Would you all like to see ISO 4548-12 Oil Filter Lab Testing Comparison, Efficiency & Capacity, Pressure vs Flow, Bubble Point, and Burst?

[Ascent Filtration Testing]

I would look at the particles under a microscope to identify them and get a size. If that doesn't work a send a sample to a lab and run a mass spectrometer for composition.
If you want a size and distribution run a sample through a particle size analyzer (I plan on getting a used Horiba).
Use the microscope size in tandem with a batch sample run it through a particle sensor for good measure and count for the sizes from the analyzer test.
Run an efficiency test on the partially used filter with the discovered particle sizes. See if they are captured. If not select the most efficient filter you can find and plumb two of them in parallel with a deltaP gauge to indicate when to change them.

All I can help you with here is testing the current filter new and used for efficiency and capacity to discover if it is trapping the particle sizes of interest, and testing a better replacement filter for the same size particles. Also testing a batch sample of dirty fluid and discovering numbers and sizes of particles in it to select a better media.

 

[ABN_CBT_ENGR]

good strategy to find a custom solution to a custom problem

 

[bbhero]

Excellent, what time and time zone? I will make sure I am free.

I made good on this. .

Andrew and I talked for over a half an hour. We had a great conversation.

I believe that Andrew is a legit really, really good guy.

It was an honor and privilege to talk with him. And I told him that on the phone.

And yes... One if the first things I did was to apologize for how things may have appeared on here by what I had posted.

He was great and I mean great about that.

I told him that if he came to my area... The Fat Tuna restaurant with awesome and I mean awesome crab cake and fried shrimp would be the place where we'd meet up and have dinner.

He is really, really a good guy... And I am very, very pleased that he has joined us on here.

 

[Ascent Filtration Testing]

I made good on this. .

Andrew and I talked for over a half an hour. We had a great conversation.

I believe that Andrew is a legit really, really good guy.

It was an honor and privilege to talk with him. And I told him that on the phone.

And yes... One if the first things I did was to apologize for how things may have appeared on here by what I had posted.

He was great and I mean great about that.

I told him that if he came to my area... The Fat Tuna restaurant with awesome and I mean awesome crab cake and fried shrimp would be the place where we'd meet up and have dinner.

He is really, really a good guy... And I am very, very pleased that he has joined us on here.

It was indeed a real pleasure to speak to you sir. I look forward to breaking bread with you when I find myself in your neck of the woods, or you in mine.

 

[bbhero]

^^^^^^^^^^^^^

Amen to that man.

I look forward to learning more about all that you do...

And like I told man...

You are doing it... You are more than able or capable... One step, one day at a time

You have a lot to be proud of... And you are building a way for yourself. That's awesome.

 

[ZeeOSix]

ABN_CBT_ENGR said:
Are you assuming a constant laminar flow of equal velocity over the entire surface of the filter area with no accounting for density changes as back pressure increases and the venturi effect increases? That can create a transition zone with turbulent flow which "can" have erratic effects on a given media or geometric design depending on the flow path.

I do enjoy my fluid dynamics, and I bet we will have a great conversation. In the scope of what we do at AFT it is purely academic to answer these types of questions. We figure out what it does, if or when it does it. We leave it up to the manufacturers filter engineers to figure out how and why.

The Mann+Hummel/Purolator paper showing the efficiency loss vs time/loading hockey stick graph also showed computer modeling that tried to explain why/how it was happening. It was basically a way to design filter media using a model that correlated loading and particle retention best as possible to what they saw in the ISO testing to try and mitigate any filter efficiency loss with loading.

 

[Ascent Filtration Testing]

The Mann+Hummel/Purolator paper showing the efficiency loss vs time/loading and the hockey stick graph also showed computer modeling that tried to explain why/how it was happening.

I know a company that spent millions of dollars trying to use finite element analysis to simulate filter performance, only to come up with nonsensical data. Obviously we may have a bias here at AFT for physical testing, but my thoughts on FEA with respect to filter performance is not very favorable at this time.

 

[Ascent Filtration Testing]

The 15 and 20um efficiencies over time had many of those peaks and valleys over short time segments making it look like the particle sensor was broken lol

 

[OilReport99]

I know a company that spent millions of dollars trying to use finite element analysis to simulate filter performance, only to come up with nonsensical data.

How to interpret the data, usually, is as important or more as the data itself. But I agree with you. The tool for the job must be carefully selected to get the desired results. You can spend millions in FEA to come up with the perfect coffee brew and fail. All it takes is a sip by a coffee guy

 

[ZeeOSix]

I know a company that spent millions of dollars trying to use finite element analysis to simulate filter performance, only to come up with nonsensical data. Obviously we may have a bias here at AFT for physical testing, but my thoughts on FEA with respect to filter performance is not very favorable at this time.

True ... that's the pitfall of modeling. But M+H/Purolator seemed to have relatively decent correlation between the modeling and measured ISO performance data. If the model works to a decent degree, it's better than designing by trial and error when designing or trying to improve filter media. Of course lab testing (and/or field testing) is always used to verify/qualify the design, and also improve the model.

 

[ABN_CBT_ENGR]

I know a company that spent millions of dollars trying to use finite element analysis to simulate filter performance, only to come up with nonsensical data.

Then most likely the issue was with the people doing the FEA- not the FEA.

Granted, there are 2 types of people; those who swear by the FEA and those who swear at them. Both views have merit but like a polygraph, they have legitimate uses and will get you 80-95% there WHEN you plug the correct information into the VM stresses and the proper equations for the problem to be solved.

Obviously we may have a bias here at AFT for physical testing, but my thoughts on FEA with respect to filter performance is not very favorable at this time.

That's not an entirely unfounded bias and I agree actual testing trumps every 'model" but as stated, the problem with this company was most likely the people doing it.

 

[Ascent Filtration Testing]

Test Parameters
Flow rate to be determined, as I have 3/4" sanitary stainless steal tube on my machine and not sure if we want to switch it out to 1/2" SS tube.
Ingress = 5 mg/L
I can change the micron sizes any time also if anyone has a size they want to see.

 

[ZeeOSix]

^^^ Can your particle counters measure below 15u ?

 

[Ascent Filtration Testing]

Then most likely the issue was with the people doing the FEA- not the FEA.

Granted, there are 2 types of people; those who swear by the FEA and those who swear at them. Both views have merit but like a polygraph, they have legitimate uses and will get you 80-95% there WHEN you plug the correct information into the VM stresses and the proper equations for the problem to be solved.


That's not an entirely unfounded bias and I agree actual testing trumps every 'model" but as stated, the problem with this company was most likely the people doing it.

^^^ Can your particle counters measure below 15u ?

4-70um for the ones we are using here.

 

[ZeeOSix]

5 mg/l of ISO 12103-1 A3 test dust is the ISO standard, but wondering if you plan on doing any testing (only on one filter pair) to verify if the test acceleration rate skews the efficiency outcome to any degree ?

 

[ZeeOSix]

4-70um for the ones we are using here.

I think measuring down to 4~5u range would be worth doing, as some of the high efficiency filters can typically catch a good percentage of particles way below 15u.

 

[Ascent Filtration Testing]

Then most likely the issue was with the people doing the FEA- not the FEA.

Granted, there are 2 types of people; those who swear by the FEA and those who swear at them. Both views have merit but like a polygraph, they have legitimate uses and will get you 80-95% there WHEN you plug the correct information into the VM stresses and the proper equations for the problem to be solved.


That's not an entirely unfounded bias and I agree actual testing trumps every 'model" but as stated, the problem with this company was most likely the people doing it.

 

[Ascent Filtration Testing]

I share a dinner table with a FEA sim engineering director, she is much more qualified than I on this hot topic, and this is going to get me in serious trouble from my experience. I cannot give any farther opinions on filter modeling as I value my life lol

 

[ZeeOSix]

I share a dinner table with a FEA sim engineering director, she is much more qualified than I on this hot topic, and this is going to get me in serious trouble from my experience. I cannot give any farther opinions on filter modeling as I value my life lol

... LoL.

 

[ABN_CBT_ENGR]

I cannot give any farther opinions on filter modeling as I value my life lol

Then you better tighten up every chance you get