Gear Shifting Theory

[Dave 53]

Got a new close ratio 6 speed in my 818 track car and pretty much immediately spun a main rod bearing from over revving it and finding out it had less then optimal bearings. Among the many lessons learned, I concluded I needed to take an analytical approach to shift points.

Please review my theory for errors in logic or math.

I found an on-line calculator that shows RPM and speed for each gear. It shows the RPM range in 3rd, 4th and 5th gear is basically 1,500 RPM.

I overlayed (by eyeball) this 1,500 RPM range over the dyno graph to maximize area under the curve. This is my old dyno curve. Perhaps we can tune the new engine around this theory.
HP is maximized between 5,200 - 6,700 RPM
Torque is maximized between 4,800 - 6,300 RPM

My rev limier is set at 7,400 RPM. The exception to these shift points would be when an up shift would only last a few seconds before having to down shift. For example, if I'm in 4th gear and hit 6,300 RPM, I'll keep it in 4th up to 7,400 RPM if that can avoid a quick (maybe 5 seconds) 4-5-4 shift.

I'll get a LED bar shift light and program it so that no light show until 4,800 RPM. If no light showing, I know I need to down shift.
1 green light at 4,800. In the torque curve
2 green lights at 5,200. In the HP curve.
1 yellow light at 6,300. Out of the torque curve.
2 yellow lights at 6,700. Out of the HP curve.
Red at 7,000.
Flashing red at 7,200.
Rev limiter at 7,400.
(This is the system I used on my shifter kart).

Now, how to drive it? Stay in the toque curve? Stay in the HP curve? Stay in the extremes, 4,800 to 6,700? Split it, 5,000 to 6,500?

In any case, the engine should be happier not seeing repeated 7,200+ shifts.

gear rpm.jpgdyno rpm.jpg

[SBC BUILDER]

So many factors in determining the "perfect" shifting point. That is the problem with using dyno numbers for shift points. Real world testing is the best thing to do. It also depends on your driving style-late or early breaking, how fast the engine revs after the shift etc,. Close ratio or wide ratio transmission. Even more so if you are going to multiple tracks with varying lengths. Very high sustained RPM is for drag racing not track racing. The choice of a camshaft/heads and induction system has to be figured out to the best possible combination. Most modern engine produce maximum torque between 5500 - 6600 RPM. Personally I would not go above 7000 RPM as you are just putting stress on the entire engine. F1 engines bearing clearance's are so tight they have to heat up the water and oil before IT WILL EVEN TURN OVER. A high volume oil pump is better than high pressure in long distance engines. It is a real science these days making the right combination for your car. I am trying out this software for tire,trans and final drive ratios.

https://www.blocklayer.com/rpm-gear

[Dave 53]

There are many factors that determine what the dyno graph looks like, but once the car is built and those factors are fixed, what are the factors that determine when to shift? We are talking about a race car looking for maximum acceleration. All I'm coming up with is rpm and am I on a section of track approaching a braking zone that I'm better off holding a high RPM for a second or two vs. doing a quick up then right back down shift. Especially when I still have another 700 rpm on the table.

So many different engine designs, it's impossible to make any generalizations. The dyno graph and when to shift a Mazda Miata will look completely different then say a Porsche GT3. But I think (am I wrong?) the goal when it comes to when to up or down shift is always the same - maximize area under the curve.

How do you do "real world testing"?

[J R Jones]

It has been stated that a successful race car is the best combination of compromises. That can be said of a successful hot lap as well.
In SCCA Nationals and Trans Am we used differential gear sets from 4.56 to 3.5 to set redline RPM on the end of main straights and no over revving. We did not change transmission gears to suit corners like pro cars.
Racing 30+ cars of near equal performance meant you were within inches of cars on all four sides. That required all one's attention and shifting was relegated to sensory inputs and muscle memory. When one could not out-accelerate a competitor you had to steal a corner by corner line position or heavy late braking to throw a competitor off his line. That may lead to a slow corner but the other guy was slower.
With a track day challenge I propose that you should chose between up shifting or feathering the throttle, but not over rev the engine.
On your metrics, I have followed Hobby Racer and Sgt Gators threads, and my observation is that the Subaru G-rotor oil pump cavitates at high RPM, probably 7000RPM. That could be part of your engine failure.
I would set the rev limiter at 7000RPM or a bit less if you are cautious. I would redline at peak power (6700RPM). Attempt to put your downshifted RPM in the torque range of 4800-5000RPM.
As a side note Hobby and Gator have used dry sumps for scavenge oil. We used four stage dry sumps with one stage (gear pump) for pressure to avoid the G-rotor pump.
jim

[Dave 53]

When I raced go karts, we played with final drive ratio just as you described. On road courses, the goal was redline in 6th gear while in the draft. Lose the draft and you were pretty much screwed because not only did you lose the draft, now you don't have the right gearing to be alone.

