When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Here’s a nerdy post for anyone interested in engine efficiency, which RX8s are of course known for.
An engine’s Brake Specific Fuel Consumption (BSFC) is the amount of fuel it costs to get a certain amount of power. I think of it as it takes X grams of fuel to get X horsepower. If you’re accelerating on an on ramp, you want the number to be low, so you burn the least amount of fuel per oomph as you accelerate. If you’re cruising on the highway at a set speed, you still want the number to be low because it takes a certain amount of power to fight wind resistance and drag. BSFC changes for each engine depending on where you are in the rpm band and how far the throttle pedal is pushed down. For NA piston engines, it’s great at low rpm and high throttle. Conversely, it’s terrible at high rpm and low throttle. Learning the BSFC chart lets you know how and when to shift gears for fuel efficiency. It also makes me feel better about pushing the fun pedal down, knowing that it’s not costing me as much fuel to accelerate fast as it would if I accelerated gradually.
I was curious what the BSFC contour looks like for my 2006 RX8. I grabbed my CAN logger and measured Absolute Load, Mass Air Flow, Lambda, and RPM over several drives. I calculated fuel flow from lambda and mass airflow, since I couldn't find a fuel flow CAN signal. I crunched the numbers and made this plot. For reference, a Ford Ecoboost engine has minimum BSFC around 260 g/kWh.
I don't know how useful this data is to anyone, but I hope it was interesting. If anyone has measured something similar or has ideas on how to do it better, let me know!
Edit: After a couple nights of thinking on it, I revisited the Absolute Load CAN signal I was logging. The signal comes in as a percent and I assumed 0% was 0 Nm and 100% was peak engine torque (assumed to be around 210 Nm). This is not the case, especially at 0% pedal, where it hangs around at 20%. If anyone knows the mapping between absolute load percent and actual brake engine torque, please post! Otherwise, I'll assume absolute load percent maps linearly from -20 Nm to 210 Nm and scrap all the points at 0% pedal.
2nd Edit: I corrected the data for Absolute load, assuming it maps linearly from -20 Nm to 210 Nm. The BSFC plot has been updated and is now much more believable. I'd guesstimate it's accuracy to 15%, since the engine torque isn't exactly known and the fuel flow numbers could be off due to using the O2 sensor for the calc. I'd say the plot is good enough to get a feel for where this rotary engine is efficient. If you're accelerating on an onto an on ramp, keep it between 3000 and 4000 rpm for the best efficiency.
Last edited by c-freeP0; 05-09-2020 at 08:05 PM.
Reason: correction to graph
Here’s a nerdy post for anyone interested in engine efficiency, which RX8s are of course known for.
An engine’s Brake Specific Fuel Consumption (BSFC) is the amount of fuel it costs to get a certain amount of power. I think of it as it takes X grams of fuel to get X horsepower. If you’re accelerating on an on ramp, you want the number to be low, so you burn the least amount of fuel per oomph as you accelerate. If you’re cruising on the highway at a set speed, you still want the number to be low because it takes a certain amount of power to fight wind resistance and drag. BSFC changes for each engine depending on where you are in the rpm band and how far the throttle pedal is pushed down. For NA piston engines, it’s great at low rpm and high throttle. Conversely, it’s terrible at high rpm and low throttle. Learning the BSFC chart lets you know how and when to shift gears for fuel efficiency. It also makes me feel better about pushing the fun pedal down, knowing that it’s not costing me as much fuel to accelerate fast as it would if I accelerated gradually.
I was curious what the BSFC contour looks like for my 2006 RX8. I grabbed my CAN logger and measured Absolute Load, Mass Air Flow, Lambda, and RPM over several drives. I calculated fuel flow from lambda and mass airflow, since I couldn't find a fuel flow CAN signal. I crunched the numbers and made this plot. For reference, a Ford Ecoboost engine has minimum BSFC around 260 g/kWh.
Per my data, which I have some reservations about, the higher in RPM you go the worse your BSFC gets. You get pretty good BSFC between 3000 and 5000 RPM, so if you're driving for fuel efficiency you should stay there (big surprise).One thing that doesn’t make sense to me is how the BSFC is better for lower torque. Normally the throttle blade creates significant losses at low torque, so I don't really trust my CAN signals here.
I don't know how useful this data is to anyone, but I hope it was interesting. If anyone has measured something similar or has ideas on how to do it better, let me know!
Edit: After a couple nights of thinking on it, I revisited the Absolute Load CAN signal I was logging. The signal comes in as a percent and I assumed 0% was 0 Nm and 100% was peak engine torque (assumed to be around 210 Nm). This is not the case, especially at 0% pedal, where it hangs around at 20%. If anyone knows the mapping between absolute load percent and actual brake engine torque, please post! Otherwise, I'll assume absolute load percent maps linearly from -20 Nm to 210 Nm and scrap all the points at 0% pedal.
tldr: The graph is wrong and I need to retake the data but will repost when it's gathered
Love nerdy posts. My data showed best cruising efficiency @ 2500-3000rpm for sustained speed cruising. No graphs but my OBD2 scanner and torque app showed fuel efficiency best in that range @24mpg over a 10 second reading. Not 100% if that’s true or not but that’s the RPM I tend to cruise at because of that reading.
Interesting . I have logs that may help ..... full throttle logs from 2500rpm where I have absolute load recorded and virtual dyno numbers (within 5% accuracy)
Brettus, dyno data with absolute load would be pretty useful, but I assume the pulls were recorded at WOT. To really take the plot to the next level, you'd need absolute load versus torque for part throttle as well. We could further complicate things with transmission efficiency. I think I'm satisfied with where the plot is for now if you guys are. I was mostly just curious what the BSFC chart looks like for a rotary versus a piston engine. That and I wanted to play with my CAN logger .
I do have absolute load recorded for some of my dyno pulls . I also have Virtual dyno which, despite what Team says, I've found I can get within a few % points of a real dyno.
Unfortunately though my current engine isn't stock (or even close to stock) so 1/2 throttle pulls for a stock engine is not something I can do right now.
05-04-2020, 09:07 PM
