Going with the Flow: Streamlining Design and Manufacturing through Computational Fluid Dynamics

Medical Device & Diagnostic Industry MagazineMDDI Article IndexOriginally Published May 2000FLOW CONTROLComputer simulations based on advanced mathematical models are helping product developers improve fluid-flow devices.

May 1, 2000

12 Min Read
Going with the Flow: Streamlining Design and Manufacturing through Computational Fluid Dynamics

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally Published May 2000


Computer simulations based on advanced mathematical models are helping product developers improve fluid-flow devices.

Now nearly 30 years old, the discipline of computational fluid dynamics (CFD) is concerned with the development and application of computational procedures designed to simulate and predict fluid flow around or inside structures of arbitrary shape. By taking advantage of sophisticated mathematical models and high-speed computing, CFD simulations enable the complex effects of fluid forces to be calculated and the resulting data can then be used for practical engineering purposes.

0005d26a.jpgFigure 1. CFD was used by Medtronic Inc. to optimize the design of a pump for bypass procedures>

CFD analysis is streamlining the design and manufacture of medical devices by providing a better understanding of the fluid-flow and heat/mass-transport phenomena that affect device performance and manufacturing processes. During device design, CFD provides detailed performance assessments, reducing the need for costly experimentation. Using CFD analysis, manufacturing processes can be optimized from both a quality and cost standpoint. CFD is also essential for simulating human physiological flows that interact with devices or affect drug delivery procedures, since such flows are difficult to measure using animal or physical models.

Many biomedical applications involve fluid flow and heat/mass transport in the body and in devices. Some examples include aerosol drug delivery, blood pumps, artificial heart valves, blood oxygenators, filtration devices, needles and catheters, tubing, and diagnostic equipment. Transport processes can include the effects of electrical fields, osmosis, multiple phases, deposition of particles, deformation of solid regions surrounding fluids, and chemical reactions. CFD analysis offers details of fluid velocities, pressures, solute or particle concentrations, temperatures, stresses, and heat/mass fluxes throughout the flow domain.

Computed flow parameters can be displayed in different formats—including color-coded graphics—which helps provide insight into physical mechanisms affecting the operation of a particular device. As part of an analysis, engineers can easily alter model geometry, boundary conditions, or material properties to determine the effects on the system under study. As a result, CFD is well suited for conducting parametric studies, making it possible to evaluate far more design alternatives than with a build-and-test method, thereby allowing for faster performance optimization and significant reduction of design cycle time.


While experimentation using physical and animal models will continue to play a paramount role in the testing of medical device designs, it has some important disadvantages that explain the increasing emphasis many device manufacturers are placing on computer simulation. Experiments take a long period of time to perform, are expensive, and, in certain cases, may involve risks to animal or even human subjects. For these reasons, device manufacturers are turning to computer simulation for evaluating the relative performance of various design alternatives and ensuring that only the best ones reach the market.

0005d26b.jpgFigure 2. Flow simulation in a spinal needle revealed potentially harmful differences in anesthetic distribution.

Another problem with physical testing concerns the limited quantity and quality of the data that are generated, which are obtained only at those locations where measurements can be made. Computer simulations, on the other hand, can provide as many calculations of as many relevant parameters, and in as many locations, as the analyst requires. Numerical simulation also eliminates the drawback of data scatter caused by difficulties in maintaining uniform experimental conditions.



An increasing number of firms are employing CFD as awareness of the technology and its potential benefits spreads among device manufacturers.

Pump Flow Characteristics. One current example of the applicability of CFD to medical product development is its use by Medtronic Inc. (Minneapolis) in the design of blood-handling devices. The company has implemented CFD software as an integral part of the design-optimization process for blood pumps and oxygenators, which assume the role of the heart and lungs, respectively, in open-heart bypass surgery.

A design objective for such devices is to minimize thermal and mechanical stresses that can damage blood cells in the artificial circulation. A type of blood pump manufactured by Medtronic, the BioMedicus Biopump, comprises three rotating cones that produce a centrifugal effect for drawing blood from an inlet port. Optimization of the pump's performance required prediction of the magnitude and distribution of pressure, temperature, and shear stress fields, and of the residence time of fragile blood cells within the pump cavity.

0005d26c.jpgFigure 3. Using CFD to simulate blood flow within aneurysms helps predict their growth and the danger of rupture.

Flow in the pump was simulated using a geometrically accurate computational domain to determine the critical flow parameters (Figure 1). Using CFD, the company was able to determine and evaluate the effects of flow characteristics they had been unable to explain using standard test methods—including a complex recirculation pattern between adjacent rotating cones.

According to Medtronic, pressure distributions predicted by CFD matched experimental predictions quite well. Within six months of purchasing CFD technology, the hardware and software expenses were recouped through improvements in product design efficiency and a reduction in time to market for new product designs. The ability to make changes late in the design cycle based on CFD results was particularly valuable, since building and physical testing of prototypes at this stage is very expensive. Medtronic has continued to use CFD simulations for design optimization of other medical devices.