I went down the oil pump wormhole. My understanding is cavitation is only an issue when oversizing to a 12mm pump. If the pump is too big, the relief valve is overactive which then holds the oil in the rotor instead of flowing it through which causes the oil to overheat which causes the cavitation.

I can't find any stories of issues with a properly sized and relief valved pump. Not to say the 10mm pump I had when the bearing failed didn't have something to do with it. Wonder if it would have failed with a 11mm or 12 mm pump.... Anyway, now it's a 11mm pump with stock pressure relief.

[Bob Cowan]

Lower gears (higher numerically) multiply torque. First gear multiplies torque a lot more than 4th gear does. Overdrive gears divide torque.

HP doesn't really matter in this case, so don't worry about that.

That being said, holding it in gear after the torque band starts to fall will make you a little faster, because you're gaining greater multiplication of the torque curve.

Also, when you do upshift, the engine will be higher in the torque band when you get back on the gas. Again improving your acceleration.

Holding it in gear to near the redline before a corner entry is smart. Every shift costs you time.

[J R Jones]

Interesting, with cars and AMA motorcycle racing, we never geared for draft. We "breathed" the engine in draft expecting to do a power pass.
There was the time at Brainerd MN when the Buells chasing a Ducati had a 3 MPH deficit down the straight and through T-1. The Buell would track at 158 but the front would slide drifting the Ducati through the turn at 161 MPH. Longest black skid-mark in our experience.
BTW in my experience cavitation is the result of low pressure (suction) on a dynamic surface. As in a propeller, the backside is high pressure, pushing the water or air; the front side goes negative and that is where the cavitation damage is with water. In a G-rotor I would say cavitation results from inability to pull oil into the rotor. Bigger pump = bigger suction.
jim

[RPGs818SNA]

Dave 53,

I don’t know much about racing, but I do enjoy math problems. Needless to say, there’s a lot more that goes into shift strategy that math doesn’t cover, but I hope you find this helpful, or at least entertaining.

I think you want to apply the most torque available to the wheels, and to up shift when applied torque falls below what you could get in the next higher gear. I found a dyno chart for a stock 2006 NA 2.5L engine that went up to 7100 rpm, some 5 speed gear ratios, and assumed a 3.7 final ratio with 24” tires to plot axle torque in each gear versus speed. This shows when the torque has fallen off so far that the next gear would be better. For this engine, it’s around 6500 rpm.

Axle torque vs mph legacy.jpg

Here’s your data plotted the same way. I assumed a 3.7 final and 24” diameter tires.

Axle torque vs mph 818.jpg

Unfortunately your dyno numbers stop at 6500 rpm, before the axle torque falls below what the next gear would offer. But you can get a sense of where it’s likely to cross and map the speed back to the rpms. If you post your new engine dyno numbers, I’ll be happy to update the chart, especially if you can get a little closer to 7000.
RPG

[Dave 53]

So much interesting discussion!!!

I raced 2-stroke 80cc shifter karts on road courses and sprint tracks. We didn't have a redline per say. Instead, we had a point at which the power fell off like a light switch and the engine just wouldn't spin any faster - 13,500. Gear for clean air 95 mph and there is nothing left if the draft is 100mph. Gotta gear for 100 mph, but if you're in clean air with that gearing, top speed is 90.

I work on ships and have studied propellers. The term "cavitation" often gets loosely thrown around to describe any loss of propeller efficiency (most commonly when the propeller isn't fully submerged - that's called aeration). Even at under 100 rpm, the pressure on the front of the propeller can be low enough to boil the water into steam which causes cavitation. I have only a basic understanding of rotary pumps which is entirely different than a propeller (or centrifugal pump). As a practical matter, I was scared away from the Subaru 12mm pump because of "cavitation" stories.

RPG, I am still trying to get my head around the practical implications of torque multiplying. I will certainly share my new dyno graph with you when I get it. Thanks!

The chicken or the egg question.... (I'm certain my tuner will have an answer). How do we tune the engine to match the transmission?

Hoping to have a first start on the new engine tomorrow.

"Horsepower sells cars. Torque wins races"

[J R Jones]

I did development for Mercury Marine, including props and jet pumps. Lab testing of those devices is in chambers 12+ feet deep so that the water pressure eliminates cavitation (like a well pump). Spinning shallow or on the surface is problematic Cavitation is low pressure pulling a bubble on the blade surface. When the bubble collapses as the low pressure normalizes, the metal can be eroded. The contributing factors involved are surface speed and pitch, high pitch and high surface speed is more cavitation prone. Submarines get away with outrageous propellers because of water depth.

