Processing and Performance Attributes of Olefin Block ......shows that the long-term fatigue behav-ior of OBC foams departs from that of EVA and POE foams with the same cure state

Processing and Performance Attributes of Olefin Block Copolymers in Crosslinked Foams

Dow Elastomers

White Paper

Updated: March, 2014Authors: Kyle G. Kummer, Research Scientist, The Dow Chemical Company Gloria Stucchi, Marketing Specialist, The Dow Chemical Company Jose M. Rego, Product Leader, The Dow Chemical Company Shaofu Wu, Sr. Research Scientist, The Dow Chemical Company

2

Abstract

The material of choice in the foam industry for many years was ethylene vinyl acetate (EVA) copolymers. This dominance was based on its flexibility, adhesion performance, and foaming window. In the last ten years, ethylene-a-olefin interpolymers (or polyolefin elastomers [POEs]) have been integrated in foam formulations as a means to increase processability and cost effi-ciency. In recent years, the discovery of INFUSE™ Olefin Block Copolymers (OBCs) has increased the benefits of using ethylene-a-olefin interpolymers in foam applications. This paper shows that the characteristics that make INFUSE™ OBCs of interest in crosslinked (XL) foam systems are increased softness, improved shrinkage, and compression set resistance at elevated temperatures. The paper also shows that the long-term fatigue behav-ior of OBC foams departs from that of EVA and POE foams with the same cure state. The room temperature data allowed one to draw the following conclusions: (i) OBC foams showed a more elastic response than EVA and POE as demon-strated by lower final strain and faster recovery after dynamic testing; the elastic response of OBC foams was attributed to its block architecture, and (ii) materials based on short chain branching (OBC and POE) seemed to recover faster than the foams based on highly branched EVA.

Figure 1: Schematic Diagram of Short Chain Branching Between POEs (top) and OBCs (bottom)[1]

Figure 2: Apparent Viscosity versus Shear Rate for a 2.5 MI EVA, 5.0 MI OBC, and 5.0 MI POE (190°C, 15% strain)(1)

Random Copolymers

Olefin Block Copolymers

Less Comonomer and Higher Density

More Comonomer and Lower Density

Soft Blocks

Hard Blocks

0.1 1 10 100

100,000

10,000

1,000

Appa

rent

Vis

cosi

ty, P

oise

Apparent Shear Rate, 1/s

EVA

POEOBC

(1) Data per tests conducted by Dow. Test protocols and additional information available upon request. Properties shown are typical, not to be construed as specifications. Users should confirm results by their own tests.®™Trademark of The Dow Chemical Company (“Dow”) or an affiliated company of Dow

3

Introduction

OBCs have polymer chains with blocks of semi-crystalline “hard” and elasto-meric “soft” segments. Figure 1 (page 2) shows the chain structure for random copolymers like POEs and the blocky structure of OBCs [1].

Higher levels of peroxide, relative to EVA or POE foam formulations, are necessary to generate the same cure state in OBC-based foams. The difference in cure has been hypothesized to be due to the higher level of branching in the soft segments, which can cause a higher level of chain scission that competes with the formation of polymer crosslinks [2].

Figure 2 (page 2) shows the difference in apparent viscosity between a 2.5 melt index (MI) EVA, 5.0 MI OBC, and 5.0 MI POE. The narrow molecular weight dis-tribution of the POE and OBC limits the shear sensitivity of these polymers. EVAs have considerable amounts of long chain branching and show significantly more shear thinning behavior during high shear rate processing like injection mold-ing. Figures 3 and 4 show the melting and cooling behavior of EVA, OBC, and POE as measured by differential scanning calorimetry (DSC)(10°C/min in N2). Table 1 lists the following key properties of the three materials: density in grams/cubic centimeter, MI, peak melting tempera-ture (Tm), and crystallization temperature (Tc). Note that the OBC shows a signifi-cantly higher Tm and Tc compared to the EVA and POE.

The batch temperature must reach the Tm of the OBC to adequately melt and mix the ingredients during compound-ing. Care must be taken to make sure the batch does not exceed the decomposition point of the peroxide and blowing agent during mixing or the molding step.

