Improvement in Bean Quality During Harvest

1996/97 CORE PROJECT PROGRESS REPORT

Project Title:Improvement in bean quality during harvest

Investigator:John Smith, Machinery Systems Engineer

Objectives for 1996 Harvest Season:

1.Pursue application of the automated quality inspection system developed by Perstorp Analytical Corporation for quality inspection of dry edible beans.

2.Evaluate the prototype harvest system introduced by Pickett Equipment Co. for harvesting dry edible beans.

Background

Visual inspection of dry edible beans for the quality parameters of color, size, shape, and mechanical damage can be very time consuming, and is subject to inconsistency of human output.  Bean processors and buyers develop inspection procedures appropriate for their particular operations.  A particular concern for this project is that most mechanical damage to dry edible beans occurs during the combine operation.  Other than a very ‘course’ visual inspection, the combine operator generally does little to determine mechanical damage caused within the combine until the beans are delivered to the bean receiving facility.  The operator could take a measured sample, sort or screen broken beans and bean parts, weigh the good and broken beans, and calculate the percent mechanically damaged beans.  This process requires an accurate scale and time and is rarely done.  Another option available to the combine operator is to stop combining, collect  a bean sample, and take the sample to the buyer for inspection.  This takes excessive time and is rarely done.   Perstorp Analytical Corporation recently developed image analysis instrumentation for the automated inspection of wheat and other common grains.   This equipment is fast, consistent, and is being considered for official grain inspection of certain cereal grains.   Perstorp Analytical Corporation was approached to determine if this instrumentation would apply to inspection of dry edible beans.

In 1996, Pickett Equipment Company introduced a new dry edible bean harvest system.   This system makes the next somewhat logical move to incorporate their successful ‘One Step’ rod cutter with the combine for a one operation harvest.  One current trend in bean harvest is to cut beans when they are much drier, reducing the period of time between cutting and combining.  In certain situations, some growers cut in the morning and combine in the afternoon to reduce the risk of wind or rain damage to the beans after cutting.  The purpose of this portion of the project was to evaluate a proto-type machine in actual harvest conditions.

Results

Seed Inspection

Personnel at Perstorp Analytical Corporation expressed interest in evaluating their current instrumentation for inspection of edible beans.  Bean seed samples were prepared at the University of Nebraska Panhandle Center for initial evaluation.  These samples included all good seed as determined by a visual inspection, all broken seed, foreign material commonly found in bean samples, and samples composed of different proportions of good seed and broken seed.  Separate sets of Great Northern and light red kidney samples were sent to Perstorp.

The analyzer unit consists of a sample transport mechanism which divides the sample into individual seeds, a balance for weighing an accumulation of seeds, a color video camera, and a computer with software.  The software requires initial instructions for defining good seed, broken seed, color differences, and foreign material.  These instructions are accomplished by running many variations of each category through the instrument for definition.  When an actual sample is run through the machine, the software compares each seed or object with pre-defined images.  Each object is sorted into appropriate categories and weighed, automatically.

Perstorp used the Great Northern bean samples to provide very preliminary instruction to the machine.  Five samples, composed of different proportions of broken and good seed, were inspected by the machine.  Results were as follows:

Perstorp made up a sixth sample and ran the same sample through the instrument five times to develop a sense of repeatability.  Results were:

The current instrument was designed for small seeded grains such as wheat and rice.  To inspect the bean seed, Perstorp disassembled part of the instrument to accommodate the larger seed and modified part of the routine procedure.  However, the above very preliminary results were encouraging.  With only limited reference imaging, the machine results were very close to visual inspection for five samples with different proportions of broken seeds.  Repeatability was also very good.

To proceed with further evaluation, the instrument would require physical modification to accommodate the larger bean seed.  Perstorp decided at that time (November, 1996) that their resources could better be used to develop and refine their instrumentation for grains that appeared to offer more potential for immediate financial return to their company.  They decided not to do further work with dry edible bean seed for at least a period of one or two years.  The project stopped at this point.

This instrument is currently being used for official inspection of wheat, rice, and perhaps several other cereal grains.  Based on this very preliminary data and on information supplied by Perstorp, the instrument is fast, accurate, eliminates human subjectivity, and does have application for inspection of dry edible beans.

