Prediction of Service Time

Final products should do their job over many years often in deleterious environments. Therefore, it is desirable to have a proper method for prediction of the service time at hand. 

---

A Concise Introduction to Additives for Thermoplastic Polymers by Johannes Karl Fink. Published 2010. ISBN 978-0-470-60955-2. Written in an accessible and practical style, the book focuses on additives for thermoplastic polymers and describes 21 of the most important and commonly used additives from Plasticizers and Fillers to Optical Brighteners and Anti-Microbial additives. It also includes chapters on safety and hazards, and prediction of service time models. 

For details of the test methods used to measure physical properties reference is made to Handbook of Plastics Test Methods or the more recent Handbook of Polymer Testing [2, 3]. Standard tests have their limitations most were intended for quality control rather than prediction of service performance and produce arbitrary rather than fundamental measures of the properties. They do have the advantages of making data compatible with others and often have known reproducibility. In many standard methods the user is encouraged to opt for standard or preferred conditions which may not have relevance to the service conditions of the product. It is then sensible to base the testing on standard methods but to use more relevant conditions of, for example, time, temperature or stress. 

From the viewpoint of prediction of service lives, the photochemical deterioration processes of polymers used as paints and finishes are theoretically analyzed based upon unsteady state dynamics. Theoretical results are compared with experimental data under natural and accelerated exposure. Infrared spectra and scanning micrographs show that the deterioration proceeds continuously inwards from the surface, but differently with the exposure conditions. Parabolic (/t ) law was derived approximately for the increase in the depth of the deteriorated layer of polymers with time. Paying attention to the influence of the deterioration of polymeric finishes, the parabolic law involving a constant term was also derived for the progress of carbonation of concrete. These parabolic laws well predict the progress of deterioration and explain the protective function of finishes on reinforced concrete. 

Information on the natural weathering behaviour of joints is very useful. By combining this information with data from accelerated laboratory tests, some realistic predictions of service-lifetime may be made. Theoretical models of the pattern of bondline saturation of joints as a function of time of environmental exposure provide a useful appreciation of the possible extent of problems (e.g. Fig. 4.21). The process of joint failure, as observed in practice or in the laboratory, is frequently non-diagnostic i.e. it rarely reveals the true cause, or the series of stages, leading to deterioration or failure. 

Laboratory accelerated weathering devices have been used for more than 80 years with increasing importance concomitant with the development of more weather-able materials and the need to determine in a short time the effects of natural exposures over prolonged periods. The importance of these devices lies in their ability to accelerate the weathering processes imder controlled and reproducible conditions. They are particularly useful in research and development of new polymeric formulations. They are also used for quality control and specification testing. However, their application to prediction of service life under use conditions is still under development (see section on Laboratory Accelerated Versus Natural Weathering). 

In linear poly(ester)s, predictions of the end of service time can be based on the relationship between the ductile-brittle transition and the entanglement limit of the molecular weight. However, for crosslinked poly(ester)s, this method is not applicable. The reasons for complications in the degradation behavior arise from the diffusion control of hydrolysis kinetics, heterogeneity of semi-crystalline polymers, and variation of the hydrophilicity with the hydrolysis conversion. 

Weathering Resistance Prediction of service life, although one of the most important areas of testing, is probably the most uncertain. Cable jackets of synthetic rubber used in the Distant Early Warning (DEW) Line in Northern Canada in the 1950s were forecast to have a 20-year lifetime. Most are stiU in service. Similar synthetic rubber wire coverings were used at that time for automotive wire harness when... 

Estimates of service life are usually made either by natural or simulated trials or, most commonly, by accelerated tests with extrapolation to predict performance at longer times under less severe conditions. An alternative approach is to subject the product to environmental exposures which equate to the whole design life, and then to assess performance by real or simulated service tests (the end performance assessment). The exposures usually have to involve accelerated procedures and can be composed of several environmental agents applied simultaneously or sequentially. 

Most plastic components are designed on the basis of experience, and that experience should be the first source of information. However, the exact conditions of service history are rarely known and are generally less harsh that the extremes for which the product has been designed. The materials are likely to have changed in the time between the design of a product and its successor. Past failures are important, but could have been for reasons unconnected with the material. In spite of these cautions, predictions based on experience are no more likely to be uncertain than those based on accelerated trials. 

The purpose of the trial also affects the choice of degradation agents and the parameters used to monitor degradation. For comparison and quality control purposes, single agents are most frequently used. For prediction purposes multiple agents are more likely to be representative of service, but at the same time they make extrapolation rules more complicated. The parameters measured in trials to predict lifetime must be those critical to service, but in many instances of comparison or quality checks the choice of parameter can be heavily influenced by experimental convenience. 

The standard methods for testing creep, the elongation and possible rupture of a plastic under sustained load, are ISO 899-1 [34] for tension and ISO 899-2 [35] for flexure. Tests last typically for 1,000 hours or six weeks. Tests at higher temperatures may be required either because of a higher service temperature or to provide a prediction of longer term behaviour by time-temperature shifting. 

Recently, a method for predicting the remanent life of a reinforcing geotextile was proposed [1] in which the strain to failure of a sacrificial sample was divided by the current creep rate. This requires verification. However, very few methods have so far been proposed or used for monitoring plastics in service and at the same time providing a numerical prediction of their remaining life. The reason for this is not just that the methods are likely to be expensive and complicated, but that there are few applications of plastics which can compete in risk and replacement cost with a high temperature boiler or aircraft structure. 

After each process is separately understood, interactions can be put together to provide a life prediction of the durability of the component under actual service conditions. One approach is to use multiple regression analysis to develop predictive equations for failure times in which several parameters (e.g., temperature, relative humidity, and stress) are treated as independent variables.6... 

The straight-line method may be applied on the basis of units of production or predicted amount of service output, instead of life years. The depreciation may be based on miles, gallons, tons, number of unit pieces produced, or other measures of service output. This so-called unit-of-production or service-output method is particularly applicable when depletion occurs, as in the exploitation of natural resources. It should also be considered for properties having useful lives that are more dependent on the number of operations performed than on calendar time. 

Mechanical Characterization of Sulfur-Asphalt. The serviceable life of a pavement comes to an end when the distress it suffers from traffic and climatic stresses reduces significantly either the structural capacity or riding quality of the pavement below an acceptable minimum. Consequently, the material properties of most interest to pavement designers are those which permit the prediction of the various forms of distress—resilient modulus, fatigue, creep, time-temperature shift, rutting parameters, and thermal coefficient of expansion. These material properties are determined from resilient modulus tests, flexure fatigue tests, creep tests, permanent deformation tests, and thermal expansion tests. 

Cullander, C. and Guy, R. H. Visualizing the pathways of iontophoretic current flow in real time with laser-scanning confocal microscopy and the vibrating probe electrode. In R. C. Scott, R. H. Guy and J. Hadgralt, eds. Prediction of Percutaneous Penetration, Vol. 2, IBC Technical Services, London, 1991. 

---