Roof Truss modification, getting more head room

[livetoride21]

Hi everyone, I need a little help. I'm currently looking to get some more headroom out of my detached garage (see here http://www.garagejournal.com/forum/showthread.php?t=310259 )
So that eventually I can maybe put a lift in, but for now it will just be so I can drywall the ceiling and not have the space feel cramped.
My current roof trusses look like this ;

A standard truss for a gable roof.
I'm looking to either modify them to be like this;

or like this
(scissor truss variation)

Can anyone here enlighten me of problems I could have with either of these structures? The second one is essentially a standard scissor truss with more bracing, so I have little worry about strength, but the first is a very simple modification and will allow me more headroom, so if it is strong enough I would like to go that route.

I will be rebuilding these trusses one at a time from the inside of the garage. The roof will not come off.

 

[ssdave]

Do you have a roof truss design program that you've run these through? No way to just guess at them, they need to be analyzed, and if the reactions check out, then they are good to go.

If you don't have that capability, (I'm guessing not, since you asked the question here) you need to find someone who does to run the calc's. This isn't a get some internet opinions sort of problem to solve.

The other problem will be what will you use to do the joint connections? One of the assumptions of the truss analysis is that the joints are made with the industry standard plated truss connectors.

 

[livetoride21]

Do you have a roof truss design program that you've run these through? No way to just guess at them, they need to be analyzed, and if the reactions check out, then they are good to go.

If you don't have that capability, (I'm guessing not, since you asked the question here) you need to find someone who does to run the calc's. This isn't a get some internet opinions sort of problem to solve.

The other problem will be what will you use to do the joint connections? One of the assumptions of the truss analysis is that the joints are made with the industry standard plated truss connectors.

Thank you for the concern. I am a mechanical technologist by trade, so I'm not afraid to do simulation and calculation when I need to. I just wanted to throw the ball in the court to see if anyone had any big roadblocks to throw out there.

I ran the stress analysis on both my current truss design and my proposed design (second one) in autodesk inventor and the wall deflection looks to be acceptable (about 20% more than the current truss design)
I just need to find the proper material specs for Spruce Framing lumber to get the proper deflection in inches. Right now I get around .09" wall deflection (if they were on a pivot) with 50 PSF roof load. I'd imagine it would be quite a bit more than that. But at the very least right now I have a good comparison between existing and future.

Now of course, these calculations are assuming the Truss is essentially one solid piece, with no loose connections, but once again it is assuming this on both the old and new.

So right now I'm pretty confident all will be well!

Can anyone chime in with youngs modulus, the shear strength, yield strength (parallel to grain) and tensile strengh (parallel to grain) of spruce framing lumber?

thanks!

 

[ez-duzit]

...right now I'm pretty confident all will be well!...

I'm reminded of the line "Anything is possible when you don't know what you're doing."

Trusses are engineered. And built by approved specialists under strict specifications and quality assurance. What you are proposing seems to be to build them, in situ, without proper engineering or inspection.

 

[ssdave]

Hmmm... Kind of different way of going about it, the problem with wood is it doesn't act like an ideal material, so different methods are used than for steel or similar materials. But, should work if you use conservative enough values.

I did a quick google search, and found several sources that had the values you need. Just google youngs modulus and spruce and similar and you'll find a bunch of data.

Good luck!

 

[ssdave]

I'm reminded of the line "Anything is possible when you don't know what you're doing."

Trusses are engineered. And built by approved specialists under strict specifications and quality assurance. What you are proposing seems to be to build them, in situ, without proper engineering or inspection.

Somewhat agreed, but the design method being used actually will work and what he is doing is valid if he uses appropriate values and is conservative enough. Using a mechanical model like he's doing isn't usually the way wood is designed, but at least in theory is fine. It will be important to use fasteners that rigidly restrain the joints so that the mechanical model is valid. I don't know what those would be if you don't have access to the plated truss plate jointer machine; it will be real hard to make a ridgid joint in place.

The problem with wood is it doesn't react ideally, and it is a non-uniform material, so you can't accurately model it like steel. So, design values and methods take that into account, and are appropriately conservative. The quick search I did gives values that he's asking that are all over the place, reflecting the non-uniform nature of the material. I'd have a hard time picking which value to use, and would probably end up using a huge safety factor.

However, if you restrict the bending and deflection of the wood members to a very small amount, it should act somewhat like an ideal material, so the method would be valid. The higher the stress and the more the deflection, the further from an ideal material the wood will act. So, with an appropriate over-design, the method should work. In reality, design numbers for wood are held very low to account for the non-linear behavior and wood material defects, and effectively, wood design methods will end up with about the same result.

