- Joined
- Jul 22, 2023
- Messages
- 147
Can a high HTHS ever be detrimental to engine protection?
- Thread starter ZZman
- Start date
The curve is shaped that way because the oil pump is in pressure relief above 800 rpm, but isn't in pressure relief below 500 rpm. This sharp knee-point in the curve is where pressure relief starts occurring.
Here are some flow curves from Melling for some oil pumps for GM V-8's. These flow curves match the pressure curves from your Z06 quite well. Your oil pressure-rpm curve is shaped the way it is primarily because it's similar to the oil pump's flow-rpm curve.
Bearing self-pumping and factors that affect flow-dP linearity will have an effect as well, but they're all very minor relative to the effect that the pump's flow rate has on oil pressure.
The two Melling pumps from the figure above have PRV settings of only 43 psi, which is higher than the OEM setting. Others are rated at only 20 psi or 33 psi. These GM V-8 engines are almost always operating with oil pump outlet pressures higher than the PRV setting, typically even at idle.
On engines like this, the oil pressure gradually increasing to ~double the PRV setting is by design. If the PRV had too little restriction, the oil pressure would be too low for the bearings at high rpm.
I highly doubt the pump is in pressure relief when the PRV is set to crack open at 70 PSI. How could the pump be in pressure relief at 800 PSI with hot oil. A flow of 4 GPM isn't enough oil flow to create 70+ PSI on the oiling system/pump output to make the PRV crack open.The curve is shaped that way because the oil pump is in pressure relief above 800 rpm, but isn't in pressure relief below 500 rpm. This sharp knee-point in the curve is where pressure relief starts occurring.
Here are some flow curves from Melling for some oil pumps for GM V-8's. These flow curves match the pressure curves from your Z06 quite well. Your oil pressure-rpm curve is shaped the way it is primarily because it's similar to the oil pump's flow-rpm curve.
Bearing self-pumping and factors that affect flow-dP linearity will have an effect as well, but they're all very minor relative to the effect that the pump's flow rate has on oil pressure.
The curves in that Melling graph shows the flow jumping from 0 to 500 RPM instantly. If they could recorded data in microseconds during that short time period, you'd see a more defined initial start-up flow curve between 0 and 500 RPM. The flow obviously isn't on the same linear trend line with the rest of the graph in that RPM zone
The curves also show a near linear flow output vs RPM as expected for a PD pump, and they also show a distinct knee-over at high RPM when the PRV starts to open. The stock pump behaves linear flow output vs RPM too until it starts cavitation at high RPM.
"Bearing self-pumping and factors that affect flow-dP linearity will have an effect as well, but they're all very minor relative to the effect that the pump's flow rate has on oil pressure."
Bearing self-pumping (or any other increasing flow resistance reduction with RPM) can make an upward arcing RPM vs P curve, which you would see with a fixed flow resistance, straighten out and roll over. You need to do some RPM flow vs P plots of various components to see that curve shape for any fixed flow resistance - the flow vs P curve will arc exponentially upwards as the pump flow increases linearly like shown in the Melling flow vs RPM graph. Looking at pump flow vs RPM is not the same curve shape and looking at flow vs P through a flow resistance that the pump is forcing oil through.
When you plot the flow vs RPM curve through a fixed resistance, the RPM vs P curve will not be linear, but will exponentially increase with linearly increasing RPM pump flow. The only thing that will make the flow vs P curve straighten out to look semi-linear is an increasing reduction of the system flow resistance as RPM increases. A reduction of flow rate due to pump slip or even the PRV opening (which is reducing pump output to the engine) will only lower the curve on the graph while keeping the same curve shape. Do a flow vs P graph for an orifice for instance, or even a straight pipe. Even if the flow rate goes down, the shape of the flow vs P curve doesn't change (ie, it doesn't flatten out).
