D. W. STANLEY,’ G. P. PEARSON and V. E. COXWORTH Faculty of Food Sciences, University of Toron to, Toron to, Ontario, Canada

EVALUATION OF CERTAIN PHYSICAL PROPERTIES OF MEAT USING A UNIVERSAL TESTING MACHINE

SUMMARY-The lnstron tester served to evaluate physical properties of uncooked rabbit and beef muscle including work of rupture, breaking strength, break elongation elasticity and stress re- laxation. These methods measure variations in muscle type, aging and post-mortem treatments comparably with shearing instruments. Shank showed higher tensile properties than tenderloin, less elasticity and lost more applied stress. With rabbit, the breaking force of longissimus dorsi unre-

strained during rigor was .237 lb/g ? 7.5% for samples 5.0 cm by 0.2-0.5 cm2 while restrained muscle gave _ 168 f 9.9% and also exhibited higher elasticity and break elongation. Post-mortem aging decreased tensile properties and elasticity. Psoas muscle, characterized by more coextensive fibers, had higher tensife properties than iongissimus dorsi.

INTRODUCTION MEAT TEXTURE is an important factor in consumer acceptance and, as such, should be able to be assessed accurately. Texture, used here to connote mechanical properties rather than coarseness or fine- ness, can be measured objectively while tenderness implies subjective evaluation and must ultimately be organoleptically determined through appraisal of charac- teristics such as juiciness, ease of tooth penetration and residual fibers.

Presently, the most widely used objec- tive method used for evaluation of meat texture involves various means by which force is applied via a blunt edge and the amount of force required to shear the sample is observed. This procedure is open to criticism on several grounds. Szczesniak and Torgeson (1965) empha- size the importance of such variables as orientation of muscle fibers, sample tem- perature, speed of shearing and blade dullness. They report variations in corre- lation coefficients between Warner-Bratz- ler shear values and taste panel reports varying from no significance to very high significance. Usually only the maximum shear force is measured and not the slope of the shear force curve which has been

1 Present address: Department of Food Science. University of Guelph, Guelph. Ontario, Canada

suggested to be more meaningful (Szczes- niak and Torgeson, 1965). Shearing de- vices are inherently empirical in nature and, as pointed out by Sharrah et al. (1965), it is not clear that these instru- ments measure the same characteristics in meat as do sensory panels. Another com- mon complaint is large standard devia- tions, over 20% of the mean in some cases, for replicates. Attempts have been made to correct shear force values for variations in sample dimensions by both Davey and Gilbert (1969b) and Pool and Klose (1969). Both groups found that shear force was more accurately ex- pressed as a function of linear dimension rather than cross-sectional area and it may now be possible to mathematically adjust data to account for this variable. Perhaps the most serious objection from a theoretical viewpoint has been raised by Pool and Klose ( 1969) who suggest that meat samples subjected to shearing stress are distorted to the point that part of the applied shear force is altered to a tensile stress of the stretching fibers. The separa- tion of fibers is due more to tensile force perpendicular to the blade than shear parallel to the blade. Another point would seem to apply to the Kramer-type shear press which supposedly measures compression as well as shearing. Only the fibers in immediate contact with the shearing bars receive the entire stress. Other areas obtain less and are com-

pressed so that the total area is much smaller prior to rupture of any fibers. It appears that most of the work done is required to express fluid from the sample.

Clearly a method for assessing meat texture free from the above complica- tions would be welcome. The first step in finding a better method is to study various physical properties of muscle to see which might be likely to correlate with sensory ratings. In an effort to choose a versatile test that might be explained theoretically and for which instrumentation already existed, it was decided in this work to apply a stretching force parallel to the muscle fibers. One criterion that can be measured using this technique is extensibility or break elonga- tion which is the length muscle fibers must be stretched to produce breakage. This method was applied to single beef muscle fibers by Wang et al. (1956) and a highly significant negative correlation was obtained when extensibility was com- pared to organoleptic tenderness.

Elasticity may also be measured by this procedure. Muscle, along with many other biological tissues, can be regarded as viscoelastic in that it resembles a combination of an elastic solid and a viscous fluid. Viscoelasticity can be recog- nized by stress relaxation which is the decay of stress with time if a material is stretched to constant extension.

