32x48 timberframe plans.. In need of some review

[Ernest T. Bass]

Might be kind of a pain to cordwood around, but certainly a possibility. (The infill will basically wrap around the outside of the timbers with shorter log ends.) Perhaps each corner could have a pair of diagonal beams running down to the nearest post in either direction. Could look kind of cool, but there would be glass in the way of a couple of them..

[Don_P]

Just like a fenceline, every bay does not have to be braced, every plane needs to be adequately braced.  The rest of the bays in that plane collect and have their lateral load resisted by this rigid panel. If every exterior wall has a braced bay and if the roof is a rigid diaphragm, the forces will be resisted by the stiff bays, as long as they are up to the task.l

Off topic; It's no different for piers for that matter, if folks would brace corners or bays they would have a building that might survive a wind or seismic event.

[glenn kangiser]

Ken Kern published old USDA info that conservatively rated mud such as cob at 100 psi dry for around 14400 lbs per square foot.  The wood is generally good for about a thousand psi, so it should make some pretty substantial bracing.

I recall him stating that a 12" cob wall weighed around 2000 lbs per lineal foot at 8 feet high, leaving 12400 or so for supporting the roof, live load etc.

Not a direct engineering study but some good rule of thumb info anyway.

[Don_P]

Interesting, I've not seen any design values for cob. That would be the compressive strength of a mud wall. The strength you gave for wood was in bending, the strength we were wondering about with a cob wall is it's shear, or racking strength.

I'm not sure what to think of that 100psi number , the allowable compressive stress for a rough, uncoursed, rubble stone masonry wall with type N mortar is 100 psi. For wood it is around 300 psi (in perp. to grain compression, a log wall). A wall with a 14,400 lb load per foot is going to take one healthy footing.


Further aside; I know the design has moved on but I did find a neat program, Beamboy 2.2 is a free download and can handle the combined loading of a 10" offset post sitting on a uniformly loaded floor. It still failed, but hey we have confirmation. With that combination of loads the point of maximum bending moment had drifted a very little right, but still directly under the post, 8 thousandths of an inch

[glenn kangiser]

That was not the actual load - just what Ken Kern said was available per foot for supporting other loads after it's own roughly 2000 lb dead load.

I find that Portland cement added to mud decreases the strength of it.  I does well with sand though.

Cob made of 30% clay and 70% sand including the aggregates in the clay, is really hard - about equivalent to sandstone.  I realize we are dealing with a variable material here also.  Good cob /clay is very strong.  An improperly chosen or silty clay would likely not be that strong.


Ken was also a bit miffed that the USDA/gov sold out to big business and quit serving the people.  He published some of their old earth building information upon which they had done some pretty extensive testing.  Properly made cob is pretty well monolithic if done right and is much stronger than adobe, due to the reinforcment from the straw throughout .. much like fibermesh in concrete.  It has been tested in Canada and survived an earthquake shaker table test - seems it was up to a 7 Richter scale test.

I have had my crane outrigger on this clay - same as used in my cob, and when dry or slightly damp it will support the cranes 60000 foot lb loading on an 8 foot setting - or even a 4 foot setting  - I'm saying loaded to tipping.  This is on a 6" round disk on the bottom of the outrigger, so If I figure it right, that is around 15000 lbs on 1/4 of a foot or 60000 lbs per foot. Nothing that will be accepted for building and not while wet, but just an observation.  The undisturbed clay here will support it pretty well as water will not soak into it.  A squishy part of an inch or so and it is solid.  I admit that this clay is much tougher than the damp pliable clay most places have. 

Some clays were rated at more than the 100 lbs -seems to 300, so the USDA was conservative in Ken's opinion.

These numbers I am referencing will likely not be allowed anywhere in codes.  Just bringing up info that the trades do not allow to be taught any more.

[glenn kangiser]

Thanks for the info on Beamboy 2.2, Don - I'll check it out.

[Don_P]

Did Kern give any other numbers for cob? Do you have titles for any of the old USDA pubs?
Glenn, I ran into it out west years ago, do your concrete guys have to send in samples to the lab from each pour? If so see what it would cost to run some cob samples. I do remember a comment from a fellow in Afghanistan that their mud walls could absorb a tremendous amount of small arms fire.

Default position when building though, if you can't quantify it as a load path find one you can quantify and make sure it is up to snuff. ... build it stout with something you know about  .

[Don_P]

Woohoo, just got the truck up here for the first time in a month. If you don't hear from me it was not as fun going back down  

The range of compressive strengths of common masonry items in the codebook range from about 50psi for things like low strength block up to 720psi for coursed granite and type M or S mortar.

This is some of what beamboy can do, it was interesting to me anyway..
Here is Andrew's drawing of the situation the beam we were wondering about is the 8'4"  one with the gambrel post landing on it... offset 10" from the post below.


