Fluidic System Design Is No Easy Task

A number of factors must be considered when designing a system containing pumps, valves, and connectors.

William Leventon

March 1, 2007

14 Min Read
Fluidic System Design Is No Easy Task


The wide choice of pumps available makes it difficult for designers to choose the appropriate one.

In hospitals, labs, and factories, many types of equipment depend on fluid flow. And that critical flow depends on the design of fluid-handling systems. Designers of these systems must answer many tough questions. For example, how should a system be configured? What components should be used? And what companies should supply them?

For help in answering these and other questions, read on to review advice from experts on pumps, valves, and connectors. They explain how you can develop the best and most cost-effective fluid-handling systems for medical devices, diagnostic equipment, and production machinery.

Component Considerations

A key part of any design process is the selection of components. To help them make the right choices, designers should draw a diagram of the work flow they want to automate with their system, suggests Mark Coffey, product manager for Tecan Systems Inc. (San Jose). The company sells pumps and valves for diagnostic equipment.

A diagram of the flow path is helpful in deciding what valve configuration—T, bypass, distribution, etc.—is right for a system. A plumbing diagram also helps designers determine whether they need single- or multichannel pumps. In addition, it can help them figure out how many pumps and valves a system requires, according to Tor Bjornson, Tecan's director of engineering.

When selecting pumps and valves, designers must look at the materials used to make the so-called wetted parts. These parts must be made of materials that are compatible with both the system fluid and any fluids that might be used to clean the system. Once compatible materials are specified for a component, be wary when suppliers suggest changes to these materials, warns Mark Connell, staff engineer for Micropump Inc. (Vancouver, WA), which sells miniature pumps for medical devices.

“You may have a supplier who says, ‘we have to make a very minor modification to a material. Our chemists have looked at it, and they're sure there will be no problems,'” Connell says. “Then, later on, you find that the parts are swelling when exposed to a certain disinfectant. And that swelling could cause a failure.”

Designers must consider system requirements when selecting pumps and valves. Other factors include component types and the compatibility between the component material and the system fluid. Products shown here are from (from top to bottom): Micropump Inc. (Vancouver, WA), KNF Neuberger Inc. (Trenton, NJ), and Hargraves Technology Corp. (Mooresville, NC).

In addition to material and fluid compatibility, component selection depends on system characteristics such as flow, pressure, and power. Besides determining the characteristics of their systems, designers should rank those characteristics in order of importance, advises Mark Garland. He is product manager for Parker Hannifin's Pneutronics Division (Hollis, NH), which supplies miniature pumps and valves to medical device manufacturers.

A ranking system helps designers make critical trade-offs when selecting components, Garland says. For example, a valve with a soft elastomer seal will provide good leak integrity but may also wear faster than a seal made of a harder material. By ranking characteristics by their importance to a particular application, a designer has a guide in deciding how much of one to sacrifice to achieve another.

Margin of Error

Garland also recommends adding what he calls a “margin of error” to the system requirements. Doing so can help ensure that pumps and valves can handle their responsibilities in the field. On the other hand, Connell notes, the margin of error, or safety factor, can't be too large, because “there's a cost penalty when you buy components with more capability than you really need.”

There's no industry standard that can be used to determine the right safety factor for an application, says Connell. Rather, he explains, the size of the safety factor should depend in large part on how well the designer understands the application.

The first step is to have a very good understanding of how all the parameters interact and what the worst-case situation would be, he says. “Then you won't need nearly as big a factor of safety as you would if those things were less well known.”

Relatively large safety factors may be assigned to medical device components that are critical to patient safety. So before selecting parts, a designer should perform a risk analysis that breaks the system down into individual components. It is then possible to see whether failure of any of them would adversely affect patients, advises Joel Bartholomew, product development manager for the OEM division of B. Braun Medical Inc. The Bethlehem, PA–based company supplies valves to medical device firms.

A valve deemed critical to patient safety may have the same performance specifications as a noncritical component. But the device manufacturer may require that 100% of critical components be inspected by the supplier. Such a requirement could make a critical valve significantly more expensive than noncritical valves with the same specifications, Bartholomew notes.

Bartholomew recommends that before designers select a supplier for critical components, they conduct audits of potential vendors. These audits should include evaluations of their manufacturing and quality systems. “Get to the sites of these vendors and look at what they're doing and how they're doing it,” he says.

Different Component Types

Liquid pumps, such as these by KNF Neuberger, can be used to transport fluid from one point to another in a medical system.

Besides determining the requirements for the components in their systems, designers must often decide between different component types. For example, the fluid in a system could be moved from one point to another with a liquid-handling pump. Or a designer could use a vacuum created by a gas-handling vacuum pump to pull the liquid through the system, notes David Vanderbeck. He is business development manager for KNF Neuberger Inc. (Trenton, NJ), a supplier of pumps for medical devices.

“The nice thing about the vacuum pump is that it's never in contact with the fluid,” Vanderbeck explains. “So it might be a good option if you're using a fluid that's aggressive.”

Like pumps, valves come in many different types that must be evaluated by designers. One option is the common check valve. But when check valves operate in fluid-handling systems, a “dead volume” of fluid often remains in the valve housing, says Chris Powers, OEM product support manager for Hamilton Co. The company, based in Reno, NV, sells valves for clinical diagnostic applications.

Instead of a check valve, Powers says, designers should consider a plug valve featuring a rotating mechanism that seals the port. The operation of a rotating plug valve reduces fluid-system problems such as dead volume and hysteresis, he notes.

Valve selection is certainly important for designers of many medical and diagnostic devices. However, it is also important for those designing adhesive-dispensing systems used in medical device manufacturing. For example, a piston valve made of materials that are compatible with a certain UV-cure adhesive may open and close in a way that causes turbulence in the adhesive, resulting in the formation of microbubbles in the fluid.

On the other hand, the operation of the same piston valve may not cause bubbles to form in other adhesives. So designers must consider both operation and material compatibility when choosing valves for dispensing systems, explains Claude Bergeron, business development manager for EFD Inc. The East Providence, RI–based company sells valves for dispensing applications.

Connection Issues

System designers must choose connectors carefully. For frequent connection and disconnection, quick-connect or luer fittings, such as these by Value Plastics, may be a good choice.

Valves and pumps aren't the only components that must be chosen when designing fluid-handling systems. Designers must also select connectors that meet system requirements. For example, some applications severely limit the size of the connectors and other components. In cases like this, notes Jim Pisula, designers could choose relatively small connectors made of strong materials that can handle high system pressures. Pisula is vice president of marketing for connector maker Value Plastics Inc. (Fort Collins, CO).

System designers consider several other factors besides size before choosing connectors, he says. For one thing, how often will the system be connected and disconnected? Applications requiring frequent connection and disconnection may call for a quick-connect device or luer fitting. On the other hand, a tube fitting might be right for a system that will be connected just once.

Another important consideration is the potential for misconnection with other equipment in the same environment. For example, Pisula notes, designers of patient-monitoring devices might want to use a connector that cannot be hooked up to IV systems, because the two are often used in the same place and misconnections can be “catastrophic.”

If patients will be using a connector, system designers may also want to consider a product that's easy to understand and use. “Some connectors are very competent technically, but they're a pain to use,” Pisula says. From a marketing standpoint, he says, a better choice might be a slightly more expensive product that makes connection easier for the end-user.

Other Features

Good fluidic system design depends on more than just pumps, valves, and connectors. Other features can improve the performance of a system and its key components. In dispensing operations, changes in system parameters, such as adhesive viscosity and ambient temperature, might require corresponding changes in deposit size. Many times, plant personnel shut down the dispensing system while determining the correct deposit size and then adjust system variables to make the desired change in the deposit amount. Since the system is not producing parts while this is going on, these shutdowns are costly to device makers.

EFD's dispensing valve controllers can work with PLCs to automate the dispensing process.

According to EFD's Bergeron, a better way to change deposit sizes is to equip systems with dispensing valve controllers that work with programmable logic controllers (PLCs) or computers operating an automated dispensing process. These controllers can make required deposit-size changes in seconds, without shutting down dispensing systems. In addition, he says, they can make very fine adjustments to valve pulse time. Such adjustments result in more-precise deposit changes than those produced by common mechanical adjustments.

Other fluid-system features focus on the operation of system pumps. For example, sensors can be combined with a brushless dc motor and a PLC to control pump speed. Working together, these components determine the optimal pump speed at any given time. They also keep the pump running fast enough to meet the needs of the host device, but no faster. Pumps that slow down whenever possible last longer than pumps constantly running at excessive speeds, Vanderbeck notes.