Evaluation of Pickett Equipment Co. Harvest System

Picket Equipment Co. manufactured two prototypes of their new harvest system for the 1996 harvest season.  One was field tested in Idaho.   The second prototype, a 20 ft. wide unit,  was brought to Nebraska and assembled to the Deere and Co. Model 9500 combine loaned to the University of Nebraska by Deere and Co.  The role of the University of Nebraska was to observe the performance of the system and make recommendations for future development.  Targeted areas of concern included ability of the combine to lift and handle the header without damage to the combine, hydraulic capacity of the combine, height control of the header, field loss, and capacity of the header-combine combination.

The Pickett harvest system includes desiccation of the field for harvest (if necessary) and the combine header.  The combine header is basically a Pickett rod cutter mounted to the front of the combine.  The header includes a rod undercutter, a finger pickup drum to lift the plants from the rod to the header auger, and the header auger which brings the plants to the center of the header and the feederhouse chain.  Height control wheels and vine dividers were mounted to each side of the header.

The header was mounted to the combine at the Panhandle Research and Extension Center.  The necessary hydraulic connections were made and the unit operated in a bare field.  The combine was able to lift, turn, and push the header in the field without problem.  Hydraulic capacity was adequate.  The combine and header were moved to the Holyoke, CO and Grant, NE area where beans were being harvested early in the season.  The combine was operated in a total of seven fields, five of which were observed by personnel from the University of Nebraska.   These fields were harvested during a period extending from August 28, 1996 through September 24, 1996.  Other than the first field, soil conditions were unusually wet after the first week of September.

In general, performance of the Pickett Harvest System could be described as good in two of the fields and unacceptable in the other three fields that were observed by the University of Nebraska.  Following are main problems and positive aspects of the unit tested in 1996:

-The system would not handle wet soil.  The rod does not cut the bean root but instead pulls it from the soil.  Any attached soil is taken into and through the combine.  The Deere model 9500 combine has no provision for effectively eliminating wet soil.  The result is, at best, the soil will end up in the grain tank.  At worst, the soil will build up in the clean grain elevator, and grain tank loading auger.  In one field with a high population of volunteer corn, the grain tank loader auger was plugged and sheared off the drive pin after only several hundred yards through the field.  This was an excessively wet field with an excessive amount of volunteer corn.

-Height control for the header was not adequate.  As soil conditions varied across the field, the rod depth varied, causing poor cutting or lifting into the finger drum.  If soil conditions were uniform and ridge shape was uniform, depth control was adequate.

-The header would not adequately pick up plants where the plant population was low or where plants were poorly developed without ample vines.  Where the plants were normally large and with some vine structure, pickup was adequate.

-All fields had been sprayed with paraquat to desiccate both weeds and green bean plants.  Not all plants had been killed.  The green foliage created a potential problem for moving material through the combine and for staining the bean seed.

-Field loss was very low when the machine was operating correctly.  In these areas field loss was estimated to be less than ½ bu/A.

-Field capacity or field speed was considered good when the machine was operating correctly.  Field speed was generally between 4 and 6 mph.

-Generally the bean seed was between 13 and 15% moisture which would facilitate lower seed damage within the combine if the bean receiving facility can accommodate this moisture level.

-The 20 ft header required considerable power to move through the soil.  This combine had the rear wheel drive option and it was required in most fields to maintain adequate traction.

-Growers who observed the system seemed generally interested in this harvest concept.  Major points of interest included elimination of one operation, elimination or reduction of risk from wind and rain, and potential to combine at a higher seed moisture level.

Upon review of the 1996 experience with the prototype in Nebraska and Colorado, Pickett Equipment Co. planned to address four problems:

1.Excess soil taken into the combine.  They plan to reduce this by better rod depth control and by a development of screens in the clean grain cross auger, currently being studied by the University of Nebraska Panhandle Center.

2.Depth control.  They plan to strengthen the frame, and modify the depth gage wheel mechanism.

3.Missing small plants.   Addition of more rows of fingers in the pickup drum should allow better pickup.

4.Excessive power requirement.   Improved rod depth control should allow the rod to be operated at less depth, requiring less power.

5.Uniform killing of weeds and bean plants with approved desiccants.  Pickett planned to visit with weed control specialists and chemical companies to learn more about proper application of approved desiccants.

The concept prototyped by Pickett Equipment Co. appears to meet the needs of some growers who are looking for a one pass harvest operation.  Further development work must be done to control depth of the rod, eliminate excessive soil passing into the combine, improvement in the pickup drum, and better performance of the desiccant.   Enough grower interest has been expressed to Pickett Equipment Co. that they plan to develop improved prototypes for testing in 1997.

Acknowledgment

I would like to thank the Nebraska Dry Bean Commission for providing partial funding for these projects.