I looked at some Subaru oil pump tech on line and find challenges:
The pump runs at crank RPM not cam (half) speed.
The pump is large diameter with high (fps) surface speed.
The pressure relief valve discharges foamy oil into the oil sump to pump inlet feed. I suspect that is the major fault in the system. High pump RPM results in more pressure and more regulation.
Going to the bigger pump will create more volume (GPM) and higher fluid velocity in the existing passages. This will result in more pressure and more regulation, a mixed blessing.
A potential improvement would be to discharge the regulated oil into the pan, to de-air. The existing package does not favor that concept.
Gear pumps in dry sumps are less susceptible to cavitation by design.
jim

[Dave 53]

5/17/23 Blew up 5 speed transmission at Sonoma.
6/10/23 6 Speed installed.
6/11/23 Blew up motor at Thunderhill.
7/31/23 New motor tuned.
8/6/23 Laugna Seca.

Since 5/17/23, my 818 has been a full time job between the new transmission and motor. But, just got it tuned today and heading to Lauga Sunday!!! Did a few improvements that I'll share later.

The 6 speed is close ratio. 1,500 RPM between gears. Given the 1,500 RPM gear spread and the new dyno graph, my question for the forum is, at what RPM's do I up and down shift?

Dyno July23.jpg

[RPGs818SNA]

Dave,

Here’s an update to the axle torque vs mph chart for your 6 gears and your new tune. Since your torque curve ends at 6500, I estimated torque at 6600 so the lines would cross. I've assumed a 3.7 final ratio and 24" diameter tires.

Axle torque vs mph Dave's 818.jpg

Here are the up shift points that would maximize acceleration:

1 to 2, 6600 rpm or more, 35 to 37 mph
2 to 3, 6600 rpm, 57 mph
3 to 4, 6600 rpm, 72 mph
4 to 5, 6600 rpm, 94 mph
5 to 6, 6600 rpm, 118 mph

Downshift into the gear that makes your rpm between 4300 and 6600, since for most up shifts your rpm will be about 4300 in the higher gear.

These are at best initial guidelines. Actual race car drivers will be able to give much better advice than I can.

Best of luck,
RPG

[Dave 53]

RPG, Thanks for the data points! I think actual race car drives use data like this.

I've been staring at this for over an hour now and I think I get the concept and I'm excited to input my real world data. Can you share the graph generator you used?

[Sgt.Gator]

I have a slightly different approach, more in tune with SBC Builder. Dyno shift points are only a model in very specific non real world. My approach requires some form of datalogging. I have 4 total dataloggers in the 818 that each does some things really well, but none of them do all things well.

I go out and drive the track using intuitive "feel" shift points. I build speed, the right gear changes with speed. I try to get a couple of consistent laps. I watch the predictive lap timer, it gives instant feedback on whether shifting sooner or holding the rpm near redline is faster.

In your example of do you shift up because you are at redline, then have to shift 2 seconds later back down again vs reducing throttle to stay just below redline, the predictive lap timer will tell you that in 2 laps! Then it's judgement call if the 0.25 second faster busy shifting is worth the wear on you and the car.

I take a short break, partly for me, and mostly so there is a clear place in the dataloggers so I know the next data will be different with experimenting gear changes. While taking the break I review the Garmin Catalyst which gives me my 3 biggest "opportunities" (segments) for faster laps. I go back out and try using different shift points in the 3 opportunities. This can get a little complex because it probably involves different brake points and carrying more apex speed too.

Back in for a break, maybe just the Garmin, but probably look at the Video Vbox too. One more session again to verify where to shift gears and if all is good move on to optimizing brake points and possibly different lines. Until you get the gear shift points down it's really hard to fine tune all the other stuff.

FWIW:
Garmin Catalyst. Driver Improvement. Best value/$ of all of them. Nothing else you do on your car will lower lap times faster for the same $$.

Racelogic Video VBOX HD2. Driver Improvement. It was the gold standard but the Catalyst has better AI, is easier to understand what it's trying to tell you (because the AI simplifies it all down) and is cheaper. The VBOX is still the best by far when two drivers share a car during an Endurance race. There is some learning of how to understand squiggly lines. I really like them both and I get failover if one runs out of storage or some other glitch crops up. Like forgetting to hit Start on the Catalyst.

