【作者】 胡吉永；

【导师】 丁辛； 王如彬；

【作者基本信息】 东华大学 ， 纺织工程， 2008， 博士

【摘要】 织物的触觉质感狭义上指人手触摸织物时的感知觉(简称织物手感),是一个多维感知属性概念,每一维感知属性称为基本质感。织物的触觉质感决定了织物的质量品质,指导织物的定价,是纺织品生产和销售领域的关键评价指标。然而,织物的触觉质感是一种主观评价的结果,不仅受到评价者的文化背景、个人喜好和情绪等主观因素的影响,而且定量标准难以掌握。由于织物触觉质感的形成机制尚未揭示,以至迄今为止织物的感官评价没有相应的国际标准或国家标准,严重地制约了织物触觉质感信息的交流和评价结果的客观性。定量、科学地表征织物触觉质感是纺织领域的科学问题。大量的工作试图通过仪器对织物基本物理属性的测试来代替相应指标的感官分析,建立包含织物基本物理属性指标、基本手感和综合手感三者之间的统计演化模型,但这些研究迄今取得的工程化效果甚微。不成功的主要原因之一是这些研究未能从人的认知规律出发,忽视织物触觉质感形成的生理过程和触觉质感的认知心理学规律,而将织物用途和服装款式视为分类的重要依据,建立纯数学的统计关系。研究表明,触觉形成的生理过程和认知心理学规律都涉及高度复杂的非线性现象,难以用简单的解析关系表示。迄今为止,仅屈指可数的几位学者定性讨论了织物与皮肤接触时的诱发生物电活动,但没有揭示纺织品触觉质感的形成过程和人的感观识别能力的有限性,不能满足感官工程参数化设计需求。本课题将从认知科学角度开展对织物触觉质感的初级认知机制的研究,通过分析与触摸法评价织物触觉质感固有范式等效的生物力学问题,针对纺织品的特征建立指尖与柔性物体接触的有限元模型,结合现已发现的人触觉识别、辨别物体的认知规律,对织物触觉质感与基本触觉质感的概念、内涵、组成及其感知机理进行探讨,把织物触觉质感的主观评价这一工程问题从触觉认知科学的角度给予研究。课题的基本假设是:织物表面几何属性、材料属性以及指尖生理结构主要影响慢适应I型感受体(简称SAI感受体,包括Merkel感受体细胞及其换能器)所感测、辨别的应力应变模式。以此为前提,课题的研究内容为:(1)从触觉认知科学角度对已有狭义织物手感的研究进行归类,建立织物触觉质感感知属性空间的概念,这是前人未做过的,是本论文的基础工作。从触觉认知科学角度,论文提出了织物触觉质感及其感知属性空间的概念,并对已有织物手感的研究按认知心理学和认知生理学的范畴进行归类,这对于当前以追求统计演化数学模型的预测精度的主流研究来说,在研究思路和方法上是一突破。结合已知的触觉识别硬质物体的机理,论文评价了对织物软硬感和粗糙感的各类研究,从认知科学角度分析现有的研究存在的问题,在此基础上,构建表征织物触觉质感的一般框架。(2)建立包括多尺度、多层次解剖结构特征的指尖有限元模型,并证明已有模拟指尖的弹性半空间模型、“水床”模型,及表示两弹性体接触的Hertz模型不适合模拟指尖触压物体。这为后续工作中探索具有柔性纹理表面的织物性能(柔软性和粗糙性)的认知机理选择了一个有效的工具。以建立的多尺度、多层次有限元模型,讨论了指尖受硬质的平面、圆柱、圆锥、“T”型块刺激作用时,软组织的机械响应与在体生物力学实验的一致性,及与皮下触觉感受器诱发放电率的相关性。同时,研究发现,在将外载荷集中于指尖敏感点及其邻域的过程中,指骨扮演了十分重要的角色。这表明,在研究指尖软组织的结构特征对触觉感知的影响时,弹性半空间模型或“水床”模型不适合。进一步通过指骨替换实验发现,在描述指尖触压织物表面的接触状态时,经典Hertz接触模型将导致较大的误差。给出了表征平纹机织物正交各向异性力学属性的本构关系,并以平纹机织物为原型,建立了指尖触压均质、线弹性柔性平板的有限元模型,讨论所建立的指尖模型解释手指触压柔性体时的生物力学现象的能力。研究证明,由于柔性物体的表面可变形性,手指通过触摸方式识别、辨别硬质物体和柔性物体时,分别基于不同的变形模式。(3)证明指尖触觉感测系统的机械敏感性,以及在评价织物机械属性时触觉感测与仪器测量之间存在本质差异。从织物机械属性的初级认知角度,纠正了当前构建织物机械属性的人体感测和仪器测量之间关系时的片面认识。依据评价织物触觉质感时手指的固有运动范式,以触压法评价织物软硬性为例,结合指尖及嵌于指尖的触觉感受器的生理特征,建立了与触压过程等效的集参化生物力学模型,并以此提出描述触觉感受体对作用于皮肤表面的外刺激敏感程度的指标—机械敏感性。以机械敏感性为因变量,分析了指尖触觉感受器关于不同织物软硬性的辨别能力,以及辨别同一织物时人与人之间、个人不同体区之间感测能力的差异。研究发现,手指感知织物的软硬性时,机械敏感性阈限取决于表征织物非线性压缩变形的常数与指尖软组织相应常数的比。其中,实验室用仪器的测试过程等效于指尖软组织非常硬时的情况,其机械敏感性接近“1”。从这种意义上,在感测织物软硬性时,人体机械刺激感测系统与仪器感测系统之间存在本质差异,从而使得统计意义上差异显著的仪器测量值,人指尖的触觉感受器并不一定能分辨。然而,当前在织物手感研究中,研究者往往默认二者感测阈值相当。考虑到集参化模型以指尖触压织物的固有范式为依据和基本力学原理为理论基础,研究结果可推广至其它织物基本机械属性的人/机测试差异分析。(4)提出表征感觉系统感受性的指标—感觉敏感性,证明人对织物软硬性的触觉辨别能力有限,织物表面属性影响软硬感的初级认知,及柔软性感知过程遵循触觉流编码理论,促进了对当前主流织物手感模型预测效果的理论解释,指出现有织物柔软感的虚拟再现技术的缺陷。基于触摸法评价织物表面软硬性的手指运动范式,结合触觉感知的心理物理规律,引入表征触觉系统感受性的指标—感觉敏感性,研究了织物柔软性的感觉敏感性随织物属性和个人触觉系统属性的变化规律。研究发现,在织物柔软感的高级认知过程中,感知者触觉系统的感觉敏感性(或感觉阈限)不同于初级认知过程中触觉感受器的机械敏感性,前者不仅与后者有关,还与刺激模态及感觉形态有关。这形成织物属性的主观估值的固有离散性,有别于仪器测试的高度可重复性。则理论上,织物触觉质感的主观估值为某种概率分布,并不是一个平均特征值。然而,在大多数纺织品感官分析中,人们往往把仪器测试或模拟预测的显著性差异等同于人体感官分析的感觉阈值和感觉阈限,忽视了人体触觉系统感受性的有限性。其结果是,以数学统计关系把两个敏感尺度不同的平均估计值连接在一起,既不符合生物学规律,也违背了基本的因果关系论,在研究方法上偏离人的认知规律。本课题将首次从认知科学角度,揭示现有的织物触觉质感预测模型鲁棒性差的根本原因。为了具体分析织物表面属性对软硬性触觉认知的影响,引入代表平纹机织物的正交各向异性力学属性于指尖/柔性平面有限元模型中,分析了织物的三个特征物理属性(即:表面摩擦系数、压缩模量和面外泊松比)对软硬性初级认知的影响。研究发现,织物低应力下的初始压缩模量和表面摩擦属性越高,感知者对织物柔软性的触觉印象越弱,且这种触觉印象的形成符合触觉流编码理论。(5)参照SAI感受体以潜在适宜刺激对硬质纹理表面几何属性的编码机理,审定现有的编码模式解释织物表面几何属性感知过程的可行性,证明应变能密度为符合神经系统信息加工效率的适宜刺激变量,并以Logistic规律编码织物表面几何属性和材料属性。这是一项基础性的研究成果,有利于推动基于人体触觉感测机理的织物表面属性的仪器化表征方法的发展。通过引入代表平纹织物表面的周期性纹理于指尖/柔性平面有限元模型,探索了SAI以适宜刺激对织物表面纹理属性的编码模式。研究证明,指尖SAI感受体邻域软组织的应变能密度是织物表面几何形貌的映象,而非同型像,且映象对比度取决于织物的初始压缩模量和表面几何属性的复合效应;应变能密度的空间变化与这些属性特征值之间满足Logistic函数关系。同时,基于适宜刺激强度—外围神经响应满足线性关系的假设,确定了描述织物表面粗糙感至少需要四维变量,即:表面形貌波动幅度、轮廓峰—峰中心间距、单峰宽度和表面压缩模量。这些结论有助于实现触觉质感的参数化设计和虚拟再现。研究结果表明:现有对织物手感的评价、预测模型中,由于忽略了触觉感知系统的感知、辨别的有限性和纯数学模型的数值显著性之间的差异,使所建立的数学模型掩蔽了人体感觉系统辨别能力的局限性,其结果往往扩大了感官感知范围和分辨能力。由于几何属性和材料属性的相互影响,对于织物这类具有表面纹理的薄片状柔性复合体,其软硬感、粗糙感的初级认知模式,不同于以砂纸为代表的硬质纹理表面粗糙感的认知模式、也不同于以几何维度相当的柔性体的软硬性感知。本课题使用的研究方法具有新颖性,它试图构建感官感知和生理模拟之间的桥梁,通过每一项具体的研究内容,以从织物物理属性指标到触觉感受体的近端刺激及认知规律对触觉通道的感受性予以量化。这个角度的认识有助于获得工程化设计模型,从而使设计者有可能集中精力于几个关键刺激属性,得到令消费者更加满意的纺织品。迄今为止,虽然人们对触觉的主要生理机理已深入了解,但关于机械换能过程的许多实用的细节还了解得不透彻,尚不能定量预测人们怎样感知作用于指尖的机械刺激。本论文的结论将有助于推动触觉质感的定量表征。同时,由于皮肤和物体表面之间微妙的相互作用,本课题的研究将有利于理解SAI感受体探测物体表面几何特征和材料属性的机械信号转导过程。更多还原

