Medtech Makers

Miniaturization and the Implantable Enclosure—A Medtech Makers Q&A

A look at the factors for why the housing, not the electronics, drives your device's size, cost, and schedule.

Released By PrecisionX Group

By Sean Fenske, Editor-in-Chief

As active implantables shrink—neuromodulation, leadless cardiac, CGM, implantable sensors—the titanium enclosure becomes the limiting factor on device envelope. For these applications, deep-drawn seamless housings beat machined-and-welded assemblies on internal volume, wall thickness, and hermetic joint count. These are just a few of the factors that must be considered for these types of projects.

Ultimately, OEM design teams should make numerous decisions early to avoid a redesign at qualification. These involve material selection, the internal components, charging method, communication protocol, and a number of other aspects that must be settled before the housing can be finalized.

To help further explain these choices and how they can impact the final design of the implant, Christian Bolcar, VP of AI and Engineering at the PrecisionX Group, participated in the following Q&A. In this interview, he responds to questions around many of the aforementioned options and explains what may determine which route is right for a specific project.

Sean Fenske: When it comes to the size of medical implants, which aspect is more challenging to miniaturize, the enclosure or the electronics? Why?

Christian Bolcar: Although we can reduce the size of our enclosures, there’s only so much we can do, and a lot of that comes down to the wall thickness relative to the size of the case.

On a larger case, the wall thickness is small compared to the case’s overall size. As the case gets smaller, the wall thickness doesn’t decrease linearly with it. The ratio changes, and the case behaves and forms differently. That’s one constraining factor.

That wall has to balance several things. It has to suit wireless power transmission and heat. It has to stay within the material’s natural ability to form without tearing, which depends on how hard you draw relative to the part: the reduction per stage, the corner radii, and the sheet thickness compared to the part’s size. And it gives the part its structural integrity, which comes from its thickness.

Heat pushes the wall thinner. With wireless charging, the field induces eddy currents in the titanium wall, and a thicker wall generates more heat. ISO 14708-1 limits the outer surface of an implant to 2°C above body temperature. So as the form factor gets smaller, the wall can get thinner, down to a specific limit, and that’s actually a benefit.

Forming and strength set that limit. We form under pressure, drawing the material over the punch and into the form die over several stages. If the wall gets too thin, it tears or thins unevenly at the corners and the bottom radius, and it loses structural integrity.

The wall isn’t the only thing that takes up space. Every weld joint needs a weld land, clearance from the heat-affected zone, and a leak path that has to be qualified. Components need clearance from the wall, for insulation and to stay clear of weld heat. Feedthroughs take space too, and they present their own sealing challenge.

The electronics are a different story. Implantable electronics are different from mainstream electronics. For CPUs and GPUs, the focus has partially shifted toward power efficiency rather than a smaller form factor. And in implantable electronics, the battery is often the main volume driver, not the logic.

Fenske: What are the common missteps you see from OEMs when specifying the housing for a medical implant?

Bolcar: I’d say a lot of the time it’s radii and inside corners. Material tends to collect at corners. A regular rectangle forms easily because the material flows from the outside in. But take a shell or cover with four corners where one corner is inverted. Now you don’t have four corners; you have five, and the material for that inverted corner has to come from somewhere. It comes from the two adjacent edges. An inside corner can’t be fed the way an outside corner is, so you get thinning, tearing, or wrinkling, and it can force extra draw stages or a machined-and-welded feature.

Wall thickness tolerancing is another, starting with the material from the mill. The industry supplies material at tolerances relative to the nominal thickness, so as the thickness goes down, the tolerance gets tighter, and we can get a tighter spec. But thinning is natural to the drawing process. We start thicker, and as we draw down, the walls near the bottom get thinner while the walls at the top get thicker, because material is being pulled up. The flat bottom under the punch stays close to the starting thickness. OEMs should know the wall won’t be one uniform thickness. One tight wall tolerance everywhere asks for something the process won’t deliver.