The source of elasticity in muscle is not yet completely clear. Muscle seems truly elastic up to about 3% extension of muscle length (Bate-Smith, 1939) but beyond this point the stress-strain curve is nonlinear. Work done on glycerinated muscle by Hoeve and Willis (1963) indi- cates that at the molecular level elasticity is related to a phase change of the fibrous proteins from an oriented crystalline state to a randomly coiled, amorphous state.

2564OURNAL OF FOOD SCIENCE-Volume 36 (1971)

PHYSICAL PROPER TIES OF MEA T-257

Under certain nonphysiological condi- tions of high temperature and strong salt solutions an essentially rubber-like elastic- ity, characteristic of the amorphous state, was observed. This was thought due to melting of the original crystalline fibrous proteins. Perhaps this explains the obser- vation of Guth (1947) that the stress- strain curves of resting muscle and rubber are significantly different in that muscle corresponds to rubber stretched out so much that the chain molecules are mark- edly oriented.

More recently Hoyle (1968) postu- lated a mechanism of muscle elasticity. It was once thought that the sarcolemma or cell membrane was mainly responsible for elasticity in muscle, but Hoyle reports that elasticity is present at lengths which do not stretch the sarcolemma and that elasticity is present in fibers from which the sarcolemma has been dissected away. Also, in some kinds of fibers the sarce lemma contributes only 1 S-20% of the total elasticity. The role of tendons in elasticity of muscle is discounted and Hoyle concludes that the individual sarco- meres must be the major source of elastic material. Both myosin and actin filaments are reported as inelastic and a new sarco- mere component is postulated, the T-fila- ment or thin filament, which has been seen in electron micrographs of the gap region between the actin and myosin of heavily stretched fibers. It is proposed that these T-filaments run from Z-line to Z-line and besides playing a passive role as an elastic element may also be involved in contraction.

The purpose of the research reported herein was to investigate physical proper- ties of meat and develop methods for their evaluation. The results show that the instrument used is capable of discem- ing variations in physical properties of uncooked muscle. It is possible that one or more of these tests will prove useful as a predictor of meat tenderness.

EXPERIMENTAL Apparatus

The lnstron universal tester (Instron Engi-

Fig. l-Measurement of physical properties of muscle using the lnstron tester.

neering Corporation, 2500 Washington Ave., Canton, Mass.) was selected to perform various tests upon uncooked meat all of which applied force parallel to the fibers. The Instron has been described in detail by Boume et al. (1966) and White (1970) as a research tool to study the theological properties of food materials. Basically, this instrument is used by mounting the sample in a gripping mechanism which is connected to a strain gauge. Changes in force applied to the sample cause the beam of the strain gauge to deflect. The output of the gauge is fed to a strip-chart recorder which draws a force-distance curve for every test. In these ex- periments a Model TT was used with the fol- lowing operating parameters: crosshead speed- 12 in./min; chart speed-12 in./min; cell-tension load cell ‘c’, 1 lb full deflection; jaws-type 2A fiber clamps.

Procedure Breaking strength of a sample was deter-

mined by mounting it in the jaws of the In- stron, initially 3.5 cm apart, and applying force by the downward movement of the crosshead. Jaw slippage was held to a minimum by wrap- ping the ends of the sample in moistened strips of fabric prior to mounting. Breaking anywhere along the sample except at the jaws was found to be acceptable. To measure elasticity, or the ability to recover after deformation, the Instron was programmed for extension cycling. At max- imum jaw separation the sample was elongated to 115% of its rest length (i.e., 3.5-4.0 cm). The muscle bundle was tested for 1 min which involved about 40 extension cycles. Relaxation or the loss of stress at constant extension was measured in a similar manner; the crosshead was moved downward and the sample was held at the same maximum separation for one min- ute.

