding ties at the plate end puts three joints at this point... M&T at the post, rafter seat, and M&T for the tie beam.  Of course if I use a jowled post I would have more "room" for the joints.

I have to note that my workshop has the tie beam, post, and plate laid out identical to my woodshed drawing and there has been no problem in20 years.  Of course the roof is conventionally framed but the total load would be the same... just spread out a little more.

Also, I'm sure there is a logical explanation I'm missing but I haven't seen any post-plate M&T joints where the tenon is laid out closer to the inside of the building.  It seems to me that putting it there would better resist that tendency for the plat to "roll" due to the forces from the rafters.  The fact that I haven't seen any tells me it's a bad idea.  That, or I'm that much smarter than countless timber framers throughout history who just never thought of it!  What are the odds  :D   

[Don P]

Thank you, bouncing ideas around makes my mind go different places, I'm not sure if that is a help or hindrance  :D

Canted struts in the ends would work too, maybe up to a collar tie? Just brainstorming. At this scale the siding there will also do that work though.

On moving the post tenon inboard, it would probably make the tension perp worse, although in Levin's article "Joinery decisions" he shows a double tenon on the post end for that reason. Sobon had seen it in old work but dismissed the detail as being used for that purpose, but thought it was to resist shear from rafter thrust.

[everythingwood]

Don,

I'm going to tie the top plates together at each bent after all.  I don't plan on using the upper deck often, and mainly for winter storage of my canoes and other odds and ends.  I definitely don't want collar ties.  If I do I won't have any headroom at all.

Hopefully my chainsaw comes back from the shop soon. We finally got some snow on the ground so time to start skidding trees!  I should have pictures to post come springtime. 

The woodshed is the practice/warm-up for a pavilion, complete with an outdoor fireplace and pizza oven in 2022.  I had a fairly simple 16 x 20 structure in mind. Of course while scouting trees the other day I saw a few crooked ones that would make a beautiful cruck frame so plans may change!  I have always loved the organic look of the cruck frame and, without side walls, I think it would be simpler than a "traditional" timber frame.

Thanks again for your comments and ramblings    

[Don P]

Structurally that works. I'm reminded of one barn I visited that had ties across the loft at an uncomfortable height. At some time they got in someone's way and they were cut out to make it easier to cross the loft. That didn't cause the ultimate collapse but it sure contributed heavily, and it took 2 or 3 generations.

If you do go that route think about putting the tie on top of the plate, the plate and post will not experience any thrust, there is a rigid triangle sitting atop a frame.

This is a good resource;
https://www.ncptt.nps.gov/wp-content/uploads/2004-08.pdf
Take a look at chapter 2 for those types of roofs.

Then go back to Chapter 1 and take a look at fig 15, the dutch barn. If the rafters are continuous and stiff enough, and the ratio of inboard to outboard lengths are about the same, the rafter is balanced on the posts. The rafters are basically propped cantilevers. Under uniform load ... if there is an unbalanced snow load things can go sideways in that thinking, so I can look at it a couple of ways.

[everythingwood]

I considered a Dutch barn but I really like the monitor style roof with the sidelight panels under the eave of the main frame.  And since it's all going to be open on the sides I wanted to go with the conventional treated lumber lean-to's to protect the timber frame from the weather.

But your recommended reading did get me thinking (a dangerous thing).  I already had the publication but  when I looked again at chapter two I noticed Figs. 25-27 and realized that duh, the top plate doesn't have to be flush with the ties!  The most pressing issue in my mind was having all those joints stacked up on the corner... especially since I was planning on using saddle joints for the rafters.  If I raise the plate a few inches I have a lot more room for the joist pocket on the end.  Then I looked more closely at fig. 22 and decided I get even more room to work if I slide the plate about 8 inches outboard from the post!  Now I can go with shorter rafter tails and still get a good overhang.  since the roof is wider I also get slightly more headroom in the loft.

I think this would work for load and stability.
  

[TroyC]

 

 

Saw your drawings and reminded me of my barn. It was neglected for at least 50 years. Left side was rebuilt because a huge pine fell on it during part of the cleanup.

