**4.2 Thermoplastic polyolefin blends (TPO) and dynamic vulcanizates (TPV)**

a. Commercial grades based on EPDM/PP

Most of the commercial available thermoplastic vulcanizates (TPV) are produced from polyolefin blends, in particular EPDM/PP, by the process of dynamic vulcanization. TPV exhibit several advantages over simple thermoplastic polyolefin blends (TPO). Due to selective crosslinking of the EPDM rubber phase, almost all material properties are improved. Especially compression set, creep, stress relaxation and swell behaviour are highly important for automotive applications, e.g. all kinds of sealing systems. As shown in Figure 8 for different types of commercial TPO and TPV, dynamic vulcanization has strong impact on relaxation behaviour (Reid et al., 2004). Whereas the stress - temperature - curve of a simple TPO blend exhibits a strong decrease of stress with increasing temperature, the decrease of stress of a TPV material of comparable hardness is significantly lower. As it becomes also obvious from Figure 8, the differences between TPO and TPV depend on hardness; the lower the hardness, the bigger the differences, and vice versa. Only one significant peak is observable in the relaxation spectra of the TPV, which is assigned to the melting of the PP matrix.


Table 1. Results of the TSSR tests as represented in Fig. 6.

In case of low hardness TPO a smaller peak appears at about 40 to 60 °C, which might be related to the -relaxation process of the PP phase. At higher temperatures, when the PP ma-

Characterization of Thermoplastic Elastomers

At a fixed EPDM volume fraction of

the content of crosslink agents.

by Means of Temperature Scanning Stress Relaxation Measurements 359

Because crosslink density is one of the most important parameters of elastomers and dynamic vulcanizates, the aim of this study was to investigate the crosslink density of peroxide cured TPV based on EPDM/PP. Until now, there are only few methods available for the determination of crosslink density (Eisele, 1979; Grinberg, 1999; Zhao, 2007). Most of these methods require high effort and are not suitable for daily use in product development or quality control. For this reason, a new test method for the determination of the crosslink density of dynamic vulcanizates was introduced. The suitability of this method was examined at the example of polyolefinic model compounds, which were dynamically vulcanized by means of a peroxide cure system (Vennemann, et al., 2006).

The TPV samples were produced in a two-step mixing process. In a first mixing step preblends containing EPDM and varying amounts of the crosslink agents were prepared at 100 °C and a rotor speed of 40 rpm. After a mixing time of 3 minutes the pre-blends were removed from the mixer and immediately cooled down to room temperature to avoid scorch of the material. The composition of the different pre-blends is summarized in Table 3. In a second mixing step the TPV compounds were produced, by melt mixing the pre-blends with the PP homopolymer. The composition of the TPV compounds is compiled in Table 4.

The morphology of the samples was investigated by transmission electron microscopy (TEM). TEM micrographs of the phase morphology of selected TPV samples are shown in Figure 9. Samples for these micrographs were cut parallel to flow from an injection moulded plaque, stained with ruthenium oxide, and cryo-microtomed into thin sections. In Figure 9 (left), the co-continuous phase morphology of the thermoplastic polyolefin blend (TPO) is visible, where the EPDM is shown as dark and the PP is shown as light areas. Phase inversion already occurs at low curative content (1 phr peroxide/1 phr co-agent) as it becomes obvious from the TEM micrograph obtained from TPV2 as presented in Figure 9 (right). Here, the elastomer appears as dark discrete particles with diameters of less than one mi-

Sample ID E0 E1 E2 E3 E4 E5 E6 EPDM 100 100 100 100 100 100 100 DTBPIB - 0.5 1.0 1.5 2.0 2.5 3.0 TRIM - 0.5 1.0 1.5 2.0 2.5 3.0

Sample ID TP0 TPV1 TPV2 TPV3 TPV4 TPV5 TPV6 Ex\*) 100 100 100 100 100 100 100 PP 30 30 30 30 30 30 30

Table 3. Composition of the pre-blends in parts per hundred rubber (phr)

\*) Ex stands for the corresponding pre-blend, e.g. TPV1 contains pre-blend E1 etc.

Table 4. Composition of the TPV compounds in parts per hundred rubber (phr)

*EPDM* 0.77 the compound compositions differ only in

cron, embedded in the continuous PP phase (lighter colour).

trix is melting, no significant peak can be detected in case of TPO, because the stress is almost zero due to relaxation. The results of the TSSR tests represented in Figure 8 are compiled in Table 1. For similar hardness TSSR - Index and T50 - values of the TPV are significantly higher than those of TPO materials.

Fig. 8. *Left:* Force - temperature curves and relaxation spectra obtained from commercial TPO and TPV of 75 Shore A hardness. *Right:* Force - temperature curves and relaxation spectra obtained from commercial TPO and TPV of 40 Shore D hardness.

b. Model compounds of peroxide cured TPV based on EPDM/PP

Phenolic resin cure systems are mostly used for the production of commercial TPV materials. Although the phenolic resin cure system normally results in products with better compression set and oil swell than equivalent compositions produced with other cure systems, e.g. hydrosilation, peroxide, or sulfur, peroxide cured TPV materials were introduced into the market, recently (Reid et al., 2004). In EPDM/PP based systems serious problems result from degradation of the PP matrix, which may lead to deterioration of mechanical properties of the TPV. Thus, during the dynamic vulcanization by use of peroxides not only crosslinking of the EPDM rubber phase occurs, but more or less of the free radicals produce degradation of the PP matrix (Loan, 1967; Dickens, 1982), which has to be taken into consideration, also. Therefore, the producers have to find out the right balance between crosslinking of the EPDM phase and degradation of the PP phase; both reactions are initiated by free radicals, created by the curatives.


Table 2. Results of the TSSR tests as represented in Fig. 8.

trix is melting, no significant peak can be detected in case of TPO, because the stress is almost zero due to relaxation. The results of the TSSR tests represented in Figure 8 are compiled in Table 1. For similar hardness TSSR - Index and T50 - values of the TPV are

Fig. 8. *Left:* Force - temperature curves and relaxation spectra obtained from commercial TPO and TPV of 75 Shore A hardness. *Right:* Force - temperature curves and relaxation

Phenolic resin cure systems are mostly used for the production of commercial TPV materials. Although the phenolic resin cure system normally results in products with better compression set and oil swell than equivalent compositions produced with other cure systems, e.g. hydrosilation, peroxide, or sulfur, peroxide cured TPV materials were introduced into the market, recently (Reid et al., 2004). In EPDM/PP based systems serious problems result from degradation of the PP matrix, which may lead to deterioration of mechanical properties of the TPV. Thus, during the dynamic vulcanization by use of peroxides not only crosslinking of the EPDM rubber phase occurs, but more or less of the free radicals produce degradation of the PP matrix (Loan, 1967; Dickens, 1982), which has to be taken into consideration, also. Therefore, the producers have to find out the right balance between crosslinking of the EPDM phase and degradation of the PP phase; both reactions are initiated by free

Sample Shore -Hardness σ<sup>o</sup> T10 T50 T90 TSSR -

TPV- 76A 76A 0.9 52 114 151 0.66 TPO - 75A 75A 1.2 37 52 101 0.49

TPV - 46D 46D 4.7 41 83 146 0.54 TPO - 42 D 42D 2.0 38 68 136 0.49

MPa °C °C °C Index

spectra obtained from commercial TPO and TPV of 40 Shore D hardness.

b. Model compounds of peroxide cured TPV based on EPDM/PP

radicals, created by the curatives.

Table 2. Results of the TSSR tests as represented in Fig. 8.

significantly higher than those of TPO materials.

Because crosslink density is one of the most important parameters of elastomers and dynamic vulcanizates, the aim of this study was to investigate the crosslink density of peroxide cured TPV based on EPDM/PP. Until now, there are only few methods available for the determination of crosslink density (Eisele, 1979; Grinberg, 1999; Zhao, 2007). Most of these methods require high effort and are not suitable for daily use in product development or quality control. For this reason, a new test method for the determination of the crosslink density of dynamic vulcanizates was introduced. The suitability of this method was examined at the example of polyolefinic model compounds, which were dynamically vulcanized by means of a peroxide cure system (Vennemann, et al., 2006).

The TPV samples were produced in a two-step mixing process. In a first mixing step preblends containing EPDM and varying amounts of the crosslink agents were prepared at 100 °C and a rotor speed of 40 rpm. After a mixing time of 3 minutes the pre-blends were removed from the mixer and immediately cooled down to room temperature to avoid scorch of the material. The composition of the different pre-blends is summarized in Table 3. In a second mixing step the TPV compounds were produced, by melt mixing the pre-blends with the PP homopolymer. The composition of the TPV compounds is compiled in Table 4. At a fixed EPDM volume fraction of *EPDM* 0.77 the compound compositions differ only in the content of crosslink agents.

