A Multi-Purpose Fruit and Vegetable Processor for Long Duration Space Missions

Transcription

1 A Multi-Purpose Fruit and Vegetable Processor for Long Duration Space Missions R. Paul Singh, D.C. Voit, and D.E. Barrett Department of Biological and Agricultural Engineering Department of Food Science and Technology University of California, Davis, CA To date the dietary requirements of the astronauts during space missions have been met mostly with a supply of prepackaged foods. The constraints imposed by the design of the spacecraft and weight limitations at lift-off have influenced the type and quantity of foods suitable for each mission. Technologies such as freeze-drying and low and intermediate moisture foods received considerable attention for such applications. In the future, lunar or planetary food systems for long duration space missions will require a paradigm shift. While pre-packaged foods may be suitable for short term missions, long duration missions will require growing, processing and preparing foods in transit and on location. Moreover these functions will need to be carried out in controlled life support environments. The design of food processing systems for such environments will be influenced by a number of constraints, including volume and mass, energy and water use, waste handling, equipment noise, operational safety, and crew time. The safety, acceptability and nutritional health aspects of foods will require careful attention. To address some of these issues in the design of food processing equipment, the U.S. National Aeronautics and Space Administration (NASA) supported a multi-year research project at the University of California, Davis, to construct a multi-purpose fruit and vegetable processing machine. The design and operation of this machine focused on some of the expected constraints of a long duration space mission. Voit et al. (2006) described the first generation of a multipurpose fruit and vegetable processor (MFVP) for manned mission to Mars. Based on follow-up studies, a prototype processor was fabricated and tested to process tomatoes into various end 1

2 products. In this paper, we provide additional details of the prototype and experiments conducted to determine its operating performance. Equivalent system mass (ESM), a metric used by NASA to compare alternate technologies for space missions, was calculated for the prototype. Figure 1 illustrates the second generation of the MFVP prototype, into which the following changes were incorporated. 1) The size reduction device was operated in the vertical operation to allow gravity feed as low gravity will be present on the Mars surface. Drive Motor Slicer Dicer Extractor Dice Tank Steam Venturi Juice Tank Membrane Module Pressure Pump Figure 1: Layout of basic operations of MFVP 2

3 2) The hot break procedure was performed using direct steam injection. A steam venturi acts as a stationary pump to agitate and heat the tomato dices. 3) Vacuum transfer was selected to transport the tomatoes through the finisher screen. Based on the results presented by Hand et al. (1955), finishing in a vacuum environment is expected to reduce the nutrient loss. Connections between the high pressure pump and the RO module are not illustrated in the drawings because flexible tubing was used. Figure 2 illustrates the divert chutes and valves which are used to remove product following any operation. Figure 3, 4, 5, 6, and 7 show photographs of the final MFVP fabricated with components tested for their individual performance. Slice Divert Dice Divert Divert Valves Figure 2: Layout of divert system and valves of MFVP 3

4 Figure 3: An assembled MFVP with different fabricated components Figure 4: Feed mechanism for diced tomatoes 4

5 Figure 5: Tanks used for tomato juice and concentrate Figure 6: Transfer components for juice into membrane assembly 5

6 Figure 7: Valves used to direct flow into desired chambers A process flow diagram was created to view the system in entirety in Figure 8. The illustration shows the processor operations fully integrated using a direct steam injection device as a hot break in lieu of an ohmic heating process. 6

7 (manual) Slicer M1 Water in WS1 (manual) CV3 VT3 Slice Removal (manual) Size Reduction System Dicer M1 (manual) Pulper M1 Pomace Removal (manual) Dice Removal (manual) Line Vac Hold Tank #1 T2 (manual) Hold Tank #2 Heating System VT1 PT2 T1 BV1 BV4 BV4 VT2 (manual) BV2 BV3 CV1 CV2 Steam in R SV1 SV2 SV3 RO Membrane Module High Pressure Pump M2 PT1 AS1 Concentration System Air In R AS2 AS3 AS4 AS5 C1 Water Removal (manual) Pulp/Conc Removal (manual) Figure 8: Process flow for MFVP A Contrologix PLC system from Allen-Bradly was selected for the MFVP due to the availability of parts and simplicity of design. The PLC system was specified based on the signal input and output requirements in Figure 8. The processor is capable of step control, while supplying durability for a long operation life to ensure that additional testing will be possible in the future. Table 1 summarizes the sensors required for basic process control. 7

