Dynamic Settlement 2

[Althalus]

I've received a client request that I've never been asked to look at before. A fellow engineer said that it was "just a number" and didn't really represent anything in the real world.

CLIENT REQUEST:

Provide calculations for soil settlement due to dynamic load conditions.

I'm going to guess that the (client) theory was that the short-term loading (wind/seismic) will cause the soil to compress like a spring and cause some permanent deformation as well as some temporary deformation which will spring back. And we need to calculate that portion that does not spring back.

Is this a thing? I had always thought that whatever permanent effects that dynamic loads had on the soil was taken up in the long-term settlement already.

Any thoughts?

 

[Greenalleycat]

Feels like we are missing some important context

Here, the most common method to calculate soil bearing capacity implicitly references settlement as it's based off testing that relates to 25mm of ultimate settlement
A factor of safety of 2 is applied for ultimate loads and 3 for serviceability which somewhat arbitrarily limits you to the 5-15mm settlement range
'Dynamic' load cases (wind/EQ) overturning + gravity bearing pressures are compared against these reduced capacities as per standard structural design theory
Huzzah, you have implicitly accounted for dynamic effects in your design

This may not meet your client's brief though
1/ It doesn't explicitly calculate the settlement expected - it just implicitly 'works' to within acceptable levels
2/ It doesn't provide any consideration for elastic vs plastic components of the settlement
3/ It doesn't consider effects from risky soils like peat
4/ It doesn't allow for tricky effects such as liquefaction expression or liquefaction-induced consolidation (or similar effect)

 

[geotechguy1]

There is liquefaction which can be triggered in coarse grained soils, there is cyclic softening which can be triggered in fine grained soils. Small cyclic loads without stress reversal can also accumulate strains resulting in settlement or possibly even triggering large strains if you have a strain softening soil. Cyclic loads can also cause some settlement in dry sands.

The thing is it depends, what are the cyclic loads, what are the soils, what are the settlement criteria.

 

[Althalus]

1/ It doesn't explicitly calculate the settlement expected - it just implicitly 'works' to within acceptable levels

I think this ^^ was what my coworker was talking about. Your wording is much more clear.

Thanks.

 

[TigerGuy]

It can also be due to equipment vibration

 

[Greenalleycat]

Hopefully this link works - this is the paper I'm referring to
Yes, the standard practice for settlement calculations is based on one dude doing some vague testing 50 years ago - but it seems to work
I'd actually misremembered this one - the curves are for a max of 7.5mm deflection, FOS 3 down to 2.5mm - not 7.5mm FOS up to ~25mm as I originally was thinking


[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1726605373/tips/Stockwell_-_Determination_of_Allowable_Bearing_Pressure_Under_Small_Structures_duz4d2.pdf[/url]

Edit: you may also find this paper, particularly Section 2, to be relevant to your disucssion

https://fl-nzgs-media.s3.amazonaws.com/uploads/2016/07/84_Dec2012_NZGeoNews.pdf

 

[geotechguy1]

I've only ever heard structural engineers talk about 25mm as a deflection at the ultimate bearing capacity. I'd expect it to be 5-10% of your foundation width at geotech ultimate. Almost all practical problems are controlled by settlement; conventionally 'allowable' bearing capacity (or ultimate / 3 or 4) was taken as a pressure where 25mm of settlement would occur. Sometimes it will still be presented that way but by determining the applied pressure at which that much settlement occurs rather than the ASD method of dividing by 3 or 4. Also has not really relevant to settlement induced by dynamic loads.

 

[Greenalleycat]

The Building Code here has a requirement to limit deflections to max 25mm over 6m under serviceability conditions
To be honest, that seems like a pretty high limit, I'd be quite miffed if my house was 25mm out, but that's what it says...

My 2c on these highly fudged methods - conventional residential construction has plenty of ground clearance so the absolute magnitude of settlement is not super important
Differential settlement is far more important
Modern foundations typically are so stiff and strong that, unless you're sitting on highly variable soils, the foundation should be able to redistribute the loads and constrain differential settlement
Also, the design loads under most foundations are really low - even soft soils can comfortably handle them - so the theoretical ground bearing is usually just a somewhat arbitrary number that forms a one-sentence check in the calcs
It is mainly critical under shallow piled foundations where the area of the footing will be closely tailored to your theoretical bearing capacity
The easiest thing to do here is just add 50mm to your footing or whatever - sure it's a bit of extra cost, but not that much, and it helps everyone sleep at night

I also am a fan of some form of compaction under the foundation prior to pouring - I don't know how you could numerically prove it, but I've always assumed that this takes out the worst of any immediate settlement in softer soils

 

[dik]

It's the differential settlement that generally causes problems.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[geotechguy1]

Greenalleycat, in the context of residential foundations in NZ; I think it can be rationalized.

