Weld lines in nylon 6 melt-blended nanocomposites.

INTRODUCTION

In recent years, layered silicate nanocomposites have attracted a great deal of interest from researchers, being a new class of composite that contain nanoparticles dispersed in a polymer matrix [1]. Nylon 6 nanocomposites are usually comprised of montmorillonite-containing silicate layers that are ~1 nm in thickness and are preferably fully delaminated and uniformly dispersed within the polymer matrix [2]. Nylon 6 organoclay nanocomposites can be produced by either melt blending, in situ polymerization [3], or solution [4] mixing techniques. Melt blending is the most common method for preparing nylon 6 nanocomposites, with the organoclay directly mixed with the molten polymer. The melt blending technique is economical and straightforward, and while maximum exfoliation of organoclay is not usually achieved, such a combination with nylon 6 produces a relatively high level of delamination [5].

The injection molding process involves the injection of a polymer melt flow into a cavity mold where the melt cools and solidifies to form a plastic product and is a three-phase process comprising filling, packing, and cooling phases. Weld lines are a source of mechanical weakness and their formation is an unavoidable defect in injection molding due to the complexity of modern product design. The formation of weld lines inevitably occurs when there are cores, pins, and multiple gates, all of which divide the molten polymer flow pattern in the cavity mold. While weld lines can be deleterious in homopolymer moldings, the problem can be amplified in two-phase systems, such as glass-reinforced thermoplastics [6].

Weld lines also consist of v-notches at the weld line surface, which is often forgotten as one of the possible causes of weakening, potentially acting as stress concentrators by increasing the stress level at the bottom of a v-notch [7]. Chang and Faison [8] reported that v-notch appearance and weld line locations depend on the design of mold and product, and subsequently processing conditions. The appearance of weld lines is found to be more visible on one surface than the other of the molding. Using optical microscopy, fan-shaped v-notches are noticeable at the edges of the weld lines. It was reported that the fan-shaped appearance is caused by the polymer melt fronts impinging at the weld region [8]. The different appearance of both sides of the specimen is attributed to the geometry of the specimen and the location of the mold parting line, through which air escapes from the advancing melt fronts.

A number of remedies for controlling processing conditions have been investigated in order to enhance the molecular diffusion at the weld line boundaries in injection-molded products [9]. The most significant processing parameters that have been identified in polymers such as polystyrene, polypropylene, polycarbonate, and polyoxymethylene (POM) are, in decreasing order of importance: melt temperature, mold temperature, holding pressure, holding time, and injection velocity. Temperature was found to be the dominant factor influencing the weld line strength. Liu et al. [10] investigated the temperature corresponding to the healing ability of a material at the weld line by determining the level of molecular entanglement after different temperature and time treatments. As expected, the chain segments have an increased ability to flow and knit when there is sufficient temperature and time.

The purpose of this study was to investigate weld line behavior in injection-molded tensile specimens of nylon 6 nanocomposite containing organically modified montmorillonite (organoclay), which was produced by the melt blending technique. Nylon 6 nanocomposites consist of extremely thin and high surface area, nanometer-scale dispersions of layered silicates. The low loading weight of organoclay is expected to achieve high stiffness and strength due to their potentially high aspect ratio (e.g., 50-500) and the unique intercalation/exfoliation characteristics of silicate platelets in nylon 6 nanocomposite. Nylon 6 nanocomposites containing only 4 wt% of well-exfoliated silicate layers can yield mechanical properties comparable to a 20 wt% mica-filled nylon 6 composite [2].

