Relationship Between the Geometric Parameters of Rock Fractures,the Size of Percolation Clusters and REV

Modeling fractured rocks with numerical methods requires some derived parameters, among which the fracture network connectivity and the size of the representative elementary volume (REV) are both of crucial importance. Percolation and REV analyses were made by the RepSim code. The program uses input parameters such as fractal dimension of the fracture midpoints (D _c), length exponent (E) and relative dip (α ^r) data. For percolation analysis, the relative sizes of the largest percolation clusters have been calculated by stochastic realizations of the simulated fracture networks with different parameter triplets. Furthermore, fracture networks can be classified into three major types on the basis of their (E,D _c,α ^r) parameters. For the REV calculations, the porosity of the generated fracture network was calculated. The derived REV size of a fracture network depends essentially on input parameters and shows a decreasing tendency with increasing D and E and vice versa. The method mentioned above was tested on both metamorphic samples of the Pannonian Basin and Variscan granitoid rocks of the Mórágy Complex. Percolation values predicted for the Mórágy granite are highly sensitive to alterations in the input parameters. The amphibolite bodies displayed a modeled fracture network with 80 to 90% of all fractures being interconnected, while the largest achievable percolation cluster size of gneiss is less than 10%. The REV size of the amphibolite is about 20 m as a result of connected fractures filling the whole body, while gneiss has lower porosity and higher REV (approximately 70 m).     

Direct-count estimates of fungal and bacterial biovolume in dead leaves of smooth cordgrass (Spartina alterniflora Loisel.)

Two types of hyphal-extraction, direct-count methods of estimating fungal biovalume in standing-dead, autumn leaves of Spartina alterniflora were compared with a clearing+staining method which does not require homogenization. Bacterial biovolume also was estimated, by an acridine-orange direct-count method. Type of homogenization had little effect on measured fungal volume, but counts made using water-soluble-aniline-blue epifluorescence were consistently lower than those made using phase-contrast (by 6–10x). Clearing+staining could not be used to estimate hyphal lengths, but was of use in estimating total ascocarp volume (=0.06 mm³ per mm³ of leaf). Estimated fungal hyphal volume was approximately 0.27 mm³ per mm³ of leaf. Bacterial volume was <3% of fungal volume.     

Efficacy of minnow traps for sampling mummichogs in tidal marshes

We evaluated two types of wire-mesh minnow traps for sampling mummichogs (Fundulus heteroclitus) from intertidal rivulets on the surface of a salt marsh. In field trials, mean catch (no. of fish trap⁻¹) did not differ significantly between trap types nor across soak times ranging from 30 to 240 min. Catch rate (no. of fish trap⁻¹ min⁻¹) was significantly (K-W ANOVA on ranked data, H_4df=24.79; p<0.001) greater at shorter soak times, and larger size classes (>45 mm TL) of fish were relatively more abundant (Kolmogorov-Smirnov test, p<0.0001) in collections from double-funnel cylindrical traps (volume=10,330 cm³) compared to rectangular traps having about twice the volume and a single funnel cylindrical traps (volume=10,330 cm³) compared to rectangular traps having about twice the volume and a single funnel opening. We estimated entry and escape rates by adding marked fish to traps at initial densities of 30, 60, and 120 trap⁻¹ and measuring changes in the numbers of marked and unmarked fish in each trap type at soak times ranging from 15 to 120 min during the flood stage of the tidal cycle. Escape rate was almost, twice as great from the two-funnel cylindrical traps (0.77 fish min⁻¹) than from the single-funnel rectangular type (0.42 fish min⁻¹) but entry rates did not differ significantly (K-W ANOVA, H_4df=0.06; p=0.813) between trap types. Escape rate increased and entry rate decreased as fish density within traps was increased. Escape rates were highest immediately following immersion of the traps, but approached a constant value after 30 min. Trap-specific rates of entry and escape can be affected by a variety of factors including physical characteristics of different trap types, method and timing of trap deployment, and fish behavior. Catch data from minnow traps are unlikely to be useful in measuring relative abundance of mummichogs among habitats unless very short soak times (≤60 min) are used and calibration tests are performed. Use of longer soak times virtually ignores the dynamics involved in determining catch, and is unlikely to provide accurate information on either abundance or size distribution of mummichogs from tidal marsh habitats.     

