- 28 February 2007 -

Engineering plastics in hot water applications

If your system runs hot (water, that is) then plastic components may be the best choice for long-term performance. Frank Heessels, global technology manager at GE Plastics explains the science that goes into creating materials that can take the heat.

Over the years, parts made of engineering thermoplastics have been successfully shown to operate in many different environments where strength, stiffness, temperature stability, chemical resistance, longevity and complex designs are required.

New developments in hot water fluid engineering applications confront the product developer with a challenging combination of all of the above-mentioned requirements. The need for very long-term performance – up to 15 years or more – of material exposed to hot water and pressure means that standard material datasheet properties will usually not be comprehensive for application development, as they are typically based on shorter-term data. Relevant material data is scarce and the product needs to be engineered and tested with care. This article describes various conditions in fluid engineering and some approaches to them from a materials perspective.

Design

The design of the component will significantly influence the final performance of the part. The design freedom that plastics offer greatly enhances the designer's ability to integrate multiple components and functions in one part. Parts used for fluid engineering tend to be complex in shape, and often tubes as well as holes will be integrated in the part. Consequently, it will be difficult to avoid knitlines in stressed areas.

If a fibre-reinforced material is used to mould the part, the fibres may not align with the direction of the highest principal stress. In order to get the most out of the material it is important to understand these effects, which will often require structural and mould-filling analysis using computer-aided engineering.

It is very important to control the moulding of these parts with sufficient holding pressure to achieve the best possible material properties. Note that the tensile dumbbell sample used for determination of standard material properties is actually a very simple part to mould, and has strong fibre alignment in the pulling direction, which results in relatively good performance compared to the more complex actual shapes. For these reasons, the results from testing the dumbbell may not provide accurate measurements of the actual component's properties.

The following material properties usually need to be addressed in the mechanical design of fluid engineering components:

Static strength

Often a high initial static strength is required to withstand abusive assembly by an installer who is used to metal components. If this abusive assembly happens at the beginning of the life of a component, then traditional engineering design is often used, using tensile properties at the required temperature. However, the component stresses that result from assembly may not be present during the whole lifetime of the part. If the design allows for it, these stresses may relax, especially at elevated temperatures (see below, under “long-term creep”). If abusive forces are too high, the design flexibility of plastics often allows a stronger design, or one may consider changing to a more plastic-friendly assembly method.

A further requirement for high initial static strength could be to anticipate long-term performance requirements that cannot be adequately simulated. By applying a margin of safety, the manufacturer can help ensure that long-term stress does not lead to failure. However, for some materials, long-term material strength may differ significantly from initial properties, meaning even high safety factors may not be adequate. It is important to always test the final component under conditions representative of actual usage.

Long-term creep displacement and strength

Many fluid applications are loaded with a constant internal pressure, which translates to a constant stress in the part. As a result, thermoplastic materials will slowly deform (creep). If this deformation removes the cause of the stress or allows the stress to redistribute to stiffer parts of the structure (e.g., when assembly causes a local deformation) then the deformation rate and stress will decrease (stress relaxation).

Stresses caused by a constant internal pressure in a fluid component are often unable to relax and will continue to cause creep over the lifetime of the part. If these stresses are relatively low, the deformation rate will decrease over time. However, if stresses are too high the deformation may grow to a critical level and ultimately cause spontaneous failure of the part, a phenomenon called creep rupture. Fibre-reinforced resins, which are often used to improve the creep performance of a material, can be effective given the remarks made in the paragraphs about design and water resistance.

Creep curves showing the relationship between applied stress and resulting deformation of tensile bars are available for many resins. Creep rupture data for the maximum allowable long-term constant stress in a material is much rarer. A very elaborate but useful method is to generate data according to ISO 9080, a standard that includes a methodology to extrapolate data generated to a 50-year lifetime.

Fatigue

Fluctuating stresses cause a very different material behaviour. These stresses result from the sudden impact loads that occur when a water tap is quickly closed, causing fluctuating pressure in the water (“water hammer”). Another source of fluctuating stresses is the difference in thermal expansion caused by temperature changes on dissimilar materials connected together, or a temperature distribution within one part.

The cumulative damage mechanism that causes fatigue failure is not easy to characterize. Design-related conditions such as type of stress and material thickness affect fatigue performance; for example, thinner walls usually result in a longer fatigue life for the same stress. Consequently, available data like Wöhler curves on tensile bars can at best only be used to compare materials. The actual fatigue performance needs to be tested using the final application.

Water and chemical resistance

General rules for the resistance of a particular plastic against chemicals are difficult to formulate. As with many other conditions, it is ultimately the length of exposure to the chemical that will define whether the component will have sufficient resistance. A higher temperature and continuous stress acting on the part accelerate the effect of a chemical on material strength.

Water acts as a chemical in this sense and actually there exist many plastics – for example, PA, PBT and PC – that degrade in hot water. Again, the temperature and exposure time will define whether a material is suitable for an application or not. Some semi-crystalline resins like PA and PPS have better resistance against many chemicals than amorphous resins like PPE and PS. However, GE's Noryl PPO* resin provides hot water resistance that is actually better than that of many amorphous resins with otherwise similar performance. Further note that glass fibre-reinforced resins may show some strength reduction in hot water because of degradation of the interface between the resin and the glass fibre.

Basically, chemical resistance needs to be tested for every possible combination of plastic and chemical. Check with the manufacturer of your plastic components and see what data they've collected on the chemicals in your process.

Testing

Because of the many interactions between environmental conditions and the design of many fluid applications, it is important to verify the actual performance of the final part with a suitable testing program. A challenge with many fluid applications is the need to represent a lifetime using a test that takes much less time. The content of a testing program will depend on the actual requirements, but from a material perspective there are a couple of considerations:

• Testing at elevated temperature can often accelerate the effect of chemical (and water) exposure, and a constant load or pressure. For example, if a part is tested at operating temperature and two temperatures above that (e.g., 10°C and 20°C above operating temperature) and a linear relation exists between the temperature and the time until a performance criterion (e.g., failure or burst pressure strength), then the relationship between time and temperature may be used to extrapolate the data generated at elevated temperature to a longer lifetime. It is generally good practice to verify this approach by comparing the test predictions with the actual performance under operating conditions.

• In the case of fluctuating stresses, increasing the temperature is generally not a good choice because many materials are more ductile at elevated temperatures, which may significantly reduce the speed of the damage mechanism. In this case, it is often a good alternative to simply increase the frequency of pressure changes to represent a longer lifetime.

www.geplastics.com

* Noryl, Verton, Ultem, Staramide, Noryl PPO, LNP, and Lexan are trademarks of General Electric Company.

 

 

 

 

To receive news and features like this direct to your inbox sign-up for the World Pumps enewsletter.

Simply register your details to receive a fortnightly roundup of the latest news from the pump and sealing industries direct to your inbox.

If you would like to advertise in the World Pumps enewsletter please contact our sales team.

If you have some news for the pump industry or would like to comment on any of the articles on this site, contact our editorial team.

You can also access the full list of contacts here.