EVALUATION OF TOMATO CANNING PRACTICES USING FLAME STERILIZATION

SHERMAN LEONARD, GEORGE L. MARSH, GEORGE K. YO=, J. R. HEIL and T. WOLCO'I"

Department of Food Science and Technology University of California Davis, California 9561 6

Received for Publication February 28, 1978

/ ABSTRACT

To compensate for variations in canningpractices, canned acid foods are systematically overprocessed to assure commercial sterility. For maximum yield and quality, common processing variables such as fill weight, headspace, consistency of covering medium and vacuum conditions should be standardized and controlled so that overprocessing can be minimized. I t was found, for example, that in terms of integrated sterilizing values (I.&',':,,, process lethality can decrease 30-40% when tomato fill weights in 303 x 406 cans increase from 7 to 15 oz. Evaluation of fruit condition showed that f i rm tomatoes can achieve 36% higher lethality than soft tomatoes for the same amount of energy input. Approximately 40 rpm (can) appeared to be most suitable for the flame processing of tomatoes packed in juice into 303 x 406 cans. By operating the commercial flame sterilizer at full capacity, a significant increase in internal can temperature can be achieved.

INTRODUCTION

Variations in acidity, raw material and filling conditions, complicated by variations in the resistance of spoilage-causing nonpathogenic or- ganisms, have made standardization of thermal processing for acid foods such as tomatoes difficult (NCA 1968a). Most of the canning industry continues to use processes that were designed to compensate for gross variations in raw material and filling conditions, and for unusually high loads of spoilage-causing organisms. Since processes for tomatoes were designed in 1941, mandatory inspection and minimum acceptable quality standards for harvested tomatoes have been established in the State of California (Canning Tomato Inspection 1975). The overall condition of processing tomatoes had markedly improved as the delay from harvesting to processing has been reduced. In 1966 the acidification of canned to- matoes was legalized (FDA 1966), allowing for the control of pH in canned tomatoes.

313

314 SHERMAN LEONARD ET AL.

With the introduction of Stkriflamme in California (Filper Corp., San Ramon, CA), processes for flame sterilizing canned whole peeled tomatoes had to be designed and evaluated (Leonard et al. 1975a,b,c). The work concentrated on juice-packed whole peeled tomatoes in 303 x 406 cans. It was found that when filling conditions were uniform and the tomatoes were not overprocessed, more of the original quality, weight and nutrients were retained than in the comparable commercial product (Leonard et al. 1975~). Following the work of Leonard et al. (1975a,b,c), York et al. (1975) confirmed the findings of earlier researchers that acidifying canned to- matoes to pH < 4.3 reduced the probability of flat sour spoilage by spores of B. coagulans. A process reagirement was determined in terms of integ- rated sterilizing values (I.S,,J based on the initial concentration of spores that could survive a heat treatment without the ability to cause spoilage.

The effort to date concentrated on methodology and the evaluation of Stkriflamme’s place in the California canning industry. In this research, several canning and processing variables were evaluated in terms of their influence on thermal processing, to permit re-examination and modifica- tion of present practices in the canning of whole peeled tomatoes.

EXPERIMENTAL

Microbiological Methods

The methodology and test organisms were the same as described by Leonard et al. (1975a,b,c), that is, spores of B. coagulans ATCC No. 8038 were cultivated on pH 4.5-5.0 tomato juice nutrient agar. The thermal resistance of the spores in tomato juice, as defined by the phantom thermal death time curve, was logio D = 7.565-0.037 T, where T was the test temperature in degrees F. Dl:i = 0.526 min and z = 27” F were calculated with this equation which had a correlation coefficient r = 0.980 for seven determinations.

Process lethalities were determined in terms of IS.::. values.

I.S.::,= I h i z (logioa - logiob) where Dziz = 0.526 a = initial number of spores (control can) b = number of spores that survived the process (processed can).

Average I.S.2”:2 values were calculated by averaging individual quantities for a and b. F::zvalues were calculated from the single point heat penetra- tion data using the General Method (NCA 1968b).

The test organisms were inoculated into the juice only, which was adequate to show the relative influence of the conditions tested.

