Schmertmann Method using CPT vs Using Modulus of Elasticity 5

[sworddark90]

Greetings all , I am having a confusion in regards to which equation is more accurate for settlement calculations for granular soils.

- Smith's Elements of Soil Mechanics ,

- Principles of Foundation Engineering Braja M.Das . I have conducted a comparison using both methods , here are my results for a square shallow foundation:

Using equation listed in Principles of Foundation Engineering Braja M.Das

Using equation listed in Smith's Elements of Soil Mechanics

 

[pelelo]

I was not aware of Smith's formulae, however I have used Das' formulae.

The modulus of elasticity in both methods is computed using 2 different ways, you will never obtain the same answer using both procedures.

You need to keep in mind there are many E values correlations, so not necessarily the one in the DAS spreadsheet is the correct one. FHWA has E values correlations as well.

 

[EireChch]

As peleo - you are using different modulus correlations. No calculation is better than the other.

Also, for 2x2m foundation, depth of influence is 4m. Layer thickness of 1m is not suitable. Should be 0.5m thick or less

 

[geotechguy1]

Considering how much uncertainty there is in the actual applied load, the stress distribution and the stiffness 14 and 20mm seem like basically the same answer to me.

 

[EireChch]

geotechguy1 - unfortunately the reviewers on our projects dont see it like that.

24mm = pass, 25mm = fail.

crazy but thats the way it is

I can see their point though, if a spec is drawn up saying 25mm is the limit, then a line is drawn in the sand

 

[geotechguy1]

EireChch - I forget, are you still in NZ or back in North America now?

I definitely encountered the 25mm rule thing while working in Canada. Definitely crazy. If you calculate 24mm, even if everything else is equal probably you have at least +- 50% error in stiffness parameters due to natural variability, limited data, strain / stress dependency in stiffness, flaws in modeling, measurement errors, limitations in the model (usually the model being used has a linear stiffness relation) etc.

Pragmatically though my experience was that in Canadian practice everything was so absurdly conservative that it was more likely you calculate 24mm and the settlement is actually 2.4mm, because you used 5 MPa stiffness for a hard or dense soil and the structural engineer provides loads that are 10 times what they actually are.

 

[EireChch]

I am in the Middle East for the last few years. Its a cut throat industry to be honest. There are geotechnical engineers that represent all stakeholders. Clients geo, main contractor geo, sub contractors geo, third party geo etc.

Currently we are representing a ground improvement subcontractor. We get reviewed and challenged on every point, from every other geo too. Its stressful and difficult at times but also great because it really makes you need to know your onions!

Completely different to the UK or NZ where I worked. We would normally issue a report and it would go into the abyss. Occasionally we would get some very generic comments from the architect or QS. Or if we did a design it was essentially taken as gospel.

Yes I do see some similarities, the main one being that the structural engineer often wants us to assess settlement based on ULS loads and not SLS load.

Regarding stiffness, we are somewhat conservative but not overly. We are working for specialist sub contractor and having Zone Load Test results we can see our E = alpha x qc relationship results in an alpha of 8-10, we are proposing 5-7. however clients geo is trying to cap us at 3.5! Based on this we are calculating settlement that is several time small than our field results.

 

[pelelo]

Like geotechgu1 stated when computing settlements there are too many uncertainties: stiffness parameters, data, boring depths (most of the time borings are not deep enough), etc.

Unfortunately, If you ask 10 engineers to compute settlements, you will get 10 different answers, based on assumptions.

 

[Ron]

Precision in settlement calcs is like measuring with a micrometer, marking with a crayon and cutting with an axe.

 

[BigH]

Years ago - in the magazine Ground Engineering - there was an article that looked at some 15 different methods of determining settlements in sand. The basic conclusion was to pick three that you are comfortable with and then average them and use that as your estimate.

In a book in which the late Dr Nitin Som (one of the eminent Indian geotechnical engineers - and graduate of Imperial College) was oe of the authors, they looked at settlements in clays by consolidation theory, stress path theory and one other - and it was amazing the wide range of values they obtained. I remember an "old guy" (now of which I have joined the group) who said, when I was first starting out - "If you actually calculate the settlement within 30%, you've had a good day."

 

[EireChch]

Big - we are forced to calculate settlement to a decimal place....

 

[BigH]

EireChch - you have to be ... (well) kidding me! especially if you are using mm!

 

[EireChch]

No joke, people really loosing sight of what geotechnical engineering is.

 

[Tianll]

The methods of subsoil settlement calculation are always different resulting from different error of theoretical parameters, obtained both from site drilling and laboratory tests, as no two leaves are the same in the world!

 

[BigH]

EireChch - my sympathies