Masticatory Deformation Mechanism And Texture Evaluation Of Viscoelastic Food Based On Masticatory-like Platform

In recent years,in order to meet the needs of various groups of people,new foods of all shapes and sizes have emerged,but the resulting food safety issues are a cause for concern.As an important indicator of food quality,food texture can effectively help companies to develop their products,promote the improvement of transportation,storage and sales,and thus promote human oral health to a certain extent.Therefore,it is of great theoretical significance and application value to explore efficient and accurate methods for texture detection and taste evaluation of food products based on the deformation process of food products during mastication.This paper is oriented toward viscoelastic food materials,and conducts research on the deformation mechanism and texture evaluation methods of food products in oral processing.The details are as follows:To investigate the deformation mechanism of viscoelastic food materials from the perspective of oral processing.The basic theory of viscoelastic material intrinsic structure model is elaborated,and the mechanical properties of the corresponding food products are investigated.Using TMS-PRO mass spectrometer as the experimental platform,the compression experiments were conducted at two rates for different ratios of gels to analyze the trend of stress-strain curves and compare the mechanical properties of different samples;then the stress-time curves of stress relaxation experiments were fitted by a seven-cell generalized Maxwell model to characterize the viscoelasticity of the material.The experimentally obtained food materials were input into the finite element simulation,and different chewing conditions were set to simulate real oral processing.The results show that the tooth action area has a greater effect on the destruction of viscoelastic foods compared to the chewing time,and the fragmentation effect is better under the grinding action.A food texture detection method based on the masticatory imitation platform is proposed.According to the human mouth and jaw system and physiological characteristics,three types of movements of the lower jaw during mastication are summarized,and the 6-PSS parallel mechanism is selected as the core of the masticatory experimental platform according to the human mouth and jaw system,and the initial construction of the platform is completed by installing relevant components after the degree of freedom verification calculation.The platform was set up with a positioning posture transformation coordinate system,and the kinematic analysis was carried out to find the mapping relationship between the displacement of the mandibular part and the change of the connecting support chain.Based on the previous research,we plan the bionic motion trajectory of the platform and judge the rationality of the trajectory through the motion replication of the virtual prototype in Adams software;based on the planned trajectory,we arrange an effective platform control scheme and driver to achieve accurate bionic mastication control.Applying the texture profile analysis method and installing the corresponding mechanical sensors in the upper and lower teeth,the platform can be used for texture analysis while bionic motion,and under the experimental comparison with the texture meter,it is shown that it can meet the demand of detecting the texture of common viscoelastic foods.The neural network technique was used to establish the texture evaluation model of viscoelastic food products.The prepared gelatin gummy samples with different compositions were used as the research objects,and the physical parameters were measured by imitation chewing platform,and the taste scores were obtained by evaluators tasting,and the sensory experimental results were processed by using fuzzy comprehensive evaluation method to obtain the comprehensive taste score of each sample.Then the GA-BP neural network evaluation model was built with the physical properties parameters as the input quantity and the comprehensive evaluation score as the output quantity,and the sample size was expanded by adding virtual samples.Under the training comparison of multiple models,the best topology of the network was determined to be 6×8×1,and the model under this structure had a better fit with a maximum error of 1.61%.Finally,three commercially available gelatin soft candy samples were selected for validation,and the prediction errors were also within ± 5%.The texture detection and taste evaluation method proposed in this paper for viscoelastic foods can not only compensate for the drawbacks of instrumental detection and improve the accuracy of quality analysis,but also provide methodological reference for texture detection of other types of foods.At the same time,the built bionic chewing platform can also be used for research and exploration in dentistry.