CRS vs Icremental Consolidation 2

[davecooper]

I have noticed that the constant rate of strain consolidation test will typically show a higher Pc or max past pressure than the incremental. Can the standard construction typically used for the incremental consol be used for the CRS? I have had clients mention to me that the CRS tests showed slightly overconsolidation when incremental consols run on the same project showed normally consolidated. Has anyone read and data about this or done some research on it?

Dirtdoc1

 

[Focht3]

Please excuse my long message - you hit close to a topic of great interest to me.

CRS and incremental consolidation tests do yield different results, but my expectation would be the opposite of the statement in your question. However, I do not run CRS tests because we do not have enough need in this area for fast consolidation testing, so my impression may be wrong.

I'm sure that you already know that CRS uses pore pressure measurements and the applied stress to calculate the effective stress at any given moment. (Assuming that pore pressures develop during the test.) While the computer code may make corrections for machine and temperature, it cannot separate the three main compression components: elastic compression, primary consolidation and secondary consolidation.

"Standard" incremental tests typically have the same problem, since measurements are taken too infrequently to separate the three main compression components, either.

In my view, the main difference between "standard" incremental tests and CRS tests is the amount of secondary consolidation in the final result. You can't get from one result to the other, either, without making a lot of assumptions. And it is my belief that the effect of preconsolidation is exhibited in the secondary consolidation component of the incremental consolidation test. As a result, I am unlikely to buy CRS test equipment.

 

[jdmm]

If you refer to CAM Clay theory it says that the position of the yield surface is stress path dependent. The effective stress path during quick loading of a CRS is different than an incremental (drained test). As a result we would expect the yield surface to be different. Based on theory I would expect that the CRS would predict at Pc' less than the incremental test due to an effective stress path lower than the incremental test (agree with Focht3) but I haven't looked at this in detail. You should consider the type of test relative to the type of loading. ie. use the type of test that best reflects the design conditions.

 

[davecooper]

Focht3 & JDMM,

Thanks for the response guys. I see where Focht3 is going with the differences due to secondary compression in one test but not the other. Same thing with the total vs effective stress path. I don't think that is an issue since what we are trying to acheive in both methods is an accurate measurement of effective stress. By measuring the pore-pressure (in the CRS test) we get effective stress and by letting the pore-pressure dissapate (in the incremental test)we get effective stress. So in both cases we get effective stress. By subtracting out the pore-pressure (in the CRS trest)we should get the same value as if we had allowed the pore-pressure to dissapate by allowing it to go into secondary consolidation (in the incremental test). Man that is a mouth full! But I'm not sure. That is a very good point.

I believe that the Pc of the CRS test is a bit higher because we are not disturbing the sample by doubling the loads. If I take an incremental consolidation test (on soft marine clay) and use small load increments I get a higher Pc also. That would indicate that the CRS test causes less disturbance and maybe the Casagrande Construction was developed to counter that disturbance. Therefore my question is: has there been or should there be a different method of determining the maximum past pressure for a CRS test?

Best Regards,

Dirtdoc1

 

[Focht3]

I had hoped to avoid this. Oh, well.

Dirtdoc1:

The differences are not due to "disturbance" due to the loading steps. And I can prove it - with your help!


Do you have any data acquisition equipment that you could hook up to your incremental load setup and measure sample movement to 0.0001 inch or less (accurate and reliable)? If so, run an incremental consolidation test on a high quality sample and take deflection measurements every second for every load increment. (I know - that's a lot of points!) For each load increment,

1. Calculate the immediate movement. Subtract it from every data point for that load increment.
2. Calculate the primary consolidation settlement using square root of time method. Then figure the primary consolidation settlement vs time and subtract those values from the appropriate net movements left after step 1.
3. Plot the remaining values - secondary consolidation versus time. Also plot these values vs effective stress (which is derived using data from step 2) and combine with curves from other load increments.

