Known best filter for subie FA engines?

[Fa24wrx]

Have to look closer at the context of that statement. I think what it means is that using an aftermarket part will not void the entire warranty on the vehicle. In other words, using a Fram filter wouldn't void the warranty on the differential, or the warranty on the wiper motor, etc. That's the whole crux of the Magnuson-Moss warranty act. But, if an aftermarket part was proven to be the cause of the failure and damage, the dealership certainly could deny the factory warranty.

Yea, I get that. I guess the main thing is: could subaru claim this filter caused an issue. I guess the safest thing would be use oem filter.

 

[ZeeOSix]

Yea, I get that. I guess the main thing is: could subaru claim this filter caused an issue. I guess the safest thing would be use oem filter.

Technically, they only could claim that if they can prove that the filter was the root cause of the failure. That goes for any aftermarket filter or part used during the warranty period.

 

[Fa24wrx]

Technically, they only could claim that if they can prove that the filter was the root cause of the failure. That goes for any aftermarket filter or part used during the warranty period.

Some days I just wanna say f*** warranty and get an intake, tune, and send it.

 

[ZeeOSix]

I wouldn't worry about using the Endurance during the factory warranty period.

 

[Fa24wrx]

I wouldn't worry about using the Endurance during the factory warranty period.

Meaning you wouldn't use it or you would use it and not worry about it?

 

[twX]

Simply let it warm up and get the oil thin before you wail on it, and bypass pressure is irrelevant for all reasonable purposes.

Many aftermarket filters on Subarus will bypass even with a warm engine. Here are some filter dP calculations for an FA20 turbo engine. The filter restriction data is from either the Purolator data sheets, Ascent Filtration dP testing, or Brand Ranks dP testing. The scenario modelled is close to a worst case for bypassing, with the oil just thick enough for the oil pump to start entering pressure relief (0W-20 at 85-90°C), and at high rpm when oil flow is at a maximum. With much colder oil, dP will be similar, but at much lower rpm.

All of the FRAM filters (shown in red) have a dP above their bypass ratings. The Extraguards are especially bad and could be bypassing continually in most operating conditions. The Purolator filters are all well below their bypass ratings.

The restriction of the FE7317 would be similar to the XG3593A in the table below, at 15.7 psi, and its bypass opens at only 11 psi. Your naturally aspirated Subaru engines have lower oil flow rates, which would lower the dP a bit, but it still would be borderline at best, even without any filter loading.

Now, I don't think some semi-regular bypassing is going to cause any issues in normal operating conditions. Most engines should live a long life with only bypass filtration. However, if the oil is heavily contaminated with a lot of large particulate, it could lead to excessive wear or clogged oil passages or screens. There are reasons that engines are designed with full flow filtration these days.

 

[Bellavita]

I’m on my 9th Subaru. I can tell you from over 800k total miles between 7 EJs, an FA24DIT and an FB25, that the whole internet lore about bypass pressure and filtration is completely overblown. I’ve used Wix. I’ve used Fram. I’ve used Purolator. I’ve used Napa. Not once have I had an engine issue related to filtration, and I’ve never lost an engine either.

All of my past EJs were sold in good running condition, and I still have 1 EJ, 1 FA24, and an FB25. They ALL get the FE7317. I prefer the multilayer synthetic filtration media along with robust wire backed construction. I had used the XG7317 until First Brands cheaped out on it.

Simply let it warm up and get the oil thin before you wail on it, and bypass pressure is irrelevant for all reasonable purposes. Once we clear that hurdle, there is a clear advantage (as @ZeeOSix has multiple threads on) that higher efficiency filters keep oil cleaner, and cleaner oil leads to lower wear and longer engine life. The OEM filter is not exactly a paragon of efficiency.

Well stated, thank you.