AIM PDM 32. I use the AIM data for car health monitoring, it draws data over the CANBUS from my LINK ECU. But not all of the data in the LINK. It's a good driver optimizer as well, but you have to learn to read squiggly lines, even more so than the VBOX. AIM was where I started on the datalogging to improve myself idea 10 years ago. I rarely use it for driver improvement anymore.

LINK ECU software. All things engine datalogging. Mostly gibberish to me! I download the logs and send them to my tuner for a quick review to make sure all the parameters are where they should be. Knock events, boost control, AFR, those type things.
Hope this helps!
Gator

[RPGs818SNA]

Happy to help, Dave. I enjoy learning about things using mathematical models.

The graph is simply data points calculated in Excel and plotted using its built in features. I used your torque plot, your gear ratios from your first post, a 3.7 final ratio from a Subaru discussion post, and 24 inch tires because that’s the biggest the front wheels will accommodate. If your final ratio or tire diameter is different, I can adjust the chart easily.

[Caution: boring details follow.] To get the data points, I copied your torque results, added a finer grid, and read the torque for every 210 rpm increment from 2610 to 6390. Next, for each gear, I multiplied crank torque times that gear ratio times the final ratio to get axle torque. Then I divided RPM by gear ratio and final ratio to get wheel RPM. I multiplied that by 2 feet and Pi, divided by 5280, and multiplied by 60 for miles per hour. Then I just plotted the MPH and torque numbers for each RPM to get the chart.

Note that my model only addresses the car. The most important factor in racing is driver skill, for which Gator has kindly shared his method of acquiring it.

Have fun with your 818,
RPG

[Dave 53]

I'd be most appreciative if you recalculated using a final drive of 3.9000 and a wheel diameter of 24.5 inches.

I'm looking for a base line to start from with my new drivetrain and how to program the LED shift lights. It seems to me simple acceleration down a straight is mathematical without much difference between theory and reality - upshift at 6,700 RPM for example. The challenge is, when to deviate from theory in the turns. Is it better to hold a gear below theoretical downshift RPM or do a quick down then back up shift? Do I do a quick up then down shift or hold a gear past theoretical shift point and touch the rev limiter for a second? The answer is yes. But, only time on the track will tell how to work it in real life.

I've never heard anyone that had a Garmin Catalyst complain about it. Unfortunately, with a new engine and transmission, I've already blown past my 818 budget for the next couple of years. I get a lot out of watching my videos and I identify plenty of low hanging fruit. And watching videos of others. Before I would just wait until I got home, but now I'll start bringing a laptop to review after each session.

O L D video from my karting days at (Sears Point) Sonoma. No tach light, down shift. One tach light means "in the pipe" as we called it. Lights blinking, up shift. There was no power "out of the pipe" and it wouldn't rev past redline, so there was no discussion about holding a gear.

[RPGs818SNA]

Sure Dave. Here's the chart revised for 3.9 final ratio and 24.5” tire diameter. Just to be sure, here are the gear ratios I’m using.

3.636
2.235
1.761
1.346
1.062
0.842

Axle torque vs mph Dave's 818 revised.jpg

RPG

[GoDadGo]

This has been a very interesting thread for me, especially since I'm not a road racer by any means which makes me have a few questions:

1. Being a former drag racer (Hobby Only) don't road racers want to stay in the peak toque curve as much as possible?
2. If the Subaru's engine has oil flow, pump cavitation and oil control issues, isn't the solution a dry sump system?
3. Couldn't adding a large pressurized oil accumulator reduce oil starvation with either a wet or dry sump system?
4. Have you considered adding a low pressure switch to kill the engine if pressure drops?

Sorry to ask silly and/or stupid question but if you can't keep the oil flowing smoothly any engine will be doomed.

Sorry That You Are Spinning Bearings!

[Sgt.Gator]

This has been a very interesting thread for me, especially since I'm not a road racer by any means which makes me have a few questions:

1. Being a former drag racer (Hobby Only) don't road racers want to stay in the peak toque curve as much as possible?
2. If the Subaru's engine has oil flow, pump cavitation and oil control issues, isn't the solution a dry sump system?
3. Couldn't adding a large pressurized oil accumulator reduce oil starvation with either a wet or dry sump system?
4. Have you considered adding a low pressure switch to kill the engine if pressure drops?

Sorry to ask silly and/or stupid question but if you can't keep the oil flowing smoothly any engine will be doomed.

Sorry That You Are Spinning Bearings!

I'll take a quick stab at these:
1) In theory at the peak area of the curve under the combined TQ + HP on the dyno plot.

2) Yes, dry sump is essential if you have wide R comp tires and aero downforce. However the first Gen 86 platform cars (BRZ- FRS) have no oil problems racing in SCCA T4 because they race on skinny tires (225 on 7" wide wheels) and no aero at all. No dry sumps. I don't know any that use a Accumulator even they are allowed.