Rotating journal bearings decrease the flow resistance to the pump as their speed increases. I'm not buying the PRV being open at 800+ RPM because I think the RPM vs OP curve roll over would essentially stop or even roll downward some when the pump flow knees over noticeably like shown in the Melling flow vs RPM graph at around 6500 RPM. And besides, decreasing the flow volume doesn't change the shape of the RPM vs P curve through the flow path resistance - a change in flow will only raises or lower the curve on the graph axes. A change in the flow path resistance is what changes the shape of the flow vs P curve.
Don't know where that table came from with the low PRV settings, but it doesn't align at all with the info in this Melling video. I'll call Melling tomorrow and see it they can clarify what the real story is.The two Melling pumps from the figure above have PRV settings of only 43 psi, which is higher than the OEM setting. Others are rated at only 20 psi or 33 psi. These GM V-8 engines are almost always operating with oil pump outlet pressures higher than the PRV setting, typically even at idle.
On engines like this, the oil pressure gradually increasing to ~double the PRV setting is by design. If the PRV had too little restriction, the oil pressure would be too low for the bearings at high rpm.
Last edited:
That part of the curve is linear, but the curve as a whole from zero rpm is not. This is true regardless of what the shape of the curve from 0-500 rpm actually looks like. There has to be a knee-point there even if we can't see the exact details of it. At high rpm, the flow could be levelling off with the Melling pumps due to pump slip and other factors.
The curve for an ideal PD pump is a straight line that intersects zero, where flow rate is directly proportional to pump speed. Flow rate per revolution for an ideal PD pump should be a constant. The OEM pump at 500 rpm is shown to flow 8.0 GPM / 1000 rpm, whereas at 6200 rpm it flows only 1.1 GPM / 1000 rpm, so nowhere close to an ideal PD pump.
The pump info came from here. There's a link to the application chart pdf if you scroll down.
I'd ask Melling about how they define the PRV pressure and whether it's consistent for everything they publish. It may not always be the initial opening pressure of the valve. I'm curious about the diameter and thickness of the gerotor as well, since 4 GPM is a lot of flow for 500 rpm and would require quite a large pump unless it's geared to spin faster than the crankshaft speed.
This is a plot of oil flow through a fixed flow resistance. The orange line is what happens when the flow is reduced through the same flow resistance - like due to linear rate pump slip going from 5% to 20% in this example. It's the same flow vs P curve because the flow resistance is constant, so the same flow gives the same P. But the orange curve is just a lower max pressure due to the flow being reduced due to pump slip. You still get the same flow vs P curve - they lay on top of each other. The "GPM No Slip" column would be the flow from a perfect PD pump with no slip as RPM increases.
An increasing PRV opening would also lower the curve in a similar manner. It doesn't flatten out or roll the curve over, it just lowers the max P on the graph. The only thing that will flatten out or roll the curve over as the flow increases is a reduction in the flow path resistance. If you do some plotting you will see the behavior. Do some plots comparing flow vs P for a fixed resistance, then for the same flow rates through a resistance that decreases as the flow increases. The knee in the curves below at around 2.0- 2.5 GPM is from the flow going from laminar to turbulent flow.
An increasing PRV opening would also lower the curve in a similar manner. It doesn't flatten out or roll the curve over, it just lowers the max P on the graph. The only thing that will flatten out or roll the curve over as the flow increases is a reduction in the flow path resistance. If you do some plotting you will see the behavior. Do some plots comparing flow vs P for a fixed resistance, then for the same flow rates through a resistance that decreases as the flow increases. The knee in the curves below at around 2.0- 2.5 GPM is from the flow going from laminar to turbulent flow.
Last edited:
I plotted the OEM LS pump curve from the Melling graph, then also calculated what the flow vs RPM would be based on the 0.96 in^3 per revolution volume output spec for the standard OEM flow M295 with 100% pump efficiency (zero slip). Then applied an assumed 95% efficient pump at 500 RPM and the pump efficiency decreasing to 80% at 6500 RPM. The annotations are on the graphs.