That ties into tolerance stacking. Some customers tend to over-define, and the stack doesn’t leave room for us to actually hit the wall thickness. References are fine as long as they’re defined properly. Choose the dimensions you actually need to hit, usually the sealing surface, the weld joint, and the feedthrough interface. Put real tolerances on those and leave the less important ones as references.

Material is the next one. Sometimes a grade gets specified that can’t be formed easily. Titanium comes in different phases with different purposes. Commercially pure titanium suits drawn parts, while stronger alloys like Grade 5 and Grade 23 are usually machined. Specifying only a grade name is another gap. The spec should call out the full material: the ELI (extra-low interstitials) grade where it applies, the ASTM F-number, interstitial limits, and lot traceability.

A few others we see include carrying over a grade or wall from a previous device without rechecking it, and prototyping with a process that won’t be used in production and expecting the same results. Cleaning and passivation validation also belong in the enclosure spec.

Finally, geometry matters for MRI. During a scan, the induced RF field concentrates at the ends and tips of a conductive implant, and the heating happens in the tissue right next to it. Leads are the main concern, but sharp edges and points on the housing are worth avoiding for the same reason. Size matters too: the ASTM test method for passive implants notes that a device smaller than 2 cm in every direction may not need RF-heat testing at all.

Fenske: What fabrication technologies or techniques are used to form the enclosure? What’s the difference? When and why is one used over the other?

Bolcar: Our deep drawing process itself is a typical deep draw. Where we differ is our draw ratios, which are proprietary to us. The draw ratio is how much you reduce per stage. Push it too hard and the part tears; too conservative and you add stages.

Beyond that, a lot of it comes down to the tooling. Depending on the size of the enclosure, our draw tooling may be hard-machined on our ultra-precision hard tooling mill, which holds extremely tight tolerances and produces good finishes. For longer draws, we use EDM to produce the draw dies. Our punches are ground. The polish on the punch and die matters, because titanium galls against tooling. So does air relief, which keeps air or vacuum from being trapped between the punch and the part. Trapped air can cause bulging, collapse, or trouble stripping the part off the punch.

We also use pressure rings to hold pressure on the blank as it’s drawn into the cup. They control how the material feeds in. Too little pressure and the wall wrinkles. Too much and it tears or thins too far.

Machining is the other way to make a can, and it can’t do the thin walls drawing can. A can is typically end milled. As the wall gets thinner, it loses stiffness. The cutter pushes it away, it springs back, and chatter can drive it into the tool until it eats through the wall.

Drawing has its own limit too. If the thinning goes too far, the material rips. So, both processes have a floor on wall thickness from different causes. Drawing is limited by tearing, and machining is limited by the wall losing stiffness and chattering into the cutter. Drawing generally reaches thinner walls than machining.

Fenske: What types of materials are used for an implantable housing? Why is one used over the other? What challenges are presented with each?

Bolcar: It tends to be feature-driven: weldability, radio frequency limiting, and what the device needs to do.

Titanium welds well under an inert shield. Its limitation is RF. Titanium conducts, so it blocks RF and charging fields, and getting signals or power through takes a feedthrough, an antenna window, a ceramic section, or an antenna in a plastic header on top of the can, which is common in pacemakers.

Ceramics like alumina and zirconia are RF-transparent and don’t heat up under charging, but they’re brittle and costly. They can’t be welded, so they’re brazed to a titanium ring. Silicon carbide is an emerging RF-transparent option.

Within titanium, the decision is which grade, and that comes down to its phase.

Alpha phase is commercially pure titanium, or CP, Grades 1 through 4. It’s essentially unalloyed. Its crystal structure is hexagonal close-packed, which gives it fewer easy ways to deform than steel and makes it direction-dependent. Even so, it’s ductile enough to draw cold. CP is the most formable, the most biocompatible, and the lowest in strength. Grades 1 and 2 are typical for drawn cans: Grade 1 is the most formable, and Grade 2 is the common workhorse with a bit more strength. CP is gummy to machine. The chip strings and smears instead of breaking.

