Portal Frames 1

[civeng80]

Does anyone have any knowledge or can they comment on industrial buildings with concrete tilt panels connected to rafters (either UB or open web trusses) with rigid connection to the panel so that the concrete panel (or part of it) acts as a beam column and where shear walls are not possible at the ends of the building (e.g. because of large openings). If so what would be the effective column width of the column/panel?
Can anyone tell me if there is any software/literature available on this subject?

Thanks !!

 

[hokie66]

My recommendation would be to use steel portal frames, so steel columns. Could possibly use a single leg portal, column on one side only. I think trying to use the concrete panels as portal columns is pushing the system too far.

 

[civeng80]

Thanks hokie66

I do agree with what your saying, but I have seen it being used. Im from Australia and the standard codes are AS4100 for steel and AS3600 for concrete. The rafters were open web trusses which were faily light and spanned about 15m to a central wall with another 15m span on the other side. effectively a propped portal.

Its a clean looking building inside of course with no steel columns. I just wonder how it was analysed and designed ??

Thanks!!

 

[hokie66]

Sorry, I don't think I can help, at least for your case with no shear resisting elements in the ends. I am in Australia also and quite often see things in my travels that don't look right with structures under construction. Tilt wall construction has taken most of the market for light industrial and storage buildings, but I honestly don't know who is designing them or how some things are justified.

 

[BarryEng]

Have a look at the CIA (Concrete Insitute of Australia, not the US spy agency) manual for tilt up concrete. They have examples showing various configurations of tilt up wall with rafters for the roof & no columns (except for large openings in the shear panel).

When I get back to work, I will find the exact reference & post it here.

 

[civeng80]

Thanks Barryeng

I think I know which reference your refering to. CIA precast concrete handbook right? Example 5.1b. However lateral forces are transfered to shear walls at the end thru roof bracing. One question while Im on this one, what do you think of the connection from rafter to panel, looks a little mickey mouse to me. Is there another more substantial connection available? Also what happens if there are large openings at the ends of the building?

 

[BarryEng]

There is a CIA pub on tilt up conc.

Try also CACA May97 - tilt up const notes - ISBN 1 876278 10 2

& tilt up const bldgs Jun98 Des & co9nst guide (UK) - ISBN 0 7210 15336

These pubs give connection details that may be of use to you.

 

[nicam]

tilt up panel design is very easy and is very economical to build. You don't need to fix (as in moment connection) the rafters to the panels, just shear connection. You don't want to transfer moments to the panel. the roof bracing (think of it as a horizontal truss) is the key to the design.

I usually create a model in microstran to size up all the members, then use inducta software to do the panels.

if there is a large opening at one end then I normally restrain the end panels at the base to resist over turning.

I hope this helps.

 

[hokie66]

Agree with nicam that you don't want to try to make a moment connection between rafter and panel. The concentrated moment would cause distress in the panel. And as he implies, you may be able to provide for overturning by fixing the bases of the panels. One way of doing this is to cast your footings with slots to receive the panels, and grout them in. Of course this makes the footings large and the panels heavier than usual.

 

[civeng80]

Thanks BarryEng, nicam and hokie66,

I do agree with you nicam and hokie66, portalizing a tilt panel would be pushing the system a bit and I certainly wouldn't do it even for short rafter spans. But the case I've seen its the only way it would have been done, since there is no roof bracing at all and the end walls have very limited panel guts to use as shear walls. Probably a very couragous structural engineer there right hokie66? As I said I wouldn't do it and I'm glad that you guys wouldn't either.
Now Im a civil engineer turned commercial builder I do my own designs (for my own projects) and am very confident in the typical portal frame system. Load bearing tilt panels are ok probably more for the smaller commercial shops, but I would still hesitate to use them in the larger building where forklifts and truck movements occur. I'd like to know what you guys thoughts are on this?

Also for load bearing panels as I mentioned to BarryEng there seems to be a faily flimsey connection with a small angle bracket bolted to the lower flange, why not just have a tee piece connected to the web? (panel can normally be adjusted by props to align the holes. What do you guys think about this?

Anyway thanks a lot guys you have been very helpful, a Merry Christmas to you all and my apology for the long rant.