Like pump life, the life span of valves depends on the amount of work done by the components. So Hamilton's Powers recommends that system designers reduce the movement of valves to extend their life. One way to do this is to design fluid-handling systems in ways that minimize valve rotation. For example, he says, say a designer is considering a system in which one valve rotation draws in 500 µl of fluid and another dispenses 500 µl. Instead, the designer could create a system that draws in 5 ml of fluid and then dispenses ten 500-µl amounts. In the second system, one dispensing rotation of the valve takes the place of 10 valve rotations in the first system.

As for component placement, Powers recommends that designers position valves as close as possible to both the reservoir and the fluid destination. The shorter the tubing lines running between valves and other points in the system, he says, the less hysteresis and lost fluid volume in the lines. Put together, those factors reduce the chances of inaccurate dispensing volumes.

Although short tubing lengths improve the accuracy and performance of fluid-handling systems, other design considerations may affect the placement of components and, therefore, the length of the lines running between them. For instance, Bjornson says, designers may want a pump in a certain location to improve user convenience or the appearance of an instrument. So trade-offs are often made that sacrifice pump or valve performance to meet some other design objective.

In the design of adhesive-dispensing systems, a key objective is making it as easy as possible for the dispensing tip to reach the target area of a medical device. Sometimes, a simple system change can prevent major dispensing headaches. For example, Bergeron says, imagine that the shape of a two-component subassembly makes it difficult to get the dispensing tip to the target area of a device. Manufacturing engineers may be able to move the dispensing step up in the production process so that it's done before the subassembly is created.

In some cases, B. Braun will recommend that designers change their tubing sizes so their systems can use standard valves. The recommended changes can be as little as 0.005–0.010 in. in diameter. “If they choose not to do that, then we have to use custom molds, as well as single-station assembly and other low-volume assembly techniques” in some cases, Bartholomew says. “And that could easily double or triple the cost of the valves compared with those we make in large volumes.”

Ready-Made versus Custom Parts

In some cases, suppliers may be able to solve one designer's problem with a feature developed for another customer, notes Bruce Johnson, Micropump's vice president of sales. Because they're ready-made, such components can usually be obtained for a lower cost than custom designs. However, Johnson adds, they can only be attained by companies that aren't a direct competitor of the firm for which the component was originally developed.

Suppliers that are involved with a customer's development process early on may be able to steer the designer toward standard components and away from custom versions. “If you design a tubing port that doesn't fit the tubing port of an existing valve, someone will have to modify a component for you,” Bartholomew says. “That means you'll have to spend money on things like a mold, validation, and qualification testing. And it's going to take you a lot longer to get your product to market.”

Of course, custom products are sometimes what designers really need. In recent years, Idex Health & Science (Pocasset, MA) has been producing custom manifolds that include pumps and valves as well as different types of sensors. Tested before shipping, these subassemblies prevent problems that often occur when designers try to integrate the individual components themselves, explains Brad Besse, IVD market manager for Idex. The manifolds also help designers reduce R&D costs and speed up product development, he adds.

According to Vanderbeck, customized products account for more than 80% of KNF Neuberger's business. “We might start with something a customer has seen in a catalog or on our Web site,” he says. “Then we modify that product to meet the application requirements.”

Whether a component is custom or standard, delivery to the customer shouldn't mark the end of the designer-supplier relationship, according to Vanderbeck. Rather, he says, designer and supplier should engage in an ongoing dialogue that lasts throughout the process of developing a fluid-handling system. The reason: Pumps that are well suited for an initial design may not be well suited for the final design, resulting in inadequate performance and short pump life.

“During the design process, there are many changes to the designer's system,” Vanderbeck explains. “We need to understand those changes and adapt our [components] to the changing requirements.”


Selecting pumps and valves for fluid-handling systems brings a host of design considerations to the forefront. For example, it's not enough to consider only component types. Designers must also consider the materials used in those parts, and whether there will be any undesireable interaction between those materials and the fluids in the system. It's also important to examine different component types to find the one that best suits the system requirements. Connections, dispensing systems, and other features can be important factors as well. The good news is that the suppliers of all of these components can often help address design problems early and competently.

Copyright ©2007 Medical Device & Diagnostic Industry

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

You May Also Like