Table 1: Density, Melt Index, Tm, and Tc of EVA, OBC, and POE(1)

Density (g/cm3)/ MI (g/10 min) Tm (°C) Tc (°C)

EVA (28% VA) 0.94/2.5 74 54

OBC 0.877/5.0 122 102

POE 0.870/5.0 62 44

Figure 3: DSC Heating Curve (10°C/min)(1)

Figure 4: DSC Cooling Curve (10°C/min)(1)

Hea

t Flo

w, W

/g

Temperature, °C

-80

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0-30 20 70 120 170

EVA Heat

POE HeatOBC Heat

Temperature, °C

Hea

t Flo

w, W

/g

-80 -30 20 70 120 170

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0

EVA Cool

POE CoolOBC Cool

(1) Data per tests conducted by Dow. Test protocols and additional information available upon request. Properties shown are typical, not to be construed as specifications. Users should confirm results by their own tests.

4

(1) Luperox is a registered trademark of Arkema Inc.

Compounding and Molding Recommendations

Mixing Foam Compounds

Similar to mixing any thermoset elas-tomer compound, mixing EVA or POE compounds for crosslinked injection molding (XL-IM) foams requires adequate shear input (for dispersing particulates) combined with good temperature control. One option is to use a traditional internal or continuous mixer for distributive mix-ing, followed by a roll mill for dispersive mixing. For typical EVA or POE com-pounding, all the ingredients are added to the internal mixer and the compound is mixed at 100-110°C for 3-5 minutes, minimum. The compound then drops onto a warm (90-100°C) roll mill, where it is mixed for another 3-5 minutes.

Similar techniques can be used when adding OBCs, but the mixing tempera-ture needs to increase to 115-125°C to completely melt the hard segments of the OBC and promote good dispersion of the ingredients. It is important to note, however, that higher tempera-tures during mixing can cause scorch, which is defined as premature curing of the compound. To reduce such occur-rences, a portion of the dicumyl peroxide (DCP) typically used in the formulation should be replaced by a scorch-protected peroxide such as Luperox(1) DC40P-SP2.

Depending on the actual temperature and time needed to fully melt and dis-perse the ingredients, up to 30 percent or more scorch-protected peroxide should be incorporated into the DCP.

Unlike a traditional thermoset elastomer, however, the XL-IM foam compound must be pelletized, granulated, or diced so it can feed in the hopper of a mold-ing machine. Granulation is the least desirable option – the result is usually so fluffy, it free flows poorly, if at all. The use of a dicer to pelletize directly from a strip as it leaves the roll mill is another option, though dicers are rare. The easi-est and most trouble-free approach is to feed the material from the mixer or roll mill into a short length to diameter (L/D) ratio extruder with an air- or water-cooled die face cutter and keep the batch temperature above the polymer Tc prior to pelletizing.

Another option for compounding is the use of a continuous compounding extruder (single screw co-kneader or twin screw extruder [TSE]) equipped with accu-rate pellet and powder feeders. The single screw co-kneader type compounder is designed to minimize shear (and excessive heating) while providing good dispersion of ingredients. In a TSE, the screw design must provide high energy input for mix-ing while limiting shear heating of the compound. Consult with the equipment manufacturer for a recommended screw design. Again, there may be a need to introduce a scorch-protected peroxide to maintain a broad processing window.

The MI of the foamable compound should be in the 3-6 range for general purpose injection molding. Lower MIs may be feasible for some parts, but reducing polymer MI will increase shear heating and reduce flow during molding.

Injection molding of crosslinked foams parallels the conventional injection mold-ing process for any thermoset polymer. Pellets which have been compounded with blowing and crosslinking agents, filler, pigment, etc. are fed into a single screw injection unit. Barrel temperatures are set above the melting point of the polymer, and below the temperature where significant blowing agent and/or peroxide decomposition occurs. Naturally, the appropriate barrel tem-perature is a function of polymer melting point, molecular weight, cure kinetics, and half-lives of the blowing agent and peroxide. For EVA and POE midsole compounds using DCP, typical barrel heat settings are in the 90-95°C range.

The recommended profile for compounds containing OBC increases to 110-115°C. Processing slightly below the OBC melt-ing temperature allows for shear heating to completely melt the OBC phase. The temperature profile should be adjusted to optimize the process.

Inevitably, blowing agent decomposition will occur, requiring the use of a shut-off nozzle to prevent “drool” between shots. The molten but still un-foamed compound is injected into a heated mold, where it is held under pressure while the blowing agent and peroxide decompose. Although most machines inject using the conventional reciprocating screw tech-nology commonly used in thermoplastic molding machines, some older units fill the mold by rotating the screw, resulting in poor control of shot size and fill speed. These machines are sensitive to variation in pellet feeding.

5

Because the incoming polymer is inher-ently gassy, filling is generally improved by pulling a vacuum on the mold cavity. Evacuating the gas from the mold also helps reduce “dieseling” (burning) in difficult to vent sections of the mold. Mold pressurization is ensured by a delay in retracting the nozzle from the mold, allowing time for the narrow gate section to crosslink and prevent cavity back flow. Cure time is a function of the maximum wall thickness, and generally ranges from 4-15 minutes. Midsoles typically run about 6-7 minutes cure time. As a result of the long cure times, a single injection unit can service multiple molds. XL-IM foam molding machines typically contain 6 or 10 (or more) clamp stations per injection unit. Mold temperature is usu-ally set about 170°C, but can range from 160-185°C. Temperature control must be accurate within ± 1°C to ensure consis-tent expansion.

When the mold opens, the parts literally jump out of the cavities as they expand. This vigorous self-ejection can toler-ate even deep undercuts in the mold, allowing considerable design freedom. Expansion ratio is defined as P / M, where P is some final part dimension, and M is the dimension of the same feature in the mold. Outsoles are typically produced at expansion ratios of 1.20 to 1.40, sandals from 1.4 to 1.6, midsoles from 1.60 or higher, and floats at 1.80 or greater.

Instantaneous mold release, however, is critical. Any hang-up causes anisotropic expansion, resulting in misshapen parts. Often, molds are sprayed with a silicone release agent after each shot. The hot parts are soft and easily deformed until cool. Many molders take extra time in production to pass the parts through a controlled temperature conveyor oven set at 50-60°C to prevent curling by non-uniform cooling tunnels and allow for dimensional stabilization.

Foam Molding Experience

The following observations were made while molding POE and OBC compounds on a single station IM machine:1. Release improves with increasing

peroxide (unlike other thermoset processes).

2. Expansion ratio is less dependent on peroxide level than with bun foam – blowing agent level dominates.

3. Poorly mixed compound with undis-persed filler and blowing agent results in foam with coarse cells and high expansion for the level of blowing agent. Large cells tend to expand more easily than small cells. Expansion is a function of mold geometry, as thick sec-tions expand more than thin sections. A possible explanation is that the thick skin inhibits expansion in areas with high surface area to volume ratios.

Materials are many times characterized by simple mechanical responses such as compression, tensile, or bending modes. Thermal properties are also frequently part of such characterizations. Regardless of content, however, this type of charac-terization is often insufficient to describe the real performance of the material

for the specific application. This is believed to be the case for XL foams used in footwear applications (midsoles or outsoles), which demand property reten-tion over not only thousands of load and unload cycles, but also a wide range of temperatures. Investigations to evaluate these types of performance are typically referred to as fatigue testing [3, 4].

The performance of XL foams is also dependent on the temperature of the environment in which they are used. Midsole foams, for example, can be subjected to extreme cold and elevated temperatures depending on the climatic conditions. For this reason, the properties of the XL foams need to be determined at different temperatures [5, 6].

It is very desirable for foamed material to show stable performance, including durometer hardness and dynamic proper-ties across a wide temperature range. Dynamic testing provides information about the structural changes as a func-tion of cyclic fatigue loading conditions. Other properties, such as dynamic modu-lus, set, and recovery relate to how XL foams change with cyclic loading/unload-ing at various temperatures [4, 7].

The research discussed in the Experimental Details section of this paper (page 6) is focused on studying fatigue performance after 100,000 cycles of EVA, POE, and OBC-based foams with the same cure state across a broad temperature range.

For the sake of clarity, reference [8] will also be summarized after the Experimental Details section.

6

Experimental Details

Materials

The raw materials used in this study are listed in Table 2. The formulations stud-ied are given in Table 3.

Testing

Testing was done on XL foams (with skin) from compounded ingredients. All foam samples were prepared by The Dow Chemical Company (Dow) using a Kobelco internal mixer, Reliable 12-inch roll mill and Carver press.

Cure State: 3.0 g samples of the com-pounds were placed into a moving die rheometer (MDR) to measure the cure behavior at 180°C over 10 minutes. The final cure state was calculated by subtracting the maximum torque value (MH) from the minimum torque value (ML) measured during the test.

Density: XL foams, with the skin layer, were weighed to the nearest 0.1 g, and their volume was determined by measuring length, width, and thickness to the nearest 0.01 cm. After removing the skin layer, the remaining foam core was also measured to the same 0.1 g and 0.01 cm dimensions, and the foam density was calculated.

Hardness: The durometer hardness of the foams was measured at several temperatures to determine the effect of lower temperatures on hardness. A standard kitchen refrigerator/freezer was used and modified to reduce the change in temperature during the data collec-tion step. A hand-held durometer was used to measure the Shore A hardness. A minimum of five readings were collected per sample per temperature.

Table 4: XL Foam Formulations from Reference [8](1)

Ingredient (PHR(5)) EVA OBC POE

Elvax(2) 265 100 – –

INFUSE™ 9500 OBC – 100 –

ENGAGE™ 8200 POE – – 100

Luperox(3) DC40P-SP2 4.4 4.4 4.4

Celogen(4) AZ-130 4 3 5

Activators 0.4 0.3 0.5

CaCO3 5 5 5

Table 2: Raw Materials(1)

Ingredient Density, g/cm3 MI, g/10 min Supplier

Elvax(2) 265 (28% VA) 0.960 2.5 DuPont

INFUSE™ 9500 OBC 0.877 5 Dow Chemical

ENGAGE™ 8200 POE 0.870 5 Dow Chemical

Luperox(3) DC40P-SP2 – – Arkema Inc.

Celogen(4) AZ-130 – – Lion Copolymer

Zinc Oxide – – Zinc Corp

Zinc Stearate – – Fisher

CaCO3 – – IMERYS

Table 3: Formulations Studied in This Paper(1)

Ingredient (PHR(5)) EVA OBC POE

Elvax(2) 265 100 – –

INFUSE™ 9500 OBC – 100 –

ENGAGE™ 8200 POE – – 100

Luperox(3) DC40P-SP2 2.0 3.5 3.0

Celogen(4) AZ-130 2.5 2.5 3.0

Zinc Oxide 0.25 0.25 0.30

Zinc Stearate 0.25 0.25 0.30

CaCO3 5 5 5

(1) Data per tests conducted by Dow. Test protocols and additional information available upon request. Properties shown are typical, not to be construed as specifications. Users should confirm results by their own tests.(2) Elvax is a registered trademark of E. I. du Pont de Nemours and Company(3) Luperox is a registered trademark of Arkema Inc.(4) Celogen is a registered trademark of LION COPOLYMER GEISMAR, LLC(5) Parts per Hundred Resin

7

Mechanical Properties – Tensile and Tear: The skin and foam layers were removed before testing their tensile and tear mechanical properties according to ASTM D638 (Tensile) and ASTM D624 (Type C tear). The sample thickness was approximately 2 mm. At least five speci-mens were measured for each layer to determine the average result.

Compression Set: Compression set (CSet) was measured per ASTM D395 method B under conditions of 25 percent compres-sion at 50°C for 6 hours, and 23°C for 22 hours. Two buttons were tested per foam and the average reported.

Foam Shrinkage: The XL foams, includ-ing the skin, were cut using a 7.62-cm x 7.62-cm die. The mass and thickness of the samples were determined. Then, they were placed on non-stick sheets in separate pre-heated ovens at 70°C and 100°C for 24 hours. Selection of the two test temperatures was based on tem-peratures that the foams may encounter post-production (shipping/storage conditions and high heat applications). Sample dimensions were re-measured again after cooling 24 hours at room temperature and the shrinkage was calculated using the equation:

% Shrink = Aa

Ab

x 100

where Aa = surface area after shrinkage and Ab = surface area before aging.

Fatigue Testing: Fatigue was measured with a hysteresis method that allows measurement of different properties simultaneously, viz., stress, strain ampli-tude, stiffness, stored and lost energies, material damping, and cyclic creep behavior. A servo-hydraulic test machine (MTS 8100) with a digital controller and a software package for the evaluation of the hysteresis loop was used in this study. The software package makes it possible to digi-tize the hysteresis loop of material, which is measured continuously, and a mid-curve is calculated in the hysteresis loop.

Specimens measuring 3.81 cm x 3.81 cm were cut from the XL foams. These specimens were subjected to compression mode sinusoidal oscillations in the range of 5-75 psi load at a frequency of 2 Hz for up to 6,000 cycles. This load simulates the pressure on the foam exerted by a person weighing 180 lbs. Each sample was run at three test temperatures (-20, 23, and 40°C). It should be noted that at the higher test temperature, there is a chance that foam shrinkage can occur during testing for long periods.

On the original test [8], the MTS tun-ing parameters were set as follows: Proportional (P) = 38; Interval (I) = 3.8. With these settings, the equipment was meeting the set endpoints on the initial part of the test but as the geometry and modulus of the samples changed, the equipment became unstable and only 100,000 cycles could be measured. This was overcome by adjusting the tuning parameters to the following settings: P = 25; I = 2.5.

Summary of Reference [8]

As mentioned earlier, this section will be dedicated to summarizing some of the key findings of the first study [8]. The formulations studied are listed in Table 4 (page 6).

Cure State

Figure 5 (page 8) shows the cure behavior of the three compounds at 180°C for 10 minutes. The OBC compound showed lower cure state than the EVA and POE, although the final expansion ratio and density of the XL foams were similar.

The cure state difference between OBC and POE can be explained by consider-ing the influence of a-olefin comonomer concentration (or short chain branching level) on cure state. This is shown in Figure 6 (page 8), with the approximate soft segment density of OBC highlighted by a star. It is apparent from this figure that the lower the density, the higher the comonomer concentration (and short chain branching level) and the lower the cure state. This explains the lower OBC cure state (versus the POE) since the soft segment of OBC has a lower density (or higher short chain branching concentra-tion) than the POE.

8

Compression Set

The room temperature and 50°C com-pression set (CSet) results are shown in Figure 7.

The OBC foam showed slightly higher CSet at room temperature compared to the POE and similar CSet to the EVA. This could be explained as a result of the difference in cure state. However, at elevated temperature, the OBC showed improved CSet performance versus both EVA and POE. At this temperature, a high proportion of the EVA and POE crystals melt, which decreases the overall net-work density. This is not the case for OBC due to the intrinsic higher melting point of these materials, and this explains the results obtained.

Durometer Hardness

Figure 8 (page 9) shows the change in Shore A hardness as the temperature decreases from room temperature to -20°C. The rate of change in hardness is lowest for the OBC foam followed by the POE and EVA. The delta change for each sample across the temperature range tested is listed in the figure. The signifi-cance of this result is the OBC remains soft as the temperature decreases and dimensionally stable as the temperature increases from ambient conditions. This could be related to the glass transition temperature, which is < -50°C for POE and OBC, but -28°C for EVA.

Figure 7: Comparison of Compression Set at 23 and 50°C(1)

Figure 5: Cure Behavior of the EVA, OBC, and POE Compounds (180°C for 10 min)(1)

Figure 6: Influence of Short Chain Branching Level (or Comonomer Concentration) on OBC Cure State(1)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0.00 2.00 4.00 6.00 8.00 10.00

Torq

ue, d

N-m

Cure Time, minutes

EVA

POE

OBC

6.76

5.03

3.92

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

0.850.860.870.880.890.9

Torq

ue, d

N-m

Polymer Density, g/cm3

Approximate OBC SoftSegment Density

0

10

20

30

40

50

60

70

80

20 25 30 35 40 45 50 55

Com

pres

sion

Set

, %

Temperature, °C

EVA

POE

OBC


9

Dimensional Stability with Temperature

Figure 9 shows that OBC foams remain dimensionally stable after aging up to 24 hours at 100°C. The effect of increasing temperature caused the POE and EVA foams to shrink, which contributed to their change in hardness measured after aging. Little change was observed for the OBC sample. This is also explained as a result of materials’ thermal properties.

Fatigue Resistance versus Temperature

The effect of temperature on the hyster-esis behavior of the EVA foam is shown in Figure 10. The EVA foam dynamic modulus decreases significantly from low tempera-tures (-20°C) up to ambient test conditions (23°C), which supports the observed change in EVA foam hardness versus temperature results shown in Figure 8.

The dynamic modulus of all materials as a function of temperature at a fixed number of cycles (1,000) is shown in Figure 11 (page 10). The dynamic modulus is defined as the ratio of stress differ-ence to strain difference between the minimum and maximum points on the hysteresis loop.

In summary, the main conclusions from this first study were: (i) the OBC foam showed improved compression set and dimensional stability at elevated temper-ature; (ii) the dimensional stability versus temperature of OBC foams is remarkable compared to the change in density and hardness associated with the EVA and POE foams as temperature increases; and (iii) the fatigue behavior of limited dura-tion dynamic testing showed that the dynamic modulus of OBC and POE foams were not sensitive to temperature in the range studied, whereas the EVA foam was highly sensitive at low temperatures.

Figure 9: Change in Density Before and After Aging at 70 and 100°C for 24 Hours(1)

Figure 10: Effect of Temperature on Hysteresis Loop at 5,000 Cycles for EVA Foam Samples

Figure 8: Effect of Temperature on Durometer Hardness(1)

0

10

20

30

40

50

60

70

-20-100102030

Har

dnes

s, S

hore

A

Temperature, °C

EVA

POEOBC

0

100

200

300

400

500

600

700

800

900

1,000

20 30 40 50 60 70 80 90 100

Den

sity

, Kg

/m3

Temperature, °C

EVA

OBC

POE

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.002 0.004 0.006 0.008 0.010

Displacement, m

Load

, kN

EVA at -20°CEVA at 23°CEVA at 40°C


Sample Sh A Change Tg, °CEVA 25.7 -28OBC 7.5 -62POE 10.8 -53

10

Results and Discussion

Figure 12 shows that the formulations in Table 3 (page 6) did produce the same cure state for all materials studied here. This also translated into slightly different overall densities (with skin and foam), as shown in Figure 13.

The dynamic stress versus strain at -20 and 40°C are shown in Figure 14 (page 11). A couple differences in how these foams behave relate to the increase in displace-ment (% Strain) before the maximum stress is reached as the number of cycles and test temperatures increase. The slope of the Stress versus % Strain loop increases with the number of cycles, which is a measure of the dynamic modu-lus shown in Figure 15 (page 11).

It is apparent from Figures 14 and 15 (page 11) that, as the number of cycles increases, the foams studied exhibit an increase in dynamic modulus or, in other words, are hardening. This could be due to the foams’ inability to recover com-pletely during the load-unload cycle. This dimensional setting accumulates cycle by cycle and renders a change in the materi-als’ modulus. This also explains the shift in strain with the number of compression cycles. It would appear from Figure 11 that this shift is more pronounced for POE, followed by EVA, and then the OBC. This could be explained by the block molecular architecture of OBC, which provides addi-tional crosslinking density (and stronger crystals) at the test temperature (in this case room temperature) and increases the elastic response of the network.

Figure 11: Dynamic Modulus versus Test Temperature After 5,000 Test Cycles(1)

Figure 12: Torque as a Function of Cure Time for All Formulations (180ºC for 10 min)(1)

Figure 13: Density with and without Skin for All Foams Tested(1)

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

-20 -10 0 10 20 30 40

Dyn

amic

Mod

ulus

, MPa

Temperature, °C

EVA

POEOBC

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0 2 4 6 8 10

Torq

ue, d

N-m

Cure Time, minutes

EVA

POEOBC

100

110

120

130

140

150

160

170

180

EVA OBC POE

Den

sity

, Kg/

m3

Foam Sample

Density, FoamDensity, Skin


11

Figure 14: Hysteresis Loops for 5,000 and 100,000 Cycles for the EVA, POE, and OBC Foams Tested(1)

Figure 15: Dynamic Modulus versus Fatigue Cycle(1)

0.02.04.06.08.0

10.012.014.0

100 1,000 10,000 100,000 1,000,000

Dyn

amic

Mod

ulus

, MPa

Fatigue Cycle

Tested @ 40°C

0.02.04.06.08.0

10.012.014.0 Tested @ 23°C

0.02.04.06.08.0

10.012.014.0

EVA

POEOBC

Tested @ -20°C

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 20 40 60 80 100

Stre

ss, M

Pa

Strain, %

EVAOBCPOE

0 20 40 60 80 100Strain, %

0 20 40 60 80 100

Strain, %

0 20 40 60 80 100Strain, %

EVAOBCPOE

EVAOBCPOE

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Stre

ss, M

Pa

EVAOBCPOE

5K Cycles at -20°C 100K Cycles at -20°C

5K Cycles at 40°C 100K Cycles at 40°C


12

Further confirmation of the higher elastic response of the OBC foam is provided in Figure 16, which shows the recovery (in %) as a function of time (in hours) up to 500,000 hysteresis cycles.

Little difference in recovery was observed at -20°C. As the test temperature increased, however, it became apparent that the OBC foam had a higher recovery rate, followed by the POE and EVA foams. After room temperature testing for 5 hours, the OBC foam has recovered 51 percent of its initial dimensions, whereas the POE and EVA foams have only recovered about 40 percent.

