Physical testing procedures of extruded products with complex geometries.

Determining optimal cure time and precise quality control testing of profiles with complex extruded geometries presents unique challenges. Modulus provides a good indication of final properties. However, the traditional methods of press curing slabs and testing dumbbells is flawed, since it is not representative of a final product's processing, curing or geometry. It is also important to have a test that can provide feedback to control the process.

A technique is explored which uses the shaping and curing of a real part as a preparation for modulus testing. With an automatic density measurement and specially developed instrument fixtures, a part can be tested to accurately measure modulus and act as a quality control tool. The effect of production curing times and temperatures on product stress-strain properties will be evaluated. The modulus results will be analyzed to optimize the extrusion and curing process. The defined test is simple to perform and interpret.

Nearly all compound development begins with a series of laboratory tests. Although it is expected that this laboratory development will predict factory performance, significant differences often exist. The functions of factory mixing, batch variation and production shaping and curing rarely duplicate the same functions in the development laboratory. One significant difference is the shaping and curing of the production part versus a slab cured laboratory sample. Whereas a production part is not only a different geometry, it is most likely cured at a higher temperature than the laboratory press cured slab. The production part has also seen different heat history due to differences in processing. Finally, a production part may be cured under pressure as in compression, transfer or injection molding, under pressure as in autoclave curing, or without pressure as in LCM or microwave and hot air curing. These differences may result in significantly different physical properties of the production part versus the laboratory cured sample.

Many production parts are tested in their completed form. For example, vibration mounts, seals and o-rings are often tested in compression to simulate the part's performance. Extruded or molded window seals may be subjected to an insertion/extraction test to simulate its use in the field. It is common, therefore, to see performance specifications for testing a finished part in compression, tension or shear. Special apparatus are often required for these tests. The testing of parts directly is of great benefit, since a part which is used in compression, for example, may be tested directly in this mode. The test result is a cumulative result of processing, shaping and curing of the part, and is generally indicative of its final performance. If a part fails the test, the cause of the failure may not be readily apparent. The part may not meet shape and size specifications, thereby causing the failure. The material itself may have been out of tolerance, causing the failure, or it may not have been cured properly, resulting in the failure. Therefore, it is often necessary to perform additional tests to determine the causes of failures.

To relate to the state of cure and material properties, stress-strain tests may be performed. Production parts may be tested by sectioning and obtaining a specimen such that it may be tested using standard procedures. It is a common practice to prepare a part so that a standard or mini dumbbell specimen may be cut. The key to stress-strain testing is the ability to accurately determine the cross-sectional area of the specimen. This is one reason that the sectioned part is of such interest, in that the cross-sectional area may be obtained by standard means.

There are many parts that do not have a direct end use specification, but where a test of the part itself may be useful. This test of the factory-produced part could be used to correlate to development or as a production control test. However, the test must be fast and easily applied to be useful.

The following experiment describes an apparatus and a procedure for testing an extruded profile for stress-strain properties. The test results are related to the laboratory.

Description of procedure

An extruded profile presents two distinct problems when it comes to evaluating its stress-strain properties. The first is a method to hold the specimen in the testing instrument. Most sample grips clamp the specimen, creating a pressure or stress point, as shown in figure la. To overcome this, sample grips similar to those used for cord testing were designed and are shown in figure 1b. In this design, the stress is distributed along the circumference of the grips, reducing the stress at the grips and eliminating premature failure at the grips. The grips with a specimen are shown in figure 1c. A standard contact extensometer, not shown, is used to determine elongation.

[FIGURE 1 OMITTED]

The second problem associated with testing an extruded profile is that of calculating the area of a complex cross-section. The cross-sectional area of simple profiles, such as those of extruded tubing or solid roping, may be easily determined from several dimensional measurements. For more complex profiles, the nominal cross-section may be calculated from an engineering drawing, but this does not take production variance into account.

The proposed method used for these tests allows the area of any extruded profile to be determined and used for stressstrain calculations. A simplified calculation for the area of a complicated profile may be determined by taking the density and mass of a specimen of a precise, known length. Given the formulae for density (ref. 1) (1) and volume (ref. 2) (3), the cross-sectional area may be calculated according to formula (5) derived as follows:

D = W/V (1)

hence: V = W/D (2)

V = A * H (3)

from (2) and (3) W/D = A * H (4)

therefore: A = W/(D*H) (5)

Where: D = density, g/co; W = mass, g; V = volume, cc; A = area of the base of a regular shape; and H = height of a regular shape.

When the test is run, the operator simply needs to enter the sample mass, W, its density, D, and precise length, H, into the stress-strain program and the results will be expressed in engineering units of stress, Pa.

This procedure is fast and accurate. The theoretical density may be used if there is very little process variation. For the most representative result, independent of batch and process variation, the density may be experimentally determined for each specimen. It is important to insure that the specimen does not trap air during the density measurement. Examples of profiles which may be used in this method are shown in figure 2.