The morphology of the samples was investigated by transmission electron microscopy (TEM). TEM micrographs of the phase morphology of selected TPV samples are shown in Figure 9. Samples for these micrographs were cut parallel to flow from an injection moulded plaque, stained with ruthenium oxide, and cryo-microtomed into thin sections. In Figure 9 (left), the co-continuous phase morphology of the thermoplastic polyolefin blend (TPO) is visible, where the EPDM is shown as dark and the PP is shown as light areas. Phase inversion already occurs at low curative content (1 phr peroxide/1 phr co-agent) as it becomes obvious from the TEM micrograph obtained from TPV2 as presented in Figure 9 (right). Here, the elastomer appears as dark discrete particles with diameters of less than one micron, embedded in the continuous PP phase (lighter colour).




\*) Ex stands for the corresponding pre-blend, e.g. TPV1 contains pre-blend E1 etc.

Table 4. Composition of the TPV compounds in parts per hundred rubber (phr)

Characterization of Thermoplastic Elastomers

Sample ID

Shore A

(Vennemann et al., 2006).

CS 22h /125°C

As described before, temperature coefficient values

by Means of Temperature Scanning Stress Relaxation Measurements 361

increases also and is indicating an improvement of "rubber like" – behaviour. In Table 5 are also values of Shore A hardness, compression set, tensile strength and elongation at break summarized. From Figure 10 it can be also observed that rupture of the samples occurs if the peroxide content of the TPV increases. This is accompanied with a decrease of the TSSR T90 values. Unlike this, the force-temperature curve of the TPO sample approached zero, without rupture of the sample. The rupture of the samples can be explained with the degradation of the polypropylene matrix by peroxide, which is also a well known phenomenon. Due to the consumption of peroxide by the PP matrix, crosslink density of the dispersed EPDM particles is reduced. Figure 11 contains normalized force temperature curves obtained from thermoset rubber samples, prepared by static vulcanization of the pre-blends. The behaviour differs strongly from that of the TPV

samples, presented in Figure 10, although the crosslink systems are identical.

Table 5. Mechanical properties and TSSR results of the TPO and TPV samples

tensile strength

TPO 61 87.9% 1.9 102% 52.2 99.1 159.4 0.59 TPV1 69 74.2% 3.4 263% 53.7 106.7 159.3 0.62 TPV2 76 64.5% 6.3 328% 57.2 121.3 160.4 0.66 TPV3 74 62.5% 8.1 381% 61.6 131.4 156.9 0.72 TPV4 75 61.2% 8.1 366% 61.7 130.9 154.8 0.73 TPV5 77 60.5% 7.9 335% 63.3 133.3 149.2 0.76 TPV6 77 59.2% 8.1 309% 64.4 134.0 145.7 0.78

of stress – temperature curves, which were obtained from TSSR measurements at TPV (Figure 10) and thermoset rubber samples (Figure 11). The crosslink densities of all samples were calculated according to Eq. (6) and Eq. (8) and are plotted against the total amount of curatives in Figure 12. It should be noticed, that the compositions of the thermoset rubber samples E1 – E6 are identical with the rubber phase of the corresponding TPV samples. Thus, the crosslink density of TPV1 can be directly compared with sample E1, TPV2 can be compared with E2 and so on. Figure 12 clarifies that with the same amount of curatives a higher crosslink density is achieved in the thermoset rubber than in the rubber phase of the corresponding TPV compound. This result confirms the assumption that the crosslink density in the rubber phase of the TPV is reduced, due to partial consumption of the peroxide by the PP matrix. By comparison of the number of moles of curatives used in the recipe, the crosslink of efficiency of the cure system can be verified, also. The obtained results of crosslink density were also compared with the reciprocal swell ratio of the samples, which is a measure of crosslink density. It was shown that the crosslink density obtained from TSSR measurements, correlates well with the reciprocal swell ratio

elongation @ break

MPa T10 T50 T90 RI

TSSR

°C °C °C

were determined from the initial part

Fig. 9. TEM micrographs obtained from TPO (left) and TPV2 (right).

Fig. 10. Selected normalized stress – temperature curves of TPV samples obtained from TSSR measurements. TPO (solid line), TPV2 (dashed line), TPV4 (dashed dotted line) and TPV6 (dotted line).

In Figure 10 selected normalized force – temperature curves obtained from TSSR tests are presented. The influence of dynamic vulcanization is clearly recognizable from the shape of the curves. Whereas the uncrosslinked TPO sample exhibits the strongest stress decrease, this is significantly reduced in the case of the TPV samples and thus, higher values of T10 and T50 (Table 5) are obtained. This behaviour was expected, because dynamic vulcanization is well known as a process to improve the stress relaxation properties of polyolefin blends. Due to the reduced stress relaxation, the area below the normalized force–temperature curve increases. Consequently, the rubber index RI

Fig. 9. TEM micrographs obtained from TPO (left) and TPV2 (right).

0,8 μm 0,8 μm

Fig. 10. Selected normalized stress – temperature curves of TPV samples obtained from TSSR measurements. TPO (solid line), TPV2 (dashed line), TPV4 (dashed dotted line) and

In Figure 10 selected normalized force – temperature curves obtained from TSSR tests are presented. The influence of dynamic vulcanization is clearly recognizable from the shape of the curves. Whereas the uncrosslinked TPO sample exhibits the strongest stress decrease, this is significantly reduced in the case of the TPV samples and thus, higher values of T10 and T50 (Table 5) are obtained. This behaviour was expected, because dynamic vulcanization is well known as a process to improve the stress relaxation properties of polyolefin blends. Due to the reduced stress relaxation, the area below the normalized force–temperature curve increases. Consequently, the rubber index RI

TPV6 (dotted line).

increases also and is indicating an improvement of "rubber like" – behaviour. In Table 5 are also values of Shore A hardness, compression set, tensile strength and elongation at break summarized. From Figure 10 it can be also observed that rupture of the samples occurs if the peroxide content of the TPV increases. This is accompanied with a decrease of the TSSR T90 values. Unlike this, the force-temperature curve of the TPO sample approached zero, without rupture of the sample. The rupture of the samples can be explained with the degradation of the polypropylene matrix by peroxide, which is also a well known phenomenon. Due to the consumption of peroxide by the PP matrix, crosslink density of the dispersed EPDM particles is reduced. Figure 11 contains normalized force temperature curves obtained from thermoset rubber samples, prepared by static vulcanization of the pre-blends. The behaviour differs strongly from that of the TPV samples, presented in Figure 10, although the crosslink systems are identical.


Table 5. Mechanical properties and TSSR results of the TPO and TPV samples

As described before, temperature coefficient values were determined from the initial part of stress – temperature curves, which were obtained from TSSR measurements at TPV (Figure 10) and thermoset rubber samples (Figure 11). The crosslink densities of all samples were calculated according to Eq. (6) and Eq. (8) and are plotted against the total amount of curatives in Figure 12. It should be noticed, that the compositions of the thermoset rubber samples E1 – E6 are identical with the rubber phase of the corresponding TPV samples. Thus, the crosslink density of TPV1 can be directly compared with sample E1, TPV2 can be compared with E2 and so on. Figure 12 clarifies that with the same amount of curatives a higher crosslink density is achieved in the thermoset rubber than in the rubber phase of the corresponding TPV compound. This result confirms the assumption that the crosslink density in the rubber phase of the TPV is reduced, due to partial consumption of the peroxide by the PP matrix. By comparison of the number of moles of curatives used in the recipe, the crosslink of efficiency of the cure system can be verified, also. The obtained results of crosslink density were also compared with the reciprocal swell ratio of the samples, which is a measure of crosslink density. It was shown that the crosslink density obtained from TSSR measurements, correlates well with the reciprocal swell ratio (Vennemann et al., 2006).

Characterization of Thermoplastic Elastomers

HDPE

0,0

EPDM/PP based materials.

0,2

0,95

1,00

1,05

0,4

0,6

normalized force F/F0

0,8

1,0

1,2

by Means of Temperature Scanning Stress Relaxation Measurements 363

c. Advanced TPV based on EPDM/HDPE designed for hart/soft combinations with

Until now, most of the commercial available thermoplastic vulcanizates are based on the EPDM/PP system. Thus, good adhesion is achieved to components made of polypropylene (PP), due to inherent compatibility of the TPV matrix and commonly used PP materials. In combination with other polymers the bonding strength is more or less poor, particularly in case of more polar engineering thermoplastics. Even with other polyolefins bonding strength to TPV may deteriorate because of incompatibility of both partners. PP and HDPE are generally considered immiscible (Shanksa, et al., 2000). Thus, adhesion at the interface of EPDM/PP based TPV and HDPE or UHMW-PE components is lower than in the case of

