Keywords

1 Introduction

Plastic Extrusion could be a high-volume producing method wont to turn out product of continuous lengths of constant cross section like pipes and tubes. To optimize the manufacturing processes, several strategies are developed and used over the years. Amongst all processes Design of experiments is that the most generally used technique. This is often one in all the foremost thoroughgoing approaches in product and method development. The foremost perspective to DOE is Full Factorial Design and Taguchi technique. Orthogonal array could be a technique recommended by Taguchi to review impact of many control factors over entire parameter space. Exploitation OA technique we tend to need lesser variety of experiments to be conducted, however Full Factorial style needs sizable amount of experiments to be conducted. To measure the performance characteristics which are deviating from the expected target value the concept of the loss function counseled by Taugchi. The value of this loss function is further regenerate into signal-to-noise(S/N) ratio. There are 3 S/N ratios of common interest that are: a. nominal-the-best, b. larger-the-better, and c. smaller-the-better. In present work smaller-the-better kind of control function, was employed in scheming the S/N magnitude relation. Akash Dip industry is a manufacturer of variety of best in class drip irrigation pipes, located close to Renapur, Dist. Latur-413527 (India).

2 Literature Review

Gadekar et al. [1] found that the inappropriate setting of operational parameters, recent machineries, compounding material with wrong proportion, poor testing and examination of material are the causes for the scrap within the company. Improper setting of operational parameters kept on an outsized share of the explanation for scrap and nonconformity of the merchandise. Taguchi’s technique of design of experiment for setting of optimum operational parameters is employed for minimizing scrap rates. Rao et al. [2] executed the design and optimization of extrusion process using Taguchi method and finite element analysis of temperature and extrusion load during extrusion of aluminum 6061 alloy. Taguchi L9 design matrix used for experimental run. Analysis of variance (ANOVA) technique is used to check the importance of the input variables on the output responses. The optimal process parameters are obtained using Taguchi’s approach. Kumar et al. [5] used Taguchi technique for optimization of process parameters in extrusion blown film machinery. To review the performance parameters on tensile strength orthogonal array and signal/noise ratio enforced. For higher quality characteristics bigger S/N ratio is chosen. Therefore, level with the greatest S/N ratio was elite. In these analysis four factors like melting temperature, winding speed, extrusion speed and pressure were elite for conducting experiments. By use of acceptable orthogonal array experiments were conducted. Afterward by measuring the tensile strength, Signal to Noise ratio was calculated. Exploitation MINITAB 17 computer code Main impact plots for S/N ratios were drawn and optimum parameter values were obtained. Narasimha et al. [7] targeted on to search out the basis causes for the prevalence of defects and wastes in plastic extrusion method. To spot the basis causes of those defects the cause and effect diagram was enforced. The important causes were the take off speed, temperature, vacuum pressure, screw speed and material properties as are often seen from cause and effect diagram. Exploitation Taguchi’s principle of loss function the quality loss for the present performance variation was calculated and demand for improvement was verified.

3 Problem Statement

The present analysis work is administered at Akash Drip Industries, Renapur. The categories of defects determined with frequency are as shown in Table 1. The subsequent information is employed for investigation purpose.

Table 1 The categories of defects with frequency data for-1 month observation
Full size table

Product: In-line Drip Pipe

Size: Diameter = 16 mm, Wall Thickness = 0.45 mm, Length = 300 m (Fig. 1).

Fig. 1
figure 1

In-line drip pipe

Full size image

4 Methodology

In present work design of Experiments (DOE) using Taguchi’s design for optimization of the process parameters is employed. The fundamental steps related to design of experiments (DOE) using Taguchi approach:

  1. i.

    Recognition of objective function to be optimized

  2. ii.

    Recognition of noise factor

  3. iii.

    Recognition of the control factors and their levels

  4. iv.

    Selection of appropriate orthogonal array

  5. v.

    S/N calculation and experimentation

  6. vi.

    Prediction of optimized setting from Taguchi analysis

  7. vii.

    Prediction of optimum control factors levels and its performance.

4.1 Recognition of Objective Function

Our objective is to minimize the uneven wall thickness defect of drip pipe. Hence, objective Function: Smaller-the-Better is employed to reduce the response. S/N Ratio for this function:

$$ S/N = - 10\log_{10} \left( {1/n\sum\limits_{i}^{n} {y_{i}^{2} } } \right) $$
(1)

where, n = Sample Size, and y = % defect in that run.

4.2 Recognition of Process Parameters

The factors that have an effect on the drip pipe dimensions and cause the defect of uneven wall thickness are as per Table 2.

Table 2 Factors that have an effect on production of drip pipe
Full size table

The factors that considerably have an effect on the performance are taken into consideration in constructing matrix experiment. All alternative factors are considered as Noise Factors. The subsequent 3 parameters that are most affecting are designated for performing the matrix experiment:

  1. (i)

    Take-off speed (m/min)

  2. (ii)

    Temperature 1(°C) (Barrel zone 1)

  3. (iii)

    Temperature 2(°C) (Coating zone 1).