 

[SubieRubyRoo]

Many aftermarket filters on Subarus will bypass even with a warm engine. Here are some filter dP calculations for an FA20 turbo engine. The filter restriction data is from either the Purolator data sheets, Ascent Filtration dP testing, or Brand Ranks dP testing. The scenario modelled is close to a worst case for bypassing, with the oil just thick enough for the oil pump to start entering pressure relief (0W-20 at 85-90°C), and at high rpm when oil flow is at a maximum. With much colder oil, dP will be similar, but at much lower rpm.

All of the FRAM filters (shown in red) have a dP above their bypass ratings. The Extraguards are especially bad and could be bypassing continually in most operating conditions. The Purolator filters are all well below their bypass ratings.

The restriction of the FE7317 would be similar to the XG3593A in the table below, at 15.7 psi, and its bypass opens at only 11 psi. Your naturally aspirated Subaru engines have lower oil flow rates, which would lower the dP a bit, but it still would be borderline at best, even without any filter loading.

View attachment 217531

Now, I don't think some semi-regular bypassing is going to cause any issues in normal operating conditions. Most engines should live a long life with only bypass filtration. However, if the oil is heavily contaminated with a lot of large particulate, it could lead to excessive wear or clogged oil passages or screens. There are reasons that engines are designed with full flow filtration these days.

Your response, on paper, makes some sense in absolute worst conditions. I don’t see any real-world correlation though; just like with the 0w20 hysteria, the mountains of failed engines have never materialized. However, considering that the oil temps in my Ascent are 198-210*F when warmed up, there is less risk than implied, because I mentioned to let the engine/oil get to operating temp before really putting the coal to it.

I’ve also read probably every single Subaru filter thread (after the third one, out of morbid curiosity alone; how board members can continually have this discussion at least 3x/month and kschachn only occasionally posts all the relevant threads) and I understand and agree with the underlying physics. BUT, aftermarket filters probably outsell OEM filters at 15:1 or better. That would imply tens of thousands of worn out/failed engines if your claims applied universally.

There’s no shift in wear metals or failure rates between people who say they only use OEM vs those who use aftermarket; so if we want to be completely fair and factual, we can say poor OEM filtration does not impact overall engine health, and that low aftermarket bypass valves do not, either.

Me, since I don’t do dumb things when the engine is cold, and therefore, by the time the engine is warm it has been filtered a couple hundred times on every startup, feel the risk of bypass for a tiny fraction of total operating time is far outweighed by allowing a greater percentage of particulate to continually circulate. Cleaner oil 99%+ of the time is going to outweigh a temporary circulation of particles while the bypass is open, and as soon as it closes, those particles are then captured by the more efficient filter.

You’re certainly welcome to use OEM; if anything, every additional data point on either side of the equation (OEM vs Aftermarket) simply further solidifies the fact that there’s no detriment to using aftermarket filters with lower bypasses.

 

[ZeeOSix]

Many aftermarket filters on Subarus will bypass even with a warm engine. Here are some filter dP calculations for an FA20 turbo engine. The filter restriction data is from either the Purolator data sheets, Ascent Filtration dP testing, or Brand Ranks dP testing. The scenario modelled is close to a worst case for bypassing, with the oil just thick enough for the oil pump to start entering pressure relief (0W-20 at 85-90°C), and at high rpm when oil flow is at a maximum. With much colder oil, dP will be similar, but at much lower rpm.

All of the FRAM filters (shown in red) have a dP above their bypass ratings. The Extraguards are especially bad and could be bypassing continually in most operating conditions. The Purolator filters are all well below their bypass ratings.

The restriction of the FE7317 would be similar to the XG3593A in the table below, at 15.7 psi, and its bypass opens at only 11 psi. Your naturally aspirated Subaru engines have lower oil flow rates, which would lower the dP a bit, but it still would be borderline at best, even without any filter loading.

Now, I don't think some semi-regular bypassing is going to cause any issues in normal operating conditions. Most engines should live a long life with only bypass filtration. However, if the oil is heavily contaminated with a lot of large particulate, it could lead to excessive wear or clogged oil passages or screens. There are reasons that engines are designed with full flow filtration these days.

