4. Conclusions

the particle, the PTV algorithm for non-spherical particles was applied. Figure 13d shows particles pair detection and its centroid, which allows us to determine particles' velocity for a

A settling column was used in order to observe the hologram evolution over time. Almost 100 holograms were processed each recording time. The times recorded were t = 0, 10, 20, 30, 45, and 60 min. Figure 14a shows particles distribution for holograms at time t = 0, and it can be

Figure 14c shows the distribution of particles at time 30 min, where an increase in size of flocs is observed attaining a maximum of 180 □m and a mean value of 80 □m. It is also observed that the shape of the distribution is log-normal, similar to theory. Figure 14d shows clearly a non-

Figure 13. Results of analysis of a digital hologram. (a) Original digital hologram; (b) processed digital hologram; (c)

Figure 14. Characteristics of flocs in settling column (a) frequency distribution of sizes at t = 0 min; (b) spatial distribution

of flocs at t = 0 min; (c) frequency distribution of sizes t = 30 min; (d) spatial distribution of flocs at t = 30 min.

observed that maximum size is 160 □m, with a mean diameter of 70 □m.

uniform distribution of particles in a hologram.

126 Applications in Water Systems Management and Modeling

reconstructed hologram, and (d) size and shape of detected particles.

sequence of holograms.

A model to estimate the floc settling velocity was calibrated for flocs obtained from aquaculture recirculation tanks that cultivate trout. The model was able to reproduce the values of settling velocity which varies between 0.01 and 0.025 m/s. For all the recording times analyzed there is a maximum settling velocity for flocs of diameter of 600 μm.

The representative values of the parameters used to determine fractal dimension are proposed in this research according to the experimental results. These values depend on floc density and vary with experimental time as flocs become more porous. The values found in this research apply to flocs coming from trout cultures in high level locations, i.e., 2800 masl.

The practical findings for aquaculture recirculation tanks design is that residence times should be short in order to minimize the presence of very large flocs. Middle size flocs settle faster. In designing the central settling device, which functions according to the hydrociclons principle, the up flow velocity should be less than 0.01 m/s in order to diminish the flow of sediments toward the recirculation deposit.

A method to obtain the suspended sediment concentration profiles for rivers with mainly cohesive sediments was presented. It is necessary to take some representative samples and using a rotating annular flume defines a steady state of flocculation after long-term runs. The most suitable method to analyze size and settling velocity of flocs are optical methods, PTV and microscopy. This research shows that the settling velocity can be accurately calculated with Eq. (6) in order to obtain an appropriate estimation of the Rouse number ZR. This allows us to properly determine the suspended sediment concentration profiles in rivers carrying a large amount of cohesive sediments.

Non-intrusive optical techniques are a suitable tool to characterize cohesive sediments, because they do not destroy flocs and allow for microscopic analysis. More advanced optical methods, like DHPIV, are showing good results for floc size and shape determination, thus in the future they will be the best method for cohesive sediment analysis.