4.3 Recognition the Control Factors and Their Levels

The extrusion setup has various thermocouples connected to the extruder and coating machine. The most important temperatures are BZ1 and CZ1 that are to be set and alternative are to be accrued incrementally within the increments as in extruder BZ1 and BZ2 increments are by nearly 5 °C, in extruder SC, MZ1, DZ1 by 25 °C and in coating zone increments by 25 °C. For optimization purpose base temperature BZ1 (Temperature T1) and CZ1 (Temperature 2) is chosen as control factors. Because the material provider changes the defects incidence range and the parameter settings also changes. Hence there’s need to optimize these control factors for every of the provider. So, experiments are designed differently for every material provider and process parameters are optimized consequently. The factors and their levels are shown Table 3.

Table 3 Control factor levels for three suppliers
Full size table

4.4 Selection of Orthogonal Array

The Taguchi optimization technique starts with choice of associate accetable orthogonal array. The minimum number (min N) of runs within the array may be determined as follows:

$$ N_{ \hbox{min} } = (L - 1)F + 1 $$
(2)

where, L = number of levels, F = number of factors.

where, L = 3 and F = 3.

Now Taguchi’s array is chosen for experimentation having minimum seven numbers of experiments. Total 9 experiments were administered as a result of the foremost appropriate orthogonal array for experimentation having three factors and three levels is L9 array.

4.5 Experimental Programme

With relation to the above OA, nine experiments were conducted with their factors and their levels as mentioned in Table 3. The defect uneven wall thickness for every experiment was measured and % defects for every supplier calculated. For every supplier want one separate experimental array of nine runs. Thickness variation of the pipe was measured using the screw gauge. Each of the above nine experiments were conducted three times (27 experiments for every supplier) to account for the variations which will occur because of the noise factors. The temperature instrument panel to manage temperatures, the machine control panels to manage takes off speed and different parameters and also the die section of machine are shown in Figs. 2, 3 and 4 severally.

Fig. 2
figure 2

Temperature control panel

Full size image
Fig. 3
figure 3

Machine control panel

Full size image
Fig. 4
figure 4

Die section

Full size image

5 Results and Discussion

Experimental work is administered at Akash Drip Industries, Renapur, Latur for 3 different material suppliers particularly Vardhaman Polymers Ahemadnagar, Lata Industries Latur, Bansal polymer Hyderabad. The experimental results are mentioned in Tables 4 and 5. Sample Calculation:

Table 4 OA with control factors for three suppliers
Full size table
Table 5 Percentage of defects for different suppliers
Full size table
$$ \begin{aligned} & {\text{The}}{\mkern 1mu} {\text{percentage}}{\mkern 1mu} {\text{of}}{\mkern 1mu} {\text{defect}}{\mkern 1mu} {\text{per}}{\mkern 1mu} {\text{shift}} \\ & \quad = \left( {{\text{Number}}{\mkern 1mu} {\text{of}}{\mkern 1mu} {\text{defective}}{\mkern 1mu} {\text{products}}{\mkern 1mu} {\text{having}}{\mkern 1mu} {\text{uneven}}{\mkern 1mu} {\text{wall}}{\mkern 1mu} {\text{thickness}}} \right)/ \\ & \quad \left( {{\text{Total}}{\mkern 1mu} {\text{number}}{\mkern 1mu} {\text{of}}{\mkern 1mu} {\text{units}}{\mkern 1mu} {\text{produced}}{\mkern 1mu} {\text{per}}{\mkern 1mu} {\text{shift}}} \right)*100 \\ \end{aligned} $$
(3)

It was found that, total 65 units are produced per shift out of that 7 defective units having uneven wall thickness.

% defect = 7/65 * 100 = 10.769% … (Putting values in Eq. 3).

After conducting the experiment, the results were converted into S/N ratio values. For these experimentation and analysis using Taguchi results are adopted with the assistance of Minitab-18 software package. Smaller-the-better kind of control function was employed in calculating the S/N ratio. Main effects plot for the main effect terms viz. factors take-off speed, temperature 1 and, temperature 2 are shown in Figs. 5 and 6.

Fig. 5
figure 5

Main effects plot for means

Full size image
Fig. 6
figure 6

Main effects plot for S/N ratios

Full size image

6 Confirmation Experiment

Confirmation experiments conducted using the factor levels obtained by S/N ratio plot, using Taguchi method, for each supplier the level of factors found different. Total three experiments were conducted and their % defects values were checked. It can be seen that the results are having minimum defects (Table 6).

Table 6 Confirmation experiment
Full size table

7 Conclusion

The Following conclusions are drawn from the current study:

  1. (1)

    The experimental results show that the temperature and take-off speed are the main parameters which will have an effect on uneven wall thickness defect.

  2. (2)

    The optimum values for influential factors for lesser percentage defects in drip pipe extrusion are obtained from Taguchi’s methodology for various material suppliers.

  3. (3)

    Using optimized values of factors percentage defects for uneven wall thickness reduced from average of 4.19, 4.103, 5.04 to 1.38, 1.66, and 1.18 severally for 3 material suppliers particularly Vardhaman Polymers Ahemadnagar, Lata Industries Latur and Bansal Polymers Hyderabad.

  4. (4)

    The current work helps to boost the quality of product and to optimize drip pipe extrusion process.