Moar!
Moar!
In reply to series8217 :
WOW, your last post was quite a read! I appreciate your approach to developing this car and the supporting systems. May i ask, what is your day job?
Super impressive! I need to read up more on the stand alone CANbus ABS stuff you used here. I've even heard of people retrofitting them on C6Zs to use aftermarket engine management with stand alone ABS.
You also really show the need to sort and refine a car to get good track times, vs. the mindset of "hey, let's just throw a bigger engine and more horsepower at it and surely it'll be faster", and then find out that nahh..not necessarily. I'm totally guilty of this.
Double, Double, Oil & Trouble
You might remember a long time ago that I was reporting power loss from the engine under some circumstances. I thought it was heat related, but never had the data to prove it. Well, now I do... and it turns out it's not heat related. Instead, it's actually being caused by me improving my driving and the setup of the car. With higher lateral G's, the problem has gotten much worse. The symptoms so far have been:
All of those symptoms point to windage-induced power loss and oil aeration.
Even worse, I analyzed oil pressure vs. RPM vs. throttle position and found that with any significant lateral G force the oil pressure drops independently of the other two. This can be a really bad combination if the oil pressure is low but RPM and throttle are high.
Here lateral G's are on the X axis, oil pressure is the Y axis, and the color is load. More orange/red is more engine load. What we don't want to see is a lot of load with little oil pressure. As soon as we get off the center of the chart (0 g's, so straight line), oil pressure drops independently of load.
We can also change the color channel to RPM. Ideally, as the colors go from blue (low RPM) to green->yellow->orange (increasing RPM) the dots should be hgiher on the graph (higher oil pressure). We see that relationship around the center of the chart, where lateral acceleration is low... but as we increase lateral G's we start to see green -- and even some orange (high RPM) -- dots at very low oil pressures.
This means I have not just an aeration problem, but starvation too.
To investigate further, I removed my oil pan and took a look at the stock oil control system.
Here you can see the stock windage tray and oil pump pickup. The LQ1 stock pump is the same high flow oil pump and high pressure spring as the GM 60-degree V6 performance manual recommends for race applications. This is needed because the LQ1 is DOHC and needs more oil flow to supply the heads.
The stock windage tray is simple but complete. It substantially separates the oil sump from the rotating assembly, and has a couple louvers to encourage oil to exit the windage cloud and return to the pan. There are also some separate baffles bolted to the engine block to protect some of the oil drainback holes from windage. Nice.
First I checked the oil dipstick markings to see where the normal fill level is relative to the rotating assembly.
The bottom hole ("low") on the dipstick is flush with the bottom of the stock windage tray.
The top hole ("full") is 1 quart higher and puts the oil level above the tray. Presumably, once the engine starts the oil level drops a bit as some oil is in the heads and flowing back. The important thing to note here is: do not overfill! Even one quart high is too much. It will just get whipped up by the crankshaft.
I experimented with filling the pan with water and sloshing it around by hand to see how the oil flows through the various baffles.
Here you can see what the stock fill level looks like with the pan at a 45-degree angle, simulating 1G right sweeper (the pic is looking from the front of the car). Yeah, the water is spilling out the end of the pan... at 1G with the stock oil fill level, we are already pushing oil into the rotating assembly. The windage tray does act as a baffle to some degree, but given a long enough corner the oil is going to migrate through it.
Digital Dissection
Next, I 3D-scanned the rotating assembly, block, windage tray, and oil pan so I could virtually dissect it and learn how it behaves under different operating conditions.
Volume in the sump is 4.7L for this analysis. In reality there will be even less in the sump because some will be in the heads in the process of draining back, and more will be entrapped in the windage cloud around the crankshaft.
First, let's look at a right hand sweeper with sustained 1.5G lateral acceleration. Here, we see the right side of the pickup is starting to get exposed. The oil volume is the big hatched polygon. Red is the submerged pickup. The lower density hatched area is the exposed pickup.
Fortunately, the engineers at GM accounted for this by creating a small baffled sump area that feeds the exposed end of the pickup tube.