[FIGURE 2 OMITTED]

Experimental

A natural rubber compound was cured in a 25 m salt bath, LCM, at temperatures of 190, 200 and 210[degrees]C and line speeds of 16, 19, 22, 25 and 28 m/min. The line speeds, LCM residence times and LCM temperatures are given in table 1.

Laboratory press cured specimens were prepared at conditions shown in table 2. Cure times were selected to be approximately 1.5 times the tc90 value determined from a curemeter test at the appropriate temperatures.

Density measurements were taken using a densiTECH+ according to ASTM D297 (ref. 3). Stress at 200% elongation was measured using a tensiTECH according to ASTM D412 (ref. 4). The production specimens were precisely die cut to 5 cm and the cross-sectional area calculated according to formula (5). Standard cross-sectional area was calculated for slab cured specimens.

Cure tests were done on a rheoTECH MDPT at 1.67 Hz, 7% strain and temperatures of 190, 200 and 210[degres]C according to ASTM D5289 (ref. 5).

Results and discussion

Standard laboratory slabs were cured and tested. The results are shown in table 2. As expected, as cure temperature increased, the stress at 200% elongation decreased. It is, therefore, difficult to extrapolate the results to the fmished cured product based upon the results of the laboratory cured data. The most obvious difference is the higher temperature. Because of the extremely short cure time at the elevated production temperatures, laboratory slabs could not be accurately cured. The second, perhaps less obvious, difference is the mode of curing; cured under pressure in the lab or cured without pressure in production. Another difference is the heat history that the production batch sees in the extrusion and curing process versus the lab milling and curing. Finally, the geometry of the slab-cured and production-cured products differ greatly. The slab is a uniform 2 mm in thickness, whereas the production part varied in thickness from 1 mm to greater than 5 mm. The laboratory data do, however, give a general basis for how to evaluate the production modulus.

Curemeter test results are shown in figure 3. If the cure curves of figure 3 are examined, a window can be drawn encompassing curing (an increasing slope), cure stability (a slope of near zero) and reversion (a negative slope). There exist an optimum cure time and a maximum cure torque at each temperature. These data, combined with the stress strain data, give a second basis of comparison for evaluating the factory cured product.

[FIGURE 3 OMITTED]

The stress at 200% elongation for the LCM cured samples is given in table 1. The data are plotted against LCM residence time in figure 4. When comparing the stress at 200% elongation of the LCM cure to that of the slab, it is evident that the values are different, but approximately of the same magnitude. As expected, the stress at 200% elongation decreases as the LCM temperature increases. It is also seen that the tests at 190[degrees]C exhibit increasing modulus, the results at 200[degrees]C are stable and the results at 210[degrees]C are decreasing or reverting.

[FIGURE 4 OMITTED]

These data compare well with the cure test observations, as shown in figure 3. The combined use of the laboratory data gives a guide to the expected data for production parts, but does not fully define either the production curing speed or the final production modulus. The use of the described method and apparatus gives a better understanding of the production process and material properties, and this allows for a quick determination of the properties. The optimum LCM line speed and, therefore, the extrudate's residence time at each temperature can be determined from these results.

Summary and conclusions

The determination of line speeds and state of cure for extruded profiles is critical to assuring a product's performance. Stress-strain properties of the extrudate often provide a means to look at the properties of a product in order to assure the proper state of cure. A method for determining the cross-sectional area of a complex extrusion was shown using the density, length and mass of the specimen. Specially designed sampie grips allowed the stress-strain properties of a length of the extruded profile to be measured without premature failure of the specimen. The combined use of the cross-sectional area calculation and specially designed grips allow a fast and accurate method for determining stress-strain properties of a complex profile, providing feedback for process and quality control.

References

(1.) ASTM D1566-03, "Standard terminology relating to rubber, "ASTM International, vol. 9.01, 323 (2003).

(2.) A. Schwartz, "Calculus and analytical geometry," 2nd Edition, Holt, Rinehart and Winston, New York, 1969.

(3.) ASTM D297-93, "Standard test methods for rubber products-chemical analysis," ASTM International, vol. 9.01, 5 (2003).

(4.) ASTM D412-98a, "Standard test methods for vulcanized rubber and thermoplastic elastomers-tension," ASTM International, vol. 9.01, (2003).

(5.) ASTM D5289-98a, "Standard test method for rubber property-vulcanization using rotorless cure meters," ASTM International, vol. 9.01, (2003).

Table 1--production conditions and results I

             LCM
LCM line   residence           Stress @ 200% elongation, MPa
 speed       time
 m/min.      min.      190[degrees]C   200[degrees]C   210[degrees]C

   16        1.56          5.61            5.00            4.12
   19        1.32          5.51            4.96            4.23
   22        1.14          5.40            5.00            4.49
   25        1.00          5.05            4.89            4.68
   28        0.89          5.03            5.04            4.65

Table 2--laboratory press cured results

Cure temp.   Cure time   Stress C@ 200% elongation
[degrees]C     min.                 MPa

   160         7.50                6.64
   170         3.50                6.30
   180         1.60                5.25