[everythingwood]

Now that's a barn smiley_clapping 

I would build another barn over it just to preserve it!

Really though, I love those old structures... it's a testament to the builders that they lasted so long despite the lack of care.

[everythingwood]

And a large pine fell on it during "cleanup"?  sounds like someone needs a refresher course in tree felling  :D

[TroyC]

Did that 'refresher course' after the felling. Learned what I did wrong (I think). The old barn had two huge pines that had grown next to the roof, one in back, one on the left side. The pines were 60+ yrs old. Actually took one of them to a local sawyer, had it milled to original dimension sizes, used that to repair left side. Other one I milled myself after I bought the mill.

You are right, a testament to the old builders. This old barn sits on basic brick piers. The side supports are pretty well gone, but considering they've been in the ground over a hundred years, it's amazing. I use it mostly to store extra wood out of the rain. The sides are great for implement storage.

[TroyC]

everythingwood- I was just looking at your dimensions on your front view. Think carefully about making the sides only 5' wide. I made my repair about 9' wide, used 12' sheet metal on roof. Sometimes I put tractor under it. Wish it was another foot wider! On a 5' end a blowing rain will certainly wet anything underneath. If I stack wood under the left side against the main building, it mostly stays dry, but I don't think it would with a 5' cover. Anyway, love your plans!

[everythingwood]

The width of the sides is a function of the height of the building.  I wanted to keep the monitor roof style and proportion.  For every foot I increase the width of the lean-to the headroom at the side wall decreases.  However, since I am now adding plate ties I believe I can safely increase the post height to gain more headroom in the loft.  Since the center frame will be taller I'll be able to raise the height (and width) of the lean-to's as well.  I'll be increasing them to 6-feet.  Their sole purpose is for drying  firewood (stacked four rows deep) and four-foot wide stickered lumber piles.  I figure the solid roof and only having one side open to the weather is still a vast improvement compared to the typical steel roofing scraps weighted with cinder blocks.

And keep working on that timber felling... I stay away from the risky stuff but I like to play with aiming trees in the woods.  I have attended trainings through work where the instructor took a 2-foot diameter oak tree with a lean that had to be approaching 10 degrees and felled it against the lean using nothing more than technique (plunge cutting with a "trigger" release) and lots of wedges.  He was a true artist with the chainsaw.

  

[everythingwood]

So Don P... I was hoping you could give your opinion on this.

I have thought about lengthening the ties and installing the plate outboard of the posts (see comment #8 ) and raising the plate above the tie beam.  I'm doing this primarily to make more room for the joints where post, plate, and rafter meet together.  However, by attaching the plate to a tie beam it's no longer held on by a tenon, but by two pegs (one at the corner).  What is your opinion on this arrangement?   

Option number 2 is going without rafter ties and using an appropriately sized ridge beam at the peak then installing a post at each bent from the ridge beam down to the tie beam.  As I understand it a post supporting a ridge beam will carry half of the roof load.  I assume this reduces the thrust and load on the plates by half.  

Unfortunately I see there are calculators for uniformly loaded beams and point loads.  I calculated the roof load at 55 lb/sf (40 snow and 15 for the roof) and ran it through for mixed red oak (per your suggestion).  I calculated a point load of 3960 for a center bent. Using an 8x10 beam it passes for the point load.  I was also going to make a loft on each end of the structure for storage.  How do I figure a combination of point and a uniform load on the beam? 
   

[Don P]

This was on my computer, it is crude and you can look it up for better traditional examples but I was playing with a simple way to do a Norwegian grindverk style frame. They solved the thrust problem by putting the plate inboard.


 

If you put the plate on the ties outboard you could drop it into a notch or cog in the ties but think about the plate overturning especially as it gets deeper than wide.

Correctly you would figure the bending moments and add them and check at the point of max moment. A midspan point load produces twice the bending moment of a uniformly distributed load, sooo, you can take half the uniform load and add it to the point load and then run the center point load calc.

