STITCHESt SEAMS AND STITCHINGS 1.3 FORMATION OF TYPE 301 LOCKSTITCH 1.4 THE NEEDS OF STITCHES AND SEAMS 1.5 THE CONCEPT OF SEAM STRENGTH AND 1.6 SURVEY OF PREVIOUS WORK 1.7 REASONS FOR THE PRESENT WORK LITERATURE

In gen(-, ral ,a seam may f ai I as a result of longitudinal or transverse loading. Previous workers have reported transverse seam strength to be a prime requirement in most cases of Joining woven fabrics by the means of sewn seams. Transverse seam breaks can be divided into two types: Type I in which only the sewinq thread breaks (with Lhe fabric undamaged); and Irype II in which only the fabric breaks (with the sewing thread undamaged). Most previous workers have tested seam. -, trenqth in woven fabrics by usinq sampIcs in which the warp ur the ureft yarns ran straight from jaw to seam to jaw; this configuration of seem loading, however, is uncommon in practical qarments. Accordi rin I y, for Fý(-, am ,; t ren(It Iii rvest Iq at i OnF; to Iia vc, a ry T (-, I I practical applicability, J-t isnecessary to examJ. ne searris in whJch t)ie fabric HireaJs run at an anglu, of bias to the line of the sea., u; such being the configuration of fabric and seam in most garments. The objectives of the present work were as follows: (a) to investigate empirically the effects of angle of bias, gauge length and sample width on Type I and Type II lockstitch seams and unseamed fabrics over a range of woven falDrics of simple structures; (b) to develop a generalized thecretical mathematical and I)hysA'ca3 model in an attempt to explain tle relationships ý, etween seam ý; trength and parameters such as angle of hias, gauge length, width of sealn, fabric type and construction; (C) to attempt to understand the cmplexities of a sewn seam in woven fabrics under load that an optimwn configuration of the various parameters investigated (together with a consideration of garment design and pattern cutting) could be achieved for any given practical situation. The experimental results showed a marked effect of the angle of bias together with the interaction of the ratio G/VI! 'W' shaped curves were evident under almost all of the situations except where the G/W1 ratio was below half. A basic mathematical model, developpd from previous work, hashown very good agreement. with the experimental results for a square-sett plain-weave fabric. Further modifications have been made to extend the application of the model to non-square-sett and non-plain-ioeave fabrics. The thcoretical and experimental restilts fit to a reasonable standard. In order that the experimental. arid theoretical works could be related to practical situations, a. survey of common garments, and the angles of seam/fabri. c bias contained in them, was undertaken. Spvcral indiistrial applications have been -, uggested. Another theory for predicting the peak angle, the only unknown parameter in the model, has been -. -, uqgc-sted in the final chapter and can be seen to work well within the li-mitatLon of the fabrics used in this work. G is defined to be the gauge-length (jaw to jaw for unseamed--sample-s or jaw to seam for seamed samples). W is defined to be sample width.