Comparison of size effect for different types of strength tests

SummaryComparison of Size Effect for Different Types of Strength Tests Different theories have been proposed to explain and predict size effect. Notable is Weibull's Weakest link theory. In addition various theories have been founded on strain energy consumption at failure. The present paper suggests a theoretical approach, based on energy considerations and mode of failure, which holds for diverse types of mechanical testing. Maximum size effect is assumed to be associated with failure through development of a single fracture plane and a lack of size effect is associated with failure affecting a volume of material. The quantitative expression of the theoretical approach is based on the relationship between load at failure (P) and cross sectional area of the specimen (A):P=KA ⁿ wheren expresses size effect andK is a constant.The theoretically lowestn-value equals 0.75 expressing maximal size effect whereas the theoretically highestn-value is 1.00 expressing lack of size effect.n-values evaluated from published data for various mechanical tests indeed lie between these limits and appear to be related to type of test, material properties and specimen shape. Point load tests and Brazilian tests are generally associated with large size effects. This may be attributed to the test conditions which promote development of single fracture planes. On the other hand relatively small size effects are generally found for uniaxial compression tests. Failure in these tests characteristically occurs through multiple fracturing and crushing and therefore affects a volume of material rather than a single plane. Brittleness seems to be associated with large size effects and ductility with small size effects.With 4 Figures     

On statistical models for fission track counts

The statistical basis for the usual analysis of fission track counts obtained by the external detector method is discussed and illustrated with examples. A consequence is that if any observed correlation between counts of spontaneous and induced tracks is due to heterogeneity in the density of uranium, then the model proposed by McGee and Johnson (1979)for assessing the experimental error is inappropriate and results based on it could be misleading. The same remark applies to the method proposed by Johnson, McGee, and Naeser (1979).     

Grain size distribution expressed as tanh-functions

The grain size distribution within a unimodal sediment can be described as a lognormal distribution when the distribution is formed by only one process. However, most sediments are formed by more than one process giving polymodal sediments. Polymodal sediments have to be described as the sum of several normal distributions, one for each process involved within the formation. Grain size distributions are usually interpreted with the help of graphical methods. Interpretations of polymodal sediments require mathematical methods. In mathematical terms a unimodal sediment can be described as a tangential hyperbolic function (tanh) and a polymodal sediment can generally be described by the sum of two or three tanh-functions. The tanh-method is a tool for identifying and estimating the number of modes within a grain size distribution and helps interpret the processes involved within the formation of a deposit. The mathematical method can also be used to computerize sediment data, allowing storage with just a few numbers. Different samples can easily be compared and classified. Also, this method could be a valuable tool for calculations of various sediment parameters both in geotechnology and hydrogeology.     

Extended CFD–DEM for free‐surface flow with multi‐size granules

Computational fluid dynamics and discrete element method (CFD–DEM) is extended with the volume of fluid (VOF) method to model free‐surface flows. The fluid is described on coarse CFD grids by solving locally averaged Navier–Stokes equations, and particles are modelled individually in DEM. Fluid–particle interactions are achieved by exchanging information between DEM and CFD. An advection equation is applied to solve the phase fraction of liquid, in the spirit of VOF, to capture the dynamics of free fluid surface. It also allows inter‐phase volume replacements between the fluid and solid particles. Further, as the size ratio (SR) of fluid cell to particle diameter is limited (i.e. no less than 4) in coarse‐grid CFD–DEM, a porous sphere method is adopted to permit a wider range of particle size without sacrificing the resolution of fluid grids. It makes use of more fluid cells to calculate local porosities. The developed solver (cfdemSolverVOF) is validated in different cases. A dam break case validates the CFD‐component and VOF‐component. Particle sedimentation tests validate the CFD–DEM interaction at various Reynolds numbers. Water‐level rising tests validate the volume exchange among phases. The porous sphere model is validated in both static and dynamic situations. Sensitivity analyses show that the SR can be reduced to 1 using the porous sphere approach, with the accuracy of analyses maintained. This allows more details of the fluid phase to be revealed in the analyses and enhances the applicability of the proposed model to geotechnical problems, where a highly dynamic fluid velocity and a wide range of particle sizes are encountered. Copyright © 2015 John Wiley & Sons, Ltd.     

