Millipore Cleanliness Test Experiences

rstocum

Involved In Discussions
My company is implementing a cleanliness requirement from our biggest customer. The parts with the requirement are fuel valve components and assemblies for automotobile engines. The requirement is called "Millipore cleanliness testing". The Millipore part is actually the name of a company that pioneered the testing methodology, and is in the business of providing the equipment.

We ended up purchasing the necessary equipment from Millipore because it was a designed system that fits together right out of the box.

The test process is called gravimetric filtration. Basically, you clean the parts in a solvent (we are using kerosene). You collect the kerosene the parts were brushed clean in, and filter it through a specified filter under vacuum pressure. Since you weighed the filter before use, and weigh it again after drying it out, you get the weight of the contaminants present from the difference between the weights.

Our initial testing to shake out the process indicates that we are compliant to the basic requirement, which is less than 3.0 mg of contaminant per part on the part we are currently testing. We fall short on the maximum particle size allowed (.010" x .008"). I have been finding particles as large as .017" x .019".

We have improved component storage to reduce our atmospheric contaminant exposure. We improved on the overall contaminant weight, but the particle sizes were not affected. We have filtering and preventive maintenance systems in place in our part cleaning operation that should prevent this particle contamination from that source. We will be auditing those systems to see if there are problems there. The parts are cleaned prior to assembly, and assembled in a room separated from the production floor to maintain cleanliness. The containers the components are kept in during assembly are cleaned in the same cleaning process as the parts themselves.

The cleanliness testing results are consistent across batches, so the problem is likely to come from the system, not some special cause.

Any ideas?
 

Kales Veggie

People: The Vital Few
Re: Anyone with Millipore cleanliness test experiences?

I have dealt with similar issues in the past and a lot of work was done to solve it.

I have quite a few questions.

1) what is your background noise? (particles from your measurement system, could be the reason for your "consistent" reading)
2) what is the cleanliness of the kerosene?
3) what are the results if you wash down the same part 5 times?
4) how are you controlling your wash process and wash fluids?
5) Have you analyzed what element(s) the particle are? This should point you towards the origin of the contamination.​
 
Last edited:

rstocum

Involved In Discussions
Re: Anyone with Millipore cleanliness test experiences?

I am new to this testing, so I will answer as best I can. Please let me know what I am leaving out.

The measurement system as far as I know, consists of:
Labware (beakers, filter flask, Buchner funnel, stainless steel wash tub,
brushes)
Filters
Drying Oven
Balance for weighing filters & samples
Microscope (20X)

I am not sure how you define background noise, or what I would do to isolate it?

We wash all the labware in a separate stainless steel sink under hot water - no detergents. The balance has a shroud to keep the weighing plate isolated from air movement, and is kept in the lab isolated from production. The oven is brand new, and also in the lab. The parts washing (for millipore test) is also done in the lab. The lab is kept clean with mopping and wiping down once per week, and is atmosphere controlled.

The kerosene is filtered through the same filters used to take the samples (5 micrometer nylon).

We haven't done one set of the same parts five times yet. We have done the same set twice and the overall weight improved, but similar particle sizes were still present.

The production parts cleaning operation is a semiautomatic water tank cleaning system. One wash tank (using detergent), two rinse tanks, and then a blower chamber to dry the parts. All the tanks and the blower chamber have filters. I haven't confirmed how small a particle those filters will pass yet, but I an doing that, and millipore testing the water in the tanks today. The tanks have ppm of contaminants measured daily, and pH weekly.

The contaminants are various metallic & non-metallic, but mostly non-ferrous metallic, which makes sense given our product mix.

I have found that parts straight out of the wash system have much less overall contamination, but that some of the same size particles are still present.
 
D

Dean Frederickson

Re: Anyone with Millipore cleanliness test experiences?

rstocum,
One thing I noticed is you haven't said anything about checking for magnetism, one of our customer spec's requires you check for magnetism of the part because that can cause particles to stick to your valve (we manufacture tubes) and not be flushed out during your cleaning process. We use small stainless steel balls, plug the end of a tube add solvent and sst balls and agitate to knock any particles loose that our solvent (we use mineral spirits)doesn't clean out of the tubes. You can use a gauss meter to check for magnetism. Hope this helps.:cool:
 

BradM

Leader
Admin
Re: Anyone with Millipore cleanliness test experiences?

Are you sure the analytical balance has the sensitivity and such to assess the difference in weight? Are you controlling all the factors for the balance-air flow, vibration, etc.? If you haven't calculated an uncertainty for your balance, I would recommend it.

I might perform a pre-post verification check on the balance too.

How sensitive is the process during drying? Possibly a uniformity test would be necessary on the oven to detect hot spots, or in this case, cold spots. It's possible the filter is not being dried out sufficiently. Hopefully this is an electric oven. I don't think you want a gas-fired oven with Kerosene in there. Even with electric, make sure you have adequate ventilation, and sufficient air changes on the oven. I know it's brand new, but it needs to be qualified (in the basic sense) that it will accomplish what you wish. How the filters are placed in the oven may have a significant effect also.

Oh yeah, remember those filters are not perfect. I know the Millipore come with a certificate, but there is some variance in them. If this is a critical process for you (sounds like it is) might want to think about investing in a filter tester. Or, develop the process to manually test the filters, pre and post.

Hope something helps.:D
 

rstocum

Involved In Discussions
Re: Anyone with Millipore cleanliness test experiences?

Dean,

We don't have a magnetism requirement in our customer's millipore spec. Our assembly is non-ferrous.

Brad,

We are not experiencing a problem with the sample weight. The balance is level, and is sensitive to the degree required in the spec. (.1 mg). It has a shroud to keep air currents from throwing the measurement off. The oven is electric. The auto ingnite temperature of kerosene is 410 F, and we are drying at 150 F for 10 minutes per the spec. We are doing great on sample weight. We are supposed to be under 3mg per part. We are allowed to average a multi part sample (we are doing 10 pieces), and we get .6 mg per part. The problem is that the overall sample includes several metallic particles that are over the acceptable particle size limit which is .008" x .010". I have measured particles as large as .019" x .023" in our vision system. I know that the particles come from our process. I am currently suspicious of the cleaning operations and our storage methods.

To update other details, the filters on our cleaning system will pass particles small than 10 micrometers. The preventive maintenance may not be as scrupulous as I had hoped. The air filter on the drying section has not been changed according to schedule. We are fixing that problem today, and re-testing.
 

BradM

Leader
Admin
Re: Anyone with Millipore cleanliness test experiences?

Ok, so if I am to understand, you produce these products, then clean them using the kerosene. You weigh the filters to see how much "stuff" came off the parts. The specification is that you should not collect any greater than a certain specified weight of "stuff", at a certain size.

If that is accurate, I would be concerned about the large particle size also. Is it possible that the kerosene cleaning process is removing more off the parts than desired? Is the "stuff" in the filter all bad stuff, or could it contain good stuff?

Essentially, it is sounding like the problem is at the point of cleaning with kerosene, and going backwards from that. Is that accurate?
 

rstocum

Involved In Discussions
Re: Anyone with Millipore cleanliness test experiences?

Yes, that is accurate, in my opinion. The particles are a mixture of aluminum & brass, which is a good reflection of the chips we cut in our manufacturing areas. I don't believe that bits as big as .019" x .023" are floating around in the air here, but I can't discount the possibility yet.

I don't believe the kerosene is removing anything that isn't already loose on the parts. The current assembly I am testing is all aluminum. There is a rubber o-ring that goes into it, and under the microscope, there is no damage to the o-ring from washing in the kerosene. In addition, there is no "stuff" that appears to be rubber particles in the cleanliness samples.

I am concerned at this point that our cleaning system is the source. The possibility exists that contaminants in the wash & rinse tanks could "hang around" without being filtered by collecting in corners and against seams inside the tanks. We drain and rinse weekly, but I am doubtful that the tanks are actually scrubbed as the PM sheet calls for. I am also almost certain that the filters in the drying chamber are not up to the cleanliness requirements for our millipore test.

I am testing these theories by rinsing parts fresh from the wash system under a spray wand right before starting the millipore test. I will let everyone know asap.
 

BradM

Leader
Admin
Re: Anyone with Millipore cleanliness test experiences?

Excellent follow-up post, rstocum!:agree1: I truly appreciate you keeping with the thread on this one. Let us know what you find out about the cleaning process, or any other things that might be of significance.
 

rstocum

Involved In Discussions
Re: Anyone with Millipore cleanliness test experiences?

The spray wand idea did not work. The overall sample weight went down, as expected, but there were still particles present that were over the limit.

The cleaning system is not up to the cleanliness standard, and millipore tests of the rinse tank water prove it. The cleaning system is set up for the vast majority of our parts, and cannot be modified to accomodate the few assemblies that are subject to this millipore requirement.

Does anyone have any suggestions for cleaning methods (systems) that would keep the particle size under .008" x .010"?

My other choice is to go to the customer's engineering and quality groups and ask for a review of the requirement. We have been providing these assemblies for years with no cleanliness problems that have been communicated to us. The customer is a large corporation (read heartless and slow on the uptake), and is unlikely to relax the requirement. Finding a way to meet the requirement without busting the budget would be the ideal.
 
Top Bottom