Medical Device Link.

DESIGN

When a development team takes an idea from a need identified by the medical community to an effective product, a significant amount of testing is required. Anyone who has been involved in this process knows that testing takes time and money and can reveal problems with a design near the end of the development process when making changes is difficult and expensive.

Design teams are integrating the use of computer simulation into the development process to reduce the schedule impact and cost of physical testing. Simulation, also called analysis, virtual prototyping or virtual testing, uses computer software to create a virtual representation of not only the object being designed, but also the loads and constraints that act upon the object during real-world use. By breaking the real-world situation into a set of equations, the software can solve the problem quickly before physical hardware is made, thus making a significant impact on the design process earlier and for less cost then testing. However, as with any tool, proper implementation is critical to obtaining success. Experience with using analysis driven design (ADD) as a foundation for product development has led to the establishment of five simple rules for its effective use.

Rule 1: Choose the right tool and learn it

Figure 1: Deflection analysis of an endoscopic actuation joint. The assembly was modelled with contact to get accurate loading rather then applying an approximate load.

The success or failure of ADD is often driven by the capabilities of the software chosen to do the analysis, and the ability of the users to utilise the tool in a timely and accurate manner. Many organisations have chosen tools based on their initial ease of use or short learning curves and ignore the need for a general-purpose tool that allows its user to simulate the real world more closely. A recommended approach is to use a general-purpose tool that provides strong associative computer-aided design (CAD) integration and offers the ability to set the level of simulation to the appropriate level (see rule 3 below) and the breadth of capability to model most real-world situations that a medical device may experience. By making the investment in a more expensive solution, and taking the time to learn this tool, almost any problem can be addressed without struggling to make the tool fit the problem.

Rule 2: Use analysis to answer questions

As with physical testing, it is critical that anyone who does virtual testing begins by deciding what questions need to be answered. Too often the analyst focusses on doing a certain type of simulation and loses sight of the fact that he/she is trying to answer a question posed by the design team. The goal of this rule is to avoid the natural tendency to want to do a certain type of analysis and to focus on the analysis methods. Much more benefit can be derived by first establishing what design questions need to be answered and then deciding what type of simulation is best to answer those questions, followed by a development of the methods needed to obtain the proper answers.

Rule 3: Use the proper level of simulation

The tendency in modern finite element simulation is to take a brute force approach to creating virtual models that represent a level of detail that is not needed to answer the questions the engineer is asking. Modern software, especially that built into CAD systems, allows the engineer to quickly convert a three-dimensional design model into a simulation model with little effort. The problem with this is that the resulting model is often too large or does not have the required fidelity
to represent the virtual test being conducted correctly. Engineers need to know where and how to simplify the analysis model and set it to the greatest efficiency for the questions to be answered by the analysis.

Rule 4: Model reality

Figure 2: The design team needed to know which endoscopic device joint was the strongest. Simulation was used to try a variety of designs and to select the best shape in a matter of hours without physical testing or any significant assumptions.

An easy mistake to make in conducting analysis as part of a development process is to assume away important aspects of the model or to use a tool that cannot model those aspects. When using simulation to answer design questions, an analyst must decide what aspects of the virtual representation are important to answering the question. If these are left out, then the answer given is really the solution to a different problem than to the issue of what the device will experience in the real world. This can lead to expensive surprises when physical testing is conducted.

Rule 5: Interpret the results objectively

The final rule deals with how the results of analysis are used in the design process. Finite element analysis is good at producing a massive amount of numerical information about the behaviour of a physical object. It is, therefore, easy to find numbers that justify a predecided conclusion. A good design team will evaluate the results objectively and interpret them to find the best answer to the questions posed. This often involves bringing in an outside party who is unfamiliar with the analysis to look at the results and offer his/her opinion.

Successful deployment

Medical device development poses some unique challenges for the design team because of the harsh operating environment, size constraints and material restrictions that are inherent to the application of these products. This makes the cost of physical testing even higher and the importance of answering design questions earlier in the process more important. By effectively deploying the right tool, in the right manner, for the right problems, the design team can make ADD an integral and beneficial part of the design process. This will reduce the time and money spent on physical testing and get the product to market sooner and with better performance.

Eric Miller is Director of Analysis & Design Technologies at Phoenix Analysis and Design Technologies Inc., 7755 S. Research Drive, Suite 110, Tempe, Arizona 85284-1803, USA, tel.+1 480 813 4884, e-mail: eric.miller@padtinc.com, www.padtinc.com.