Comparison of Two Probability Distributions Used to Model Sizes of Undiscovered Oil and Gas Accumulations: Does the Tail Wag the Assessment?

Undiscovered oil and gas assessments are commonly reported as aggregate estimates of hydrocarbon volumes. Potential commercial value and discovery costs are, however, determined by accumulation size, so engineers, economists, decision makers, and sometimes policy analysts are most interested in projected discovery sizes. The lognormal and Pareto distributions have been used to model exploration target sizes. This note contrasts the outcomes of applying these alternative distributions to the play level assessments of the U.S. Geological Survey's 1995 National Oil and Gas Assessment. Using the same numbers of undiscovered accumulations and the same minimum, medium, and maximum size estimates, substitution of the shifted truncated lognormal distribution for the shifted truncated Pareto distribution reduced assessed undiscovered oil by 16% and gas by 15%. Nearly all of the volume differences resulted because the lognormal had fewer larger fields relative to the Pareto. The lognormal also resulted in a smaller number of small fields relative to the Pareto. For the Permian Basin case study presented here, reserve addition costs were 20% higher with the lognormal size assumption.     

Estimates of parameters of grain-size distribution from weight frequencies

When grains of a sediment sample are separated by sieving into a number of size classes, the weight of the grains belonging to a particular size-class is called the weight frequency of that class. That the weight frequencies cannot be used as simple frequencies for the calculation of the mean and standard deviation of size of grains is well known. A method is developed in this paper for estimating these two as well as a third parameter, named shape parameter, by minimizing a quadratic form that arises naturally as an analogue of the ² statistic. Two fully worked out numerical examples, with simulated data, are presented to illustrate the method. A computer program in FORTRAN language is also appended. Comparative study shows that the quicker conventional method used by geologists may produce reasonably good estimates of standard deviation when the sample size is large, but the estimates of mean may show large deviations.     

A note on the generation of periodic granular microstructures based on grain size distributions

This short communication discusses an algorithm suited for the generation of periodic microstructures of granular media. Its particular features are a user‐defined grain size distribution, a representative volume element which is intrinsically periodic ab initio and a user‐defined termination criterion, controlled by an increase of volume fraction. For low densities our particle packings resemble fluids or gases, while we aim to reach for rather dense particle packings, similar to granular solids. The generated microstructures can thus be readily incorporated into large multiscale simulations, e.g. on the integration point level of a finite element analysis of a particular sand or concrete. The individual grain size distribution of the granular medium is incorporated through the introduction of different growth rates governing the final particle size distribution. We briefly sketch the generation of the representative volume element within a serial event‐driven scheme and demonstrate how periodic boundary conditions are ensured throughout the representative volume element generation process. The potential of the suggested algorithm will be illustrated through the generation of two different periodic multi‐disperse microstructures. They are based on different given grain size distributions, one for a quartz sand with a low non‐uniformity index and one for concrete aggregates classified as A32 by the German standard norm DIN 1045 to have a rather large variation in grain size. Copyright © 2007 John Wiley & Sons, Ltd.     

