Steel Roofed Garage in Ireland

[mikkymo]

Good Morning to you all from Ireland!

Over the past year I’ve lurked, read and read some more about achieving isolation but I don’t want to start without getting advice from the forum. I hope there's enough info here, if not, I'll happily add it!

I have a detached garage in the UK which is ‘post and panel’ concrete construction and is nearly completed. The panels are well sealed with flexible mastic/sealant and all gaps will be filled with cement mortar before any work is done. The base of the garage is 12” concrete with steel mesh. Under that is 12” hardcore, 2” blinding/quarry dust, 4” EPS Insulation (Kingspan). The internal dimensions are 6m x 5m (19.5’ x 16.6’) and the wall heights at the sides are 2.2m (7’). The garage has a raised tie truss system for additional ceiling height in the middle portion, raising the ceiling height to 2.6m (8.5’) for 75% of the span. At the front there is an insulated roller door measuring 4.2m x 2.1m (14’x7’). Not perfect by any means, but it’s what I have.

The side door in the images below was omitted.





I have a little stream which runs behind my house and lots of trees which add to the background noise level. Typical noise levels for my neighbourhood are 45-50db. I want to use the garage as a man cave for working on cars, tinkering with engines and also as a place where my kids can make noise, hit things with hammers and learn how to skin their knuckles like I did with my Dad. In terms of how much isolation, ideally I’d love to use an angle grinder at 10pm and hear next to nothing outside. Data I’ve come across suggests an angle grinder produces around 110db when cutting. I’m possibly hoping for too much then! I’d be very happy to eliminate general talking and music being played at 80db.

My current plan is to build a freestanding 2x4” stud wall on neoprene rubber (2” away from the outside wall) around the 3 side walls of the garage, fill with 6” of fibreglass insulation, then cover with 3/4" OSB, bitumen roofing felt and then 1/2" drywall. For the roller door wall, I need occasional access for cars so a permanent structure won't work. I'm planning on making a freestanding, removable wall of 12 off 2.5x4.5’ ‘stackable gobo’ type boxes made from 8x1” boards covered in OSB and 2 layers of drywall each side. Each box will be keyed to fit together, sealed with neoprene rubber seals and clamped with latches/toggle clamps. The remaining 3 feet will be a door with similar construction. Something along the lines of this (but solid both sides):



With the roller door, that may result in a 3 leaf construction so my plan is to build them with fabric on the side nearest the roller door and see how it performs first.

Windows will have removable window plugs made from 8x1" timber frames, OSB and Drywall.

Like any newbie with a little knowledge, I’m fairly confident in the wall construction and I’ll have to work on maximising the non-ideal situation with the removable wall. The ceiling however is a serious concern as the garage has a ‘box profile’ steel roof. At the minute, this roof is completely vented so that moist air won’t condense on the inside and drip onto the garage floor below. My intention is to use Genie Clips and Hat channel on the underside of the trusses and hang a layer of OSB, bitumen roof Felt and drywall from that. Above the ceiling between the trusses will go 6” of fibreglass insulation.

From my research, I need Mass-Air-Mass to reduce the sound transmission, with the mass as isolated as possible. In my case, I think the OSB, Felt and drywall will give me reasonably good mass and the genie clips and hat channel provide isolation. The air gap is filled with insulation, but my outer leaf (trusses and steel roof sheet) is neither massy or sealed.





The roof sheets are screwed onto 2x2" battens which are then nailed onto the trusses.

To stop rot and condensation, I’ve considered adding a 'closer' board under the battens and at the eaves. My hope is that I can seal the attic, but leave the roof sheet vented to the outside.



Seems common to do with rigid insulation, but I'd propose to do it with OSB, or OSB and drywall if it helped.





What do you think? Do I gain anything by sealing the attic? Once the inner ceiling is up, there's not much I can do to seal it.

If you’ve read this far, I appreciate it. There's a beer waiting for you at the bar in heaven.

