Shift Towards ULS from ASD for Foundations

[GalileoG]

I had originally posted this thread on the Foundation Engineering forum but did not receive any responses. I had deleted that thread and I am posting this here. I hope it is within the technical confines of this forum.

Here in Canada, there is a strong shift from ASD design towards ULS/SLS design for foundations. Almost all geotechnical reports that I have been receiving lately do not provide any ASD numbers. I really like this move, as the substructure design is now more consistent with the superstructure design, ASD design of the latter having been discontinued many decades ago.

I'm currently updating a spreadsheet that I have developed for spread footings to allow for the ULS/SLS design method. I do have several questions I would like to pose though - would really appreciate guidance on these matters:

1. When a geotechnical report states "ULS Bearing Pressure = 100 kPa" - does that already include the 'Resistance Factors for Shallow Foundations' shown on Table K-1 of the NBCC Structural Commentaries 2010? Table K-1 gives a resistance factor of 0.5 for "Shallow Foundation - Vertical Resistance By Semi-Empirical Analysis using Laboratory and In-Situ Test Data". Would I have to multiply the ULS value provided by 0.5 and check against my ULS load combinations?

2. When performing checks for uplift/overturning - I believe I should be using ULS load combinations now as well. What factor of safety should I be applying here? With ASD, the requirement was FS = 2 - does this change with ULS load combinations? Doesn't a FS defeat the purpose of ULS/SLS design? Can't seem to wrap my head around how to check for overturning/uplift using the new method.

3. It is often cited that it is bad practice to have any net uplift/tension on the footing. From what I understand, the rationale behind that recommendation is that the footing would be more prone to settlement when under constant tension. Does it not make sense then that this recommendations be applied to SLS combinations only as opposed to SLS and ULS?and not to ULS? In other words, I can design for net tension if I encounter it under ULS combinations, but should try to avoid net tension with SLS combinations? Does this rationale make any sense?

4. The SLS load combinations from the NBCC Structural Commentaries do not make any rational sense to me. These combinations are: 1) 1.0D + 0.5L + 0.5S, 2) 1.0L + 0.5S, 3) 1.0S + 0.5L, where 0.5L becomes 1.0L at storage areas. Do these not seem like odd service load combinations, perhaps slightly unconservative? Your thoughts?

I have two more questions that are unrelated to the ULS/ASD discussion, but I did not want to create new threads for them, these questions are:

5. One geotechnical report I have recently recieved states that I can neglect the weight of the pile when performing a gravity load check to determine pile length (this really helps, as sometimes the weight of the pile alone can be greater than the external load demand.) But can someone please explain why this is the case?

6. How effective is insulation (bead form) or a sonotube wrapped around a pile/pile cap in acting as a bond breaker against frost heaving? If a geotechnical report specifies a frost heave value of 65kPa acting over a certain depth, how much would this value be reduced to if insulation/sonotube is provided?

Thanks again in advance.

 

[JoshPlumSE]

Galileo -

I was disappointed that no one responded to that other thread. I was hoping to glean some knowledge out of any replies. Perhaps you are asking too many questions in one post. I can only respond with any real expertise on a few of them, but I will give it a shot.

1) I'm not a Canadian Engineer (I practice in the US). But, my belief is that the only way to determine if the resistance factors have been applied to is talk with the geotech directly.

2) I don't use the Canadian codes, so I'm not sure. However, the US codes used to require a 1.5 safety factor for overturning (service level), but then eliminated this when they changed the load combinations to 0.6D + 1.0 WL.... effectively building the safety factor directly into the load combination. It's possible that the Canadian guys did the same. But, we need someone with more Canadian design experience to reply definitively.

3) Your language on this question is imprecise. I tend to design foundations for full bearing contact for all "standard operation" loads (dead load, live load) et cetera. But, I allow partial uplift for seismic and wind.

 

[DamsInc]

I would also appreciate some clarification on this issue.

Regarding #2, I once asked my boss the same question. As a fresh university graduate in Canada I had no exposure at all to ASD, and safety factors were never used in our design classes. When I commented one day that it didn't make sense that we would include a safety factor after using the material and load factors he looked at me like I was insane "of course you use a factor of safety!" He had not been doing much design work in recent years and I was very skeptical of his understanding of the differences between the two systems.

