Comments on “Using the Impulse-Response Pile Data for Soil Characterization”

This is another post on a paper (linked to from here) which cites my work, in this case two of them: Improved Methods for Forward and Inverse Solution of the Wave Equation for Piles and Closed Form Solution of the Wave Equation for Piles. The concept is simple but the execution, not so much, and as with anything with geotechnical engineering there are pitfalls on the way to a usable solution.

We start with an existing technology: low-strain integrity testing of piles. A simple example of this is shown above, it’s the Pilewave program from Piletest. (Yes, I’m aware that it’s the Windows 3.1 version, if you’re interesting in running DOS and Windows 3.1 programs to save on the expense of “new” engineering software, you can visit Partying Like It’s 1987: Running WEAP87 and SPILE (and other programs) on DOSBox.)

With that distraction out of the way, note that, as the stress wave goes down and back up the pile, there is attenuation due to the interaction with the soil. In the simple demo of Pilewave, the soil resistance is constant along the shaft. But…if we could determine that the pile didn’t have defects which reflected waves, could we use information from the soil attenuation to determine the type of soil surrounding the pile at any given elevation? The answer in principle is “yes” and this paper, although not unique, it is an interesting step forward.

Pile Integrity Testing is a low-strain technique. That’s in contrast to the high-strain methods we’re used to in pile driving analysis. This one takes a leaf from the seismic refraction method (which will be featured as before in Soils in Construction, Seventh Edition) which is also a low-strain technique, as it is a geophysical method. The idea is that the pile acts as a probe into the soil; the response to exitation can be inversely analysed to determine the types of soils around the pile. As the paper notes, if you divide up the pile into enough “layers” the actual soil layering itself (based on the properties returned to you by the method) will basically emerge from the data.

As is generally the case with inverse methods, the solution is complex; it is described in the paper. There are a few comments that I would like to make as follows:

• His governing equations are similar to the Telegrapher’s Equation used in Closed Form Solution of the Wave Equation for Piles and include a strain term but lack a damping term. Usually a damping term is necessary to model the energy dissapation into the soil; whether that applies to this problem remains to be seen.
• The toe model he used was inspired by Closed Form Solution of the Wave Equation for Piles, which I discussed relative to its origins and influence by Alain Holeyman in my recent post Comments on “Nonlinear fictitious-soil pile model for pile high-strain dynamic analysis” and “Fictitious soil pile model for dynamic analysis of pipe piles under high strain conditions.”
• Driven piles are subject to compaction and disturbance at the soil-pile interface; how this affects the results remains to be seen. The difference in soil response based on rate effects also will need to be addressed.

I hope that this research continues; I think it has potential.