As you say the modern foundation types (at least if you're using one of the typologies associated with TC1 / TC2 which seem to be common) are fairly robust. Some of the settlement happens during construction. The 25mm differential criteria is high but, we deal with alot of challenging ground in NZ. Our American / Canadian / British / Australian colleagues would probably refuse to build on a significant amount of the land in NZ, but there is really not much choice here - you're either sitting on a swamp or sitting on or under a landslide (or a volcano...or in a tsunami zone...or subject to cliff erosion / coastal retreat...or trying to build a house straddling a fault (someone asked me to do this once) .

That being said, if you're on TC1 or TC2 type land building a residential house in NZ, you are probably doing something that has been proven by many other trials (i.e. all the other houses), so unless you're doing something out of the norm like building a two story or three story house in a sea of one story houses a 'well this has worked before' approach is probably somewhat justifiable (they use fancier words in things like Eurocode / British Standards but essentially they allow for 'experience' based approaches for low risk / low complexity projects). I don't really trust scalas but it gives you a rough idea of the shallow soil conditions and an idea if you need some undercuts or compaction. Also, for one house with hundreds of other houses with similar foundations in similarish ground conditions - there is already some reasonable proof that things 'work' even if we can't always explain why with a 3d plaxis model or something.

 

[Greenalleycat]

Completely agree with you (especially with how utterly rubbish our land mostly is)

I do have a soft spot for Scalas though. I think you just need to calibrate your expectations though - the data is laughably vague and prone to soooo many errors that it's impossible to get meaningful errors
However, if (as you say) you stay within accepted practice and use the Scala as more of a risk assessment tool then you're great
I do quite a bit of shallow geotech work but am a structural engineer through and through - my philosophy is that I shouldn't leave a site until I know the form that the foundation needs to take
The Scala + hand auger combo does an excellent job at this on almost any site (for resi)

 

[geotechguy1]

I think of scalas like a profiling tool mostly. It gives you an idea. It's OKish in coarse grained soils. In fine grained soils, best to do some shear vanes and use the scala for profiling (it can be quite useful to use a scala to quickly work out how thick a crust layer is - the numbers may be more questionable in fine grained soils but usually you can tell a distinct transition between a stiffer crust and some goop underneath).

In Canada I would run around with this bearing probe that you could push max 25mm into the ground and then use to get a capacity, for residential inspections. Comically useless, it was basically just an upgraded version of the old school 'boot test', or a piece of rebar and a hammer. Actually, I think the piece of rebar may have been more useful - at least you might be able to whack the thing in a bit further. That being said, all those sites had full basements so it was just for show.

 

[Greenalleycat]

The profiling/transition thing is a good point
I love the change in sound as goes in - probably has fucked my hearing, but it's professionally satisfying to here the ping change and know that the soil has just changed...
I personally feel that the structural designer should be the one doing the testing (pipe dream, I know) - it's such an imprecise test that it's mainly just useful for the designer to understand the overall risk profile of the site
The ground bearing thing is just an arbitrary exercise we do at the end...

 

[geotechguy1]

Yeah similar to a pile driving hammer right, you can tell from the sound

The closer you are to the testing, the more comfortable you are with the design. I spent my early career 'doing everything' so to speak (all the scoping / pricing, carry out field work, do the logs, write the geotech report). Was less conservative as a result. Now I am two or three layers removed from it and I've started questioning whether I trust any of the results. Every now and then someone sends me a log for 4 meter hand auger with all 'UTP shear vanes' in a swamp next to a CPT with QC of 0.5 MPa all the way down and I get nervous...

 

[Greenalleycat]

I see that come through in the reports that we get from other geotechs too - the conclusions just get progressively more and more conservative
Somehow every site has like 400-600mm of topsoil on it so no standard foundation solutions work blah blah....
Doesn't pass the sniff test on most jobs but it is very hard for us to question their site logs as we are so removed from the process and have no actual evidence

I've started putting less text in mine and just putting in photos of all the soil changes
That way it's pretty obvious what I consider "topsoil" vs "dark brown silt"
I view the purpose of a geotech report to be as short as possible whilst still providing the key criteria that is needed by the person using it
For almost all of the jobs that I do that is 1) a soil profile 2) a recommended bearing depth 3) a bearing value 4) identification of any important geotechnical risks 5) a recommended foundation system
That can be easily covered in 2 pages