8 Table 1: Motors, Valve and Sensor requirements for the Prototype MFVP Motors M1: Variable Speed 3 Hp AC Motor Controls Slicer, Dicer and Finisher/Sieve M4: Variable Speed 2 Hp AC high pressure pump Valving BV1 Electric solenoid valve for direct steam injection BV2 3 way valve to control product flow and close off steam tank BV3 Electric solenoid valve for vacuum line BV4 Water valve for membrane flushing BV5 Electric modulating diagram valve with feedback control to control TMP SV1 Steam solenoid 1 SV2 Steam solenoid 2 SV3 Steam solenoid 3 AS1 Air Solenoid 1 AS2 Air Solenoid 2 AS3 Air Solenoid 3 AS4 Air Solenoid 4 AS5 Air Solenoid 5 WS1 Water solenoid 1 (for cip) Sensors T1 Temperature sensory for hot break process feedback T2 (optional) temperature monitoring to predict, countdown process concentration time PT1 (optional) Pressure transducer 1 to monitor/control TMP PT2 (optional) Pressure transducer 2 to monitor/control TMP C1 (optional) Electrical conductivity monitoring to monitor status of membrane material Extensive trials were conducted with the prototype to assess its technological performance (Voit, 2005). The performance and output of each operation were characterized. Key results obtained from these trials are as follows: Size Reduction It was found that increases of the rotational speed of slicer blades increased the observed damage of tomato slices. Operated at 12.6 m/s tip speed, the percent torn slices by count was 10.5% and by mass was 11.5%. These values increased to 36.9% by count and 41.0% by mass when operated at 31.5 m/s. Additionally, it was observed that a minimum rotational velocity of 12.6 m/s was required to slice firmer tomatoes. The line loss (product remaining in the line following shut down) and drain yield of slices were also affected by changes in rotational velocity. When operated to process 2 kg at 12.6 m/s, the line loss averaged 10 grams while operation at 31.5 m/s yielded a mean line loss of 20 grams. The added difference was believed to be a result of added tissue damage. 8

9 Drain yield showed similar losses of slice integrity with a drain yield of 94.6% at 12.6 m/s and 90.0% at 31.5 m/s. The rotational dicer design was not capable of controlling the dice distribution by adjustment of the feed auger s rotational velocity. The geometric mean diameter of diced tomatoes was between 13.6 and 15.2 mm with a geometric variance between 12.4 and 14.0 mm. The qualitative performance of the system was, however, notably affected by rotational speed. It was observed that a minimum velocity of 4.8 m/s tip speed was required to ensure consistent flow of tomato across the dicer open area. This agreed with the results of the texture analysis. The Kramer shear force at 4.8 m/s dicer tip speed was 13.6 kg-f significantly higher than at 1.2, 2.4, and 3.6 m/s. These results indicated that the operation of the dicer with higher rotational velocity results in more reliable product flow and significantly greater product integrity. Line loss measurement agreed with the qualitative observation that high tip speed induced superior flow of tomato. An average line loss of 870 grams was observed at when 2 kg was processed at 1.2 m/s compared with an average loss of 202 grams observed at 4.8 m/s. Increasing the rotational speed range to 9.6 m/s, failed to illustrate that control of dice size may be achieved by altering the rotational velocity of the feed auger. Extractor tip speed affected the consistency of the extracted tomato juice. Results, quantified by rotational rheometry and Bostwick, illustrated that operation at 1.2 m/s yielded thinner juice than 4.8 m/s. The mean apparent viscosity at 50 s -1 was Pa s at 1.2 m/s compared with Pa s at 4.8 m/s. Additionally, the line loss decreased from 560 grams at when 2 kg was processed at 1.2 m/s to 240 grams at 4.8 m/s and the solids content of the juice and pomace increased to 4.95% [w/w wb] and 17.87% [w/w wb], respectively. In addition to higher total solids juice extracted, a higher rotational velocity was less prone to separation when stored. Juice extracted at 1.2 m/s readily separated in less than 1 hour. Rotational velocity was not shown to significantly affect the resulting ph, soluble solids or electrical conductivity of the juice. Increased rotational velocity to 9.6 m/s, suggested that further increases in consistency could be achieved. Ohmic heating 9