Neat nylon 6 is used as the control material to highlight the effect of organoclay in the polymer matrix. In addition, the work will identify the principal process factors important for nylon 6 nanocomposites when a weld line is present. It has also been reported that the decrease in mechanical properties for weld lines containing conventional fillers is due to the fillers aligning perpendicular to the flow direction at the weld line [6]. This work seeks to determine whether the flow orientation of organoclay and matrix morphology similarly influences mechanical properties in the weld region. Single-end-gated samples were also made to determine the effect of the presence of a weld line in the tensile specimens. The Taguchi experimental design approach is used in this study to help identify the most important influences on weld line strength because it provides significant information from relatively few experimental runs [11]. A modified [L.sub.16] orthogonal array was designed, according to the experimental-parameter-design method previously developed [12, 13]. The signal-to-noise (S/N) ratio, a sensitivity test proposed by Liu et al. [10] and Turng and Kharbas [14], was also used to identify the principal process factors and the optimum molding conditions, while an analysis of variance (ANOVA) was used to evaluate the effect of the various process factors on weld strength in an orderly sequence [12]. Failure surfaces from the tensile tests were investigated to understand the influence of the resulting polymer morphology on tensile properties.

EXPERIMENTAL

Material

Neat nylon 6 and nylon 6 nanocomposite were obtained from RTP (Winona, MN). Both materials have an average molecular weight, Mn, of 18,000. The nylon 6 nanocomposite containing organically modified montmorillonite (organoclay) was produced by the supplier by a melt blending technique.

[FIGURE 1 OMITTED]

Thermogravimetric Analysis 

The precise weight of inorganic clay content in the commercial nylon 6 nanocomposites was measured using a thermogravimetric analysis (TGA), Setaram TG92 at a heating rate of 30K/min from 50-1000[degrees]C in an argon gas flow.

Injection Molding

Injection-molded specimens of the materials were injected with a Battenfeld (South Elgin, IL) model Plus 350-75 35ton injection molding machine (screw diameter = 30 mm, L/D ratio = 13.5:1, maximum injection pressure = 109.4 MPa) along with a Yann Bang (Taiwan) model YBM-I-P mold temperature controller.

Tensile Specimens

Dumbbell-shape tensile specimens of double-end-gated cavities were used in this study (see Fig. 1). In single-end-gated specimens, there are no weld lines present and they are referred to as single-end-gated samples, while in double-end-gated specimens a weld line is formed at the middle of the tensile bar and it is referred to as a double-end-gated sample.

Experimental Design

A modified [L.sub.16] orthogonal array with three levels (except back pressure) was used to schedule the injection molding experiments [13] (Table 1). Eight different process factors were evaluated, namely, injection pressure, holding pressure, holding time, back pressure, screw rotational speed, melt temperature, mold temperature, and cooling time. Several injection trials were conducted to obtain upper and lower parameter setting conditions prior to the main experiments to ensure reproducibility in sample quality. These conditions then served as a guide to preset various processing parameters for further experiments (Table 2).

Tensile Testing

Tensile testing was conducted after 1 week of storage of the molded tensile specimens to allow equilibration under atmospheric conditions. Tensile properties were measured according to ASTM D638-87b using an Instron (Norwood, MA) 4505 with a 5kN load cell and a crosshead speed of 50 mm/min. The ultimate tensile strength of single-end-gated samples is referred to as bulk strength, while the failure strength of double-end-gated samples with weld lines will be called weld strength.

Optical Microscopy and Environmental Scanning Electron Microscopy 

The weld line of double-end-gated tensile specimens was observed by reflection optical microscopy using an Olympus (Lake Success, NY) model PMG 3. Fracture surfaces of tensile failure specimens were examined on an environmental scanning electron microscope (ESEM), FEI model Quanta 200 at low vacuum, 20 kV under a back-scattering condition.

RESULTS AND DISCUSSION

Thermogravimetric Analysis (TGA)

The thermogravimetric analysis results show that the inorganic clay content of the nylon 6 nanocomposite is 3.5 [+ or -] 0.5 wt%. Since this material is obtained commercially, the precise type and amount of organoclay surface treatment (layered silicate and organo-ion) is unknown, but it often amounts to ~30 wt% organo-ion in most commercial samples. It is likely that between 4 and 5 wt% of organoclay was blended with the nylon to form the nanocomposite.

