The Ins and Outs of Laser Process Validation

Laser micromachining is a versatile process that can be used for an array of medtech applications. Yet when using it, it is crucial to select the appropriate method of process qualification or validation for your specific application.

Scott Marchand Davis, Resonetics

Laser micromachining provides micron-level precision material removal and texturing in metal and polymer parts, allowing for a range of feature geometries to support different functions for both diagnostic and therapeutic applications. Typical applications include hole array drilling for embolic or diagnostic filters, catheter hole or window drilling for drug delivery or sensor exposure, precision wire stripping for neurovascular applications, coating ablation for solder pad exposure, cutting for stents, and texturing for tissue adhesion.

Validation Deliverables and Requirements

With such a range of processes, the key deliverables for laser micromachining can vary significantly. For many applications, dimensional and simple visual inspection provide assurance of a successful and stable process. For others, the effects on polymer chemistry (for example bioabsorbability) or material tensile strength can be critical. Selecting the appropriate method for process qualification or validation becomes an active interaction between the laser micromachining supplier and the life science device manufacturer, to ensure that functional requirements of the component are met reliably and consistently.

From a regulatory perspective, ISO 9001, ISO 13485, and 21 CFR Part 820 all include requirements for process validation. For ISO 9001, the requirement is to validate processes "where the resulting output cannot be verified by subsequent monitoring or measurement." ISO 13485 has the same language, but also includes a specific requirement to validate software used for production or service provision if it could "affect the ability of the product to conform to specified requirements." 21 CFR Part 820 also includes a similar requirement, with the major difference being that validation is required "where the resulting output cannot be fully verified by subsequent monitoring or measurement." Thus, from a regulatory perspective, the key requirement for process validation is the degree to which the process output can be verified after completion of processing.

Based on these requirements, only those laser processes whose quality cannot be verified after processing must be validated. This could include processes where material properties are important (necessitating destructive testing in most cases) as well as secondary processes such as cleaning or sterilization that cannot be verified by inspection or testing. Conversely, processes requiring purely dimensional and cosmetic inspections need not be validated from a regulatory perspective, and could be qualified by a simpler process. Dimensional inspection of laser micromachined parts is often costly and time-consuming, and process validation may be warranted economically by demonstrating a high process capability to justify sampling inspection rather than inspecting 100% of the parts.

How the Validation Works

Laser micromachining process validation follows a path similar to other equipment-based process validations, typically including an Installation Qualification (IQ), Operational Qualification (OQ), and Performance Qualification (PQ), as well as a Gage Repeatability and Reproducibility study (Gage R&R). By preference, the Gage R&R is performed before the OQ, but this order may vary based on availability of material.

Figure 1
Figure 1: A Gage R&R that needs additional work.

Gage R&R typically consists of approximately ten sample parts measured by two or three operators over two or three repeated cycles; the number of operators and cycles varies based on the level of difficulty of the measurement process. Because of the small feature sizes of laser micromachined parts, measurement is often challenging; in some cases, automated optical coordinate measurement machines can be used, but these require software programming and this software in turn must be validated. Use of automated inspection, either inline or offline, is often an efficient option for higher volume parts or parts with many features requiring measurement (such as filters with many holes).

Performing a dimensional Gage R&R provides assurance that the measurement system error does not take up a significant fraction of the part tolerance, which would affect the ability to achieve a reasonable process capability. For particularly stringent visual or cosmetic inspections, an attribute Gage R&R can be performed. Such a Gage R&R requires samples with known defect modes as well as good samples; assembling the requisite parts can be challenging, but establishing consistency among multiple inspectors to a subjective standard can be important for demanding applications.

Figure 1 shows a Gage R&R that needs additional work--the reproducibility element, showing variability between operators, is too high. This type of result would indicate a need for more robust procedures or fixtures for the measurement, and might also indicate a potential for discrepancies between the laser micromachiner's measurement and the customer's measurement of the same part. In these cases, it is critical to establish a consistent measurement methodology throughout the supply chain to prevent either rejecting good parts or accepting nonconforming ones.

IQ is required when new equipment or tooling is part of the process. Depending on the particular production model, laser micromachining may occur on a more flexible workstation with the capability to rapidly reconfigure for multiple parts, or on a dedicated workstation built specifically for a single part process. In the latter case, the IQ process can consist of verification of the required machine inputs, such as power, process gas, exhaust, and software controls, as well as outputs such as beam positioning, beam power or energy fluence, and vision system or metrology data. For a pre-existing system, IQ can be much simpler or even not required if neither tooling nor the workstation itself requires modification to run the part.

