Vaulted my ceiling!

[Rickenbackerman]

Last year I got a Direct Lift four post lift, but with my 8' ceilings and conventional trusses, couldn't really do much with it. So I vaulted the center bay of my 24x36 detached three car garage by installing two LVL's, a double 2x12 ridge beam and 2x8 rafters, and then removing the three trusses above the center bay. It came out pretty damn good if I do say so myself! Having the lift to do this work was a HUGE help, too - I built two platforms that sit in between the runways, giving me one huge work surface that I can set at the perfect height.

I didn't get a permit, but I'm an engineer and I sized all of the lumber myself. The hardest part was getting those LVL's up there, them things is H E A V Y. I cut the trusses out last night - I cut the bottom chords first, and they only pulled apart about 1/16", which (in my mind) means that I did good and everything is nice and stiff. These pics are crappy, but better than nothing, right?

I still need to sweep up and stick my car on it and see how high I can go, my firebird is pretty low and I think I might be able to go to full height - the bottom of the ridge beam is 11'2".

Only had one little "whoopsie!"...

Piece of one of the trusses fell, bounced off the lift and made a bee-line for the back window.

 

[Rickenbackerman]

I should also mention that I spent a TON of time planning this out, and about a weeks worth of google searching on figuring out the best way to do this is what brought me to this site in the first place. And now I'm hooked. And planning on running gas and more juice out there, redoing the electrical, insulation and OSB walls, metal ceiling, etc... you know how it goes!

 

[IONH]

That's awesome. I want to do the same to my kitchen/living room. I had already planned on removing the wall with LVLs and transfering the weight to my basement, but then I decided on vaulting it instead.

Can you point me to any good sites you found with info? I have found some about putting in collar ties and attaching to those which allows for the outer walls to support the weight previously supported by the center... but those threads usually go to someone saying to get an engineer etc. I will get an engineer regardless, but for this early of a stage I would like to just read about it.

Definitely post more pictures. I also have a garage like yours where the roof is blocking the lift. I opted for a scissor lift instead which takes up too much floor space. I'd still go for a shorty lift if I had about 3-5 more feet of head space.

 

[blkhonda1991]

what is the ridgebeam sitting on?

 

[Rickenbackerman]

A notched 2x10. I put both hurricane clips and two horizontal metal straps across the 2x10 both above and below the ridge beam to prevent it from trying to kick sideways.

 

[Rickenbackerman]

Can you point me to any good sites you found with info? I have found some about putting in collar ties and attaching to those which allows for the outer walls to support the weight previously supported by the center... but those threads usually go to someone saying to get an engineer etc. I will get an engineer regardless, but for this early of a stage I would like to just read about it.

I haven't found much info online at all, actually. I did see a few about going 1/3 of the way up with collar ties but with my situation that would have only given me about another foot. ...and yeah, there's always the guys that jump in and say "don't touch it, get an engineer!!!", etc... If the light is good when I get home I'll take some more pics.

 

[IONH]

I haven't found much info online at all, actually. I did see a few about going 1/3 of the way up with collar ties but with my situation that would have only given me about another foot.

I believe it's 1/3 the way down. I was going to say that your ties are nearly at the ridge.

Did you have an engineer sign off on this approach? Any permits? Will your inspector actually be coming out to inspect the work during/when finished?

 

[Rickenbackerman]

I think you're right, 1/3 down. Either way, I wasn't comfortable with that approach, I'd think it'd be less stiff than the way I did it. I don't have any collar ties at all - just some pieces at the peak cut to bear on the rafters and give the ridge a flat spot to push up on.

No permit, no sign off, no inspection. I do structural work for NASA - wood is a bit different than A286 alloy, but I'm pretty confident in what I've done.

 

[Rickenbackerman]

The other thing I forgot to mention is that I kept the LVL completely separate from the existing truss next to it, it's only carriage bolted to it at the ends. This means I can keep an eye on the relative displacement of each LVL and truss so see if things start to sag at all. And if it does (for example, if we get blasted with 4' of snow this winter), I can stick a lally column under the center of each beam for temporary support.

 

[IONH]

My thought on placing the tie so far up is that there is more leverage/sag on the rafter below it. Originally that would not be an issue because of the ceiling joists which kept the walls together.