It just seems crazy that a PD pump would be over designed that much to put out so much more flow than what the engine is supplied. That would mean the pump is basically relieving most of it's available flow output in the higher RPM range - the flow difference between the blue and orange & gray lines gets crazy large near redline. The blue curve between 0 and 1000 RPM is out of bed with what the ideal output flow is based on the volume per rev spec.
Last time I called Melling, I was asking the guy about that Melling graph (I had that one saved a long time ago), but he couldn't tell me how the pumps were tested to generate that data. Maybe I can get someone that knows more about it. I'm going to ask if these pumps are basically in relief above ~1200 RPM as seen in my graph. That would be a crazy way to over design a PD in my opinion. I'm sure the pump rotates at the same RPM as the crankshaft, but I'll ask to make sure to verify if my orange and gray lines are accurate.
Anyway, on a side note ... even if the pump actually is in pressure relief not much above an idle, and the pump output volume vs RPM is like shown and going through a fixed resistance, the RPM vs P curve will still be increasing exponentially like in post 65. The engine's oiling system would still need to have a decreasing flow resistance as the RPM increased in order for it to roll over and look like my Z06 RPM vs P curve. If this is accurate, the pump flow rate only goes from 4 to 7 GPM between idle and 6500 RPM, and with a small flow change like that the cumulative journal bearing side leakage at higher RPM could certainly have a more profound effect on the oil supply pressure as seen on the oil pressure gauge.
PS - My journal bearing calculations for all the crank, rod and cam bearings (close to the dimensions of the LS), shows a total combined side leakage of 1.04 GPM of flow at 5400 RPM. The bearing side leakage from the self pumping aspect essentially acts like a variable flow mini-scavenging pump on the supply pressure that causes it to decrease as the RPM increases.
Good discussions and appreciate your inputs.
Last edited:
I can think of a few reasons why a oiling system might be designed this way.
An engine will require a certain minimum oil pressure regardless of engine rpm, for hydraulic devices like VVT components or chain tensioners. The figure below provides a good example of oil pressure requirements of an engine, with minimum required pressures in blue, and excessive pressure/flow shown in the yellow areas. On this Toyota engine, minimum acceptable main gallery pressure at idle is said to be only 7 psi to ensure proper functioning of the VVT system. Main bearings were also said to be a limiting factor. I'm not sure if the GM engines would require higher minimum pressures. Is there something about pushrod valvetrains that would require high pressure or flow at idle?
An advantage of the GM design is that oil pressure above idle rpm will not be very sensitive to changes in viscosity. When the oil gets very hot, oil pressure at the bearings will only be reduced by a small amount, which would help prevent cavitation and bearing failure. It should also reduce the effect of oil pressure drops from oil aeration. On a engine like a Subaru, oil pressure can quickly get marginal at high rpm when oil temperatures start to get extreme, since oil pressure will be almost directly proportional to viscosity when the pump isn't in pressure relief.
Another advantage of the GM design is that since oil pressure isn't as sensitive to oil temperature, hydraulically activated oil jets can be better controlled to start opening at the desired rpm over a wider range of oil temperatures, instead being open even at idle when the oil is cold.
Another advantage is that since the PRV setting is low, cold oil pressures will be lower. Cold oil pressures on Subarus can be in excess of 140 psi, and that's measured downstream of the filter and oil cooler. This requires designing oil passages, gaskets, seals, and oil filters to be able to handle these pressures.
The main downside of the GM design is a small reduction in engine power output and fuel efficiency. Another downside is that excessive restriction in the oiling system, like from a clogged oil filter, will result in significantly less oil flow and less pressure at the bearings.
