Researchers and sewing thread manufacturers had different approaches in developing thread consumption calculations. Thread length required for a unit length of a stitch is the most frequent measurement (Ukponmwan et al. 2000) which provides flexibility to calculate the thread consumption for a given length of stitch. Measuring the actual length of threads in the stitch is one prominent method which gives correct results if measuring standards are properly followed. Actual lengths of thread are measured by unraveling the formed stitch carefully under correct tension of threads. Unraveled threads are measured following standard testing methods since they are in-build with crimp. As an example, the French standard NF G07 101 is used to determine the unraveled thread lengths (Jaouadi et al. 2006). Though unraveling method gives an accurate value, the stitching operation has to be performed several attempts to improve the accuracy by optimizing sewing parameters. The measuring process is time consuming and it requires testing cost for material, and equipment and people. Test operator should be highly skilled person to obtain exact values of yarn lengths without exposing them to excessive tensions or distortions.
With these drawbacks of the actual measurement of thread consumption, researchers approached on prediction techniques applicable to determine thread consumption as quick and viable solution. Value prediction charts, mathematical formulae, thread length ratios, predictive algorithms based on past data, learning algorithms and software solutions (Jaouadi et al. 2006) are several methods available in literature which use to predict the thread consumption. When analyzing the prediction values of each techniques for a given stitch length with same input parameters in same stitch configuration, it has shown a significant scattered behavior as well.
Jaouadi et al. (2006) reveals that parameters such as stitch density, fabric thickness, thread linear density and seam width are mainly being used to predict sewing thread consumption. In practice, garment manufacturers calculate thread requirement based on consumption estimate charts provided by specific thread suppliers (Carr & Latham 1994; Amann Group 2010; American & Efird Inc 2007). The charts calculate thread consumption using different variables and assumptions. They either estimate approximated thread consumption per garment or thread requirement per unit length of the stitch. Since these charts can be used under given specific scenarios; i.e. for given set of stitch densities or for a given thickness of fabric etc., it provides less flexibility in using for varying stitch densities and fabric thicknesses.
Introduction of consumption ratios is another prediction technique which decides ratio of thread amount with respect to the geometry of the stitch. Different ratios were derived as per the stitch types though these ratios have limited to one stitch density value. To overcome these disadvantages, thread suppliers subsequently developed consumption ratios for selected stitches which facilitate calculation of thread consumption for given set of stitch densities and fabric thickness values (Amann Group 2010; American & Efird Inc 2007).
Most of leading thread suppliers currently uses software packages to improve the efficiency of thread consumption calculation with different formulae and ratios. These software packages are capable on calculating thread consumption for varying parameters such as stitch lengths, stitch densities, fabric thickness etc. Majority of stitch-classes used in the apparel sector have been defined in these systems. Parameters such as seam widths are also included with the appropriate stitch type. Unlike previous techniques, thread consumptions are calculated separately for needle threads, bobbin threads and looper threads. Though certain software solutions address thread properties such as ticket number for thread consumption calculations, majority of solutions have not addressed some critical physical properties which associates stitch formation.
Previous studies specify parameters such as stitch type, stitch density, fabric thickness, and seam width determine correlation to the thread consumption as per the stitch type being considered (Samuel & Poojitha 2010; Jaouadi et al. 2006). Experimental study of Jaouadi et al. (2006) on cloth thickness and stitch density revealed a significant effect on thread consumption with a positive correlation for stitch types 301, 401, 504 and 516. In addition to fabric thickness and stitch density, yarn count of the thread has influenced on thread consumption. Further, they revealed that fabric type do not have a statistically significant effect on thread consumption (Jaouadi et al. 2006).
In previous studies by Rengasamy and Samuel (2011) on needle thread tension for lock-stitch 301, four tension peaks have been signified by measuring the online tension of stitch formation. They reported that the threads are extended dynamically at the tension peaks mainly when needle descends and penetrate through the fabric during the stitch formation (Rengasamy & Samuel 2011). Especially in stitch tightening; the last step of the stitch formation produces highest tension peak also apply an extension to the sewing thread (Weimer & Mitschang 2001; Rengasamy & Samuel 2011). Thread tension was identified as one of the important parameters which influence the quality of the seam in garment construction. Krishnan and Kumar (2010) have attempted developing reliable tension measuring devices to measure on-line tension of the running thread. Ferreira et al. (1994a, [b], [c]) have studied thread tension on lock stitch sewing machine and investigated the tension behavior of both the needle-thread and bobbin-thread.
Some researchers have approached theoretical techniques such as neural networks, statistical and geometrical methods to calculate sewing thread consumption (O'Dwyer & Munden 1975; Hayes 2001; Kennon & Hayes 2000; Jaouadi et al. 2006; Ukponmwan et al. 2000; Amirbayat & Alagha 1993). Jaouachi et al. (2012) developed a method to calculate the consumption of the sewing thread of jean pant using taguchi design analysis. The recent efforts of thread consumption research (Jaouachi et al. 2012; Rasheed et al. 2014; Ghosh & Chavhan 2014) demonstrate that still researchers tend to examine on better formulae to minimize the errors occur in using the existing formulae. However, thread tension has not been taken as a parameter for any of the above studies. This study investigates the influence of thread tension to the thread consumption of lock-stitch 301 and chain-stitch 401 while incorporating the parameters which had been already been examined by previous authors (Jaouachi et al. 2012; Rasheed et al. 2014; Ghosh & Chavhan 2014).