06 Jul The Bubble Challenge Part II : Minimizing the Risk of Bubbles
Here’s a truism about medical devices: It’s impossible to design an application that handles fluids and expect not to have to address the trouble with bubbles. They’re a bit like taxes: They’re unavoidable, but with advance planning you can minimize the impact.
This is the second post in a two-part series, and the more uplifting of the two. In this one, we’ll be strategizing solutions.
In the first part, we saw why bubbles pose a big risk. For devices that deliver medicine directly into a patient, bubbles can be deadly if they end up in the bloodstream. For machines that pump, wick, or pipette specific volumes of fluids, bubbles can sabotage precise measurements. Bubbles show up in tubing, at junctures, in agitated containers; they can move, change size, and end up in inconvenient corners of your device’s geometry.
But there’s hope! Bubbles may be preventable. And they can usually be detected. Here, we’ve put together a few big-picture solutions that can help you outsmart the menace of bubbles in device design.
PREVENT BUBBLES IN THE FIRST PLACE
Geometry. For most applications, it’s best to design a fluid path geometry that is easy to prime and discourages bubble formation in the first place or at least allows bubbles to break free from surfaces and be flushed out before they inflict any damage. Choose configurations where kinks or bends in the tubes are unlikely to develop, and avoid sudden expansions in cross section shape and ‘dead volumes’ that can’t easily be fully wetted out.
Junctions. Pay attention to how you design joints and seals between fluidic components, and how you intend to secure or bond them together. Subtle changes in the fluid path near a junction can have a large impact on bubble behavior, so it is important to design these interfaces well. Reduce gaps between joining parts where air can get trapped; or design the bond so the adhesive seamlessly fills the gap. Consider lengthening remaining gaps to creating a smooth transition so fluid can fill the gap area.
Materials. Choose wisely: Select materials and surface finishes that help to prevent bubbles getting stuck along the way. For macro-scale systems this means smooth ID plastic tubing and parts with polished surfaces compatible with your fluids. For microfluidic systems, contact angle properties become very important, and can be influenced by a variety of factors like surface finish and treatment, sterilization method, part aging, and storage conditions.
We designed a device recently that needed to produce repeatable droplets of liquid, but found that drops were forming irregularly on the tip, and creating lots of droplet size variability. Video of the process revealed that droplets were not wetting the outside surfaces of the dropper tips uniformly due to surface finish variability and contamination buildup over time. A new material was chosen for the dropper tip which reduced the sensitivity to these surface irregularities, allowing for complete droplets to form reliably.
Remember your user. Users can unintentionally introduce bubbles while priming a device, changing tubes, exchanging cartridges, or inserting needles. First, optimize the device and workflow to promote bubble-minimizing fluid handling behaviors. Then, where possible, provide the user with immediate positive feedback that they have handled the fluid correctly to reinforce those behaviors. Clear materials can allow the user to inspect for bubbles in real time. Or automated devices can take over the handling and inspection of fluids following user interactions by way of embedded sensors with smart algorithms, assuring the user their behaviors are ‘bubble-free”.
CHOOSE THE RIGHT SENSOR(S)
You can’t always prevent bubbles, so when they happen, how to find them? For the contamination-sensitive medical industry, the market offers plenty of non-contact sensors. These employ ultrasound or optics which work well for gross air in line detection, but they can be difficult to tune for applications which require higher sensitivities. There will always be a balancing act between sensitivity to the presence of air and the occurrence of false alarms, particularly when there may be variability in the interface between the disposable and the sensor. Upscale bubble sensing methods use machine vision, requiring proper illumination to image the areas of interest and then processing to “see” bubbles. Match your sensor to the specific needs of your application.
RE-EVALUATE: CAN YOU TOLERATE THE BUBBLES YOU GET?
Sometimes, the design team will have a better idea of the bubbles it can’t allow than the bubbles that actually come about in an application. In trying to answer the question, “How serious are these bubbles?”, you need to be able to define “these bubbles”. Installing bubble detection on early prototypes and then running parametric bubbling tests can help to characterize the application-specific bubble field early in development. Once the team develops an understanding of what to expect in bubble size, location, and prevalence, the team can then evaluate how serious the mitigation needs to be. This may result in redesign of the fluidic path(s), or reassessment of the requirements to hone in on just the bubbles that cause adverse effects. For example, after installing detectors for very small bubbles in one application, the team learned that there were a tremendous number of very small bubbles, and that those kinds of bubbles (thankfully) had no measurable effect on the accuracy of the instrument. So relaxing the requirement was the way to go, after demonstrating a basis for doing so. The team implemented a smart bubble sensing subsystem that achieved the required sensitivity to critically sized bubbles, reduced nuisance alarms, and maintained the intended throughput.
TEST, AND TEST AGAIN
The last solution tip is simple: Identify all relevant fluid handling scenarios in your application and test them. A comprehensive test plan for any development project should include a focus on the most expected operating modes and likely worst-case conditions, as well as consideration of corner cases where certain combinations of variables could produce unexpected results.
To investigate air bubbles in a fluid handling system, this will mean getting a solid grasp on all the combinations of variables which could impact bubble behavior including flow speeds, fluid pressures, hydraulic resistances, fluid types, fluid interfaces, user interactions, and manufacturing variabilities to name a few. It can be quite a test matrix to tackle, but the process yields invaluable insight into bubble behaviors in your system.
Finding and assessing the impact of unwanted bubbles can be mission-critical to new product design. By dealing with the bubble issue early in the design process and then challenging your design along the way, it’s possible to – forgive the pun – deflate the risk.
Mariano graduated summa cum laude from University of Maryland, Baltimore County, with a BS in Mechanical Engineering. In a previous life Mariano also received a BS in Psychology from Towson University and worked in early childhood development and special needs therapy. When he’s not at work, you can find Mariano enjoying family time with his wife and three children, or riding a vintage motorcycle through Baltimore County landscapes.
Latest posts by Mariano Mumpower (see all)
- The Bubble Challenge Part II : Minimizing the Risk of Bubbles - July 6, 2017
- The Bubble Challenge Part I: Finding the Trouble with Bubbles - May 25, 2017
- Please Excuse the Growl - March 20, 2013