Dear structo's,
When analysing fault gouge, it is important to consider the origin of such
a gouge as a fundamental part of the interpretation and the significance
of the results of the measurements.
Fault gauges in granites or sandstones form by cataclasis, a process
whereby matrix grains are ground into a fault rock. Essential parameters
in the brittle domain are the normal stress on the shear zone (fault) and
the distance along which shear has taken place. Temperature is not so
important in the brittle domain, but does become importanat at high
temperature zones. I focus on the brittle domain and for example HC
reservoirs where fault gouges are important.
Cataclasis begins at low stress levels (i.e. shallow normal faulting) and
is then associated with increasing porosity and permeability.
With increasingly higher stress levels the porosity becomes increasingly
reduced and finally a cataclastic gouge forms a paste in which even SEM
analysis (3000x mag) can not identify remaining porosity.
An other complicating factor is the composition of the matrix rock. Quartz
grains are strong, but feldspars and clays are weaker and will break
earlier than quartz grains.
Clearly the scope for variations in grain size distribution is very wide.
Regards, Dirk
> Dear Lei Zhang,
>
> We performed experiments on intact granitoid rocks in the laboratory and
> compared the resulting grain size distribution to that of natural fault
> gouge. We used thin sections and image analysis on images taken with an
> electron microscope plus optical microscope to study the (2 dimensional)
> grain size distribution for the fault gouge. For both the naturally and
> experimentally produced fault gouge we found two different D-values, with
> a
> transition at a few micrometers (the exact value depends on the
> mineralogy). The D-value is higher for the larger grains than for the
> small
> grains, probably because of a change in the main comminution mechanism.
> Healing of the fault gouge reduces the D-value again. These results were
> published in a couple of papers a few years ago, let me know if you do not
> have access to them. There is a paper by Fernlund (1998) where she
> compares
> the grain size distribution obtained by sieving to image analysis.
>
> With kind regards,
> Nynke Keulen
>
> 2012/2/15 ÕÅÀ× <[log in to unmask]>
>
>> Hi, everyone,
>>
>> I have a question in the particle size analysis if we can use the laser
>> particle analysis data to calculate the fractal value(D) of the gouge
>> like
>> the method below (Storti et al., 2003). However, laser particle analysis
>> data can gives us larger particle size distribution from mm to ¦Ìm. the
>> curve of the particle size distribution is not linear, there is an
>> inflection, which often occurs at several micron, so this means the
>> particle distribution is not self-similar organization below this size
>> value.
>>
>> The sieved method to obtain the different grade of particle size.
>> The weight of residual material in each sieve was transformed to the
>> equivalent particle number by assuming that the grain shapes can be
>> approximated by spheres, and we suppose that the material in each sieve
>> has the same diameter as the mesh aperture of the overlying sieve. A
>> density of 2670 kg/m3 was used for computing the weight of the reference
>> spheres.
>> then we can obtain the number of the material in different grade of
>> particle size (Storti et al., 2003).
>>
>>
>> Lei Zhang
>>
>> 2.15.2012
>>
>
|