Printer Friendly

Cluster criteria: Auburn develops optical sensor for measuring discrete particle flows.

Two centuries ago, Italian scientist Giovanni Battista Venturi developed a mass flow measurement device for incompressible fluids based on the characteristic pressure drop that fluids exhibit when accelerated to a higher velocity.

As this concept has evolved, measuring mass flows of discrete particles has become a common agricultural practice. Many materials--including fertilizers and manure and feed pellets--are transported and dispersed in granular form.

Although mass flow measurement devices are available for spraying applications, no feedback device exists to measure a granular fertilizer spreader's true output. This lack of a feedback mechanism for granular spreaders has for decades limited the uniformity that could be attained in aerial and ground-based applications.

Another area that involves mass flow measurements is in monitoring peanut, cotton and grain yields. Traditional yield monitoring methods, particularly for cotton, are based on a single light source--usually a laser beam--and an optically matched receiver. Mass flow information is retrieved from the amplitude of the receiver signal.

However, this method is prone to errors due to alignment problems, vibration, temperature effects and contamination. A poorly understood light interruption mechanism uses an empirical approach that relies on calibration to make ends meet.

As an alternative to this traditional method, an optical mass flow sensor has recently been developed by the department of biosystems engineering at Auburn University. The optical mass flow sensor's features include:

* A spatially non-coherent light source is used instead of a laser and eliminates alignment and vibration problems.

* The timing signal produced by clusters of particles passing the sensor is used instead of the receivers' analog output. This timing is insensitive to light level--as long as there is sufficient intensity--which eliminates contamination and temperature effects. Also, the sensor data's digital nature reduces information storage needs.

* The sensor has 30 interruption lines spaced 0.04 inch (1 millimeter) apart instead of a single laser beam. The interruption area is bounded so extrapolation to larger areas is simple.

* The well-developed theory of interruption processes involves particle cluster arrival at the sensor described using a Poisson distribution--as used in queuing theory. The theory can also describe the probability of cluster emergence with certain lengths.

* The sensor needs no calibration when considered for use in a hypothetical situation where particles have uniform diameter. In a more realistic case of particles with different diameters, a single calibration constant will be needed based on the particles' diameter distribution.

How the system operates

Particle clusters are dropped through the optical sensor (Figure 1) and interrupt light layers formed by two sensor arrays. Each array contains 30 OptoSchmitts--digital on/off switches--spaced 0.4 inch (10 millimeters) apart. When the light beam hits an OptoSchmitt, output is high. When the light beam is blocked by a cluster, output is low. All OptoSchmitts in an array are connected with a logical "AND" function, which means that if one in an array is blocked, the entire array registers low output. This arrangement forms two light layer planes rather than one interruption line.

[FIGURE 1 OMITTED]

Cluster length is obtained from the event timing during light beam interruption. Cluster velocity is measured from the time it takes the cluster to travel from the top light layer to the bottom. Cluster length is determined using the velocity combined with the total time a cluster blocks either array. The formula for this determination is:

CL = b[[[Delta][t.sub.p]/[[Delta][t.sub.f]]][m]

where:

CL[m] Cluster length

b[m] Distance between light layers

[DELTA][t.sub.f][m] Total time to move cluster from top layer to bottom layer

[DELTA][t.sub.p][s] Total duration cluster blocks either light layer

Because the OptoSchmitts could allow smaller particles to slip through without being detected, a lens system has been added to magnify the cluster shadow by a factor 10, similarly to a slide projector principle.

Reconstructing the image

Graphs of sorted cluster lengths in a typical experiment involving 4,000 identical particles identified 2,187 singles, roughly 500 doubles and a small number of triples and quadruples. The sensor detected a total 2,962 clusters. The original 4,000 particles in the experiment can be reconstructed using queuing theory based on the formula:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where:

N estimated number of particles

E total number of measured clusters

[N.sub.0] total number of measured Singles

[alpha] material specific constant ('l' for identical diameter particles

This formula can be roughly verified by inserting the number of the previous experiment data using [alpha] = 1 (identical diameter particles):

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Because the total number of original particles has been reconstructed, mass flow can be computed by translating the mean particle diameter into a volume and multiplying by the material's true density.

The normalized number of reconstructed particles is shown in Figure 2. The graph's left side shows that the flow density is low so a high accuracy of reconstruction may be expected as nearly every particle is measured individually. On the graph's right side, density becomes high and accuracy remains high. The total error in mass flow measurement was 3 percent, even under high mass flow conditions.

[FIGURE 2 OMITTED]

In precision agriculture, enabling technologies such as global positioning systems and global information systems have matured. However, sensor technology is still in puberty. A need exists for robust, reliable sensors that can function in harsh environments and require little or no calibration. This goal can be achieved by shifting emphasis toward fully digital designs and away from analog designs that use calibration as a safety net.

[ILLUSTRATION OMITTED]

ASAE member Tony E. Grift is a professor in the department of biosystems engineering, Auburn University, 213 Tom Corley Building, Auburn, AL 36849, USA; 334-844 3545, tegrift@eng.auburn.edu.
COPYRIGHT 2002 American Society of Agricultural Engineers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Grift, Tony E.
Publication:Resource: Engineering & Technology for a Sustainable World
Geographic Code:1USA
Date:May 1, 2002
Words:954
Previous Article:Old drains, new challenges: Illinois engineers study base flow in tile-drained watersheds.
Next Article:Fertilizer sense: new techniques boost nitrogen efficiency for corn.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters