Choosing Materials for Medical Devices
Using the right materials can be critical to medical device performance. Dr Sarah Egan of Granta Design outlines technologies that can help designers with the important materials selection process.
When choosing materials for a new product, designers also need to consider the impact of this choice on factors such as biocompatibility, qualification, regulation and cost. Given the significance and complexity of materials selection for devices, best practice in this area has received surprisingly little attention. But this situation is changing and technologies that can save companies years of work are being increasingly used in the industry.
Device designers typically use a range of methods to choose a material. The simplest is "doing what we did before,” that is, take a familiar material and apply it in the new situation, particularly if the application is similar to previous products. They may also ask colleagues, consultants or suppliers for advice. In addition, increasingly, they use resources such as online materials property databases to perform structured searches for materials that meet their specification.
Pros and cons of the traditional approach
These typical approaches have many advantages. They get the job done quickly, which is a critical factor when time-to-market is important. They produce workable results. Furthermore, relying on what has been used before is low risk: regulatory or qualification problems seem less likely and it does not require new expertise.
For many medical device organisations materials innovation is not a core competence compared with, for example, developing the therapeutic or engineering aspects of their products. They do not invest in extensive materials teams. Given these facts and the strict regulatory environment, it is unsurprising that designers are often satisfied with a relatively conservative approach to materials selection.
But this approach has disadvantages. Most obviously, you cannot be sure that you have made the best selection. You cannot have full confidence in your choice. Your product may be sub-optimal in performance or cost, and it may lose out to a more innovative competitor. Also, without a systematic selection process it can be hard to explain and justify a material selection at a later date. This leads to problems if you need to refine product design. In particular, if you must replace the material, for example, due to a supply or regulatory issue, then you have to start from scratch in your analysis. In general, anything that diminishes the "auditability” of design is a negative for medical device manufacturers.
Even using materials databases does not solve these problems. Although databases offer a more quantitative and auditable approach, they often contain incomplete data and focus on searching for materials with specific properties, rather than exploring the combinations of properties that determine materials performance. Designers can miss viable options, or produce results that meet their primary search criteria, but introduce additional problems, for example, toxicity or limitations on the processing route. Database searches often create the illusion of an exhaustive, systematic, auditable approach, without delivering the reality.
So how can we meet these three criteria: an exhaustive search; a systematic, repeatable method; and auditability in a manner that is quick, low risk and viable for organisations without large materials teams?
Methodologies for rational materials selection are not new. Professor Mike Ashby of Cambridge University first published the standard text on the subject, "Material Selection in Mechanical Design,” in 1992. His method divides design requirements into
- Function: What does the component do?
- Constraints: What essential conditions must be met?
- Objectives: What variables then remain to be minimised or maximised?
Typically, there are many constraints and a few objectives. Rational selection takes the "universe” of available materials, screens it based on the constraints, then ranks the remaining materials against the objectives. A formula known as a "performance index” can be used for ranking. This gives the combination of properties to minimise or maximise and thereby optimise a specific objective for a given design scenario; for example, to minimise the volume of a beam in bending for a given stiffness. Indices can be complicated and non-intuitive, but they have been derived for a wide range of standard design situations. In theory, all the designer need do is pick the right index for each of the design objectives and work out which material provides the best trade-off between them.
Imagine that we need to choose a material for the handles of some electrosurgical forceps. The function of the handles is to enable a surgeon to hold and manipulate the forceps. The handles must not deflect too much, therefore, must be relatively stiff. In engineering terms, we could think of them as short beams limited by stiffness. Now we specify some constraints: the material must be biocompatible, sterilisable by steam autoclave, tough so that it does not fracture easily, an electrical insulator, and so on. Then we establish some objectives; we can vary volume and cost somewhat, but prefer both to be small.
Figure 1 shows the performance index for minimum volume for a beam limited by stiffness plotted against the index for minimum cost. Only materials that passed all constraints are shown. The best choices are towards the bottom left of the graph.
Figure 1: A performance index for minimum volume limited by stiffness plotted against the index for minimum cost
The first thing to note is that there is no clear winner. First, the designer must decide how to weight volume versus cost. But the graph makes it easy to see that glass fibre polyarylamide is a good compromise. Second, the designer can see the "landscape” of possible materials and explore alternatives. Third, the rational method has already screened out materials that do not meet the constraints of the application, therefore we can avoid problems due to properties not related to our design objectives.
Finally, consider what would have happened in a more straightforward database search. We would probably have searched for the minimum "cost per kilo” rather than the more subtle "cost per unit of function” studied here. We would have found it hard to construct a search that explores the impact of modifying the handle design, which is what we have done by looking at the impact of varying volume. Thus, the rational approach can result in different conclusions and gives much greater confidence in those conclusions.
Why isn’t everyone doing it?
If the logic is clear, why isn’t everyone doing rational selection? Of course, the approach has some practical challenges:
- You need data on the "universe” of materials
- Data must be normalised so that sensible comparisons can be made
- Data cannot have gaps or materials excluded for no better reason than they have merely been missed
- You need information beyond the traditional "engineering properties” found in databases or handbooks such as price, biocompatibility and sterilisability
- It is complicated to efficiently apply constraints andto study the trade-off between objectives
- How do you pick the right performance index?
The barrier to overcoming these challenges has been high and the motivation to do so has been low, thus traditional approaches to selection have sufficed. But these factors are changing. Rational selection software is now relatively mature. Moreover, medical device organisations face pressures that are driving them to look harder at materials selection or that create more demand to replace materials. These include the need to differentiate themselves in competitive global markets, regulatory requirements that demand greater auditability, disruption to materials supplies and price volatility.
Enabling rational selection
Granta Design is a Cambridge University spin-off founded by Mike Ashby in 1994. In the intervening period materials selection has become a standard element of many university engineering courses. Therefore, an emerging generation of engineers and designers is familiar with the principles. The tools for rational selection and supporting technology have developed substantially in two major areas: data and software tools.
The MaterialUniverse database, developed at Granta and Cambridge University, provides engineering, economic and environmental properties on more than 3,000 materials. It has been designed for selection. Data is comparable and there are no gaps, and unknown values are estimated using proven data models. In recent years, a medical version has been developed that adds properties such as toxicity, biocompatibility and sterilisability. The database covers the full range of commercially available materials types with the properties that device designers need. It includes estimated data and does not describe specific grades, which is the right level of detail for an initial screening that identifies the best classes of material. Designers can then use one of the more specialist databases such as CAMPUS Plastics database or their own in-house data to make final choices.
The second element is software. The company’s CES Selector has a simple user interface that makes it quick for users to specify constraints on materials choice. A graphical Performance Index Finder (Figure 2) lets designers pick a design scenario that is close to their application and they do not need to learn the mathematics. The software produces clear materials property charts that make it easy to visualise and trade-off materials options.
Figure 2: A graphical performance index finder allows designers to select a design scenario that is close to their application
On-line information resources and handbooks still play an important role. Several resources are available providing extensive information on cardiovascular and orthopaedic materials and devices, and on human biological materials (Figure 3), including those from ASM International. These resources help designers to specify their selection problem accurately and to explore in greater depth the candidate materials generated during selection.
Figure 3: Today’s software allows designers to specify their selection
problem and explore in greater depth the candidate materials generated
Re-using materials knowledge
There is another important strand to this story. As stated above, companies can apply in-house data during selection and it is important to remember the value of auditable selection. The problem is that many organisations do not have good sources of in-house data and that, even if they had a rational approach to selection, they would not be able to recover the details of this work at a later date. Medical device companies are often project-focused in their development. Information and knowledge developed in one project may not be captured and made available to future projects. This is particularly true of materials information, where the lack of a large corporate materials team means that there is no natural place for this information to gather. This runs counter to the advice of the US Food and Drug Administration, which encourages "product lifecycle management” principles in device development, believing that knowledge from one product should be applied to future generations.
The medical industry could perhaps learn from aerospace. Aerospace organisations have large materials and process teams and have driven the design of materials information management software that systematically captures and maintains materials information and makes it available enterprise-wide. The results are less time wasted looking for the right data, less duplication of materials tests and better informed materials decisions. This is now a proven technology. Granta’s GRANTA MI software, for example, is used by companies including Rolls-Royce, Boeing and NASA. The good work of the aerospace industry means that other sectors such as medical devices do not need large materials teams or lots of expertise to build effective systems. This software can be implemented "out of the box” for any size of company and is now used in medical device companies.
A new choice
The medical device industry is beginning to successfully apply materials information technology that can help with exhaustive, repeatable and auditable materials selection. It is using software developed in other sectors to help it re-use materials knowledge between projects. Whether these approaches will gain widespread acceptance across the industry remains to be seen. But, with more engineers coming out of universities who have been trained in the key concepts, and forward-looking medical device companies seeking a competitive edge, the chances are high that materials selection will be getting more attention.
Dr Sarah Egan is Product Manager Medical at Granta Design Ltd, Rustat House,
62 Clifton Road Cambridge CB1 7EG, UK,
tel. +44 (0)1223 218 000,
If you found this article interesting please
Sign Up Now for your free subscription to