|
|
|
 |
||||
Alabama is the second largest producer of seedlings in the country, growing more than 237 million seedlings per year in 1997. The number of seedlings per nursery varies between 2.5 and 6 million per year. An important characteristic of a seedling is the root collar diameter (RCD) as shown in Figure 1 (right). It is a measure of the seedling’s survivability potential and also the parameter on which the seedling price is based. Nurseries measure the RCD of thousands of seedlings each year to obtain a statistical measure of the crop value in the field. Traditionally,
the RCD is measured by hand with a slide micrometer, a tedious and time
consuming task. AAES researchers in the Auburn University Department of
Biosystems Engineering developed an alternative that measures the RCD
automatically. The user can place a seedling on a measurement table and
at the push of a button the RCD is measured and stored in a spread sheet. The
principle of measuring the seedling diameter is shown in Figure
2. A single diverging infrared light beam is used in combination
with two optical sensors. When
a seedling is blocking the light beams, the signals emerge. The velocity
of the seedling is obtained from the time difference between interruption
of layer 1 and 2, respectively. The RCD follows from multiplying the velocity
by the total time a seedling blocks either light layer. The constant
b is the distance between the light layers, which is a constant
as long as the distance from the seedling passage path to the sensors
is constant. The formula used to determine the RCD is:
D:
Diameter oftheobject (RCD) (m) b:
Distance between centers of optical sensors (m) The
timing signals are measured using a dedicated board and automatically
put in a spread sheet. The
first prototype of the sensor consisted of an arrangement where the user
manually slid the seedling along a set of sensors (Figure
3). This concept proved to be cumbersome, because a human being is
not capable of maintaining a constant velocity during measurement. The
assumption that the distance between the seedling and the sensors is constant
does not apply here; it depends on the diameter itself. Another fact is
that the orientation of the seedling is often not exactly vertical, which
means that the measured diameter is larger than the true RCD.
The manual sliding device was calibrated with drill bits of diameters
[shown in millimeters (mm) and also inches]: 3.17 mm (1/8 inch), 3.57
mm (9/64 inch), 3.97 mm (5/32 inch), 4.34 mm (11/64 inch), 4.74 mm (3/16
inch), 5.12 mm (13/64 inch), 5.53 mm (7/32 inch), 5.92 mm (15/64 inch),
6.33 mm (1/4 inch), 6.69 mm (17/64 inch), 7.91 mm (5/16 inch), and 9.46
mm (3/8 inch). The diameters of the drill bits were measured with a slide-micrometer
with an accuracy of 1/100 mm. The calibration curve is shown in Figure
4. The standard deviation among 10 measurements of the smallest object
(3.17 mm) was 0.01 mm (0.3%) and increased linearly to 0.07 mm (0.7%)
for the largest diameter (9.46 mm). One
problem with the manual sliding device is that the distance between the
sliding path and the center of the seedling is not constant. This will
introduce errors because the optical beam used is not parallel. This error
can be virtually eliminated by placing the sensors in a balance configuration
as shown in Figure 5. A dual
opposing set of sensors is placed such that errors due to lateral movement
of the seedling are compensated by the opposing sensor set. To
eliminate the errors introduced by the skill of the user, a new concept
was developed where the seedling is placed on a table and the sensors
themselves are moved. Proper alignment of the seedling is aided by projecting
a laser line on the measurement table. Also, the balanced configuration
shown in Figure 5 was incorporated in this new design. To move the sensors
with respect to the seedling, the device uses a counterweighted bridge
principle as shown in Figure 6.
The seedling is the horizontal wood-textured bar and the sensors are the
white dots in the black square. In idle, the sensors are below the surface
of the table. When released, the counterweight on the left drops freely
and the sensors move upward under a constant acceleration. Because of
this acceleration a new formula was developed to obtain the RCD as follows:
b
:Distance between centers of optical sensors (m)
a:
Acceleration of the sensors (ms-2) The
sensor as developed here could have a great potential for nurseries to
reduce a tedious task inexpensively at a total hardware cost of around
$600. |
|||||
: Flank
separation signal (s)
: Pulse
width of a single sensor signal (s)
:
Diameter of the object (RCD) (m)