Comparison of Nylon 6 and Nylon 6 Nanocomposites

Figures 2 and 3 show the tensile stress-strain curves for neat nylon 6 and the nylon 6 nanocomposite. The tensile specimens were injection-molded under standard processing conditions. Table 3 summarizes the ultimate tensile strength and strain-to-failure for the neat nylon 6 and the nanocomposite, both with and without weld lines. The clear difference is that the weld line in the nanocomposite is weaker than the single-end-gated samples, whereas in the neat nylon the strength is not significantly affected.

[FIGURE 2 OMITTED]

Figure 2 demonstrates that the neat nylon 6 is tough and ductile, even with a weld line present, and that the presence of a weld line causes little variation in tensile properties. The double-end-gated sample does not fail at the weld region but rather randomly along the gauge length of the tensile bar specimens. Clearly, the interdiffusion of the polymer molecules at the weld region is extensive for the neat nylon 6, and the weld line defect is visually observed with difficulty on the tensile specimen. Both the single- and double-end-gated samples of neat nylon cold-draw under a uniaxial tensile force, and elongate to a significant strain (greater than 132%) before failure.

Figure 3 shows that the nylon 6 nanocomposite exhibits rigid and brittle behavior, even without a weld line present. The brittleness of the nylon 6 nanocomposite demonstrates that even a low loading of organoclay in the nylon, i.e., 4-5 wt%, is able to influence the mechanical properties significantly. Clearly, the inherent size of the organoclay plays a dominant role in the property performance. The double-end-gated samples with weld lines have even lower tensile strength and strain-to-failure compared to the single-end-gated samples. A surface notch is observed at the weld line of the double-end-gated samples, due to air being entrapped at the weld line interface between the two opposing flow fronts. When the flow fronts impinge at the weld line, the entrapped hot air will also generate local heating and could cause some polymer degradation at the weld line and lead to the poor weld strength. The notch present at the weld line of the nylon 6 nanocomposite (Fig. 4a), which was estimated to be about 60 [micro]m in depth, could also act as stress concentrator and initiate crack propagation. A fan-shaped appearance of the notch can also be seen at the weld line edges. This differs from the neat nylon 6, where the notch was not observable in the double-end-gated samples (Fig. 4b), although the point of joining of the flow fronts was still observable as a line in the photograph. The greater notch depth in the nylon 6 nanocomposite may also illustrate the fact that the nylon nanocomposite exhibits a higher viscosity and melt strength, as both the neat nylon 6 and nylon 6 nanocomposite materials were processed under similar conditions in this work.

[FIGURE 3 OMITTED]

In addition to surface notch formation, the low nanocomposite weld strength could also be caused by organoclay agglomeration, where the organoclay is oriented and located at the weld line and acts as a stress concentrator, causing weakening (see Fig. 5). This is clearly different from the neat nylon 6, which has no nanoparticles in the polymer matrix. Examination of the tensile fracture surfaces of nanocomposite samples shows that the double-end-gated tensile specimens (which invariably failed in the weld region) leave a smoother fracture surface (Fig. 6a) than the single-end-gated samples (Fig. 6b). Organoclay orientation in a high shear flow field during injection molding can induce a high degree of nylon chain orientation due to the fountain flow at the flow fronts having an extensional component normal to the overall melt flow direction. When both the flow fronts impinge at the weld line, the organoclay and polymer chains may orient perpendicular to the overall flow direction, making it a potential source of weakness. The poor weld strength of micron-sized, fiber-filled polymers have been reported to be due to perpendicular alignment of the filler at weld line plane and to their nonuniform distribution [7]. In the weld line region, the nanoparticles could inhibit molecular diffusion across the weld interface and/or promote planar molecular orientation parallel to the weld line plane, and hence lead to incomplete molecular entanglement [15].

[FIGURE 4 OMITTED]

The planar orientation of nanoparticles may also hinder molecular chain motion on solidification and inhibit optimal crystalline development and orientation in the weld line region. Indeed, it is also well known that incorporation of organoclay in nylon 6 can enhance the [gamma]-phase crystal structure compared to the [alpha]-phase crystalline phase of neat nylon 6, which may itself affect the physical and mechanical properties [16]. It was found that the organoclay enhances the [gamma]-phase crystal structure from ~2% (neat nylon 6) to ~30% (nylon 6 nanocomposite), while maintaining [alpha]-phase crystal structure at between 25% and 50% in nylon 6 nanocomposite [17].

Further possible explanations for the poor weld strength of the nylon 6 nanocomposite could be that higher moisture absorption is likely at the weld line area of the samples following solidification and before testing. Nylon is a hygroscopic material due to the polar amide groups. The planar orientation of the nanoparticles provides additional surface area and possibly voids in the weld line region, thus encouraging the weld region to absorb moisture at a rate faster than for the rest of the molded part, a similar effect having been noted for glass fiber-filled nylon [18].

Because of this particular sensitivity of the tensile strength to weld lines, the following investigation was undertaken to determine the processing conditions under which the nanocomposite material could be processed to minimize the weld line effect.

Principal Process Factors Determination

The significance of the effect of the processing factors can be evaluated using an S/N ratio and (ANOVA). Both the S/N ratio and ANOVA use data from measurements of the ultimate tensile strength of the injection molding.

S/N ratio. In this section the S/N ratio is used to determine the degree of response due to variations of each parameter under consideration. It is used to evaluate the significance of process factors in determining the ultimate tensile strength, and to determine an optimum molding condition that achieves the highest tensile strength. The S/N ratio formulation is:

[S/N] ratio = -10 [log.sub.10]([1/n] [n.summation over (i=1)] [1/[y.sub.i.sup.2]]),

where n is the number of specimens tested in each run and [y.sub.i] is the measured ultimate tensile strength.

Figure 7 shows the effect of the various process factors on the ultimate tensile strength of single- and double-end-gated samples of the nylon 6 nanocomposite. The results show that the injection pressure ([DELTA]S/N = 1.6) and mold temperature ([DELTA]S/N = 1.3) are the principal process factors for the single-end-gated samples, while the mold temperature ([DELTA]S/N = 1.3) and melt temperature ([DELTA]S/N = 1.1) are the most significant process factors for the double-end-gated samples.

As expected, the mold temperature has a profound effect on the bulk and weld strengths. Malguarnera and Manisali [19] investigated the effect of melt temperature, mold temperature, injection speed, and injection pressure on the tensile properties of polypropylene. They also found that increasing the mold temperature increased the ultimate tensile strength of both the single- and double-end-gated samples in polypropylene. The effect of mold temperature may be explained by considering that when the polymer melt begins to enter the cavity, it starts to lose heat as it contacts the cool mold, and relatively cool flow fronts impinge at the weld lines, reducing molecular interdiffusion across the polymer-polymer interface. A high mold temperature prolongs the molten state of the polymer melt during the injection stage, enabling the melt to more easily fill the rest of the cavity, generally increasing the bulk strength, as well as the weld strength. High weld strength is thus enhanced in the double-end-gated samples if the melt temperature remains sufficiently high for a long enough period of time to allow the molecular chain segments of the opposing flows to become entangled, particularly when considering the effect of the possible planar orientation of nanoparticles at the weld line interface, which may inhibit such cross-diffusion [7].

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

The injection pressure plays a significant role in determining the bulk strength of the single-end-gated sample, but not so in the double-end-gated samples. The mold temperature has one of the highest S/N ratios for both single- and double-end-gated samples, indicating the importance of this temperature in achieving microstructures with high strength. This is underlined by the finding that for the double-end-gated samples the melt temperature is the second most important process variable, while for the single-end-gated samples melt temperature is important to a lesser degree than injection pressure. The apparent increased importance of injection pressure for the single-gated samples is probably due to the fact that the strength of the nonweld line-containing samples is essentially dependent on the microstructure obtained along the entire gauge length, whereas the weld line-containing samples are critically dependent on the microstructure at the weld line itself, with diffusion across the weld line being of paramount importance. The injection pressure affects injection velocity, and thus shear rates and crystalline orientation, and so the strength. Such a phenomenon is less important in the double-end-gated samples where weld line behavior dominates strength.

[FIGURE 7 OMITTED]

Table 4 shows the optimum molding condition for single- and double-end-gated samples. This optimum molding condition is based on the maximum tensile strength achieved as a function of parameter setting. The results were derived from the various levels of process conditions listed in Tables 1 and 2. It can be seen that the single-end-gated molding condition is slightly different than for the double-end-gated molding configuration. The optimum molding condition for the double-end-gated configuration is achieved at a higher melt temperature, but at a lower injection pressure than for the single-end-gated configuration. This reinforces the observation that increased melt temperature enhances weld line strength due to molecular diffusion and entanglement at high temperatures [20, 21]. The reason for slightly lower injection pressures in the case of weld lines may be due to the levels of molded-in-stress building up at the melt fronts. The double-end-gated sample has a shorter maximum flow length, compared to the single-end-gated sample with the shorter flow length requiring lower injection pressures. High injection pressure at shorter flow length may lead to higher molded-in-stresses, particularly at the weld line where melt fronts impinge on each other. Thus, the double-end-gated sample prefers a lower injection pressure to achieve higher weld strengths.

ANOVA. Table 5 shows the ANOVA results calculated from the ultimate tensile strength data. The variance ratio (denoted F) is the ratio of a given factors mean to its mean square error [12]. The F value shows the importance of the process factor in influencing the process response [13]. A (relatively) large F value means the effect of the particular process factor is large, compared to the error variance. The significance of each process factor on the weld strength of injection-molded parts can therefore be judged by the F value. The calculated results show that the various process factors, in decreasing order of importance in affecting weld strength, are: mold temperature (18.1), melt temperature (12.8), back pressure (3.7), holding time (3.5), screw rotational speed (2.5), injection pressure (2.0), holding pressure (1.6), and cooling time (0.4). The mold temperature has the highest F value, meaning that this processing factor has the highest influence on the average weld strength and causes the greatest property change, while the cooling time has the lowest F value, and plays only a small role in enhancing weld strength.

CONCLUSION

This work is the first report of weld line behavior in nanocomposites. Neat nylon 6 shows little variation in tensile properties in the presence or absence of weld lines. Conversely, nylon 6 nanocomposites formed by melt-blending show higher tensile strengths, but much lower strain-to-failure, when compared with neat nylon 6. The strain-to-failure of neat nylon 6 retains a much higher value than the nylon 6 nanocomposite, even for double-end-gated samples. Clearly, in neat nylon 6 the molecules knit strongly across the weld line, while the inclusion of a low concentration of organoclay has a significant effect on the tensile properties. The organoclay increases the tensile strength, but adversely reduces the strain-to-failure. This may be due to the presence of notches at the weld line region, interference with the molecular diffusion across the boundary, or as a result of variation in orientation or type of crystallinity due to clay platelets.

An experimental design based on the Taguchi approach was used to identify the effect of the various process factors on the molding of nylon 6 nanocomposite. The modified [L.sub.16] orthogonal array with three levels is a simplified experimental design, and by running 16 experiments it allowed identification of the principal process factors and the optimum molding conditions for either single or double-end-gated samples. The ANOVA calculated results show the significance of the process factors on the double-end-gated samples, and so the effects on weld strength in decreasing order of importance are: mold temperature, melt temperature, back pressure, holding time, screw rotational speed, injection pressure, holding pressure, and cooling time. Current work is investigating the nature of crystalline and organoclay morphology at the weld line by a range of characterization techniques.

REFERENCES

1. P.C. LeBaron, Z. Wang, and T.J. Pinnavaia. Appl. Clay Sci., 15, 11 (1999).

2. A. Usuki, Y. Kojima, M. Kawasumi, A. Okada, Y. Fukushima, T. Kurauchi, and O. Kamigaito, J. Mater. Res., 8, 1179 (1993).

