150 to 6750 m

The first five ranges are the most useful for studying metal powders used in the P/M industry. The actual sensor dimension (height in Fig. 2) is one- larger than the largest particle size for which the sensor is rated; consequently, sensor blockage is minimized (Ref 3, 4).

The microprocessor scales the sensor output pulses and sorts them into 23 size ranges, or channels, in a geometric progression of 1 to 1.18 by using 24 analog comparators. Within the limits of the physical size of the sensing zone, the thresholds can be adjusted to change the widths of the channels to provide more or less detail of certain areas of the particle size distribution, or to extend the range slightly. The threshold adjustment circuits are housed on a printed circuit board that is changed when the sensor is changed to monitor a new particle size range, for example.

The system memory accumulates the number of pulses generated between each pair of thresholds, and the central processing unit reduces the data, calculates the output data, and controls the printer and plotter output functions.

Accuracy and Interpretation. Instrument-related factors must be considered, even with perfectly dispersed, spherical particles (Ref 5). The oversized particle indicator has another purpose in addition to detecting large particles that exceed the range of the sensor--it indicates an overconcentrated solution. If several particles overlap while passing through the sensor, the oversize indicator flashes, and the oversize counter is incremented. If this tendency occurs frequently (more than once or twice per analysis), the solution may need to be diluted, or the sensor may be blocked. If dilution does not eliminate the oversize indication, but does reduce its frequency by an amount proportional to the dilution, the large particles are most likely part of the actual distribution.

Even if dilution eliminates the oversize indication, small particles may overlap and skew the distribution toward the coarse side. To eliminate this occurrence, a maximum recommended number of particles per cubic centimeter of liquid is given for each sensor. High-concentration sensors that allow twice the concentration to be analyzed by restricting the volume of fluid, and therefore the occurrence of two particles in the sensing zone at the same time, are available.

To increase detail in some parts of the distribution and to extend the analysis range, channel thresholds can be adjusted. Adjustable thresholds, however, can slip out of adjustment. The appearance of a sharp peak and an adjacent dip in the same place on several different derivative distributions may indicate the need for threshold adjustment. When this condition occurs, counts are being placed in the channel above or below the proper one. For the cumulative percent under or over distributions, the effect is cancelled when both of these channels are passed. Threshold adjustment can be performed with a printed circuit extender board and a digital voltmeter. The overall detection level of the instrument should be checked periodically using calibrated spheres.

Settling of particles in the sample reservoir can be reduced by using a more viscous fluid such as glycerine. Pure glycerine, however, may be too viscous for the vacuum to draw and also may prevent the dispersion of particles that tend to agglomerate. Glycerine can be thinned with water, but antisettling benefits are diminished.

A more effective way to eliminate settling is to use reservoir shapes, such as round-bottom beakers, and stirrer configurations that maximize turbulence and prevent "dead" spots in the stirring current, without the generation of foam or bubbles that appear as particles in the sensor. Even if reservoir suspension is ideal, particle settling can still occur in the sampler tube (the tube connection between the reservoir and the sensor).

The higher the linear velocity of the fluid, the larger the size of the particles that can be entrained and carried along. The linear velocity of the fluid in the sampler tube is determined by the rate of flow through the restricted sensing zone and by the diameter of the sampler tube. The small sensor has a smaller sensing zone; consequently, the flow rate through the sampler tube is lower, causing the larger particles to settle before reaching the sensor. This tendency can be controlled by fitting the sensor with a smaller sampler tube. This increases the linear velocity of the fluid and keeps the larger particles entrained.

Small particles may have a tendency to loosely agglomerate due to moisture, or van der Waal's forces, for example. Stirring the water suspension of such a powder may be only partially successful in separating these particles. Addition of a surfactant or type IC dispersing aid to the powder before water is added helps the water wet the particles and allows them to separate with the stirring action. Ultrasonically vibrating the solution for several minutes also helps separate particles.

Effect of Lubricants. Many metal powders for P/M use are prelubricated by the powder manufacturer. Typical lubricant additions range from 0.5 to 1.0 wt%. While the presence of lubricants or other chemical additives or particle surfaces may inhibit wetting and cause agglomeration, lubricants have another significant effect. Lubricants or other chemical additives are much less dense than metal particles. Consequently, a small amount of lubricant by weight may contribute greatly to the particle count (population) distribution.

When the volume distribution is calculated from the population distribution, the small amount of lubricant (by weight) is weighted the same (per particle) as the metal powder. In this case, the volume distribution is not the true weight distribution. The same occurs during the analysis of a powder mix in which the constituents have different specific gravities. Single-component powders provide the best test results, unless only the population distribution is desired. Powders should be unlubricated, or the lubricant should be removed prior to analysis.

0 0

Post a comment