Research Of Tenderization And Quality Evaluationmethod On Raw Roast Beef

Beef is one of the major red meat in our domestic consumption. Since new century, with the improvement of living standards and introduction of western culture, beef consumption has being increased dramatically. Roasting is one of the main methods of beef consumption. But due to Chinese beef industry lagging behind, the quality of raw roast beef are disordered and baseless. It is necessary to set up an effective method that can be used to the raw beef's tenderization and quality evalution. The traditional quality grade standards and evaluation methods usually use chemical analysis, instrumental detection and sensory evaluation and other destruction detection methods to determine the product's quality, which consumes a lot of time and money. And the development of beef industry are limited because that. As a kind of non-destructive, sensitive and rapid detection method, the near-infrared reflectance spectroscopy(NIRS) has received widespread attention in meat industry.In this dissertation, different beef cuts were acted as the objects of study. Response surface methodology was used to model and optimize the tenderization methods of raw roast beef firstly. Then we used the domestic raster scan of near infrared spectroscopy instruments combined with the chemical analysis, instrumental analysis and sensory analysis of beef cuts to develop the optimal protocol for NIRS on-line evaluation of roast beef's quality, determinate the quantitative quality factors of raw beef and predict the grades of beef. The major results and conclusions are summarized as follows:1. Response surface methodology was used to model and optimize responses. Enzymatic specific activity, phosphate concentration and time were selected as the main tenderizing parameters. The models were established based on optimal conditions for the Warner-Bratzler shear force(WBSF) and sensory evaluation of the production. The results showed that the two models were both fine(P<0.01)and R2 were 0.9278 and 0.8985 respectively. As a result, Optimum parameters were papain activity 9.44U/g,concentration of phosphate 18.35 mg/m L and tenderizing time 9.09 h. By the function of papain and phosphate, the Warner-Bratzler shear force can be remarkably reduced and a higher sensory evaluation score can be achieved.2. The chemical composition, processing properties and sensory evalution of the raw meat were analysed by principal component analysis and correlation analysis. The key factors which can affect the quality were determinated, and the quality evaluation and grading methods in raw meat of roast beef were set up. The results showed that the cumulative variance contribution rate of crude protein content, ether extract content, moisture content, WBSF, WHC and a* reached 80.78%.That showed the six parameters can reflect the quality of raw meat. And the regression analysis model of quality score and the six parameters was established. According to the score, the raw meat of prepared roast beef can be divided into three grades(S, A, B, respectively).3. The domestic raster scan of near infrared spectroscopy instruments combined with the chemometric methods were used to probe the ability of NIRS predicting pre-conditioning raw meat quality and grades with different states. The results showed that NIRS had a higher prediction accuracies to crude protein content, ether extract content, moisture content when samples were minced,R2 c and RMSECV were 0.93 and 0.004(crude protein), 0.88 and 0.003(ether extract), 0.77 and0.005(moisture). R2 p and RMSEP were 0.80 and 0.007(crude protein), 0.67 and 0.006(ether extract),0.68 and 0.007(moisture) respectively. WBSF, CL and WHC were better when the samples were detected non-destructively. R2 c and RMSECV were 0.88 and 0.33(WBSF), 0.96 and 0.006(CL), 0.91 and0.021(WHC). R2 p and RMSEP were 0.77 and 0.473(WBSF), 0.83 and 0.013(CL), 0.69 and 0.04(WHC)respectively. Besides, the partial least squares discriminate(PLS-DA) model was established in this study to discriminate differences of beef quality. The accuracy of calibration, validation and prediction reached 95.00%, 93.33%and 80.00% respectively.