Wind: ASCE 7-10 versus 7-05 2

[TomBarsh]

ASCE 7-10 introduced a new load combination reduction factor of 0.6 for wind in paragraph 2.4 (ASD) versus a factor of "1.0" used in 7-05.

The Commentary provides some explanation of the basis for this change. But the end result seems to be that for a nonbuilding structure designed to paragraph 29.5 the base shear is reduced by 40% from same structure analyzed under 7-05 rules. I wonder if this was fully intended by the Committee. Any comments? Thank you.

 

[Hacksaw82]

WillisV - well yes and no. Under 2005 ASD I applied 90mph, factored it by 1.0 and backed down the material stress to make it 'safe'.

Your description works for LRFD only.

It used to be rational. We'd factor up loads for LRFD, and factor down to allowable stresses for ASD. Now with the new wind code, we factor up the wind loads for ASD, and factor down the wind loads for LRFD.

Luckily, we have all these fancy analysis programs to calculate this all for us in a black box, so we don't have to be concerned about actually understanding what's going on.

 

[TXStructural]

Wait!
The 2010 winds are ACTUAL, maximum winds for a certain return interval. The interval is different for different categories of structure. A load factor is not applied since there is no upward increase due to variability - these are maximums. Floor loads are factored upward because of potential variation in actual loads, NOT for the purpose of providing a factor of safety for anything except load variation.

If the structure is designed for a 2010 wind, it is designed for that wind. If you use a lower value, you get a weaker-than-design-strength structure (or cladding, or whatever.) If the cladding does not need to resist the full design wind for some reason, and the code allows, then design for less.

The 2005 winds are nominal winds, which could be expected to be exceeded at some interval, which is why they were factored up.

 

[WillisV]

@TXStructural - yep.

 

[dcarr82775]

I am with TXS and WillisV.

I can not grasp what people are complaining about. The equations are basically the same, you just use a higher V. If you want service loads you multiply by 0.6. Geez.

 

[Hacksaw82]

Heck, I understand where all the factors have moved, and how to use the new method. I'm just asking why the change was made? What was wrong with what we were doing before?

In most buildings (in the mid-west, where I practice) the lateral drift from wind will control. Unlike what dcarr said above, to get service loads, you actually need to use the 'service' wind maps from the appendix. So now, we have four wind maps for each building category, and four more for the deflections of each building category. Eight wind maps, and new load combinations too.

I know the loads didn't change (in most cases), and again, I get the method and how to implement it. I'm just concerned we are adding complexity, and therefore the chance of errors.

Does anyone have any literature on what facilitated this change? I'm trying to get the old grey hairs at the office to buy into the code changes, and I'm having trouble making them see the more than just change for the sake of change.

I was lucky enough to take a wind course in grad school by Peter Irwin (principal of RWDI and major contributer to the wind provisions). I didn't get the impression from him the 2005 code was a terrible way to do things. Why did the code committee make this change?

 

[TXStructural]

The publishers of the code books like this kind of change ($$), but there is surprisingly little input by those publishers into how and what gets published. The main purpose of the change was to make all lateral forces similar, in that now wind and seismic are based on the actual maximum credible event loading. Both are based on a "reasonable" return interval, based on experience, statistics, and the best guess of people whose job it is to study and estimate such things.

Since we design strength for maximums, it makes sense to use a maximum value. The change to maximum winds values allows easy maintenance of these values as experience increases and modeling improves. Any use of factors to reduce these values is arbitrary, which should lead us to publishing a "service level wind" map instead. This would give us one more map which would be essentially the 2005 wind map, but it could be based on actual winds with a 50 or 100 year interval, like floods.

The use of multiple maps is a more technically correct method than an arbitrary multiplier. The map for more "important" buildings is simply for a longer timeline, meaning that the probability of the same winds is greater and a higher wind is possible. The philosophy behind the current wind levels is that we go with the highest winds during the interval. Whether the methods to get these numbers is valid, or whether doing so is truly needed, are entirely different issues. I think we could really benefit from agreeing that designing strength for all buildings for winds with a 700 or 1700 year return interval is adequate, and it needn't vary by category of occupancy.

One of the things that explicitly weighs in to ICC code changes is "will this change increase the cost of construction". The way this change was done allowed them to answer "No", whereas making a wholesale change to the design level loads would have either made structures weaker or made them more expensive.