The floor has a 40psf uniform load, it's bedrooms I assumed. The post has an 11,788 lb point load coming down. This is what I entered into beamboy;


This is the output;

Maximum bending moment is directly under the post, if you look at the bottom graph the maximum deflection is to the right of the post. I entered the information for the 8x8 beam and the middle graph is giving the maximum bending stress in the beam as 1240 psi. That's what we would check against the allowable "Fb" for the species and grade of timber. Depending on quality, red spruce allowable stress is in the 575-900psi range. So that beam was too small. Not trying to belabor what we've moved beyond but it is a neat tool.

[glenn kangiser]

Did Kern give any other numbers for cob? Do you have titles for any of the old USDA pubs?
Glenn, I ran into it out west years ago, do your concrete guys have to send in samples to the lab from each pour? If so see what it would cost to run some cob samples. I do remember a comment from a fellow in Afghanistan that their mud walls could absorb a tremendous amount of small arms fire.

Default position when building though, if you can't quantify it as a load path find one you can quantify and make sure it is up to snuff. ... build it stout with something you know about  .

I mentioned to Whitlock the other day that bullets would not go through the earthen walls.  Just joking around but pretty well true.  I tried a hammer and chisel for about 20 minutes to chop through a cob wall Sassy built for a stove pipe out of the underground complex.  A small hole was all I had before I went to get the rotohammer and finish it in another 10 or 15 minutes of chiseling.

Got my "The Owner Built Home" book out to look up the info for you, Don.

Going from memory about 7 years ago I was off on details a little but not too bad.

What I was referring to apparently came from Chapter 17, Adobe Block starting at page 141.  I wish I could put this material online - it is so good, but a few excerpts for education is fine.  

"One effective lobby by the American Lumberman's Association to the US Department of Commerce succeeded in a complete shelving of the earth wall program."

"With unquestioning loyalty and unbelievable ignorance, building inspectors say "No " to earth wall construction in this country, yet, according to structural tests made at the School of Engineering, Christchurch, New Zealand, a foot length of soil cement wall, 8 inches thick will carry over 21 tons at failure.  The weight of each lineal foot of wall 8 feet high is approximately one fourth ton.  That leaves 20 3/4 tons for roof weight and safety. The Australian Commonwealth Experimental Building Station found that the compression strength of an adobe block is in excess of 25 tons to the square foot.  Our own Farm Security Administration claims 33 tons compression strength for these blocks.  This is actually about  ten times the strength needed for conventional roofing weight."

He goes on to explain about bond beams etc.

Cob is monolithic -straw reinforced and much stronger than adobe.

He mentioned the report in the 1940  Bureau of Standards as pictured above.

A chart from Bureau of Standards - 1940 - not too good - sorry bout that.



The above Kern book is out of print but often available used.  

Interesting note... our John Raabe - took the picture that is on the cover of the Scribner's version of  "The Owner Built Home" and that book is the reason I am here.  My son loaned me one of the original hand written copies and I wanted to find out more.

[glenn kangiser]

Interesting rammed earth article 1946 for educational purposes

[Beavers]

  article, thanks Glenn!

I'm really finding the whole cob thing really interesting...really getting the itch to try building something with it. 

[glenn kangiser]

My pleasure, Beavers.  The hardest part is just getting started and doing it.  Make mud - have fun - talk the gals into a bit of mud wrestling. 

I use my Bobcat to mix the cob, but the traditional method is by your feet. http://www.weblife.org/cob/index.html

Don, I forgot - yes - we do testing here but Ken Kern gives full info on testing strengths of adobe - cob etc. in the same chapter.  He details it out using a lever and tells how to do the math to calc out the strength.  It would be a lot cheaper way to test various mixes.

[glenn kangiser]

Note - cob, rammed earth, adobe, compressed earth block, are all various methods of earth building that differ in methods but use soil, aggregate, stabilizers and straw or other reinforcement

[Ernest T. Bass]

Very cool stuff! I feel like the cob would probably provide plenty of bracing to the timber frame, but the cordwood aspect might ruin the monolithicness of it, plus we will be using chopped straw to make it more pliable. While I still think it'd be strong enough, a couple of corner braces would give complete peace of mind...

I'll post some pics of my second attempt at the frame after I get some sleep.. That beamboy program looks cool, but probably not Mac-compatible..

[Don_P]

I was wondering the same thing... whether the cordwood could somehow roll if the wall were racked. My gut says that contained withing the frame it would take quite a force but what that is I don't know. Working some other bracing in there couldn't hurt.

I did find one of the old publications googling, haven't read it yet but it sounds like others want to read more too so this is it's link;
http://www.aaronhauser.com/rammed-earth-books/farmers-bulletin-no-1500-rammed-earth-walls-for-buildings/

[Ernest T. Bass]

8x8, 8x12, joists are 4x8, 3' oc.  

