Progress 07/01/13 to 02/28/14
Outputs Target Audience: The target audience for the proposed particulate mass flux sensor is the food process industry. We were able to reach out to several large companies such as Pepsicola, Kraft, Tate & Lyle, General Foods, etc., and inform them about our product. We were able to get endorsement letters from two of these companies for our Phase II effort. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? At Purdue University, a graduate student helped in measuring the mass attenuation coefficient of food ingredients. How have the results been disseminated to communities of interest? Preliminary results from the project have been shared with a few processed food manufacturers. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts What was accomplished under these goals?
1. Introduction Continuous processes are inherently less expensive than batch processes, since it consumes lower energy and enables better control of total output. If the capability to measure particulate mass flux of particulate is available, then continuous line manufacturing of food will be feasible. This will be a significant boon to the food manufacturers as it would lower costs, while simultaneously increasing product consistency. This project demonstrated the feasibility of a mass flux particulate sensor during the Phase I work. 2 Technical Objectives and Work Tasks The technical objective of the proposed Phase I project was to evaluate the feasibility of a particulate mass flux sensor for the food industry. Three work tasks were proposed to answers these questions. The three tasks were (1) the design and construction of the X-Ray system, (2) the development of the mass flux algorithm, and (3) validation of the system. 3 Results of the Phase I Work The first task was to build an X-Ray system to obtain the mass flux of particulate. En’Urga Inc. had developed an X-Ray system to measure extinction in oil/water pipe lines under a National Science Foundation project. Essentially the sensor consists of a soft X-Ray source (10 to 50 KeV) at one end and six linear arrays (3 horizontal and 3 vertical) on the other end of a lead encased aluminum box. The horizontal arrays are used to obtain the total planar concentrations and the vertical arrays are used to estimate the velocity of the particulate flow. The frequency of the system was 200 Hz. Two different configurations of the X-Ray system were developed during the Phase I work. In the first configuration the particulate was allowed to fall freely through a pipe that passed through the hole shown on the base plate. In the second configuration, the system was tilted on its side and a conveyor belt carrying particulate was passed through the hole. The second task was to obtain the mass flux of the particulate from the extinction measurements. There are two subtasks for achieving this objective. The first is to obtain the velocity of the particulate flow from the extinction measurements. The second is to obtain the mass attenuation coefficients of various common food ingredients. Velocity was obtained using a cross correlation technique. The results show that the velocity of the particulate on a moving belt was measured within 1% of the belt speed. The second subtask is to obtain mass attenuation coefficients for various common food ingredients. Though extensive tables of the mass attenuation coefficients are available for all elements and common compounds, very little data is available for food materials. The major components of food materials are carbohydrates, proteins, lipids, salts and minerals. Foods representative of each category were selected for the determination of their x-ray linear attenuation coefficients (µ). The true density of these components was separately measured using an air-comparison pycnometer for the determination of the x-ray mass attenuation coefficients (µ/ρ) of the food components. Two instruments at Purdue University were used for the measurements. The first is a Micro CT (µCT 40, Scanco Medical Inc). The second is an x-ray digital radiograph (RapidStudy EDR6, Sound-Eklin, CA). In general, the attenuation coefficients decrease with an increase in X-ray energy for all the food components. The salts have the highest attenuation coefficients and the oils have the least amongst the food components. The interesting thing to note is that oils have a lower attenuation coefficient than water. The attenuation coefficients for the corn vary with the form of the corn. For instance, Corn-Wet germ has lowest attenuation coefficient while corn Fiber has the highest. Furthermore, the variation in attenuation for the different forms of corn has different slopes. This implies that utilizing extinction at multiple energy levels may provide information on individual components from a multi-component mixture. The last task of the Phase I project was to evaluate the system experimentally. A commercial feeder from Tec-weigh (Model QC-10) was used to feed the particulate down a 4 inch tube or to place the particulate onto the conveyer belt. The feeder was first calibrated to provide the mass flux of the food items that were tested during the Phase I work. The mass fluxes of different ingredients are reasonably linear with the input frequency of the particulate feeder. Then the sensor was turned on and extinction data collected for flour, milk powder, sugar, rice, corn-germ, and corn-gluten in the chute configuration. The extinction levels were typically less than 10%. The extinction data ws radially integrated. Using the total extinction, the mass attenuation coefficient, and the velocity of the particulate obtained from the sensor, the mass flux of particulate was estimated. The results indicate that the X-Ray sensor estimated the mass flux very well for all the basic food ingredients. The average absolute error across the entire set of measurements is 5%. The next set of experiments involved food ingredients being moved via a conveyor belt. The same food ingredients were dispensed onto the belt using the particle feeder. In all cases, the speed calculated by the X-ray sensor was within 1% of this value. This is very important since this implies that in industrial applications, the customer does not have to know the speed of his conveyor belt or ensure that the speed is maintained at a constant value for accurately estimating the mass flux of the particulate. The absolute error over these twenty five measurements (5 measurements for each food ingredient) is 4.9%. Based on these measurements, it is very clear that the proposed sensor can provide the mass flux of particulate in either configuration within an accuracy of 5%. On closer examination of the data, it is clear that the highest flow rates also have the highest errors associated with them. This is probably due to beam hardening (soft rays being preferentially absorbed by the material). To test this hypothesis, two sets of experiments were carried out. The first was just to use a higher power for the X-Ray beam. The absolute error at 50 KeV is 3.9% while that at 30 keV is 5.4%. However, just raising the tube voltage is not a solution in every case since as the tube voltage increases, the signal (extinction) also decreases. The same effect was also observed by placing a thin aluminum plate in front of the source (causing the X-ray beam to be hardened). With the introduction of the aluminum plate, the absolute error dropped even further to 0.8%. Therefore, for the selected ingredient (rice) the error obtained was less than 1% for both the chute configuration as well as the belt configurations. This completely demonstrates the feasibility and robustness of the method. 4. Conclusions The following conclusion can be drawn from the Phase I work. 1. The results of the Phase I work show that the X-Ray sensor was able to obtain mass flux within +/- 5% for many common food ingredients and processed food. 2. Preliminary data suggests that if beam hardening effects are taken into account, then the accuracy can be improved to an average error of +/- 1% 3. The results of the Phase I work completely demonstrated the feasibility of using soft X-rays for the development of a particulate mass flux sensor for the food industry.
Publications
|