I don't know if this will help in your circumstance, as I am unfamiliar with the construction and organization of information in the
PFMEA documentation in your use... but I can describe how I addressed this problem for a manufacturing customer. I generated several baseline manufacturing process
FMEA that were tied to the process (generally equivalent to major pieces of manufacturing equipment) rather than product.
This would be akin to having a PFMEA for an injection molding process as opposed to having a PFMEA for each housing.
- The PFMEA informative documentation included a process flow. The process flow was a semi-abstract diagram that mapped the distinct process steps of types of manufacturing processes, as opposed to individual components that pass through the process.
- Typically: many steps in the general process for each component all lead to a similar "end result" of a non-conforming part.
- The different effects of a non-conforming part were analyzed in a distinct section.
Some explanation:
(1) The steps of the map were approximately mapped 1:1 with a work-order traveler used as pieces moved between different stations/area on the manufacturing floor. We expanded the map to include "process steps" that might not be part of any
particular work order, such as equipment prep, maintenance, shutdown.
(2) All the steps in the process map had sections in the PFMEA worksheet, but most did not have any part-focused analysis. Instead, these sections direct the reader to review the "non conforming product" part of the worksheet. We did include the practices and controls that were in place for those specific process steps. For example: a setup step might have a risk control such as checking fluid levels, which might otherwise be cumbersome to call out as a risk control somewhere else in the process map.
(3a) Generally, the "non-conforming end result" section was the section of the process map where the physical work (on a part) is actually done... and where the mental work of analysis happens. For a deburring process, it would be when the worker applies the deburring tool to the part... as opposed to when prepping the grinder/knife/whatever or a post-deburring wash/rinse. It is important to note that this is where there need to be analysis of non-conforming outputs of the typical "too much, too little, wrong ____" as well as any special failure modes for unique design requirements (e.g. EMC/EMI shielding that may apply only to some parts).
(3b) The "non-conforming end result" will look very much what the
DFMEA has, except that for PFMEA it is almost always the case that the manufacturing process isn't actually
controlling design risks, it is more appropriate to say that the manufacturing process has controls in place to avoid
introducing risks.
(3c) The organization of this non-conforming end result section
will have output-specific lines. For my client: we categorized parts by product line, as the most serious (patient/user) risk was generally the same for all the parts that were produced via any given process. If there were special characteristics with unique risks for certain components, we sub-categorized. It is necessary to maintain a mapping of what parts belong to what lines of analysis.
I can offer some
specific advice about deburring: Deburring is (in my experience) one of the few manufacturing process steps where it makes a lot of sense to have both "too much" and "too little" lines of analysis in a PFMEA, because the effects can be quite different for each. Just as an example: too much grinding can leave gaps for fluid ingress, too little can lead to sharp edges, pinched wires. In contrast, a punching process that leads to non-conforming parts is unlikely to lead to as many different effects.