The measurements obtained from these methods are summarized in Figure 1. The fol- lowing tensile properties were evaluated; break- ing strength or breaking load in lb force/g sam- ple; break elongation or strain required to rupture the sample as a percentage of the orig- inal 3.5 cm sample between the jaws; specific work of rupture or the area under the stress- strain curve in inch-lb/g sample. Time effects measured included elasticity which was taken as the area under the stress-strain curve following 1 mm of cycling to 115% elongation as a per- centage of the initial area, and relaxation ex- pressed as the amount of stress loss in 1 min at 115% elongation as a percentage of the original stress. Materials

The samples used in these experiments con-

Fig. 2-Regression of muscle breaking strength on sample weight.

sisted of commercially obtained, uncooked tenderloin and shank beef muscle. Rabbit sam- ples were paired longissimus dorsi (LD) muscles from young female animals treated during rigor to produce one muscle that had been excised, unrestrained, and allowed to contract freely while the control was restrained on the carcass at rest length. Both samples were refrigerated at 0-S°C for 24 hr post-mortem. For details of the post-mortem method and procedures used to measure sarcomere length and fractionate protein extracts consult Buck et al. (1970). Fol- lowing rigor the contracted muscle was excised and both muscles subjected to physical testing.

Samples were prepared by cutting the meat to 5.0 cm length and about 0.2-0.5 cm2 cross- sectional area. The latter was only approximate due to the irregularities of biological material and the samples were also weighed prior to test- ing. Experiments were conducted in a condi- tioned room at 70°F. Samples were kept in an ice bath prior to use.

RESULTS & DISCUSSION THE IRREGULAR SHAPE of the stress- strain curve for muscle, an example of which is depicted in Figure 1, is thought to represent physical rupture of muscle fibers occurring at different loads. The fibers with the lowest break elongation will fracture first. Once a few fibers have ruptured, breaking is accelerated since the increasing load is spread over the remain- ing fibers, increasing the specific load per fiber at an ever increasing rate. The point is finally reached where the increase in load due to the breaking of a fiber will cause another fiber to break immediately and this process continues until the whole specimen ruptures. This happens at a load which is less than the sum of the breaking loads of the individual fibers (Morton and Hearle, 1.962).

Very small psoas muscle fiber bundles, less than 0.5 mm diameter, were used in an attempt to obtain a stress-strain curve free from the effects of connective tissue and fat. Even small amounts of connec- tive tissue running the length of the sample could have a noticeable effect on breaking strength and break elongation because of this material’s very high break- ing strength (Abrahams, 1967). These results were similar to those found for larger muscle bundles and did not resem-

Fig. 3-Influence of sample length on breaking strength of beef tenderloin fibers.

25%JOURNAL OF FOOD SCIENCE-Volume 36 (1971)

Table 1-Phvsical orooerties of beef muscle.

Tensile properties Time effects Specific work Breaking Break Stress

of rupture strength elongation Elasticity relaxation Muscle (inch-lb/g) (lb/g) (%) (%I (%) Shank .327 ,912 28 28 70 Tenderloin .061 .341 20 36 60

ble the biphasic curves characteristic of fibrous proteins such as wool (Chapman, 1969) or spun soy fibers (Stanley, unpub- lished data). This may be interpreted as meaning that the stress-strain curve for muscle reflects other constituents than the fibrous proteins, perhaps the sarco- lemma.

Averaged data for all animals are pre- sented in Figure 2 and indicate that for the range of sample sizes used, breaking strength is proportional to sample weight at constant length and hence to cross- sectional area. The lines shown are regres- sion lines plotted through the origin. Larger samples of rabbit LD muscle were required to obtain consistent results be- cause in this muscle, in contrast to the beef samples and rabbit psoas muscle, microscopic examination showed the fibers are not completely parallel to one another. Tenderloin, a quite tender mus- cle, has a somewhat higher breaking strength per unit weight than rabbit. This may also be explained by a failure of the rabbit fibers to be completely parallel or run the entire length of the sample which leads to fibers being pulled over one another rather than breaking. Standard deviations for replicates of a single animal were between 15 and 20% of the mean for beef muscle and slightly higher for rabbit. Again, it is felt that this is due to irregularity in the rabbit muscle.

