BOBISTHEOILGUY FILTER TESTS

[Brian Barnhart]

I’m very new to this sight, and I’m no mechanical or hydraulic engineer. Nor do I intend to offend or stir up trouble. However, it seems to me that this test is set-up wrong, at least as far as simulating an engine lubrication system, and evaluating oil filters. In an engine, much of the system’s restriction lies in the oil passages and bearing clearances, most of which lie after the oil filter. The only restrictions in this test set-up appear to be the filter, fittings, and hose. That may be why the filter pressure drops appear so high.

It’s very possible that this set-up pumps much more oil than a typical engine application, causing large pressure drops across the filter. This could be verified with flow data.

It seems what’s needed is a proper orifice (to simulate oil passages and bearing clearances) after the filter. This would reduce oil flow to a level closer to a typical engine, better distribute pressure drops, and allow the generation of data that’s more representative of engine applications.

Brian

 

[MNgopher]

Regardless of pressure drops across the bearings and the filter, or the bearings restricting flow to the filter, what these tests show is the relative amount of pressure drop across the filter. It seems reasonable to assume that this is a relative measure of the filter restriction, regardless of what the true limiting factor in the system is.

I'm not looking at the tests in the light of at x psi, I get x psi pressure drop. Instead, look at it from the perspective of Brand A drops x psi, brand B drops y psi, and therefore brand ? is less restrictive than brnad ?. Couple that with removal efficiencies (when availible from the manufacturer

) and you get a measure of the filters that no can opening test can duplicate.

 

[MolaKule]

Mngopher is correct.

The test is not to simulate the complete engine lubrication system but to determine pressure, flow, and restriction potentials of the filters being tested.

 

[BOBISTHEOILGUY]

Brian, You've got a point but I think you're not seeing the whole picture.

I have a 302 ford oil pump(new out of the box) attached to a 90vdc treadmill motor with a home built p/s, 4heavy diodes, cap on ac/ side, variac transformer(controls voltage input to diodes) diodes convert a/c to dc, and 3electrolitic caps on dc output to smooth out motor cycling.

The oil pump is attached via a 5/8th hose from the output, directly into the oil filter base(black hose). the output(red hose) same 5/8th and it is 3ft in length to simulate length of an average engine, given there is two channels, rod bearings and cam shaft area, I used the 5/8th to somewhat compensate for channeling through a motor, then I attached at the end of the red output hose a ball valve to setup a flow restriction similar to what a rod bearing would produce. I set the valve to 40lb input pressure to the filter with the motor using 30vdc input and a gutted out fram filter with 0 restriction, showing less than .5lb drop across it. Then we start the motor with test filter, and ramp motor up to 40lb pressure input on each filter with the Variac control not the valve.

Hope that helps explain the idea behind the test we are doing. We are basically looking at the drop across the filters with 40lb input. One thing to keep in mind as well is that so far all the filters(not bypass systems) were all purchased for my car a 1.9Lt ford engine, so the basic idea is that all filters should be presenting themselves somewhere in the same resemblance of drop set by ford engineers in this case.

 

[unDummy]

You need to tach that motor. Variac is being too fair by allowing each filter its own "power setting" and initial 40psi. Just wondering which filter is a HP pig, which one isn't, and which strains the oil pump the most. Stressed oil pump will be less efficient(leak more?).

Another thing to consider is a slow ramp up to see which filter starts bypassing earlier. 10/20/30 psi setting would also be interesting. Anymore oil pressure might be risky. Be careful. An eyefull of oil sucks.

Did you swap the full/bypass filters in that dual mount? if possible swap the center threads?

 

[Pablo]

Can we question the 40psi?

Don't most pumps crank more than this?

 

[Dan4510]

40 psi is a good point to test. I would say most stock motors of the old six and v-8 varieity rarely put out more than this stock. My slant six puts out 43 psi tops with m1 10w-30 and mobil 1 filter.

Big block chyrslers stock put out about 45 psi at the most. High performance were more, but i dont know how much.

