It is frightening how poorly the engineers must calculate or alternatively, the low requirements, the arena in Vänersborg was inaugurated in June I believe, and now parts of the roof have collapsed, a construction costing 300 million. Many old buildings should have collapsed first; looking at the roofs that have collapsed (7-8) around us, it is "modern 80-90-00" buildings that have collapsed, and it was roughly then that engineers started to "calculate"
http://ttela.se/start/vanersborg/1.739824-taket-rasade-pa-arenan
http://ttela.se/start/vanersborg/1.739436-tak-rasade-pa-vargon-alloys
http://ttela.se/start/vanersborg/1.722592-tak-pa-bussgarage-rasade-in
http://ttela.se/start/vanersborg/1.729067-tak-pa-garagelanga-rasade
http://ttela.se/start/vanersborg/1.739824-taket-rasade-pa-arenan
http://ttela.se/start/vanersborg/1.739436-tak-rasade-pa-vargon-alloys
http://ttela.se/start/vanersborg/1.722592-tak-pa-bussgarage-rasade-in
http://ttela.se/start/vanersborg/1.729067-tak-pa-garagelanga-rasade
Last edited:
We usually have snow here in Dalarna, but shoveling roofs is probably quite uncommon.JanneL said:The most remarkable thing I think is that no one seems to have understood that you might have to shovel the roofs. Or don't you usually have snow down there?
Indeed. We have snow here in Ångermanland too. Good times, but when roofs start collapsing like they do down there, maybe one should consider what it depends on and take action.FredrikR said:We usually have snow here in Dalarna, but shoveling roofs is probably uncommon.
Last edited:
I think it's because they have completely different snow zones there.
And it's dimensioned weaker due to it.
Then it's not unthinkable that the snow that has fallen down south is wetter and heavier.
300kg/m2 is what our garage roof should withstand according to the roof truss drawings, that's a good pile of snow.
And it's dimensioned weaker due to it.
Then it's not unthinkable that the snow that has fallen down south is wetter and heavier.
300kg/m2 is what our garage roof should withstand according to the roof truss drawings, that's a good pile of snow.
I believe that in many cases people have built based on what they think looks like it usually does without calculating it, but for sports halls, it's likely that someone made a mistake during construction. You know how it sometimes goes; builders think it's over-dimensioned, and even if it says to bolt a beam with 4 bolts tightened to a certain torque, they might think two bolts are enough or they hand-tighten it to what they think should hold.
And they don't consider that there might be a reason why something should be tightened in a certain way with the correct number of bolts. I really think that someone might have wanted to get a little extra piecework bonus, thinking it's over-dimensioned anyway
When you calculate constructions, there are many safety factors. So, if you calculate it and the execution is correct, a roof in areas like Dalarna or northern Uppland should be calculated at 2.5kN/m2, and then certain factors come into play. First, µ depends on the roof pitch, and if we're calculating less than 30 degrees, it becomes 1.1 in some parts and 1.6 at the ridge, so on average, it's about 1.3. Then you add safety factors because it can be more than you calculate for, and for a regular roof resting on a roof beam/heart wall to the ridge, it's roughly like this (with regular lumber and assuming you don't know the quality, you should calculate at K12 (12MPa)).
(now I'm showing roughly)
2.5kN*1.3(depending on roof pitch)*1.3(safety factor) = 4.225kN/m2 in load
A 12MPa rule supports at least 95% of the wood in the class with a normal distribution curve, so many K12 can even handle over 20MPa, providing a pretty big margin. Then you add the m and n factor (1.25 for wood and 1.2 for the roof or upper floor, plus a time factor for wood and snow load at 0.75) -> (12*0.75)/(1.25*1.2)=6MPa to be used in calculations. Then the safety margins you MUST calculate with give ((4.225/6)/(2.5/12))~ 3.4 in the safety factor. So, if calculated by an engineer, it holds, and if the wood quality holds (not the 2-5% that fall below), and the load is only for a week, then the roof beam can withstand a snow load of 8kN/m2, which is approximately 3m packed snow or say 2m wet snow (but if it's wet snow, it slides off unless you have very flat roofs). With the new snow that has come here, it's about 5-6m of snow, and that thickness hasn't occurred.
