compression index of soil? 2

[boffintech]

We have an engineer in the office that takes a certain amount of joy in testing me. He returned some consolidation test results to me with a note asking for the Cc (compression index) of each soil sample. The what? He claimed it could be solved for with the info from the tests. I searched ASTM D-2435, Standard Test Method for One-Dimensional Consolidation Properties of Soils, and found nothing. Is this guy pulling my chain or is there really a formula for this? And if so, where?

 

[BigH]

Why don't you look at your introductory soils books. Cc is nothing more than the slope of the consolidation curve beyond the preconsolidation pressure. Now of coure, we have several scenarios here - the laboratory Cc value, the estimated field Cc value. Then, there are the sensitive soils that actually appear to have curvilinear results beyond the range. So - for the lab Cc value, draw a slope along the "virgin" compression line - then take the value as the diffence in the void ratios (using a maginitude difference on your load - say use 1 and 10 tsf; or 2 and 20 tsf. Now, if your lab data was plotted as strain vs log p', then, the curve is the compression ratio, CR, or Cc/(1+eo). Hence, knowing eo and CR, you can calculate Cc. Similar arguments are for the Recompression index or the recompression ratio. (that part of the e-log p') curve that is before the preconsolidation pressure. Now, if he was really yanking your chain, he might have given you the data as a plot of e-p' rather than e-logp' so you might need to replot the data to determine the Cc value.

p.s. Now, yank his chain and ask him about alternative methods of determining the preconsolidation pressure other than Casagrande's method. (For your information, this was discussed in detail in a paper I found on the internet, but I saw many years ago in Geotechnique - plot loge-logp and it is the intersection of the two before and after curves on the plot.

 

[boffintech]

For the lab Cc value, draw a slope along the "virgin" compression line - then take the value as the diffence in the void ratios (using a maginitude difference on your load - say use 1 and 10 tsf; or 2 and 20 tsf.

OK, this what I have:
void ratio at start of test is .93 eo
void ratio at end of test is .72 ef

P1 = 1.0 tsf and e1 = .95
P2 = 10 tsf and e2 = .67

the Cc = .210

 

[BigH]

Taking that both are on the "construction" slope - then Cc would be 0.28 (0.95-0.67). The computation is on a "fake" line that is used just for the purposes of determining the slope - as you see from your data, the "chosen" 1 and 10 tsf are outside the tests e values.

 

[boffintech]

After more thought I wonder if the P1, e1 and P2, e2 represent points on the consolidation curve then how could the initial and final void ratios be within the e1 and e2 range? That withstanding, having only 2 points on the consolidation curve does not appear to allow for the calculate Cc. What if the points are on the over consolidated portion of the curve, or the normally consolidated portion, or one point on each? If both points are on the Nc portion of the curve then (.95-.67)/log(10/1) = .28 Right?

 

[jdonville]

boffintech,

If you could report the test results to us in a tabular form (just the e and applied pressure values should suffice) then we could help you interpret them.

You are correct. Both points used for estimating Cc from 1-d consol test results need to be on the NC or virgin compression curve.

ASTM D2435 shows examples of how to obtain the Cc and Cr/Cs slopes, depending on the method used to perform the test.

Jeff


Jeffrey T. Donville, PE
TTL Associates, Inc.

http://www.ttlassoc.com

 

[BigH]

The method I indicated, again, is a "line" that is constructed along the virgin compression curve. To make the calculation simple you use a "factor of 10" between two points - it is a mathematical interpretation for EASE of computation. It doesn't matter whether the "point" is really on your laboratory curve or not so long as the "line" is along the virgin curve - whichever method you use - using "actual" values or the mathematical model - will give you the same result - just that how I indicated is easier and quicker to do.

 

[Dirtguy4587]

Just to put my 2 cents in here - I often use different methods and compare the results. The Schmertmann Procedure and Cassagrande Procesure both are commonly used. They attempt to correct for changes in void ratio caused by "sampling disturbance". They may be of interest to you during yout