A frame doesn't get much more basic..   I have the tension braces drawn as though they planed into the corner posts at the same height, but they would be staggered at different heights to retain the post's strength.

[Don_P]

Braces are normally considered to be working in compression, that's why they are in pairs, the one in tension is really riding and the one in compression is working. Wood shines in compression, steel in tension.
Your long braces reminded me of a German frame. Check these out;
http://trailridgetimberframestravel.com/germany1.htm

Call me boring, I like looking at the roof delivering load straight down the posts, very strong, no thrust, looks good.

I'd been thinking along another path, trying to clearspan the upper floor.
This is without the upper posts and purlins. The diagonal is just to make the software work, no load. The black vertical lines are loads not members. The loads are not in relation to your building this is all just a thinking exercise. There is the tie across the curb roof at the pitch break, and the upstairs walls just happen to be sloped steeply. Sloping the walls (the steep pitches of the roof) means there is a horizontal as well as a vertical reaction. This is thrusting so the upper floor is also a tie. Essentially you have drawn a pair of structural ridges and the other way is tied rafters. Just another thought.


The floor system is all in one flat plane, that might be necessary. If not, stacking the 8x8's on the 8x12's then the joists on top of that avoids alot of joinery or hangers, and is stronger.

[Ernest T. Bass]

I like the idea of stacking the layers, but that creates a lot of time-consuming voids to fill, above interior walls and such. Also, might the timbers have a tendency to twist? I would like to make a simple version of this thing, (http://www.timbertools.com/Products/LignaTool.html), if possible for quickly hanging the joists.. The trick would be finding a large enough dovetail bit.

For some reason the timber framers call those corner braces "tension braces", so I was just throwing their lingo around...

So, if the 2nd floor system acts as a tie, the upper floor posts and beams might not even be necessary? It might just make the rafter layout easier to have them there anyway, but could we use smaller timbers, rather than hoisting an 8''x12''x12' 20' up in the air without a crane?

[glenn kangiser]

I was wondering the same thing... whether the cordwood could somehow roll if the wall were racked. My gut says that contained withing the frame it would take quite a force but what that is I don't know. Working some other bracing in there couldn't hurt.

I did find one of the old publications googling, haven't read it yet but it sounds like others want to read more too so this is it's link;
http://www.aaronhauser.com/rammed-earth-books/farmers-bulletin-no-1500-rammed-earth-walls-for-buildings/

Thanks Don, That's better than I did.

Note that I did a rammed earth wall on my RV garage lower level.

http://countryplans.com/smf/index.php?topic=1166.0

[glenn kangiser]

I found the zipped links to 3 books on the previous page - download them if you like.

http://www.aaronhauser.com/rammed-earth-books

[Don_P]

Thanks for posting those Glenn, I'll check them out. We own the copyright to gov't pub's, copy and post 'em somewhere to keep them alive  . I've seen high compression numbers several other places.

Andrew, that's early concept, we can run the loads and see how it looks. What is your design snow load?

I've done drop in dovetails, never again. As the width of the joist shrinks narrower the tail gets narrower in the beam's mortise. The beam is also getting narrower and wanting to pull away from the joist end. It doesn't fail but the tail can withdraw enough to get your pinkie in the gap  If you want to go that route the beam needs to be sized with the mortise depth removed. The stacking method requires lots of blocking, the twisting potential is about the same as with a shrinking dovetail. If there is slop twisted grain can unwind the timber as it dries. A tusk tennon is structurally probably the best joist/beam joint but would take a set of hands on every joist as the beams are brought in to slip each one into a pocket in the center, neutral axis, of beam height. It's mortise doesn't break the top or bottom "strap" of the beam but is in the center, low stress, area. If you can get a copy of Cecil Hewitt's "Historic English Carpentry" on interlibrary loan there is a good discussion and illustrations of this.

[Ernest T. Bass]

Andrew, that's early concept, we can run the loads and see how it looks. What is your design snow load?

Code says 70 psf. In reality, we get about half of what most of the U.P. gets, being just a couple miles from the lake..

I've done drop in dovetails, never again. As the width of the joist shrinks narrower the tail gets narrower in the beam's mortise. The beam is also getting narrower and wanting to pull away from the joist end. It doesn't fail but the tail can withdraw enough to get your pinkie in the gap 

I was wondering if the tapered shape of the tenon that the router jig makes would help prevent that problem... As the joist narrows, wouldn't it simply follow gravity down into the mortise, maintaining a tighter fit? Tightly wedging the tenons from the top is supposed to prevent pullout as the timber drys, but that wouldn't be possible with the short 1.5'' tenon made by the router jig.

If you want to go that route the beam needs to be sized with the mortise depth removed.

If the joinery is fairly tight, and the mortises are on the upper (compression) section of the girder, wouldn't the beam retain most of its strength?