At this point, the POE foam shows an acceleration of recovery and after 24 hours, both the OBC and POE foams have recovered about 70 percent of initial dimensions while the EVA foam has only recovered 50 percent. The acceleration of recovery only happens for the EVA foam after 96 hours when it reaches 67 percent. The rate of recovery must certainly be related to the molecular architecture of the precursor materials. It can be hypothesized here that the block architecture of OBC decreases the level of setting during dynamic testing, and that a short chain branching architecture (OBC and POE) is faster at recovering from the fatigue cycles than the highly branched EVA. Recovery is important in applications like footwear or protective padding where you want to keep the material from compressing over time and losing its ability to cushion.

Figure 16: Recovery of Dimensions as a Function of Time After 100,000 and 500,000 Cycles of Dynamic Testing(1)

20

40

60

80

100

20

40

60

80

100

20

40

60

80

100

0.01 0.1 1 10 100 1,000

Rec

over

y, %

Time, hours

EVA

POEOBC

100K Cycles @ -20°C

500K Cycles @ 23°C

500K Cycles @ 40°C


13

The samples used to test dynamic per-formance are presented in Figure 17. The top image shows sample thickness prior to testing and the bottom image shows the same samples after testing at 40°C for 500,000 cycles and allowing time to recover. The important point here is that all three samples looked like the EVA sample in the bottom image as soon as the test was completed. Over time, the OBC and POE samples recovered part of their original thickness while the EVA sample did not.

Figure 17: Before and After 500,000 Cycles of Dynamic Testing at 40°C and Recovery(1)

EVA OBC POE

EVA OBC POE

After

Before

(1) Data per tests conducted by Dow. Test protocols and additional information available upon request. Properties shown are typical, not to be construed as specifications. Users should confirm results by their own tests.(2) Luperox is a registered trademark of Arkema Inc.

Conclusions(1)

The XL-IM foam process is a growing tech-nology for producing lightweight parts to a net shape in one step. Processing of OBCs requires some adjustment in proc-essing conditions. For example:• Higher compounding and molding

temperatures are required when proc-essing OBCs.

• Replacing a portion of the typical dicumyl peroxide used with a scorch-protected version such as Luperox(2) DC40P-SP2 reduces the tendency to scorch during processing.

• The batch temperature should be kept above the OBC crystallization tempera-ture (105°C) prior to pelletization.

• Blends of different MI polymers are rec-ommended to adjust the flow properties of the compound.

• Blends of different density polymers can be used to adjust physical and mechani-cal properties.

The dynamic test results showed the fol-lowing behavior:

(i) The hysteresis loops of all foams tested showed an increase in dynamic modulus and a shift to higher strain as the number of cycles accumulated. This was explained as a result of the inability of these materials to recover from the deformation in the time scale available between cycles.

(ii) OBC foams showed a more elastic response than POE and EVA as demon-strated by lower final strain and faster recovery after dynamic testing.

(iii) The elastic response of OBC foams was attributed to its block architecture.

(iv) Short chain branching seemed to provide a faster recovery than highly branched polymers, as demonstrated by the longer time associated with the EVA foam recovery.

14

References

[1] S.V. Karande, Y.W Cheung, C.F. Diehl, and M.J. Levinson, SPE ANTEC Technical Papers, pp. 1000-1004 (May 2006).

[2] K.G. Kummer, J.M. Rego, and S. Wu, “Dynamic Properties of Crosslinked Polyolefin Foams,” SPE ANTEC 2012, Orlando, Florida, p. 2366 (April 2012).

[3] M. El Fray and V. Altstadt, “Fatigue behavior of multiblock thermoplastic elastomers. 1. Stepwise increasing load testing of poly(aliphatic/aromatic-ester) copolymers,” Polymer, Vol. 44, pp. 4635-4642 (2003).

[4] R. Verdejo and N.J. Mills, “Heel-shoe interactions and the durability of EVA foam running-shoe mid-soles,” Journal of Biomechanics, Vol. 37, pp. 1379-1386 (2004).

[5] M. El Fray and V. Altstadt, “Fatigue behavior of multiblock thermoplastic elastomers. 2. Dynamic creep of poly (aliphatic/aromatic-ester) copolymers,” Polymer, Vol. 44, pp. 4643-4650 (2003).

