Roof Trusses with Huge RC column supports 2

[mes7a]

There are existing huge RC columns above the roof (0.5 x 0.5 meter size). Instead of another floor. We will just put trusses to support a lightweight plastic or metal roof. There are 4 bolts of about 14mm at the corner of each column (see pictyure picture). If baseplate would be attached to it with the I-beams welded to it. What would be the problem if there is very strong wind or hurricane. Would the bolts just break? I read that in lightweight roof trusses and I-beams.. the support needs to be flexible so all the forces won't be concentrated on the 4 bolts and the support would just bend. What connections would I need to do?

 

[BAretired]

Hollow block walls are definitely qualified as firewalls in Canada. Vertical cells are reinforced and filled with concrete as required. If you place a bond beam at the 3m level and tie into your roof diaphragm, you will be satisfying the 3m maximum unsupported height requirement. The parapet will simply act as a cantilever carrying wind load and the weight of parapet will be supported by the wall below.

If you maintain your present roof framing arrangement, you don't need an edge beam other than a typical purlin to carry a half panel of roof deck. It would be normal practice to bolt the edge purlin to the bond beam at about 1200mm o/c.


BA

 

[mes7a]

Hollow block walls are definitely qualified as firewalls in Canada. Vertical cells are reinforced and filled with concrete as required. If you place a bond beam at the 3m level and tie into your roof diaphragm, you will be satisfying the 3m maximum unsupported height requirement. The parapet will simply act as a cantilever carrying wind load and the weight of parapet will be supported by the wall below.

The reason a bond beam at the 3m level can't be tied into the roof diagphragm is because there must be spacing between the bond beam and roof diaphragm to serve as rain gutter as the following architectural detail shows.

If you maintain your present roof framing arrangement, you don't need an edge beam other than a typical purlin to carry a half panel of roof deck. It would be normal practice to bolt the edge purlin to the bond beam at about 1200mm o/c.

If you will see in the roof framing plan I posted prior which i'll reproduced below (unchanged):

You will see that the columns at the lower end are not aligned but are diagonal. This is why the edge beam has to be able to carry the rafters at least at the lower part of the picture. A purlin may not take the load.. unless you mean it can take the rafter load at the left and right side where the rafter frames into the edge beam (which you suggested a purlin is enough)?

Also our firewall is only 6 inches in thickness.. so how to insert a steel inside. unless you mean very small purlin.. because the original HSS perimeter beam is also 6" being as stated before 250x150mm x 6mm thick.

Anyway. My architect said the HSS rafter at size 250mm x 100mm x 8mm thick is expensive because it is 8mm thickness and special order and suggest what if trusses instead. In your place. Is the mentioned HSS size really special order and expensive.. would trusses be cheaper instead of the rafter in my framing plan? Again noting the actual metal roof won't be touching the perimeter beam because of the rain gutter so the metal roof (which you call roof diaphragm) is just supported entirely by the rafter and purlins crossing it perpendicularly (as the framing plan above shows).

Thank you a lot.

 

[BAretired]

The reason a bond beam at the 3m level can't be tied into the roof diagphragm is because there must be spacing between the bond beam and roof diaphragm to serve as rain gutter as the following architectural detail shows.

Maybe you could use the rain gutter as a structural element to provide lateral support to the wall. Discuss with architect.

You will need an edge beam because of the offset columns at Y3 and Y5. It may be worth looking at the possibility of spacing your rafters at 2000 o/c and spanning steel deck parallel to the ridge beam. That uses more rafters but saves the cost of purlins.

HSS 250 x 150 is a standard size in Canada. HSS 250 x 100 is not standard. It is usually better to stick to shapes which are listed in the steel handbook.

BA

 

[mes7a]

Maybe you could use the rain gutter as a structural element to provide lateral support to the wall. Discuss with architect.

You will need an edge beam because of the offset columns at Y3 and Y5. It may be worth looking at the possibility of spacing your rafters at 2000 o/c and spanning steel deck parallel to the ridge beam. That uses more rafters but saves the cost of purlins.

HSS 250 x 150 is a standard size in Canada. HSS 250 x 100 is not standard. It is usually better to stick to shapes which are listed in the steel handbook.

