The primary drivers for planning your QC strategy are:

• Designing the consumable based upon the expected production throughput and QC requirements
• Ensuring functional performance by evaluating the critical specifications prior to shipment
• Lowering cost of goods (COGs) by preventing the scrap from low manufacturing yield
• Monitoring the build quality to catch potential process drift or other indicators of issues with upstream processes

The strategy should consider the different types of tests to perform and their sampling rate.

Often the most reliable test method for many consumables is a destructive test. However, this approach may not directly measure the consistency of the manufacturing process and can mean reduced yields, or a decreased test rate that under-samples and does not detect failures.

Implementing non-destructive tests have their challenges as well. These tests may be laborious to conduct and their results potentially not repeatable. A common example of this is leak testing, which entails pressurizing the vessel and measuring the pressure decay rate over a set period. It can be a prohibitively slow process for in-line QC.

The alternate leak test approach of an off-line batch process will not only create complexity in the manufacturing workflow, but also increase the amount of scrap because an entire batch of faulty parts were produced before detection. A leak test may also weaken the thermal bond resulting in a back-to-back test with different outcomes.

The starting point when developing a testing strategy is to:

1. Identify the manufacturing processes that are critical to meet quality specifications and their required process capability (e.g., bonding processes such as adhesives or thermal and ultrasonic welding)
2. Use tools like a Process Failure Modes Analysis (PFMEA) to identify their failure modes
3. Determine what needs to be measured to identify failures and how it will be measured

These steps should be completed during the conceptualization of the part as this will affect the final design of the consumable.

When determining the measurement method, it is advantageous to use established methods. Visual inspection, by operator or vision system, is a common QC method implemented in manufacturing with a strong precedence in many industries. However, not all failures can be detected visually. Other more challenging methods may be needed.

For example, creating a microfluidics channel requires a seamless weld with no misalignments which calls for an extensive level of process development work. The assemblies are typically welded using different parameters determined in a Design of Experiment (DOE) – such as pressure, weld speed, power, and materials – to understand how variations in those parameters might affect the quality of the weld. Each weld condition will need to be evaluated using a destructive measurement method like pressure or flow testing of completed assemblies to ensure functional performance.

During this process, non-destructive methods, such as a visual inspection of the seams, should also be evaluated to identify correlations between the destructive and non-destructive methods. If a direct relationship can be determined, the non-destructive method can be used in manufacturing, and may even be able to be automated for continuous production testing.

Over time, process parameters can be modified to ensure consistent quality. As the development teams receive prototype parts, getting feedback on the functional effectiveness of those components is essential to monitoring and improving the consistency of the process. This approach will help to build a manufacturing knowledge base of the process parameters that reliably produce a good, functional part as well as the normal variation in the process.

QA/QC implementation strategies vary by production volume of the manufacturing process. Ideally, you would measure and control every process variable and non-destructively test every process step; however, this would be prohibitively expensive in most cases. Therefore, selective risk-based sampling coupled with statistical process control prior to release of finished product is commonly implemented.

The goal is to have a hundred percent confidence in the consistency and quality of what has been produced in the batch; however, a hundred percent inspection is generally not economical or practical – and may even be impossible.

Best practice is to evaluate the critical areas throughout the process at the component, subassembly, then assembly levels. For example, evaluating molded parts independently before assembling ensures that failures are identified early.

This avoids having to troubleshoot, disassemble, re-work and reassemble completed units. Failures become more expensive the further downstream in the design and manufacture process they happen.

For consumable assemblies, planning out the in-process QA/QC strategy enables an inherent confirmation that each operation has been performed correctly and consistently. The approach also enables each QC step to be developed in parallel with the assembly process itself.

One example is incorporating a load cell into a dispensing station. The load cell not only confirms that the correct amount of material has been dispensed, but also prevents failures that may occur during the heat sealing process if a well is overfilled. An added benefit is also decoupling fill failures from the causal chain during troubleshooting of the heat sealer.

Timing of the QC implementation is critical. If you introduce it too soon, the manufacturing process or consumable designs might still be in development requiring additional changes to the QC process or fixturing. Implementing can delay production, or, even worse, allow faulty consumables to reach customers.

The optimal timing is determined by engaging with your chosen contract manufacturer early and often. Even sending the preliminary strategy will provide you with valuable early feedback and allow them to prepare.