But when calculating something in steel, you have more control and apply smaller safety factors, but even then, you dimension in the end for double what you have in the tables, which are quite generously assessed from the start.
Looking at the hall where the roof collapsed, it's a section that gave way, and I imagine it's because when they attached the beams between the arches, they didn't fasten them correctly. So when one gives way, those around follow suit since they get overloaded. If it's bolted or welded, I'm convinced it was done in a way not intended, and if one end fails, it breaks loose at the other end when the roof comes down.
Sure, people can miscalculate, but there are large margins to accommodate both some material faults and more snow than anticipated (what you calculate for should happen once every 50 years, and to be safe, you add a safety factor of 1.3, making it 30% more than the maximum should occur once every 50 years, plus additional safety for materials and the structure's location).
Oops, it got a bit long now, but I believe some have not calculated things correctly, or some build based on what looks good without calculating it. Where the calculations are correct, the execution has to be wrong for it not to hold :blushing:
And they don't consider that there might be a reason why something should be tightened in a certain way with the correct number of bolts. I really think that someone might have wanted to get a little extra piecework bonus, thinking it's over-dimensioned anyway
When you calculate constructions, there are many safety factors. So, if you calculate it and the execution is correct, a roof in areas like Dalarna or northern Uppland should be calculated at 2.5kN/m2, and then certain factors come into play. First, µ depends on the roof pitch, and if we're calculating less than 30 degrees, it becomes 1.1 in some parts and 1.6 at the ridge, so on average, it's about 1.3. Then you add safety factors because it can be more than you calculate for, and for a regular roof resting on a roof beam/heart wall to the ridge, it's roughly like this (with regular lumber and assuming you don't know the quality, you should calculate at K12 (12MPa)).
(now I'm showing roughly)
2.5kN*1.3(depending on roof pitch)*1.3(safety factor) = 4.225kN/m2 in load
A 12MPa rule supports at least 95% of the wood in the class with a normal distribution curve, so many K12 can even handle over 20MPa, providing a pretty big margin. Then you add the m and n factor (1.25 for wood and 1.2 for the roof or upper floor, plus a time factor for wood and snow load at 0.75) -> (12*0.75)/(1.25*1.2)=6MPa to be used in calculations. Then the safety margins you MUST calculate with give ((4.225/6)/(2.5/12))~ 3.4 in the safety factor. So, if calculated by an engineer, it holds, and if the wood quality holds (not the 2-5% that fall below), and the load is only for a week, then the roof beam can withstand a snow load of 8kN/m2, which is approximately 3m packed snow or say 2m wet snow (but if it's wet snow, it slides off unless you have very flat roofs). With the new snow that has come here, it's about 5-6m of snow, and that thickness hasn't occurred.
But when calculating something in steel, you have more control and apply smaller safety factors, but even then, you dimension in the end for double what you have in the tables, which are quite generously assessed from the start.
Looking at the hall where the roof collapsed, it's a section that gave way, and I imagine it's because when they attached the beams between the arches, they didn't fasten them correctly. So when one gives way, those around follow suit since they get overloaded. If it's bolted or welded, I'm convinced it was done in a way not intended, and if one end fails, it breaks loose at the other end when the roof comes down.
Sure, people can miscalculate, but there are large margins to accommodate both some material faults and more snow than anticipated (what you calculate for should happen once every 50 years, and to be safe, you add a safety factor of 1.3, making it 30% more than the maximum should occur once every 50 years, plus additional safety for materials and the structure's location).
Oops, it got a bit long now, but I believe some have not calculated things correctly, or some build based on what looks good without calculating it. Where the calculations are correct, the execution has to be wrong for it not to hold :blushing: