Operational Qualification of a CNC machine

Q

QltyGrl

Hi,

I work for a very small medical device manufacturing company (< 10 employees). We are currently working on ISO 13485 certification and I have been assigned the task of validating all of our machines, processes, products, etc. The task is very daunting and I do not come from a QA background but am a recent transplant into the field. I am really struggling with designing validation protocols and although I have read the GHTF guidance document and have done a lot of reading on using IQ/OQ/PQ as a validation tool I still am very confused when it comes to OC and how to design a protocol for a CNC machine.

I understand that IQ has to do with installation, training, preventative maintenance, etc. and I understand the with PQ you are qualifying the machines ability to consistenly product a conforming product but I am totally lost when it comes to OQ.

I would greatly appreciate any and all assistance. If anyone has an example they can share of a validation protocol written for a CNC machine that would be wonderful. Please keep in mind that I am no statistician and am still learning a lot about the world of quality.

Thanks in advance for all assistance! :thanx:
 
A

arios

For OQ/PQ purposes you could design a sample specimen (e.g. a jig with various shapes, depths, profiles and contour) and have your CNC producing such part. You will assess your equipment capacity if the specimen meets your expected requirements for radii, dimensions, etc. You could use a CMM to inspect or lay-out your specimen

For IQ, confirm your CNC equipment is installed as per the manufacturer's recommendation (level, power, phases, inlet pressure, etc). You may want to confirm vibrations in the surrounding area (if they exist) do not adversedly affect your equipment's performance.

P.S. Based on your OQ results you may determine which machine operations are worth to keep under regular dimmensional inspections and which ones not. For instance, if your CNC produces pretty accurate radii, then you may have a rationale for not inspecting such types of shapes as your equipment has proven capacity to producce them.

I almost forgot, please consider having a verification of the proper transmision of data from your laptop with the design data to the CNC's. This data transmision could be part of a software validation
 
Last edited by a moderator:

Bev D

Heretical Statistician
Leader
Super Moderator
OQ is one of the most misunderstood aspects of FDA regulations (guidance) for process and equipment validation.

technically speaking the requirement is for all processes whose final output cannot be verified by subsequent measurement. (aka 'special processes') It also includes defects or failures that may occur after use - reliability type of Problems.

Practically speaking most medical device companies also perform OQ on all processes and equipment such as CNC machining.

OQ is essentially verification of the ruggedness of your process settings. You identify the process inputs (including raw material inputs such as hardness) and their min/max settings. OQ tests the ability of process - inclduign the equipment - to produce product that meets all specifications at the 'extremes' of the process settings. Typically this is a fractional factorial with the critical factors and any interactions directly tested while varying the other factors 'randomly'. For newbies this may be somewhat hit-or-miss. Over time these types of test structurese become more surgical as the organization gains knowledge and expertise in experimental design. Some approaches I've seen are to create lots where all factors are set at the low limits and then a lot where all factors are set at the high limits. Another approach is to use engineering knowledge and judgment to appropriately mix and match the levels of the settings depending on their effect on the output.

A precursor to the OQ lots is to perfroma a screenign experiment and Response Surface to determine the critical factors and their levels.
 
J

JaxQC

Arios, what is you opinion on a ballbar test in place of the sample specimen for the OQ?
 
Top Bottom