.jpeg)
Sediment and water leave the basin and enter the fluxometer through the white funnel at the top. The looped hose coming out of the bottom right of the collection cylinder is a self-actuating siphon. As fluid rises above the apex of the loop, the siphon 'closes' and sucks fluid out of the collection cylinder. Three electrodes have been inserted into the collection cylinder. The black electrode near the funnel is a ground, the white electrode carries an applied voltage from datalogger, the red electrode closes the circuit when fluid contacts it. The datalogger records the mass of the apparatus (supplied with a thin beam load cell, partly visible as a bronze fixture at the base of the collection cylinder) when fluid contacts the white electrode, and again when fluid contacts the red electrode. Differencing these measurements yields the mass contained between the electrodes. I measure the volume between electrodes prior to each run, and calibrate the load cell (its signal is voltage) with known masses. With known volume, known mass, time required to fill collection volume, and assuming the density of solids = 2.65 (quartz; substrate is silica flour) and the density of water = 0.997 (varies slightly with temperature), the mass flux rates for water and sediment can be calculated. Hence, each data point is an average flux over the collection volume (usually around 90 to 130 cm^3). It's a rather simple system, gets dirty and is subject to wear and tear, and is susceptible to extreme vibrations in the laboratory, as well as more locally from momentum in the falling fluid. I have routed the falling fluid from the basin down a few strings, which seems to dampen and smooth the entrance of fluid into the collection cylinder. The datalogger can take measurements every 1/8 second. To remove fluctuations in the mass measurements, I take 5 measurements when each electrode closes, and average them. The signal is still a bit noisy, even over a collection time ranging from 10 to 30 seconds. Some of the noise creeps in from slightly different closures of the circuits, perhaps because of a drip hanging from the electrode.
Chris Ellis, one of the brilliant engineers on staff at St. Anthony Falls laboratory, helped me with the design of the fluxometer, and wrote the circuit monitoring code that operates measurement collection and storage in the datalogger. Thanks, Chris!