I just spent a couple of hours re-reading several of the ISO specifications related to “Millipore” testing including ISO16232-3 and -7. I put the word Millipore in quotes here because although these are normally called Millipore specifications, none of them mentions the word Millipore. (Millipore happens to be the name of a company that developed the testing concept and is a prominent supplier of the filters that are often used to collect contaminants for evaluation as described below).
Why did I read these specifications again for the umpteenth time? Because I am desperately searching for something that succinctly defines what a Millipore test is and how it should be performed. I thought I might have missed something in the previous umpteen minus one readings. Why? Because I continue to see test results on the same parts that vary wildly from lab to lab and even within the labs of the same facility. In my mind, a test performed to a specification should provide a definitive, meaningful and repeatable result.
Millipore testing is a widely utilized method to measure cleaning effectiveness. A cleaned part is re-cleaned using a method superior to the method being evaluated. A part cleaned using an automated spray washing machine with an aqueous detergent, for example, might be re-cleaned using a hand-directed high pressure spray with a low surface tension solvent like mineral spirits to remove any contaminants not removed by the aqueous spray washing procedure. The run-off from the second wash is collected and then filtered to isolate these contaminants for quantification. Quantification can be by weight of the particles or by counting the number and size of particles collected on the filter. The following comments apply primarily to the particle counting method.
A typical “Millipore” specification requires that there be fewer than a specified number of particles in each of several size groups and/or type categories collected in the re-wash for a part to be acceptable. In some cases, the specification will differentiate between contaminants of concern and those that are not. For example, metallic flakes within a certain size range might not be acceptable while fibers falling in the same size range may be of no concern – but we’ll get to that later. For now, let’s just concentrate on counting and sizing particles.
A lab charged with evaluating the cleanliness of a part usually receives the following – –
- Cleaned sample parts bagged to maintain cleanliness.
- A table defining the allowable number of particles in each of several size categories for a part to be acceptable.
- Instructions to apply particular ISO specifications or, occasionally, a more detailed test procedure.
You say, “So what’s the problem?” Well, just about everything! It’s a little like my grandmother’s recipe for “the best apple pie” that has all of the ingredients listed but none of the quantities or how to mix and bake them. My drift is that even with strict application of the elements of the ISO specifications, the results can vary considerably based on interpretation and procedure. This may lead to false rejection or acceptance of a part based on cleanliness as each lab applies the specifications in a different way.
The real problem is that the ISO specifications do not really define the test. They set guidelines and limits but as long as you stay between the foul lines, the test is considered to be valid. But, there are so many variables that have an effect on the outcome that are not controlled!
Before everyone goes berserk, let me say that there facilities that have produced and adhere to very detailed procedures for “Millipore testing” which do produce meaningful and repeatable results, at least within that facility. Unfortunately, these represent a minority and each operates in isolation.
Any testing instrument or procedure requires calibration to provide a repeatable result. I have looked but, until now, have not found a way to calibrate the testing in my lab to assure that it will give the same result that my customer might get in their tests. SO I NEED YOUR HELP!! Is anyone out there aware of a source for a calibrated dirty samples for testing? By “calibrated samples” I mean samples or groups of samples that have been doped with a precisely known level of contaminant and packaged in a way that assures the contamination remains consistent until tested? I can find calibrated dirt but no samples with dirt applied in a precise way to act as a standard for comparison.
Meanwhile, I would love to hear from those of you who think I am barking up the wrong tree! If my observations above are wrong, let me know with a comment below or an email directly to me at email@example.com. I’m anxious to hear from you!