0 20 40 60 80 100 120 140 160 180

temperature T / °C

Fig. 13. TSSR force – temperature curves of a model TPV based on EPDM/HDPE ( ) and

Only a few papers have been published about dynamic vulcanization of EPDM/HDPE blends (Gosh et al., 1994) (Machado & van Duin, 2005), until now. A strongly increased compound viscosity as a result of dynamic vulcanization has been reported, particularly at higher EPDM contents. Therefore, one aim of this work was to optimize the blend composition of EPDM/HDPE dynamic vulcanizates with respect to mechanical properties and processibility, in order to produce TPVs with improved rubber elasticity and compatibility to rigid HDPE thermoplastics (Vennemann et al., 2009). In Figure 13 TSSR force - temperature curves of a model TPV based on EPDM/HDPE and of commercial TPVs based on EPDM/PP are shown for comparison. From the initial part of the curve it becomes obvious, that the entropy elastic behaviour of the EPDM/HDPE material is more pronounced than of the

 EPDM-X+PP commercial / 73 Sh A EPDM-X+PP commercial / 80 Sh A

EPDM-X+HDPE / 78 Sh A

miscible polymers and may have a negative effect on the functionality of the part.

20 25 30 35 40

commercial TPVs based on EPDM/PP of 73 ( ) and 80 ( ) Shore A.

Fig. 11. Selected normalized stress – temperature curves of thermoset rubber samples obtained from TSSR measurements. E1 (solid line), E3 (dashed line) and E5 (dashed dotted line).

Fig. 12. Crosslink density of thermoset rubber samples E1 - E6 () and TPV1 - TPV6 samples (■) as obtained from TSSR – measurements.

Fig. 11. Selected normalized stress – temperature curves of thermoset rubber samples obtained from TSSR measurements. E1 (solid line), E3 (dashed line) and E5 (dashed dotted line).

Fig. 12. Crosslink density of thermoset rubber samples E1 - E6 () and TPV1 - TPV6 samples

(■) as obtained from TSSR – measurements.

c. Advanced TPV based on EPDM/HDPE designed for hart/soft combinations with HDPE

Until now, most of the commercial available thermoplastic vulcanizates are based on the EPDM/PP system. Thus, good adhesion is achieved to components made of polypropylene (PP), due to inherent compatibility of the TPV matrix and commonly used PP materials. In combination with other polymers the bonding strength is more or less poor, particularly in case of more polar engineering thermoplastics. Even with other polyolefins bonding strength to TPV may deteriorate because of incompatibility of both partners. PP and HDPE are generally considered immiscible (Shanksa, et al., 2000). Thus, adhesion at the interface of EPDM/PP based TPV and HDPE or UHMW-PE components is lower than in the case of miscible polymers and may have a negative effect on the functionality of the part.

Fig. 13. TSSR force – temperature curves of a model TPV based on EPDM/HDPE ( ) and commercial TPVs based on EPDM/PP of 73 ( ) and 80 ( ) Shore A.

Only a few papers have been published about dynamic vulcanization of EPDM/HDPE blends (Gosh et al., 1994) (Machado & van Duin, 2005), until now. A strongly increased compound viscosity as a result of dynamic vulcanization has been reported, particularly at higher EPDM contents. Therefore, one aim of this work was to optimize the blend composition of EPDM/HDPE dynamic vulcanizates with respect to mechanical properties and processibility, in order to produce TPVs with improved rubber elasticity and compatibility to rigid HDPE thermoplastics (Vennemann et al., 2009). In Figure 13 TSSR force - temperature curves of a model TPV based on EPDM/HDPE and of commercial TPVs based on EPDM/PP are shown for comparison. From the initial part of the curve it becomes obvious, that the entropy elastic behaviour of the EPDM/HDPE material is more pronounced than of the EPDM/PP based materials.

Characterization of Thermoplastic Elastomers

negative.

mated.

gression.

by Means of Temperature Scanning Stress Relaxation Measurements 365

Although TSSR measurements are significantly faster and easier to perform than other methods to characterize the crosslink density of thermoset rubber and TPV, further improvement of the method is desired from industry, in particular by reduction of the test duration, without deterioration of the reliability of the results. High potential for the reduction of test duration is included by the isothermal relaxation period. During the isothermal test period, which lasts normally 2 h, the short time relaxation processes occur and the sample reaches a quasi equilibrium state before the non-isothermal test starts. If the isothermal test period is reduced to shorter values, the entropy effect will be partially compensated by stress relaxation and thus the obtained temperature coefficient is diminished, systematically. This problem can be solved by considering the effect of stress relaxation on the experimentally observable temperature coefficient . Generally, the total value of can be divided into two parts, as described in Eq. (10). It should be mentioned, that the entropic part of is positive, whereas the contribution of stress relaxation is