【Abstract】 Tactile textures of fabric,in a narrow sense of the word,are perceived by human hand, namely fabric handle.Its space consists of multi-dimensional perceptual attributes,of which any principal dimension is called as the primary tactile texture.Tactile textures of fabric determine its quality and performance as well guide price.And then,it is considered as the key index evaluating the performance in the production of textiles and their sale.However,tactile textures of fabric are perceived by human,and the perception is affected by both its own properties and many subjective factors,involving cultural background,personal preference and personal emotion,et al.As a result,it is difficult to establish a criterion quantifying tactile textures.The status is owed to the lack of information for the physiological mechanism of fabric touch sensation,so that there are no specific and prevalent international or national standard for sensory evaluation on textiles,which leads to poor objectivity and difficult communication of evaluation results.Quantifying tactile textures of fabric is a scientific problem in the field of textile engineering.Various attempts are made to substitute instrumental testing for human sensory evaluation on fabric physical properties,and to develop all kinds of statistical relationship between principal physical variables,primary handle and total handle. However,the performance is poor in the parameterized design of new textile products. The reasons for failure in practice are that the developed models are based on statistical mathematics and classification into clothing end-use and style,and neglect the physiological mechanism of sensation and perception,which involves complex nonlinear phenomenon.Up to now,only do numbered researchers record and quantitatively analyze the evoked firing rates of cutaneous mechanoreceptors when human skin contacts with fabric surfaces,and the information for how to sense fabric properties,whether geometric or material,and tactual limitations as well,is lack.Importantly,the lack has much impact on parameterized design of fabric touch sensation.Therefore,this dissertation focuses on the primary cognitive mechanism of fabric touch sensation.By reviewing the biomechanical and electrophysiological phenomena related to paradigms of human touching fabric surfaces to extract perceptual attributes,this stuty makes a parameter-lumped and a Finite Element(FE)biomechanical model equivalent to the specific paradigm.Based on the developed models as well the recognized cognitive mechanism in human identifying and discriminating objects by touch means,this dissertation introduces and explains the tactile textures of fabric,and also discusses the underlying physiological mechanism of sensation and perception.In this sense,sensory evaluations on fabric physical properties in the filed of textile engineering are understood in the perspective of tactual cognitive science.This dissertation considers a central hypothesis that declares that the geometric and material properties of both fabric and human fingertip will significantly affect the patterns of stress and strain that are sensed and discriminated by SAI populations(the Merkel cell complex,and its transducers).Based on the hypothesis,the following specific contents are studied:Firstly,this paper reviews and classifies the existing recognition and understanding of fabric handle from the point of