Sorry if the images are a little too big - still working on how to reduce the image sizes for posting here.

All advice gratefully accepted!

Thanks!
Mikky

[mikkymo]

I should add that the room will be a practice room as well - with 3 kids under 5, I haven't played amplified electric guitar for years now and it would be great to have a space where I can jam with my brother and friends. Testing of my 100 watt Marshall with some amazed looking children would suggest that 85db is the magic figure for jamming.

Am I better nailing the OSB to the underside of the battens to maximise rigidity, or adding the 'beef layer' as a separate, cleat held, more flexible layer? Drywall doesn't seem like a good idea in there as any moisture dripping from the underside of the roof will destroy it.

Once again, all help appreciated!

[Soundman2020]

Hi Mikky, and Welcome!

Your project is a little different in purpose from a studio, but still the same in technique and materials. Sound is sound, and blocking it is the same.

Data I’ve come across suggests an angle grinder produces around 110db when cutting.

Right, but most of that is in the mid and high frequencies, which are not so terribly hard to isolate as low frequencies. 110 dB coming from a drum kit is really, really hard to deal with. 110 dB from an angle grinder: less so.

My current plan is to build a freestanding 2x4” stud wall on neoprene rubber

It does not need to be on a rubber base. Floating a wall is really hard to do properly, and useless if not done properly. Plus, you don't need to do that in you case anyway.

then cover with 3/4" OSB, bitumen roofing felt and then 1/2" drywall.

Forget the felt. That's a very, very old technique that is not very effective, expensive, and not even very predictable. Just use the OSB and drywall, and you'll be fine.

For the roller door wall, I need occasional access for cars so a permanent structure won't work.

Then you have a problem....

Each box will be keyed to fit together, sealed with neoprene rubber seals and clamped with latches/toggle clamps. The remaining 3 feet will be a door with similar construction. Something along the lines of this (but solid both sides):

I do not have high hopes for that to work... how would you ensure a continuous air-tight seal and continuous, consistent surface density where that meets the walls, floor and ceiling? It is also basically just a single leaf, so how would you tie that into the two-leaf system you are talking about for the rest of the room?

Windows will have removable window plugs made from 8x1" timber frames, OSB and Drywall.

Forget the windows, You say you are building this place yourself, and it is not done yet, so just don't put in any windows in the first place. That will save you a lot of time, money, and effort, and will make it a lot more likely that you can get decent isolation.

Like any newbie with a little knowledge, I’m fairly confident

...

At the minute, this roof is completely vented so that moist air won’t condense on the inside

Right... which is why it should stay that way! Don't play with things that were designed for a specific purpose, if you don't want trouble down the line...

My intention is to use Genie Clips and Hat channel on the underside of the trusses and hang a layer of OSB, bitumen roof Felt and drywall from that.

Here too, forget the felt. Not justifiable.

but my outer leaf (trusses and steel roof sheet) is neither massy or sealed.

Yup... in other words, it is the way it was designed to be...

I’ve considered adding a 'closer' board under the battens and at the eaves.

You could, but that's a LOT of work, and expense...

It would be much, much easier to just nail OSB to the bottom of the existing joists, put drywall on that, call that your "middle leaf", then build a new ceiling across the top of your new walls. simple, fast, effective, less expensive.

Testing of my 100 watt Marshall with some amazed looking children would suggest that 85db is the magic figure for jamming.

With small kids around, that's a good setting. Don't want to damage their hearing so early in life: they'll likely do it to themselves anyway when they are teenagers, with their iPhone earbuds turned up to 140 dB, but at least you'll be protecting them in the intervening years by keeping things under control in the rehearsal space.

Am I better nailing the OSB to the underside of the battens to maximise rigidity, or adding the 'beef layer' as a separate, cleat held, more flexible layer?

You are better off not changing anything about the roof itself, and just leaving it the way it was originally designed.

