Found: New Japanese Method for Measuring Hybrid Power

Already done by @PencilGeek who posted some dyno runs:

This is why making a video of a 2016 maximum acceleration was such a brilliant hack. We still have some unknowns, the roll-down coefficients, but these should be available in January. But I also expect we'll see dyno runs within six months.

Toyota released a presentation showing acceleration is more linear with the sound of the CVT. Then there is: Nissan alters all CVTs to act less like a stretched rubberband

. . .
To give drivers the option of feeling gearshifts while on the road, Nissan is adding its D-Step Shift Logic feature to the CVTs in multiple vehicles. Steve Powers, Nissan's senior manager of powertrain performance, told Autoblog the system forces the transmission to "hold a ratio and then shift" to simulate the way that a traditional automatic would. It's simply a change in software, but the company "can't do it to older CVTs," he said, because it would require changes to transmission logic, as well. According to Automotive News, the upgrade is coming to the 2015 Versa, Versa Note (pictured above), Sentra, V6-equipped Altima, Pathfinder and Quest. "We're rolling it out to all programs," said Powers.
. . .​

Bob Wilson

 

Let's not be too hasty until we get some hard dyno number or better coast-down coefficients.

Bob Wilson

 

Source: 2016 Aston Martin V8 Vantage GT First Drive Review

To Argonne National Laboratory’s Mike Duoba and his SAE J2908 (Hybrid System Net Power Rating) task force, the net power of a hybrid powertrain is material for serious debate. Rating a car’s combustion engine by its lonesome is easy peasy by current J1349 and J2723 standards: Test the engine as it’d be dressed and installed for customer consumption, right at the crankshaft. But for a hybrid system—where two (or more) propulsion components and two (or more) energy storage mediums are in play—its energy management can be highly unpredictable.​

For years, engine shaft power to the transmission was the metric . . . and then they put a hydro-mechanical, automatic transmission between the engine and wheels. They were so inefficient that the transmission fluid had to pass through a radiator cooler loop. Yet these mechanical monsters were ignored by the automotive press and 0-60 times and efficiency were often rated for the mechanical transmission cars.

Net horsepower to the wheels, dyno runs, are the gold standard. But the draft appears to use shaft power from the engine and traction battery power, not the wheel values. There are still constant velocity joints, wheel bearings, and tires to pass through . . . tires being the big unknown.

I would be OK with coupling the drive wheels to a dyno coupling. Something as simple as hydraulic pumps or electric motors would work for me. This would eliminate the tire-roller unknowns. But keeping the tires has one big advantage . . . finally measuring tire rolling resistance.

Then one last change, speed stepped MPG chart. Use random number generators to:

1. Set target speed order, 15-80 mph, and 0 (aka., random stops)
   
2. Set target speed duration, 5-10 minutes

About the tires, the same dyno with the tires bolted to a weight for loading would work very nicely.

Bob Wilson

 

SAE J2908 for 2010 Prius (Gen 3, ZVW30) using HEM mini OBD, 200 ms. sampling rate:


Approaching the Brindley Mountain upgrade, I was stopped at a traffic light with a minivan on the right. When the light change, the minivan and I moved out briskly down a 1/4 mile slope to the base with an estimated speed between 65-70 mph. Then I floored the accelerator and soon was climbing at 91 mph. At the crest there is a gentle turn and as I was lifting off, the 'traction' warning light briefly flashed. It looks like that triggered an engine torque pulse while it was still spinning at 5100 rpm, the power spike at the end that slowed the engine down to ~4,000 rpm. I've attached a section of the recorded data.

My eyes suspect the 2010 and 2016 Prius share many of the same control laws. The maximum power may be speed limited because the first two seconds saw a lethargic spin-up of the engine. This is consistent with the 2016 Prius, maximum acceleration curve seen in the Nov 8 demo drives that remained flat until Jason reach 89-90 mph and started to rise. Regardless, my eyes see maximum power in the 117-120 hp range that tapers off as the traction battery charge is depleted.

I have Nuvi recorded GPS data to integrate with these metrics which will add speed and altitude changes.

Bob Wilson

 

You might look for some 'Easter Eggs' in the attached data file. For example, injector volume times ICE rpm versus MAF and EGR position during the climb. You can see where the car switches to fuel enrichment during the climb.

When I first found SAE J2908 (draft) and the Japanese version, it really looked close to a net HP. The only thing missing was how to account for the transaxle, CVT, wheel bearing, and tire drag. Fortunately, the basic metrics can be found in the OBD stream. Just the HEM mini is a little complex.

The HEM package has an encrypted database of pretty much every Toyota OBD for just about every model. It has a sophisticated way to poll the ECUs to determine which commands are valid. This is used to 'enable' the OBD commands from the database. Unfortunately, it takes some field work to trim the database down to the dynamic parameters . . . 'some assembly is required.' It also has issues with multi-value returns.

Some OBD data records contain two metrics avoiding timing delays (aka., race conditions) such as returning combined rpm and torque needed for power calculations. I've put the dual-data return OBD values aside for now as there are ways to handle race conditions in the data at some loss of timing. I'm hoping to find combined voltage and current from the traction battery.

Right now, I'm tweaking HEM before running a maximum acceleration test like Ad Hoc 0-60 mph | PriusChat. Only I'll go to at least 100 mph instead of doing a hill climb. I'll also get a total weight with driver before the test. I am surprised no one else has thought to replicate this test with a Gen-3, Gen-2, or Gen-1. It will also allow faster and more accurate collection of mph vs MPG data in some interesting speed ranges.

Anyway, enjoy the 'Easter Eggs' as more will follow.

Bob Wilson

 

So this is a work in progress but my first cut at 2001-03 Prius performance:

• 90-94 hp net power after 70 mph
  
• ICE power increases by rpm which is speed dependent
  • ICE power limited by MG1 torque which is nearly constant for each run
• First run after forced charge has traction battery energy at end of run
• Second run without forced charge runs out of battery energy

Without some way to increase MG1 torque, we can not get more power from the engine. This is probably the same limitation of all Prius.

Bob Wilson