This article clarifies, on the one hand, whether additively manufactured components meet today’s regulatory requirements in medical technology and, on the other hand, whether these components meet the specified properties in terms of manufacturing accuracy and strength.
The advantages of additive manufacturing are generally known. But can we benefit from these advantages in medical technology? Which questions would have to be clarified before additive manufacturing can be used reliably? We answer these questions based on a concrete project that IMT solved on behalf of our customers.
For a customer’s device, a dynamically rotating system had to be developed within a short period of time. This system had to meet the following requirements:
Table 1: System requirements
To accomplish this task, we considered the following possible manufacturing technologies:
Based on the above requirements, the blue colored criteria were most decisive for the choice of the manufacturing technology: time, quantity, precision, stress, complexity, aesthetics, cost, …
We decided to use a combination of additively manufactured components post-processed by turning for the following reasons:
Figure 1: Reasons for choosing additive manufacturing
Additive manufacturing can be used to realize components of any complexity that would not be feasible through conventional manufacturing processes.
If you choose additive manufacturing technology, you clearly have a time advantage over, for example, an injection-molded component. Since no injection mold must be built, which usually takes 10-14 weeks, this naturally allows for much shorter iteration cycles.
It is clear that additive manufacturing is not suitable for very large quantities of, for example, over 10,000 units per year. But below that number, it can be worthwhile to use additive manufacturing. Once the manufacturing process has been validated, production can be scaled up with several machines and the same quality can be produced regardless of location. Which reduces transportation costs, time, and logistics efforts.
Manufacturer specifications show strength values that are quite comparable with semi-finished products.
Additive manufacturing reaches its limits when it comes to high precision requirements. To still achieve the necessary precision, certain areas of components can be reworked using conventional manufacturing processes such as turning. For this purpose the components are produced in oversize during additive manufacturing.
The choice of manufacturing technologies above now led to the fact that certain challenges had to be clarified.
On one hand, the production-suitable design for additive manufacturing played a central role in the entire design process. In this way, an economical solution could be achieved with the “Think Additive” approach.
On the other hand, confidence in additive manufacturing for medical devices was non-existent.
There were concerns about the following:
The lack of confidence in additive manufacturing had to be systematically addressed by clarifying the mentioned concerns. The definition of a clarification sequence was a key factor driving time and cost. For this purpose, short iteration cycles were defined using the “fail-early” approach.
In the following chapters, the chronological order of the clarification of the respective challenges and the lessons learned are presented.
To provide this evidence, possible materials were evaluated which were biocompatible according to information from 3D printing manufacturers. The evaluation resulted in the following selection:
Using the designated manufacturing technology, functional prototypes were then fabricated with these materials and sent for biocompatibility pretesting at Suva Analytik.
Table 2 summarizes the pretests results regarding VOC («Volatile Organic Compounds»), VVOC («Very Volatile Organic Com-pounds») and PM («Emission of Particulate Matter»):
Table 2: Results of the biocompatibility and purity pretests
Due to the detectable formaldehyde concentration, SLS material was not pursued further. For time and financial reasons, this result was not investigated further. Accordingly, all further investigations were carried out exclusively with the MJF HP 3D High Reusability PA12 material.
The surface of MJF manufactured components has a certain rough/porous structure. The task was to clarify the extent to which sealing with an O-ring is possible on this porous surface and what leakage rates can be achieved. Furthermore, it was necessary to find out how well MJF components are spannable and what accuracies can be achieved.
Figure 2: MJF structure in 5x magnification
It turned out that on this surface a sealing could be achieved with not too high tightness requirements. Leakage rates of < 1l/min at an applied pressure of 80mbar were achievable.
MJF components can be reworked in a spannable way without any problems. Some shrinkage cavities were visible here, which would certainly be visually disturbing in visual applications, but were functionally irrelevant for the designed press fit. A tolerance of ±0.02mm with a diameter of 4mm was achieved for the bore, which was reworked by turning.
To verify the specified strength properties of the MJF components, tensile tests and accelerated aging tests under load were carried out. The tensile tests were performed with specimens according to DIN 527. A tensile strength of > 43MPa was verified, which corresponds to the manufacturer’s specifications of 48MPa.
In the accelerated aging test, specimens were printed using MJF technology and the press fit was reworked to achieve the required surface finish and accuracy. Steel pins were then pressed into the specimens.
Figure 3: Aging tests under load
In accordance with ASTM F1980-16, the aging tests were carried out in a climatic chamber. Four time-intervals were defined to measure the transmissible torque of the specimens. In the case of thermoplastics, relaxation was expected, i.e. a decrease in the transmissible torque over time. However, practice showed exactly the opposite. Why this behavior of MJF PA12 is so, could not be further investigated.
