This is one of the toughest questions in rock mechanics. Many 'answers' may be far from the truth - which we will never know anyway. But here goes: what if 2D UDEC-MC or UDEC-BB are allowed/available. Represent the jointing as sketched by an engineering geologist, digitize, and apply scaled JRC, JCS values using block size scaling. Convert the non-linear BB to MC for the joints if UDEC-BB not available. Try to model with some discontinuous jointing, and use numerically glued joints. Can estimate a lower-bound c for the intact bridges/glued joints from a linear tangent to the tensile and UCS Mohr circles, c =1/2(sigma c x sigma t)E0.5: construct and check, also estimate phi. If a continuum analysis is to be performed with linear MC or non-linear HB be aware that degrading c and mobilizing phi is more correct than the usual addition. If the building is very high and you are under higher stress, the more correct curved envelope for the critical intact bridges is found by moving 1 x UCS along the normal stress axis and drawing the critical state max. shear strength Mohr circle, with radius = UCS, therefore max possible shear strength numerically similar to UCS (Singh et al 2011). Sigma 1 (max) = 3 x UCS for stronger more correct curvature than HB. Note also stress-dependent deformation moduli might be relevant for a high building.

If you have performed triaxial tests on rock, you may use that data to calculate the Hoek-Brown model. You may find how to calculate MC & HB model para

mter in following paper

https://link.springer.com/content/pdf/10.1007/s00603-012-0276-4.pdf

Thank you Sadiq Shamsher

Dear Ahmed Hussein Olat ,

Commonly triaxial results in rocks don't follow a lineal behaviour in the space s1-s3. Therefore, in rock mechanics we commonly use non-lineal envelopes as Hoek Brown rupture criterion. Fitting your data to H-B criterion will give you the intact rock properties, properties that generally we scale to rock mass (scaling given by answer of Sadiq, here the more recent version Article The Hoek-Brown failure criterion and GSI – 2018 edition

This scaling tries to take into account the effect of discontinuities in rock mass properties (by means of GSI parameter). Therefore, about your question, I believe that do or not the reduction, depends on the size-scale-relation between the pressure bulb and the typical rock mass block size (see figure 19 in the paper that I mention)

Thank you Cristobal Valderrama

Witold Bogusz thank you very much

Assuming that the correlations between the (triaxial) or uniaxial strength of rock with the rock strength parameters are correct, you have to account for the followings when choosing the rock strength analysis or design parameters:

1- The rock type.

2- The real RQD values of the rock in the field.

3- Rock with higher RQD values are better treated as mass.

4- Rock having lower RQD values, while being highly weathered and fractured, are better treated as soil having high cohesion and friction values.

5- Reducing the cohesion and friction values is a good practice, but you should follow specific correlations for the reduction.

5- Having good quality rock cores do not mean having a good rock layer, it all depends on the RQD or the Core Recovery.

This is one of the toughest questions in rock mechanics. Many 'answers' may be far from the truth - which we will never know anyway. But here goes: what if 2D UDEC-MC or UDEC-BB are allowed/available. Represent the jointing as sketched by an engineering geologist, digitize, and apply scaled JRC, JCS values using block size scaling. Convert the non-linear BB to MC for the joints if UDEC-BB not available. Try to model with some discontinuous jointing, and use numerically glued joints. Can estimate a lower-bound c for the intact bridges/glued joints from a linear tangent to the tensile and UCS Mohr circles, c =1/2(sigma c x sigma t)E0.5: construct and check, also estimate phi. If a continuum analysis is to be performed with linear MC or non-linear HB be aware that degrading c and mobilizing phi is more correct than the usual addition. If the building is very high and you are under higher stress, the more correct curved envelope for the critical intact bridges is found by moving 1 x UCS along the normal stress axis and drawing the critical state max. shear strength Mohr circle, with radius = UCS, therefore max possible shear strength numerically similar to UCS (Singh et al 2011). Sigma 1 (max) = 3 x UCS for stronger more correct curvature than HB. Note also stress-dependent deformation moduli might be relevant for a high building.

Thank you Tarek Nageeb Salem and Nick Ryland Barton

Indeed rock cohesiveness varies from sample to sample and very difficult to generalize in Rock Mechanics. I agree with the suggestion of Dr.Bolton that Represent the jointing as sketched by an engineering geologist, digitize, and apply scaled JRC, JCS values using block size scaling. Convert the non-linear BB to MC for the joints if UDEC-BB not available.

Indeed rock cohesiveness varies from sample to sample and very difficult to generalize in Rock Mechanics. I agree with the suggestion of Dr.Bolton that Represent the jointing as sketched by an engineering geologist, digitize, and apply scaled JRC, JCS values using block size scaling. Convert the non-linear BB to MC for the joints if UDEC-BB not available.

Thank you, Binay Samanta

I think it is reduced by a factor take into account the variation of drainage from laboratory and field, the size of sample (scale effect), discontinuities features, rate of loading and others.

Thank you Moataz A. Al-Obaydi

Mr Barton and other friends explained good details. Finally, under all parameters there are will be c

@Ahmed Hussain Olat hi.