The point‐estimate method with large numbers of variables

Rosenbleuth's point‐estimate method has become widely used in geotechnical practice for reliability calculations. Although the point‐estimate method is a powerful and simple method for evaluating the moments of functions of random variables, it is limited by the need to make 2ⁿ evaluations when there are n random variables. Modifications of the method reduce this to 2n evaluations by using points on the diameters of a hypersphere instead of at the corners of the inscribed hypercube. However, these techniques force the co‐ordinates of the evaluation points farther from the means of the variables; for a bounded variable, the points may easily fall outside the domain of definition of the variable. The problem can be avoided by using other techniques for some special cases or by reducing the number of random variables that must be considered. Copyright © 2002 John Wiley & Sons, Ltd.     

Classification and analysis of palynodebris from the Palaeocene sediments of the Forties Field

Sedimentary organic matter behaves as sedimentary particles and this study attempts to interpret its occurrence in the Palaeocene sands and mudstones of the Forties Field of the North Sea. New nomenclature and classification of this palynodebris are presented and are used to make quantitative counts of the palynodebris from nearly 600 samples of known sedimentary character. The new categories are further refined by objective testing of the data by Principal Components Analysis and Cluster Analysis. It is argued that two types of sedimentary organic matter are of particular value because they are preferentially deposited in different environments: palynowafers are most commonly deposited in the submarine fan lobes and channel complexes and amorphous matter occurs mostly in the lower energy basin plain sediments.     

The spatial and temporal distribution of grain‐size breaks in turbidites

Grain‐size breaks are surfaces where abrupt changes in grain size occur vertically within deposits. Grain‐size breaks are common features in turbidites around the world, including ancient and modern systems. Despite their widespread occurrence, grain‐size breaks have been regarded as exceptional, and not included within idealized models of turbidity current deposition. This study uses ca 100 shallow sediment cores, from the Moroccan Turbidite System, to map out five turbidite beds for distances in excess of 2000 km. The vertical and spatial distributions of grain‐size breaks within these beds are examined. Five different types of grain‐size break are found: Type I – in proximal areas between coarse sand and finer grained structureless sand; Type II – in proximal areas between inversely graded sand overlain by finer sand; Type III – in proximal areas between sand overlain by ripple cross‐laminated finer sand; Type IV – throughout the system between clean sand and mud; and Type V – in distal areas between mud‐rich (debrite) sand and mud. This article interprets Types I and V as being generated by sharp vertical concentration boundaries, controlled by sediment and clay concentrations within the flows, whilst Types II and III are interpreted as products of spatial/temporal fluctuations in flow capacity. Type IV are interpreted as the product of fluid mud layers, which hinder the settling of non‐cohesive grains and bypasses them down slope. Decelerating suspensions with sufficient clay will always form cohesive layers near to bed, promoting the generation of Type IV grain‐size breaks. This may explain why Type IV grain‐size breaks are widespread in all five turbidites examined and are commonplace within turbidite sequences studied elsewhere. Therefore, Type IV grain‐size breaks should be understood as the norm, not the exception, and regarded as a typical feature within turbidite beds.     

Variations in rates of nucleation and growth of biotite porphyroblasts

Differences in rates of nucleation and diffusion‐limited growth for biotite porphyroblasts in adjacent centimetre‐scale layers of a garnet‐biotite schist from the Picuris Mountains of New Mexico are revealed by variations in crystal size and abundance between two layers with strong compositional similarity. Relationships between fabrics recorded by inclusion patterns in biotite and garnet porphyroblasts are interpreted to reflect garnet growth following biotite growth, without substantial alteration of the biotite sizes. Sizes and locations of biotite crystals, obtained via high‐resolution X‐ray computed tomography, document that of the two adjacent layers, one has a larger mean crystal volume (9.5 × 10^?4v. 2.4 × 10^?4 cm³), fewer biotite crystals per unit volume (232 v. 576 crystals cm^?3), and a higher volume fraction of biotite (23%v. 14%). The two layers have similar mineral assemblages and mineral chemistry. Both layers show evidence for diffusional control of nucleation and growth. Pseudosection analysis suggests that the large‐biotite layer began to crystallize biotite at a temperature ～67 °C greater than the small‐biotite layer. Diffusion rates differed between layers, because of their different temperature ranges of crystallization, but this effect can be quantified. The bulk compositional difference between the layers, manifested in different modal amounts of biotite, has an effect on the biotite sizes that is also quantifiable and insufficient to account for the difference in biotite size. After these other possible causes of variation in crystal sizes have been eliminated, variability in nucleation and diffusion rates remain as the dominant factors responsible for the difference in porphyroblastic textures. Numerical simulations suggest that relative to the small‐biotite layer, the large‐biotite layer experienced elevated diffusion rates because of the higher crystallization temperature, as well as increased nucleation rates in order to achieve the observed size and number density of crystals. The simulations can replicate the observed textures only by invoking unreasonably large values for the thermal dependence of nucleation rates (activation energies), strongly suggesting that the observed textural differences arise from variations between layers in the abundance and energetics of potential nucleation sites.     

