Timber frame porch roof

[Tom S]

Good morning all. 

It's my first post here so be gentle. lol

I would like to build a roof over my ground level deck. The roof would be 32' wide with an 4/12 pitch and extend 16-(18' with 2' overhang) from the house (attached to the house on one end). I'd like to construct timber frame king post trusses to support the roof and if possible have no center post on the 32' span. I have looked at the calculators on the site here but would love a second opinion. I'm thinking of going with 8x8 posts on helical piles and 8x for the truss. Snow load in my area is 40lb/f2. What size Timbers would I need for the top and bottom chords? Also what size beam for the 16' span for the trusses to rest on?

Thanks for the help. I'm looking forward to getting started and learning as I go but am tripping over this in order to submit for a building permit. 

[Don P]

Hi Tom, Welcome

We've been talking about this on another thread, start there if you haven't already.

This is a screenshot of one truss calc, each grid square is 2', add 2 zero's to all loads and forces.
I've set it up crudely at 4/12 and hung 3,000 and 3,500 lbs at the respective panel points trying to round relatively close ... crude.



 

Things are worst where the tension and compression meet in the heels. It looks like that joint needs to be good for around 15,000 pounds.

Those are axial loads, the forces running along the axis, down the length, of the timbers. Blue is compression, columns. Red is tension, bowstrings.

The loads on the roof are creating those internal axial forces within the timbers. We know enough to begin design of the joints but there is a difference in that truss diagram and the real world. I'm going to probably apply the load along the top chord rather than to just the panel points. (if you can restrict the loads to only the panel points, I salute you). For the rest of us, we are applying a bending load to the timber as well as that internal axial load created by the rigid connection of columns and bowstrings.

Whew, I'm heading for turn 4.
The top chord is in "combined load". It has an axial force and a bending force on it.

Stand up a yardstick. Push down on it, that is compression. Note the axial COMPRESSION force required to make it buckle to the side. Release. Restrain top and bottom but no compression. Push on the yardsick from the side at the 18" and note the BENDING force required to displace it to its serviceable point.

We can have half that amount of axial combined with half the allowable in bending, or 1/4:3/4, etc. We cannot ask the members to do double duty, the combined effect must be less than 1 (allowable strength).

It is hard to motivate looking at freezing rain  :D

[Don P]

Here's my kingpost calc, it is my feeble attempt at stuffing those equations into code;
Untitled (forestryforum.com)

If there are 3 trusses, the one in the middle is the most loaded. 32' wide x a swath of roof 4' fore and aft of it, 8' deep .. 256 square feet x (40psfLL + 10psfDL) = 12800 lbs total on the truss. I think the rest of the inputs are self explanatory. Size of the members depends on species and grade. Are you capable of sawing a 32' bottom chord?

That calc is better than the rough graph one. The bottom chord is a bowstring with 15,000 lbs of tension on it. Bury those heels tight.

This is one joint, for one think about what happens as things dry. But the next structural checks would be sizing the notch to avoid localized crushing of the joint. And shear in the relish beyond the heel notch, the area in red in the pic sliding off and taking the top chord with it.


 
We need to resist 15,000 lbs of shear in that zone. The allowable depends on species, for example I'll use 150 psi allowed. 15,000/150=100 square inches ow shear resisting area minimum req'd. 100/8" wide= I need at least 12.5" of relish beyond the joint, in the direction that chunk of firewood wants to bust. Just sayin.

I would make a steel bottom chord on top of or buried within the timber one.


 

That would have been a whole lot easier/ stronger if I had made a solid strap from side to side and welded it together. I've also used rod and turnbuckle which is great for drawing things tight. A truss isn't a dead hunk of chunks in bending, it is tuned up to pluck a note. You are building an instrument of construction  ;D.

Checks in that pic would be crushing of the heel at the steel, provide sufficient area. And group action of fasteners from the NDS for those rows of 1/2" lags. That was kind of a stream of consciousness wander through some of the options and checks.

I'm happy, as you can tell, to help think through this. Whether that passes down at the building dept is up to them, we're coloring outside the lines. hmm, outside their lines anyway.


edit;
If you still have the truss calc open, notice the bottom chord tension at 4/12. Now change the roof pitch to 12/12, click enter and notice the tension... about 1/3 of a 4/12. And the top chord probably just failed.
The polite term is, its an iterative process trying out the possibilities.

Reboot, I'm in the weeds  :D

[Tom S]

Thanks for the reply. I redid the calculations with a bit more understanding. 

The reason for the 4:12 roof pitch is to not have this roof peak above the current roof peak. The two peaks will meet at 90 degrees. Current roof pitch is also 4:12. I want the eves of both roofs to line up. 

If possible I'd like to get away with a 2 piece bottom chord that is joined at the king post and pegged as per the photo below. This would surely bring down the cost of the Timbers. 

The structure will rest on the exterior wall of my house at one (gable) end. So having a center truss and an outer truss make things simpler. I was hoping to not have a post mid span (on the sides) but the calculators didn't like that. If there is a post under each truss then there would be no weight for a beam along the eves to support. 