3) Yes with a wet sump. No value with a dry sump. Accumulators are a band aid....they start working after the oil pickup is already sucking up air!

4) You can, and many aftermarket ECUs have that option built in. Killing the engine is good for complete loss of pressure like a oil line to the cooler failing. Or they can drastically reduce power if oil pressure drops below your threshold but not shut down the engine.

[J R Jones]

The BRZ/GR86 research video that Sgt Gator provided earlier was exceptionally well done and great information. In earlier threads (St Gator?) I have seen oil pressure degradation at high RPM.
As I mentioned to Dave, that suggests G-rotor cavitation, and I feel that is exacerbated by the regulated oil being re-introduced into the oil pump feed from the pan. The G-rotor gets foamy oil under high suction, a recipe for cavitation.
A solution I have envisioned is blocking the OEM regulator discharge path and create a new path to the oil pan.
Alternatively eliminate the OEM regulator and plumb an adjustable external regulator to an existing oil galley. Plumb the external regulator discharge to the oil pan.
The end in mind is non-foamy oil exclusively to the oil pump.
jim

[Dave 53]

This has been a very interesting thread for me, especially since I'm not a road racer by any means which makes me have a few questions:

1. Being a former drag racer (Hobby Only) don't road racers want to stay in the peak toque curve as much as possible?
2. If the Subaru's engine has oil flow, pump cavitation and oil control issues, isn't the solution a dry sump system?
3. Couldn't adding a large pressurized oil accumulator reduce oil starvation with either a wet or dry sump system?
4. Have you considered adding a low pressure switch to kill the engine if pressure drops?

Sorry to ask silly and/or stupid question but if you can't keep the oil flowing smoothly any engine will be doomed.

Sorry That You Are Spinning Bearings!

1. I assume yes, stay in the torque sweet spot, but road racing, there are times when it might make more sense to hold a gear out of the sweet spot vs. doing a quick up then right back down shift (or down then right back up).

2. Dry sump is the gold standard, but very very expensive. My research concludes pump cavitation is due to using too big of an oil pump. Subaru EJ (what 818's use) oil pumps come in 3 sizes, 10mm, 11mm and 12mm. It seems cavitation only happens with 12mm pumps. OEM is 10mm and I have an 11mm pump. Instead of a dry sump, I have an Arctangent oil control plate that is sandwiched between the oil pan and block. Unfortunately, I don't think they are made any more. It keeps the oil from backing up into the heads. In addition, I have the Killer B pan and oil pickup tube. An aftermarket oil pickup tube is essential because OEM tubes are prone to failure. Also essential is a good air oil separator. I have gotten myself dizzy on a skid pad and don't seem to have any oil issues. I think the Arctangent plate is a game changer.

3. I don't remember my research, but I do recall I talked myself out of an accumulator when I went down that wormhole a couple of years ago.

4. I have two oil pressure sensors. One for a gauge and the OEM warning light sensor. The gauge won't clearly display momentary drops in pressure. The warning light sensor is very sensitive and will trigger with even a split-second drop below 10 psi. I have a bright high profile light on my dash. It will be impossible to pick up a momentary drop from the light because I will be focused on the turn, but in reviewing video, the light has never illuminated even in long sweeping turns. I have since added a feature that the light will flash for 15 seconds if it is triggered for even a millisecond so I will have feedback immediately after a turn if I had a drop during the turn. My dumb OEM ECU can't accommodate an engine shut down feature, although after figuring out how to make the 15 second flashing light work, I'm certain I could hack a solution to shut the engine off. Higher end stand alone ECU's can be programed to respond as desired to any number of issues - oil pressure, water temp, etc. You still have to grab the clutch and from my karting days, I can tell you it's impossible to grab the clutch fast enough to do any good.

I think I spun a bearing because it turns out the engine had substandard bearings and I was over-revving it. What I have learned from all this is that I don't need to rev the **** out of the engine! And now I'm using top shelf bearings.

[J R Jones]

Cavitation or not, is speculation without a lab test. Cavitation is more likely with increased RPM, increased surface speed (diameter) and increased pressure differential (inlet and outlet).
10mm, 11mm, and 12mm pumps have consistent RPM and surface speed.
The pressure differential, with the same pump inlet plumbing, increases with rotor thickness.
The regulator masks reading cavitation. At RPM if the pump creates too much pressure it regulates, and discharges (foam) into the inlet side. That can cause cavitation, loss of pressure, and the regulator stops discharging into the inlet side. A Catch 22.
A test of cavitation would be an RPM performance curve without a regulator. RPM and pressure are tracked until pressure flattens or drops. Inlet restriction and lift must be representative.