EVALUATION OF TOMATO CANNING PRACTICES USING FLAME STERILIZATION 3 15

Canning

Round tomatoes were freeze peeled using the method of Leonard and Winter (1974). Unless otherwise specified, the tomatoes were packed with 4.3" B inoculated juice in 303 x 406 enameled tomato cans with 75 lb bodies and ends. The cans were vacuum closed under 15 in. vacuum.

Tomato Fill Weight.-Firm, uncut tomatoes of uniform size were used. Fill weights were within+ 0.1 oz Of 7,9,11,13 and 15 oz. Inoculated juice was added to 16.5 oz net weight. The cans were flame sterilized in a commercial cooker, a continuous laboratory cooker, or a simulator capable of processing one can at a time. To compare the results from the 3 machines, each with different processing conditions, the results are re- ported in percentages of the maximum I.S.i&alues achieved in the respec- tive processes. Six replications were made for each treatment.

Headspace.-For evaluating headspace, the cans were filled with 10.5 oz firm tomatoes and covered with inoculated juice to 14,14.5,15,15.5,16, 16.5 and 17 oz total weights. The final closing weight determined headspace. Results are given in terms of I,S.,:7,values achieved under constant processing conditions.

Tomato Quality.-Cans were filled with 10.5 oz all firm or all soft whole tomatoes. Tomatoes were covered with inoculated juice. Closing weight was 16.5 oz. The severity of the process was varied to reveal the influence of tomato condition on the lethality acquired.

Covering Medium.-Ten and one-half oz round firm tomatoes were covered with 6 oz of 4.6" B single strength tomato juice, 8.5" Brix tomato puree, or 13.5" Brix tomato puree. The 30 sec puree flow in the Bostwick consistometer measured 12.8 cm for the 8.5" B, and 6.4 cm for the 13.5" B purees. Process lethalities were measured in terms of I.S.,:7,values on five replications.

Thermal Processing

Heat penetrationdata were obtained in laboratory models ofsteriflamme using methods described by Leonard et al. (1975a). Count reduction and integrated sterilization (IS.) were measured with spores of B. coagulans ATCC No. 8038 in both laboratory and commercial models of flame cookers. The severity of the processes was varied by changing the flame intensities in the high flame section, by changing can rotation to give desired processing temperatures (in the cans), or by changing the holding times for a given processing temperature and can rotational speed.

316 SHERMAN LEONARD ETAL.

Can Rotational Speed.-The influence of can rotation was tested at 20,40,60,80 and 100 rpm. The cans were filled with 10.5 oz firm tomatoes and covered with 6 oz inoculated juice. Residence times and flame inten- sities were constant, only can rotation was varied. Five replications were made.

Temperature Measurements

Can end surface temperatures were estimated using Tempilaq (distrib- uted by Omega Engineering Inc., Stamford, Conn.) temperature indicat- ing liquids. The results obtained were reliable to 2 3°F. Internal juice temperatures were measured using 11/16 in. copper-constantan ther- mocouples (0. F. Ecklund, Cape Coral, Fla.) installed through the center of the lid. Internal juice temperatures were measured continuously in the single can flame cooker. In the commercial flame cookers, due to unsafe accessability after the steam section, internal temperatures were meas- ured with thermocouples only after the high flame and holding sections. Thermocouples were installed through the can bottom prior to filling with tomatoes. The filled cans were placed in the processing line. At the end of the high flame section, and after the holding section, these cans were removed, and temperatures were read through a plug-in read-out poten- tiometer calibrated in O F.

RESULTS AND DISCUSSIONS

Using the count reduction method of process evaluation, it was possible to show what influence the canning and processing variables have on the lethal effect of a given time-temperature treatment. The “ F values calcu- lated from single point heat penetration data from the samecans are not reported because they did not show differences for the canning and proces- sing variations.

Most of the findings reported could apply to all processing equipment which facilitate heat transfer to the canned food by intermittent or con- tinuous agitation (can rotation).