Figure 4: Behavior of MJF PA12 material under load in aging test
With this knowledge, it was now possible to work towards a first production batch. In three iterations, the design was optimized down to the smallest detail in terms of production, assembly, and process technology. A build job definition was created with precisely defined acceptance criteria. The build job consisted of 103 part-sets, a fender roll angle and four tensile test bars.
Figure 5: Abstract representation of the construction job and associated test items
The acceptance criteria for a build job were defined as follows:
Subsequently, the part sets were classified into the 3 groups (see Figure 6).
Figure 6: Abstract representation of the parts sets in the construction job
The design was tolerance engineered based on published manufacturing accuracies, though we have tried to stay more on the safe side.
Figure 7: Manufacturing tolerances in the original design
The three sets of parts mentioned above were measured using computer tomography (CT) and a corresponding measurement report was prepared. The following points stood out during the measurements:
Gravity pulls the component lengthwise and thus also affects the detail of the edge areas.
Figure 8: Detail of the edge areas of the printed part in the CT image
Drop-in points / repeatability
Despite a design that was true to production, there were regions with drop-in-points (Figure 9).
Figure 9: Drop-in points of the printed part in the CT image
It was also not possible to maintain the required form and position tolerances over the three defined part sets.
Figure 10: Shape and position tolerance over the three defined part sets
Based on these results, the tolerance calculation was analyzed again to define the form and position tolerances more accurately for production.
Figure 11: Process-related manufacturing accuracy of “raw” MJF components after correction
As can be seen, the general profile tolerance has been increased from 0.3mm to 0.5mm. In addition, the references to the covers have been removed, so “Best-fit” applies. Position tolerances have also been removed, as they were already covered by the general profile tolerance. It can also be seen that the horizontal x-y plane can be manufactured more accurately than the vertical z plane.
In medical technology, it is common to be able to fall back on validated processes, which is also required by authorities such as the FDA. It was decided to apply an analogous process validation strategy (as is common in the injection molding process) to the MJF manufacturing process.
Specifically, we performed the following qualifications:
Table 3: Validation of process stability
The “Operational Qualification” was already carried out based on the first production batch. The findings from the FAI components (CT) led to the explained drawing changes.
The statistical evaluation of the critical mass showed that “cmk” values of ≥1.33 can be achieved with the selected form and position tolerances.
With these findings, the statement can now be made that MJF can in principle be used as a complementary manufacturing technology for the medical technology application described, because:
Now there is one question left: What about the reliability of MJF components?
To clarify this point, 50 systems were built using the MJF components from the first production batch.
In order to check the reliability of the assembled systems in the endurance test, a test rig was developed which allows 32 systems to be evaluated simultaneously. The systems were also operated under increased temperature and increased electrical voltage in order to accelerate the tests accordingly.
Two test patterns were defined, with the cyclic pattern being most relevant for the initial statement of reliability in order to proceed with the systems into more advanced overall system verification.
Test pattern 1:
Here the goal is to run the systems consistently for 25’574 hours. The systems are taken off the test bench at cyclic intervals to verify their functionality. Currently 6’300 hours of operation have been accumulated without failures.
Test pattern 2:
Here the goal is to operate the systems cyclically for 3.4 million cycles. Here, too, the systems were taken off the test bench at cyclic intervals to check their functionality. The 3.4 million cycles could be verified for all systems without any problems. At present, over 20 million cycles have already been completed without failures.
To learn the limits of the system and to know the margin from the specified range to failure, it was decided to perform additional HALT tests. The following “surveys” were performed:
As can be seen, the system meets the requirement with a sufficiently high safety margin.
The reliability of MJF components has been proven. For the requirements mentioned above, it could be systematically deduced that MJF in combination with post-processing (by turning) can be a complementary manufacturing technology for certain applications in medical technology for small to medium production series.
It could be systematically deduced that additive manufacturing with MFJ technology is a complementary manufacturing technology for medical devices. On one hand, the MJF technology fulfills all relevant regulatory requirements. On the other hand, a design that is both production- and load-oriented withstands the stresses over the service life. Regarding manufacturing precision, we demonstrated the limits of MJF technology, but also how precision can be increased by reworking.
From an economic point of view, it is imperative to use the advantages of additive manufacturing in the development and production of medical devices.
Have we sparked your interest in “Additive Manufacturing”? Do you have further questions around this topic in your projects? Then please contact us.
Leave your e-mail address here and we will inform you as soon as we publish new contributions.