3. K. Yano, A. Usuki, A. Okada, T. Kurauchi, and O. Kamigaito, J. Polym. Sci. Part A: Polym. Chem., 31, 2493 (1993).

4. R.A. Vaia, B.B. Sauer, O.K. Tse, and E.P. Giannelis, J. Polym. Sci., Part B: Polym. Phys., 35, 59 (1997).

5. J.W. Cho and D.R. Paul, Polymer, 42, 1083 (2000).

6. S.K. De and J.R. White, Short Fiber-Polymer Composites, Woodhead, Cambridge, UK (1996).

7. S. Fellahi, A. Meddad, B. Fisa, and B.D. Favis, Adv. Polym. Tech., 14, 169 (1995).

8. T.C. Chang and E. Faison III, J. Inject. Mold. Tech., 3, 61 (1999).

9. K. Tomari, S. Tonogai, T. Harada, H. Hamada, K. Lee. T. Morii, and Z. Maekawa, Polym. Eng. Sci., 30, 931 (1990).

10. S.J. Liu, J.Y. Wu, J.H. Chang, and S.W. Hung, Polym, Eng. Sci., 40, 1256 (2000).

11. J.C. Viana and A.M. Cunha, J. Inject. Mold. Tech., 6, 259 (2002).

12. D.C. Montgomery, Design and Analysis of Experiments, John Wiley & Sons, New York (2001).

13. G.S. Peace, Taguchi Methods: A Hands-On Approach, Addison-Wesley, New York (1993).

14. L.S. Turng and H. Kharbas, Polym. Eng. Sci., 43, 157 (2003).

15. S.G. Kim and N.P. Suh, Polym. Eng. Sci., 26, 1200 (1986).

16. T.D. Fornes and D.R. Paul, Polymer, 44, 3945 (2003).

17. J. Tung, R.K. Gupta, G.P. Simon, G.H. Edward, and S.N. Bhattacharya, Polymer, in press.

18. R.M. Belz and E. Lingle, ANTEC Annual Tech. Conf., 3511 (2002).

19. S.C. Malguarnera and A. Manisali, Polym. Eng. Sci., 21, 586 (1981).

20. M. Christie, Plast. Eng., 42, 41 (1986).

21. A. Gennaro. Plast. Rubber Proc. Appl., 9, 241 (1988).

J. Tung, G.P. Simon, G.H. Edward

Cooperative Research Center for Polymers, Department of Materials Engineering, Monash University, Clayton, Victoria 3800, Australia

Correspondence to: J. Tung, e-mail: [e-mail]

Contract grant sponsor: Cooperative Research Centre for Polymers (CRC-P).

TABLE 1. Modified [L.sub.16] orthogonal array an experimental design.

                 A                  B                C        D Back
Run no.  Injection pressure  Holding pressure  Holding time  pressure

 1       A1                  B2                C2            D1
 2       A2                  B2                C2            D1
 3       A3                  B2                C2            D1
 4       A2                  B1                C2            D1
 5       A2                  B3                C2            D1
 6       A2                  B2                C1            D1
 7       A2                  B2                C3            D1
 8       A2                  B2                C2            D2
 9       A2                  B2                C2            D1
10       A2                  B2                C2            D1
11       A2                  B2                C2            D1
12       A2                  B2                C2            D1
13       A2                  B2                C2            D1
14       A2                  B2                C2            D1
15       A2                  B2                C2            D1
16       A2                  B2                C2            D1

              E           F           G              H
Run no.  Screw speed  Melt temp.  Mold temp.  Cooling time

 1       E2           F1          G3          H2
 2       E2           F1          G3          H2
 3       E2           F1          G3          H2
 4       E2           F1          G3          H2
 5       E2           F1          G3          H2
 6       E2           F1          G3          H2
 7       E2           F1          G3          H2
 8       E2           F1          G3          H2
 9       E1           F1          G3          H2
10       E3           F1          G3          H2
11       E2           F2          G3          H2
12       E2           F3          G3          H2
13       E2           F1          G1          H2
14       E2           F1          G2          H2
15       E2           F1          G3          H1
16       E2           F1          G3          H3

Key to symbols is in Table 2.