Alpha-beta phase covers Grade 5 (Ti-6Al-4V—6% aluminum and 4% vanadium) and Grade 23 (Ti-6Al-4V ELI). These grades dominate orthopedic and load-bearing implants. Aluminum is used because it strengthens the titanium through solid-solution strengthening, it’s low density so you get strength without weight, and it keeps the alpha phase stable at higher temperatures. Vanadium makes the alloy heat-treatable and adds back toughness. The same aluminum that adds strength makes the alloy resist the draw. Alpha-beta grades have high yield strength, low uniform elongation, and high springback. When machined, they break chips better than CP, but they hold heat at the cutting edge and react chemically with the tool, which wears tooling fast.

Grade 23 is extremely hard to form. It forms like Grade 5. You can compare it almost to glass, where it’s hard to move without heat. The comparison is about forming, not brittleness, since Grade 23 is tough. In design terms, alpha-beta grades end up machined or formed with a heavier wall, while CP allows a thinner drawn wall at lower strength.

Grade 9 (Ti-3Al-2.5V) is the middle ground. It’s stronger than CP, with better formability and weldability than Grade 5.

Beta phase has a body-centered cubic structure. Beta alloys are specialty and not typical for implant cans.

On biocompatibility, Grades 5 and 23 are accepted in implants because they form the same protective oxide layer. There is a long-term concern about aluminum and vanadium ions. ELI lowers oxygen and iron to improve toughness and fatigue resistance. It doesn’t reduce the aluminum or vanadium. Ti-6Al-7Nb (ASTM F1295) replaces the vanadium with niobium, and it’s used mainly in orthopedics.

Fenske: What advantages are realized from having a complete implantable solution provided by one supply partner rather than sourcing components or having assembly performed by multiple vendors?

Bolcar: It may seem like having multiple supply partners spreads the risk. But with a single source that has high-precision capabilities like ours, you deal with one person and one knowledge base, and we get to a solution quicker.

Other companies can’t do what we do as a one-source supplier because they don’t have the correct equipment, knowledge base, or resources to learn, which we do. They have to outsource specific processes under their engineering master or process sheet, and when they sub out processes, they have less control over timeline and quality.

That shows when something goes wrong. When issues happen with sub-suppliers, and in the real world they do, it’s harder to investigate the root cause and make a fix. In a corrective action, you have less influence on a sub-supplier. With one source, all the process owners can be gathered at once. With sub-suppliers, it’s very difficult to get a meeting of the minds quickly. It becomes a logistics issue that can delay delivery, which is critical, especially when you need parts quickly.

With one partner, you get one design-for-manufacturing review covering forming, machining, and finishing, one first article and one owner for any corrective action, traceability from material lot to finished can in one quality system, and faster engineering changes.

We have deep drawing, precision and Swiss machining, and progressive stamping, along with mass and hand finishing, cleaning and passivation, and welding, including laser welding. Leak and hermeticity testing is available when required. All of it runs under our ISO 13485:2016 quality system.

Fenske: When it comes to the timeline for an enclosure, what should OEMs expect? What changes when using multiple vendors for the process?

Bolcar: There isn’t one typical lead time, because the timeline scales with complexity. As complexity grows, so does the timeline, especially when we’re doing things no one has done before.

With multiple vendors, you have less control over the timeline. Any issue means getting process owners from different companies together, and that becomes a logistics issue that can delay delivery. With one source, they can be gathered at once.

Fenske: Do you have any additional comments you’d like to share?

Bolcar: Bring the enclosure in when the device envelope is set, not after the electronics freeze. Early on, share the target envelope, the charging scheme (frequency, power, and coil location), the annual volume, and the sealing method.

Customers shouldn’t be afraid to push the limits. We’re able to invest in technologies and to help develop technologies that push the industry to the next level. That’s the whole point of PrecisionX. We engineer with our customers. We don’t just manufacture.

Click here to learn more about the PrecisionX Group >>>>>

Request more information from PrecisionX Group

Keep Up With Our Content. Subscribe To Medical Product Outsourcing Newsletters