Our architect said HSS with thick 8mm are more expensive than I-beam. Is it true in your place HSS are generally more expensive than I-beam for the same sections or strength?

My colleague and I will redesign everything as I-beam (rafters and perimeter beam). And we will put the perimeter beams on top of the columns.. what my colleague suggested is welding the dowels of the parapet reinforcement to the I-beam perimeter beam itself. What is your comment and experience about welding 10mm dowel/rebars to the I-beam vertical to serve as parapet hollow block inner reinforcment?

 

[hokie66]

I will make a few comments, as this roof structure is like Australian practice in some respects.

1) Yes, hollow sections when used as beams are generally more expensive than wide flange sections.

2) I assume the concrete columns are cantilevered, and you are not using a roof diaphragm. If your roofing is the metal sheeting, its purpose is to shed water, not to serve as a diaphragm.

3) Using concrete block for a parapet atop a steel beam makes no sense. You could use a fascia truss, fabricated from relatively light members, and clad it with a compressed fibre cement product to achieve the fire rating you want. Support the fascia truss on upstand brackets off the rafters. The fascia beams on X1 and X3 are then superfluous, except where you have the offsets at Y3 and Y5.

4) As you mentioned seismic, projecting elements like masonry parapets are one of the main offenders in loss of life. If you must have a reinforced masonry parapet, suggest you provide a concrete beam for it to start on, or else extend the reinforced masonry wall from below, and brace it at the roof level.

Not solving all your connection problems, just giving a few suggestions as to the way we would do it in Australia...

 

[mes7a]

I will make a few comments, as this roof structure is like Australian practice in some respects.

1) Yes, hollow sections when used as beams are generally more expensive than wide flange sections.

2) I assume the concrete columns are cantilevered, and you are not using a roof diaphragm. If your roofing is the metal sheeting, its purpose is to shed water, not to serve as a diaphragm.

3) Using concrete block for a parapet atop a steel beam makes no sense. You could use a fascia truss, fabricated from relatively light members, and clad it with a compressed fibre cement product to achieve the fire rating you want. Support the fascia truss on upstand brackets off the rafters. The fascia beams on X1 and X3 are then superfluous, except where you have the offsets at Y3 and Y5.

4) As you mentioned seismic, projecting elements like masonry parapets are one of the main offenders in loss of life. If you must have a reinforced masonry parapet, suggest you provide a concrete beam for it to start on, or else extend the reinforced masonry wall from below, and brace it at the roof level.

Not solving all your connection problems, just giving a few suggestions as to the way we would do it in Australia...

Yes, the concrete columns are extended from foundation without any splice and size reduction. Yes, the light roofing purpose is to shed water.. not serve as a diaphragm. You have a point that a firewall parapet standing on exposed I-beam doesn't make sense. I can imagine the I-beam melting and the firewall parapet falling inside, yikes! So we'll design either the fascia truss.. or extend the wall from below and brace it at roof level. Brilliant idea.. yours and BaRetired. But still we need one meter or so of perimeter steel beam to support the rafters at the offset at Y3 and Y5.. If not steel beam.. how do we support the rafters at the offset? Thanks.

 

[hokie66]

At Y3, you could just make the fascia truss on grid X1 strong enough to pick up the rafter. Then, at Y5, you can again either extend the whole truss to the column, or provide a beam fabricated within the truss which would extend to the column. Either way, as this member would extend beyond the roof, you would have to enclose it with cladding. When columns don't line up, you have to be a bit inventive.

 

[mes7a]

At Y3, you could just make the fascia truss on grid X1 strong enough to pick up the rafter. Then, at Y5, you can again either extend the whole truss to the column, or provide a beam fabricated within the truss which would extend to the column. Either way, as this member would extend beyond the roof, you would have to enclose it with cladding. When columns don't line up, you have to be a bit inventive.

Do you have example of any fascia truss that is very strong? That you see in books, references or pictures. This is better than lateral support because the latter would extend the roof sheeting higher to brace the wall (if brace is below roof sheets).. I can't imagine extending the brace outside or above the roof sheets because this would create many leaks later on with many flashings to waterproof the brace extending outside/above the roof sheets.

