# Rheological test to characterize injection molding.

The viscoelastic properties of elastomers will determine not only their final product performance but their processing properties as well. Several rules-of-thumb have been established in the industry for investigating their performance and processibility of elastomers. These rules however, have not faired well in predicting processing by injection molding. The time scale for injection molding (as opposed to compression molding), higher shear rates and complicated flow properties suggest a more sensitive dependence on rheology properties. To this end, several elastomers are evaluated by DMS during cure and compared to the processing evaluation offered by Peacock et al's complimentary paper (ref. 1).

Experimental

Materials used for this study are identical to the resins studied by Peacock, Bussem and Hertz (ref. 1). Six 1owmooney (28-40) 34% ACN nitrile elastomers, containing curatives only, were compounded in the gum state. All resins were cured to completion at 350[degree]F (177[degree]C). The cure package contained 5.0 pans zinc oxide, 1.0 part stearic acid, 1.5 pans sulfur and 1.5 parts MBTS.

Dynamic mechanical properties were determined using a modified Rheometrics Dynamic Spectrometer (RD-II). In order to obtain the most information during the cure process, a multi-wave technique was employed. For this technique the imposed strain deformation is the sum of several sinusoidal responses are represented by the time dependent strain presented below.

[MATHEMATICAL EXPRESSION OMITTED]

In this expression i represents the imposed strain levels, coo is the primary frequency and ai is the multiplier for the primary frequency. Strains were selected so as to be in the linear viscoelastic region at all times ([Epsilon][sub.i] set to 2.5%) while the fundamental frequency and multipliers were selected so as to give eight frequencies; 1, 2, 7, 10, 20, 40, 70 and 100 radians/second. A typical strain history is presented in figures 1 and 2. Through cross correlation Fouirier transformation, the viscoelastic properties at each frequency are then determined in a fraction of the time it would take to measure each frequency separately.

The elastomers were provided in the form of a thick specimen, typically 1/4 to 1/2 inches thick. These specimens are cold pressed to a thickness of 2 mm. A 25 mm round specimen is then cut from the pressed sample. The parallel plate fixture is then preheated to 177[degree]C at which point the zero gap reference is set. The specimen is then placed between the plates and compressed to an exact gap of 2 mm. After 2-1/2 minutes of heating, testing is initiated. A multi-wave test is conducted every 20 seconds for a total elapsed time of 15 minutes.

Results and discussion

The objective of this study is to determine the theological basis for evaluating the process performance of the elastomer, keeping in mind the fact that other properties as described by Peacock et al (ref. 1) are important as well. The shear storage modulus, G', and the shear loss modulus, G", are presented in figures 3 and 4 respectively.

Elastomers D and F do not cure in a fashion represented by the remaining resins. The shortage of data for Elastomer F is due to slippage in the plates. The data past 350 seconds is therefore eliminated. Specimens A, B, C and E demonstrate typical viscosity and moduli cure profiles.

Peacock (tel. 1) suggests that resins A and D demonstrate strong potential for mold fouling in injection molding. As seen in figures 3 and 4 above, these two resins are different in their processing behavior. The qualitative differences in the G' cure curve and Peacock et al.'s S' cure curve measured on the Monsanto MDR 2000E are consistent. However, there are differences in the ranking of final modulus values when comparing these two techniques. The qualitative differences are unable to suggest a correlation between potential mold fouling in injection molding SinCe it iS best correlated with resin chemistry. The quantitative dynamic mechanical spectroscopy results suggest that the cure rate for these materials is high and may lead to complications during injection molding depending on the residence time associated with the injection molder. This would suggest that the mixing time scale during processing is longer than the reaction time scale leading to phase separation of process residuals and compounding ingredients. This argument would predict that similar problems will occur with elastomers E and B as well. Work is in progress to evaluate these resins. Elastomer D has a much longer time before it reaches its maximum cure rate and may not process well since it has the potential to undercure.

The derivative of the storage modulus with time is presented in figure 5. As a first approximation, the storage modulus can be treated as the shear modulus for the elastomer network; that is, G' can be used to approximate G. G is typically defined as NkT where N is the number of crosslinks per volume, k is the Boltzman constant and T is the absolute temperature. For the time derivative of the storage modulus we have the following expression.

[MATHEMATICAL EXPRESSION OMITTED]

This suggests that the maximum change in the number of crosslinks produced during cure occurs at the maximum associated with the rate-of-change in the storage modulus. The value at this maximum can be used to determine the number of crosslinks produced per second.