The area where the * is keeps a reserve of oil that bleeds through the opening at the arrow. This gives a supply of oil to the pickup tube until the reservoir is depleted. I don't know how much volume is entering the reservoir during a steady state corner in this direction, but there are a few places that supply to it --- anything scavenged by the windage tray, the volume returning from the timing chain area, and some of the oil drainback holes from the heads.
Now left turns get pretty nasty.
The pickup is almost completely exposed at steady state, and a good 1/3 of the crankshaft is submerged in oil.
The only thing helping us here is the "ceiling" baffle that partly encloses an area around the pickup. Unfortunately it is mostly open to permit the pan to be installed in one piece over the pickup and pickup tube. I don't expect it to be very effective. In theory, the volume shown here in orange is kept wet by these baffles, at least for a little while.
The crankshaft is basically submerged in oil for every corner at maximum lateral G's, which saps power and causes aeration of the oil, decreasing its ability to protect the engine bearings. Since aerated oil is compressible, this also causes the hydraulic valve lifters to pump down. In addition, the oil is flung up onto the cylinder walls, coating them with more than the piston rings can control and leading to excess oil consumption and oil smoke out the exhaust.
In addition, the oil pump pickup is exposed under any steady state corner at maximum lateral G's, even with the stock baffles in the pan. Under transient conditions the baffles and windage tray certainly help, but given enough time oil flows away from the pickup and it starts to suck air.
Wet Sumps Are Hard
Unfortunately there isn't a complete solution other than going to a dry sump system. I'm not prepared to invest in a dry sump for this engine, as I intend to swap in a higher output powerplant in the next year. So I want to do what I can to make the engine survive with a wet sump.
I reached out to Kevin Johnson of Ishihara-Johnson Crank Scrapers for advice. He recommended a close-cut teflon crankshaft scraper to help strip oil from the windage cloud around the engine and return it to the sump. This should help reduce aeration and windage-induced power loss. He also recommended adding directional screening to the windage tray to prevent droplets from bouncing off the tray and back into the rotating assembly.
I intended to reuse the stock windage tray, but a friend advised me that a later version of the OHV 60-degree V6 had a more performance-orientated tray with more louvers. I bought one and successfully test-fitted it. Just needed to bend the oil dipstick tube by a few degrees. The only disadvantage of this tray is it doesn't wrap around the side of the oil pump. Not sure how much that actually matters. Did the GM engineers get lazy (or the bean counters get in the way?) or did they discover that it wasn't useful or was perhaps even detrimental to have the windage tray extend all the way to the flywheel end of the sump?
LZ9 windage tray on the left, LQ1 tray (stock for my DOHC engine) on the right.
Here's the LZ9 tray installed in my LQ1
The LZ9 tray doesn't cover the crank throw that's under the oil pump, whereas the LQ1 tray is extended over that area. I thought maybe the LZ9 pan incorporated baffles in that area instead. As it turns out, the LZ9 pan has no baffles at all. The only oil control feature in the LZ9 (aluminum pan OHV 60-degree V6) is the windage tray.
I sent the tray to Kevin Johnson for him to add some directional screening. He's also building a teflon crank scraper for my engine.
My goal with the windage tray and scraper is to help solve aeration related oil pressure loss, and windage related power loss and oil consumption.
The cornering-related oil pressure loss is going to be a much tougher nut to crack. I am considering installing an accusump, but Kevin pointed out that if my oil is aerated at the time of pressure loss, the accusump is going to get continuously refilled with aerated oil...
At the very least, I'll have the scraper and windage tray installed before the next track day so I can see if they help with the aeration problem.
Q&A
AngryCorvair (Forum Supporter) said:In reply to series8217 :
WOW, your last post was quite a read! I appreciate your approach to developing this car and the supporting systems. May i ask, what is your day job?
I used to work in automotive (mostly ADAS) and now I build automated farm equipment.
Pics were broken in my last post... not sure why that keeps happening. They showed up fine in the editor, then when I checked my thread a few days later just a broken link. 
I think I fixed them be re-uploading.
Stalking you here about unknownTCM_Torque_189 and TransmissionInputSpeed while my pcm account activation is still pending. My current project is TCU integration for rusEFI, what's your interest in GM CANbus?
You'll need to log in to post.
