Validation of manufacturing processes is extremely important, given our highly regulated environment. Here’s a primer to get you started on the subject.
Steve VanderKooi, Kyle Kulwicki, Kevin Roberts, PTI Engineered Plastics
Validation uses objective evidence to prove that a process is capable of consistently producing a product that meets pre-defined requirements, including customer and user needs, functional specs and regulatory hurdles.
The thought of validating new or modified existing processes can raise your anxiety level. To reduce your worries, here are some answers to basic questions about this not-always-difficult challenge:
1. Why do I need to validate?
The short answer is that it is a federal regulation. A longer answer is to provide objective proof that a process is capable of consistently making parts to a specified confidence level.
2. What do I need to validate?
When the results of a process cannot be fully verified by inspection and testing, validation is required. And when the output of a process isn’t proven to meet acceptance criteria, it too likely requires validation.
3. When do I begin validation?
The validation process starts once you have a working design and a manufacturer capable of completing the validation process. At this point, you need to start the validation paperwork (IQ, OQ, PQ and possibly more). Keep in mind that design impacts validation, so the validation process should be considered from the earliest brainstorming stage.
4. How much is “enough” validation?
The process is validated with objective evidence of high confidence of producing a good product. In process capability statistical terms, Cp and Cpk must exceed, say, 1.66 for a risk level 1 product (arbitrarily chosen for this example) with an appropriate product sample rate during validation – including all secondary and tertiary documentation and testing.
5. Are there different levels of validation?
In a word, no. All aspects of the process must be validated to ensure that they result in the intended form, fit and function and that the process is capable of consistently meeting the required level of quality. This all ties back to the risk level of the part.
6. What is the difference between a validated and non-validated product?
A non-validated product does not have an assurance level of quality outside of in-process inspections, meaning that defects could be found within the manufactured lot. A validated product, on the other hand, will have a high level of quality assurance; no defects will be found in the lot when validation is done correctly.
7. Does my product design impact the validation?
Product design has a large impact on validation activities, depending on the critical dimensions of the part. As an example, an injection-molded tongue depressor might only have three critical dimensions that need a validation: Length, width and thickness. On the print for the product it is called out as a Class 1 device which requires a Cpk level above our arbitrary 1.66. Its validation revolves around those dimensions and that confidence level.
If we look at the whole design, there are round edges, but those were not called out as critical dimensions, so now they only need to be confirmed once, on a part to print. Fewer measurements, most taken without a coordinate measuring machine (CMM), means a shorter validation time.
Now let’s say you are making a housing that has action in it, a circuit board and over-molded metal inserts. The print has three notes that call for action once assembled, plus 45 critical dimensions. The product is also called out as a Class 2 device requiring a Cpk level above 2 (another arbitrary example).
This product is much more complicated, so the validation will take substantially longer due to the higher number of measurements needed to confirm that the process is robust and meets requirements.
8. What can reduce the cost and timing of validation?
To reduce validation cost and duration, optimize as many factors as possible – material, part design, mold design, injection-molding machine, etc. The material selected must not include extra additives, the material must not be a specialty grade and it must be readily available on the open market. The part design must have realistic dimensions, with the critical-to-quality dimensions identified only as needed. (Avoid details that add no value, such as a sunroof in Antarctica). The injection-molding machine selected must be repeatable and reliable. Last but not least is a robust injection-molding process that focuses on the parameters of the plastic that made the part, not the machine parameters.
To reduce duration, look at the entire pie chart. Typically, measurements and the mold-build stage take the most time. Anything that reduces the time spent in those two stages will save the most on the full validation time. Working 24/7 on those parts of the process will provide give great gains to reduce cost.
Steve VanderKooi is a quality engineering manager, Kyle Kulwicki is a process engineer, and Kevin Roberts is a validation engineer at PTI Engineered Plastics (Macomb, Mich.).