 

[hawkaz]

The ULS method sounds like another way to turn a 2 line hand calc into a four page computer printout.

Why do we persist in taking the simplest of tasks and complicating them to absurd points. Practicing engineers need to get involved in the code writing process. We need balance to the academicians that are un-necessarily complicating this.

 

[SkiisAndBikes]

Galileo,

Here is my understanding of the Limit States Design (LSD), as related to the ULS and SLS numbers being provided by the geotechs. I would confirm any information with your geotech.

Essentially the shift is to bring the geotech parameters in line with the LSD methodology used for steel, concrete, wood design, etc... The relative resistances of the structural materials get factored down, the loads get factored up.

The geotechnical resistance at the ULS should be developed using the unfactored parameters of the soil. In the bridge code this is sometimes also refered to as the ultimate geotechnical resistance.

The factored geotechnical resistance shall be the ultimate geotechnical resistance multiplied by the relevant resistance factor, i.e. 0.5 for bearing. (akin to using resistance factors of 0.65 for concrete, 0.9 for steel, etc..)

Therefore with your example of being given a ULS of 100 kPa, I would be designing for a factored geotechnical resistance of 50 kPa, using factored structural loads.

With regards to your #4. The load combinations have been developed based on statistics. The code writers have decided that it is very unlikely you will experience full snow load conditions at the exact same time as full occupancy conditions. Keep in mind that these are minimum load combinations where past experience, and specific requirements from other related codes (i.e. concrete/steel/wood/masonry codes) may dictate more stringent load combinations.

Clear as mud?

 

[shobroco]

Okay, GalileoG, I am also Canadian & I'm going to give you my opinion, to be weighed by your engineering education & experience. Contrary to hawkaz's opinion, limit states is far more practical and logical, and less work, than working stress. I agree that the NBC requirements are ambiguous and contradictory on occasion, and that's where our education and experience comes into play. If we decide that our design meets the intent of the code, and the necessary life safety standards and serviceability standards, then we put our name on the design and our neck on the line. I've had a tough time getting straight answers from geotech guys many times, but it seems to be precisely because of what you are saying: they can't decide if they are giving us ultimate capacities for us to factor or working capacities that we end up re-factoring. I've found that you have to speak very slowly to them. You sound like you understand what you need to know, so try to make the guys supplying you with information be very clear.

For your two trailing queries, 5)could be because the pile weight is close (not equal) to the weight of soil displaced and 6) insulation is only effective if placed to eliminate frost, sonotubes may or may not be effective, but a double layer of poly and a smooth surface will eliminate any frost heave.

 

[SkiisAndBikes]

Galileo,

Re your #6. There is frost heave and frost adhesion, both different.

A double layer of poly or greased poly or smooth vertical surface can eliminate frost adhesion against a vertical surface, but will do NOTHING for frost heave.

Properly designed and detailed extruded polystyrene insulation (NOT bead board) can reduce or eliminate frost heave.

 

[Elite7]

+1 @ Hawkaz.

 

[Ron]

ASD, whether applied to wood, steel, aluminum, concrete or any other construction material is intuitive and straightforward. Using a limit state approach is neither; however, it became the system of choice in teaching long ago.

I agree with hawkaz.

Sorry Galileo that I'm offering nothing to your questions; however, I do agree with Josh on one aspect....ask the geotech what he used.

 

[SAIL3]

Agree wholeheartedly with hawkaz and Ron, however, it's too late now to unscramble this mess.
Something that originated in academia and sold as better representing the true behavior material/loads in engineering belies the fact that it's just rearranging load and resistance factors based on so-called enlightened statistics...
resulting in more opaqueness, more complicated and less useful/intuitive for the practicing engineer in the real world of engineering..
intuition is not a high priority in academia..more research/grants chasing a more empherial concept is..
we, as practicing engineers, are out of the loop and the results show this....
Sorry for kinda highjacking the thread, but this subject hits a raw nerve.....