Accelerated testing, 1.23

Accelerated testing can be seen as the least attractive option because of the inherent cost and difficulties of making reliable predictions. Nevertheless, in a great many circumstances accelerated tests have to be relied on at least in part. 

An account of available prediction techniques is included in the Practical Guide to the Assessment of the Useful Life of Rubbers (17). In general there are two stages to modelling the degradation process  

Using these relationships the change in property on exposure to longer times and lower levels of the degrading agent can be predicted. Clearly, the success of the process is critically dependent on the validity of the models used but, as mentioned above, extrapolation models have not generally been comprehensively validated. 

By far the best known model is the Arrhenius relation which relates temperature and reaction rate through  

Where K(T) is the reaction rate for the process E is the reaction energy R is the gas constant T is absolute temperature C is a constant. 

---

An extensive survey of accelerated test methods for anticorrosive coating performance which emphasizes the need to develop more meaningful methods of testing has been pubUshed (129). The most powerful tool available is the accumulated material in data banks correlating substrate, composition, apphcation conditions, and specifics of exposure environments with performance. 

The relative susceptibHity of several commercial aHoys is presented in Table 8. The index used is a relative rating based on integrating performance in various environments. These environments include the harsh condition of exposure to moist ammonia, Hght-to-moderate industrial atmospheres, marine atmosphere, and an accelerated test in Mattsson s solution. The latter testing is described in ASTM G30 and G37 (35,36) and is intended to simulate industrial atmospheres. The index is linear. A rating of 1000 relates to the most susceptible and zero designates immunity to stress corrosion. 

Corrosion Resistance. The environment to which a plated part is to be exposed should be a part of any definition of corrosion resistance. Problems arise in testing a part in its intended environment in part because of the long time period required. In many plating processes, corrosion resistance is direcdy proportional to the plate thickness, so a specification on plate thickness is a much faster method of indirectiy measuring corrosion resistance. In specifying a corrosion resistance requirement for the production of plated goods, accelerated tests are used especially if plate thicknesses caimot be related to corrosion protection. 

Sa.lt Spray Tests. One of the older accelerated corrosion tests is the salt spray test (40). Several modifications of this imperfect test have been proposed, some of which are even specified for particular appHcations. The neutral salt spray test persists, however, especially for coatings that are anodic to the substrate and for coatings that are dissolved or attacked by neutral salt fog. For cathodic coatings, such as nickel on steel, the test becomes a porosity test, because nickel is not attacked by neutral salt fog. Production specifications that call for 1000 hours salt spray resistance are not practical for quahty acceptance tests. In these cases, the neutral salt spray does not qualify as an accelerated test, and faster results from different test methods should be sought. 

Salt spray tests, humidity tests, and other accelerated tests, some usiag sulfur dioxide and carbon dioxide, have shown favorable results for tin—2inc ia comparison with 2iac, cadmium, and fin deposits. Chromating improves the performance. 

As of this writing the 2inc alloys are too new to have actual corrosion resistance data, except for that based on accelerated tests. Zinc—nickel usually shows better results than 2inc-cobalt in salt spray tests. The reverse is tme when the Kesternich test is used. Tin—2inc performs well in both salt spray and Kesternich tests, but appears only to equal 2inc plating and 2inc—nickel in humidity tests. 

Electrical Properties. Electrical properties are important for the corrosion protection of chip-on-board (COB) encapsulated devices. Accelerated temperature, humidity, and bias (THB) are usually used to test the embedding materials. Conventional accelerating testing is done at 85°C, 85% relative humidity, and d-c bias voltage. Triple-track test devices with tantalum nitride (Ta2N), titanium—palladium—gold (Ti—Pd—Au) metallizations with 76... 

Work on these and other new techniques continues, and it is hoped that a truly reliable, accelerated test or tests will be defined. 

The early carbon trap and SHED methods measured two components of evaporative emissions. Hot soak emissions were measured for a one hour period immediately after a vehicle had been driven on a prescribed cycle and the engine turned off. Diurnal emissions were also measured during a one hour event where the fuel tank was artificially heated. The one hour fuel temperature heat build was an accelerated test that was developed to represent a full day temperature heat build. 

Kurzprtifung,/. short teat, accelerated test, kurz-skulig, a. (Min.) short-columnar, -schlies-sen, v.t. (Elec.) short-circuit. Kurzschliffhalskolben, m. flask with short ground neck. 

In considering the corrosion of magnesium and its alloys it is important to examine the methods available for assessing corrosion tendencies and particularly those known as accelerated tests. Tests carried out by immersion in salt water or by spraying specimens regularly with sea-water are worthless as a means of determining the resistance of magnesium alloys under any other than the particular test conditions. Extrapolation to less corrosive conditions is not valid and even the assessment of the value of protective measures by such means is hardly possible. The reason is to be found in the fact that corrosion behaviour is directly related to the formation of insoluble... 

It was established that significant changes in resistance took place at the transition temperature and consequently sharp changes in protective properties. The resistance always fell with an increase in temperature and this may provide an explanation for the fact that accelerated tests using the same corrosion cycle, may not produce the same results if carried out at different temperatures. 

The behaviour of samples under the actual conditions of service is the final criterion, but unfortunately such observations take a long time to collect and assess, and the cautious extrapolation of data from accelerated tests must be relied on for forecasting the behaviour of anodised aluminium in any new environment. 

The most widely used accelerated tests are based on salt spray, and are covered by several Government Specifications. BS 1391 1952 (recently withdrawn) gives details of a hand-atomiser salt-spray test which employs synthetic sea-water and also of a sulphur-dioxide corrosion test. A continuous salt-spray test is described in ASTM B 117-61 and BS AU 148 Part 2(1969). Phosphate coatings are occasionally tested by continuous salt spray without a sealing oil film and are expected to withstand one or two hours spray without showing signs of rust the value of such a test in cases where sealing is normally undertaken is extremely doubtful. 

The use ot rotating discs to carrv itsi spaiiiicns has been extended to studies rtf protective coatings in what are considered to be accelerated tests o( such coatings tor service underwater... 

Some investigatorshave advocated a type of accelerated test in which the specimens are coupled in turn to a noble metal such as platinum in the corrosive environment and the currents generated in these galvanic couples are used as a measure of the relative corrosion resistance of the metals studied. This method has the defects of other electrolytic means of stimulating anodic corrosion, and, in addition, there is a further distortion of the normal corrosion reactions and processes by reason of the differences between the cathodic polarisation characteristics of the noble metal used as an artificial cathode and those of the cathodic surfaces of the metal in question when it is corroding normally. 

In conclusion it must be emphasised again that all the tests used are accelerated tests and only provide information on susceptibility to intergranular attack under the precise test conditions prevailing. They are quality control tests that may be used to demonstrate either that heat treatment has been carried out adequately or that a steel will withstand the test for a certain sensitising heat treatment. 

Streicher , however, considered this approach to be unsound and pointed out that the short duration of the potentiostatic studies carried out by France and Greene cannot be used to predict long-term behaviour in service. The prolonged dialogue between these workers was well summarised in the review article by Cowan and Tedmon who concluded that these particular potentiostatic tests cannot be regarded as accelerated tests for service environments and that predicting future industrial service for periods longer than the test is not advisable. 

---