10 Ohmic heating of reconstituted tomato juice showed that the electrical conductivity of tomato juice varies linearly resulting in a non linear heating rate. A temperature increase from 18.7 C to 82 C was achieved in 22 minutes for 6 Brix juice. This was close to the calculated value of 27 minutes for a temperature increase from 20 to 85 C which was based on electrical conductivity values obtained from literature sources. The temperature observed by a glass thermometer agreed with results from a sheathed T- type thermocouple, however a variability of up to 9.8 C occurred within the heater for static reconstituted tomato juice. Electrical conductivity was also found to vary linearly with temperature at 7.5 and 10.5 Brix. The rate of heating for diced tomatoes did not follow the same trend as the reconstituted juice. The reference, 0 C, electrical conductivity was S/m for 4.5 Brix tomatoes compared with S/m for 6 Brix reconstituted juice. In addition, the electrical conductivity versus temperature was non-linear and appeared to follow a quadratic growth. It was found that solid particles heated more rapidly than liquid in the static heater due to the high packing fraction of dices. A reduction of electrode separation increased the heating rate and the power requirement of the operation. A peak power requirement of 5.3 kw at an average temperature of 81.4 C was observed and additional heating was beyond the capabilities of the experimental facility. The higher power requirement of the ohmic heating operation added to the ESM value of 2033 for the component. Concentration Experiments with direct tubular reverse osmosis (RO) confirmed that concentrations exceeding 20.7 Brix is achievable. Operating in a batch recirculation mode results in a long decay of permeate flux as the osmotic pressure of the tomato juice increases. Solute rejection values exceeded 96% when membranes remained intact, providing a simple method for automated detection of membrane failure. An initial flux value of 47.6 ± 6.98 L/m 2 h was obtained with operation at 50 C and 48.3 bar. Operation at ± 33.3 L/m 2 h C resulted in a higher initial flux 96.5 ± 33.3 L/m 2 h, however, repeated use of the membrane resulted in membrane failure evidenced by solute rejections less than 90%. 10

11 Additional tests compared the effects of the operational parameters by permeate flux and power requirement. Increased cross flow velocity did not result in detectable increases in permeate flux but did increase power. The power requirement increased from 1.1 kw to 1.6 kw when the cross flow velocity was increased from 1.0 m/s to 5.2 m/s at 41.4 bar and 75 C. On the other hand, increased transmembrane pressure did result in large increases in permeate flux and small increases in power. A permeate flux of ± 0.8 L/m 2 h was observed for operation with 7.5 Brix juice at 75 C and 20.7 bar. When the pressure was increased to 55.2 bar, the permeate flux increased to ± 1.03 L/m 2 h. Power increased from 1.2 kw to 1.4 kw with the pressure increase. Similar results were observed at 25, 35, 50 and 65 C for 7.5 Brix and 18 Brix. Temperature also significantly affected the permeate flux but was not found to increase the power requirement. Operation with 7.5 Brix juice at 25 C and 55.2 bar yielded a permeate flux rate of 34.5 ± 0.49 L/m 2 h, compared with the near three-fold higher flux rate previously noted at 75 C, ± 1.03 L/m 2 h. The combined increases in permeate flux by operating at higher temperature and pressure are projected to result in lower energy of concentration by reducing processing time. Additional tests found that concentration at 30, 50 and 70 C did not result in concentrates with significantly different apparent viscosity regardless of total processing time, but were successful in providing superior performance data to benchmark future improvements. Equivalent System Mass The ESM of the MFVP is transient, because unit operations are cycled through as the batch process proceeds. However, because the systems are designed to function as a batch in series, it can be considered as three separate values. The equivalency factors used in calculation of the MFVP ESM are summarized in Table 2 (Hanford 2004). The magnitude of the power component is expected to decrease if nuclear power is selected as the power source. If this becomes the case, the value will reduce to 86.9 kg/kw (Levri et al. 2000). Table 2 summarizes the mass, volume, power and cooling requirements of the MFVP. Levri et al. (2000) notes that mass components should include tools, 11