**4.3 Determination of crosslink density by means of rapid TSSR - tests** 

 

 

Under the condition of sufficiently long duration of the isothermal testing period, the influence of stress relaxation on the initial slope of the non-isothermal stress - temperature curve is negligible. But, if the isothermal test period is shorter, the contribution of relax can no longer be neglected. For this reason a theoretical approach was developed, which is suitable to estimate the contribution of the isothermal stress relaxation. In this approach it is considered that under isothermal conditions the decay of stress can be described by the empirical function given in Eq. (11), which fits the experimental values very nicely.

( ) *<sup>c</sup>*

Differentiation of Eq. (12) with respect to time t and by considering the linear relationship between t and temperature T leads to Eq. (13), from which the theoretical contribution of stress relaxation relax on the experimentally obtained temperature coefficient can be esti-

*c*

Where β is the heating rate and tmax is the duration of the isothermal test period. The empirical parameters a, b and c of Eq. (12) are easily calculated from least - square fits by non-linear re-

The entire stress - curve, including the isothermal part, is presented as a function of time in Figure 15, as obtained from a TSSR - test of a standard commercial TPV sample of about 70 Shore A. In contrast to the standard test procedure, this test was performed with reduced duration of the isothermal test period of 60 minutes. As can be seen from the magnified part

 *b t* 

1

max *c*

*relax*

*relax entro py* (10)

*t a bt* (12)

(13)

Fig. 14. AFM phase images of TPVs based on EPDM/PP (left) and EPDM/HDPE (right)

This observation corresponds to results from DMA measurements, in particular loss tangent, which is significantly lower at room temperature in case of EPDM/HDPE compared with EPDM/PP (Vennemann et al., 2009). At higher temperatures, above 100 °C, the EPDM/HDPE exhibits stronger stress relaxation due to the melting of the HDPE matrix. But, at moderate temperatures, up to 70 °C, compression set and rubber elasticity of these materials are equal or even better than existing products based on EPDM/PP. Thus, TPVs based on EPDM/HDPE might be an alternative to well known EPDM/PP products if enhanced rubber elastic properties are required. Furthermore, in hard/soft combinations with HDPE these compounds may have advantages over other TPEs due to their good compatibility, which results in excellent adhesion. The phase morphology of TPVs based on EPDM/PP and EPDM/HDPE having the same blend ratio and similar composition are shown in Figure 14. In these AFM phase images the EPDM rubber phase is light coloured, whereas the thermoplastic hard phase appears dark. In the case of the EPDM/PP sample (left), sharp boundaries between the light coloured EPDM domains and the dark appearing PP matrix phase are recognizable. The size of the EPDM domains is less than 5 µm, which is typical for this type of material. The AFM phase image obtained from the EPDM/HDPE sample differs significantly from the EPDM/PP sample. Obviously, micro-phase separation also occurred, but the phase boundaries are not as sharp, as in case of EPDM/PP. Apparently, both polymers are interpenetrating each other at the phase boundaries, because of good compatibility of EPDM and HDPE. This might be also an explanation for higher interaction and higher viscosity of the EPDM/HDPE material, which results in better rubber elasticity but also in poorer processibility.

Recently, a novel powdery EPDM/HDPE material has been developed, which can be produced by means of a special two–step mixing process (Vennemann et al., 2009). Consistency and processibility of the powdery raw material is similar to UHMW-PE powder. That means, the only way of processing is compression moulding. But, due to the almost identical consistency and the very good compatibility of the EPDM/HDPE powder and the UHMW-PE, the production of double-layer or multi-layer plates is possible, by a simple compression moulding process. Thus, hard/soft combinations can be produced easily which combine the extraordinary high toughness and abrasion resistance of UHMW-PE with the rubber like behaviour of a TPV, based on EPDM/HDPE.

Fig. 14. AFM phase images of TPVs based on EPDM/PP (left) and EPDM/HDPE (right)

results in better rubber elasticity but also in poorer processibility.

rubber like behaviour of a TPV, based on EPDM/HDPE.