Anyway, looks great and I am glad to see someone else doing things like this themselves. I wouldn't vault my entire ceiling as I have a big I beam in the middle allowing my free span the full ~32 foot depth. I would vault just one side if I were to do it at all.

My house on the other hand.. I would do the whole thing and get rid of that pesky load/wall entirely.

 

[Kevin54]

Just making sure, but you did nail the roof sheeting back down to the new rafters right?

 

[familytruckster]

It looks great, exactly how I am planning on replacing my trusses in one bay.

Just making sure, but you did nail the roof sheeting back down to the new rafters right?

On the same line nailing and gluing the new rafter to the old truss top beam should be sufficient.

I would suggest blocking the LVL on the ends to keep it from rotating.

 

[Rickenbackerman]

Just making sure, but you did nail the roof sheeting back down to the new rafters right?

I did not, what's the concern there? To keep them from rotating? The new rafters are carriage bolted to the old top chords - two at each end and one in the center. And the LVL's are carriage bolted as well, two at each end. I also plan on adding hurricane straps at the ends of ALL of the trusses/rafters in the garage.

More pics coming shortly...

 

[Rickenbackerman]

 

[IONH]

I see the jack studs on the outside wall. I hadn't noticed the "jack rafters" (if that's even a term) from the collar tie down. I had not seen this listed on the few sites which had any information. Did you do this to avoid sag in the roof?

What size were your original rafters and what size are those jack rafters?

 

[knedgr]

what carries the load to the ground?

 

[Rickenbackerman]

I see the jack studs on the outside wall. I hadn't noticed the "jack rafters" (if that's even a term) from the collar tie down. I had not seen this listed on the few sites which had any information. Did you do this to avoid sag in the roof?

What size were your original rafters and what size are those jack rafters?

I'm not following you. I'm not sure what you're calling "jack rafters"? As I said before, I don't have any collar ties, just those "house shaped" wedge-things up at the peak. Those are just to give the ridge beam a flat spot to push up on.

 

[Rickenbackerman]

what carries the load to the ground?

The load is transferred from the LVLs to the header at the front (which is a doubled 2x12 with a 3/16" steel flitch plate sandwiched in between - wicked strong), and the wall at the back (which has additional studs nailed in under each LVL to carry the load).

 

[IONH]

I'm not following you. I'm not sure what you're calling "jack rafters"? As I said before, I don't have any collar ties, just those "house shaped" wedge-things up at the peak. Those are just to give the ridge beam a flat spot to push up on.

Looks like mini collar ties at the top where the LVL is attached below.

A jack stud is a second stud sister'd in next to a regular stud to help transfer load down. I called your secondary new rafters "jack rafters" because I don't know if there is a better term for them. I believe those would assist in keeping the walls from bowing out.

 

[buening]

Nice work. Below are my initial concerns from a fellow structural guy:

- the 2x12 or whatever you used for the transfer columns from the ridge beam to the LVL may be a problem, as this will have bending moment from the ridge beam in the end 1/4s. The way it is set up it is bending about its weak axis. If you place a straight edge like an 8' level and notice any bowing, then I'd double up that 2x for more column depth.

- the LVL is coped at its ends in order to fit into the soffit/top wall plate. The LVL appears to have sufficient depth to withstand the bending moments at midspan, but it also needs to have enough shear capacity at its ends. Shear capacity is related to the beam depth as you likely know being a fellow structural. You don't have your location in your profile, but with enough snow load I fear the LVL will shear off at the ends. You only have a couple inches of beam depth at its bearing location. If you don't get snow, then wind should be considered.

- On one end of the LVL it appears you added another wall stud, but make sure that is enough. I assume you did a free body diagram of the entire system. You will get a little distribution from the 2x4 top plate below the LVL, but to be conservative I'd assume all load is applied to the 2 or 3 studs you have supporting it.

- The other end of the LVL appears to rest on the garage door header, which it likely won't have the capacity to support such loading. Double check that

- What fasteners are being used to attach the ridge beam to the transfer column? My scale may be off but it appears screws were used. Make sure they are ones recommended for that plate and not general purpose screws.

All in all I think it should be fine as long as you don't get extremely high winds (south florida) or heavy snow loads (northern US).

Edit: took awhile to type out and didn't see above few comments

 

[Rickenbackerman]

Nice work. Below are my initial concerns from a fellow structural guy:

EXCELLENT post!

