Interpreting oedometer tests.

[Aluminium_Soil]

Hi all,
Just trying to clarify something.

As the title suggests I am attempting to derive a characteristic Mv value for a cohesive soil using the oedometer test.
I have read online, people sometimes average given Mv values for all incremental pressures and take an average.
This seems way too obvious to be inaccurate for design purposes (perhaps I'm wrong).

Anyways I am struggling with interpreting the results.
I have an Mv value for each pressure increment, doubling from overburden up to 5x the expected load for the soil, then unloading back to overburden.
Is it as simple as selecting the Mv value for the known load of the structure?

The problem exists where the Mv value at 50 kPa on the lab test is way higher compared to 100/150 kPa?
(I understand its a function of a larger initial height change in the sample, where PwP dissipates which becomes gradually smaller overtime).
So if using the Mv value for a house designed to bear at 50 kPa, when calculating consolidation settlement this almost always returns excessive settlement and cannot be limited to 25 mm.
My thought is that surely a house designed to bear 50 kPa obviously a lighter load, would settle less than a 100 kPa house but when using the Mv for 100 kPa the house settles less?!?

Maybe I am missing something or would like to understand how you interpret oedometer results.
Any help is greatly appreciated - I'm at a loss after going through 6 books.

 

[EireChch]

Post the odeometer plots.

How much larger is Mv at 50kPa compared to 100kPa?

Maybe you have a disturbed sample and the 50kPa loading is removing some disturbance up to a certain point before actually starting to apply true loading. This will give you a dodgy Mv value for that increment.

Also, have you assessed settlement using cc, e0 etc.?

 

[Aluminium_Soil]

Thanks for the reply EireChch,

Results:

If using say around 40 kPa for a lightly loaded structure (0.25 m2/MN) - this would return a rather high result.
Although when using a heavier structure at (0.16 M2/MN) would return a lower settlement result.

Would you propose to schedule tests a different way?

Settlement was assessed using Poed with Skemptons/Bjerrum geological coefficient, then summing up the layers for consolidation.

 

[geotechguy1]

Can you share your settlement calculations? Over what area / footing width did you apply the 50 kPa load? How did you calculate stress distribution with depth?

 

[EireChch]

plot the full test pressures also. i plotted in elogp v void ratio and you have straight line data. no noticeable preconsolidation stress. if you have more pressures i can plot more and this will give an idea of the quality of your sample/test

whats your sample depth?

and re settlement calc. what load are you applying on mv of 0.25 and 0.16?

gg1 is right, we need to see a settlement calc

 

[Aluminium_Soil]

Thanks both,

With regards to the amount of data points, is it normal to only have five pressures?
Thought this was standard - would you suggest testing more points?

Sample depth was at 1.00 m bgl.

Apologies if I didn't make it clear, what I am trying to say is how would I select the appropriate Mv value?
Wouldn't settlement calculations given less settlement with a larger load as there is a smaller Mv value compared to the larger Mv value for a lighter load?
Surely wouldn't a lighter load cause less settlement regardless of calculation?

Quite new to the game so have a lot of time in books rather than experience.

 

[BigH]

Aluminium Soil - why use Mv . . . why not just use Cc/1+eo etc. - i.e., the pressure vs void ratio curve

 

[Aluminium_Soil]

Hi BigH,

What is the difference between using Mv and using Cc for settlement analysis?

Could you point me in the right direction for a good understanding on the analysis? Thanks.

 

[BigH]

Aluminium Soil: I just think the Cc method is a bit more straightforward - you don't have an "issue" with "changing" Mv values - you need Cr and CC (and if you plot Log Pressure vs Strain, you don't need to determine Cr or Cc as the values will be Cr/(1+eo) and Cc/{1+eo).

 

[EireChch]

I would agree with BigH and use Cc/1+e0 method. I think its far less susceptible issues with a bad reading. For example you are using Mv based on one data point. However with cc/1+e0 method you estimate Cc (and Cr) based on a number of data points.

In an ideal world this should not matter as you should get the exact settlement from using Mv or Cc/1+e0 methods. However i do think the above to be true. Using the Cc/1+e0 also eliminates the issue of which pressure to use as loading up to p' (pre consolidation stress)is along the Cc line while after p' loading is along your Cr line.

I have looked at your data and it is difficult to use. As your sample is from 1m depth it is very hard to interpret the p' (pre consolidation stress). Ideally you should have 2-3 data points below you exiting overburden stress. The reason for this is that if there is p' then you should start to see it after overburden stress. This should give you a nice rounded curve, similar to ones in textbooks which are more easily interpreted. Again all this does assume you have a good sample, if you have a bad sample then all bets are off.

You have started your first stage at 20kPa. Overburden is 10-20kPa at base of this layer, considering water table. I would have started at 5kPa, then 10kPa, 20kPa.

Your technicians / lab may moan at this as the loadings are so small however it is possible to do.

 

[geomane]

@EireChch, do you ever run into issues with the sample swelling when you use very low consolidation pressures for your first couple of points?

 

[EireChch]

Yes we have run into that. It is typically more coming in overconsolidated clays. In these cases the pre consolidation is obvious higher than current effective strees so it doesn’t become as big an issue.

But then sometimes it can be avoided.