TABLE 2. Parameter settings for the experimental design.

                                                Parameter setting
Process factor          Unit        Symbol  Level 1  Level 2  Level 3

Injection pressure      MPa         A        19       25       31
Holding pressure        MPa         B        19       22       25
Holding time            s           C        15       20       25
Back pressure           MPa         D         9       16      --
Screw rotational speed  rpm         E        30       55       70
Melt temperature        [degrees]C  F       268      280      290
Mold temperature        [degrees]C  G        40       60       80
Cooling time            s           H        20       30       40

TABLE 3. Tensile properties of nylon 6 and its nanocomposite, with and
without weld lines.

                         Ultimate tensile
Material         Sample  strength               Strain-to-failure

Nylon 6          Single  51.4 MPa [+ or -] 5.0  135.7 [+ or -] 10.0%
                 Double  50.9 MPa [+ or -] 7.0  132.2 [+ or -] 15.0%
Nylon 6          Single  73.6 MPa [+ or -] 5.0    4.1 [+ or -] 0.3%
  Nanocomposite  Double  52.7 MPa [+ or -] 5.0    3.2 [+ or -] 0.3%

TABLE 4. Optimum molding condition achieved for single and double
end-gated samples.

                                             Optimum condition
Process factor          Unit        Symbol   Single  Double

Injection pressure      MPa         A         31      19
Holding pressure        MPa         B         25      25
Holding time            s           C         25      25
Back pressure           MPa         D         16      16
Screw rotational speed  rpm         E         70      70
Melt temperature        [degrees]C  F        280     290
Mold temperature        [degrees]C  G         60      60
Cooling time            s           H         20      20

TABLE 5. Weld strength and ANOVA calculated results for nylon 6
nanocomposite produced as double end-gated samples.

                                                        ANOVA
Process                     Weld strength (Mpa)    Sum of  Degree of
factor     Level  Run no.  #1    #2    #3    #4    square  freedom

Injection  1       1       60.2  62.1  56.3  59.5   54.2   2
           2       2       53.2  55.1  52.2  57.0
           3       3       60.9  48.5  59.5  56.3
Holding    1       4       63.2  62.0  49.3  61.1   56.4   2
           2       2       53.2  55.1  52.2  57.0
           3       5       60.4  59.0  61.8  55.3
Holding    1       6       57.4  53.4  58.6  60.0   58.2   2
           2       2       53.2  55.1  52.2  57.0
           3       7       55.2  58.8  63.3  61.9
Back       1       2       53.2  55.1  52.2  57.0   63.1   1
           2       8       52.0  61.9  62.5  63.6
Screw      1       9       50.7  61.7  65.0  65.4   89.4   2
           2       2       53.2  55.1  52.2  57.0
           3      10       60.4  58.7  60.8  58.2
Melt       1       2       53.2  55.1  52.2  57.0  110.9   2
           2      11       58.9  61.2  57.5  58.8
           3      12       62.0  65.0  59.0  61.1
Mold       1      13       59.6  59.4  56.7  57.0  157.7   2
           2      14       64.8  60.0  65.5  62.7
           3       2       53.2  55.1  52.2  57.0
Cooling    1      15       15.1  55.1  57.5  61.1    8.4   2
           2       2       53.2  55.1  52.2  57.0
           3      16       52.8  60.1  54.8  56.9

                                  ANOVA
Process           Mean sum of      Mean sum of
factor     Level  square (factor)  square (error)  F value

Injection  1      27.1             13.7             2.0
           2
           3
Holding    1      28.2             18.0             1.6
           2
           3
Holding    1      29.1              8.4             3.5
           2
           3
Back       1      63.1             16.9             3.7
           2
Screw      1      44.7             17.9             2.5
           2
           3
Melt       1      55.4              4.3            12.8
           2
           3
Mold       1      78.9              4.4            18.1
           2
           3
Cooling    1       4.2             10.7             0.4
           2
           3