 

[hokie66]

No, you don't brace the fascia truss. The chords, top and bottom, are typically channels, web horizontal, sometimes hot rolled, sometimes cold formed, depending on the situation. The chords span between the upstands which are cantilevered from the rafter ends.

 

[mes7a]

No, you don't brace the fascia truss. The chords, top and bottom, are typically channels, web horizontal, sometimes hot rolled, sometimes cold formed, depending on the situation. The chords span between the upstands which are cantilevered from the rafter ends.

I meant I know either to use fascia truss or brace it. We would choose fascia truss instead of bracing it. Anyway. We will calculate the fascia trusses and whether it can be tiny enough to have sufficient concrete cover. If not sufficient. Maybe better not to have a firewall parapet at all than to have one that would topple over in the event of fire (from exposed steel beam).. lol.. A lot of thanks to all the people who helped here.

 

[hokie66]

I don't know what you mean by "sufficient concrete cover" on a fascia truss. There is no concrete on a fascia truss.

 

[mes7a]

I don't know what you mean by "sufficient concrete cover" on a fascia truss. There is no concrete on a fascia truss.

Oh. I don't know where to buy "compressed fibre cement" to clad it with as you stated.. we don't have it locally.. so I plan to enclose it with sufficient concrete cover to act as fire protection. Remember in beams and columns, the rebars are protective by concrete cover. Won't this be sufficient??

Because if not sufficient (by empirical computations). We may have to ditch the whole parapet firewall. Because it can be more dangerous if it falls over from melted beam considering what's below are just thin roof sheeting for water to glow over gutter.

 

[mes7a]

I can't remember ever using steel beams to support concrete walls so I can't answer your question. If you need a parapet above the steel roof to serve as a continuation of the firewall, it would seem to make sense to simply continue the lower wall to the top of the parapet. The wall would need to be laterally braced at the roof level and I'm not sure you can achieve that with your lightweight plastic or metal roof.

Rafters may be connected to "something flexible" but there is no mandatory requirement to do so that I am aware of.. If the rafters are connected to the edge beam, they could be sized to take the wind moment from the edge beam but that cannot be done if you attach the edge beam to the face of the concrete column.

The idea of breaking the wall to accommodate an HSS is simply the wrong concept in my opinion. If the purpose of the edge beam is to provide lateral support to the wall, then put it inside the wall so that the wall can pass through to the top of parapet uninterrupted. The HSS would be carrying horizontal load from wind but the parapet would bear on the wall below and the steel beam would carry no gravity load.

BAretired. I've been reading your messages over and over again the past 2 days.. about 30 times. I thought when you mentioned to put the beam inside the wall. I thought you were referring to within the wall.. that was why I was asking the beam is same size as the wall (how could the beam be inside the wall). But I realized today you were referring "inside the wall" to the other side not outside. I think my last resort is to use your solution before ditching the whole firewall idea. Because I don't have the concrete fibre cement Hokie required to clad the fascio trusses. I'd like to ask the following which I don't quite get. You said " If the rafters are connected to the edge beam, they could be sized to take the wind moment from the edge beam but that cannot be done if you attach the edge beam to the face of the concrete column". Are you saying the edge beam would not be connected to the column? Or if it would be connected at top of column. It's the same as your "that cannot be done if you attach the edge beam to the face of the concrete column".. because if welded on top of column. The edge beam can't transfer the moment to the rafter.

If you were mentioning about edge beam hanging with the wall itself. Of course the rafters can't be rested on it. When you said "they could be sized to take the wind moment from the edge beam" (I read this over 35 times already). You meant the parapet would rotate ("wind moment")? transferring this to the beam.. but the only way for this to happen would be if the parapet rested on the beam. But if it is just beside the parapet.. how can it transfer wind moment (making the beam rotate)? You also mentioned to extend the rafter up the wall as lateral support (separate idea to all this). I'd just like to explore your idea of the rafter being sized to take the wind moment from the edge beam. Just please clarify what you meant by it (like whether the beam should be welded on top of column) before I'll share the idea to our team of 5 structural designers tomorrow who will compute the members sizes and dynamics of it all. This is our last resort. Thank you.