The physical properties of uncooked, commercially obtained beef muscle are presented in Table 1. Each number repre- sents the average of 28 determinations

The effect of sample length on break- ing strength of beef muscle is shown in Figure 3. These results may be explained by the so-called “weak-link” effect which predicts a fiber will break at its weakest point and the longer the fiber, the greater the statistical possibility of its having a “weak-link” (Morton and Hearle, 1962). This leads to the breaking strength of a specimen decreasing as the test length is increased. It is possible that over the region where the curve is linear the fibers are mainly running the entire distance but that at greater lengths the force is utilized more to cause the fibers to slide over one another.

on two animals. The magnitude of dif- ference between the specific work or breaking strength for two muscles is about what might be expected using a shear press (e.g., see Wang et al., 1956). Wang’s results were supported in that a greater break elongation was found for the tougher sample. Since an extensio- meter was not used with this work, it is possible that jaw slippage may have af- fected these results, but because of the rather large elongation found, it seems likely that this factor is of slight impor- tance. Average standard deviations for elasticity and relaxation measurements were about 10% of the mean. The more tender cut had a greater ability to recover after deformation and lost less stress during mechanical conditioning. It was found that when the log of stress is plotted against relaxation time, an initial fast decay occurs, followed by a linear decrease. Thus, stress decay in muscle is an exponential function of time.

The physical properties of rabbit mus- cle were evaluated and the results in Table 3 are the average of 12 deterniina- tions on one or two animals. The differ-

These initial results with beef were indicative that the instrument was capa- ble of discerning gross differences in physical properties of meat. Next we investigated the rabbit system in which the post-mortem conditions could be better controlled. In all trials the unre- strained muscle had shorter sarcomeres than its restrained twin. This difference in sarcomere length proved highly signifi- cant. Total extractable protein soluble at an ionic strength of 0.55 was measured for both treatments and the protein extracted from unrestrained or con- tracted muscle was significantly less than that extracted from restrained muscle. Lowering the ionic strength of the pro- tein extract to 0.23 precipitated a crude actomyosin fraction. Significantly more actomyosin was precipitated from re- strained muscle. These results are summa- rized in Table 2 and are similar to those found by other investigators using these post-mortem treatments (Herring et al. 1965, 1967a; Buck and Black, 1967, 1968; Buck et al., 1970).

Table Z-Characteristics of post-mortem rab- bit muscle.

Rigor treatment Unrestrained Restrained

Sarcomere length 01) Total extractable

1.90 2.34**

protein (mg/g) 58.0 79.4* Actomyosin (mgjg) 24.5 40.7*

**Treatments significantly different at 1% level.

*Treatments significantly different at 5% level.

ence in breaking strength for the two treatments is comparable to that found for shear force values of cooked muscle by Buck et al. (1970). Break elongation was higher for all rabbit trials than for the tough sha& muscle which indicates that a simple inverse relationship between exten- sibility and tenderness does not hold in all cases. Again, elasticity was higher and re- laxation lower for the presumably more tender or restrained muscle. To measure animal variation 7 pairs of LD muscles were evaluated for breaking force only. A total of 140 tests were performed for both treatments. The contracted muscle averaged .237 lb/g and the restrained muscle .168. Standard errors of the mean for the two treatments were 7.5% and 9.9% respectively. A paired “t” test showed the difference in the means to be highly significant (P < 0.01).

Since the degree of contraction is the only difference between the muscle pair, the theory has been proposed that con- traction causes a greater degree of overlap between the thick and thin muscle fila- ments and leads to a higher concentration of actomyosin in the unrestrained sample (Herring et al., 1967a, b). This is the concept of actomyosin toughening and is supported by electron microscopic evi- dence (Carlsen et al., 1961). Experiments attempting the direct measurement of actomyosin by salting out have been unsuccessful in demonstrating a higher concentration of actomyosin in the con- tracted sample (see also Buck et al., 1970). However, this appears to be a fault of the method used rather than the theory (H. 0. Hultin, private communica- tion).

Although at one time it was thought that post-mortem stretching was effective because of thinning of connective tissue (Buck and Black, 1968) this now seems unlikely considering the results we have obtained. These clearly show a greater breaking strength for the unrestrained muscle even when the force is applied parallel to the fibers.

Table 3 also presents data showing the effect of post-mortem aging on the phys- ical properties of rabbit muscle. Measure- ments made 2 hr, 1 day and 8 days

PHYSICAL PROPERTIES OF MEAT-259

Table 4-Correlation between methods used to evaluate physical properties of meat.

Correlation Comparison coefficient

Table 3-Effect of muscle type, aging and post-mortem treatment on physical properties of rabbit muscle.