An improved method for computer simulation of geochemical neutron activation experiments

A computer method has been developed that simulates gammaray spectra of geochemical thermal neutron activation experiments. The input for the simulation includes the designation of rock type, neutron flux, sample volume and density, irradiation time, time of detection start, duration of monitoring, and number of analyzer channels. The output of the simulation includes either two-or threedimensional computer drawn plots of the resulting (simulated) gammaray spectra. The threedimensional plots (energy vs counts per sec vs time of detection) allow the investigator to observe the apparent appearance and disappearance of photo peaks, annihilation peaks, and pair production peaks of the various isotopes as a function of time. Various combinations of irradiation, delay, and detection times result in various spectral patterns. The resultant patterns accurately simulate the real activation experiments run under the same time conditions. The simulated spectra are useful for identifying peaks obtained from real spectra, helping in radiochemical separation decisions, and, in general, in designing the neutron activation experiment.     

An experimental investigation of sand–mud suspension settling behaviour: implications for bimodal mud contents of submarine flow deposits

The settling behaviour of particulate suspensions and their deposits has been documented using a series of settling tube experiments. Suspensions comprised saline solution and noncohesive glass‐ballotini sand of particle size 35·5 μm < d < 250 μm and volume fractions, φ_s, up to 0·6 and cohesive kaolinite clay of particle size d < 35·5 μm and volume fractions, φ_m, up to 0·15. Five texturally distinct deposits were found, associated with different settling regimes: (I) clean, graded sand beds produced by incremental deposition under unhindered or hindered settling conditions; (II) partially graded, clean sand beds with an ungraded base and a graded top, produced by incremental deposition under hindered settling conditions; (III) graded muddy sands produced by compaction with significant particle sorting by elutriation; (IV) ungraded clean sand produced by compaction and (V) ungraded muddy sand produced by compaction. A transition from particle size segregation (regime I) to suppressed size segregation (regime II or III) to virtually no size segregation (IV or V) occurred as sediment concentration was increased. In noncohesive particulate suspensions, segregation was initially suppressed at φ_s ～ 0·2 and entirely inhibited at φ_s ≥ 0·6. In noncohesive and cohesive mixtures with low sand concentrations (φ_s < 0·2), particle segregation was initially suppressed at φ_m ～ 0·07 and entirely suppressed at φ_m ≥ 0·13. The experimental results have a number of implications for the depositional dynamics of submarine sediment gravity flows and other particulate flows that carry sand and mud; because the influence of moving flow is ignored in these experiments, the results will only be applicable to flows in which settling processes, in the depositional boundary, dominate over shear‐flow processes, as might be the case for rapidly decelerating currents with high suspended load fallout rates. The ‘abrupt’ change in settling regimes between regime I and V, over a relatively small change in mud concentration (<5% by volume), favours the development of either mud‐poor, graded sandy deposits or mud‐rich, ungraded sandy deposits. This may explain the bimodality in sediment texture (clean ‘turbidite’ or muddy ‘debrite’ sand or sandstone) found in some turbidite systems. Furthermore, it supports the notion that distal ‘linked’ debrites could form because of a relatively small increase in the mud concentration of turbidity currents, perhaps associated with erosion of a muddy sea floor. Ungraded, clean sand deposits were formed by noncohesive suspensions with concentrations 0·2 ≤ φ_s ≤ 0·4. Hydrodynamic sorting is interpreted as being suppressed in this case by relatively high bed aggradation rates which could also occur in association with sustained, stratified turbidity currents or noncohesive debris flows with relatively high near‐bed sediment concentrations.     