I have seen Subaru aftermarket youtube videos speculating on this and they admit testing in-car is work intensive, an engine dyno would be easier. Block the regulator and do the RPM vs pressure tests rotor to rotor.
Question: If you get spec oil pressure from 10mm, why go bigger and increase regulation/cavitation?
jim

[Dave 53]

Three months ago, I didn't know there were 3 sizes of oil pumps. Curious, I went down the worm hole. Like many subjects, there was information - some probably good and most probably not and nothing conclusive. The only conclusion seems to be that it's an unsettled subject in regard to "built' engines. Then I found myself in need of an oil pump and now a decision needs to be made. Would a bigger pump be better for me? What am I doing compared to a daily driver grocery getter that is obviously fine with a 10mm pump? I've got a "built" engine with a giant oil cooler that I'm flogging on racetracks.

No, I didn't seem to have oil issues before. It wasn't broke until it was broke. Would I have spun a bearing if I had a bigger pump? Probably defiantly, but....

Not being sold on any given argument and perhaps it's misguided, but I concluded that there is *something* to going to a bigger pump on a track car, but one can go too far, so I settled on a momma bear 11mm pump. If I had the money, I probably would have gotten a fancy IAG or RCM 11mm pump.

It would be great if someone would publish the test you propose.

[mikeinatlanta]

IMHO: Way too many assumptions being applied to draw any conclusion.
Spun due to overrev?
If oil starvation, due to cavitation?
Oil pump size?

We'll start with the third. Only way to know is to data log oil pressures. Without this information you can't possibly know whether your oil pump is too small or too big. Do you build pressure @ the classic 10 psi per 1000 rpm? This is how you determine pump size, not high rpm pressure loss.

Even if data log shows pressure loss, it can be very difficult to tell (impossible without testing) whether it's cavitation due to pump design or aeration due to high windage. Then of course one needs to determine exactly how braking, acceleration, and cornering loads impact it.

Spun a bearing due to overrev. A bearing spun due to stress will show indications of cap walk. What was the history of the old motor and what analysis was done to the failed parts?

Of course, the easy answer is to put on a dry sump and call it a day, but without some valid data you are flying blind.

[J R Jones]

A larger pump will create more outlet pressure and more inlet suction (duh). No matter the outlet potential, the regulator will modulate at "X".
Enlarging the inlet plumbing will minimize suction.
The S-Gator BRZ/GR86 tests evaluate inlet plumbing/baffling and it is not encouraging.
The you-tube video I watched justified a bigger pump because they opened-up bearing clearances. I do not get that (loose) objective, and they had no follow-up results.
jim

[Dave 53]

100% flying blind with a voice in my right ear saying "come over here" and another voice in my left ear saying, "no, come over here." I can read my oil pressure gauge, but then I had an engine builder question the accuracy of my oil pressure gauge. (Mike at GST, Hayward, CA).

I went down the oil pump worm hole before I needed to buy one because I enjoyed doing so and learning about it. Made a decision for better or worst on a 11mm pump. And I will continue to enjoy learning about this subject and the stimulating conversations! I only hope I'm making worthwhile contributions.

I wish I had the cash for a dry sump. A friend tells me the solution is just get a 911. In one respect, I'm glad my engine failure was user error because that's easy to fix. I don't think a bigger pump or even a dry sump would have saved me from myself. I had 25 track days on the car. I'll be the Arctangent oil control plate / Killer B and now 11mm oil pump test bed. Take away the user error and so far, so good...

Now that the pump is in, I'll go do some laps. What would be appreciated from you all is tips on how to evaluate things during or after a track day.

The hobby for me is learning and DOING (Laguna Seca this weekend!). Sometimes you gotta take a **** or get off the pot.

Now I'm going down a worm hole on what kind of tape to use on my radiator ducting. Home Depot has 6 different flavors.

[mikeinatlanta]

Engine Masters has some good episodes on oil level, pump sizing, windage, oil pressure, and oil viscosity. Suggest recording every episode and review any one on oil. Not perfect info but beyond the vast majority of engine builders.

[mikeinatlanta]

Dave, You don't happen to know my brother Troy Messer from your carting days do you? He's out of west Texas but headed west racing a bit.

[Dave 53]

Thanks for the Engine Masters tip. I don't recall the name Troy Messer. I did race in the National Road Race (big tracks) Championships around 2000 at Thunderhill, CA (placed 5th - I'm pretty proud of that). That event had folks from all over the country and I do remember a couple of teams being from Texas.