Tensile properties Time effects Specific work Breaking Break Elast- Stress

Aging Post-mortem of rupture strength elongation icity relaxation Muscle period treatment (inch-lb/g) (lb/g) (%I (%I (%) LD 2 hr - ,092 ,344 34 41 41 LD 1 day restrained .045 .195 35 39 41

unrestrained .053 .253 31 32 50 LD 8 days restrained .038 .159 31 17 52

unrestrained .065 .241 33 28 48 Psoas 1 day restrained ,078 ,293 16 20 66

unrestrained .191 .512 21 16 69

post-mortem showed a greater than two- fold reduction in breaking force and specific work of rupture over this period for restrained rabbit samples along with a similar decrease in elasticity. Aging gener- ally had less effect on unrestrained sam- ples. Elasticity decreased to a greater degree in the restrained muscle, perhaps as a consequence of sustained stretching. Relaxation increased over this period but not to the extent of the other indices. Recently it has been reported that aging causes changes in the microscopic appear- ance of the myofibril. The Z-lines disap- pear completely and the A bands length- en at the expense of the I zone (Davey and Gilbert, 1967, 1968, 1969a; Fukaza- wa et al., 1969). The breaking strength of a muscle fiber may be related to the propensity of the Z-line area to rupture and thus be a valid measurement of tenderness in aged meat.

The results obtained when rabbit psoas was treated similarly to LD may be found in Table 3 as well. After one day post- mortem both restrained and unrestrained samples had consistently higher tensile properties than were found with LD. Break elongation was lower and approxi- mated that found in beef. Following the general pattern elasticity was lower and relaxation higher in the unrestrained treatment. Also, removing the muscle and allowing it to undergo the rigor period without restraint increased the tensile properties measured in these experiments.

Microscopic examination of rabbit psoas revealed that its oriented fibrous structure closely resembled the beef mus- cle rather than the more random fiber alignment of LD. That the fiber orienta- tion can influence mechanical properties is evident from the results of Corey (1970) who, working with a model sys- tem constructed from spun soy fiber, found networks composed of parallel

fibers held together with a gelatin binder exhibited higher plastic response, total strain and stress relaxation but a smaller elastic region when compared to random or perpendicular networks. These results agree well with the data given here for psoas and LD muscle.

After the results had been obtained, it was of particular interest to note the degree of independence of the tests used. Correlations were calculated and the coef- ficients are presented in Table 4. Work of rupture and breaking strength are so strongly related that it should not be neccessary to measure work directly. Elongation seems independent of either of the other tensile properties while elasticity appears independent of the ten- sile properties. Stress relaxation, while only marginally related to elasticity, shows a significant positive correlation with breaking strength and negative cor- relation with elongation. This might be related to structure if stress relaxation can be thought of as the proclivity for fibers to rupture and slide over one another while break elongation is a measure of the resistance of the sample toward elonga- tion.

These results suggest there may be several advantages to the type of tests described. Several different and independ- ent values may be obtained from one instrument as compared to the less versa- tile shearing devices. The methods seem as reproducible as those currently em- ployed. It is possible that physical proper- ties can be related to muscle structure as has been done for other protein fibers so that the technique will develop into more than an empirical tool. It is evident that tenderness in meat is due to the subtle interaction of many factors, one of which is texture. An important question is which of the many parameters available should be measured as a predictor of

Breaking strength vs specific work of rupture

Breaking strength vs break elongation

Breaking strength vs elasticity

Breaking strength vs stress relaxation

Break elongation vs elasticity

Break elongation vs stress relaxation

Elasticity vs stress relaxation

+0.99**

-0.22

-0.17

+0.74**

+0.26

-0.61*

-0.40 **Coefficient significant at 1% level.

*Coefficient significant at 5% level.

tenderness. Correlations of physical prop- erties with subjective evaluations should be helpful in this regard. It seems unlikely that one test would be adequate for all cases.

REFERENCES Abrahams, M. 1967. Mechanical behavior of

tendon in vitro. A Dreliminarv retort. Med. Biol. Engng. 6: 433:

Bate-Smith, E.C. 1939. Changes in elasticity of mammalian muscle undergoing rigor mortis. J. Physiol. 96: 176.

Bourne. M.C.. Mover. J.C. and Hand. D.B. 1966. Measurement of food text& by a universal testing machine. Food Technol. 20: 170.

Buck, E.M. and Black. D.L. 1967. The effect of stretch-tension during rigor on certain physi- cal characteristics of bovine muscle. J. Food Sci. 32: 539.