[6] K. Bruckner, S. Odenwald, and J. Heidenfelder, “Temperature dependence of midsole materials,” Taylor and Francis Group, London, United Kingdom, p. 361 (2010).

[7] F. Kleindienst, B. Krabbe, K. Westphal, and M. Grandmontagne, “Temperature influence of varying midsole hardness on functional properties,” 5th Symposium of Footwear Biomechanics, p. 54 (2001).

[8] K.G. Kummer, J.M. Rego, S. Wu, and V. Juarez, “Dynamic Properties of Crosslinked Olefin Foams,” SPE Foams 2011, Iselin, New Jersey (2011).

The principles of Responsible Care® and Sustainable Development influence the production of printed literature for The Dow Chemical Company (“Dow”). As a contribution towards the protection of our environment, Dow’s printed literature is produced in small quantities and on paper containing recovered/post-consumer fiber and using 100 percent soy-based ink whenever possible.

NOTICE: Any photographs of end-use applications in this document represent potential end-use applications but do not necessarily represent current commercial applications, nor do they represent an endorsement by Dow of the actual products. Further, these photographs are for illustration purposes only and do not reflect either an endorsement or sponsorship of any other manufacturer for a specific potential end-use product or application, or for Dow, or for specific products manufactured by Dow.

NOTICE: No freedom from infringement of any patent owned by Dow or others is to be inferred. Because use conditions and applicable laws may differ from one location to another and may change with time, the Customer is responsible for determining whether products and the information in this document are appropriate for the Customer’s use and for ensuring that the Customer’s workplace and disposal practices are in compliance with applicable laws and other governmental enactments. Dow assumes no obligation or liability for the information in this document. NO WARRANTIES ARE GIVEN; ALL IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE ARE EXPRESSLY EXCLUDED.

NOTICE: If products are described as “experimental” or “developmental”: (1) product specifications may not be fully determined; (2) analysis of hazards and caution in handling and use are required; (3) there is greater potential for Dow to change specifications and/or discontinue production; and (4) although Dow may from time to time provide samples of such products, Dow is not obligated to supply or otherwise commercialize such products for any use or application whatsoever.

NOTICE REGARDING MEDICAL APPLICATION RESTRICTIONS: Dow will not knowingly sell or sample any product or service (“Product”) into any commercial or developmental application that is intended for:a. long-term or permanent contact with internal bodily fluids or tissues. “Long-term” is contact which exceeds 72 continuous hours;b. use in cardiac prosthetic devices regardless of the length of time involved (“cardiac prosthetic devices” include, but are not limited to, pacemaker leads and devices, artificial hearts, heart valves, intra-aortic

balloons and control systems, and ventricular bypass-assisted devices); c. use as a critical component in medical devices that support or sustain human life; ord. use specifically by pregnant women or in applications designed specifically to promote or interfere with human reproduction.

Dow requests that customers considering use of Dow products in medical applications notify Dow so that appropriate assessments may be conducted.

Dow does not endorse or claim suitability of its products for specific medical applications. It is the responsibility of the medical device or pharmaceutical manufacturer to determine that the Dow product is safe, lawful, and technically suitable for the intended use. DOW MAKES NO WARRANTIES, EXPRESS OR IMPLIED, CONCERNING THE SUITABILITY OF ANY DOW PRODUCT FOR USE IN MEDICAL APPLICATIONS.

This document is intended for global use.Published March, 2014.© 2014 The Dow Chemical Company

®™Trademark of The Dow Chemical Company (“Dow”) or an affiliated company of Dow®Responsible Care is a service mark of the American Chemistry Council. Dow is a partner in the American Chemistry Council Responsible Care initiative. Form No. 788-09901-0314X SMG

SMG 12521

dow.comdowelastomers.com

North AmericaU.S. & Canada 1 800 441 4369 1 989 832 1426Mexico + 1 800 441 4369

Latin AmericaArgentina + 54 11 4319 0100Brazil + 55 11 5188 9000Colombia + 57 1 219 6000Mexico + 52 55 5201 4700

Europe/Middle East + 800 3694 6367 + 31 115 672626Italy + 800 783 825

South Africa + 800 99 5078

Asia Pacific + 800 7776 7776 + 603 7965 5392

Documents

Processing and Performance Attributes of Olefin Block ......shows that the long-term fatigue behav-ior of OBC foams departs from that of EVA and POE foams with the same cure state