Zero RF Leakage MRI Balun Component via PEI Injection Molding

The Challenge

An MRI balun component lives inside the radio-frequency environment of a magnetic resonance scanner, where the part must not perturb the RF field or leak energy that would corrupt the image. The component was molded from unfilled polyetherimide (PEI), an amorphous high-performance polymer, because its dielectric behavior is stable and predictable and it introduces no conductive or magnetic artifact into the field. Material purity and dimensional stability were therefore non-negotiable, not nice-to-have.

Dimensional stability under the scanner environment was the core difficulty. PEI has a glass-transition temperature around 217 °C and excellent high-temperature retention, but it is hygroscopic and its dimensions are sensitive to moisture and to internal mold stress. A part that was dimensionally correct on the bench but shifted as it absorbed moisture or relaxed internal stress would change the RF behavior of the balun and degrade image quality. The program tolerance was ±0.05 mm on features that had to stay put for the life of the scanner.

Internal stress was the hidden risk. PEI is processed hot, at a melt of roughly 340 to 425 °C with mold temperatures of 140 to 175 °C, and if the fill or pack is unbalanced the part freezes in residual stress that later relaxes and moves dimensions. For a balun that is exactly the failure mode to avoid, so the mold flow had to be tuned to produce a low-stress, uniform part rather than merely a filled one.

The program was a medical production engagement with a real quality gate. The client was a leading Tier-1 medical imaging OEM, the volume was 15,000-plus units per year, and the schedule ran 8 weeks from design-for-manufacture to a PPAP release. That meant the part had to be right at first article and capable of repeatable production, not tuned by hand on the floor.

Qualification for a scanner environment is exacting. The balun sits inside the magnetic resonance field for the service life of the machine, so the part had to be shown stable in dimensions and in dielectric behavior, not merely acceptable at first article. Any relaxation of internal stress or moisture uptake that moved the ±0.05 mm features would shift the RF performance the customer relied on.

Volume and lead time left little slack. At 15,000-plus units per year with an 8-week DFM-to-PPAP window, the part had to be right at first article and capable of repeatable production. The hygroscopic nature of PEI meant drying and process control could not be assumed; they had to be specified and held, or the dimensional stability claim would not survive contact with real manufacturing.

Dielectric stability is the real spec. Inside the scanner the balun must not perturb the RF field, so the PEI part had to hold both its dimensions and its dielectric behavior part to part and over time. Because PEI is amorphous and hygroscopic, the path to that stability ran through drying, through low-stress molding, and through verification that the part would not drift as it relaxed or took on moisture in service.

High processing temperature is a double edge. PEI is molded hot, at a melt around 340 to 425 °C with mold temperatures of 140 to 175 °C, which gives the part its thermal class but also makes internal stress easy to lock in if fill or pack is unbalanced. For a balun, that stress is a latent dimensional shift, so the process had to be tuned for low stress rather than merely for a filled cavity.

The program also had to respect the scanner stack. The balun located into the RF assembly where its dielectric and dimensional stability decided image quality, so the molded properties had to hold relative to the field, not just to the part drawing. A component perfect in isolation but drifting in the field would still degrade the image, so the ±0.05 mm stability was managed against the scanner environment, not the single part.

The Solution

We selected unfilled PEI specifically for its stable dielectric performance and inherent flame behavior, and we treated dimensional and moisture stability as design inputs from the start. The part was dried to the correct level before molding so the resin entered the barrel at a known moisture state, removing a major source of dimensional drift between shots.

Mold flow was tuned to avoid internal stress that could later shift dimensions. Gate location, fill rate, pack pressure, and cooling were balanced so the PEI filled symmetrically and froze without locked-in stress, and the mold temperature was held in the 140 to 175 °C band to keep the amorphous part dimensionally consistent. The result was a low-stress part whose ±0.05 mm features stayed stable after molding.

Process discipline supported the RF requirement. The molding parameters were locked and documented, and the dielectric-relevant dimensions were inspected so the balun behaved the same part to part. Because PEI is inherently stable and low-outgassing, the part met the clean-material needs of the scanner environment without additives.

The 8-week window from DFM to PPAP was met through a steel-safe review and a validated tool. First-article inspection confirmed the ±0.05 mm features and the dimensional stability, and a PPAP package backed the release to the 15,000-plus-unit yearly volume.

Inspection tied dimensions to RF performance. First-article measurement confirmed the ±0.05 mm features, and the low-stress molding was verified so the part would not drift as it relaxed or equilibrated moisture. The dielectric-relevant dimensions were checked so the balun behaved the same part to part inside the field.

The tool and process were released through PPAP. The 8-week window was met with a validated tool and a locked, documented process, and the material handling (drying, melt and mold temperature control) was encoded so the 15,000-plus-unit yearly volume reproduced the first-article result. A spare strategy kept the program available.

We treated stability as a design input. The resin was dried to a known moisture state before molding, and gate, fill, pack, and cooling were balanced so the PEI froze without locked-in stress, holding the ±0.05 mm features stable after molding. The mold-temperature band of 140 to 175 °C was held so the amorphous part stayed dimensionally consistent shot to shot.

We tied inspection to RF performance. First-article measurement confirmed the ±0.05 mm features and the low-stress molding was verified so the part would not drift as it relaxed or equilibrated moisture, and the dielectric-relevant dimensions were checked so the balun behaved the same part to part in the field. The 8-week DFM-to-PPAP window was met with a validated tool and a locked, documented process.

We qualified the balun against the RF environment. The dimensions and the low-stress molding were verified so the part stayed stable in the field, and the dielectric-relevant features were checked so the balun behaved consistently inside the scanner. The 8-week DFM-to-PPAP window delivered a tool and process the Tier-1 medical imaging OEM could release with confidence.

The Result

The program delivered RF-leakage-compliant balun components with consistent dielectric behavior, molded from unfilled PEI and held to ±0.05 mm. The low-stress molding approach kept the critical features dimensionally stable so the part did not perturb the scanner RF field across its service life.

The leading Tier-1 medical imaging OEM received a PPAP-released part inside the 8-week DFM-to-PPAP window, and the 15,000-plus-unit yearly volume ran from a documented, repeatable process. The component met the material-purity and stability bar required inside the MRI environment.

First-pass yield met the program target, and the low-stress PEI part held its ±0.05 mm features with stable dielectric behavior, which is what kept the balun from perturbing the scanner RF field across its service life. The RF-leakage requirement was met on the production part, not just on a sample.

The leading Tier-1 medical imaging OEM received a PPAP-released component inside the 8-week window, with the traceability and stability evidence the scanner environment demands. The validated approach gave a baseline for follow-on MRI components using the same unfilled PEI discipline.

The balun component met the scanner-environment requirement on the production part. Molded from unfilled PEI and held to ±0.05 mm with stable dielectric behavior, it did not perturb the RF field across its service life, and the RF-leakage requirement was met by the molded part rather than by a sample or a coating. The 15,000-plus-unit yearly volume ran from a repeatable process.

The leading Tier-1 medical imaging OEM received a PPAP-released component inside the 8-week window, with the traceability and stability evidence the scanner demands. The validated unfilled-PEI discipline gave a baseline for follow-on MRI components and a documented basis for the material-purity and dimensional-stability claims the application requires.

The balun performed as an RF system component inside the scanner. By holding the ±0.05 mm features with stable dielectric behavior relative to the field, it gave the 15,000-plus-unit yearly volume the image-quality support the application required, and the PPAP release let the medical imaging OEM carry it into production with the traceability the environment demands.