- the 2x12 or whatever you used for the transfer columns from the ridge beam to the LVL may be a problem, as this will have bending moment from the ridge beam in the end 1/4s. The way it is set up it is bending about its weak axis. If you place a straight edge like an 8' level and notice any bowing, then I'd double up that 2x for more column depth.

Good call. I'll keep an eye on that. Those columns are actually pretty short though, about 26" from the LVL to the bottom of the ridge beam.

- the LVL is coped at its ends in order to fit into the soffit/top wall plate. The LVL appears to have sufficient depth to withstand the bending moments at midspan, but it also needs to have enough shear capacity at its ends. Shear capacity is related to the beam depth as you likely know being a fellow structural. You don't have your location in your profile, but with enough snow load I fear the LVL will shear off at the ends. You only have a couple inches of beam depth at its bearing location. If you don't get snow, then wind should be considered.

I'm in Maryland. I sized everything to 40psf total. Since I have light roofing and will only be using lightweight materials for the ceiling, I went with 10psf dead and 30 live. I admit the ends of the LVL's do look scary, and I was worried about that too, but there is only ~300psi of shear there under full load. I don't know what the shear allowable of the LVL is, but the bending strength is 2600psi, so I just assumed I'm good. I might nail a chunk of 2x4 under the end of each LVL to bring the shear load out into where the beam is a little deeper anyway.

- On one end of the LVL it appears you added another wall stud, but make sure that is enough. I assume you did a free body diagram of the entire system. You will get a little distribution from the 2x4 top plate below the LVL, but to be conservative I'd assume all load is applied to the 2 or 3 studs you have supporting it.

I'm a bit of a nerd. I did all my calcs in MathCAD. Being a fellow engineer, if you've never used it, you're missing out. We're talking about 720lbs at the end of each LVL under full load. I assumed three studs would support that (50psi).

- The other end of the LVL appears to rest on the garage door header, which it likely won't have the capacity to support such loading. Double check that

As stated above, that header is a doubled 2x12 with a 3/16" steel flitch plate sandwiched in between. I'm not sure what the load capability is, but the garage has been there for about 30 years and there is ZERO sag in that header, so I just went for it. Advice on that one?

- What fasteners are being used to attach the ridge beam to the transfer column? My scale may be off but it appears screws were used. Make sure they are ones recommended for that plate and not general purpose screws.

10 penny nails and the big U-shaped hurricane clips. The horizontal metal straps are also nailed in, but they're just there to keep the transfer column from wanting to split.

All in all I think it should be fine as long as you don't get extremely high winds (south florida) or heavy snow loads (northern US).

I do plan on adding hurricane straps EVERYWHERE too. Just for my own sanity.

 

[M-technik-3]

That's a good idea I might be tempted to do something of this nature to my garage.

 

[Rickenbackerman]

For you and anyone else who might be tempted - DISCLAIMER - this is not a "how to". Although I admit I did not, everyone should consult an engineer (I am one) and go through the proper permit and inspection process to ensure that their roof doesn't cave in.

 

[Falcon67]

I just went over this in my "new shop thinking" and will add it to my build thread, but the short info is that a "collar tie" is a small board - typically 1x6 - no more than 1/3 down from the ridge board that helps keep the rafters against the ridge. It should be used every 4' or so.

Raise the joist more than 1/3 above the plate or drop the collar tie more than 1/3 (rules of thumb ONLY) and it changes to a "Rafter tie" which is a completely different deal. Rafter ties put one hell of a load application on the plate, rafter and ridge, to the point where (based on roof loads) you cannot put enough nails in the joints to sustain the load. You also have to upsize the rafters - calculators are out there to figure the adjustment. In one case I penciled, I'd have to upsize the rafters by 45% - from 2x6 to 2x10, basically.

A full vault requires an engineered ridge beam with sufficient support at each end and should be sized (based on what I've read so far) to carry at least 1/2 the roof load. In my sketch, a small 24x22 area at 20 psf figures up to over 6000 lbs. I figger this means that instead of transferring roof loads down the rafter to the plate and putting a side load on the top plate, a vault places mucho load on the ridge, which you then have to transfer down to the foundation. Not so much goes down the rafter to the plate. The ridge has to carry a bunch of the load to keep the walls from pushing out. Some of the rafter is just along for the ride.