12 contingency, and consumable components. For the size reduction component there are no known consumables. Motor life is expected to exceed the mission duration of 2-3 years with intermittent use and no unique tools are required. For the hot break process, as before, there are no consumable parts. The membrane concentration will, however, have consumable parts. The number of spare membrane sets required will be determined by the number of concentration cycles required. At this time, the total number of concentration cycles anticipated is not known, however a spare set of membranes will at 0.5 kgs additional weight along with 0.15 m 3 volume. Table 2: ESM values for MFVP sub-systems. Mass (kg) Volume (m3) Power (kw) Cooling Crew Time (hr) Size Reduction N/A Hot break N/A Concentration N/A Volume also is expected to include any spares or contingencies (Levri et al. (2000). Power is generally considered as the peak power (Levri et al. 2000). This is detrimental for an ohmic heating process in this application. For cooling, Levri et al. (2000) note that it is generally assumed that all energy added must be removed as thermal energy. Therefore, the cooling component is set equal to the power component. Lastly, the crew time requirements were not included based on discussions with French (2004). The MFVP process is being automated so that crew time requirements will be small. Long duration testing of the process is recommended to quantify the maintenance requirements. Table 3 summarizes the ESM values of each component for each of the subsystems. For the size reduction system, the mass may be largely reduced with the use of alternative materials. The prototype system was constructed using 316 stainless steel. This material, although dense, means maximum durability and lower cost for prototype verification. Use of a less dense material such as titanium with a density of 4.5 g/cm 3 compared with stainless steel at about 7.2 g/ cm 3, would reduce the ESM of that component to about 57.5 kg. 12

13 Table 3: ESM values for MFVP sub-systems ESM Coef. Size Reduction ESM Hot Break ESM Concentration ESM Weight N/A 93.0 (kg) (kg) (kg) 75.5 Volume (kg/m 3 ) (m 3 ) (m 3 ) (m 3 ) Power 228 (kg/kw) (kw) (kw) (kw) Cooling 146 (kg/kw) (kw) (kw) (kw) Crew Time 1.25 (kg/cm-hr) N/A N/A N/A N/A N/A N/A ESM NCT (kg) (kg) (kg) ESM Total (kg) For the hot break process, power and in turn cooling based on the assumptions was a large contributor to the ESM value. The thermal process consumes a large amount of energy by nature. Although no phase change occurs, the energy is required in a short period of time. To reduce the ESM of the hot break component, it is necessary to either heat the product slower, although not desirable due to the adverse affect of enzyme degradation, use existing thermal energy within the module or allow for thermal stockpiling of energy for use. Levri and Finn (2001) discuss the possibility of hot and cold stream matching in advanced life support systems. Although the results presented are for steady state operating conditions, a certain amount of preheating may be obtainable for the hot break process. Another alternative is to use a slow operating steam boiler activated a day prior to use of the MFVP. The slow operating boiler would have a relatively low power requirement, thereby reducing the ESM component largely in the hot break unit operation. The MFVP processor will not by nature be used daily. Therefore, a steam boiler could slowly generate and store the necessary steam for cooking and cleaning of the MFVP. Use of thermal hot break methods such as direct or indirect steam allow for waste heat reuse as discussed by Levri and Finn (2001) and Levri and Finn (2000) in the steam boiler as well as prior to the actual hot break process. For the concentration operation, the ESM value is low, when compared to alternatives such as steam evaporation. For the individual components, the mass is large and could be reduced using lighter materials such as titanium. In addition, if the full array of vegetable crops were decided, a pump could be more accurately sized to reduce the total power and thus energy of the process. Steffe (1984) notes that proper sizing of pumps for non-newtonian fluids can result in significant energy savings. The application of a power law is discussed and is applicable in this case. 13