I'd be particularly careful in the middle design shown (the collar tie type one) because the bending force on the rafter below the collar tie would be quite high, and the wood may fail differently than an ideal material would. Wood also isn't a truly elastic material, and sags from deflection under load set and become permanent, and then progress with the next loading cycle. Thus it is increasingly important to limit deflections to almost nothing. The scissor truss design should be more valid with the mechanical model being used to analyze it, as the deflections would be low and the truss action would put the members into tension, which counteracts bowing tendencies.

You need to get allowable design values from building code, and apply loadings in accordance with building code to get a good representative design. The load factors in code have been developed specifically to take into account the long range deflection tendencies of wood structures.

 

[ovilla]

Keep it simple!
Go to a truss company!
Give them your roof slope and pitch or the stamped codes on your existing trusses.
Tell them you need scissor truss HALVES and ask for the hardware/method for assembling them on site.
Sister them to an existing roof truss.
Cut out everything below your new trusses.
Done!
Order your lift!

If you're overtly worried about handling the weight of snow or just want things beefed up, ask them to bump up your trusses. They will most likely build them out of 2X6's instead of 2X4's, or even bigger stock.

 

[matt_i]

How are you going to join the wood-to-wood connections?

 

[William Payne]

I know a guy who had a wooden framed garage built with open ceilings without trusses. However running the length of the garage at the peak of the roof is a huge wooden beam.
His logic was if they could wooden housing with big open peaked ceiling then they can do the same with a garage.

 

[larry4406]

Keep it simple!
Go to a truss company!
Give them your roof slope and pitch or the stamped codes on your existing trusses.
Tell them you need scissor truss HALVES and ask for the hardware/method for assembling them on site.
Sister them to an existing roof truss.
Cut out everything below your new trusses.
Done!

I've done this in a house with the roof shingles on but no drywall. We changed from a tray ceiling to a vaulted ceiling.

We have a member on site here goes by Medeek who designs trusses for a living. Maybe he can help you
http://www.garagejournal.com/forum/showthread.php?t=191720

 

[livetoride21]

Hmmm... Kind of different way of going about it, the problem with wood is it doesn't act like an ideal material, so different methods are used than for steel or similar materials. But, should work if you use conservative enough values.

I did a quick google search, and found several sources that had the values you need. Just google youngs modulus and spruce and similar and you'll find a bunch of data.

Good luck!

Thank you for the reply, the unfortunate thing when searching on google, is going to three different sites will yield 3 different values for material properties. However youngs modulus seems to be pretty close to the same all over. I'm going to do some more searching tonight.

Somewhat agreed, but the design method being used actually will work and what he is doing is valid if he uses appropriate values and is conservative enough. Using a mechanical model like he's doing isn't usually the way wood is designed, but at least in theory is fine. It will be important to use fasteners that rigidly restrain the joints so that the mechanical model is valid. I don't know what those would be if you don't have access to the plated truss plate jointer machine; it will be real hard to make a ridgid joint in place.

The problem with wood is it doesn't react ideally, and it is a non-uniform material, so you can't accurately model it like steel. So, design values and methods take that into account, and are appropriately conservative. The quick search I did gives values that he's asking that are all over the place, reflecting the non-uniform nature of the material. I'd have a hard time picking which value to use, and would probably end up using a huge safety factor.

However, if you restrict the bending and deflection of the wood members to a very small amount, it should act somewhat like an ideal material, so the method would be valid. The higher the stress and the more the deflection, the further from an ideal material the wood will act. So, with an appropriate over-design, the method should work. In reality, design numbers for wood are held very low to account for the non-linear behavior and wood material defects, and effectively, wood design methods will end up with about the same result.

I'd be particularly careful in the middle design shown (the collar tie type one) because the bending force on the rafter below the collar tie would be quite high, and the wood may fail differently than an ideal material would. Wood also isn't a truly elastic material, and sags from deflection under load set and become permanent, and then progress with the next loading cycle. Thus it is increasingly important to limit deflections to almost nothing. The scissor truss design should be more valid with the mechanical model being used to analyze it, as the deflections would be low and the truss action would put the members into tension, which counteracts bowing tendencies.

You need to get allowable design values from building code, and apply loadings in accordance with building code to get a good representative design. The load factors in code have been developed specifically to take into account the long range deflection tendencies of wood structures.

This is the exact problem I ran into with the material data. it's all over the place as wood is a Transverse Isotropic material, so it's quite hard to get its data set on paper. Hence comparing current to what I'm looking to do.
The second drawing I put up, I didn't have very high hopes for, and when modeled had way too much deflection under load. The third drawing is very promising, having only about 20% more deflection under load, and only two stress points in compression.

I've done this in a house with the roof shingles on but no drywall. We changed from a tray ceiling to a vaulted ceiling.

We have a member on site here goes by Medeek who designs trusses for a living. Maybe he can help you
http://www.garagejournal.com/forum/showthread.php?t=191720

I will have-to contact him to get his opinion, thanks!
Kevin