Buck. E.M. and Black. D.L. 1968. MicroscoDic characteristics of cooked muscle subjecied to stretch-tension during rigor. J. Food Sci. 33: 464.

Buck, E.M.. Stanley, D.W. and Comissiong, E.A. 1970. Physical and chemical character- istics of free and stretched rabbit muscle. J. Food Sci. 35: 100.

Carlsen. F.. Knappeis, G.G. and Buchthal. F. 1961. Ultrastructure of the resting and con- tracted striated muscle fiber at different degrees of stretch. J. Biophys. Biochem. cyto1.11: 95.

Chapman, B.M. 1969. A mechanical model for wool and other keratin fibers. Textile Res. J. 39: 1102.

Corey. H. 1970. Texture in foodstuffs. Critical Rev. Food Technol. 1: 161.

Davey, C.L. and Gilbert, K.V. 1967. Structural changes in meat during aging. J. Food Tech- nol. 2: 57.

Davey, C.L. and Gilbert, K.V. 1968. Studiesin meat tenderness. 6. The nature of myofibril- lar proteins extracted from meat during aging. J. Food Sci. 33: 343.

Davey. C.L. and Gilbert, K.V. 1969a. Studies in meat tenderness. 7. Changes in the fine structure of meat during aging. J. Food Sci. 34: 69.

Davey. C.L. and Gilbert, K.V. 1969b. The ef- fect of sample dimensions on the cleaving of meat in the objective assessment of tender ncss. J. Food Technol. 4: 7.

Fukazawa, T., Briskey. E.J., Takahashi, F. and Yasui. T. 1969. Treatment and post-mortem aging effects on the Z-line of myofibrils from chicken pectoral muscle. J. Food Sci. 34: 606.

Guth. E. 1947. Muscular contraction and rub berlike elasticity. Ann. N.Y. Acad. Sci. 47: 715.

260--JOURNAL OF FOOD SCIENCE- Volume 36 (1971)

Herring, H.K., Cassens, R.G. and Briskeu, E.J. Morton, W.E. and Hearle, J.W.S. 1962. “Physi- 97. 1965. Sarcomere length of free and re- cal Properties of Textile Fibres.” The Tex- White. G.W. 1970. Rheology in food research. strained bovine muscles at low temperatures tile Institute. Manchester. J. Food Technol. 5: 1. as related to tenderness. J. Sci. Food Agric. Pool, M.F. and Klose. A.A. 1969. The relation Ms. received 8/16/70; revised g/24/70; accepted 16: 379. of force to sample dimensions in objective 9129170.

Herring, H.K., Cassens. R.G., Suess, G.G., Bnm- measurement of tenderness of poultry meat. gardt, V.H. and B&key. E.J. 1967a. Tender J. Food Sci. 34: 524. ness and associated characteristics of Sharmh. N.. Kunze, M.S. and Pangborn. R.M. The authors wish to thank their colleagues stretched and contracted bovine muscles. 1965. Beef tenderness: Comparison of sen- for valuable discussions throughout the course J. Food Sci. 32: 317. sory methods with the Warner-Bratzler and of thii investigation and to express appreciation

Herring, H.K., Cassens. R.G. and Briskey. E.J. ;.9~.;~8Kramer shear presses. Food Technol. to the President of the University of Toronto, 1967h. Factors affecting collagen solubility Dr. Claude Bissell, for the financial support prcr in bovine muscles. J. Food Sci. 32: 534. Szczeshiak,’ AS. and Torgeson, K.W. 1965. vided through his auspices. This work was sup-

Hoeve, C.A.J. and Willis, Y.A. 1963. Elasticity Methods of meat texture measurement ported in part by Government of Canada of the fibrous muscle proteins. Biochemistry viewed from the background of factors af- National Research Council Grants ##G.R. 21 and 2: 279. fecting tenderness. Adv. Food Res. 14: 33. #577-40.

Hoyle, G. 1968. In “Symposium on Muscle,” Wang, H., Doty. D.M., Beard, F.J., Pierce, J.C. Presented at the 30th Annual Meeting of the eds. Ernst. E. and Straub. F.B., p. 46. and Hankins, O.G. 1956. Extensibility of Institute of Food Technologists in San Fran- Akademiai Kiado, Budapest. single beef muscle fibers. J. Animal Sci. 15: cisco.