I've done a couple of studios exactly as outlined above, with three-leaf systems up top, and it works well. Don't play with wild animals, and don't play with roofing systems: they can both bight you hard...


- Stuart -

[mikkymo]

Thanks for the reply Stuart!

I appreciate the 'real world' advice - a lot of the forum posts I've read with a similar structure suggest loading drywall and OSB everywhere, without considering the ventilation aspect for the roof. The mention of an intentional triple leaf is heresy on the 'home theatre' type forums.

Nailing OSB to the underside of the trusses will be much much easier than what I was considering - and easier to seal too which should hopefully attenuate the mid/high frequencies.

I think as a start, I'll put up my inner wall (bolted to the floor?), OSB the underside of the trusses to seal the 'attic' and see what sort of isolation that gets me. Then I can look at adding genie clips and drywall or an independent ceiling.

My problem is that my trusses are 'raised tie' type, so the ceiling is like /-----\ with sloped areas along both sides. The wall height at the sides is 7' and the central height (across 75% of the span) is 8'6". Not so simple to build a new ceiling straight across the walls. I guess I could engineer something up to match the inside profile of the ceiling and rest on the inner walls.

The freestanding wall is a work in progress. If it works, great, if not, I'll have a portable spray booth or a load of saw horses to play with.

Thanks!

Mikky

[Soundman2020]

the mention of an intentional triple leaf is heresy on the 'home theatre' type forums.

Yup, they do go a bit overboard... "...out of an abundance of precaution..." as the authorities say, when they do something silly as a result of ignorance!!

A three leaf system isn't a sin. There's no commandment against it in the Bible, and no law against it either! It's just preferable not to do it if you can avoid it, but in those cases where you can't avoid it reasonably, you just compensate for it, and go on your way. "Compensating" basically means two things: 1) put more mass on the "middle" leaf (optimally, it should be as much as the mass of the other two leaves combined), and 2) increase the size of the air gaps on either side of that middle leaf. If you do that intelligently, then a three-leaf system is no worse than a two leaf.... except that it cost you more to build it, and takes up more space. Acoustically, there's no difference. In fact, a 3-leaf is BETTER at isolating mids and high frequencies!

Nailing OSB to the underside of the trusses will be much much easier than what I was considering - and easier to seal too which should hopefully attenuate the mid/high frequencies.

Yup. But sealing is more about overall total isolation, than about just mids and highs. Sealing is the difference between "I have great isolation" and "I have no isolation".... It's that simple. Sealing is imperative.

as a start, I'll put up my inner wall (bolted to the floor?), OSB the underside of the trusses to seal the 'attic' and see what sort of isolation that gets me.

That would probably be a mistake. If you go that route, then you have already attached your inner-leaf walls to your middle-leaf ceiling, which is actually attached to your outer leaf walls.... that's a major flanking path, and you would not get good low frequency isolation like that. If you go with a 3-leaf the way I described it, you would need to make sure that the inner-leaf does not touch any other leaf at any point. Not even a single nail or wood splinter can join them. They must be totally separate, "fully decoupled" (except for the floor).

Then I can look at adding genie clips and drywall or an independent ceiling.

You would not be able to do that with just isolation clips, as that would create a very thin ai gap between the middle and inner leaves of your ceiling, which is the opposite of what you need, especially for a 3-leaf system. And adding new joists plus drywall is also not an option, unless you first decouple the inner-leaf walls from everything else...

My problem is that my trusses are 'raised tie' type, so the ceiling is like /-----\ with sloped areas along both sides.

Yes, I noticed that. I have designed a few studios like that on purpose, to gain more ceiling height inside the rooms. For example:

OWLA--trusses-10-and-ridge-beam-image-ENH-SML.jpg

OWLA-Raised-tie-trusses-and-blocking-1-SML.jpg

OWLA-USA--V2-S055-middle-ceiling.png

That was for a garage conversion for a customer in LA

The one below is for a ground-up build, also in California, but more up north:

FRKUS-20150319_163944-Roof-trusses-in.jpg

FRKUS-20150320_190203-Sheathing-on-SML.jpg

I guess I could engineer something up to match the inside profile of the ceiling and rest on the inner walls.