14 By comparison, breadmakers were analyzed for their operation on long duration space missions by French and Perchonok (2004). It was found, using the same equivalency factors, that breadmaker ESM values ranged from kg to kg. Although the breadmaker with the lowest ESM was selected, French and Perchonok (2004) discusses the importance of human interface and overall equipment design in the selection of successful technology. Weiss et al. (2003) found lower values of ESM for bread makers, and kg, however the researchers noted that when technologies have multiple uses the ESM value unfairly compares a technology. For the MFVP, although the ESM sum is high, application of the processor for processing other produce such as carrots, cucumbers, radishes and potatoes will increase the applications. In summary, experimental trials with the second generation MFVP have provided new and useful information on its operating performance. The operating conditions of the MFVP for the desired product characteristics have been investigated. The ESM values calculated according to the NASA guidelines provide a quantitative measure to evaluate different components of the processor and seek future modifications. References French, S.J. (2004). National Aeronautics and Space Administration, Johnson Space Center, Houston. Personal communication. French, S.J. and M.H. Perchonok (2004). Metric Evaluation of Technologies for Long Duration Space Missions: Application to Bread Makers. Journal of Food Science, 69 (5), R139-R142 French, S.J. and M.H. Perchonok (2004). Metric Evaluation of Technologies for Long Duration Space Missions: Application to Bread Makers. Journal of Food Science, 69 (5), R139-R142 14

15 Hand, D.B., Moyer, J.C., Ransford, J.R., Henning, J.C. and Whittenberger, R.T. (1955). "Effect of processing conditions on the viscosity of tomato juices." Food Technology. 9: 228. Hanford, A.J. (2004). "Advanced Life Support Baseline Values and Assumptions Document," Rep. No. NASA/CR Lockheed Martin Space Operations, Houston, Texas. Levri, J.A., Vaccari D.A. and Drysdale, A.E. (2000). "Theory and application of equivalent system mass metric," Rep. No New Jersey NASA Specialized Center of Research and Training and Stevens Institute of Technology, New Jersey. Levri, J.A. and Finn, C.K. (2000). "Energy reuse by matching hot and cold streams in advanced life support systems," Rep. No. 00ICES-170. Orbital Sciences Corporation and Stevens Institute of Techonology, NASA Ames Research Center, Moffet Field, CA. Steffe, J.F. (1984). "Problems in using apparent viscosity to select pumps for pseudoplastic fluids." Transactions of the ASAE. 27(2): Voit, D.C The development and evaluation of a multi-purpose fruit and vegetable processor of a long duration space mission. M.S. Thesis. University of California, Davis. Voit, D.C., M. Santos and R. P. Singh Development of a multipurpose fruit and vegetable processor for a manned mission to Mars. Journal of Food Engineering. 77(2)

16 Weiss, I., Ozen, B.F., Hayes, K.D., Mauer, L.J. and Perchonok, M.H. (2003). "Comparison of Equivalent System (ESM) of Yeast and Flat Bread Systems," Rep. No Department of Food Science, Purdue University. Acknowledgements This research was supported by a NASA contract (NRA #02-OBPR ). 16