The stacking method requires lots of blocking, the twisting potential is about the same as with a shrinking dovetail.

Yet another possibility is metal connectors, as we used on our house, but I feel that good wood joinery would be stronger, cheaper and more aesthetically pleasing...

[glenn kangiser]

I agree, Don.  It bugs me when I can't find the books and studies we paid for, and it bugs me more when it is hidden or done away with because of industry lobbying.

[Don_P]

If we are the owners of those documents, did they hide them or did we lose our homework?
But yes I agree 

I was wondering if the tapered shape of the tenon that the router jig makes would help prevent that problem... As the joist narrows, wouldn't it simply follow gravity down into the mortise, maintaining a tighter fit? Tightly wedging the tenons from the top is supposed to prevent pullout as the timber drys, but that wouldn't be possible with the short 1.5'' tenon made by the router jig.

The ones that withdrew were that shape, same house as the brace above, cut by a Hundegger, my understanding is that cutting station on the machine is like a universal router or end mill, doing the same thing as the template on your jig. Their template is computer code controlling the router but the result is the same. Looking at it that way your jig is a bargain  .  The floor deck holds the joists up on the beam too so the joists don't really sink deeper is what I figured. The wedging would work if you can both hold the beams from drifting as it seasons AND have access to keep the wedges snug until they quit driving. The crush to fit joinery options don't work.

If the joinery is fairly tight, and the mortises are on the upper (compression) section of the girder, wouldn't the beam retain most of its strength?

No, you will hear that over and over, it's wrong.
The bottom fiber is in overload long before the top "seats" and takes its share. The joist will shrink in width.

I had used the section modulus of the remaining cross section after the mortise had been made, an inverted T shape. Much more conservative than you. Dr Schmidt, a university PE, presented at that same TF Guild conference in Roanoke I spoke of earlier.  He did destructive testing with full sized beams and found the most accurate way to look at it was to take the rectangular section between mortises and that is the actual beam. For example with any type of drop in joists or purlins where the mortise cuts through the top edge, the remaining width on the top projected down the depth of the beam is the beam to figure on. An 8x8" timber with 4x6 joists mortised in 2" from each side is effectively a 4x8 in the way it breaks. That removed a good bit more strength than what I had considered safe, I wasn't alone in the room  .

Exception from Dr Schmidt;
Unless using a mortise that removes wood only in the neutral section. A soffit tennon fills that bill but the tenon can be reinforced by sloping the upper face, making a tusk tenon. The mortise slopes up to the top corner, it doesn't take a whole lot of fiber there in the beam to support the compression. Assembly takes a pair of hands on every joist as the beam comes in though. Doing that you can take full design value of the remaining section, the 8x8 less mortises.

Timber joinery is many things but even the old timers kept their options open. Old churches often have iron in high tension areas. The anti steel thing lives largely between some peoples ears, it's an option and shouldn't be dismissed automatically. Nothing wrong with trying to stay all wood first.

The purlin plate;
It looks like the upper roof is 18' wide X 12' bay spacingx 80 (70 live+10dead) psf=17,280 lbs
Half of that is resting on each beam, I'm coming up with an 8x12.
That was calling it a structural ridge.

If the load is travelling down the steep rafters though that purlin plate, it could be anything, or non existant, as long as the steep roof to upper roof connection is adequate. I think if the steep rafters had a birdsmouth under the bottom and up the outer face of the beam they would each act as a post, at that point if you jerked the vertical posts out, where would it go? The posts do form a triangle...bracing, sometimes you see purlin plate posts canted off plumb out to the roof plane as well.

This is with vertical posts, the black dots are "nodes", pivot points and also where I've hung the roof loads. Add a zero to all numbers and this should be close to correct forces for rafters at 2' spacing if it were bearing just on the outer walls, still thinking... It looks like you need to have about 1700 lbs worth of "strapping" across the floor in each 2' strip, fasten the flooring down good, stagger the joints.

Those are not our true support conditions  

[Don_P]

OK, Viewing it as a clearspan truss with no posts, if the rafters are on 2' centers I looks good with 2x10's in #2 SPF or equivalent. There is about 2560 lbs vertical load under each rafter foot.

I've been playing with a program that let's me enter more support conditions. If the posts and beams take the load there is a 1440lb load from each rafter set on the purlin plate beam, carry that down the beam and that load theoretically travels down the internal posts. The tension across the floor drops way down. I don't think that will uniformly happen. At some point in the purlin plate's span and depending on its stiffness it will deflect enough that the rafters will deliver some portion of that load down them to the outer beam. So I would design the outer perimeter as if the posts and purlin plate were not load bearing. This is conservative but with that logic the outer 8x12's are overspanned for that load, and you've mentioned a possible later shed roof. I'd try to fit midspan posts and cut those spans in half.