Evolution of pore structure and hydraulic conductivity of randomly distributed soluble particle mixture

The plane strain behavior of particulate mixtures containing soluble particles was investigated by conducting both laboratory tests and numerical analysis. To perform the laboratory experiments, soluble mixtures were prepared using photoelastic disks and ice disks with diameters in the ratios (D_{ice disk}/D_{photoelastic disk}) of 0.5 and 0.7, and the evolution of the force chain and pore structure was monitored during the dissolution of the ice disks. Subsequently, numerical analysis was conducted by using the 2‐dimensional discrete element method for the soluble mixtures, and it was compared with the experimental results. Additionally, parametric studies were implemented by varying the particle size ratios between the soluble and non‐soluble particles and the volumetric fraction of the soluble particles. The results of the laboratory experiments and numerical analysis demonstrate that (1) after the dissolution of the soluble particles, the pore fabric of the specimens changed, resulting in a force chain changes, local void increases, and coordination number decreases; (2) the effects of soluble particles on the macro‐behaviors of the mixtures could be divided into 3 zones based on the particle size ratios between the soluble and non‐soluble particles and volumetric fraction of soluble particles. These zones were as follows: (Zone 1)—with a small total soluble volume, slight decrease in the in situ lateral pressure (K₀), and minor increase in the hydraulic conductivity (k); (Zone 2)—with a moderate soluble particle; the dissolution generated a honey‐comb particle structure; (Zone 3)—the total soluble volume was very large, and the high volumetric fraction of the dissolving particle collapsed the pore structure, decreasing in the in situ lateral pressure (K₀) but increasing the hydraulic conductivity (k). The horizontal stress returned to almost the original level, and the internal arching formation increased significantly with the hydraulic conductivity (k).     

Surface displacements due to batter piles driven in cross‐anisotropic media

This article derives the closed‐form solutions for estimating the vertical surface displacements of cross‐anisotropic media due to various loading types of batter piles. The loading types include an embedded point load for an end‐bearing pile, uniform skin friction, and linear variation of skin friction for a friction pile. The planes of cross‐anisotropy are assumed to be parallel to the horizontal ground surface. The proposed solutions are never mentioned in literature and can be developed from Wang and Liao's solutions for a horizontal and vertical point load embedded in the cross‐anisotropic half‐space. The present solutions are identical with Wang's solutions when batter angle equals to 0^°. In addition, the solutions indicate that the surface displacements in cross‐anisotropic media are influenced by the type and degree of material anisotropy, angle of inclination, and loading types. An illustrative example is given at the end of this article to investigate the effect of the type and degree of soil anisotropy (E/E′, G′/E′, and ν/ν′), pile inclination (α), and different loading types (a point load, a uniform skin friction, and a linear variation of skin friction) on vertical surface displacements. Results show that the displacements accounted for pile batter are quite different from those estimated from plumb piles, both driven in cross‐anisotropic media. Copyright © 2007 John Wiley & Sons, Ltd.     

Volume estimate of flow-type landslides along carbonatic and volcanic slopes in Campania (Southern Italy)

