When one of our suppliers mixes or buys a batch of rubber, a batch number is automatically assigned. But before it can be molded into usable parts (such as O-rings), the batch must be tested to make sure it is a “good batch,” i.e. that its physical properties meet specifications. Batch testing is vital in ensuring consistency among finished parts.

To test a batch of rubber’s physical properties, a sample of the material is molded into 6" x 6" x .070" slabs. These slabs are then cut into the various shapes needed to test for hardness, tensile strength, modulus, elongation, and compression set. All of these tests are described in the Physical Properties section of “Selecting the Material.”

Specific gravity is also often measured, though more as a check on compounding consistency than as a physical test. Per ASTM D 792, specific gravity (or relative density) compares the weight of a molded sample to the weight of an equal volume of water. Specific gravity (SG) is noted without units. If a material is twice as heavy as water, its specific gravity is 2. Using the specific gravities of previously-molded compounds for comparison (e.g. a material may have an SG of 0.86, or less than that of water), a manufacturer can see if a sample is consistent with prior batches.

If the tested physical properties of a batch of rubber meet all specifications, the batch is approved for production of O-rings or other articles. If the properties are not satisfactory, the batch must either be reworked (broken down and reformulated) or scrapped. Scrapping an entire batch of rubber and starting over can be very costly and is thus a last resort. But even if the compound’s physical properties are acceptable, it must still meet processing requirements in order to be ready for use in a specific molding facility.

TOOLS FOR TESTING
At one time, the Mooney Viscometer was the most common tool used to determine the processing characteristics for a given batch of rubber. Many compounders still use the Mooney to verify viscosity (which is indicative of molecular weight) when obtaining raw polymer stock. This works because a compound’s resistance to being moved by the Mooney’s internal rotor is directly linked to its viscosity. The viscometer’s previous role as the chief indicator of processing traits, however, has now been usurped by the rheometer.

There are two main types of rheometers currently in use: the ODR and the MDR. The older of these, the Oscillating Disk Rheometer (ODR, see Figure 58), builds on the Mooney Viscometer’s rotor-based design. An ODR gauges the amount of torque (twisting force in pounds per inch, lb/in, or decinewtons per meter, dN/m) needed to oscillate a rotor within the rubber sample. Whereas a viscometer rotor relies on full rotation, the ODR rotor only moves back and forth across a small arc. This oscillation is less degrading to the material than in the viscometer, where destruction of the sample is typical.

ODR test results are also more reflective of actual cure conditions because constant high pressure and the desired vulcanization temperature are maintained on the sample. As testing progresses, the sample begins to behave in predictable ways. Viscosity briefly drops as the sample first heats up, but the chemical reaction soon starts. The rubber becomes more viscous due to crosslinking of the macromolecular chains. As a result, the amount of torque that is required to internally shear (deform) the sample increases. Using this increasing torque as a gauge, the ODR plots a cure curve (see Table 24) illustrating the state of cure for a given time and temperature.

Though the Monsanto ODR was for many years the most-used rheometer, a more recent development is the Moving Die Rheometer (MDR). Whereas the ODR uses an embedded rotor to torque the rubber sample, an MDR holds the sample between a pair of heated dies (metal plates forming a cavity). As one of the dies moves across a small arc, the other die gauges the reaction torque generated in the sample. This again results in a cure curve that can show the optimum cure time for the desired blend of properties. Since the MDR does not insert a rotor into the sample, many molders feel the MDR is less intrusive to the curing process and thus more objective and accurate than the ODR.

DETERMINING CURE TIMES
Whether generated by an ODR (Oscillating Disk Rheometer) or an MDR (Moving Die Rheometer), a cure curve is essentially “torque versus time (at a given temperature).” The torque value is a direct indication of the sample’s shear modulus (resistance to shearing deformation). A number of processing characteristics can also be read, including the minimum pressure needed to make the material flow properly into the mold cavity, scorch time (prior to vulcanization), optimum cure time (typically 85 to 95% of maximum cure), and maximum cure (prior to over cure). Keeping the initial cure slightly below the maximum helps avoid over cure by allowing leeway for any necessary post cure (controlled continuation of vulcanization to finish cure, drive off byproducts, and stabilize) or inadvertent after cure (uncontrolled continuation of vulcanization after heat is removed).

Though specific vulcanization questions can be answered via a cure curve, rheometers also help molders address more general concerns about processibility and consistency. No matter what the cure curve says, “optimum” cure time is largely a matter of economics. There is no “universal” cure time for a compound. A batch of rubber may have different cure times if given to different molders, depending on their capabilities.

The old adage about time being money is especially true when it comes to cycle time (the time between a given point in one molding cycle and the same point in the next cycle; e.g. loading of raw stock, through molding and unloading of finished parts, then to reloading; see Figure 59). Generally speaking, the longer the cycle time, the more expensive the process and the more costly the part. As a cost-cutting measure, manufacturers may increase mold temperature to decrease cure time. A 20° F boost can cut cure time in half, but this is not always advantageous. Sometimes the ratio of the time the mold is open (for unloading and reloading) to the time the mold is closed and in the press allows the mold temperature to dip below what is needed for full vulcanization. Partially-vulcanized, unusable parts can result.

Again, consistency among different batches of the same material is always a concern. The cure curve can serve as a “fingerprint” for a given batch of rubber. By comparing cure curves, it is possible to see if the properties present in one batch are present in another. Because wasted processing can be costly in terms of both time and money, compounding errors are much more economically spotted in batch testing than in subsequent stages of quality control, such as vulcanizate testing.

ASSIGNING CURE DATES
Quality control is aided by the batch number that was initially assigned to the rubber. This number follows the batch as it makes its way through the manufacturing process. When the batch (or a portion of it) is molded, a cure date is also assigned. This cure date consists of the quarter and year in which the parts are molded. For example, all parts molded in January, February, or March of 2001 have a cure date of 1Q01. Parts molded in October, November, or December of 2000 have a cure date of 4Q00. Both the batch number and cure date stay with the part through to the end user, thus assuring that complete traceability is maintained.

“Batch testing is vital in ensuring consistency among finished parts.”

Figure 58

Table 24

Figure 59