LSJr tests oil filter BP function ... "OIL FILTERS: What most people DON'T know"

[johnmyster]

The flaw in your argument hasn't changed in over a year. This curve is not performed at constant pressure. So it does not represent flow in a way that anything else can be extracted. It only shows flow vs speed on an engine. It shows the response of a system (source plus sink) but not isolating the variables that would prove additional points. The pressure curve is not isoflow and the flow curve is not isobaric and we can't use them as such. Having data and using it well are two different things.

And you can't quote pump slip as a fixed percentage like 15%. It might be that at a given speed/viscosity, but it is not 15% across the board. You've seen this chart before. More complete data for a pump would look like this:

LSJ observed higher dp across the filter element hot than cold, at a given RPM. The hot filter dp was higher because flow was higher. The flow was higher because the pump efficiency picks up as back pressure goes down. The reduction in pump slip due to pressure reduction was enough to overcome the increase in pump slip do to viscosity change.

Pump slip doesn't matter when the flow vs RPM curve is on a steep increasing slope before the pressure relief starts opening, and even after the pump goes into pressure relief. Pump slip is invisible to the system when the pump is in pressure relief.

Correct. The two have the same form. Except pump slip is always present. (The orifice never closes.) So you could use pressure relief to falsely make a statement about pump slip.

Your assertion is that cold flow across the element was lower because the pump bypass was open, which supports your long standing view that pump slip is minor/negligable for PD pumps. But LSJ didn't give us all the data, so we don't know that pump bypass was open when cold. It could've been closed and pump slip could produce the same effect.

He said as the oil was warming up the dP across the filter was increasing (ref time 8:48). I say no way on that one - something is off there. If the oil becomes thinner and the flow rate stays the same through the filter, then the dP across the media will be lower - there is no physical way it would be higher. What's going on inside the engine after the filter has no bearing on the dP across the filter. The dP across the filter is solely a function of the oil viscosity and flow rate through the media.

(constant flow rate is the wrong assumption) What if some magic (pump slip) was happening where flow increased across the system due to lowering of viscosity. As in, less viscous oil went through the engine's "leaks" more easily, and thus, the PD pump was able to pump at a higher rate (lower head pressure). That could be a physical way for it to be higher. Pump slip.

 

[ZeeOSix]

The flaw in your argument hasn't changed in over a year. This curve is not performed at constant pressure. So it does not represent flow in a way that anything else can be extracted. It only shows flow vs speed on an engine. It shows the response of a system (source plus sink) but not isolating the variables that would prove additional points. The pressure curve is not isoflow and the flow curve is not isobaric and we can't use them as such. Having data and using it well are two different things.

Why would it have to be "performed at a constant pressure". This curve shows the oil pump's performance on an engine ... which is exactly what we are talking about in this thread - an engine PD pump performance on an engine. It doesn't have to prove any additional points. It's data used well, and clearly shows that LS gerotor type pumps on LS engines can go into relief at lower RPM, even with hot thin oil. If only the oil viscosity was changed, you would see the RPM where the pump starts hitting relief shift on the graph. And knowing the swept volume per rev of those pumps vs that flow curve shows the slip is ~15% in the linear (not yet in relief) portion of the curve. Once the pump goes into relief, pump slip doesn't matter because it essentially disappears from the picture. So pump slip doesn't really matter much like you think with a PD pump on an engine.

And you can't quote pump slip as a fixed percentage like 15%. It might be that at a given speed/viscosity, but it is not 15% across the board. You've seen this chart before. More complete data for a pump would look like this:

In the realm of an engine oiling system and the PD pumps used on them, I'd say they are all around 15% slip if healthy. Again, once the pump starts going into pressure relief the pump slip doesn't even matter or have any effect on the output pressure and flow from the pump. The pressure relief is then controlling the output volume and associated pressure.

LSJ observed higher dp across the filter element hot than cold, at a given RPM.

They only showed the dP across the filter with room temperature ("cold") oil at 1700 RPM. They never revved the engine any higher with "cold" oil. If so, at what time(s) in the video they did that, because I didn't see it.

So at what time in the video are you getting the claim in bold from - the claim that at a constant RPM the dP across the filter was higher hot than cold. Again, the only "cold" data point they have is room temperature oil at 1700 RPM (at time 6:50 and showing the filter dP was 0.5-0.6 PSI) waiting for the oil to get to full operating temperature to do the full RPM sweep runs. There's no way the dP across the filter is going to be higher with hot oil vs cold at the same RPM, even with an LS type oil pump in pressure relief because that pump still builds volume and pressure while in relief. That Melling pump graph shows a lot on how it works on an engine. LSJr testing never did any runs with "cold" oil, they only idled the engine at 1700 RPM ... therefore incomplete and misleading information.

The hot filter dp was higher because flow was higher. The flow was higher because the pump efficiency picks up as back pressure goes down. The reduction in pump slip due to pressure reduction was enough to overcome the increase in pump slip do to viscosity change.

The hot filter dP was higher because the only time they revved the engine over 1700 RPM was with fully hot oil. The pump even with pump slips sends lots of flow into the oiling system, and even though the viscosity is lower, the end result is more dP across the filter. I've said that a few times in this thread. The pump slip didn't matter because it still puts out lots of flow per rev, especially when the pump started going into pressure relief.

Correct. The two have the same form. Except pump slip is always present. (The orifice never closes.) So you could use pressure relief to falsely make a statement about pump slip.

Yes, it's always there ... but with a PD pump on an engine it doesn't really matter. The pump still puts out tons of flow. If the pump was sized 15% smaller with zero pump slip, the end result would be the same.

Your assertion is that cold flow across the element was lower because the pump bypass was open, which supports your long standing view that pump slip is minor/negligable for PD pumps. But LSJ didn't give us all the data, so we don't know that pump bypass was open when cold. It could've been closed and pump slip could produce the same effect.

I never said the cold oil flow dP was low across the filter because the pump was in relief - said is was low because the flow was low and filters aren't very flow restrictive (typically around 1/15th that of the oiling system). I'm not surprised by the filter only having 0.5-0.6 PSI of dP at those conditions.

Again, they never revved the engine above 1700 RPM with "cold" room temperature oil. Even if the pump was slightly into relief at 1700 RPM with room temperature oil (the pressure was ~80 PSI), it's still a fact that the flow going into the engine was low, and that's why the filter dP was low. If the oil pressure was 80 PSI without or even with pressure relief, that pump slip didn't matter. If it did, the oil pressure wouldn't be that high at 1700 RPM with room temperature oil.

How do you explain Jim Allen's on-vehicle filter dP testing and showing that with cold enough oil and high enough RPM the filter indeed goes into bypass quite often. What the LSJr video failed to do is test for those conditions and falsely claim that the filter will never go into bypass unless all clogged up (time 5:25 in the video). If they would have qualified their statement to say the filter wouldn't go into bypass if the idle RPM was low with oil not too cold, they might get away with making that claim.

(constant flow rate is the wrong assumption) What if some magic (pump slip) was happening where flow increased across the system due to lowering of viscosity. As in, less viscous oil went through the engine's "leaks" more easily, and thus, the PD pump was able to pump at a higher rate (lower head pressure). That could be a physical way for it to be higher. Pump slip.

Regardless of pump slip, when the pressure goes up at any fixed location in the oiling system that means the flow rate has also increased at that location if the viscosity is held constant. Pressure is a direct correlation to flow rate. Pump slip is not preventing more oil from flowing and the pressure increasing with increased RPM with the oil viscosity being constant. PD oil pump slip isn't even in the picture once the pump hits pressure relief. Pump slip at idle speeds could be an issue if the pump was really worn out, and the oil was very thin when hot. Otherwise, pump slip is much ado about nothing, and not the cause of things that are actually caused by something else.

 

[johnmyster]

They only showed the dP across the filter with room temperature ("cold") oil at 1700 RPM. They never revved the engine any higher with "cold" oil. If so, at what time(s) in the video they did that, because I didn't see it.

So at what time in the video are you getting the claim in bold from - the claim that at a constant RPM the dP across the filter was higher hot than cold. Again, the only "cold" data point they have is room temperature oil at 1700 RPM (at time 6:50 and showing the filter dP was 0.5-0.6 PSI) waiting for the oil to get to full operating temperature to do the full RPM sweep runs. There's no way the dP across the filter is going to be higher with hot oil vs cold at the same RPM, even with an LS type oil pump in pressure relief because that pump still builds volume and pressure while in relief. That Melling pump graph shows a lot on how it works on an engine. LSJr testing never did any runs with "cold" oil, they only idled the engine at 1700 RPM ... therefore incomplete and misleading information.


The hot filter dP was higher because the only time they revved the engine over 1700 RPM was with fully hot oil. The pump even with pump slips sends lots of flow into the oiling system, and even though the viscosity is lower, the end result is more dP across the filter. I've said that a few times in this thread. The pump slip didn't matter because it still puts out lots of flow per rev, especially when the pump started going into pressure relief.

This is your error. They didn't need to present data from an RPM sweep cold vs the same sweep hot. They had a 1700 rpm data point, cold, AND a 1700 rpm data point, hot. Two data points that were compared.

Why would it have to be "performed at a constant pressure". This curve shows the oil pump's performance on an engine ... which is exactly what we are talking about in this thread - an engine PD pump performance on an engine. It doesn't have to prove any additional points. It's data used well, and clearly shows that LS gerotor type pumps on LS engines can go into relief at lower RPM, even with hot thin oil. If only the oil viscosity was changed, you would see the RPM where the pump starts hitting relief shift on the graph. And knowing the swept volume per rev of those pumps vs that flow curve shows the slip is ~15% in the linear (not yet in relief) portion of the curve. Once the pump goes into relief, pump slip doesn't matter because it essentially disappears from the picture. So pump slip doesn't really matter much like you think with a PD pump on an engine.

It doesn't have to be performed at constant pressure unless you're trying to make isobaric conclusions from the flow data, which you did/have/are. The pressure data doesn't need to be isoflow unless you're trying to make isoflow conclusions from the data, which you did/have/are.

How do you explain Jim Allen's on-vehicle filter dP testing and showing that with cold enough oil and high enough RPM the filter indeed goes into bypass quite often. What the LSJr video failed to do is test for those conditions and falsely claim that the filter will never go into bypass unless all clogged up (time 5:25 in the video).

You and I both have the same objection - when a specific test case is used to support absolute statements. We know LSJ came from a race car world, dry sump systems with tank heaters. Cold function probably doesn't matter much to him. My experience with an LS engine was a 250k mile 5.3 in a GMT800. It spent plenty of time under 1500 rpm, making torque, and was probably a bit looser than it was (oil pump and engine clearances) from the factory. If I was an intern designing an LS engine for that application I'd probably oversize the oil pump to give sufficient hot pressure at low RPM. I have never questioned that a degree of filter bypass could occur under some conditions. Again, those conditions don't sound like my drive cycle. We can always setup an experiment to make a thing happen. I have never argued that filter bypass or pump bypass does not happen.

I have questioned your very absolute/simplistic statements that PD pumps are constant flow - where output is rigidly proportionate to RPM, regardless of factors like viscosity and back pressure on the pump...aka the factors that drive pump slip. (This is generally the statement you make when someone questions if a more restrictive filter element would impact flow.) Then as you have acknowledged pump slip, I have also questioned the absolute/simplistic statement that it's 15%, which oversimplifies the behavior and cannot be determined from the data presented. And both of these ignore that we live in a world with variable displacement oil pumps running around on the street.

 

[ZeeOSix]

This is your error. They didn't need to present data from an RPM sweep cold vs the same sweep hot. They had a 1700 rpm data point, cold, AND a 1700 rpm data point, hot. Two data points that were compared.

I don't think you've really watched the video very closely, or don't correctly recall what you saw. There is no place in the video where they show the filter dP with hot oil at 1700 RPM ... if you think so, at what time in the video are you seeing that.

If you watch the real time data on the "screen 7" around time 8:00 after a hot pull and they bring the engine back down to 2900-3000 RPM (the lowest hot RPM ever shown), the filter dP is 4.3-4.4 PSI, and the oil temp is 248-250F. The data is never shown there or anywhere else in the video at 1700 RPM with fully hot oil.

The bottom line is that it was an incomplete test without at least doing a RPM sweep with the "cold" room temperature oil.

It doesn't have to be performed at constant pressure unless you're trying to make isobaric conclusions from the flow data, which you did/have/are. The pressure data doesn't need to be isoflow unless you're trying to make isoflow conclusions from the data, which you did/have/are.

The Melling graphs of pressure and flow vs RPM that I posted earlier clearly shows that those pumps have enough swept volume to start putting it in pressure relief (depending on what pressure relief valve spring it has) around 2700-3200 RPM with 200F hot 5W-30 oil.

BTW, that isoflow graph you posted earlier shows that for a fluid with a dynamic viscosity of 100 Centipoise or more, the pump slip correction doesn't even apply. The main isoflow lines on that graph are for fluid viscosity of only 1.0 cP (they call it CPS - same viscosity unit). So when the LSJr video shows there was 80 PSI with the room temperature oil at 1700 RPM, then the pump very well could have been somewhat into relief, depending on the pump on that engine which could be higher volume than a stock pump. If it was in relief at all, then any "pump slip" is out the window and zero effect. If it's not in pressure relief at all, then there isn't enough pump slip to make any difference. Since the oil is "cold" (room temperature) the oil is thicker than at 250F, and therefore any pump slip would be less than after the oil has reached 250F.

You and I both have the same objection - when a specific test case is used to support absolute statements. We know LSJ came from a race car world, dry sump systems with tank heaters. Cold function probably doesn't matter much to him. My experience with an LS engine was a 250k mile 5.3 in a GMT800. It spent plenty of time under 1500 rpm, making torque, and was probably a bit looser than it was (oil pump and engine clearances) from the factory. If I was an intern designing an LS engine for that application I'd probably oversize the oil pump to give sufficient hot pressure at low RPM. I have never questioned that a degree of filter bypass could occur under some conditions. Again, those conditions don't sound like my drive cycle. We can always setup an experiment to make a thing happen. I have never argued that filter bypass or pump bypass does not happen.

If you read the threads about the on-vehicle filter bypass testing done by Jim Allen, you would see how possible it is to make oil filters go into bypass given enough engine RPM with cold thick oil. As I've always said, keep the engine RPM down until the oil warms up pretty good to keep the filter out of bypass.

If the LSJr test would have done a full RPM run with the oil at room temperature, I'd bet that filter would have hit bypass dP. The oil pump would have been in pressure relief, but as the Melling pump graph shows the pressure and flow volume would still increase with RPM and therefore the filter dP would also keep increasing with higher RPM.

I have questioned your very absolute/simplistic statements that PD pumps are constant flow - where output is rigidly proportionate to RPM, regardless of factors like viscosity and back pressure on the pump...aka the factors that drive pump slip. (This is generally the statement you make when someone questions if a more restrictive filter element would impact flow.)

I've talked about pump slip way before you ever popped in randomly in past threads and this one to try and discount my views. But just like now, I'm still saying it's a factor that really doesn't matter unless the pump is very worn out and the slip is excessive. And pump slip is really only going to show up more when the oil if very hot and thin, like at hot idle. It's totally out of the picture when the pump is in pressure relief.

When guys come here and claim that "Filter X" made the oil pressure do way down at an idle, then that could be a case of a worn out oil pump being sensitive to a change in filter flow restriction because the pump is worn out. Most of those cases are vehicles with tons of miles on them. I've also made it clear that the flow will basically remain the same vs RPM if the pump is not in pressure relief. If the pump is in relief, the output volume and pressure can still increases with more RPM on a simple spring loaded PRV as seen in the Mellilng graphs. I've made many statements over the years with qualifiers, but you seem to have latched on to thinking I just blurt out some "absolute/simplistic" statements, which isn't true as if that was the case we wouldn't be this far into the "technical bickering" about PD oil pumps, lol.

Then as you have acknowledged pump slip, I have also questioned the absolute/simplistic statement that it's 15%, which oversimplifies the behavior and cannot be determined from the data presented.

The Melling pump graph posted earlier, when uses along with the known swept volume per revolution of that pump, shows that pump has ~15% pump slip, and that's with 5W-30 at 200F. If the oil was at room temperature the pump slip would be less of course. I highly doubt healthy gerotor PD oil pumps on engines are sporting pump slip much worse than 15% with hot oil ... maybe 20% max. The are certainly not 30%, 40%+ unless worn out or total junk off the factory floor or the oil is 0W-8 at 300 F.

And both of these ignore that we live in a world with variable displacement oil pumps running around on the street.

Variable PD oil pumps, and the various spring and/or ECU type controls used on them is a whole world in itself, and makes discussions like this even more complicated.

 

[oil pan 4]

Ima run a larger filter with a higher pressure bypass and not worry about it.

 

[johnmyster]

I don't think you've really watched the video very closely, or don't correctly recall what you saw. There is no place in the video where they show the filter dP with hot oil at 1700 RPM ... if you think so, at what time in the video are you seeing that.

You're right, the entire discussion from 8:15 to 9:20 didn't show us the data, so we don't get to draw any conclusions from it. Data fakers, all of them.

 

[ZeeOSix]

You're right, the entire discussion from 8:15 to 9:20 didn't show us the data, so we don't get to draw any conclusions from it. Data fakers, all of them.

The test could have been done better. It tells people nothing about how the filter would react in terms of bypassing if the oil was colder and thicker. And they "conclude" (at time 5:25) that the filter would not go into bypass with cold oil, or ever go into bypass unless it was totally clogged, which isn't a very true statement. Maybe they will redo that test down the line with a full RPM sweep with at least room temperature oil. Maybe they didn't want to risk engine damage doing so ... who knows.

 

[ZeeOSix]

Guess I'm not the only one mentioning they didn't do a complete test by not showing what happens with colder and thicker oil. Comment from LSJr in that video below.

 

[CKolo]

I don’t know about you, but I don’t want this discussion to end!
Although it’s probably run it’s course. Thoroughly enjoyed the debate and totally “nerding out” and visualizing all the principles of fluid dynamics occurring in the engine lubrication system.

I’m an Engine trainer now in my career and my colleagues and I are always striving to learn more; never pretending to know everything and I have to say I picked up some really valuable tidbits from the both of you. Well done.