Recent studies on flow-type landslides in pyroclastic deposits have been performed to identify potential source areas and the main depositional mechanisms. Interesting methods for mapping landslide susceptibility have also been proposed. Since the potential volume of flow-type landslides is a measure of event magnitude, hence of considerable use in hazard assessment, we propose a method to estimate the potential volume for the morphometric analysis of 213 flow-like landslides occurred in Campania in recent centuries. First, our data show that the height, H, of the detachment and erosion-transport zones (i.e. the difference in height between the top of source area and a point, the first break at the foot of the slope, where the deposition stars to take place and the landslide loses velocity) and the area, A _f, of the same zones are linked by a mathematical function. Secondly, only part of the entire thickness of the pyroclastic material on the slope is involved. To define the potential volumes of the flow-type landslides, we analysed slopes, both in volcanic and carbonatic contexts, considering both channelled and unchannelled flow-type landslides. The most susceptible areas are identified by using a landslide-triggering susceptibility map, and then in each case the height H was estimated. This height is the difference in level between the point on the slope with highest susceptibility and the first break at the foot of the slope. Using the statistical correlation between H and A _f, both calculated for historical landslides, we evaluate the area of a potential landslide on a slope. Finally, potential volumes are calculated by using A _f and a constant thickness of the pyroclastic cover for the whole slope. This method could represent a useful tool to detect the main areas where risk mitigation works are required.     

On the influence of wavelength-dependent light scattering on the UV-VIS absorption spectra of oxygen-based minerals: a study on silicate glass ceramics as model substances

 The Mie scattering theory shows that the presence of randomly distributed submicroscopic inclusions with narrow size distribution and a refractive index n _i in a matrix with different refractive index n _m may give rise to a λ-dependent, band-like scattering (e.g., Kortüm 1969). Intensity and spectral position of this scattering depend on a combination of several independent parameters, such as size and number of inclusions, their form and n _i /n _m ratio (Kortüm 1969). Recently, it was confirmed that at a certain inclusion size and n _i /n _m ratio the scattering bands can contribute to the UV-edge in the electronic absorption spectra of garnets, influencing their colour (Khomenko et al. 2001). In natural minerals, however, it is impossible to differentiate between individual influence on scattering of the above mentioned parameters because of complex and interconnected variations in number, size and type of inclusions. Additionally, in most natural minerals variable amounts of transition metal ions are present. They may cause UV-VIS absorption due to ligand to metal charge transfer (LMCT) in the same region where band-like scattering may occur (Khomenko et al. 2001). At least some of these difficulties may be avoided in the case of some glass ceramics where number and size of crystalline microinclusions can be controlled by varying ceramization conditions such that fine-grained homogeneous microstructures are formed (e.g., James 1982; Petzoldt and Pannhorst 1991). Thus, glass ceramics of known composition, containing different amounts of microcrystals of specified size, may serve as unique patterns for the experimental study of effects caused by submicrocrystals on bulk properties of transparent solid materials, such as scattering, UV-VIS transparency and colour. Four types of parent glasses and a series of glass ceramic materials produced from them by heat treatment were investigated using transmission electron microscopy (TEM). These materials were also studied by transmission spectrometry in the range 35 000–20 000 cm⁻¹. Different inclusions, from five to several hundred nm in size, were observed in the glass matrices depending on their compositions and heating history. These inclusions represent two groups: early very small crystals of Ti, Zr oxides and relatively large crystals of stuffed high-quartz type or keatite. The absorption spectra of the glass ceramics show largely varying long-wavelength slopes of the UV absorption. UV-edge intensity correlates mostly with the size of the inclusions and changes drastically when larger keatite-type microcrystals are growing. Small variations in the UV edges also follow the early process of Ti-phase separation and nucleation. This may be explained by Ti depletion from the glass matrix and, thus, by reducing the measured intensity of LMCT in the first co-ordination sphere of Ti⁴⁺ ions. The different yellowish colourations of unheated glasses studied here are caused by this effect, whereas developing several hundred-nm-large keatite crystals leads to a strong scattering effect and a milky colour in glass ceramics. Received: 15 February 2002 / Accepted: 11 November 2002 Acknowledgements M. Garsche and A. Kisljuck generously provided the base glass and glass ceramic samples. F. Galbert and S. Herhing-Aghte, both at the Technical University Berlin, helped with electron microprobe analyses and refractive indices measurements, respectively. The Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg, provided financial support through grant La324/35. To all these individuals and this institution our sincere thanks are due.