BOBISTHEOILGUY FILTER TESTS

Gary Allan · Mar 12, 2005

quote:

That is where you assume wrong.

Well, let's think about this a little more...

quote:

The pump is going to pump all it is allowed to pump because the motor only allows so much oil to pass thru the oil channels.

Well, sorta. The pump is going to pump in relation to its speed. That FLOW will result in a pressure. If that pressure never exceeds the oil pump's internal relief settings ...100% of every drop (except for internal minor leakage) will go through the engine.

quote:

Higher pressure will push more oil thru the same channels.

Well, it is more correct to say that higher flow will result in higher pressures. The pressure is a result of the flow ..not the flow dependant upon the pressure.

quote:

It is the channels that determine the amount of oil flow.

No. Your oil pump's volume determines your flow. The restrictions of the channels may determine your pressure.

quote:

You can't get more oil going thru the pump and filter than what goes into the engine.

Please show me where I said this??

quote:

The pump pushing the oil against the oil gallery, oil passages, or oil channels (what ever you want to call them) is what gives you your oil pressure.

Absolutely!!! The FLOW that the OIL PUMP is moving produces the pressure.

quote:

What is PSID? PSI is per square inch but what is the D?

PSI "D" = differential.

When you have a positive displacement pump ..for every rpm you are going to move a given volume of, in this case, oil. The pump doesn't have any other option unless its internal relief settings are reached. Xxrpm ..xx gpm ..no two ways about it. When your oil is cold ..or your filter picks up debris ..the PSI differential across the media will increase. If the differential is in excess of the filter's bypass valve setting ..let's say 12 PSID ..that vavle will open and a portion of the oil will not go through the media. 100% of the flow still goes to the engine ..it is just divided between going through the media and bypassing the media.

Flow is king in this situation ..everything else changes to make the equations work. This isn't like a water line ..a fire hose ..or an electrical outlet. Ohm's law (adapted) applies to a certain degree ..but you've got current (flow) being the dictating element.

Winston · Apr 13, 2005

Let me add a bit more analysis to the data. Like someone else pointed out, it would be great to create a Flow vs. Pressure drop curve for these filters. This would allow us to determine when the bypass valve opens. And like someone said before the valve does not "pop" open it cracks open once the set pressure is reached, then continues to open more as the pressure increases. If you had a flow vs pressure drop curve it would look like this.

Oops, I tried to draw one with ** but it does not work.

The pressure drop across the filter will be proportional to the flow squared, then once the relief valve starts to open the curve will flatten out dramatically.

Thus the curve will ramp up exponentially, then flatten out.

Having said that, lets look at the data we do have.

Filter: PIN POUT dp
Amsoil: 40 34 6
Mobil 1: 40 30 10
K&N: 42 36 6
Fram: 40 38 2
Bosch 40 30 10
Napa Silver 38 32 6
Fram TG: 40 22 18
Napa Gold: 40 32 8
Pure1 40 28 12
Purolator+ 40 32 8
Motorcraft: 40 24 16
STP: 40 34 6

As far as Bob's testing system. It would be nice to have a flow meter but it really is not necessary. The piping/tubing on the downstream portion of his system does not change. Thus, the pressure drop of the downstream piping (POUT) is proportional to the flow squared. The DP across the filter is proportional to the flow squared. So DP is proportional to POUT. Get it? So, if I adjust the dp measured to what we would see if the POUT was equal to 30(picked from a hat). The data would look like this:

Fram: 1.6
K&N: 5
Amsoil: 5.3
STP: 5.3
Napa Silver 5.6
Napa Gold: 7.5
Purolator+ 7.5
Bosch 10
Mobil 1: 10
Pure1 12.9
Motorcraft: 20
Fram TG: 24.5

These are sorted from Low to high. Now what can we see from this data? First off, the Fram at 1.6 is broken. Then we have a group of filters at 5.x, a group at 7-10, then a 12.9 and two 2x.x filters. Pressure drop is influenced by: 1.filter flow resistance(generally more resistance equals finer particle filtering) 2. filter area (more area = less DP) 3. Can design (number and size of holes at bottom and in center core). I would say the fives have good can designs with a lot of filter area with unrestrictive filter media. Once you go up from their it is hard to predict. But, if you combine these results with what we know from Grease's study we can see some correlation to the fine filter media of the Pure One, and TG filters. The Motorcraft is an anomaly like the regular Fram. I bet the extreme number from the Fram TG is due to a poor can design....

Let the Flames Roar!

Rabid Dog · Apr 23, 2005

I just thought I'd use my first post, to bump this excellent thread. I can't wait for the third set-up, and then the synthetic tests.

Gary Allan · Apr 27, 2005

quote:

It would be nice to have a flow meter but it really is not necessary.

Well, I believe Bob ran the electric motor until he established 40 psi on the upstream gauge. This allows the media to vary the true flow of the oil. If he had merely set the flow to 40 psi downstream in all cases, then it would probably mean that the engine was receiving the same flow (assuming that it's restrictions/resistance was constant). That would then express all media restrictions on the same plane of flow and would have altered the results. Just like in a real running engine ..the downstream pressure is a result of volume and viscosity ..while the upstream pressure is an additional resistance presented to the pump and doesn't alter flow within the confines of the pump's limit on resultant pressure.

That is, only velocity of the oil changes in any of these "chokes". It has to - to balance the equations. So, I believe, that, within certain confines (namely the pump relief setting), that the rest doesn't mean a whole lot in the composite scene. All observations are resultant to the mandated physics of the model.

[ April 28, 2005, 11:34 AM: Message edited by: Gary Allan ]

Winston · Apr 28, 2005

Originally posted by XS 650.

quote:

Pressure drop across a filter element in much closer to directly proportional to flow than to the square of the flow.

hf = K*((V**2)/2g)

Ref. Cameron Hydraulic Data.

Hf = Head loss in feet
K = resistance coeficient
g = acceleration of gravity
V = velocity

This equation says that head loss (in feet) due to friction equals K times velocity squared divided by 2g.

Velocity is proportional to flow. Thus, head loss is proportional to flow squared.

This is my reasoning as to why I said it was proportional to the flow squared. Why did you say it was not?

[ April 28, 2005, 04:47 PM: Message edited by: Winston ]

Winston · Apr 28, 2005

Re: Gary's message.

It is funny Gary, whenever I read your posts I seem to get lost in the wording. I know you have said the same about my posts. Hmmmm.

I was able to follow you through your first two sentences. Then you lost me. Specifically your second sentence said;

quote:

If he had merely set the flow to 40 psi downstream in all cases, then it would probably mean that the engine was receiving the same flow

I agree and this is what we want! We want to know the pressure drop caused by the filter at a given flow rate. That data would allow us to compare all the filters.

Gary Allan · Apr 28, 2005

quote:

It is funny Gary, whenever I read your posts I seem to get lost in the wording.

Yes, I can't handle too much nowadays in absolute values and complex ritmatick..but conceptual things are still within my mental grasp. Expressing them so others can understand them is another thing. You have to have drank from the same water fountain or something, I guess.

If I setup a bench system like Bob is and manipulate the upstream pressure to 40 psi (which IIRC, Bob did) and then measure the pressure drop (agian, IIRC, Bob did) across the filter..then the flow is a function of the total pressure/resistance (or some variant of it).

If however I set the flow to a downstream pressure of 40, regardless of the upstream pressure of the filter (pump side) then we can reasonably assume that the engine (assumed to be a constant) is receiving the same flow regardless of which filter is inline with the pump.

Now, if my wording isn't too cryptic (it's a crippling affliction at times)......

At this point it doesn't much matter whether you have dense media ..light media ..restricting media ..large holes ..tiny holes ..Gaping portals ..minute orifices ....."worm holes", transporters, conduits (subspace and otherwise-transwarp or conventional)

YOU ARE AT THE SAME FLOW IN EACH AND EVERY FILTER INSTALLED INLINE (remember ..the downstream pressure is 40 PSI into a constant restriction=the engine) ..

So ..the PSID is merely expressing the velocity increase for the fluid to make it through this "resistance" ...or rather the change in velocity is evidenced by the PSID ..since flow is constant ..and therefore mean velocity must be the same. Since we have a pressure drop ..this must be where the localized velocity increase occurs.

Let me simplify this. Suppose instead of a filter ..you merely install ONE equivalent orifice?? What MUST happen for a fluid that is moving at a constant volume ..yet moves through variable diameter conduits?? It changed velocity. Same as it does going from a 4 lane to a 2 lane and then back to a 4 lane again (btw- through osmosis, I've learned that orifices aren't linear flow/resistance characteristics - but that doesn't alter the premise here - thanks to my pals here).

Now just swap in different orifices to demonstrate different filters.

My only point in this is that the restrictive nature of the filter ..or the flow potential of a filter is pretty much a moot point as long as the oil pump relief setting is not reached. They don't alter flow at all. Not one bit ..nada ..zip.

Is that better??

Ugly3 · Apr 28, 2005

Gary Allen - My only point in this is that the restrictive nature of the filter ..or the flow potential of a filter is pretty much a moot point as long as the oil pump relief setting is not reached. They don't alter flow at all. Not one bit ..nada ..zip.

I think I finally understand your point. Given that the oil pump releif does not kick in, all of the oil output of the oil pump will get downstream, either through the filter or the filter bypass.

Is that it?

Gary Allan · Apr 28, 2005

Yep. So, and with that one qualification, there should be no such thing as a high or low flow filter ..just how long one will last without being rendered useless by being perpetually in bypass.

IMHO

XS650 · Apr 28, 2005

quote:

Originally posted by Winston:

The pressure drop across the filter will be proportional to the flow squared, then once the relief valve starts to open the curve will flatten out dramatically.

Pressure drop across a filter element in much closer to directly proportional to flow than to the square of the flow.

Winston · Apr 28, 2005

Eureka. I understand now. That post by Ugly and Gary cleared it all up for me, and I agree with both of you completely.

However, if you go back to the original purpose of Bob's test (10 pages and over 350 posts ago). The purpose of the test was;

quote:

Guys, (and gals), Joe, and myself, all have been working to put a flow test together on filters. .... Remember, the better the filtration, the lower the flow? At least that was my theory, so to somewhat establish this, here is what I did.

He was trying to measure the flow restriction of filters. I was trying to massage his data so we could make meaningfull comparisons between the filter's flow restriction.

In the end, I agree with you guys and filter restriction alone is not a very relevant issue.

Winston · Apr 28, 2005

OK. So Gary established something very significant with his last post.

quote:

there should be no such thing as a high or low flow filter ..just how long one will last without being rendered useless by being perpetually in bypass.

This brings up a couple issues to me. Let me bring up one;

Gary's tests have shown that a filter is in bypass mode during initial cold starts. Wouldn't that mean that an engine will get more flow during this time if the filter bypass was at a lower pressure. (some filters have a bypass set at 8-10 psi and some at 12-15 psi) So, if I have an engine that makes a lot of lifter/bearing noise at start-up wouldn't it be helpfull to pick a filter with a lower bypass valve setting?

Gary Allan · Apr 29, 2005

You bring up a good point here, Winston. Keep in mind, however, that unless I had the oil pump in relief, I didn't see any "bypass-like" activity out of the filter.

My current feelings are that when the pump is in relief ...then the resistance (or flow potential) of the filter can play a role in "flow". This is the reason for my "qualification" ..since unless the pump was in relief, and without regard to viscosity, the same 3-4 PSID was realized. 10-30, 15w-40, 20w-50, 5w-20 ..all had 3-4 PSID whenever the pump wasn't in relief.

When the pump relief is open ..then comparative resistance allows the flows to diverge proportionally. You're no longer in a "flow-dictating" model ..but a "resistance-dictating" model.

(more redundancy to compensate for being "challenged")
When the relief is close ..all things are subordinate to flow. All apparent pressures are a result of that being "immutable" in any equation. Everything else MUST conform in line with flow being "king". Everything else has to be manipulated from that standpoint (in view and perhaps in your case- calculations).

OTOH

With the relief open ..that flow is no longer "immutable" and resistance (apparent or otherwise) is the "non-maskable" variable that DOES impact flow.

Again, bear with my in my "slowness" (my inability to cash the conceptual checks that I'm writing here ..so to speak

).

In practical terms, for the brief time that oil pump relief is open, then you may be concerned about the resistance that a filter may present to the flow. Those who have potential for being in relief often, in more severe climates, using way too heavy an oil, may want to consider this in their choices. In my tests, we don't know what a low resistance filter would show under the same conditions. It may react differently ..it may not. This is just my reasoning..and therefore not proven.

Ugly3 · Apr 29, 2005

Gary, your oil pump releif valve should be open any time you are running above some RPM (generally between 1500 and 3000). So if you put the van in 2nd gear and ran 50 or 60 MPH the engine would be above the 3000 RPM mark and the releif would be in bypass. Bet the filter does not go into bypass then either.

Gary Allan · Apr 29, 2005

My oil pump is in relief anytime it is over 82 psi. It never is after warm up ..or most times with most oils. This may vary in other engines with other oil pumps. It does with my wife's jeep. In that the pressure relief is open at anything off idle (above about 1300 rpm). In my 02 SE 4 banger ..my oil pressure only varies between 46+/- to a low of 40+/- ..so I would imagine that it too is in pressure relief very early in the rpm band. But these are measurements after the filter ..hence I have no idea whether the filters are in bypass or not. I also have no idea what a gauge reading looks like on my Caravan. I only know what the filter see's. For all I know my idiot light reading is more stable and greatly reduced (if I put a gauge there).

but .. I'll give it a try in 2nd gear and see how high the pressure goes and see how the differential looks ..fwiw

02StangGT4Foam · Dec 17, 2005

Is this experiment still ongoing? Just wondering. This is what turned me on to BITOG. Thanks.

Gary Allan · Dec 17, 2005

Bob has retired from active duty, so to speak. He sold the board to Tony (59 Vetteman). Bob checks in from time to time.

02StangGT4Foam · Dec 17, 2005

Thanks Gary for the info. That's too bad that this test hasn't continued. It is extremely informative and at the time was (and probably now still is) the most comprehensive examination into the internals of various oil filters. I hope Tony can keep it going. Thanks again.

Gary Allan · Dec 17, 2005

This may interest you.

Here's where I'm at now

I've run into a few snags here and there..

02StangGT4Foam · Dec 18, 2005

Gary,
The pic is kinda dark to really see anything but what I think are digital gauges. Is this your new setup much like Bob had? If it is I'd love to see it in light. Thanks.

BOBISTHEOILGUY FILTER TESTS

Similar threads