Example Installation Qualification Inputs and Outputs

Inputs

Outputs

Power

Beam energy

Cooling water

Laser operating temperature

Process gas

Gas flow rate

Control software

Part positioning

Safety requirements

Verification of emergency stop process

 

OQ can be a challenge for laser micromachining. While in many traditional OQ processes, the processor can select material at the high or low end of the material specification to test process boundaries, the small size and tight tolerances on raw material often make it difficult to produce the parts at all, let alone to adjust the process center point in one direction or the other. As a result, the common practice for laser micromachining OQ is to adjust process parameters to define a window within which the process can produce good parts. Typical process parameters for OQ might include fluence, focus depth, pulse rate, or scan rate, depending on the details of the process and the part geometry.

The graph in Figure 2, for example, shows the results of changes in fluence in terms of rate of material removal, an important parameter for both process speed and control of key dimensions such as the outer diameter of a skived tube.
 

Figure 2
Figure 2: Etching Rate versus Fluence

To complete the standard validation process, PQ provides assurance that the process will be reliable over larger volumes. Typically three shifts of production provides adequate statistical samples to assure process capability at nominal process parameters, although for some slower processes, additional production time may be needed to provide a large enough sample of processed parts.

In the example in Figure 3, the variability is shown well within tolerances for the dimension, establishing a capable process that can support sampling inspection.

Figure 3
Figure 3: A process for supporting sampling inspection.

One other necessary prerequisite for OQ and PQ is software validation. Since laser micromachining relies on computer control of optics and stages to achieve precise tolerances, the software providing that control must be validated to confirm it will perform as designed. Often this validation can be combined with a designed experiment to establish key process parameters and set nominal values.

Challenges in Laser Machining Process Validation

With the exception of OQ, the process described here is fairly consistent with process validation of other manufacturing processes. The execution of that process in laser micromachining, however, can provide some unique challenges. Simply measuring the output of the process can be complex for complicated geometries at a micron scale. For example, if the radius of curvature of the entry point of a laser drilled hole is important for part function, as might be the case for a flow orifice, three-dimensional profile measurement might be required to confirm the radius met customer specifications. Development of this measurement method and confirmation of its adequacy to assess part quality might add to the complexity of the qualification process.

Another potential complication of laser process validation appears when considering process and part variability. While some process variables such as beam position or fluence are relatively easy to control, others, such as cover gas flow or coating thickness, may not be so simple to manage. Designing the laser process to accommodate variability in part geometry and less-critical process conditions is crucial to ensuring a long-term capable process.

Validating processes for medical devices and diagnostics at the cutting edge of technology can also create issues with design variability and part availability. Iteration in part design may require iteration in process design as well; good communication between the micromachining supplier and the device design team helps minimize the impact of such changes. Additionally, the availability of parts for process development may be limited; particularly in the early stages of product development and with startup companies, prototype parts may be in short supply, requiring maximum efficiency with process experiments. Design of experiments techniques such as fractional factorial or Taguchi designs can provide maximum data with a limited number of samples.

A robust, validated process can provide a great deal of confidence for customers; data from validation can also be valuable when process changes become necessary or unexpected variation appears. For example, a recent validation identified a performance discrepancy between material lots that was traced back to an unqualified raw material supplier earlier in the supply chain.

In a more complex example, a wire stripping operation began to have higher fallout at final inspection. Comparison of current process data to the original validation data helped isolate the cause of the higher reject rates and establish a method for re-centering the process and returning it to its original performance.

Validation Delivers Big Benefits

While validation can be a complex and time-consuming process, its benefits are significant: not only can validation data provide support for later troubleshooting efforts, but also for reductions in inspection. Reduced inspection based on validation data can save cost and time in the manufacturing process. Validation also provides assurance that complex processes remain stable, producing parts that keep life science devices safe and ensure efficacy; this reduction in risk to both end user and device manufacturer is a critical benefit to a robust, statistically sound validation process. To reap those benefits, however, is a complex project requiring high levels of technical expertise in both quality engineering and the laser process itself.

Scott Marchand Davis is the director of quality assurance and regulatory affairs at Resonetics (Nashua, NH).

Learn more about medical applications of lasers at MD&M Philadelphia, October 7-8, 2015.

Create your own user feedback survey

500 characters remaining