Measuring Nuts and Coffee Beans Using the ColorTrend HT

The color of nuts and coffee beans can be assessed immediately after roasting or cooking, while still on the line, using the ColorTrend HT. Such measurements provide information about product quality and sufficiency of cooking or roasting. Color measurements can help reduce costs associated with color complaints and can indicate where there is a process upset, such as a temperature set too high or too low. On-line measurements allow you to react quickly to such upsets and to measure product constantly, rather than just intermittently in the laboratory.

There are a number of process variables that can impact the color of nuts and coffee beans. Typically, control parameters such as operating temperature, line speed, and dwell (or resident) time all have an effect on color. In order to determine how these variables affect color, it is important to first get real-time measurements of color that can be tracked accurately and consistently from the beginning stages of a run through to the finished product. This information helps determine the causes and effects of process variation, allowing prompt corrective action to minimize off-quality product, increase production throughput, and maximize equipment usage. As a result, plant operating costs are ultimately reduced and overall product quality is improved. On-line measurements also result in reduced laboratory labor and material sampling costs as the laboratory quality control function is reduced to infrequent auditing.


The ColorTrend HT

The Application

Typically, the ColorTrend HT sensor is mounted above the conveyor belt just after the roaster or oven. (Mounting over the oven band is not recommended.) The nuts or beans pass under the sensor in layers or piles and the instrument measures the nuts that it "sees" and provides colorimetric data that is updated frequently.

Recommended Color ScaleCIELAB or Hunter L, a, b as a full color descriptor
Recommended Illuminant/ObserverC/.

Measurement Method

1. Position the ColorTrend HT over the product on the line. Tilt the sensor 15° if the product is oiled or has a glossy appearance. A bed depth of 0.5-inch of product is required for the batch of product to appear opaque to the instrument. This can be achieved by using a plow or other planing device to flatten and spread the product to a constant depth. The ColorTrend HT viewing window should be 3-4 inches from the product surface. The instrument can compensate for height, so small differences in the presentation of the product to the sensor is not of concern as long as the minimum bed depth of 0.5-inch is maintained.

2. Configure a product setup for each type of product with the following parameters:
...*Parenthetical entries refer to terminal wording where it differs from that of the Utility Software.

...Set-up NumberYour choice
Set-up Name (Product Name)Your choice
Color ScaleCIE L*a*b*
Product Target Values (Target Values)Establish using the read button with an ideal product standard, if desired (Enter values for ideal product standard, if desired). For roasted cashews, a reasonable specification is an L* between 48 and 53. For roasted peanuts, try L*=50-70, a*=-10-10, b*=40-60.
Chart Color Index (Display Height)No (Yes)
Hitch (Enable Hitch)Off (No)
Average1
Read ModeTime Average (Time)
Sample Time (Read Time)20 seconds
Hold Time1 second
Data ModeAbsolute
Datalogging Interval1 second or more
Background DistanceRecalculate for this product
Cut-off DistanceRecalculate for this product
Units (Height Units)Your choice
Alert LimitEnter color values beyond which alert should be given
Alarm LimitEnter color values beyond which alarm state should be activated
Analog Outputs (Channel Centers/Channel Spans)If analog outputs available, enter appropriate values
FilterEnabled 0.50
Color Tracking (Track Color)On and enter limits (thresholds)
Tilt Angle15°
Standardization IntervalIf autocalibrator available, enter appropriate choice
AC FrequencyEnter appropriate value
White Tile ValuesDo not change values shown

3. Begin the run.

4. Modify the target values and limits as necessary to optimize the process. The ideal situation is one where the reading of defective product immediately causes the alarm state to be activated, but acceptable products continue to be read as usual.