One thing that is unknown is what's the flow vs dP of the OEM Subaru oil filter. I have not seen any test data showing what it is.

After diving deeper into the LS oil pumps, I don't think the Subaru oil pumps are totally out of relief until 102 PSI. The swept volume of the Subaru pump is too large IMO, and I doubt it's only 73% efficient at 6700 RPM. With the GM LS pumps, when they say it has a "70 PSI relief" they basically mean the oil pressure at the sensor will be limited to 70 PSI near redline on an LS engine. The PRV starts opening way before ~70 PSI on LS pumps - it's not the PRV cracking open pressure (you know from out previous pump discussions). If you assume the pump is totally out of relief at idle with hot oil (which I'm sure it is), that means the swept volume is 0.751 in^3/rev. If the pump was 100% efficient (no slip) it would be putting out 21.78 GPM at 6700 RPM. But per the Subaru specs, it puts out 15.9 GPM. So IMO, the pump is in relief before 102 PSI on its outlet. Can you find any technical info saying the Subaru pump actually starts cracking open at 102 PSI? The term "working pressure" in the manual could simply mean the oil pressure is trying to be limited to 102 PSI. And yes, the actual output pressure can go somewhat above the designed limit pump pressure. Guy at Melling verified that, saying even though a pump has a "70 PSI relief" (with the PRV actually cracking open at 33 PSI), some operating conditions could make the oil pressure go over that by 10-15 PSI, especially with cold thicker oil.

Without seeing an actual engine RPM vs oil pressure curve (with know oil temperature and viscosity) with more than two data points (600 and 6700 RPM), it's a guess on exactly where the pump starts relieving flow. Since the pressure per RPM drops off with increased RPM, the pump is most likely is some level of relief. If it wasn't, the oil pressure would be arching upwards quite a bit with increased RPM, even though the journal bearings are bleeding off some of the oil supply.

Note - your 63.2 nominal pump flow is in quarts. It's actually 60.2 L/min per the Subaru specs.

 

[ZeeOSix]

Here's a Melling 10295 LS pump RPM vs Flow and RPM vs Pressure through a fixed orifice that basically simulates an LS engine. I ran the same setup through the on-line dP calculator and it was very close to the curve before the pump went into relief. This Melling pump has the GM OEM swept area of 0.96 in^3/rev.

You can see exactly where the PRV starts cracking open - where the RPM vs Flow curve starts kneeing over and where the RPM vs Pressure curve stops arching upwards, and starts rolling over. The PRV that cracks open at 43 PSI limits the max oil pressure to 70 PSI (the "70 PSI relief spring" as they call it), and the PRV that cracks open at 53 PSI limits the max oil pressure to 80 PSI (the "80 PSI relief spring" as they call it).

I think the Subaru oil pump would show something similar if it actually starts going into some level of pump bypass before 102 PSI. If the Subaru pump actually put full output before the pump's PRV started cracking, the oil pressure would not roll over, but instead increase exponentially like shown in these graphs before the PRV started cracking open.

 

[twX]

After diving deeper into the LS oil pumps, I don't think the Subaru oil pumps are totally out of relief until 102 PSI. The swept volume of the Subaru pump is too large IMO, and I doubt it's only 73% efficient at 6700 RPM. With the GM LS pumps, when they say it has a "70 PSI relief" they basically mean the oil pressure at the sensor will be limited to 70 PSI near redline on an LS engine. The PRV starts opening way before ~70 PSI on LS pumps - it's not the PRV cracking open pressure (you know from out previous pump discussions). If you assume the pump is totally out of relief at idle with hot oil (which I'm sure it is), that means the swept volume is 0.751 in^3/rev. If the pump was 100% efficient (no slip) it would be putting out 21.78 GPM at 6700 RPM. But per the Subaru specs, it puts out 15.9 GPM. So IMO, the pump is in relief before 102 PSI on its outlet. Can you find any technical info saying the Subaru pump actually starts cracking open at 102 PSI? The term "working pressure" in the manual could simply mean the oil pressure is trying to be limited to 102 PSI. And yes, the actual output pressure can go somewhat above the designed limit pump pressure. Guy at Melling verified that, saying even though a pump has a "70 PSI relief" (with the PRV actually cracking open at 33 PSI), some operating conditions could make the oil pressure go over that by 10-15 PSI, especially with cold thicker oil.

Without seeing an actual engine RPM vs oil pressure curve (with know oil temperature and viscosity) with more than two data points (600 and 6700 RPM), it's a guess on exactly where the pump starts relieving flow. Since the pressure per RPM drops off with increased RPM, the pump is most likely is some level of relief. If it wasn't, the oil pressure would be arching upwards quite a bit with increased RPM, even though the journal bearings are bleeding off some of the oil supply.

The oil pump pressure relief specs for the Subaru engines do seem to be based on the cracking pressure of the relief valve, since oil pressure can get much higher than the pressure relief spec. For instance, a non-turbo FA20 will have an 85 psi pressure relief spec, but will allow oil pressure of over 140 psi at high rpm (with an oil temperature of 40°C), which should equate to >160 psi at the pump outlet. The knee-points on oil pressure curves from Subaru engines are also consistent with the pressure relief specs being the initial opening pressure.

The test conditions for the oil pump specs on the FA20 turbo are with 120°C oil and with an oil pressure of only 46 psi at 6700 rpm, so the oil pump is definitely not in pressure relief in these conditions. Oil pump efficiency can drop off pretty rapidly at high rpm. 73% volumetric efficiency at max rpm doesn't seem abnormally low.

Here's an example of a gerotor pump that drops to ~64% of ideal flow by 6,000 rpm. At even higher rpm, the efficiency would drop so much that flow would start to decrease with increasing rpm. I suspect this is the case with the Subaru pumps since oil pressure starts to decrease as the engine approaches redline.

Note - your 63.2 nominal pump flow is in quarts. It's actually 60.2 L/min per the Subaru specs.

I added an arbitrary 5% to the oil pump output since I assumed that peak flow occurs at an rpm lower than 6700 rpm, and would be somewhat higher than 60.2 L/min. It's hard to say what the actual maximum flow rate of the pump is given that we only have the one data point at 6700 rpm. It's just a best guess.

 

[Bellavita]

Your response, on paper, makes some sense in absolute worst conditions. I don’t see any real-world correlation though; just like with the 0w20 hysteria, the mountains of failed engines have never materialized. However, considering that the oil temps in my Ascent are 198-210*F when warmed up, there is less risk than implied, because I mentioned to let the engine/oil get to operating temp before really putting the coal to it.

I’ve also read probably every single Subaru filter thread (after the third one, out of morbid curiosity alone; how board members can continually have this discussion at least 3x/month and kschachn only occasionally posts all the relevant threads) and I understand and agree with the underlying physics. BUT, aftermarket filters probably outsell OEM filters at 15:1 or better. That would imply tens of thousands of worn out/failed engines if your claims applied universally.

There’s no shift in wear metals or failure rates between people who say they only use OEM vs those who use aftermarket; so if we want to be completely fair and factual, we can say poor OEM filtration does not impact overall engine health, and that low aftermarket bypass valves do not, either.

Me, since I don’t do dumb things when the engine is cold, and therefore, by the time the engine is warm it has been filtered a couple hundred times on every startup, feel the risk of bypass for a tiny fraction of total operating time is far outweighed by allowing a greater percentage of particulate to continually circulate. Cleaner oil 99%+ of the time is going to outweigh a temporary circulation of particles while the bypass is open, and as soon as it closes, those particles are then captured by the more efficient filter.

You’re certainly welcome to use OEM; if anything, every additional data point on either side of the equation (OEM vs Aftermarket) simply further solidifies the fact that there’s no detriment to using aftermarket filters with lower bypasses.

Good Morning SubieRubyRoo,
Another well written comment by you on the history of this topic on Bitog, the reality of the volume of non oem filters used on Subarus and how good driving habits win the day for filtration. Most appreciated, thank you.

 

[ZeeOSix]

The oil pump pressure relief specs for the Subaru engines do seem to be based on the cracking pressure of the relief valve, since oil pressure can get much higher than the pressure relief spec. For instance, a non-turbo FA20 will have an 85 psi pressure relief spec, but will allow oil pressure of over 140 psi at high rpm (with an oil temperature of 40°C), which should equate to >160 psi at the pump outlet. The knee-points on oil pressure curves from Subaru engines are also consistent with the pressure relief specs being the initial opening pressure.

If that was the case, and if you could obtain RPM vs oil pressure off the engine (like I did with the Z06), then you would see an RPM vs P curve like in post 30 where the P would climb exponentially until the pump PRV started opening. I'd like to see actual collected RPM vs P on a Subaru between idle and 5000 or 6000 RPM with more than just the end data points so the shape of the curve is better seen, which would tell you if and when the pump hit relief.

The test conditions for the oil pump specs on the FA20 turbo are with 120°C oil and with an oil pressure of only 46 psi at 6700 rpm, so the oil pump is definitely not in pressure relief in these conditions. Oil pump efficiency can drop off pretty rapidly at high rpm. 73% volumetric efficiency at max rpm doesn't seem abnormally low.

Here's an example of a gerotor pump that drops to ~64% of ideal flow by 6,000 rpm. At even higher rpm, the efficiency would drop so much that flow would start to decrease with increasing rpm. I suspect this is the case with the Subaru pumps since oil pressure starts to decrease as the engine approaches redline.

Agree that the Subaru pump probably isn't in relief with 120C oil that only gives 46 PSI at 6700 RPM. If the pump has that much slip (and maybe some added cavitation like in the stock LS pump), then its pumping efficiency could fall off real fast at high RPM. The other Melling curves posted in the other thread(s) showed the RPM vs flow to be pretty linear on the stock LS pump until cavitation starts. The Melling pumps are tweaked to keep from cavitating which helps keep flow from falling off a higher RPM.

 

[ZeeOSix]

To add ... It would be interesting to see the flow vs dP curve for an OEM Subaru filter. They are relatively small, so based on the media flow performance and total media area, they could be relatively flow restrictive and part of the reason why the bypass valve is set higher - along with the high pump flow factor. If they are pretty restrictive, then even with a higher bypass valve setting, they could be bypassing oil just as much as some of the aftermarket filters that have a lower bypass valve setting.

A filter's bypass valve can't be set higher than what the media and center tube can take without causing possible damage from high dP. Also, a higher filter bypass valve setting can put the pump in pressure relief sooner if the filter gets really clogged and restrictive. A worst case example would be a severely clogged filter without a bypass, which would put the pump in relief much sooner, causing much more oil flow reduction to the engine at higher RPM.

Also @Mewbs in post 13 said he's going to verify the OEM filter bypass valve cracking pressure, so that should be interesting.

 

[twX]

To add ... It would be interesting to see the flow vs dP curve for an OEM Subaru filter. They are relatively small, so based on the media flow performance and total media area, they could be relatively flow restrictive and part of the reason why the bypass valve is set higher - along with the high pump flow factor. If they are pretty restrictive, then even with a higher bypass valve setting, they could be bypassing oil just as much as some of the aftermarket filters that have a lower bypass valve setting.

The Subaru filters are small, but they pack a lot of media area into a small can. For instance, the blue Subaru OEM filters manufactured by FRAM have over 50% more media area than the FRAM branded equivalent PH7317 (830 cm^2 vs 540 cm^2), and they use the same type of media. This should result in around half the dP at high flow rates, so around 11 psi for an OEM filter compared to the 22 psi estimated for the FRAM PH9688 in the example I posted previously. The 23 psi bypass setting may seem unnecessary if the dP is only 11 psi, but it allows for a healthy margin for increasing dP from dirt loading.

Hopefully Brand Ranks will do a dP test on an OEM Subaru filter at some point so we can get some more solid data, but I suspect the OEM filters have average to below average restriction compared to the aftermarket options.

Also, a higher filter bypass valve setting can put the pump in pressure relief sooner if the filter gets really clogged and restrictive. A worst case example would be a severely clogged filter without a bypass, which would put the pump in relief much sooner, causing much more oil flow reduction to the engine at higher RPM.

This risk is mitigated by the fact that the Subaru oil pumps have high pressure relief settings, usually 85-102 psi, so even a 30 psi drop across a clogged filter when the pump is in relief won't result in dangerously low oil pressure. On these engines, the worst case scenario for low oil pressure at the bearings will always be with very hot oil when the pump isn't in relief.

 

[twX]

If that was the case, and if you could obtain RPM vs oil pressure off the engine (like I did with the Z06), then you would see an RPM vs P curve like in post 30 where the P would climb exponentially until the pump PRV started opening. I'd like to see actual collected RPM vs P on a Subaru between idle and 5000 or 6000 RPM with more than just the end data points so the shape of the curve is better seen, which would tell you if and when the pump hit relief.

Here's a pressure-rpm curve from a stock Subaru non-turbo FA20 engine (red). The curve is slightly exponential from 800 to 2000 rpm (doesn't look this way due to improper scaling on x-axis), then linear from 2,000 to 4,000 rpm.

Here's a second curve from tests done with multiple different oil viscosities. The thickest oils show obvious signs of the pump entering relief, whereas the thinner oils do not. The two thickest grades at 185°F have a clear inflection point as the pump relief starts to open.

The shapes of the curves aren't that accurate since they're interpolated from few data points, but a video of a dP test from this same vehicle shows more clearly that the inflection point is ~55 psi. This inflection point would equate to the 85 psi initial opening pressure rating of the pump PRV, if there were a 35% drop in pressure across the oil filter, aftermarket oil cooler, and oil passages between the pump and the pressure sensor, which seems reasonable for these engines.

There's no clear sign of pressure relief in the first chart (above) despite higher maximum pressure (64 psi), but that engine did not have the added restriction of an oil cooler.

The 0W-20 at 225°F (green) starts to get pretty non-linear above 2,000 rpm, which we don't see in the first chart. With a maximum pressure of 48 psi, it should not be the oil pump PRV that is causing this. One possible reason for this is that the car from the second chart has added restriction from an oil cooler, which would have an exponential pressure-flow curve. This would reduce the pressure measured downstream at the pressure sensor. The oil cooler restriction may also be why the oil pressure drops off more at high rpm in the second chart.

It's also interesting to compare that curve to the 10W-40 at 265°F (light blue), which has a kinematic viscosity at its test temperature that is just a bit higher than the 0W-20. The oil pressures are very similar up to 2,000 rpm, but differ a lot at 4,000 rpm.

What I think might be happening is that at low rpm, shear rates are low, and neither oil is significantly shear thinned. At 4,000 rpm, shear rates are high enough that the 0W-20 is significantly shear-thinned, while the 10W-40 is not. The higher temperature of the 10W-40 means that it will be more shear-resistant, and its VI is a bit lower than the 0W-20 as well. At 7,000 rpm, shear rates are high enough that the 10W-40 is also significantly shear-thinned, and oil pressure is again similar to the 0W-20.

 

[Nmr1981]

I’m on my 9th Subaru. I can tell you from over 800k total miles between 7 EJs, an FA24DIT and an FB25, that the whole internet lore about bypass pressure and filtration is completely overblown. I’ve used Wix. I’ve used Fram. I’ve used Purolator. I’ve used Napa. Not once have I had an engine issue related to filtration, and I’ve never lost an engine either.

All of my past EJs were sold in good running condition, and I still have 1 EJ, 1 FA24, and an FB25. They ALL get the FE7317. I prefer the multilayer synthetic filtration media along with robust wire backed construction. I had used the XG7317 until First Brands cheaped out on it.

Simply let it warm up and get the oil thin before you wail on it, and bypass pressure is irrelevant for all reasonable purposes. Once we clear that hurdle, there is a clear advantage (as @ZeeOSix has multiple threads on) that higher efficiency filters keep oil cleaner, and cleaner oil leads to lower wear and longer engine life. The OEM filter is not exactly a paragon of efficiency.

Yep I’ve used the above mentioned for 120k on my EJ25, I’ve also used Fram Ultra, Wix, and NAPA gold on my FA24 with no issues. Maybe if I were running 20w50 in winter I’d be concerned but with 0w-xx I’m not losing sleep over it. It’s a Subaru not a hand built supercar engine. I figured clean and fresh is best.

 

[ZeeOSix]

The Subaru filters are small, but they pack a lot of media area into a small can. For instance, the blue Subaru OEM filters manufactured by FRAM have over 50% more media area than the FRAM branded equivalent PH7317 (830 cm^2 vs 540 cm^2), and they use the same type of media. This should result in around half the dP at high flow rates, so around 11 psi for an OEM filter compared to the 22 psi estimated for the FRAM PH9688 in the example I posted previously. The 23 psi bypass setting may seem unnecessary if the dP is only 11 psi, but it allows for a healthy margin for increasing dP from dirt loading.

Hopefully Brand Ranks will do a dP test on an OEM Subaru filter at some point so we can get some more solid data, but I suspect the OEM filters have average to below average restriction compared to the aftermarket options.

Not sure what Subaru filter p/n you're referring to, but Whip City shows the Subaru 15208AA12A to have 81 sq-in (522 cm^2) of media. Yes, actual flow vs dP with oil around 12-15 cSt on some OEM Subaru filters would be beneficial info.



These two (15208AA160 & 15208AA15A) have more media - 130 sq-in (834 cm^2). I'm assuming these are the OEM filters specified for the high volume oil pumped engines. Whip City shows the PH7317 to have 102.5 sq-in (661 cm^2) of media area.

This risk is mitigated by the fact that the Subaru oil pumps have high pressure relief settings, usually 85-102 psi, so even a 30 psi drop across a clogged filter when the pump is in relief won't result in dangerously low oil pressure. On these engines, the worst case scenario for low oil pressure at the bearings will always be with very hot oil when the pump isn't in relief.

Pressure doesn't directly correlate to flow when the viscosity or system resistance increases and causes the pump to go into relief. If the oil is very cold and thick enough to put the pump in relief at pretty low RPM, then the flow rate to the engine can be cut way back. Lack of flow volume is what's detrimental to the engine. As always, keep the RPM down until the oil warms up pretty good to ensure flow volume isn't cut back drastically.

With very hot oil and the pump out of relief, the oil pressure will be lower, but all the engine components will be getting lots of oil flow volume. An engine likes that better than very little oil flow volume at high pressure - ie, the pump far into relief and flow cut way back. Oil pressure just makes oil move to where it needs to be, and the pressure feeding the journal bearings really adds nothing to their natural hydrodynamic MOFT wedge pressure as long as the supply pressure provides adequate flow to feed the beariing's self pumpng flow rate. The high pressure in the MOFT hydrodynamic wedge is built simply by the bearing rotation.

 

[ZeeOSix]

Here's a pressure-rpm curve from a stock Subaru non-turbo FA20 engine (red). The curve is slightly exponential from 800 to 2000 rpm (doesn't look this way due to improper scaling on x-axis), then linear from 2,000 to 4,000 rpm.

That looks a lot like the Z06 RPM vs OP curve. What the curve seems to show is that regardless of the oil pump, the curves are basically on top of each other, even with the two "Killer B" pumps. I'm assuming the oil viscosity was basically the same between all 3 tests. If the Killer B pumps are higher output volume per rev, then it seems they would result in more OP vs the stock pump, just as higher output LS pumps result in more OP over the stock LS pump, as show earlier in post 30. So maybe in this case, the pumps are all in pressure relief because if the RPM vs P curves are all basically the same, then the RPM vs flow of all 3 pumps has to also be basically the same.

Here's a second curve from tests done with multiple different oil viscosities. The thickest oils show obvious signs of the pump entering relief, whereas the thinner oils do not. The two thickest grades at 185°F have a clear inflection point as the pump relief starts to open.


The shapes of the curves aren't that accurate since they're interpolated from few data points, but a video of a dP test from this same vehicle shows more clearly that the inflection point is ~55 psi. This inflection point would equate to the 85 psi initial opening pressure rating of the pump PRV, if there were a 35% drop in pressure across the oil filter, aftermarket oil cooler, and oil passages between the pump and the pressure sensor, which seems reasonable for these engines.

There's no clear sign of pressure relief in the first chart (above) despite higher maximum pressure (64 psi), but that engine did not have the added restriction of an oil cooler.

The 0W-20 at 225°F (green) starts to get pretty non-linear above 2,000 rpm, which we don't see in the first chart. With a maximum pressure of 48 psi, it should not be the oil pump PRV that is causing this. One possible reason for this is that the car from the second chart has added restriction from an oil cooler, which would have an exponential pressure-flow curve. This would reduce the pressure measured downstream at the pressure sensor. The oil cooler restriction may also be why the oil pressure drops off more at high rpm in the second chart.

It's also interesting to compare that curve to the 10W-40 at 265°F (light blue), which has a kinematic viscosity at its test temperature that is just a bit higher than the 0W-20. The oil pressures are very similar up to 2,000 rpm, but differ a lot at 4,000 rpm.

Yes, looks like all 3 oils at 185F caused the pump to start relieving some flow around 2000 RPM or slightly above, causing the pressure to roll over. With the hot oil at 225-265F (closer to the Subaru service manual pump performance specs), the extreme roll-over in pressure above 4000 RPM could be the pump falling on it's face due to pump slip (inefficient pumping). As shown earlier using the Subaru service manual specs, the pump losses a lot of efficiency with thin oil at high RPM. Could be that with the colder thicker oil, the pump slip is not nearly as bad. Also, it's possible that when the pump itself is that hot, maybe the internal clearance is also effected and that's why the slip seems to get pretty bad. To me, that's what this graph seems to indicate, and why the hot oils lost pressure badly above 4000 RPM.

On a side note - I find it strange that the Subaru service manual shows oil pump flow and resulting pressure specs at 120C (248F) since that spec is primary meant for service techs to use to ensure the pump and oiling system is operating properly. It would have been more useful for that purpose if those specs where given with oil at around 100C (212F), which would be closer to normal operating temperatures for a service tech to take measurements at.

What I think might be happening is that at low rpm, shear rates are low, and neither oil is significantly shear thinned. At 4,000 rpm, shear rates are high enough that the 0W-20 is significantly shear-thinned, while the 10W-40 is not. The higher temperature of the 10W-40 means that it will be more shear-resistant, and its VI is a bit lower than the 0W-20 as well. At 7,000 rpm, shear rates are high enough that the 10W-40 is also significantly shear-thinned, and oil pressure is again similar to the 0W-20.

The 0W-20 should have less VIIs, and therefore should have less temporary shear down inside journal bearings than a 10W-40. From what I've seen, engine OP correlates well with the HTHS viscosity and not with the Kinematic viscosity. The 10W-40 will have a much higher HTHS viscosity than the 0W-20. Also, the 10W-40 is at a higher temperature than the other two hot oils.

Journal bearing self-pumping flow rates going on at high RPM could also be effecting the OP some at higher RPM - more so if the pump isn't in relief because the PRV can't adjust for the bearing self-pumping flow trying to decrease the pump's back pressure. Lots of things are going on at the same time that can have an impact of the actual RPM vs OP seen on a running engine.

 

[sw99]

I have been using all sorts of aftermarket filters in my subi's over the years and zero issues. Tons of ecores also and they are actually one of my favorites. My guess is that a large majority of the Subaru's out there on the roads today are not running OEM filters...