Some of this info can be found in the complete Southern Pine span table reference, but only in cautions.

So if you consider this option - study up or call an engineer, it's not as simple as it might look from the floor! LOL.

 

[buening]

Works for me man! Without running any calcs, my observations are just visual/gut feel. The header may work fine since it has the steel flitch plate. I don't use MathCAD much in my field (bridge design), but have found it can be a very useful tool!

For shear at the ends of the LVL, calculate your vertical reaction/shear force (V). To calculate your applied shear stress, fv = 3*V/(2*b*d). "b" is the width of the LVL and "d" is the remaining depth at the face of the wall (this is conservative, as actual shear force and depth of beam is typically calculated a distance "d" from face of wall). This is per NDS 3.4.2 I'm not sure what wood the LVL is made of, but for example the allowable service shear, Fv, for Douglas Fir South is 180psi and Red Oak is 170psi per NDS Table 4A. I believe a factor of safety is built into those PSI values.

That transfer column is pretty short, which could result in the bending transferred from the ridge beam maxing out at midspan rather than located near the ends on a longer column. If anything, install some kickers/diagonals to brace the transfer column to a board tied to other trusses if you begin to see deflection issues. All in all I think you'll be fine with everything.

 

[Rickenbackerman]

For shear at the ends of the LVL, calculate your vertical reaction/shear force (V). To calculate your applied shear stress, fv = 3*V/(2*b*d).

You are correct, I made a mistake there. I used V/A but it's 1.5 V/A, however I used the load at midspan rather than the end load, so the shear stress at the end of the LVL is actually 205psi. I think I'll definitely put some chunks of 2x4 to bring the shear point out a bit...

I'm not sure what wood the LVL is made of, but for example the allowable service shear, Fv, for Douglas Fir South is 180psi and Red Oak is 170psi per NDS Table 4A. I believe a factor of safety is built into those PSI values.

Is that an "across the grain" value rather than "with the grain"? I'd assume so, but 170psi seems incredibly low to me... I don't know what the LVL is made out of, either, although I'd think it'd have a weaker shear allowable than regular wood "across the grain", since 50% of the grain in an LVL is going the other way, since it's basically just plywood.

 

[petee_c]

very cool.

P

 

[WNYflyer]

As a fellow structural guy I will throw my two cents in and nice job by the way

- Your new construction is a post and beam structure (rafters + "ridge beam" rather than a "ridge board") therefore no collar ties required on your new rafters.

-New rafters should be tied to the old top chord based upon the allowable unbraced length of the compression edge of the new rafter which is of course relative to the load it needs to carry.

As Buening touched on,

-The coped LVL's at the door and wall. I would probably bulk those area's up by slapping some gauge steel on the side of LVL's at the coped area or build up some seated connections under the LVL at those locations to get under more depth of the LVL. The seated connections will cause eccentricities and moments though.

- I would also bulk up the post the ridge beam sits on.

- And lastly for the LVL's, where the post comes down on them. The LVL's are designed taking into account their compression edge under bending is unbraced for the entire span of the LVL? Typical load tables for lumber assume continous bracing of the compression edge is in place (i.e. roof sheeting bracing rafters, etc.). Also even if the LVL is designed to be unbraced for its entire length it should be tied off by 2x material perpidicular to it at the point the post comes down to keep that very important joint stable laterally. The 2x should run to your end walls and be tied off to the wall studs sufficiently. The steel code says that type of junction/joint has to be braced laterally and I assume wood would be no different.

 

[Falcon67]

This is all verrrry interesting. Question for Mr. Rickenbackerman - if your span for the ridge beam was -say- 24' and it was supported to the foundation on both ends, would it still be possible to use dimensional lumber or is the loading too great. I don't know enough of the math to work something from here.

 

[buening]

Is that an "across the grain" value rather than "with the grain"? I'd assume so, but 170psi seems incredibly low to me... I don't know what the LVL is made out of, either, although I'd think it'd have a weaker shear allowable than regular wood "across the grain", since 50% of the grain in an LVL is going the other way, since it's basically just plywood.

The allowable shear stress from the tables is parallel to the grain. Actual shear forces act at a 45°, but due to the grain of wood the shear forces are parallel to the grain. Since the shear acts at a 45°, this is the reason why its typically taken at "d" from the face of the support, as that is the top of the shear exit at the top of the beam.