Thanks!

 

[hokie66]

civeng80,

What you have seen indicates to me that rather than a courageous structural engineer, there was no structural engineer. Probably more like a builder who thinks no engineer is required, and a private certifier who deals in volume. Glad you are more conscientious.

Personally I prefer framed structures with the precast as cladding, but load-bearing precast is gaining in popularity in Australia, not only for tilt-up industrial buildings, but also for low to medium rise commercial buildings. Done with due care, it is another acceptable and sometimes economic system for consideration. Horses for courses.

I think the main reason for the connection you describe is to prevent prying on the panel, which I suppose supports my aversion to trying to take concentrated moment into the concrete.

Merry Christmas to you and yours.

 

[nicam]

No roof bracing at all?? Stay away from that building on a windy day!!

Like I said before the roof bracing is the key to the design.

I've design up to 37m clear span and 50m two span tilt up buildings. I used tilt up once for a load bear application, but that was with a combination of steel portal frame, with the columns on the outside of the panels, which were fixed to the columns.

As far as the the connection detail, I too asked this question when I first designed tilt up, but this has been the norm for connection panel to rafter (angled bracket) I've used this method of connection all the time without any problems. Also no moment tranfer.

 

[civeng80]

Thanks nicam

So 37m clear span with no columns? Also do you use microstran then just to design the simply supported rafters? What about end walls do you use the same rafter section at the end walls or a ledger angle supported continously ?

 

[nicam]

civeng80

with the 37m span you would use either WB for rafters or truss, I normally would use trussess, but on that ocation I used a 900WB.
End wall don't need rafters used raking angles as you mentioned.

I only us MS for larger buildings but for small rectanglar build I just use hand calc's

 

[nicam]

Also that can be the down side sometimes with tilt up, having no moment end connections with the rafters means the supports rotate and you end up with more deflection, hence bigger section required.

 

[civeng80]

Thanks nicam

900WB for a 37m span sounds about right I presume it had a ridge at the centre with a roof pitch of about 5 degrees? Was the connection to the concrete panel still an angle seat about 150 x 150 x 16thick, was there any web connector as well to prevent rotation? I think for that sought of span I would tend to think more about portal frames rather than load bearing tilt up panels. For one reason you just mentioned and the other one is that better sight control is required for tilt up panels especially for the larger building (although this is not the designers problem).

Ive been looking at castellated beams lately rather than open web trusses what do you think of those for the large span buildings?

Cheers !

 

[hokie66]

Be careful about trying to prevent rotation of a 900WB by connecting to a concrete panel. You will only succeed in breaking the panel. I agree with you that steel portals are a better solution for long spans, with the concrete panels used as cladding.

The use of castellated beams would depend on fabrication costs. The only times I have used them is for gymnasium roofs when the architect wanted that look.

 

[civeng80]

I meant lateral rotation, but come to think of it that would also prevent longitudinal rotation which would induce a crippling moment in the panel. So you would suggest just an angle bracket seat supporting such a deep rafter then (assuming this form of construction)?

Cheers !

 

[hokie66]

Sorry I misunderstood your question about rotation. The top flange does need to be restrained transversely. This is commonly done by the horizontally spanning waler which braces the top of the wall. If the waler is not at that level or close, you can use restraints bolted to the wall but not connected to the beam.

The vertical shear connection to the wall can take various forms, as long as the connectors into the wall are more or less in a horizontal line to prevent prying. The connection doesn't have to be right at the bottom of the beam. One common connection consists of an angle each side of the beam or slotted through, in turn seated on angles each side which are bolted to the wall.

 

[civeng80]

OK that makes sense.

Vertical shear connection. Do you mean 2 horizontal angles(or slightly graded to suit rafter slope) welded (say) on both sides of the web and supported on a prepared welded angle bracket with the bottom leg graded to suit the rafter slope and slotted for the web and a row of horizontal bolts connected to the wall?

Maybe use this method to connect to the top flange and restrain the bottom flange using the walers?

One further question. For a rafter supported at the bottom with an angle bracket can proper restraint be achieved from the purlins and/or flybraces?

Cheers!