The two elastomer systems with the highest rate of crosslinking are A and C. This data suggests that the structure is established over a limited time (as determined from the width of the derivative peak). This observation is consistent with the frequency sweep data. As the structure is established, the frequency dependence of the elastomers should be independent of time. An example of this behavior is presented in figures 6 and 7 where the frequency dependence is monitored during the cure process.

Both the storage and loss moduli measured during cure are sensitive measures of the change in structure. In particular, the loss modulus as a function of frequency is a good indicator of the structure's state. Similar data for the remaining resins suggests that elastomer C reaches full cure shortly after its peak crosslinking rate. Elastomer C is observed to have the highest conversion rate and processes with the highest final modulus (3.5 x [10.sup.6] dynes/[cm.sup.2]). This correlation with conversion rate is expected since it is critical for an injection molding process that maximum conversion be obtained during a relatively small window of time. The time to reach maximum conversion rates as compared to the residence time of the injection molder is most likely another critical parameter. Unfortunately, this information was not available for this study or for the complimentary study by Peacock et al (ref. 1). Further work is needed in this area.

Summary

The DMS data measured during the cure of six elastomers is used to understand the relationship between processibility and viscoelastic properties. The systems used in ref. 1 have been evaluated with resin C showing favorable properties. Our results suggest that there are two critical rheology parameters that correlate with the performance of these resins. The magnitude of the storage modulus time derivative, d(G')/dt, represents the rate at which crosslinks are formed. A fast rate over a moderate time span seems to be critical in the performance of the resin during injection molding. The time to reach this maximum reaction rate relative to the residence time of the injection molder is critical as well. Unfortunately, an exact relationship could not be determined since limited information was available about the injection molder.

Frequency sweeps produced from multi-wave analysis are a critical source of information needed to determine the "state" of cure. This technique provides a convenient method for monitoring the cure process. Further studies of the finally cured elastomers will be reviewed but are not available at this writing.

[TABULAR DATA OMITTED]

References

1. C. Peacock, H. Bussem, D.L. Hertz, Jr., presented at a meeting of the Rubber Division, American Chemical Society, Louisville, Kentucky, May 19-22, 1992.

Experimental

Materials used for this study are identical to the resins studied by Peacock, Bussem and Hertz (ref. 1). Six 1owmooney (28-40) 34% ACN nitrile elastomers, containing curatives only, were compounded in the gum state. All resins were cured to completion at 350[degree]F (177[degree]C). The cure package contained 5.0 pans zinc oxide, 1.0 part stearic acid, 1.5 pans sulfur and 1.5 parts MBTS.

Dynamic mechanical properties were determined using a modified Rheometrics Dynamic Spectrometer (RD-II). In order to obtain the most information during the cure process, a multi-wave technique was employed. For this technique the imposed strain deformation is the sum of several sinusoidal responses are represented by the time dependent strain presented below.

[MATHEMATICAL EXPRESSION OMITTED]

In this expression i represents the imposed strain levels, coo is the primary frequency and ai is the multiplier for the primary frequency. Strains were selected so as to be in the linear viscoelastic region at all times ([Epsilon][sub.i] set to 2.5%) while the fundamental frequency and multipliers were selected so as to give eight frequencies; 1, 2, 7, 10, 20, 40, 70 and 100 radians/second. A typical strain history is presented in figures 1 and 2. Through cross correlation Fouirier transformation, the viscoelastic properties at each frequency are then determined in a fraction of the time it would take to measure each frequency separately.

The elastomers were provided in the form of a thick specimen, typically 1/4 to 1/2 inches thick. These specimens are cold pressed to a thickness of 2 mm. A 25 mm round specimen is then cut from the pressed sample. The parallel plate fixture is then preheated to 177[degree]C at which point the zero gap reference is set. The specimen is then placed between the plates and compressed to an exact gap of 2 mm. After 2-1/2 minutes of heating, testing is initiated. A multi-wave test is conducted every 20 seconds for a total elapsed time of 15 minutes.

Results and discussion

The objective of this study is to determine the theological basis for evaluating the process performance of the elastomer, keeping in mind the fact that other properties as described by Peacock et al (ref. 1) are important as well. The shear storage modulus, G', and the shear loss modulus, G", are presented in figures 3 and 4 respectively.