 

[mes7a]

to continue what I was saying BaRetired.. we must not just consider wind lateral loads (which you call wind moment) but seismic loads which can sway the walls back and forth. This is the main problem if you will make it high all the way to parapet. During strong seismic movement where the ground shakes, the wall would sway to left and right.. so consider this.. noting the hollow blocks are also just weaks. Even if you can extend rafter up to brace it. The bolts may give way when seismic shaking goes to the right pulling out the connections. This is the exact reason why we need maximum wall height of 3 meters and putting stiffer below and above to avoid the walls from toppling down in seismic. So please reconsider your ideas using seismic left and right shaking to see if the parapet is still feasible. If not we will just ditch it. You are our last hope towards it. Thank you.

 

[BAretired]

I don't know why you need a firewall, but if you need a firewall, it has to be either solid concrete or masonry. Steel members covered with cladding will not give you a firewall, at least not by the definition in my code.

As for seismic events, there are design provisions in the code to cover them. I tend to agree that masonry is not an ideal material if serious seismic shaking is likely to occur. Fortunately for us, we do not have that problem in Alberta so I am not too familiar with seismic design.

If the architect is calling for a masonry firewall, then it has to be continuous from bottom to top. It cannot be interrupted by a steel beam. The wall must be laterally braced at the 3m elevation. The parapet must be cantilevered from the wall below or otherwise braced to the roof structure, possibly using diagonal braces from the top of parapet to the rafters.



BA

 

[mes7a]

I don't know why you need a firewall, but if you need a firewall, it has to be either solid concrete or masonry. Steel members covered with cladding will not give you a firewall, at least not by the definition in my code.

As for seismic events, there are design provisions in the code to cover them. I tend to agree that masonry is not an ideal material if serious seismic shaking is likely to occur. Fortunately for us, we do not have that problem in Alberta so I am not too familiar with seismic design.

If the architect is calling for a masonry firewall, then it has to be continuous from bottom to top. It cannot be interrupted by a steel beam. The wall must be laterally braced at the 3m elevation. The parapet must be cantilevered from the wall below or otherwise braced to the roof structure, possibly using diagonal braces from the top of parapet to the rafters.

It is difficult to laterally brace the wall at the 3m elevation. You are only consider wind energy going into the wall or building. What about seismic energy that can pull it to fall on the neighbor. The brace are just attached to hollow block. the block can break especially after certain seismic cyclic loading. This is the reason we always confined them inside the columns and beams (bottom and top). So even without parapet, we will need to put a perimeter beam to hold or confine the 3 meter wall below.

Anyway. I think we'll just ditch the whole firewall thing.

 

[BAretired]

It is not difficult to brace the wall at the 3m elevation. Use an edge beam flush with the inside face of wall designed to resist the lateral force from the wall caused by wind or seismic action. Drainage of the rain gutter must be done without unduly weakening the edge beam.

Ditching the whole firewall thing is not your decision to make. It is the Architect's decision. Why does he want a firewall? If the building is on or near a property line, a firewall may be a code requirement.

BA

 

[mes7a]

It is not difficult to brace the wall at the 3m elevation. Use an edge beam flush with the inside face of wall designed to resist the lateral force from the wall caused by wind or seismic action. Drainage of the rain gutter must be done without unduly weakening the edge beam.

Ditching the whole firewall thing is not your decision to make. It is the Architect's decision. Why does he want a firewall? If the building is on or near a property line, a firewall may be a code requirement.

The building is on the property line with the neighbor also on property line (almost touching each other). I'm trying to convince the architect the parapet may be unstable. The hanging column thing is unique situations that other structural engineers in the office are running out of idea. Anyway. We'll analyze your bracing idea. When you said the following which I have read 50 times already "If the rafters are connected to the edge beam, they could be sized to take the wind moment from the edge beam but that cannot be done if you attach the edge beam to the face of the concrete column."

1) you said "they could be sized"... which "they" are you referring to.. the rafters or edge beam?

