Load Combinations for SLS 8

[BigBakwas77]

I am quite clear on all the load combinations at ULS as prescribed by National Building Code of Canada. But cannot find any information on what combinations to use at SLS. If it is sum of all Dead + Live + Snow + Wind loads multiplied by one, then it is more than many ULS combinations. Is this right?

 

[skeletron]

There are no SLS load combinations in NBCC 2015. There is talk of having them returned in 2020 or the following code cycle.

My understanding is that when Canada went to limit states design in the 80s, the serviceability combinations were removed to prevent designers using a quasi-allowable stress method for design. There are serviceability factors (0.9 for snow, 0.75 for wind) but the combinations have never been explicitly implemented beyond that.

You need to consider a number of issues with SLS and deflection. Typically it is the "live load" deflection which is relevant because it will have the biggest serviceability impact. This includes live load and snow loads. Wind loads are short term and dead load can usually be cambered out. Looking at those two loads independently is often best.

 

[Agent666]

To backup what Skeltron is noting, often as a designer you are interested in the deflection after something is built, some of the loads are built in and do not contribute the effects that might be perceived at the SLS. For example for a roof the deflection the users will see or feel due to wind is simply the wind load in isolation, how much it sagged under its own self weight is somewhat irrelevant as you are talking about relative deflection being noted vs absolute deflections as far as what is perceived by the observer. Same can apply to concrete slabs, for example if a composite slab is screeded level, then the user might only really sees the additional deflection due to applied SDL and LL. Therefore you need to understand the effect you are looking at for SLS and derive an appropriate load case or loading scenario for each effect you might be looking at, for example:-

1) is it a 'visual' thing like a column swaying over under wind or earthquake

2) is it a 'felt' thing like vibration of a floor
.
3) is it a prevention of damage or loss of serviceability thing (for example too much deflection might cause a serviceability issue with respect to watertightness, too wide a cracks in reinforced concrete might cause a durability issue).

Similarly the limits to apply to each effect may also differ, for example different deflection limits apply often to different loading, for example short or long term deflection limits.

AISC design guide 3 has a number of tables such as outlined below that shows some appropriate limits and also the associated loading to consider. Many standards and codes around the world have similar guidance which is equally applicable to other regions given all buildings share similar construction forms and techniques.

Edit, due to the lack of loadfactors over 1.0, the SLS load should always be less than the ULS load.

 

[BigBakwas77]

Thank you for your valued opinions!

 

[amecENG]

There are SLS load combinations in the NBCC 2015 Commentary A Clause 27 and 28.

For differential settlement of foundations, for example, the load combination is given as:
Dead + 0.5Live + 0.2Snow

 

[ClearCalcs]

If you're designing with wood specifically, CSA O86-14 has 4 load combinations for SLS in Table 5.2.4.2. This is accompanied with a limit on deflection of L/180 in Cl 5.4.2, which is rather generous so some judgement may be required here. Note that all the load combinations include dead load. There's also a limit of L/360 on deflection from long term loads to limit creep.

After a quick look through O86-19 it seems that the load combinations have been removed in favour of "look in your building code", so these may not be applicable anymore. In Ontario at least, the OBC still references O86-14.

-Laurent

 

[r13]

I often hear people argue the inherent shortcoming of ULS is safe guarded/balanced by SLS, surprised to see they are decoupled.

 

[IDS]

I often hear people argue the inherent shortcoming of ULS is safe guarded/balanced by SLS, surprised to see they are decoupled.

I don't know the Canadian code, but from the discussion above it sounds like it does have SLS requirements, even if they have chosen not to call them that for some reason.

The SLS is just as much a part of limit state design as the ULS, and in my opinion it should be recognised as such in all limit state codes. That's why I dislike the term LRFD, which implies that only ultimate load conditions need to be considered.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[r13]

IDS,

Agreed.

 

[Settingsun]

It's common for SLS requirements not to be as tightly specified as ULS. ULS is about safety so wriggle room should be limited. SLS problems can often be rectified or simply lived with. The cost of meeting rigid SLS criteria might not be warranted so it's left up to judgment.

 

[Settingsun]

Serviceability checks are still LRFD but many of the factors will be 1.0.

On the first post in this topic, ULS loads are approx upper bound loads or even overestimates so SLS will never exceed ULS. A guide in absence of code requirement would be to pick ~2/3 of distributed live load (plus dead if interested in total deflection); or ~2/3 of a ULS environmental action plus ~1/3 of live load. Adjust as common sense dictates, eg if environmental conditions won't result in reduced live loading due to nature of structure's use.