340s had a high pressure high volume pump from the factory that was set at 80 psi. I tried just a high volume pump in my 340 when i rebuilt it and lifters clattered on high rpm runs from not enough oil pressure. It put out 45 tops with that pump and 10w-30 mobil1. I put a high pressure/high volume in it and it quit having clattering lifters. Topped out at 75 psi with that pump. All the literature warned that a beefed up pump drive had to be installed or it could be twisted off by the high pressure pump. I run 6 qts due to the windage tray in the engine.

I dont know about the new 4 bangers with all the computer controls. I may buy one of those in 2030 or so.

 

[Patman]

quote:

---

Originally posted by Pablo:
Can we question the 40psi?

Don't most pumps crank more than this?

---

At full throttle, and at 60mph on the highway, my LT1's oil pressure guage shows 40psi.

 

[Pablo]

The Volvo 245 (B21Ft)is about 60 psi cruising.....but as with your cars, this is post filter pressure. What is the pressure before the filter?

 

[BOBISTHEOILGUY]

That might be Pablo but question, what is your pressure at idle? obvoiusly you can't have 40lbs at idle, then what would you have?

 

[MNgopher]

Hey, my Jeep gets over 60psi at idle when its starting cold

Stays there for a while too until the oil heats up. When hot, It usually registers about 25 psi at idle, 55ish on cruise.

I think that 40psi is a pretty reasonable number that nearly all engines will see during operation.

 

[jason07-1zzfe]

40psi upstream of the filter does not make any guarantees about the flow being the same as an engine running at 40psi, nor ensure that all filters tested are flowing the same, simpley because upstream pressure is same.
There still needs to be flow measurement and needs to be set for each test.
I am sure you understand that 40psi downstream of the filter in an engine with significant restriction downstream of the sensor is not the same amount of flow as 40psi upstream of the filter, with no restriction after the filter. Even if there is restriction added after the filter you cannot say the flow is same, simply by monitoring upstream pressure.

 

[Brian Barnhart]

Bob,

Thanks for the additional information. I understand the test set-up better now. However, I still have concerns about flow rates.

I believe the following 3 points to be true. Correct me if I’m wrong.

1) Pressure drop across a filter is directly proportional to flow and fluid viscosity (and possibly a few other secondary things). Pressure drop across a filter may even be directly proportional to the square of the flow (which would make the flow rate even more crucial). At any rate, the higher the flow, the higher the pressure drop.

2) Filter input pressure, alone, has no bearing on pressure drop across the filter (See #1 above). However, if the input pressure is too high, the filter housing may burst.

3) Oil flow through an engine oil filter is only a few gallons per minute. If pressure drop is being measured at flow rates of significantly more than that, say, 9 gallons per minute, the pressure drop data may be misleading for judging oil filter performance in an engine application.

For example, let’s assume filter pressure drop is proportional to flow (not the square of the flow). And let’s say we have 2 filters, A and B. Filter A drops 6 psi at 9 gallons per minute and Filter B drops 15 psi at the same 9 gallons per minute. But at 3 gallons per minute, filter A drops 2 psi, and filter B drops 5 psi.

From the 9 gal/min test, one might conclude that Filter A is OK (6 psi drop) and filter B has unacceptably high pressure drop (15 psi). However, if the flow was 3 gallons per minute (instead of 9), one might conclude that while filter A flows better than filter B, either is acceptable (2 psi vs 5 psi).

Since pressure drop across a filter is directly proportional to flow (or square of flow), the flow rate needs be reasonably close to that encountered in engines. Otherwise the data gathered may be misleading and therefore might lead one to incorrect conclusions.

The test stand, as described, would seem to fit the bill. However, it appears that the ball valve setting in the outlet hose is absolutely critical because it simulates the engine bearing restriction and passages (and, therefore, nominal flow rate). Then adjusting input pressure with the power supply as described would achieve the desired results, and generate good data.

How was the ball valve setting established?

Brian

 

[BOBISTHEOILGUY]

The primary setting of the ball valve was done by putting the p/s dc vcc at 30vdc, at that level, I used a blank, no filter media, gutted out fram filter with a .5lb drop, and set the input pressure at 40lb using the ball valve. From that point on, it was not moved.

So to prove your theory, I will go out and put a filter on, set the unput pressure to 20,30,and 40lb input pressure and take a reading on all 3 steps to see if it is linear or not. Would this help you see how the filters re act based on this test?

 

[Brian Barnhart]

quote:

---

So to prove your theory, I will go out and put a filter on, set the input pressure to 20,30,and 40lb input pressure and take a reading on all 3 steps to see if it is linear or not. Would this help you see how the filters re act based on this test?

---

I’m not sure this would tell us whether the pressure drop across the filter varies linearly with flow. I believe the pressure drop across the filters would change, but I don’t think that would tell us the relationship between flow and pressure drop across the filters. I think we already know, or at least strongly suspect, that less input pressure causes less flow, and reduced flow means less pressure drop across the filter. It seems we really need to know flow to sort this out.

Thanks for the info on how the ball check valve was set. Do you have enough oil in the tank to pump it into a bucket for 15 seconds or more? If so, you could run the pump with the gutted filter at 40 psi as before, for a short time, and then measure the volume of oil pumped into the bucket in that amount of time. Then the flow rate could be calculated.

For example, if you could run the pump for 30 seconds and it pumped 2 gallons of oil into the bucket in that time, that would be 4 gallons per minute. If you don't have much oil, you could run it for less time. Just keep in mind that the shorter the amount of time, the more accurate the timing and volumetric measurement must be to keep the flow calculation error small. Hopefully the measured/calculated flow rate will be a few gallons per minute. If not, it seems the ball check valve should be adjusted until the desired flow rate is achieved. I believe this will also require changing the power to the motor to maintain the 40 psi.

Brian

 

[jason07-1zzfe]

The psi upstream of the filter is irrelevant.
As long as the flow rate is same on all tests and then you can measure the dp. Or you could set the delta p to be the same and measure the different flow rates

 

[Pablo]

Jason - In the case of the Amsoil dual by-pass with the magic spring and ball valve the prefilter pressure may not be so irrelevant. If that spring opens at X pressure then the dp would change....I just have NO concept at what pressure that spring opens...

 

[BOBISTHEOILGUY]

Well, let's consider this as a viable way to look at this...

If you have the valve set for a certain flow rate with 30vdc which is like the rpm of a motor, then we know that with no pressure restriction other than the bearing(ball valve), there is 40lbs of pressure in and out for all intense purposes.

That given, we need to keep the "rpm" or vdc set to 30 to establish the same rpm of the oil pump which should produce the same flow through the oil pump in effect and the resistance measured on the input/output pressure gauges should reflect higher psi with the same 30vdc(rpm's) of the motor. Now, keep in mind that flow is not going to be the same for each filter BECAUSE, in the oil pump, if the pressure gets over a certain point(backpressure from the filter) then there is a ball valve in the pump that will recycle it back to the input side instead of allowing it to flow through the filter, thus the difference in output flow from the pump.

With this marginal 40lb psi, I don't think we will hit that limit but it is possible. So, volume wouldn't have to be a part of the equation if the rpm is constant that would mean that volume (what ever it is) is constant, on the input side and depending on the filter restriction, it would reflect drop across each one acordingly.

 

[Brian Barnhart]

quote:

---

The psi upstream of the filter is irrelevant.
As long as the flow rate is same on all tests and then you can measure the dp. Or you could set the delta p to be the same and measure the different flow rates

---

I agree, as long as all you care about is pressure drop verses flow.

The interesting thing about this test stand is that it seems it can be set-up to simulate an engine oiling system. Once set-up, no adjustments will be required, except perhaps input pressure. Just screw on different filters and measure pressure drop. Want to test effects of a different or heavier oil? Just change the oil. Interested in the effects of oil temp? Heat or cool the oil.

If you want to simulate a higher pressure oiling system, just crank up the input pressure. Want to simulate lower pressure at idle? Reduce the input pressure.

 

[rob-the-oil-nut]

quote:

---

Originally posted by Brian Barnhart:

Just keep in mind that the shorter the amount of time, the more accurate the timing and volumetric measurement must be to keep the flow calculation error small.
Brian

---

I noted your postings for several reasons.
1) Your comments are well thought out, astute, helpful, and show intelligence and education.
2) You are new, so

3) I went to school with a Brian Barnhart, any chance you are he?
Rob