tactual cognition,and introduces the tactile textures of fabric,which is the basic work in the study.From the point of tactual cognition,this dissertation introduces the tactile textures of fabric and underlying perceptual attributes space,and then reviews and classifies the existing work of fabric handle within the framework of cognitive psychology and cognitive physiology.The present idea breaks out the general pursuit in mathematically improving the precision of prediction for fabric handle.Furthermore,this study summerizes the recognized physiological mechanism of human identifying and discriminating softness and roughness of objects in detail,including textiles,and the shortages of existing work are introduced in the view of tactual cognition.Thus,the general framework is established to characterize tactile textures of fabric.Secondly,a multi-dimensional and multi-level fingertip FE model is developed and validated.This study discovers that all of the homogeneous elastic half-space model and“water bed”model simulating fingertip,and the classical Hertz contact model as well,are unappreciated to simulate the case of human fingertip in contact with objects.This work chooses an efficient tool for exploring cognitive mechanism of tactile textures of fabric, involving softness and roughness.Based on a multi-dimensional and multi-level FE model,the mechanical responses of soft tissues within fingerpad to the loading displacement with a certain distribution, respectively,including uniform-distributed loading by a hard plane,curve-distributed loading by a cylinder,concentrated loading by punch and sequential loading by a“T”-bar, are validated by in-vivo biomechanical experiments and the evoked firing rates of SAI populations.Meanwhile,this work demonstrates that the phalanx bone within human fingertip plays an important role in focusing the external work on soft tissues in the vicinity of dense cutaneous mechanoreceptors underneath the maximum sensitive spot of fingerpad.It means that the elastic half-space model and the“water bed”model are inappropriate when the effect of structural and physiological characteristics of finger on tactual discrimination is involved.By subcutaneous tissues taking the place of bone, when the mechanical responses of fingerpad pressing toward fabric surfaces are predicted by the classical Hertz contact model,a significant deviation from the simulating resuts is discoveredOn the other hand,the orthotropic constitutive relationship characterizing mechanical properties of simple plain-woven fabric is proposed.By the protype of a plain woven fabric,a FE model of fingerpad contacting homogeneous compliant planes is developed,and the ability of the FE fingertip model in predicting the activated biomechanical responses is discussed.This study concludes that the deformation kinetics of fingertip pressing toward compliant planes is different from that case of hard objects.Thirdly,this work uncovers the mechanical sensitivity of tactual sensing system of human fingerpad,and the intrinsic difference between performance of cutaneous mechanoreceptors and that of instrumental testing in delecting fabric physical properties. In the view of human primary cognition,furthermore,it rectifies the misunderstanding of the relationship between human sensing discriminability and detectability of instrumental sensors in studying fabric handle.A parameter-lumped biomechanical model equivalent to the paradigm of fingerpad touching fabric between two fingers or between one finger and hard platform is developed,and the mechanical sensitivity,a variable characterizing the sensitivity of cutaneous mechanoreceptors is