Run the test from 0.1 tsf to max capacity, unload to 0.1 tsf and reload to max capacity. Plot primary consolidation movements vs effective stress. Notice the results track a very nice hysteresis loop. If the "break" in the plot were due to the effect of P^'_c then why does the reload track the initial loading almost exactly? It should not track beyond the sample's original P^'_c since the P^'_c for the reload is now the max capacity of the test frame.

I contend that the effects of P^'_c affect secondary consolidation, not primary. Too many engineers don't understand that a "standard" e-log(p) curve includes machine compression, immediate (elastic) movements, primary consolidation movements and secondary consolidation movements (as well as temperature effects, trimming errors, etc.)

CAM clay model description is correct; unfortunately, most geotechnical engineers don't understand CAM clay. And it doesn't really address the primary vs secondary consolidation issue, either.

 

[davecooper]

Focht3:

Are you trying to dazzle me with brilliance or baffle me with bullshit?! I think the latter. I am not even going to address what you said. Rather I will stick with some logic. The secondary consolidation has nothing to do with the Pc. The Pc is determined on the X axis. The secondary consolidation affects only the Y axis. X = load, Y = change in height. All the secondary compression can do is shift the curve up and down along the Y axis of the strain-log-p curve. That won't affect the Pc, which is determined on the X axis. You once told me to stick with what I know. Well I am. You should follow your own advice. How many CRS tests have you run? I have lost count.

By the way, any lab worth it's salt has a machine deflection correction (also known as equipment compliance)coded into it's software or spreadsheet.

Dirtdoc1

 

[MRM]

Also, I would like to add that immediate compression (of the sample) is actually not included in the consolidation test either, as Focht3 suggested. Immediate compression in the classical sense is caused by undrained distortion movements (mostly lateral) of a large soil mass due to a comparitively small footing. In fact, some call it "distortion settlement." The consolidation test is run with a confining ring so that lateral deformations cannot occur (1-d). The immediate load is taken by the pore water pressure until disappation occurs and a corresponding decrease in height occurs (primary consolidation). Not to be nit-picky, but it is a fundamentally important point worth mentioning.

 

[Focht3]

I was a little reluctant to submit my previous post. It's a complicated topic - and few engineers are willing to reassess those topics that they accepted as fact as a student. Geotechnical engineers are no exception.

Dirtdoc1:
No "turd blossoms." If you want to find out firsthand about P_C then do as I suggested. THINK about the problem, don't rely on someone else's interpretation of the data. P_C is dependent on the effective stress ("x axis&quot but (in my opinion) it does not affect the primary consolidation behavior of the soil sample.

Regarding CRS tests: I know that you have done far more of these. But that isn't the issue. Secondary consolidation is not a "constant" and does not simply shift the graph. It is a function of stress level and time, but is unaffected by the end of primary consolidation. The amount of time that a given load is left on a sample will affect the contribution of secondary consolidation to the total deflection of the sample.

It seems that you think I was making fun of you. I was not. My comments on machine deflection were intended to reflect my understanding of all the factors that could affect the test. If I offended you - I apologize. That is not what I intended.

MRM:
I've run a lot of incremental load consolidation tests myself - I am not relying on the observations of others. You may disagree with the use of the term "settlement", but the movement is quite real. The deflection dial registers an immediate movement on application of load; you can calculate the immediate deflection using square root of time techniques and subtracting the machine deflection for that load. (The immediate deflection is significantly greater than the machine deflection.) Since water is effectively incompressible and machine compression has been accounted for, this has to occur within the soil fabric. And it is not necessarily due to trimming errors; it happens on reload load increments, too.

The sample is indeed laterally confined - the same restraint assumed by Westergaard when he modified Boussinesq's work. (I hope I spelled those correctly!) The presence of lateral restraint does not invalidate my interpretation. Compare the shear modulus obtained from an incremental consolidation test to "elastic" test results (DMT, SBP, resonant column, low strain cyclic triaxial tests) for the same soils. You won't get a match, but they will be in the same range. You may observe a problem with this as the effective stress increases - due to side shear and the resulting stress rotation.