Elastomers D and F do not cure in a fashion represented by the remaining resins. The shortage of data for Elastomer F is due to slippage in the plates. The data past 350 seconds is therefore eliminated. Specimens A, B, C and E demonstrate typical viscosity and moduli cure profiles.

Peacock (tel. 1) suggests that resins A and D demonstrate strong potential for mold fouling in injection molding. As seen in figures 3 and 4 above, these two resins are different in their processing behavior. The qualitative differences in the G' cure curve and Peacock et al.'s S' cure curve measured on the Monsanto MDR 2000E are consistent. However, there are differences in the ranking of final modulus values when comparing these two techniques. The qualitative differences are unable to suggest a correlation between potential mold fouling in injection molding SinCe it iS best correlated with resin chemistry. The quantitative dynamic mechanical spectroscopy results suggest that the cure rate for these materials is high and may lead to complications during injection molding depending on the residence time associated with the injection molder. This would suggest that the mixing time scale during processing is longer than the reaction time scale leading to phase separation of process residuals and compounding ingredients. This argument would predict that similar problems will occur with elastomers E and B as well. Work is in progress to evaluate these resins. Elastomer D has a much longer time before it reaches its maximum cure rate and may not process well since it has the potential to undercure.

The derivative of the storage modulus with time is presented in figure 5. As a first approximation, the storage modulus can be treated as the shear modulus for the elastomer network; that is, G' can be used to approximate G. G is typically defined as NkT where N is the number of crosslinks per volume, k is the Boltzman constant and T is the absolute temperature. For the time derivative of the storage modulus we have the following expression.

[MATHEMATICAL EXPRESSION OMITTED]

This suggests that the maximum change in the number of crosslinks produced during cure occurs at the maximum associated with the rate-of-change in the storage modulus. The value at this maximum can be used to determine the number of crosslinks produced per second.

The two elastomer systems with the highest rate of crosslinking are A and C. This data suggests that the structure is established over a limited time (as determined from the width of the derivative peak). This observation is consistent with the frequency sweep data. As the structure is established, the frequency dependence of the elastomers should be independent of time. An example of this behavior is presented in figures 6 and 7 where the frequency dependence is monitored during the cure process.

Both the storage and loss moduli measured during cure are sensitive measures of the change in structure. In particular, the loss modulus as a function of frequency is a good indicator of the structure's state. Similar data for the remaining resins suggests that elastomer C reaches full cure shortly after its peak crosslinking rate. Elastomer C is observed to have the highest conversion rate and processes with the highest final modulus (3.5 x [10.sup.6] dynes/[cm.sup.2]). This correlation with conversion rate is expected since it is critical for an injection molding process that maximum conversion be obtained during a relatively small window of time. The time to reach maximum conversion rates as compared to the residence time of the injection molder is most likely another critical parameter. Unfortunately, this information was not available for this study or for the complimentary study by Peacock et al (ref. 1). Further work is needed in this area.

Summary

The DMS data measured during the cure of six elastomers is used to understand the relationship between processibility and viscoelastic properties. The systems used in ref. 1 have been evaluated with resin C showing favorable properties. Our results suggest that there are two critical rheology parameters that correlate with the performance of these resins. The magnitude of the storage modulus time derivative, d(G')/dt, represents the rate at which crosslinks are formed. A fast rate over a moderate time span seems to be critical in the performance of the resin during injection molding. The time to reach this maximum reaction rate relative to the residence time of the injection molder is critical as well. Unfortunately, an exact relationship could not be determined since limited information was available about the injection molder.

Frequency sweeps produced from multi-wave analysis are a critical source of information needed to determine the "state" of cure. This technique provides a convenient method for monitoring the cure process. Further studies of the finally cured elastomers will be reviewed but are not available at this writing.

[TABULAR DATA OMITTED]

References

1. C. Peacock, H. Bussem, D.L. Hertz, Jr., presented at a meeting of the Rubber Division, American Chemical Society, Louisville, Kentucky, May 19-22, 1992.

Printer friendly Cite/link Email Feedback | |

Title Annotation: | pressure testing elastomers |
---|---|

Author: | Paramasivam, M. |

Publication: | Rubber World |

Date: | Dec 1, 1992 |

Words: | 1334 |

Previous Article: | NR breaker compound containing specific chlorotriazines as adhesion promoters. |

Next Article: | Stomil. |

Topics: |