Medical Device Link.

MATERIALS

A reality check

Figure 1: Screen grab of website stress–strain data for unreinforced PBT, according to ISO 11403 Part 1.1 (click image to enlarge)

A plastics material data sheet is rather like the cover on a novel, it is not the whole story. At best it can be used to compare similar products, at worst it gives a distorted view of the truth. A typical data sheet, given freely by all the major material suppliers and easily downloaded from their websites, will give information on yield stress, impact strength, temperature of deflection under load and shrinkage. Figure 1 and Table I are examples of the typical data that are supplied in this case for unreinforced polybutylene terephthalate (PBT). For materials used in medical applications, there may also be information on sterilisation and testing according to relevant standards such as United States Pharmacopoeia Class VI and ISO 10993, Biological Evaluation of Medical Devices, Parts 1–20. It is useful to examine some of the pieces of data commonly found on a data sheet.

Yield stress

The first property often quoted is yield stress and two questions immediately come to mind. On the stress–strain curve generated to arrive at this figure, where exactly does yield occur? It is rarely a definite spot where plastics are concerned. Thus, a stress pertaining to a particular strain is usually quoted, this is often the first stress at which an increase in strain occurs without an increase in stress. The second question is: Does this point on the curve in any way relate to what will be experienced in real life? The honest answer to this is “no” or “I don’t know.” To add to the confusion, this rather arbitrary property is often shown at room temperature (23 °C) and also at a range of higher and lower temperatures. Many real life situations will have a complex stress state where a part is under various loads and to give a figure for yield under a uniaxial stress state is meaningless. Furthermore, as soon as repeated stress and relaxation is introduced, that is, fatigue, the data become even more problematic. An amorphous material such as polycarbonate will be relatively poor in fatigue, perhaps failing at below 25% of the yield stress if subjected to a typical test regime of one million cycles. A semicrystalline material such as nylon or acetal, however, will show a high fatigue limit, perhaps retaining half the yield stress figure. In-service vibration, repeated impacts and frequent use of snap fits are all examples of regimes that may be regarded as fatigue.

Impact strength

Many products have to be designed to withstand impact in use. Devices fall from surfaces and hit the floor; units are moved around hospitals and come into contact with hard walls and other rigid structures. Thus, it would seem reasonable when designing a part to look at the impact figures given on the data sheet. The most common tests conducted on plastic specimens are the Charpy and Izod tests,² both notched and unnotched. Unfortunately, the toughness of a particular part is influenced by the material, the design and other factors such as temperature and processing. Attempting to judge how much impact can be tolerated from a data sheet figure and relating this to real life is folly.

The whole area of impact analysis is fraught with danger. It is easy for a novice designer to think that because he or she obeys the rules of using good radii and smooth wall thickness transition that an unnotched figure may be used for calculations. In reality this is not so. It is easy to overlook stress concentrations that occur in moulded parts, for example, through the addition of extra stiffening ribs. Scratches may inadvertently be left on the moulding tool and complex tool shut-offs may occur on curved surfaces leaving less than perfect witness lines. Because these are on the tool, these imperfections will be faithfully reproduced on every part that is moulded and will be able to cause premature failure as a result of impact. In reality it is extremely difficult to determine what part will not be affected by in-use scratches, blemishes and wear that can allow notches to develop.

Thermal properties

Table I: Example of data supplied for unreinforced PBT. (click image to enlarge)

If any medical device that uses polymers has to encounter heat, it is reasonable to evaluate the chosen plastic by investigating thermal properties. It is unfortunate that what is written on the data sheet is confusing. It can typically state: melt temperature, temperature of deflection under load, Vicat softening temperature³ and perhaps relative thermal index (RTI). For a typical engineering polymer these figures may range from 175 °C down to 90 °C, thus it is vital to understand the meaning of each figure quoted. Temperature of deflection under load, or heat deflection temperature (HDT) as it is sometimes written, is a measure of the maximum short term temperature a material can withstand under load. RTI shows the upper use temperature that may safely be tolerated for lengthy exposure. In this latter case, the criterion is the temperature at which, after 60000 hours (seven years), the most sensitive property drops to 50% of its initial value. Normally, only tensile strength is measured. HDT and RTI measure different things and designers must be aware of this and use the data accordingly.

Shrinkage values

The size of a part moulded for the medical industry is of critical importance. Shrinkage values are often quoted on data sheets, but these are so called “global” values, that is, a general figure taking into account many experimental results. Like all the other properties mentioned in this article, shrinkage is affected by a number of factors. There is what may be termed moulding shrinkage, that is, shrinkage that occurs in a moulded part within, for example, 24 hours. It is the difference between the tool cavity size and the measured moulding size and relates to polymer crystallisation. Postmoulding shrinkage, however, is the continuing effect of crystallisation after 24 hours plus the effect of moulded-in stress. These two figures can be affected by processing considerations such as the temperature of the mould, but will also be affected by hold pressure time, another processing parameter. If there is any reinforcement in the polymer such as glass to aid stiffness, then the shrinkage may be anisotropic, that is, higher in one direction than in the other. Finally, colour can also affect shrinkage. Pigments act as nucleating agents and some colours have a bigger effect than others. The practical effect of this is that a change from one colour to another could result in a differently sized part. For example, a change from cream to brown could result in a size difference of more than 1 mm on a 300 mm long moulding.

Relevance of optimum conditions

The final and perhaps most important point is that even if the figures on a material data sheet are relevant to the medical device designer, which in many cases they are not, they are all based on some sort of test bar. The standard test bar is of uniform section with a smooth transition in width. It has no sharp corners. The gate where the polymer is injected is ideally placed to give even flow of the molten material through the test area, which will be under load. The molten polymer is fully homogenous and at the perfect processing temperature, and the cycle time is optimised to give the best mechanical properties. The test environment is benign with fully controlled temperature and humidity.

Ask the right questions

Before relying solely on information on a data sheet, the designer needs to ask: Is my part perfectly designed? Will the processing be perfect? Will the in-use environment affect the properties? In general, the answers will be “no,” “no” and “yes!” In the same way that a book should not be judged by its cover, materials should not be judged entirely by their data sheets.

John Hockey is an independent consultant and founder of John Hockey Associates, Dunstable, UK, tel. +44 1525 222 708 e-mail: jhockey@btconnect.com. Free design and material information/expertise is available on the Omnexus website: www.omnexus.com