A question for all that I thought of while just eating some eggs. When I start the engine cold, oil pressure is at its highest, almost 90 psi at idle and say 70 degrees. Of course, psi comes down as oil temp increases. That tells me the pressure relief valve is set at 90 psi or higher. Can I conclude that as long as pressure is under 90 psi, the relief valve is staying closed?

It seems the issue is what happens to the oil when the pressure relief valve opens up. If it's not opening, do we have a problem?

[mikeinatlanta]

Thanks for the Engine Masters tip. I don't recall the name Troy Messer. I did race in the National Road Race (big tracks) Championships around 2000 at Thunderhill, CA (placed 5th - I'm pretty proud of that). That event had folks from all over the country and I do remember a couple of teams being from Texas.

A question for all that I thought of while just eating some eggs. When I start the engine cold, oil pressure is at its highest, almost 90 psi at idle and say 70 degrees. Of course, psi comes down as oil temp increases. That tells me the pressure relief valve is set at 90 psi or higher. Can I conclude that as long as pressure is under 90 psi, the relief valve is staying closed?

It seems the issue is what happens to the oil when the pressure relief valve opens up. If it's not opening, do we have a problem?

He was running around 2003. I know won some championships. Recently running things like NASA time trial and racing. Won national championships in TTU and the unlimited road racing. Also I think in TT1.

You are correct that that is where your relief valve is opening. In a perfect world you would gradually build pressure and hit the relief valve at redline. Back to the very generalized 10 psi per 1000 rpm, but starting around 20-30 hot. If you are hitting full pressure way earlier then you have too big a pump and are just wasting hp. The motor in my bike (GM LS based V4) sits at 14 during idle. Same for my 427W hot rod motor.

[J R Jones]

Back in the seventies multiviscosity oils were introduced in racing. Street oil filter canisters on race engines have blown off the base plate on cold start, necessitating the HP1 Fram which is more pressure-proof.
Some dry sump engines had oil heaters used before cold start. Delivering oil quickly is a priority. Today's 5W- and 0W- oils are OEM requirements to deliver oil on cold start.
The OEM oil system (regulator) is designed to function at operating temperature. The 90PSI reading could be cold oil overwhelming the pressure relief. The actual relief pressure may be lower.
I expect your hot pressure max represents the relief valve setting. If the pressure drops at high RPM, you are ventilating at the pick-up (turns or braking) or cavitating (straight).
jim

[Dave 53]

Back in the seventies multiviscosity oils were introduced in racing. Street oil filter canisters on race engines have blown off the base plate on cold start, necessitating the HP1 Fram which is more pressure-proof.
Some dry sump engines had oil heaters used before cold start. Delivering oil quickly is a priority. Today's 5W- and 0W- oils are OEM requirements to deliver oil on cold start.
The OEM oil system (regulator) is designed to function at operating temperature. The 90PSI reading could be cold oil overwhelming the pressure relief. The actual relief pressure may be lower.
I expect your hot pressure max represents the relief valve setting. If the pressure drops at high RPM, you are ventilating at the pick-up (turns or braking) or cavitating (straight).
jim

I will data log oil pressure (oil pressure gauge in view of video camera). Then take a close look throughout a lap.

My oil temperature gauge has a low and high warning light. I have the low set at 145 degrees and go VERY easy on it until that light is off. Mostly just idle or creep slowly down my street. It's usually around 170* by the time I get to the freeway. Is there a better warm up routine?

[J R Jones]

No I have no alternative.
I used to have (low) viscosity anxiety, but all my best cars have recommended it. That was foolish of me, hot oil is like water.
BTW efficiencies are better with thin oil and 200F+ is an accepted target. Oil temp is not problematic below 270F.
BTW I built a high perf RX7 rotary for the street. Rotary's run hot and loud. I used an aluminum two-pass Griffin SBC racing radiator. I welded an oil cooler extrusion in the transfer tank and it worked very well. The oil and water temperatures warmed and cooled together. Seemed to me to be more stable than segregated cooling cores. Good packaging too.
jim

[Dave 53]

I also posted this video on my "Track Videos" thread, but for some comments relative to this thread...

Shift Points: I added LED shift lights. The first blue light comes on at 5,000 RPM when I enter the RPM "power zone." Out of the power zone at 6,500 RPM. The lights blink at 6,300 figuring I'd probably be at 6,500 by the time I reacted. Super helpful!

Oil Pump Size: Oil pressure seems to be steady at 75 psi regardless of RPM from 4,000 to 7,000 at 210 degrees. This tells me the pressure relief is set at 75 psi and there is steady relief happening. The question - is this a problem? I think I can swap back from the 11mm pump to a 10mm pump in 3 hours which really means 6 hours. Should I?

