Hi Thomas,

First of all, thank you very much for your explanation, I really appreciated it.

About the usage of GPR in order to classify the rock using the Geological Strength Index (GSI), I was interested in determining the characteristics of the descontinuities in the rock mass, mainly its spacing, since it is one of the parameters that will be using.

Konstantinos, I don't think it's possible to do a complete rock mass classification with any indirect method. For example, it's already more difficult to make a RMR classsification in a borehole sample than on an outcrop or excavation face, let alone making the classification without even seen any part of the material.

I think a best bet would be to try to obtain some parameters used in rock mass classifications, like RQD (I'm not talking about GSI specifically here).

Take a look at this discussion here

http://www.forumgeociencias.com.br/viewtopic.php?f=56&t=30 (in Portuguese),

and also at

Onodera, TF. 1963. “Dynamic Investigation of Foundation Rocks, in Situ.” in Fifth Symposium on Rock Mechanics. New York: Pergamon Press.

You can try to obtain some discontinuities information from seismic waves velocities.

Sorry for not answering your actual question, but remember that rock mass classifcations systems deal with many parameters that are evaluated in a qualitative way. The authors of those systems (RMR, Q, GSI etc) have also made warnings about trying to use the systems in some rock types and about making comparisons between the indexes. So, when trying to obtain some of those parameters in a indirect and quantitative way, you must know you are not actually obtaining the parameter itself, but something that may lead to it.

Dear Marcos,

Thank you very much for your answer. I appreciated it.

I agree with about the risks of a rock mass classification using only the GSI method, but I am doing it only to estimate the constant rock parameters of Hoek-Brown criterion (mi, s and a), once I will be simulating a non linear finite element analysis of a rock mass under a specific gravity dam. Therefore I do not want to execute a comprehensive rock mass classification of a specific site.

So, as we know, there are only two factors used in this chart method, the blockiness and the conditions of the rock joints. The first one can be determined with the RQD value, while the second can be described by characteristics of the joints.

I already have the RQD values. My concern here is the possibility of GPR give me informations about the characteristics of discontinuities with some accuracy.

(I should've detailed a little bit more my question, sorry).

Just to be clear, even with the absence of the RQD measurement, there are some correlation that allow me to estimate it, including with the rock joint spacing, which I am thinking that It could be obtained through a GPR analysis. So, this is the reason for I asking here about the potential of this geophysical method in order to classify a rock mass with GSI method.

Dear Niel,

Thank you for your answer, I greatly appreciated it! I will definitely read both of your articles with professor Dr. Barton. I think that geotechnical and geophysical engineering are two sciences that should walk more together!

Konsantinos! I believe that both methods (GPR and GSI) will give limited help. Specifically re GSI one really has to be an optimist to think that GSI actually gives strength and deformability - think of the crudeness of the first visual estimate (of GSI), and then all the black-box algebra (and supporting equations) of what follows - and the guessing of D, meaning any desired deformation modulus is obtained. Of course if one knows more or less what parameters are needed - for the dam foundation analysis, one could 'justify' them by back-figuring a GSI that would fit. But this is hardly the scientific method that classification should be. Cross-hole GPR suggested by one of your respondents sounds interesting - if wave velocity could result.

Dear Dr. Barton,

Thank you for your response, I have a deep respect and admiration for you and all of your contributions to the Rock Mechanics.

What I was propousing here is an alternativa way, not the most expesive one, to estimate the GSI value using the Ground Penetraning Radar. I know about the importance of the discontinuities on a rock mass, and how they affect the rock mass strength. When I said about defining the rock mass strength and deformability, I was referring to the parameters used in the Hoek-Brown failure criterion and Hoek & Diederichs correlation.

Maybe I sounded like "I want to resolve all the rock mechanics problems with one geophysical method", but I was just wondering how could I evaluate a rock mass using GSI method with more accuracy.

The answer depends on the depth of target. As you know very well depth penetration of GPR method is very limited. Besides, deriving the fractures within rocks based on GPS is very hard. Therefore, classification of rock type beneath a site have to be done using other geophysical methods.