That's what I normally do, yes! But I don't necessarily angle the inner-leaf ceiling: I often "step" it up at a certain point, and use some of the space for HVAC ducting and silencer boxes. If you design carefully, you can get most of the room with a high ceiling, and only some parts at the edges down lower.

The freestanding wall is a work in progress.

What is going to support the ceiling on that side of the room, when you take the wall out? What is going to provide the lateral support for the two adjacent walls when you take that wall out? This isn't just about acoustics or convenience: it is about structural integrity. You'd be letting yourself in for some major structural nightmares if you go down that path. You better get hire your structural engineer right now, as you are going to need his input on how to support and brace the rest of the room when that wall is not there. My guess is that you'll need a very major beam across the top (perhaps even an RSJ), and some type of isolated lateral support devices, such as with sway braces and/or seismic snubbers, but you should leave the engineering details of how to do that, and the specifications, to the structural engineer.


- Stuart -

[mikkymo]

Stuart,

I’ve been thinking a lot about this over the weekend and I can’t figure out an easy / low cost method of creating an independent ceiling under the existing trusses. I’m an Aerospace Engineer so I have a rough appreciation of the structural loading aspects but as you said it would absolutely require a structural engineer to complete some calcs which would be another cost. Even from checking span tables for timber joists, 5m is on the limit of what is safe, never mind adding in raised tie portions, etc.. Plus it would all have to be tied into the internal walls and around the door, plus bracing, etc. as you mention. Basically, once you factor it all in, an independent ceiling is just way beyond my budget. It would be much easier to have independent walls around the outside (bolted to the floor?) which don’t touch the outer walls at all, and a ceiling which is attached to, but decoupled from the trusses.

I’ve spoken to the company who designed the roof and they’re happy to have 2 layers of OSB and a layer of Drywall attached directly to the underside of the ceiling as they’ve accounted for a much higher ‘dead load’ on the trusses which could have included 18mm ‘sarking’ sheet over the trusses (I used battens) or heavier sheeting, some junk stored in the attic and a plasterboard ceiling.

I don’t mind the work of adding the closer panel to the underside of the battens, if it helps with isolation. It’s just a matter of copying the gaps between trusses exactly (I can make an adjustable jig to help), nailing to the battens and then sealing/caulking. There will still be space between the closer panel and roof sheet for ventilation and I can add a fan to the attic to force air to circulate if required.

Or do I add my mass to the inner ceiling instead? Am I better bulking up one layer or trying to balance the mass between the two leaves? (Leafs? Layers? Eeesh it’s a Monday morning…)

If I use genie clips and hat channel for my inner OSB/drywall ceiling, will it be sufficiently decoupled for the mid/high frequency noise I’m trying to isolate? Is the soft rubber in the Genie Clips actually worth it? Would rubber/neoprene washers and cheaper IB-1 type clips do a similar job?

Sorry for all the questions! I really appreciate the help with this!

[Soundman2020]

I can’t figure out an easy / low cost method of creating an independent ceiling under the existing trusses.

You do the same as for any other 2-leaf construction: You put joists across the top of the inner-leaf wall (not attached in any way to the outer leaf wall), and you hang the drywall from that.