This observation corresponds to results from DMA measurements, in particular loss tangent, which is significantly lower at room temperature in case of EPDM/HDPE compared with EPDM/PP (Vennemann et al., 2009). At higher temperatures, above 100 °C, the EPDM/HDPE exhibits stronger stress relaxation due to the melting of the HDPE matrix. But, at moderate temperatures, up to 70 °C, compression set and rubber elasticity of these materials are equal or even better than existing products based on EPDM/PP. Thus, TPVs based on EPDM/HDPE might be an alternative to well known EPDM/PP products if enhanced rubber elastic properties are required. Furthermore, in hard/soft combinations with HDPE these compounds may have advantages over other TPEs due to their good compatibility, which results in excellent adhesion. The phase morphology of TPVs based on EPDM/PP and EPDM/HDPE having the same blend ratio and similar composition are shown in Figure 14. In these AFM phase images the EPDM rubber phase is light coloured, whereas the thermoplastic hard phase appears dark. In the case of the EPDM/PP sample (left), sharp boundaries between the light coloured EPDM domains and the dark appearing PP matrix phase are recognizable. The size of the EPDM domains is less than 5 µm, which is typical for this type of material. The AFM phase image obtained from the EPDM/HDPE sample differs significantly from the EPDM/PP sample. Obviously, micro-phase separation also occurred, but the phase boundaries are not as sharp, as in case of EPDM/PP. Apparently, both polymers are interpenetrating each other at the phase boundaries, because of good compatibility of EPDM and HDPE. This might be also an explanation for higher interaction and higher viscosity of the EPDM/HDPE material, which

Recently, a novel powdery EPDM/HDPE material has been developed, which can be produced by means of a special two–step mixing process (Vennemann et al., 2009). Consistency and processibility of the powdery raw material is similar to UHMW-PE powder. That means, the only way of processing is compression moulding. But, due to the almost identical consistency and the very good compatibility of the EPDM/HDPE powder and the UHMW-PE, the production of double-layer or multi-layer plates is possible, by a simple compression moulding process. Thus, hard/soft combinations can be produced easily which combine the extraordinary high toughness and abrasion resistance of UHMW-PE with the

#### **4.3 Determination of crosslink density by means of rapid TSSR - tests**

Although TSSR measurements are significantly faster and easier to perform than other methods to characterize the crosslink density of thermoset rubber and TPV, further improvement of the method is desired from industry, in particular by reduction of the test duration, without deterioration of the reliability of the results. High potential for the reduction of test duration is included by the isothermal relaxation period. During the isothermal test period, which lasts normally 2 h, the short time relaxation processes occur and the sample reaches a quasi equilibrium state before the non-isothermal test starts. If the isothermal test period is reduced to shorter values, the entropy effect will be partially compensated by stress relaxation and thus the obtained temperature coefficient is diminished, systematically. This problem can be solved by considering the effect of stress relaxation on the experimentally observable temperature coefficient . Generally, the total value of can be divided into two parts, as described in Eq. (10). It should be mentioned, that the entropic part of is positive, whereas the contribution of stress relaxation is negative.

$$\mathbf{\dot{\kappa}} = \mathbf{\dot{\kappa}}\_{relax} + \mathbf{\dot{\kappa}}\_{entropy} \tag{10}$$

Under the condition of sufficiently long duration of the isothermal testing period, the influence of stress relaxation on the initial slope of the non-isothermal stress - temperature curve is negligible. But, if the isothermal test period is shorter, the contribution of relax can no longer be neglected. For this reason a theoretical approach was developed, which is suitable to estimate the contribution of the isothermal stress relaxation. In this approach it is considered that under isothermal conditions the decay of stress can be described by the empirical function given in Eq. (11), which fits the experimental values very nicely.

$$
\sigma(t) = a + b \cdot t^{-c} \tag{12}
$$

Differentiation of Eq. (12) with respect to time t and by considering the linear relationship between t and temperature T leads to Eq. (13), from which the theoretical contribution of stress relaxation relax on the experimentally obtained temperature coefficient can be estimated.

$$\kappa\_{\text{relax}} = \frac{-c}{\beta} \cdot b \cdot t\_{\text{max}}^{-c-1} \tag{13}$$

Where β is the heating rate and tmax is the duration of the isothermal test period. The empirical parameters a, b and c of Eq. (12) are easily calculated from least - square fits by non-linear regression.

The entire stress - curve, including the isothermal part, is presented as a function of time in Figure 15, as obtained from a TSSR - test of a standard commercial TPV sample of about 70 Shore A. In contrast to the standard test procedure, this test was performed with reduced duration of the isothermal test period of 60 minutes. As can be seen from the magnified part

Characterization of Thermoplastic Elastomers

geous, in particular for production control.