Catheter Design. In another application, researchers in the Office of Science and Technology (OST) at FDA's Center for Devices and Radiological Health (CDRH) used CFD to investigate the role of spinal-anesthesia catheter design in some previously mysterious neurological injuries. In spinal anesthesia, a catheter or needle is used to deliver anesthetic directly into a cavity within the spinal column. Several patients had suffered cauda equina syndrome, a form of temporary paralysis, when they were anesthetized using certain types of spinal catheters.

It was postulated that the injuries were due to a maldistribution of anesthetic that resulted in an accumulation of toxins in isolated locations within the spinal column. Studies of anesthetic dispersion are typically performed using plastic models of the spinal column in conjunction with flow-visualization methods. Because these analyses are limited in the amount of information they can provide, CDRH decided to investigate the problem using CFD.

The CFD computer simulation clarified for the first time some of the factors affecting anesthetic distribution in the spinal column (Figure 2). Graphical results from the analysis clearly demonstrated that smaller catheters (0.02-cm diam) tended to produce an uneven distribution of anesthetic compared with larger ones (0.06-cm diam). A jet of anesthetic solution coming out of the smaller catheters tended to be confined to a tightly circumscribed region of the spinal column, where it could form a pool and result in neurological injury. Larger catheters, however, produced a more widely dispersed jet that also tended to be more evenly distributed in the spinal space. The CFD analysis allowed researchers to systematically study the effects of various system geometries and operating conditions on performance. This technique can be used by FDA to evaluate similar devices submitted for approval or market clearance.

Tubing Processing. CFD can also help improve manufacturing processes for devices. For example, Vygon SA (Ecouen, France) has used CFD analysis to optimize manufacturing of medical-grade tubing. There are several factors that complicate the design and manufacture of flexible tubing products. Often, the tubes must transfer multiple fluids, each flowing in a separate lumen whose cross-sectional area controls the maximum possible flow rate. Given these constraints, flexibility and small tube size are typical goals, and the manufacturing process must be designed to yield very specific cross-sectional shapes. Such tubes are generally extruded, and the process mandates a precise die design that can reliably deliver the required shape.

Vygon used numerical flow-simulation methods to evaluate several die designs for multilumen tubes, factoring in precise control of the internal lumen cross sections. Based on imposed flow profiles and operating conditions, CFD computed the required die shape. The dies were then built and tested by Vygon, and an extrudate shape very close to the required one was obtained on the first trial. The resulting dies are more robust and stable to handle than those used previously, and they yield more reproducible products that require negligible fine-tuning. The outcome has been increased production rates and a reduction in the number of extruders needed in the company's facilities.

Physiological-Flow Modeling. Another important area for CFD is in the modeling of physiological flows, with applications in toxicology, drug delivery, and surgical and diagnostic procedures. On a more fundamental level, modeling of fluid transport in organ systems can lead to a better understanding of physical mechanisms that contribute to the development of certain diseases. As an example, researchers at Thomas Jefferson University (Philadelphia) have modeled blood flow in arteries to predict the growth and rupture risk of cerebral aneurysms.

A result of weakness in the arterial wall that manifests itself as a balloon-shaped bulge, aneurysms are believed to be caused by abnormally large flow shear stresses that can damage wall cells. Once an aneurysm is formed, fluctuations in blood flow within it can induce vibrations of the aneurysm wall that contribute to progression and eventual rupture. Simulation of blood flow with CFD enables determination of arterial wall stress levels, which in turn helps identify where the artery may be susceptible to aneurysm formation, growth, and rupture (Figure 3). As CFD analysis is further developed into a practical diagnostic tool, it is expected to dramatically improve the ability of physicians to weigh the benefits of various treatment options.


CFD is emerging as a valuable tool for medical device makers, both for improving design and manufacturing processes and for simulating physiological flows, including ones that interact with devices. Flow modeling offers engineers the ability to accurately determine the performance of design concepts, reducing the need for physical testing and building of prototypes. This makes it possible to evaluate many more designs, potentially resulting in substantial enhancements in product performance. At the same time, relatively quick and inexpensive simulation can provide faster time to market and minimize development costs.

Keyvan Keyhani is a member of the biomedical applications team at Fluent Inc. (Evanston, IL), which provides design and simulation software used to predict fluid flow, heat and mass transfer, chemical reactions, and related phenomena. Rupak Banerjee is staff scientist in the bioengineering and physical science program at the National Institutes of Health (Bethesda, MD). Shoreh Hajiloo is a senior consulting engineer at ICEM CFD Engineering (Berkeley, CA), which develops and markets software for pre- and postprocessing of applications such as computational fluid dynamics and structural analysis.

Return to the MDDI May table of contents | Return to the MDDI home page

Copyright ©2000 Medical Device & Diagnostic Industry

Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like