Even from checking span tables for timber joists, 5m is on the limit of what is safe,

I'm not sure what span tables you are looking at, but from what I can see plain old 2x8 joists spaced 16"OC will easily span 5m while supporting that type of load. In metric terms, that would be 40x190 joists spaced 400mm OC. I can't see why that would not be safe. Is there something different about building codes in the UK? Do the require higher strength joists? If so, you might need to go to 2x10's instead: plenty strong enough. You could hang a small truck from a few of those!

never mind adding in raised tie portions,

For the above, I'm assuming a flat ceiling, or a stepped ceiling. There's no need to resort to gabled ceilings for your rooms. If you design everything carefully, it's not that hard to make it all fit by just stepping up the center section, or locating only the rooms that need it in the center section, with ancillary rooms around the edges. If you've never done this type of design before, you might want to consider hiring an experienced studio designer to do it for you. PM John and see if he is available.

It would be much easier to have independent walls around the outside (bolted to the floor?) which don’t touch the outer walls at all, and a ceiling which is attached to, but decoupled from the trusses.

Correct. That is, indeed, the correct way to do it at the lowest cost, and with the highest isolation.

There will still be space between the closer panel and roof sheet for ventilation and I can add a fan to the attic to force air to circulate if required.

I would check your local building code to see if that is allowed first, then run the design past your inspector to make sure he will sign off on it. My guess would be that if your plan would need an electric fan to make it work, then it will not be acceptable. Residential roof systems generally use only natural convection. If the fan fails, you'd be in deep trouble. Two things you don't want to do, are to build something that does not pass inspection, and to build something that causes damage to your building, which perhaps won't be apparent until months or years later...

Or do I add my mass to the inner ceiling instead? Am I better bulking up one layer or trying to balance the mass between the two leaves?

For a two-leaf roof, the best isolation is obtained when the mass on leaf one is about the same as the mass on leaf two (surface density, of course: not total mass). For a 3-leaf system, as I already said, you get the best performance when the mass of the middle leaf is the same as the combined mass of the other two leaves. In other words, m1 + m3 = m2. The equations are fairly simple. Did you try plugging in values that you project you'd be using, to make sure that f0, f+ and f- are reasonable, and that TL is within the range that you want?

If you go with your plan, you would be building a 3-leaf system, so most of the mass would have to go on the underside if the rafters, not on top of the joists. This means that ALL of your mass will be on the trusses. Outer-leaf, middle-leaf, and inner-leaf. You should go back to the roofing company and explain that scenario to see if they are OK with it. It would also reduce the maximum wind loading, seismic loading, and snow loading (if applicable) for your roof, since you'd be adding mass to the actual roof deck, not just to the attic joists. Very different, structurally.

If I use genie clips and hat channel for my inner OSB/drywall ceiling, will it be sufficiently decoupled for the mid/high frequency noise I’m trying to isolate?

Assuming that you follow the manufacturers instructions, have a large enough sealed air gap, sufficient mass on all three leaves, and sufficient damping (insulation) in both cavities, then yes.

Is the soft rubber in the Genie Clips actually worth it?

Yes. Because that's what it is designed to do, and has been tested to do, by independent acoustic labs. As far as I know, the manufacturer of Genie clips do provide the test reports on their website, or if not you could request that they e-mail them to you. If the can't or won't, then forget that manufacturer and look for one that is honest about their product.

Would rubber/neoprene washers and cheaper IB-1 type clips do a similar job?

Do you have access to an independent acoustic laboratory that can do the type of testing you'd need to determine if that would work or not? Do you have the money to pay the lab for such tests? What type of rubber are you thinking of? What durometer hardness? Which Shore scale are you measuring that on? What's the compressive modulus for that rubber? What shape factor would you use? What static deflection would you need? What load would you need to get the correct static deflection? What would be the resonant frequency of that system, and how does that compare to the highest resonant frequency that is allowable for your wall? What load could such a combination safely support? If you can answer all those questions confidently, and then do the math, then you could probably do that, but you'd be responsible for getting the structural certification for your devices. Do you really want several hundred kilograms of mass hanging over your head, when it has never been tested or certified as being safe?

Or you could just buy the devices that were specifically designed and tested for this very purpose, and whose performance is already known....