Using seismic refraction method will answer your question

This is a good question and has a brilliant future. It should be noted that this method (GSI) is one of the methods for measuring rock mass properties. In fact, rock mass strength must be mainly checked for the strength of the intact rock and the characteristics of the joints. Therefore, the geophysical method should be able to investigate these issues at the desired depth. However, to investigate rock mass properties, seismic and geoelectric methods are considered by some scientists.

I have taken guidance from many excellent geophysicists for over 30 years and one consistent thread I come across is that they generally say that you need more than one technique to get an accurate result. Hence, perhaps a way of looking at this question is to ask will GPR be a useful addition to a portfolio of tests rather than in its own right.

Indeed, Fusciardi. When I was studying more about the GSI, I realized that this method are limited to the data from outcrops and boreholes. So, when I decided to ask that specific question, I was interested in seeing the feedback from this community to the possibility of using GPR to map the underground fractures in order to help the classification system.

GSI is one of the empirical approaches for RMC or Rock Mass Characterization and is helpful where in-situ rock exposures are there. However, beneath surface, Geophysical methods would be more helpful for assessing characters and strength of discontinuities.

Agreed with @ Nick Ryland Barton

Hi there, thanks everyone for the feedback on this thread. We're currently planning an investigation where characterizing the persistence of joint sets in a bluff is crucial. However, the top of the bluff is densely vegetated, and we can't drill on top of the bluff due to limited site access. In this instance, would GPR give a crude estimate of the persistence of joints at the top of the bluff?

Thanks Neil for your feedback - appreciate the insight. What about prominent joint sets? E.g aperture of joints between 5 and 10mm between competent rock mass? We're dealing with cooling fractures associated with an ignimbrite.

Rock Mass Classification sometimes very difficult by direct geological methods.

Then Rock Mass Classification is a highly subjective question by any precise geophysical methods as GPR .

Hello Philippa,

I suggest you try to get help from a GPR/and more-advanced-radar expert Dr. Khosrow Bakhtar, who works in California. Those were big apertures - as in a near-surface columnar basalt.

You can definitely measure some elements of most of the Rock Mass Classification Systems using GPR such as the spacing between the joints or RQD but you have to be careful when mapping dry or very water saturated rock as with some dry rocks their dielectric constants are going to be similar to the clay infilling between the rock joints. Therefore, you may mistake the clay infilling between the joint for solid rock. Therefore, I recommend changing the dielectric constant by introducing moisture into the rock fractures. You can find better explanation of mapping of rock joints using GPR in some of my publications:

Apel, D.B., Dezelic, V., 2005, “Using Ground Penetrating Radar in Analyzing Structural Composition of Roofs in Tunnels”. SME Journal Vol. 60, No. 7, pp. 56-60 and SME Annual Transactions.

Apel, D.B., Dezelic, V., 2005, “Evaluation of high frequency ground penetrating radar (GPR) in mapping strata of dolomite and limestone rocks for ripping technique”, International Journal of Surface Mining, Reclamation and Environment, Vol. 19 issue 04, pp. 260-275.

Derek! Thanks for reminding me of something from my distant past as a geophysics-rock mass quality reviewer. (NB, 2006). Chapter 9 conclusions, of partial relevance here, and partly along the lines of your good advice were the following:

9.12 Radar and seismic signals are sensitive to different physical parameters (electro-magnetic wave conductivity, and mechanical stiffness respectively). The respective tomograms therefore highlight different features of the rock mass. Radar may delineate permeable zones caused by pore space or by joint apertures, in slightly different locations to the low seismic velocity zones associated with the clay-filled sections. The one will usually lie parallel to the other, since higher permeability may be associated with the heavily jointed zones that are often found in the walls of faults.

9.13 Low resistivity generally correlates with zones of increased water content and frequently with higher permeability. At a site in South Korea, a series of boreholes in weathered granites were Q-logged (by NB) and later compared with resistivity tomograms in a ‘blind’ test. Sections of the core with increased joint frequency (low RQD, high Jn) did not always correlate with low resistivity and vice versa. The two parameters that did show a consistent correlation with low resistivity were low values of Jw estimated from iron staining or apparent aperture, and the high values of Ja due to sand or silt fillings, and due to clay-fillings. The latter gives ‘anomalously’ low resistivity due to the ionic effects of the clay. Water content and permeability are clearly lower in such discontinuities than in those that are sand or silt filled. This represents a potential source of error in judging the meaning of low resistivity zones