Home | Wiki | Live Chat | Dev Stream | YouTube | Archived Forums | Contact

Study Of Effiency and Power in relation to Compression Ratios and Ignition Timing


#21

Hmm… good point. I suppose that arbitrary numbers with fake numbers is technically worse because it combines the sins of being arbitrary and being a lie.
Solution = make AFR real during the engine calculation revamp?
If possible… I know timing isn’t because a timing map is much too complex to put a single number on it.
I suppose AFR could be static with just an on/off throttle position… like what the in-game system implies now.


#22

Yeah, I can’t think of a good way to make AFR real either, I mean, every ECU I’ve worked on has ended up with AFRs ranging aaaalll over the place depending on the situation. Yeah, the on/off throttle numbers documented better would probably help.


#23

I was hoping for three numbers: the leanest is idle/low cruise, 2nd is normal acceleration, third is when the secondaries open near WOT. Granted, computer controlled EFI can produce literally any mixture, but from a carburetor perspective those 3 numbers are what carb tuners take from the charts to select a tapered metering rod and primary and secondary jet size.

In the EFI world, the 3 number system might still applies to some degree–lean at low power, stoichiometric under moderate load, and rich (to a point) on heavy load. That point is limited by 1) the engine’s ability to actually burn the fuel, and 2) emissions. Excessive enrichment will kill your cat, so there’s a sweet spot in WOT enrichment beyond which you hurt reliability and produce sub-optimal power. Emission numbers start to rise as soon as you move rich of 14:1, so if you’re selling in fruinia, you’ll need to put a reasonable cap on your WOT mixture. <–Hours of immersive game play right there.

My two cents is that an arbitrary number would cut into the immersion. I should not have come off as complaining because what you have is a great thing going–Maybe an arbitrary number selector could be used to produce fuel curves in a similar manner to the suspension presets, but for fine tuning, I’d like to be able to tweak the low, middle and high end of the fuel curve with power, efficiency, emissions and reliability consequences. the game already has the first three,

One arbitrary number would make me sad. All that said, 16:1 is where best economy is. You can’t apply power at that lean setting, but a VVT engine with the magic of in-cylinder exhaust scavenging can cruise 60mph with a 16:1 mixture and a cylinder full of exhaust gas. Just a bug in your ear–I spend lots of time tweaking those engine settings. It’s addictive. One number would suck the fun out.

I’ve got to admit this is probably the most realistic vehicle builder I’ve ever played. All day I deal with real-life engine numbers, test reports, and occasionally a broken prototype, then I come home and play Automation. It’s that much fun. I also play Automation at work. But it looks like I’m working. :grin:

PS edit: I’ve been an automotive engineer for years and I could dig up some useful charts if you think you could use them. It might surprise you how efficient a 4-bbl carb can be. Gas-crunch-era vehicles in the US had owners putting some very lean-running jet/rod combinations–a lot of burned and broken engines, but a lot of good data for the 70s and 80s. 22mpg from a 4000-pound V8 cruiser with a 4-barrel carb earned bragging rights. But like most engineers, I would never release anything because I never stop tweaking :robot:


#24

Maybe for the the Low, Mid and High Load AFR, one could have a Graph (like the Power Graph), but you could add points similar to how RGB curves work, therefor making it infinitely variable.

Trying to Illustrate what I mean…


#25

Nice idea, and nice way to present a complex AFR system, but holy hell that’d be complicated to teach people. And I don’t think would add much gameplay value to things.

I think we’ll remain with our current system, but display cruise mixtures better (and probably eventually tweak the effects of AFR changes)

Also @PinballWizard79, thanks for all that feedback, that’s really useful stuff to know, even if we don’t use it all! I’d love you to share as much data and experience as you can! Really cool to have some proper engineers playing Automation, and would absolutely welcome every little bit of feedback you have (as long as you know that some of the inaccuracies may just be intentional simplifications on our part)


#26

Exacly like that.


#27

I’m always happy to ramble on about the intricacies of real life and believe me, I know all too well the limitations of a sim. There’s always an echo from the software people “It didn’t do that on the bench” when vehicles get into the field. A bench test cannot anticipate real-world conditions. Not very well. And that’s why QA has such high job security.

Presets for those who don’t want to manipulate the curve might help those who are not into fine-tuning a fuel system. the perfect model for this is already in automation (the suspension). With suspension, you can select a preset, but a player can tailor better numbers to a specific vehicle application with a little practice, even if they know nothing about suspension, and there are 8 variables.

For gaming purposes, you might have two mixture sliders instead of one–one for power and one for cruise/idle. The lower octane number that you get when switching from carb to EFI seems to reflect the controller’s ability to prevent engine knock. The sim also gives EFI engines better marks for emissions and BSFC (efficiency)–it really is a8 pretty good sim.

The only other thought I have at the moment–carburetors do not have to be so inherently inefficient in-game. IRL, a big taper on the primary metering rods, which creates a starker difference between cruise and power AF ratios will give a carbureted engine pretty good fuel economy if the operator is disciplined not to use too much throttle. Running lean of peak at low power, the efficiency of a carburetor can compete with EFI, but when vacuum falls (part throttle), the carb does not have any O2 or knock sensors to gather data from, so it must enrich to a worst-environment safe point or the engine will knock and kill reliability. Automation simulates this very well. The only missing component is the ability to lean out the idle and enrich part/full throttle. Last thought on this–if the mixture change is high and the engine transitions abruptly between very lean and very rich, you get a lightswitch spot in the throttle that is hard to manage. This would impact driveability and comfort.

How far apart the cruise and power mixture should be set is quite similar to the current sim version’s VVL curves–if you set them too far apart in-game, you get an M-shaped torque curve. You get the same thing in real life. All in all I’m enjoying the game. Things have to be simplified for the game to be playable, but nothing is misrepresented.


#28

Thanks, that’s a really interesting thing to have a think on for when we eventually overhaul all the engine calculations one day!

Hey, slightly unrelated thing. I don’t usually ask people to leave us Steam reviews, but it’d be really neat, given your unique perspective as an actual engineer, if you’d like to leave us a Steam review sometime giving your perspective on how it relates to the real world :slight_smile:


#29

Look at what I stumbled upon today for a gasoline engine at full throttle (Book from Basshuysen). Timing is set for max. torque.


For non-german speaking: On horizontal the CR, on vertical left is MEP (solid line) and on right the ‘‘effective efficiency’’ (dotted line).