2) What do you mean by wind moment? how can wind have moment unless you are talking about the lateral pressure caused by the wind? Can you call it a moment? Or are you talking about torsion? But this torsion is a problem we have thought of if the parapet were welded on top of the perimeter beam as we initially envisioned. Again.. pls clarify meaning of "wind moment".

3) Are you talking about putting 5 rafters instead of 3? Because with only 3 connecting at columns. I can't see how the edge beam can affect the rafters. But 3 rafters would be sufficient. If so.. would it able to take the wind moment from the edge beam.. by whatever wind moment you meant.

4) Are you talking of putting rafter on top of edge beam.. or edge beam on top of rafter?

 

[BAretired]

The building is on the property line with the neighbor also on property line (almost touching each other). I'm trying to convince the architect the parapet may be unstable. The hanging column thing is unique situations that other structural engineers in the office are running out of idea. Anyway. We'll analyze your bracing idea. When you said the following which I have read 50 times already "If the rafters are connected to the edge beam, they could be sized to take the wind moment from the edge beam but that cannot be done if you attach the edge beam to the face of the concrete column."

I was referring to the situation where you were bearing the parapet on top of the edge beam, a concept which I do not recommend for reasons stated earlier. In that case, lateral forces from wind or seismic activity could only be carried by torsion in the edge beam. The rafters could be sized to take that torsional moment if they were properly connected to the edge beam. The above comment was made in response to a question which you had asked, but it is not my recommendation, so please ignore it.

1) you said "they could be sized"... which "they" are you referring to.. the rafters or edge beam? The edge beams could be sized to take the torsion and the rafters could be sized to take the torsional moment from the edge beam.

2) What do you mean by wind moment? how can wind have moment unless you are talking about the lateral pressure caused by the wind? Can you call it a moment? Or are you talking about torsion? But this torsion is a problem we have thought of if the parapet were welded on top of the perimeter beam as we initially envisioned. Again.. pls clarify meaning of "wind moment".

Wind does not have moment. Wind moment refers to the moment in a member caused by wind. If the parapet is of height h and the pressure is w, the wind moment per unit length of edge beam is wh²/2. If the rafters are spaced at s, the bending moment in the rafter is swh²/2.

3) Are you talking about putting 5 rafters instead of 3? Because with only 3 connecting at columns. I can't see how the edge beam can affect the rafters. But 3 rafters would be sufficient. If so.. would it able to take the wind moment from the edge beam.. by whatever wind moment you meant.

I was thinking of spacing the rafters at 2000 o/c, i.e. at the third points of the 6m span. The rafters could bear on top of the edge beam and the edge beam would be providing lateral support to the firewall using short stubs to fill the gap required for the roof drain.

4) Are you talking of putting rafter on top of edge beam.. or edge beam on top of rafter? See (3) above.

BA

 

[mes7a]

I was thinking of spacing the rafters at 2000 o/c, i.e. at the third points of the 6m span. The rafters could bear on top of the edge beam and the edge beam would be providing lateral support to the firewall using short stubs to fill the gap required for the roof drain.

One rafter is so expensive.. we would retain only 3 spaced at 6000 o/c instead of 9 at 2000 o/c. So if we will retain 3. Is what you describe still valid where we will use edge beam on the wall resting on top of column. Then the rafter either at level or on top of edge beam at column? Why should it be on top if we will just use the column (except at the offset at front)

I know about extending the rafter to the vertical side of the wall at the column part and even diagonally brace it at the rafter center.

Use an edge beam flush with the inside face of wall designed to resist the lateral force from the wall caused by wind or seismic action. Drainage of the rain gutter must be done without unduly weakening the edge beam.

Must the edge beam be a wide flange or HSS (or what material) to flush with the inside face of wall. So I imagine the rebars/dowels inside the wall welding to part of the edge beam which is located on the surface.. then this would be welded to the 3 columns (I know the parapet would bear on existing wall and not on top of edge beam). Is connecting the edge beam on the 3 columns top enough. Meaning the rafter and edge beams won't have torsional effect or transfer on each other. The rafter only providing diagonal support on top of the wall... so how rafter and edge beam are connected doesn't matter. Thanks so much.