Use of crosslink density measurements to assess heat history in molded rubber parts.
In this article, we will show that in some instances, the question (viable/not viable/severity of heat history) may be answered with a high degree of confidence. This is done by preparing a series of air aged ASTM slabs which represent a range of samples from slightly aged to severely aged. The latter target is set objectively, with reference to SAE J2236, which defines the temperature ceiling for a compound as that temperature at which there is at least 50% retention for both tensile and % elongation.
Once prepared, the reference samples are measured for both tensile properties and crosslink density. With the resulting curve, we are able to use crosslink density to predict tensile properties, and thus answer the question: Is the rubber viable? A case study will be provided in which such a study answered a critical question concerning a customer returned sample.
A part was returned from the customer, an auto manufacturer. The part had given rise to a concern in a heat cycle test. While it passed the test, leaking neither oil nor air, portions of the part had stuck to the test shaft.
At first glance, the part did not look right at all. The thermal cycle test would be expected to challenge the rubber somewhat, but the apparent damage was severe, much more so than expected. Microhardness readings were taken and were much higher than expected. In addition, there were portions of the part stuck to the shaft (figures 1 and 2).
Ordinary application of crosslink density (XLD) testing
The most common type of testing is done to verify/disprove the contention that a particular part has been fully cured. For us, it will most often be one of our own compounds at FNST, so we will have access to uncured rubber, and also cure conditions, i.e., time/temperature at the press and, if appropriate, oven post cure.
Step 1: Prepare reference samples at values of TC50, TC70, TC90 and TC99, along with their post cured counterparts. Step 2: Subject reference samples and test samples to testing according to a reliable state of cure technique. In this case, we will use the swell technique and the Flory-Rehner equation for relative crosslink density. Since we will have prepared samples at the full range of cure conditions, we will be in a position to know what sort of sample variation to expect for a given cure condition as opposed to the sort of variation that indicates a significant difference in cure state.
How will the approach differ in this instance?
In this instance, there is not so much concern that the samples were fully cured when received by the customer. Our concern is that at some point they were subjected to far more heat history than was expected. The tested sample was quite hard. The part in question, a vacuum pump seal, had apparently left a residue on the mating shaft. Our principle concern, after establishing the correct compound was used, was to show that the heat history suffered by the part was in excess of that expected; or, to put the matter differently, that the customer's test was invalid. This is what we had to establish.
How to establish heat history
If we were examining slabs from an air aging oven, this would be straightforward; especially if we already had data showing trends in % of original values for tensile and elongation. We did not have such values at hand, not for the range of interest, so we had to produce the samples. We chose a temperature which would degrade the samples quickly. Our sample was a polyacrylate rubber, a family of compounds designed to withstand 1,000 hours of exposure to air at 150[degrees]C. With this as a guide, we established an accelerated test temperature. The rule of thumb, based upon the Arrhenius equation, is that a 10[degrees]C temperature rise translates to a twofold increase in reaction rate. Thus, (1,000 hours at 150[degrees]C) = (500 hours at 160[degrees]C) = (250 hours at 170[degrees]C), and so on. The relationship does not, of course, extend indefinitely, nor is it reliable for predicting behavior in actual samples. The rule is taken only as a guide. We tested samples to track their degradation experimentally. There is a range of time and temperature exposure such that exceeding one or the other will involve a different reaction mechanism than the one that governs typical air aged exposure. The graphs which follow show that our data range was selected in such a way that this rule was not violated.
Heat history expected
The parts in question were subjected to a thermal cycling test, with temperatures ranging from-18[degrees]C to 150[degrees]C, over a time interval of approximately 2,500 total test hours. Our seals made from this ACM material had passed this test before.
Critical values for tensile, elongation
There is a critical value for elongation and/or tensile values, i.e., loss of 50% of original for either tensile or elongation. Property loss in excess of this is arguably proof that the rubber sample is no longer viable. This is taken from the SAE J2236 standard, which addresses temperature ceilings for rubber compounds. What it means in practice: A compound that has a J2236 value of 150[degrees]C for 1,000 hours is one that retains at least 50% of both original tensile strength and ultimate elongation for this time/temperature combination.
Preparation and characterization of reference samples
ASTM slabs were prepared by press cure, followed by post cure. We prepared slabs for T0, and for 1, 2, 3, 4, 5, 6 and 7 days aging. The air aging oven was set for 200[degrees]C. Once we had all test slabs prepared, we tested them for:
* Tensile and % elongation at break
* IRHD microhardness
* Crosslink density
* 180[degrees] bends
Characterization of test samples
These samples were tested by FTIR and EDS to confirm the correct compound was used in production. They were also tested for:
* Crosslink density
For reasons noted above, they could not be tested for elongation and/or tensile values. Our intention was to produce a defensible approximation of these from the crosslink density values.
Confirmation of rubber compound
The rubber from the tested seal was compared by FTIR and SEM/EDS. The results are shown in figures 3 and 4. Both FTIR and EDS data support the claim that the rubber compound is the correct one.
Establishment of heat history
Tensile, elongation and crosslink density values were obtained for the reference samples. The results are given in table 1.
Translation of XLD to % elongation
Figure 5 will show that information on crosslink density translates to % elongation with a correlation coefficient of 0.9974, and that values of XLD greater than 1.75 E-04 represent nonviable rubber (figures 5 and 6).
Comparison to other FNST samples that passed the customer's test: Others were tested by XLD and yielded values ranging from 1.38 to 1.48 E-04. These values translate to at least 74% residual elongation, i.e., viable samples.
A happy ending
The last significant note from our applications engineer said this: "I spoke with (name withheld, an engineer for our customer) about this project today and he is beginning to strongly believe that they had a temperature related concern with the test setup."
Significance of the study
We believe this study shows that, in some cases, correlations may be drawn in a defensible manner between crosslink density and physical properties. This may be of interest to other rubber manufacturers who may have reason to suspect that their products are wrongly blamed for someone else's mistake.
This article is based on a paper presented at the 184th Technical Meeting of the Rubber Division, ACS, October 2013.
by Fred Fraser and Ryan Fleming, Freudenberg-NOK
Table 1--tensile, elongation, % original elongation and crosslink density values for air aged samples--ASTM D 573 method Exposure Tensile Elongation Elongation, Crosslink (hrs.) at at break at break % of density 200 (MPa) (%) original (moles/cc) [degrees]C 0 8.90 294 100 1.16E-04 24 7.55 293 100 1.18E-04 48 7.07 264 90 1.27E-04 72 6.45 244 83 1.36E-04 96 5.52 219 74 1.47E-04 120 5.37 178 61 1.61E-04 144 5.31 175 60 1.66E-04 168 5.20 107 36 1.90E-04 Forecast: 50 1.75E-04
|Printer friendly Cite/link Email Feedback|
|Author:||Fraser, Fred; Fleming, Ryan|
|Date:||Apr 1, 2014|
|Previous Article:||Development of new silicone rubber with high flex-fatigue resistance.|
|Next Article:||High performance Keltan ACE EPDM polymers.|