The older GM LSx series of engines (like in the C5 Z06s) don't have any hydraulic devices like VVT components or chain tensions, and don't have piston squirters ... they are basically simple "old school" engines. Piston squirters came along with the LS7. And the valve train doesn't need some crazy high oil flow volume to stay lubricated. They have hydraulic lifters in the valve train, and that's all. So I can see why the oil pumps are relatively low flow, with the stock pump max flow only being around 7 GPM at redline. Melling higher flow pumps are mainly to keep the oil from cavitating in the pump, like the OEM pumps can do near redline.I can think of a few reasons why a oiling system might be designed this way.
An engine will require a certain minimum oil pressure regardless of engine rpm, for hydraulic devices like VVT components or chain tensions. The figure below provides a good example of oil pressure requirements of an engine, with minimum required pressures in blue, and excessive pressure/flow shown in the yellow areas. On this Toyota engine, minimum acceptable main gallery pressure at idle is said to be only 7 psi to ensure proper functioning of the VVT system. Main bearings were also said to be a limiting factor. I'm not sure if the GM engines would require higher minimum pressures. Is there something about pushrod valvetrains that would require high pressure or flow at idle?
Now that I see the ideal pump output curve based on the swept volume per rev spec (orange or gray line in my graph ion post 66 above), I can see why a non-variable output PD pump with an output like that needs to be used in order to achieve a good flow rate at idle. The "PRV" on these pumps basically acts like an "output volume control valve" that's somewhat matched to the engine oiling system, and bleed off a lot of pump flow in order to cut the output flow way back at higher RPM. I tried calling Melling today, but they already closed, so will call tomorrow, and will get their take - I think when they say the pump has a "70 PSI relief spring", that basically just means the oil pressure will be maxed of around 70 PSI near redline, it's not really when the PRV starts cracking open. The other Melling info supports that too. That's now more evident when looking at the RPM vs flow output curves I compared above in post 66.
As the Figure 1 above for variable output pumps shows, a variable flow pump is more "tailored" to control the oil pressure over the RPM range better to suit the required oiling system. I'd like to see those same pump's RPM vs flow plots, because the resulting RPM vs pressure plot after putting the pump flow through a flow resistance isn't the same curve shape as just plotting the pump RPM vs flow curve. Think I've also seen where some newer engines may use an electronically controlled valve in the system using oil pressure sensor feed-back to proportionately control (not just a valve "open or closed" control method) to keep the oil pressure in the main gallery as RPM increases to a specific profile. With the right oil pressure control system, the RPM vs P curve on a specific engine could be made to be shaped any way you want it to be as RPM increases.
See table and graph below for OP vs RPM and oil temperature - yes, the oil pressure doesn't decrease a whole bunch with oil temps.
As long as the journal bearing side leakage at high RPM is well below the the extra side leakage that the bearing oil supply pressure causes, then the bearings should survive. That extra bearing side leakage flow from the oil pressure will help keep the bearings cooler and the MOFT up if the oil film stays cooler. If the oil pressure was reduced significantly due to lots of added system resistance like a very clogged oil filter, then less flow certainly could impact the bearings.
The LS engines that have DOD and other hydraulically controlled components require the higher flow version pumps to ensure they get enough flow. That's one thing Melling points out in their pump application info.
Here's some data I took to see how the OP changed as the oil warmed up. True that the idle OP didn't change much between cold and hot oil. The OP at 212F would have probably been 34 PSI if the revs were up to 950 RPM like when measured at cold idle RPM.
Speaking of the Subaru pumps, have you ever seen the oil pump output vs RPM curve - are they 1:1 driven off the crankshaft? And hows that compare to the RPM vs OP curve taken off a running engine like I did on the Z06? Pump RPM vs flow and engine RPM vs P curves won't look the same as discussed because of the journal bearing side leakage reducing the oiling system resistance with RPM. It sounds like the non-variable pumps on Subarus have a much higher PRV setting.
OP vs Oil Temp at constant 2000 RPM.
