Load Factors

[HankTexas]

Several years ago I did quite a bit of research for a presentation on wind loading. During that research I discovered that the probability of a 50 year storm occurring over a period of a 50 year structure lifetime was about 66%. This was considered unacceptable but it was understood that the load factors increased the frequency to around a 700 year storm which had around 5% chance of occurring over a period of 50 years. Before this research I was of the opinion that load factors were used as a buffer against mistakes in calculations. Although I would have never used load factors to compromise a design, I probably even mentioned the load factors as safety factors when discussing a structure’s loading. After researching the subject I had a completely different opinion towards the subject. It seems there are no ‘safety factors’ against design, just a 5% chance the wind loading may actually exceed the capacity of the structure over a 50 year period. Does anyone have any thoughts about this that I might be missing?

 

[JAE]

Are you talking about load factors as required in the US Building codes and ASCE 7?

If so - those load factors are there to provide a level of safety against the variability in the loads.
The higher the variability, the higher the load factor.
So live load uses, currently, 1.6, while dead load is at 1.2.

The safety factor against variations in capacity is the Φ factor.
This deals with variations in things like concrete strength, rebar depth, steel yield, member geometry, etc.
It also deals a bit with the accuracy of the formulations that estimate that strength.

The Φ factor, together with the load factors combine to provide an overall design safety factor on the calculated strength vs. loads, which is usually calibrated back to historically used levels of safety and historical levels of probabilities of failure.

For your wind derivation - choosing a type of storm and a probability of failure is almost a separate task in developing a model wind magnitude to design to. Others here might want to weigh in if they know how the various wind magnitudes were derived.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[canwesteng]

Load factors aren't for mistakes - they are there to account for the probability the load is exceeded. Combined with resistance (the probability of material being understrength), you get an acceptable probability of failure.

 

[MrHershey]

I don't see the safety factor and statistics concepts as incongruous.

Commonly considered building service life is 50-100 years. For a 700 year storm that means that in the building's 50-100 year service life there's a 7%-15% chance it'll ever see the design load. At least in my eye, that's a safety factor. It's not a guarantee, nothing in structural engineering is. But if you've designed things correctly then your strength is sufficiently high that it's fairly unlikely you'll exceed it at any time during the building's service life. If you want a higher 'safety factor' you could design for a longer return period storm (Risk Category III and IV buildings use a 1700 year return). If you're doing a barn and can deal with a lower 'safety factor' you could design for a shorter return period storm (Risk Category I buildings use a 300 year return).

And that's before you get into the phi factors of the material design codes on the capacity side.

Though will say there aren't safety factors against mistakes in calculations. Mistakes may get covered up because the structure never actually sees the load it was supposed to be designed for, but they're not intended to cover up calculation mistakes.

 

[SAIL3]

with climate change and the recent ever-increasing intensity of the storms, the probability of a structure experiencing the design wind loads in it's lifetime is increasing and should be taken into consideration....the codes are a compromise between economics, public safety and engineering....

 

[JStephen]

Sail3- that's something that should be taken into account in the generation of the wind speed maps, not tacked on afterwards. (Note that I said "should", not "is").
And by similar reasoning, I should be able to reduce snow loads and increase the design metal temperature for my projects as well.