[mikkymo]

Do you have access to an independent acoustic laboratory that can do the type of testing you'd need to determine if that would work or not? Do you have the money to pay the lab for such tests? What type of rubber are you thinking of? What durometer hardness? Which Shore scale are you measuring that on? What's the compressive modulus for that rubber? What shape factor would you use? What static deflection would you need? What load would you need to get the correct static deflection? What would be the resonant frequency of that system, and how does that compare to the highest resonant frequency that is allowable for your wall? What load could such a combination safely support? If you can answer all those questions confidently, and then do the math, then you could probably do that, but you'd be responsible for getting the structural certification for your devices. Do you really want several hundred kilograms of mass hanging over your head, when it has never been tested or certified as being safe?

ARRRRRRRRGGGGGGGHHHHHHHHHHHHH!!!!

You have the patience of a Saint Stuart... Your logic is faultless though!

I know it'll be far more expensive to do, but I'm going to follow your advice and look at doing the independent ceiling. It will take the existing roof entirely out of the equation and I can build something which I can be confident will hold the extra ceiling mass and will be properly isolated.

So from re-reading your other post, I nail OSB to the existing trusses, add drywall and then new ceiling under that? KISS? (Keep It Simple, Stupid!). I know I could put the new ceiling joists between the existing trusses, but then I reduce the ventilation gap for the roof so it may be a huge headache in the future.

Do I add insulation between the OSB and the steel roof sheet (leaving a large gap for air to circulate) and insulate between the new independent ceiling joists? Or just the new ceiling joists?

I’m a lot happier with the independent ceiling / room-within-a-room idea because the genie clips/hat channel is a total unknown to me. Regardless of the structural implications of using the existing roof trusses, I’d not have as much confidence in my ability to do it ‘right’ without shorting out, adding too many screws to the hat channel or too few.

Using the angled portions of the ceiling for HVAC sounds like a great idea. I’ll also see if there’s a method of creating an angled joist to cover the span and get a structural engineer to do calcs.

Thank you very much for the advice – I really appreciate it!

[mikkymo]

I’ve spoken to a structural engineer about the internal room and after saying lots of things like “you’d reduce your sound more easily using wall coverings” and “you could tie the new ceiling to the existing roof trusses for greater strength”, he mentioned a ‘torsion box’ ceiling would work well in this instance, supported by the inner side walls, as per the attachment.

So after much Googling, a torsion box is essentially a gridwork of lightweight structure with an upper and lower skin glued on. Aircraft wings and hollow core doors are all torsion boxes. If you think of a hollow core door, there’s a cheap, thin wooden frame around the outside, a honeycomb paper/cardboard structure inside and 2 wood skins glued on. If you were to lay the door out flat like a table between two blocks and stand on it, the lower skin of the door would be in tension and the upper skin in compression. The strength is all in the glue joints which will transfer the load and stop the door breaking under your weight. If one or two glued areas fail (e.g. under impact), the whole thing will potentially fail catastrophically.

My structural engineer felt that as there’s no snow or wind loads, if I made a gridwork out of 8x2s and had a plywood skin (not OSB apparently as there’s no grain/fibres to safely carry the load) glued AND screwed on the underside, I would increase the overall strength of the structure. Some flat roof systems are made this way. Adding an upper skin makes it a true torque box, but acoustically that makes it a 2 leaf system – unless I can bridge the two leaves to make a single leaf?

If I were to get the torque box designed, could I fill the gaps in the gridwork with something to make it a thick single leaf? Compressed insulation? Or keep it as a 2 leaf system? That would make my overall system:

1. Internal ceiling lower skin (plasterboard, plywood)
2. Internal ceiling upper skin (plywood)
3. OSB on underside of trusses
4. Metal roof sheet

4 leaf system? Enough to make John Sayers and Rod Gervais throw up a little? lol!

[Soundman2020]

saying lots of things like “you’d reduce your sound more easily using wall coverings”

“you could tie the new ceiling to the existing roof trusses for greater strength”

Let's hope he's a better structural engineer than he is an acoustician!!!!

he mentioned a ‘torsion box’ ceiling would work well in this instance, supported by the inner side walls, as per the attachment.

... which would be fine, except that you can't do that when you also need good acoustic isolation, because the third leaf would trash the low end.

My structural engineer felt that as there’s no snow or wind loads, if I made a gridwork out of 8x2s and had a plywood skin (not OSB apparently as there’s no grain/fibres to safely carry the load)

Are you SURE this guy really is a structural engineer? The experts on OSB do not agree with him at all:
"Due to its high mechanical properties and the orientation of the strands within panels, OSB is particularly suitable for load-bearing applications in construction and is widely used for flooring, roof decking and wall sheathing,"
and "The main merits of OSB lie in the field of its mechanical performance, which is directly related to the geometry of the strands and their orientation within the panel."
and " ... in either diaphragms or shear walls, OSB sheathing uses an engineered manufacturing process to provide superb performance in racking and deflection strength. Under all types of exterior cladding, OSB provides extra thermal resistance and acoustic control. In addition, OSB is highly workable."
and "wood frame structures with OSB-sheathed walls perform well against seismic forces or racking shear – much better than concrete or masonry. Where high winds are of concern, OSB wall sheathing provides peace of mind because of its strength and density."
I could go on, but I think you get the picture: It would seem that your structural engineer does not understand the structural properties of OSB, which is rather strange... I've never heard of a structural engineer who did not know that OSB is commonly used in structural engineering. In fact, many engineered beams are specifically made using OSB for the webbing, precisely because of its structural properties. Hmmm.... There's something fishy about a structural engineer who has not idea about OSB's structural capabilities.

Try googling "picture of structural OSB" and see what you get... It's amazing that all those structures could be built using OSB, when your guy says that OSB cannot be used for structures...

Aircraft wings and hollow core doors are all torsion boxes.

Aircraft wings and hollow doors have to carry all kinds of loads, in all kinds of different directions, including torsional loads, which is why they need to be built like that. Your ceiling only needs to carry its own weight. Therefore it can be built in the exact same way as all other similar ceilings, have already been built, millions of times over, all around the world. Including the ones I showed you above, which passed inspection just fine, and are still standing years later. In earthquake country...

Take another look at those designs: there is way more than adequate support for the load, provided by the framing itself: after all, it seems to be standing up there, all by itself, pretty darn well, with no sheathing on it! The sheathing on that adds sheer strength, not load-carrying strength. Think about it... The load carrying capability is provided by the framing, not the sheathing. No torsion box is required, because there are no torsional forces acting on the ceiling (which is different from aircraft wings and doors, where there really ARE torsional forces in action).

Numerous people have already built these structures, in numerous places, approved by real structural engineers who actually do know what they are talking about. You should probably forget about the one you already found, and look for one of those real ones, who won't go spinning you fascinating yarns about esoteric techniques that simply are not applicable to your situation. I'm not a structural engineer myself, but I have designed and built enough of these structures to understand that the guy you spoke to is just blowing smoke. OSB has no structural strength? Really? That will be major news to thousands of building contractors and millions of homeowners whose houses have been built using structural OSB. Ceilings must be built using "torsion box" design? Really? That too will be news to numerous contractors and home owners who have built theirs without torsion boxes.

All you need to do is to frame the ceiling suitably to carry the weight (dead load) safely, then sheath the framing on only one side (upper or lower), with OSB, using the correct nailing schedule. If you don't know how to do that yourself, you should hire someone who does, to do it for you.

It seems to me that you are trying to re-invent the wheel, but you are trying to make it toroidal in seven dimensions using string theory, when all you really need is just a circle with an axle in three dimensions... Wheels already exist: just copy them. Ceiling framing already exists: copy that too.


- Stuart -