I think the oil filter would have to be really clogged to have a major impacted on the oil flow and resulting OP. I used 5 different brands of oil filters on the Z06 and did some RPM vs OP checks and never saw any real difference in the curves, so a few PSI of flow dP difference between filters at high RPM wasn't noticed. But yes, if the filter was super restrictive, that could cut some flow to the engine as RPM increases since the PRV is already opened to some degree at low RPM, so that's one thing to note on these GM engines. Overall, for an "old school" PD oil pump design, they actually work pretty well as intended.
Last edited:
I ran my k24 with redline 5w30 and the hths of 3.8. Luckily the earthdreams engine is a oil dilution monster and I feel it probably dropped some. I did top off but my guess is I would like to keep it around the 3.0 to 4.0 hths.
BrendanC
$100 site donor 2024
our accord is a 7th gen so luckily we don’t deal with that.
BrendanC
$100 site donor 2024
she paid $800 for it at 120k miles. okay i’ll stopyeah yeah, rub it in. Saltin an open wound.
wemay
Site Donor 2023
It's impossible to have it both ways. You may not notice a significant difference but the lower viscosity, lower HTHS oil absolutely gets better mpg. Conversely, those running lower HTHS might not notice a significant difference in wear ("I drove that car for 300k mi on 0w20 w/o issue") but rest assured a higher HTHS does give better wear protection.
BrendanC
$100 site donor 2024
hand calculated 27mpg hasn’t changed coming from PUP 5w-20
wemay
Site Donor 2023
"My car lasted to 340K miles on 0w20 and my buddies to 340K miles on 5w40...yup, that means there's no wear difference"
Both are fallacies we commonly use to establish our points of view. But if there were no difference, lower viscosity oils would not exist - because that really is the only reason they are around. Billions of dollars and thousands of tests attest to this.
Last edited:
BrendanC
$100 site donor 2024
just say you’re a thin oil guy. i’ll go with higher protection.
wemay
Site Donor 2023
Furthest thing from the truth. I just know it can't be both ways, pretty simple really.
I get "thin guys" calling me a "thick oil guy" too. So at least balances out.
BrendanC
$100 site donor 2024
and i presented you with data. fuel economy has not changed. 27 MPG average. the woman gets on a highway and drives ~37 miles one way at 80mph twice a day.
It could be the main bearings in these engines that are the limiting factor then. MOFT in the main bearings seems to be lowest at low rpm, which seems to be the opposite of rod bearings, which have lower MOFT at high rpm. Or it could be due to the other advantages discussed.
The Subaru oil pump is driven at crankshaft speed. It's got a lower flow per revolution than the GM pumps, but it has a higher output rating since it's flow rating is based on the pressure being lower than the 102 psi pressure relief. It will only start going into pressure relief at high rpm when the oil is thicker than ~12 cST. If you compare the dimensions of the gerotor in the table with that of a GM pump, I'm sure the Subaru pump is smaller.
The only flow data I have for my turbo FA20 engine is at 600 rpm and 6700 rpm. It must be from a running engine since the pressures given correlate well to actual engine data in these conditions.
Flow per revolution is 12.3 L/min per 1,000 rpm at idle, dropping to 9.0 L/min per 1,000 rpm at high rpm, which is a 27% drop in flow per revolution by high rpm. Pressure per revolution is 8.5 psi per 1,000 rpm at idle, dropping to 7.0 L/min per 1,000 rpm at high rpm, which is only an 18% drop in pressure per revolution.
So that seems to indicate increasing restriction with rpm. The increasing restriction may have to do more with the restriction of oil passages than the bearings. Comparing your Z06 pressure curve with the Melling curves, there is the same relationship.
Increasing restriction at higher flow rates is what you would expect if the flow is turbulent enough. The higher the flow rate and lower the viscosity, the higher the Reynolds number and the more turbulent the flow will be. Oil filters and oil coolers have increasing restriction with flow rate, and it's the same for oil passages in general when the flow isn't laminar.
wemay
Site Donor 2023
Yes, "data" ok.
Similar threads
- Locked