This is so normal for rock mechanics assessments. Rock samples taken from the field for engineering structures projects should be in the best representative of the geology there.

The geotechnical parameter of the rock sample you take from the field shows the intact rock properties. What you need to do is not just find the intact rock parameters of the fresh rock sample. The correct engineering to be done here is to determine the rock mass properties of the GEOLOGICAL rock in the project area. In this sense, you need to convert the intack rock parameters that you obtained from the fresh rock sample to Rock mass parameters, for example by means of the roclab program. A good field observation and an accurate GSI assessment is essential for this transformation. The correct determination of the discontinuity characteristics of the rock unit is also very important.

However, the example you have given above did not come true to me. My advice to you would be to review the analysis using the roclab program.

Regards.

Thank you Yazgan Kırkayak . The parameters given are indeed, laboratory intact rock cohesion and angle of internal friction

Laboratory tests already suggested should be completed to arrive at conclusion of rockmass cohesion properties

Thank you Binay Samanta

You're welcome.

Actually I have been pointed out "50-300 kPa" values which you have been mentioned. These are so underestimated values for given intact rock properties.

So, these are being appropriate values for fault zone gouge maybe region within the GEOLOGICAL rock which has intact rock properties given above.

Regards.

Thank you again Yazgan Kırkayak

I am afraid I have to express doubt that a GSI assessment can ever be accurate when it is so crude a way to 'characterize' a rock mass. Then there is the obviously impressive-looking Roclab linkage to a lot of algebra and some record-long equations. Deformation modulus is not made depth or stress dependent, but it should be, and adding 'c and phi ' (sigma n tan phi) was shown long ago to be incorrect. ( Hajiabdolmajid V, Martin C D, Kaiser P K. Modelling brittle failure. Proc. 4th North American Rock Mechanics Symposium, NARMS 2000. Seattle J. Girard, M. Liebman, C. Breeds and T. Doe (Eds). Rotterdam: A.A. Balkema. 2000: 991–998, and others). Rock mechanics and rock engineering are not so easy as assumed by those who like the quick colourful plots. Rock masses around tunnels and under large loads fortunately do not develop so-called 'plastic zones' in the ways most frequently modelled with crosses and circles and colour. There is proved to be a lot of exaggeration if comparing with other methods.

Hi Mr. Nick Ryland Barton

Your evaluations and assessments on this topic are very valuable. Thank you a lot on behalf of myself.

I indirectly agree with your assessment of GSI. I suggested this method in terms of being an quick example to the solution of the problem. In terms of more detailed evaluation, RMR and / or Q-system evaluations should be made additionally according to the importance of the project. As you mentioned, the rock geomechanical parameters that we need to use in engineering analysis should be different depending on the state of engineering structures in the project (tall buildings, bridges, tunnels etc.). On the basis of these evaluations, as I point out above, a good geological evaluation and correct description of the discontinuity systems of the rock are very important.

In addition, I would like to add to some of the topics you have specified. As you mentioned, there is no change in the rock deformation modulus depending on the depth parameter entered in the Roclab program. However, this is not a deficiency or mistake of this method but an acceptance of this method. As you know, this situation is already clearly defined in the article Empirical estimation of rock mass modulus (E. Hoek, M.S. Diederichs, 2005). There are already in-situ field tests in the acceptance of the theoretical basis of the Roclab program. There is no Em variable depending on the depth because there are already rock deformation modulus measurements performed at the tunnel levels in situ and normalized correlations of these measurements. Since this theory is already based on GSI acceptance, so rock discontinuity systems are naturally included in this acceptance. As you know, the Roclab program is not a numerical program. An engineering geologist who wants to make more comprehensive analysis may be model the rock discontinuity system correctly and may determine the plastic regions accurate way in the rock units via numerical program.

In addition, the "Modeling brittle failure" article that you have mentioned above is not an admission completely ignored the c and f values calculated with hoek-brown, mohr-coulomb failure criteria or Roclab program. This article is a theoretical study for describing the acceptance of cohesion weakening-frictional strengthening for rocks with very high strength, very high GSI values and sparse cracked in intact rock condition. It should not be expected to represent all rock masses. The first paragraph in “The Hoek-Brown failure criterion and GSI” article is “The Hoek-Brown criterion was derived from the results of research into the brittle failure of intact rock by Hoek (1965) and on model studies of jointed rock mass behavior by Brown (1970)”. However, there is an article "Hoek – Brown parameters for predicting the depth of brittle failure around tunnels" related to this subject.

As a result, as you have stated, rock mechanics is not just selecting values from the colourful tables and making a crude engineering design. Good field work, adequate engineering geology and geotechnical assessment are required. Finally rock geotechnical parameters that we will use in the engineering design should be chose according to type of engineering structure and should be depend on the regional geology and structurel geology and the most suitable theory should be use in this context.

Best regards.

Excellent summary by Nick Barton.

I would argue that a C value in a MC shear strength model for discontinuous rock mass is a consequence of curve fitting, rather than a real mechacal property. Any C is possbile

Thank you Phil Dight and Yingxin Zhou