proposed.And then,in terms of the mechanical sensitivity, the discriminability of cutaneous mechanoreceptors in the mechanical resistance against compression(MRC)of fabric surfaces is analyzed parametrically,and the testing process of instrumental sensors in lab is comparable to the case of mechanoreceptors within soft tissues with high MRC.The work demonstrates that the mechanical sensitivity of cutaneous mechanoreceptors within fingerpad depends on the ratio of MRC of fabric to that of fingerpad.In this sense,the intrinsic difference exists between the ability of human sensing system and instrumental that in detecting MRC of fabric,so that the cutaneous mechanoreceptors of human fingerpad can’t always identify and discriminate the detectable differences by instrumental sensors.In the existing studies on tactile textures of fabric,however,both of them are mistakenly deemed to be comparable. Considering the common principle of the developed parameter-lumped model,the above-mentioned conclusions can be generalized to a deep understanding of the intrinsic difference between human sensing system and instrumental that.Fourthly,the perceptual sensitivity,a variable characterizing human discriminability in evaluating the mechanical resistance against compression of fabric,is proposed,and the limitation of human tactual system is validated.The effect of typical fabric surface properties on softness sensation and the encoding pattern on fabric softness by tactile flow as well is discovered.This work rectifies the previous misunderstanding of fabric handle predicted by the existing models,and gives the shortage of current virtual rendering technique of fabric softness.Based on the active paradigm of fingerpad touching fabric surfaces for softness evaluation and the psychophysical law,the perceptual sensitivity,a variable characterizing the diseriminability of human tactual system are proposed,and its changing trend with MRC of fabric surfaces and properties of tactual modality is analyzed parametrically.The study demonstrates that the perceptual sensitivity of human tactual system is different from the mechanical sensitivity in primary cognition,and depends on both mechanical sensitivity and sensory modality.It means that the sensory estimations of tactile textures of fabric are intrinsically scattered,and should obey a certain probability distribution along the mechanical intensity continuum.In most of sensory evaluation on the handle of various fabric,however,the testing significance of their instrumental recording magnitudes or their fabric handle predicted by models are mistakenly considered as the perceptual discriminability.As a result,the developed statistical relationships between subjective and instrumental estimations with different sensitivity go against human cognition and the effect-cause theory.From the point of cognitive science,thus,this study will uncover the underlying cause leading to the poor prediction of existing models for fabric handle.Meanwhile,the effect of specific fabric surface properties on softness sensation can’t be observed by the parameter-lumped biomechanical model,and then a FE model of fingertip pressing toward orthotropic woven fabric is developed.The typical physical properties of fabric surfaces,involving friction coefficient,initial compression modulus and out-plane poisson ratio,are discussed parametrically to discover their effect on cognition of fabric softness.The results demonstrate that both the initial compression modulus and the friction coefficient have a significant impact on tactual cognition of softness.And the small the friction coefficient is,the softer can is perceived