Tomato Fill Weight

The relative influence of tomato fill weights is shown in Fig. 1. Each data point represents the averaged results on 6 cans. The data indicate that as tomato fill weights increased, the effectiveness of the thermal process decreased, that is, with each one ounce increase in tomato fill weight, the processing time or temperature should be increased to compensate for the 3.2 +. 0.5 percent drop in the average I.S.::2values. It is assumed that as

EVALUATION OF TOMATO CANNING PRACTICES USING FLAME Sl'ERILJZATION 317

0

0

0

r= -0.858

a= 105.6 0

0

50 b=-3.348 0

7-p 7 i 8 9 I0 1'1 I2 I3 I4 -15 16

TOMATO FILL WEIGHT (OUNCES) FIG. 1. INFLUENCE OF TOMATO FILL WEIGHTS ON THE STERILIZING

VALUE OF A THERMAL PROCESS The r value is the correlation coefficient. The equation ofthe line is % of max. IS. ::2

value achieved in a process = 105.6-3.348 x tomato fill weight (02).

more tomatoes were packed into a can, the movement of tomatoes within the can during processing became more restricted, making heat distribu- tion within the can less uniform.

Headspace

Headspace is essential in canned foods to allow for product expansion and to facilitate heat transfer and the uniform distribution of heat in the can through agitating cookers. In this study, with constant fill of appar- ently uniform quality tomatoes, the net closing weights were more reliable and related better to the 1.S.::. values than did the actual headspace measurements. The data are presented in Fig. 2 and indicate that by decreasing headspace (higher closing weight), longer processing is re- quired to achieve a comparable level of sterility. Each % oz interval in the 303 x 406 can correspond to 3/32-4/3~ in. change in gross headspace. Cans having extreme closing weights averaged 3O/32 in. gross headspace with 14 oz, and 9.5/32 in. with 17 oz fills. The 14 oz net fill weights in commercial

318

3.0 -

2.0-

W 3 -J 4 >

r - N " cn

1.0-

SHERMAN LEONARD E T A L .

r= -0.728

a-q.516 b= -0.351

14 14.5 15 15.5 16 16.5 17 NET CLOSED WEIGHT (OUNCES)

FIG. 2. INFLUENCE OF CLOSING WEIGHT (HEADSPACE) IN CANNED TOMATOES (303 x 406) ON THE STERILIZING VALUE OF A THERMAL

PROCESS The r value is the correlation coefficient. The equation of the line is I.S.::2value =

7.516-0.351 x closing weight ( 0 2 ) .

cutouts were rare, whereas one-third of the cans examined weighed 16.9 oz or more.

Tomato Quality

When constant tomato count and net fill weights were used, the samples with firm tomatoes showed better response to thermal processing than did those which were soft to the touch. The relationship of tomato quality to I.S.::2values achieved is presented in Fig. 3.

Paired comparisons indicated that the textural properties of tomatoes significantly influenced the effectiveness of thermal processing. For a given process, soft fruits obtained significantly lower (p < 0.001) values than their firm counterparts.

EVALUATION OF TOMATO CANNING PRACTICES USING FLAME STERILIZATION 3 19

2.0- CLJ w 0

I 0 I-

I- LL 0

!z

1.0- J a 3 J

> P = (VN

clj -

0 0 0

a: 0.208

b- 0.570 ~

1.0 2 .o I.S. 2,2 VALUE, FIRM TOMATOES

INFLUENCE OF WHOLE PEELED TOMATO TEXTURE 0

27

FIG. 3. TI E STERILIZING VALUE OF THERMAL PROCESSES

The r value is the correlation coefficient. The equation of the line is IS. ::,,(soft tomatoes) = 0.208 + 0.570 0 I.S.2";: (firm tomatoes).

Covering Medium

When the influence of the packing medium on tomatoes in 303 x 406 cans was tested, under the same processing conditions the I.S. ::z values acquired decreased by 18% in the 8.5" B, and by 63% in the 13.5" B purees, as compared to the juice packs (4.6" B). Some commercial juice packs have been measured a t 7.0" B, thus, according to the above trends, unless the covering medium is controlled, the variations experienced in single strength juice packs may cause significant differences in the lethality acquired under specific processing conditions.