Anyhow, thanks for the responses.

 

[MRM]

Focht3,
I don't doubt that the immediate movement is real and measurable...I've observed it myself in all of the consolidation tests I've run.

My question to you is; do you think it's possible that the instant the load is applied to the test specimen, the movement you initially notice (before any measurable time has elapsed) is actual primary consolidation (water immediately leaving) occurring to the specimen at the very outer fringes of the sample? I think we need to remember that primary consolidation of the specimen is not completed uniformly. The specimen has a definite thickness, maybe 0.5" to 1". The outer fringes reach 99.99% primary consolidation long before the center even begins to evacuate water from the voids. I would say this is the immediate settlement that we see when we plot the Taylor curve.

Also, I don't believe the loads used in most conventional consolidation tests are not large enough to cause much particle distortion or water compression. If they were, the compression of the water would be converted to actual consolidation settlement as primary progresses, so it would be a moot point. Particle compression is too small to measure.

To add an interesting point to the secondary compression discussion in the earlier messages in this thread(at least I've always found it interesting) is that since the outer fringes of the specimen have largely completed primary consolidation long before the center, the outer fringes are actually undergoing secondary compression, WHILE the remainder of the specimen is still in the primary consolidation phase. If that doesn't frustrate us geotech types from a data interpretation viewpoint, I don't know what will!

 

[Focht3]

MRM:
I am willing to admit that primary consolidation is a possible cause of the "immediate" settlement seen in incremental load consolidation tests - but I don't think so. If it were, then it would also fit on the initial linear portion of a square-root-of-time graphing of the data. It doesn't. And I see no reason for the soil's properties to vary so drastically over a total time increment of only two seconds.

I agree with your "force magnitude" comments regarding soil particle and water compression. I would add that the loads exerted by foundations don't reach the necessary levels to cause these types of compression, either.

Your comments regarding the complexity involved in sample behavior is correct; yet it looks so simple!

 

[MRM]

Focht3,
Now if they would only pay us accordingly for having to interpret such complex soil behavior, we'd be all set! Good points, and I enjoyed the discussion. Thanks.

 

[toolpusher]

There is no question that secondary consolidation effects can lead to reduced void ratios in the standard (incremental) oedometer test, wherein typically the load is left on for a specific period of time. There has been a considerable amount of work done to identify the amount of primary and secondary compression and generally we find that more organic soils exhibit commensurately more secondary consolidation than inorganic soils do. I would caution that the incremental oedometer test is perhaps not always the ideal choice for some soil types, which have high secondary-to-primary compression ratios.

Also, assuming the load-increment ratio is maintained at 1 throughout the incremental oedometer test, secondary consolidation effects within the recompression stress range can cause a slight flattening of the end-of-load e vs log P curve; this is most problematic with silty soils and can give a sample an apparent "disturbed" consolidation response. This generally makes the Casagrande construction to determine Pc difficult and in my experience tends to lower Pc compared to CRS results.

The implication of course in this is that we presume that secondary and primary consolidation processes occur simultaneously. The CRS test is in my opinion therefore superior, if for no other choice in that it tends to minimize the secondary consolidation effect, in essence by continously loading up the sample - giving the best estimate of the unaged, or "end-of-primary" response as possible. It also tends to be a lot faster to perform. Still, I have had good success testing organic-rich offshore clayey silts in using the incremental test (the only equipment I had at hand) by not allowing compression to continue after the t100 point had been reached; especially at stresses below Pc. Use of the root-time method is very helpful in evaluating when to start the next load, rather than the log-time method.

In the CRS equipment you do have to be aware of the equipment deflection, which is produced by "stretching" of the loading system (ie. the clamping rods that hold loading head on the oedometer ring (the rods in the Rowe cell). This can easily be calibrated by pressurizing the cell and noting the change in deflection on the vertical displacement dial gauge.