I don't know what the allowable shear force perpendicular to the grain is, but due to the coping at the ends we would likely see shear failure perpendicular to the grain. My gut feel is the allowable shear perpendicular is more than parallel to the grain. The tables show a "compression perpendicular to the grain" column in which the allowable compressive stress for douglas fir is 440psi. I don't think that is this situation though.


Good input WNYflyer! The LVL can be designed either way. For example if it is intended to support floor joists for a basement, I believe the joists just rest on the LVL and thus not bracing the compression face. I do like your idea of sandwiching steel plates on each side to increase shear capacity.

 

[e-tek]

Looks plenty befey to me. Good work!

 

[Kevin54]

Forget what I said about nailing your sheeting down to the new rafters. I was under the assumption that you removed ALL of your old rafters and replaced with new. I did not see before where you sistered the new to the old.

 

[Rickenbackerman]

-The coped LVL's at the door and wall. I would probably bulk those area's up by slapping some gauge steel on the side of LVL's at the coped area or build up some seated connections under the LVL at those locations to get under more depth of the LVL. The seated connections will cause eccentricities and moments though.

Slapping some steel on the side is a better idea than what I was thinking (building up seated connections). I think I'll do just that.

- I would also bulk up the post the ridge beam sits on.

What do you think the failure mode would be?

- And lastly for the LVL's, where the post comes down on them. The LVL's are designed taking into account their compression edge under bending is unbraced for the entire span of the LVL? Typical load tables for lumber assume continous bracing of the compression edge is in place (i.e. roof sheeting bracing rafters, etc.). Also even if the LVL is designed to be unbraced for its entire length it should be tied off by 2x material perpidicular to it at the point the post comes down to keep that very important joint stable laterally. The 2x should run to your end walls and be tied off to the wall studs sufficiently. The steel code says that type of junction/joint has to be braced laterally and I assume wood would be no different.

Good call. I'll run some 2x4's from near the center out to the outer walls to brace the LVL's. Cheap insurance.

And a huge THANK YOU to another great post from a fellow structural guy!

 

[Rickenbackerman]

This is all verrrry interesting. Question for Mr. Rickenbackerman - if your span for the ridge beam was -say- 24' and it was supported to the foundation on both ends, would it still be possible to use dimensional lumber or is the loading too great. I don't know enough of the math to work something from here.

I don't know for sure without doing some calcs, but for a 24' span I'd bet my hat you'd need LVL's or an I-beam.

 

[Rickenbackerman]

The allowable shear stress from the tables is parallel to the grain. Actual shear forces act at a 45°, but due to the grain of wood the shear forces are parallel to the grain. Since the shear acts at a 45°, this is the reason why its typically taken at "d" from the face of the support, as that is the top of the shear exit at the top of the beam.

I don't know what the allowable shear force perpendicular to the grain is, but due to the coping at the ends we would likely see shear failure perpendicular to the grain. My gut feel is the allowable shear perpendicular is more than parallel to the grain. The tables show a "compression perpendicular to the grain" column in which the allowable compressive stress for douglas fir is 440psi. I don't think that is this situation though.

I don't think so either. I think I'm going to take WNYflyer's advice and slap some steel on the inside face of each LVL at each end.

 

[Rickenbackerman]

Oh yeah, I put my '99 Formula up on the lift - it went all the way up with 4" spare (at the deck lid)! I'd have even more room if I shifted the lift forward some, which I think I will...

 

[IONH]

Oh yeah, I put my '99 Formula up on the lift - it went all the way up with 4" spare (at the deck lid)! I'd have even more room if I shifted the lift forward some, which I think I will...

Can you stand under that? What's the total clearance?

 

[Rickenbackerman]

The lift is at it's max height, the locks are in the top notch. I think the bottom of the runway is at about 6'. I'm 6' tall, so it's not quite perfect, but good enough for me. It's a direct lift pro park 8S.

 

[IONH]

Adding the body clearance to the ground from the bottom of the tires, you probably have enough space to tilt your head up at least. Looks good.

 

[Falcon67]

I don't know for sure without doing some calcs, but for a 24' span I'd bet my hat you'd need LVL's or an I-beam.

Think you are right, which is enough to go on to the next idea.