by human. Furthermore,the softness is encoded by tactile flow.Thus,the virtual rendering of softness on a single physical property variable will has poor fidelity.Finally,based on the preferred or proximal stimuli of SAI populations encoding on geometrical properties of hard textures,this work validates the ability of the recognized encoding pattern on roughness in explaining the sensing process of tactile textures of fabric,and also strain energy density is the proximal stimulus to SAI populations with respect to the efficiency of human tactual system processing information from exterior stimulus.Furthermore,a model in conformation to the recognized physiological mechanism of tactual cognition is proposed to explain the relationship between physical properties of fabric,whether geometrical or material,and the proximal stimulus.This work is original,and will facilitate the development of methods characterizing fabric surface properties in conformation with human tactual cognition.By a FE model simulating fingerpad touching fabric with periodic surface textures and orthotropic mechanical behavior,this study proposes a tentative model of SAI populations encoding fabric surface properties by proximal stimulus.The focus is on the cross-interaction between geometrical and material properties in tactual cognition.The results indicate that strain energy density as the proximal stimulus to SAI populations encodes the initial compression modulus and geometrical profile features in a logistic law. And then,in the terms of the linear relationship between the evoking firing rates of SAI populations and the proximal stimulus,the dimensions characterizing fabric surface roughness sensation are at least four,namely asperity height and size,center-to-center distance and initial compression modulus.Together the above contents suggest that the limitations of whether identification or discrimination of human tactual system has been misunderstood in the existing statistical models predicting fabric handle,and the significant difference in statistic test is usually considered as the discriminable stimulus intensity by human,and the mechanical sensitivity of man-made instruments as the absolute thresholds in perception.As a result, those developed models obliterate the discriminability of human sensory system.On the other hand,for softness-hardness and roughness-smoothness of textiles,the primary cognitive pattern is different from that of hard texture surfaces and compliant objects with nearly same dimensions.The general research approach pursued in this dissertation is innovative because it seeks to bridge the gap between physiological modeling of sensory perception on fabric and its principal physical properties.With the development of specific contents,drawing upon the strengths of each model quantifies limitations of the touch modality in transforming distal stimuli(the physical properties of fabric)into proximal stimuli to cutaneous mechanoreceptors within fingerpad.This perspective may lead to an engineering performance model that would allow designers focus their efforts on critical parameters,and gain insight into more satisfied fabric products by costumers.Although the major physiological mechanisms underlying touch are well understood, many practical details of the mechanotransductive process are not yet understood well enough to quantitatively predict how people perceive stimuli at the fingertip.Furthermore, in terms of the subtle interactions between skin and object surfaces,this dissertation facilitates a deep understanding of mechanotransduction of geometrical and material properties encoded by SAI populations.更多还原