I have always used the tangent off the e vs log P curve.  If the structure is very settlement sensitive, you may want to run several tests to determine an average value, since as you point out, there is a lot of art in picking the point.

markvs,

For a recent project we convinced the DOT to allow us to perform a load/unload/reload in the middle of the test instead of the more conventional unload at the end of the full loading. This makes determining Cr much easier. However, taking Cr approximately equal to Cs is usually a good approximation.

Jeff

Jeffrey T. Donville, PE
TTL Associates, Inc.
www.ttlassoc.com

I have always performed the test with several unloads in the middle.  This give you multiple oportunities to look for Cr and can help with determining the preconsolidation pressure in odd cases.

jdonville - why did you have to convince the DOT - presumably ODOT - to permit you to run an unload/reload cycle in your test?  This was always standard procedure whenever we did/do consolidation testing.  Did you run the cycle before or just after reaching the preconsolidation pressure?

I second BigH, why did you have to "convince" the DOT to do this?  If it was Ohio DOT, let me know (privately) who it was.

In this age of auto equipment and data aquisition, I am surprised the clients are not requesting unload/reload testing as a standard item.  

As a side note, Since I purchased my equipment 5 years ago, I have found it to be very cost effective (paid the equipment off in 2 years) and the level of my information has greatly increased. I can afford to run more tests on projects. I recommend the equipment to all who can. I imagine the equipment would not fit into the travels of BigH.

   In reading Jeff's last sentence again, I agree to a degree on the Cs and Cr.  I would limit, though, the "slope" to be in the range of the end of the test Cs to less than the existing overburden pressure.  When the load is "wholly" removed, I have noticed increased rate of swelling - like a near free swell rather than a confined swell - and a straight-line estimation would be too steep.  
   As for emmgjld - congrats on getting the "fancy" equipment.  It does help - when I was last really doing the consulting geo, I found that the tests were just getting too expensive for the budgets we were able to get.  Of course, I would break the bank if necessary, but one doesn't often have such budgets - why triaxials were out of the question except in the "meaty" jobs.  As for my travels, you understand!  but, I will soon be able to utilize a double probe nuke!!
Ciao.  

ODOT only specifies the test method to be used (AASHTO or ASTM 1-D Consolidation Stds). ODOT does not give any explicit guidance or instruction on unload-reload cycles.

Panars,

As I understand the situation, the baseline consolidation test per the methods outlined above is a simple increasing load followed by the unload at the "end" of the test. The unload-reload cycles were provided at an extra charge to reflect the extra running time and effort to perform them. Thus the need for "convincing" the relevant project authority that the extra cost was worth the extra information. In actuality, I am not aware who suggested the unload-reload for that project or just how much "convincing" was needed (before I joined my firm).

BigH,

Unload/reloads for that project were performed at a standard point in the test (same for all samples tested) - which ended up near or slightly above the preconsolidation presure for most samples.

emmgjld,

Would you be able to tell us what equipment you are using?

Jeff