Bosch 3330 cut open

[NJC]

I don't believe it's the length of service that caused the failure, it's a faulty design. As evidenced by the many other Purolator posts showing the exact same problem - and many of those were within service limits.

I'm changing to a Mann ML1003 - they are also made in the USA but appear to have come out of a different factory.

 

[Sam2000]

Mann as in Mann+Hummel Purolator?

 

[Garak]

Originally Posted By: NJC
I'm changing to a Mann ML1003 - they are also made in the USA but appear to have come out of a different factory.

That is likely a Purolator Classic clone. If you don't like the Ford Tempo, you don't switch to the Mercury Topaz, as it were.

 

[Eltydiyer]

Originally Posted By: Sam2000
Mann as in Mann+Hummel Purolator?


+1 Exactly, Mann made in USA is pretty much a Purolator Classic.

It's too bad, because I heard good things about the German made Mann. It's apparently the OEM filter maker for a lot of cars.

 

[440Magnum]

Originally Posted By: ZeeOSix
Originally Posted By: PhillyJoe
That's the problem. These tears don't turn up on new filters, only used. The filters may fail due to usage, pressure, etc. but they only show up after we've used a pipe cutter, can opener or hack saw to open. Yes, we are all careful, but if I am Purolator or Bosch, I would be reluctant to accept responsibility when the opening was not under their controlled conditions. Having said that, P seems to acknowledge that a problem may have occurred.


If the tears were created during cutting open, then explain why the pleats have tears on both ends. People do not cut open both ends of the filters. And many here use a filter cutting tool, which will not damage the media. And if someone does nick the media when using a hacksaw, etc it's pretty obvious that the damage was from a tool and not from media ripping open due to oil flow force.



I think the only way the opening process is doing this is if the entire can twists right at the end of cutting and "torques" the media end-caps relative to each other. Its definitely NOT an issue of the cutter hitting the media.

 

[NJC]

Originally Posted By: Garak
Originally Posted By: NJC
I'm changing to a Mann ML1003 - they are also made in the USA but appear to have come out of a different factory.

That is likely a Purolator Classic clone. If you don't like the Ford Tempo, you don't switch to the Mercury Topaz, as it were.

Well, when I say I'm switching it meant that I had already ordered and received the Mann filters before cutting open the Bosch or realizing Mann / Purolator / Bosch were the same in this application.

Haven't written that, I may return the filters to Rockauto and get something else. This application is for my 2004 Sienna, and I don't want to worry about potentially bypassing filters from [censored] MFG filters.

 

[NJC]

I'm speculating here - but the problem could be weak media (numerous failures are possible) or lacking pitch uniformity at media to end cap bond. I haven't studied other filter designs enough to know if this is typical, but it looks like the pleats are too tight for majority of glue joint, and too wide near seam.

 

[ZeeOSix]

^^^ Yep, that's been the talk for quite some time now. The photos show a good example of how the pleat's V-spread where glued into the end cap is large on the two pleats next to the seam.

Weak media plus higher side force on those pleats from oil flowing through them is most likely causing enough stress in the media for it to fail.

I'm curious to see if filters with date codes a month from now have better pleat spacing and smaller V-spread near the seam. If so, then Purolator picked up on this too.

 

[NJC]

I have a Bosch 3330 filter on my 2004 Sienna, which I recently installed prior to knowledge of these failures. I'll run it 6 months (service interval for van) and cut and post the pics. Hopefully I can use a real camera next time too.

 

[440Magnum]

Originally Posted By: NJC
I'm speculating here - but the problem could be weak media (numerous failures are possible) or lacking pitch uniformity at media to end cap bond. I haven't studied other filter designs enough to know if this is typical, but it looks like the pleats are too tight for majority of glue joint, and too wide near seam.



I agree with ZeeO that this is the best working theory. Where the theory comes apart a little (pun intended) is that if pleat uniformity were the only factor, then EVERY entry-level Fram-built filter (Orange can and clones) would do this, because they are far more wide-pleated in places than a Purolator Classic. Just speculating here, but maybe the Fram Orange Can's end caps made of filter media actually save it. Maybe its the stress of a flexible medium glued to a rigid metal end cap. Or maybe its the Classic media is softer than the cheapy Fram media.

I dont' think its just one cause, its at least two or more contributing factors coming together.

And its shtuff like this that is driving me to pay the premium for synthetic wire-backed filters for EVERY vehicle at the moment. Mostly Royal Purple (I've cut 4 or 5 open after use and never found anything but absolute perfection), but I've currently got one Fram Ultra and one Purolator Synthetic in service also. I've cut 1 used Ultra in the past- also perfection. Give me about a year to put some miles on the Puro Synthetic, and I'll let you know how it holds up. Its on the least-driven vehicle- the Ram 1500- so it won't come off for quite a while.

 

[NJC]

Originally Posted By: 440Magnum
Originally Posted By: NJC
I'm speculating here - but the problem could be weak media (numerous failures are possible) or lacking pitch uniformity at media to end cap bond. I haven't studied other filter designs enough to know if this is typical, but it looks like the pleats are too tight for majority of glue joint, and too wide near seam.


I agree with ZeeO that this is the best working theory. Where the theory comes apart a little (pun intended) is that if pleat uniformity were the only factor, then EVERY entry-level Fram-built filter (Orange can and clones) would do this, because they are far more wide-pleated in places than a Purolator Classic. Just speculating here, but maybe the Fram Orange Can's end caps made of filter media actually save it. Maybe its the stress of a flexible medium glued to a rigid metal end cap. Or maybe its the Classic media is softer than the cheapy Fram media.

I dont' think its just one cause, its at least two or more contributing factors coming together.


I think this is the correct, multiple causes. If I had to hazard a guess and pick a culprit, I'd go with weakened media. Not necessarily too thin (as per pictures above, the media from Bosch 3330 measure .6mm or .024") but too weak - either weak design or batch, or weakened with oil saturation .. or another explanation. I did a quick glance through this gentleman's PDF, and he compares Corolla filters but also has media measurements - they seem to vary between .021" and .038" so the .024" I observed is well within the range.

http://www.corolland.com/oil-filters.html

Originally Posted By: 440Magnum
And its shtuff like this that is driving me to pay the premium for synthetic wire-backed filters for EVERY vehicle at the moment.

Myself included, I'll go for a different MFG like Wix. But gauging from other's observations, Purolator filters are VERY popular which means this problem is widespread. And it's significant.

I won't do business with companies that produce substandard product, and it should be shouted from the roof tops until something changes.

 

[ZeeOSix]

Originally Posted By: 440Magnum
I dont' think its just one cause, its at least two or more contributing factors coming together.


Most definitely. And the conditions the filter is used in also play a role. This current rash of torn media reports are most likely filters that were used over the cold winter season, which means the oil was a lot thicker on cold start-up which could have put more stress on the pleats. And if one of these filters is used on a car with a high volume oil pump and/or the driver revs the engine pretty high with cold oil, then again there is more stress put on the pleats which could make the tears happen. Many factors involved that have to add up together in the wrong direction.

 

[Sam2000]

At the risk of sounding anal, those additional contributing factors are not a cause or fault, they are circumstances that will be repeated and should be tested for.

The only fault we can see is the V shape pleat. The media is strong enough if arranged correctly or if this arrangement is correct then the media is not strong enough.

But as far as I can see, there is only one reason behind the fault happening.

The scientific / logical way of explaining this is necessary / sufficient.

Cold weather / high revs is not sufficient for the tears to happen. It is also necessary to have the wrong filter design.

Now cold weather / high revs may be necessary for the failure to occur but my memory suggests some have this fault without those circumstances.

 

[ZeeOSix]

^^^ That's why I asked in a few threads to the member who was emailing back and forth with the QA director at Purolator if they were doing extreme cold use test cases in their investigation (ie, -20 deg F start-ups with 5W-30 oil and worst case oil flow rates kind of scenarios). He made it sound like they were, but who really knows without real proof beside work of mouth.

 

[earlyre]

Originally Posted By: NJC
For REF, the date code is F06D07D2 from the Bosch 3330:


which is Purolator code for
F = Fayetteville, NC
06 = June
D = 2012
07 = 07
D = Line D
2 = Shift 2

or in English:
Built June 07, 2012; Fayetteville, NC Plant; Second Shift; Line D

 

[Nate1979]

Originally Posted By: ZeeOSix
^^^ That's why I asked in a few threads to the member who was emailing back and forth with the QA director at Purolator if they were doing extreme cold use test cases in their investigation (ie, -20 deg F start-ups with 5W-30 oil and worst case oil flow rates kind of scenarios). He made it sound like they were, but who really knows without real proof beside work of mouth.


Temps are very mild here in the PNW barely reaching freezing. I can't imagine the cold caused my filter tears.

 

[ZeeOSix]

Originally Posted By: Nate1979
Originally Posted By: ZeeOSix
^^^ That's why I asked in a few threads to the member who was emailing back and forth with the QA director at Purolator if they were doing extreme cold use test cases in their investigation (ie, -20 deg F start-ups with 5W-30 oil and worst case oil flow rates kind of scenarios). He made it sound like they were, but who really knows without real proof beside work of mouth.


Temps are very mild here in the PNW barely reaching freezing. I can't imagine the cold caused my filter tears.


It may not have to be super cold, depends on what oil weight is used and how the car is driven during a warm-up. I'm in the PNW too, and the Classic that tore on me was during the winter months use (Sep - May) per my records.

 

[KCJeep]

If they bypass is working properly the extreme cold starts should not subject the media to extra abuse, unless the media is so poor that even low, sub bypass differential pressures are tearing it.

 

[ZeeOSix]

Originally Posted By: KCJeep
If they bypass is working properly the extreme cold starts should not subject the media to extra abuse, unless the media is so poor that even low, sub bypass differential pressures are tearing it.

If there was large delta-p across the oil filter in very cold start-up, I don't think the by-pass valve would react instantaneously, so it's possible there could be pretty large short bursts of pretty high force on the pleats while the by-pass valve is trying to react. A few of those and it could cause the media to tear.

If Purolator can't accurately simulate the most extreme use cases of the oil filter in the lab (ie, Montana winter cold starts), then they may not find the cause if that is indeed a factor in the cause.

 

[TriboGeek]

Originally Posted By: NJC
I'm speculating here - but the problem could be weak media (numerous failures are possible) or lacking pitch uniformity at media to end cap bond. I haven't studied other filter designs enough to know if this is typical, but it looks like the pleats are too tight for majority of glue joint, and too wide near seam.






Thanks for this post. It very nicely and plainly illustrates the primary problem. And let me emphasize, this is indeed the PRIMARY problem. Media strength is of course a factor also, as well as operating conditions/pressures, but those would not be a problem if the pleating was correct.

In another thread I made a verbose attempt at describing how this pleat arrangement creates a problem. I can simplify it quite nicely now in hindsight, and with these above photos: The oil pressure pushes the media toward the center tube. It is plain to see that there is about 3 times as much cross-sectional area of media material in the normal pleated area (it's a lot closer to optimal, even if it might be a little tight) as compared to the "V" area. This is essentially the problem - the media in the "V" area is being subjected to 3X the shearing stress (1/3 as much material to do the same load-bearing). There are other contributing stresses created as well (bending from lack of mutual pleat support), but this is the main idea. The media in the "V" area experiences substantially higher stresses than it ought to, because of the pleat construction flaw. It is the primary problem, because with proper pleating and no other change, failure would not be happening in the failed examples.