**5. Conclusion** 

conditions of TSSR tests.

lower hardness.

tion properties of such complex systems.

by Means of Temperature Scanning Stress Relaxation Measurements 367

proaching an almost constant level at long periods of isothermal relaxation. After correction by Eq. (13), the influence of the duration of the isothermal relaxation period has almost vanished. Therefore, rapid TSSR - tests with strongly reduced isothermal relaxation period give comparable results than standard TSSR - tests, but in shorter time, which might be advanta-

The aim of this paper is to describe the opportunities of a new test method, especially developed to characterize the stress relaxation behaviour and thermoelastic properties of TPE. In contrast to conventional stress relaxation measurements the new TSSR test method is less time consuming and requires only a minimum on manual effort. Generally, three phenomena have to be considered if a stretched rubber sample is annealed under the conditions of non-isothermal TSSR tests. Stress relaxation or decrease of stress, immediately occurs after the strain has been applied to the sample. Because relaxation time constants strongly decrease with increasing temperature, stress relaxation is accelerated when temperature is scanned during a TSSR test. An opposite effect results from entropy elastic behaviour of the rubber sample. Due to increasing temperature a stretched rubber sample tends to contract and therefore the stress increases if the strain is kept constant. Furthermore, stress relaxation and entropy elastic behaviour will be superimposed by a slight increase of sample length, caused by thermal expansion. It is shown that thermal expansion of the sample is negligible, if the strain is sufficiently high. Basic equations for evaluation have been developed, taking into account the specific

The versatility of TSSR measurements has been demonstrated at several examples of commercial TPE and model compounds. The relaxation spectra of commercially available TPE based on SBC exhibit two significant peaks, which can be identified as glass transition temperature of the polystyrene end blocks and the melt temperature of an additional blend component, e.g. polypropylene. Blends of SBC and PPE were investigated to improve the heat resistance of the material. It has been shown that PPE and the polystyrene end blocks of SBC form a mixed phase with elevated glass transition temperature. The corresponding shift of glass transition temperature of the hard phase could be clearly identified from TSSR relaxation spectra. Thus, TSSR measurements are a suitable tool to determine the stress relaxa-

Results obtained from investigations of thermoplastic polyolefin blends (TPO) and dynamic vulcanizates (TPV) based on EPDM/PP and EPDM/HDPE, demonstrate the versatile opportunities of TSSR measurements to characterize stress relaxation behaviour and crosslink density. Comparison of commercial TPO and TPV of different hardness clearly show that the relaxation behaviour of the material is significantly improved by crosslinking of the rubber phase. It is also seen, the impact on stress relaxation is more pronounced for materials of

A model system of peroxide cured TPV based on EPDM/PP was investigated to determine the crosslink density of the rubber phase. By varying the amount of curatives the crosslink density of the samples has been altered within certain limits. These samples were subjected

of the curve in Figure 15 (right), the slope of the stress - curve increases immediately, after the non-isothermal test period has started. It becomes also obvious, that the slope of the non-isothermal stress - curve is partially reduced by the influence of the ongoing stress relaxation.

Fig. 15. Entire stress - curve (left) of a TSSR - test and zoomed part (right) of the curve, as indicated by the rectangle.

Fig. 16. Temperature coefficient (left) and apparent crosslink density (right), with and without correction, of a commercial TPV.

In order to determine the entropic part entropy of the experimentally obtained temperature coefficient , it is necessary to eliminate the contribution of stress relaxation by means of Eq. (13). Thus, an apparent value of crosslink density of the sample can be calculated from Eq. (8) when 0 is replaced by the entropic part entropy of the temperature coefficient. To demonstrate the influence of the correction, the experimentally obtained values of temperature coefficient are plotted as a function of isothermal relaxation period, with and without correction for comparison, on the left side of Figure 16. On the right hand side of Figure 16 the corresponding values of apparent crosslink density are presented, as calculated from Eq. (8) without and with correction by Eq. (13). Without correction, the values of temperature coefficient and crosslink density vary over a wide range, starting from negative values and approaching an almost constant level at long periods of isothermal relaxation. After correction by Eq. (13), the influence of the duration of the isothermal relaxation period has almost vanished. Therefore, rapid TSSR - tests with strongly reduced isothermal relaxation period give comparable results than standard TSSR - tests, but in shorter time, which might be advantageous, in particular for production control.