[J R Jones]

Dave, I agree with your observations. Likely the pressure relief is 75 psi and it may be venting until RPM drops.
On a slow acceleration, at what RPM do you reach 75 PSI with the 11mm pump?
What pressure does the 10mm pump achieve? I assume 75 PSI with the same regulator?
On a slow acceleration, at what RPM do you reach max PSI with the 10mm pump?
jim

[Dave 53]

I took a close look at some video data logging.

I use Motul Sport 5-40 oil.

With the 10mm pump, my general observation is, at about 220 degrees and track RPM (4k-7k), psi is steady at 50 on straights and drops to 25 in turns/hard braking. In long turns, it drops to a steady 25, then climbs back up to 50 straight. I'm thinking 25 psi is a bit low at 7,000 rpm? On acceleration, it seems to hit max psi around 3,000 rpm.

Not much data on the 11m pump. First two sessions with it were a fail and I only got one lap in at a fast warm up pace before my axle broke. But the data on that lap: Oil pressure is not a function of straights vs. turns - it holds rock steady at 75 psi regardless. 90psi under 180 degrees out of the pits. Rock steady 75 psi once warmed to 220 degrees. On acceleration it seems to hit max psi around 2,000 rpm.

Curious to see psi at around 260 degrees. It was a nice cool day at Laguna but 100 degree track days are common here in California as are 260 - 270 oil temps.

I'm surprised to see this difference in performance profiles of the two pumps (and other unknown or considered factors). It will be very interesting to compare data of the 2 pumps on the same tracks. In the meantime, based on this limited data, it seems to me the 11mm pump's steady 75 psi beats the 10mm drops to 25psi. But, I'm learning, so maybe not?

Jim (and anyone else), what conclusions can you make based on this limited data?

[J R Jones]

Dave, What I do not know is if the regulator carries over from the 10mm rotor to the 11mm rotor? Does the 11mm rotor have a dedicated regulator? 50 to 75 suggests a higher pressure regulator.
What is most encouraging is the steady pressure without drop. One would expect that drops in braking and turns is ventilation, the 11mm alone should not improve that.
Curious. jim

[Sgt.Gator]

I've done a deep dive in the Subaru oil pump stats bin because I have my own line of modified Subaru OEM oil pumps. I haven't done any marketing of them because I wanted a year of real world racing before I introduce them publicly. More on that later...
I've uploaded the Subaru stats from the factory for the 10 , 11, and 12 mm pumps for the EJ, and the stats for the FA20 for grins. There is a little confusion out there about the rotor sizes. Subaru makes the 10 mm in both a 76 and 78 mm size. The 78 is the older style. They went to a 76mm rotor diameter I believe based on efficiency. I have the SAE paper on that subject although it's not on the Subaru pump specifically, just georotor design.
The pressure relief valves open at different pressures for each size pump. The low is 71 psi, then 78 psi, then 85 psi. And of course you can raise it yourself by adding a washer(s).
The FA20 relief valve is at 102psi; and the Pressure is tested at 248 degrees! At 6,700 rpm and 248 degrees the oil pressure is 46.6 psi.

The EJ pumps are all tested at 176 degrees. You can expect at 220 degrees they would be a lot lower pressures. Also Note the high rpm stats on some are at 5,000 rpm, others at 6,000 rpm, the FA20 at 6,700rpm.
10mm 15010AA300 78mm Oil Pump Specs.JPG2008 STI DAVCS Oil Pump 11mm 78mm specs.JPG2014 WRX Oil Pump 10mm x 76 specs.JPGFA20 DIT WRX Turbo Oil Specs.JPGSubaru Oil Pump Stats.gif

[J R Jones]

Great specifications. Oh for consistent test procedures.
The margin between maximum pressure and regulation pressure is preferred. However the 78X12 is same-same.
Dave's 11mm pump max pressure @ 75 does not conform with the chart; could he have a "tighter" lube system? He is about 18PSI over and on the pressure relief spec. A confirmation with alternative measuring equipment might be appropriate.
Obviously larger diameter rotors pump more oil, but the surface speed is greater, increasing the possibility of high speed cavitation.
jim

[Sgt.Gator]

JR a lot depends on where your oil pressure sensor is located. Maybe Subaru is reading it right at the pump exit. Or maybe at the oem oil pressure switch port at the front of the block? And a lot of folks read it at the right rear port.

pump bypass (small).jpgEJ25x Cylinder and Oil Layout.jpgOil Galley Ports Subaru EJ.png