Study On The Performance Of Woven Fabric Based On The Prediction Of Physical Properties

Fabric shearing, compression and air permeability are three importantproperties closely related with the handles of sofness and fullness,formability, and the comfortability of heat and moisture. This dissertationemphasizes on predicting the three properties with structure parameters. Themain aim is to explore the baisic theories and more accurate methods forscientific designing of fabric performances. The content of this dissertationis mainly included four parts as following:1．Fabric shear propertiesThe theoretical equation expressing shearing rigidity of woven fabricwith weave structure, fabric density and the rate of friction change wasderivated. Due to the difficult quantification of the rate of friction changeand the friction difference for different types of fabrics, the experimentalequations of the rate of friction change for each type of fabric were exploredrespectively. Taking worsted fabrics and filaments fabrics as the examples,the experimental equations expressing the rate of friction change with weavestructure and contact area beween warp and weft were established.Combining the experimental equations with the theoretical equation, thepredictive equations for shearing rigidity of the two types of fabrics weremade. The accuracy of predictive equation of worsted fabric was validatedshowing the average error below 10% and the maximum error below 20%.The theoretical equation expressing fabric shearing hysteresis withweave structure, warp and weft desinity, and shearing hysteresis at theinterlacing point was derived. Worsted fabrics and filament fabrics were usedto explore the relationship between shearing hysteresis at the interlacingpoint and fabric structure parameters such as contact area of warp and weft.The predictive equations for shearing hysteresis of the two types of fabrics ?were established by combining the theoretical equations with experimentalequations. The validation for the worsted fabrics shows the max-error below20% and average error about 10%.2．Fabric CompressionFabric compression work and compression hysteresis are closely relatedwith fabric thickness. The equation expressing fabric thickness with warpand weft yarn density and weave structure was established. Validation showsthat the average error below 10% and max-error below 20%. Linear modeland exponential model were use to explore fabric compression work.. Theresults show that two methods can well predict fabric compression work withthe average error below 10% and max-error below 20%. And the linear modelis simple and predictive error is lower, while exponential model can wellunderstand the compress process and the predictive error is stable. Twomethods of linear model, structure unit model were used to investigate theloss of compression work ( compression hysteresis). The validation showsthat the predicted error of the two methods is similar, while the predictederror of structure unit model is more stable.3．Air permeability of woven fabricsThe pores distribution and the relationships between the pores andfabrics' structures, permeability were discussed. The results show that thedistribution of pore sizes has a wide range from small to big with theasymmetry of peak value towards left, and the difference of the poresbetween interfibers and interyarns is obvious. The pores have the significantcorrelations with fabric cover factor, weave float, and air permeability theHagen-Poiseuille formula has been used to establish the equation expressingthe relationship between permeability of interyarn interstices ( Q₁ ) andfabric structures. The apparent diameters of yarns were investigated to modelthe interstice of yarn better. Validation with experiment data shows thatinteryarn porosity has very big influence on fabric permeability. The stepwise regression method has been used to explore the equation betweenthe permeability of interfiber interstices ( Q₂ ) and structure parameters. Theprediction equation for fabric permeability( Q) has been developed focusingon permeability of interyarn pores ( Q₁ ) and the fabric overall cover factor( E ) . One set of worsted fabrics is used to validate the equation. It showsthat except fabrics with z/z twist yarns or twistless filaments, the predictedequation gives good accuracy with average error of about 13%. BP neuralnetwork can also be used to predict air permeability of woven fabrics and thepredictive error is similar as the methods discussed above.4．Prediction for fabric performancesThe accuracy of the predictive equations is improved by comparing withthe existed experimental equations. The existed predictive system for fabricperformances was reviewed, and the modified system supported by theconclusions ahieved in this dissertation was described. The predictiveaccuracy of basic hands and formability with the modified system wereinvestigated. The results show that the basic hands calculated with modifiedsystem more approaching to the basic hands calculated with the measuredvalues, and the judge of fabric formability is better than that judged by theexisted experimetal equations.