What is the difference in forces between the two truss designs in the sketch below? Also on the bottom truss is there a significant difference between locating the post under the bottom chord (left side) or under the top chord (right side)?  

I will most likely add 2"x6" purlins and 1/2" plywood for roof sheathing. 

I appreciate the feedback. 

[Don P]

When you make truss connections try to do it so that everything in the joint revolves around one point. You are not trying to create a moment resisting joint. Usually if you do it tends to split things or create bending moments elsewhere.

In the bottom sketch, 
First those 2 pegs with maybe at best 1000 lbs capacity each aren't going to restrain the 15,000 lbs of tension in the joint. There needs to be a steel rod or strap from side to side. On the left the lines of force do converge in the corner. On the right they do not and as you can imagine the higher the tie is raised the more bending is added to the rafter. Those designs are popular because of the tail, I see them all over the net. Think through what you are looking at. Can it actually handle the forces at work.

The upper design, you are using the heel notch to resist the 15 kips, check the shear area and crushing as mentioned above. Because the angle of compression in the joint is neither parallel or perpendicular to the grain, which we have published allowable compression for, you need to use the "Hankinson formula" to figure the allowable crushing force for the angle in the joint... you're interpolating between parallel and perp allowable strength depending on the angle. Abandon hope of getting that notch and a tenon with pegs all happening there as drawn. You won't get good bearings by the time you make it pretty. All fine and good, you need steel in that bottom tie to make me happy, too much force to go without a net.

In any of these build in some camber. As things settle into contact, sand shrink in service, this is the hot roof zone, the kingpost top chord pinch joint gets considerably tighter. The heels push out tight. if it starts out level it will "hog", sag. That looks sad. A positive camber appears stronger to the eye even though functionally they are equal. In the 24' trusses in my shop on 10' centers, there are 7 pegs each side on the splines IIRC. Again, steel, you would have at least 15 and the group effect would kill it.

What species are you using? 

[Tom S]

I haven't contacted any mills yet. I assume it will be eastern white pine available locally (Eastern Ontario, Canada). 

My thoughts about the split bottom chord being cheaper didn't take into account the sheer force of that piece. What you said makes sense. It would be easier with a one piece bottom chord. I'm guessing I would still need more than the notch and two pegs to resist the force at the ends. I'm trying to visualize the photo you posted with the steel. Does it go between the top and bottom chord? I'm not against using steel in the construction either visible or hidden. What I need is a design I can send to an engineer as I'm sure the city will request one. 

[Don P]

Sorry to be slow, an old friend has had seizures the past 2 nights, we've been walking the floor. Here is what I've been thinking, enough to get you going?

I'm getting a pass on the kingpost calc with 8x10's in #2 EWP... be careful here, ewp- throws a spray of knots every year causing a cluster every 2-3' along the timber. It is VERY easy to get #3 which will fail in a bad way at these spans. Explain and pay premium for the timbers.

I doubt you want to try to buy or move 38-40'ers. I think I'd do a discontinuous bottom chord but it is your option. Steel needs to run from a heavy "stop" at and under one heel, through the king, under the opposite heel to its' "stop". This is the bowstring and is what is carrying the bottom chord tension. the wood is decoration, 8x10 will visually match the top chords. I would camber at least an inch at the king, more would not hurt. That is where to use the worst of the timbers, the top chord is in compression with bending and that is what the calc was checking.

The start of a rough sketch, posts under/within 45 degrees of the heels , showing roughly~18"-2' overhang.

[Don P]

Here's another thought, a Fink truss with steel rods for the tension members and wood for compression members. The strut, in work of that period, would have been cast iron so steel fabricated would not be out of place their either.

[Don P]

The Library of Congress has loads of pics, if you can get the search dialed and have patience
This heel connection looks light
Wood Truss Details
Truss pics


 

An interesting heel joint here.
Cast Iron Truss Details


A king rod here, well braced and strutted;


 

[Don P]

I was sketching till the ice melted off the dish, the sun just came out  8)
Playing with that kingpost sketch;


 

I drew this, stepped back and looked at it and realized there was a clash between the engineering in the calc and what I'm looking at.

Here is the problem, this line of code and my note to self beyond the <!--.

var chordLL = ((Span/4) * LLratio) <!-- Line Length of top Chord to web

It is assuming the web is supporting the midpoint of the top chord on each side of the ridge. for 10/12 to 12/12 I think this would be a close enough assumption. At 4/12 the result is going to be way off. We had better take 4x the greater span which will overestimate the axial force and accurately estimate the bending force. The result will be overly conservative. I called it 45' (technically add 1/2 the required bearing at each end) I'm not getting a pass till the 8x top chord is over 13" deep.  (the sketch is showing 8x10's now)

On the right side I've sketched the rough outline of, I guess Warren webbing. This is probably why you usually see a Fink layout of the webbing on most shallower pitches.