Variables Related to the Can

Since the can becomes its own pressure vessel in flame sterilization, i t is relevant to discuss factors which influence the cans' ability to resist per- manent distortion (buckling) during flame processing. Four of the factors which directly influenced the cans' ability to resist buckling at a desired processing temperature were observed as follows: (1) can size, (2) vacuum

320 SHERMAN LEONARD ETAL.

condition, (3) plate weight or thickness of ends and (4) profile, i.e. depth of concentric circles on the end plates.

Other factors considered were concerned with the cans' relationship to the processing equipment and to each other.

Can Size

It was observed that as the cans' diameter increased, the structural strength of end plates decreased, and the ends became permanently dis- torted (buckled) at lower temperatures. For example, under comparable conditions, with 85 lb (0.0093 in. thick) end plates, sealed under 10-12 in. vacuum, the 4V16 in. diameter cans could be heated only to 238" F before buckling, while the 33/16 in. diameter ends resisted the internal pressures to a 259" F.

Vacuum

By increasing vacuum from 10-12 in. to 27 in. Hg, processing tempera- ture could be increased from 3 253" F to 3 275" F in a 303 x 406 can.

Plate Weight

By increasing the end plate weight by 10 lbs i.e., from 0.0082 in. to 0.0093 in., buckling occurred at 3 10" F higher internal can temperatures.

Profile

It was observed that if the can ends did not have identical profiles, buckling occurred at temperatures 3.5-5.5" F lower than with ends that were identical. The structural strength of the ends was apparently depen- dent on plate weight and profile, that is, the configuration and depths of the concentric circles. The end which buckled first had 18.5% less overall depth of concentric circles than the other end, as measured between the countersink and the flat center of the ends. The results obtained with currently used commercial lots of cans packed with whole peeled tomatoes in juice show the same trends that Casimir (1972) obtained in his model system. Variations in plate temper and other factors concerning the cans were not considered at this time.

Can Rotation

This is one of the most important variables within a machine design. Can rotation expedites heat transfer and distribution which in turn de- termine the corresponding time-temperature treatment (and therefore the size of machine) necessary for the product to be processed.

EVALUATION OF TOMATO CANNING PRACTICES USING FLAME STERILIZATION 321

2.0 - w 3 -I s c 2

1.0- -

Canned tomatoes consist of particulates which heat by conduction and a liquid which heats primarily by convection. The rate of heat transfer to the liquid may be increased by mechanical agitation caused by can rotation and thus by the movement of the particulates and headspace bubble in the can. The rate of heat transfer into the particulates can be influenced by agitation only to the extent of eliminating the stagnant layer of liquid surrounding the particulates, i.e., by minimizing surface resistance to the penetration of heat. This would be facilitated by turbulence in the juice, and/or the movement of tomatoes and headspace bubble. Turbulence, the movement of tomatoes and headspace bubble, can be increased effectively by increasing can rotation to the point of inducing any centrifugal effect.

The relationship of can rotational speed to the I.S.::2value achieved in a process for canned whole peeled tomatoes packed in juice in 303 x 406 cans is shown in Fig. 4. Minimal variation in I.S.i12values is desirable, and this

3.0 -

[ . / o

0

0 INDIVIDUAL 1.5.;; VALUES

X AVERAGED l.S.E:z VALUES

(AVERAGING SURVIVORS)

20 40 60 80 100 CAN ROTATION (RPM)

FIG. 4. INFLUENCE OF CAN ROTATIONAL SPEED ON FLAME PROCESS- ING OF CANNED WHOLE PEELED TOMATOES IN JUICE IN 303 x 406

CANS, MEASURED IN TERMS OF INTEGRATED STERILIZING VALUES

322 SHERMAN LEONARD ET AL.

occurred at 20,40 and 100 rpm. At 100 rpm, 3 out of 5 cans buckled, an$ buckling did not coincide with the higher I.S.::2values achieved. The 1.S.212 values were low at 20 rpm, but both consistent and adequate at 40 rpm (York et al. 1975). The data confirm the use of 38-45 rpm in the commercial flame cookers for processing canned whole peeled tomatoes in juice in 303 x 406 cans.