[everythingwood]

Thanks so much Don P.  I consulted the great Gazoogle and found this!

https://www.youtube.com/watch?v=G7wyqv_aPlc

It's a great slow motion virtual raising of a grinverk frame.  I think this may be just what I'm looking for.  Of course since I have ties placed lower to hold the loft floor I can go a little smaller on the plate ties.  You have given me a whole new direction to explore here!

One more thing, I'm no engineer (more like a curious "guy with a hammer") but I love getting into the weeds on a new project.  I was searching for NDS standards and found the AWC website. I'm trying to go cheap here... can I same myself $150 by downloading an older (free) version of the NDS someplace?  Math is math, I'm guessing the content on timbers hasn't changed much and newer versions typically cover new methods and the latest engineered products.    

[Don P]

I think I too qualify as a curious "man with hammer"  :D. I'm just an old carpenter.
You should be able to find a free downloadable older version of the NDS on the publications page of the NDS, my latest hard copy here is an '05. There was a major change in shear there from the previous version but generally it doesn't change much. I'll check out the video, looks cool.

[everythingwood]

I do find it curious that the only thing holding the plate in place are the small wedges that keep it tight to the post and the braces.  My first I thought was "what's to keep the whole roof from lifting off in a wind" but upon  closer inspection it's not just the brace pegs, the braces are also designed to work in tension.  Still, I wonder if there is something they are not showing in the video because that doesn't seem like much.  I would be installing a large bolt through each cross tie for insurance. 

[Don P]

Yes, when you look at traditional work also be scrolling through the NDS and figuring out how to bring it up to current standards. For instance a lag or timber screw in each rafter to plate is going to be better in withdrawal than a peg or nothing. There is much wisdom to be gleaned from looking at what has come before but connections and sizes are often less than what is done nowadays. I think these evolved from knock down boat sheds that could be easily moved and then I've seen it morph into barns and larger structures. That is all my guessing, this is a very old style of construction and my command of Norwegian is non existant but it sure seemed to be an elegant solution to a problem.

[everythingwood]

Don P.  Thanks for all  of your help.  I thought a lot about the grinverk style frame but I really wanted to do me some big mortice and tenons again!  This is may latest plan... hopefully my last as I plan on felling trees this weekend!



 

It's the same basic frame with a few changes:

• increased the height of the posts and upped the size to 8x8
• increased the width of the lean to's
• added ties at the top plate.  These will be dropped in with pinned half dovetails.  I'm thinking a two-inch deep dovetail.
• In order to avoid too much material removal at the corners I will place the principal rafters so they fall outside of, and adjacent to the ties.
• I lengthened the braces on the long axis of the building to 4 feet to avoid stacking all the joints on top of each other as well.

 

I love making sawdust.  Time to get on with it! 8)

[everythingwood]

I got my chainsaw back from the shop, so cutting list in hand, I finally hit the woods and started making timbers!



 

This is the first "volunteer" for my woodshed :D.  It's about 20" on the stump and still over 16" at 20'.  It's going to be 1/2 of a top plate and at least a few girts and braces.  I put three others on the ground but I'm still waiting for my nephew to get my tractor (Allis Chalmers WD) up and running before I can skid them out.  I pan on skidding most of them full length up to the house for processing into timbers and firewood.  I hope to have them all ready for the sawyer before the snow melts.

[Don P]

The dovetail mortise in the plate, post mortise and plate joints are looking pretty busy right there over the posts, or can you make the plate full length?

[everythingwood]

Each plate will be two-piece, with a scarf joint (half bridled with pins) located in the center bent a few feet from a post so they will not coincide with any other joinery except perhaps a brace (which will be half lap dovetails cut in after the frame is raised).  At each post there will be a mortice & tenon and the drop-in half dovetail with a pin.  For the tie beam-to-plate joint picture figure #12 on page 11 of the Sobon timber joinery paper, except that the half dovetail lap will only be 2" deep on the plate and the tie will sit above the plate.  I will pin the joint as well but the dovetail is doing the work here.  I will also extend the plate 6" past the post so I have someplace to make the rafter seat for the last rafter pair.

I did this scarf joint on my workshop but didn't pin the table and it has held fast.  For the woodshed I will add four pins to the table area.