Proximity of Cans

In retort processing the energy input is proportional to the number and mass of the units processed. In flame sterilization, the energy input can be considered constant for a specific process, therefore it would be a sound practice for the canner to run his flame cooker to full capacity. In addition to the economic soundness of operating the flame cooker filled to capacity, the proximity of cans effects conditions which cause significant increases in internal can temperatures. It was found in commercial flame sterilizers that juice temperatures in cans surrounded by others measured 3.5-7" F higher than in cans processed alone. This difference, as evaluated by the t test, is very significant (p < 0.001), and indicates that running the flame cooker to full capacity would offer the advantages of maintaining maximum production, and saving fuel by lowering gas consumption to the level necessary to maintain the designed process temperature.

CONCLUSIONS

One of the significant results of this research was to learn that single point measurements did not indicate the influence of canning and process- ing variables on the effectiveness of the time-temperature treatment given to canned, whole, peeled tomatoes. The indications are that in a rotating can the definition for the slowest heating point must be reevaluated and/or that single point measurements are not sufficiently sensitive to measure the differences which were demonstrated in the biological count reduction.

Integrated sterilizing (IS.) values showed that common processing var- iables such as headspace, fill weight, consistency, quality and can rota- tional speed all have considerable influence on the lethality attainable within a given time-temperature treatment for canned tomatoes. In pres- ent canning practice, gross processing variations are not eliminated, but the preservation of the canned foods is geared to overprocessing so that spoilage which would result from overfilling, variable quality, etc., is held within acceptable limits. By enforcing standard conditions, this compen- satory overprocessing could be replgced by precision processing which would yield more cans of foods (no overfilling) of better and more uniform

EVALUATION OF TOMATO CANNING PRACTICES USING FLAME Sl'EXILIZATION 323

quality. Precision processing should minimize thermal degradation in the food, thus more of the heat-labile attributes could be preserved per unit energy consumed.

ACKNOWLEDGEMENTS

The authors wish to express appreciation to R. L. Merson for helpful discussions, to Filper, a DiGiorgio Company for grants-in-aid to partially support this project, and to the Continental Can Co. for machinery.

REFERENCES

Canning Tomato Inspection. 1975. Private communication. State of California Dept. of Food and Agriculture. Division of Inspection Services.

CASIMIR, D. J. 1972. Container requirements for flame sterilization. In Flame Sterilization. Specialist courses for the Food Industry, No. 2, 36-44. AIFS & CSIRO, Melbourne.

FDA. 1966. Canned tomatoes: Standards of identity and quality. In The Almanac of the Canning, Freezing, Preserving Industries, (1967 Ed.), p. 101, E. E. Judge & Son, Westminister, MD.

LEONARD, S., MERSON, R. L., MARSH, G. L., YORK, G. K., HEIL, J. R. and WOLCOTT, T. 1975a. Flame sterilization of canned foods. An overview. J. Food Sci. 40, 246249.

LEONARD,S., MARSH, G. L., MERSON,R. L., YORK, G. K., HEIL, J. R., FRYER, S., WOLCO'IT, T. and ANSAR, A. 1975b. Comparative procedures for calculat- ing Stkriflamme thermal processes. J. Food Sci. 40, 250-253.

LEONARD, S., MARSH, G. L., MERSON, R. L., YORK, G. K., BUHLERT, J. E., HEIL, J. R. and WOLCO'IT, T. 1975c. Chemical, physical and biological aspects of canned whole peeled tomatoes thermally processed by Stkriflamme. J. Food Sci. 40, 254-256.

LEONARD, S. and WINTER, F. 1974. Pilot application of freeze-heat peeling of tomatoes. J. Food Sci. 39, 162-165.

National Canners Research Laboratories. 1968a. Investigating spoilage problems. In Labomtory Manual for Food Canners and Processors, Vol. 1, pp. 4-4-87, Avi Publishing Co., Westport, Conn.

National Canners Research Laboratories. 1968b. Process calculations, In Labom- tory Manu1 forFood Canners and Pnxessors, Vol. 1, pp. 220-251, Avi Publishing co., westport, Conn.

YORK, G. K., HEIL, J. R., MARSH, G. L., ANSAR, A., MERSON, R. L., WOL- COTT, T. and LEONARD, S. 1975. Thermobacteriology of canned whole peeled tomatoes. J. Food Sci. 40, 764-769.