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

I decided to study the affects of different Compression Ratios and Ignition Timings in Automation: The Car Company Tycoon Game.

Tested in the OPEN BETA UE4 (Update 2) version of the game on 03/06/2017.

Engine Specifications:
Configuration - NA Inline 6
Bore - 92.2mm
Stroke - 87.5mm
Capacity - 3505cc
Fuel - 100 Octane
Camshaft Configuration - 4 Valve per Cylinder DOHC (VVL and VVT enabled)
[Quality +7] Crank - Billet Steel
[Quality +7] Connecting Rods - I Beam Titanium
[Quality +7] Pistons - Lightweight Forged
[Quality +2] Camshaft Profile - 60
[Quality +2] VVL Profile - 100
[Quality +2] Fuel System - Direct Injection
[Quality +2] Intake - Race Throttle per Cylinder
[Quality +2] Fuel Mixture - 15
[Quality +2] RPM Limit - 8700
[Quality 0] Headers - Race
[Quality 0] Exhaust - Single 76.20mm; Cat - High Flow 3 Way; No Mufflers

- 12.1:1 Compression= 303kW; Efficiency 28.08% @4100rpm
- 12.4:1 Compression= 301kW; Efficiency 28.02% @3800rpm
- 12.6:1 Compression= 299kW; Efficiency 27.87% @4000rpm
- 12.8:1 Compression= 296kW; Efficiency 27.57% @4500rpm
- 13:1   Compression= 290kW; Efficiency 27.07% @4500rpm
- 13.2:1 Compression= 284kW; Efficiency 26.42% @4300rpm

Excel Table (Charts didn’t work…)

Some interesting Results… What do you guys and gals think?

The loss of power at higher ratios, are you sure the engine isn’t knocking?


Yes, I lowered the Ignition timing to a value that resulted in a Octane Rating of 99.9 RON
Double Checked…
The loss of power seems strange as I have never heard of it happening at high compression ratios? (In Reality)

What are you doing with the ignition timing? The more retarded the timing the less efficient the engine runs (especially below 50% timing), even if you bring compression up to meet the knock limit. Ignition timing changes also slightly affect the torque curve, so you might need to adjust valve timing (cams) to achieve maximum performance.

1 Like

What are you doing with the exhaust diameter to get the scavenging just right and gain some power?

Efficiency of the burn does not equate to efficiency of the engine–high compression engines suffer terrible pumping loss–the energy to compress is a function of the cube of pressure (I think). So you’ve got this super awesome burn, but you rob energy to compress the air/fuel. Eventually the energy robbing overruns the increased efficiency.

On fuel mixture–High compression should create more potential for power, but you have to enrich to reap the benefit, especially with a high cam profile. The lean mixture combined with the superdense air ends its combustion before the optimal time in the power stroke. A rule of thumb is just prior to EVO.

The burn time in the 12.5+ : 1 compression ratios is getting shorter and starting just as early. A (slightly) greater percentage of the combustion energy is released before the top of the compression stroke.

A high spark advance aggravates this problem further both IRL and in-game–you will notice if you increase the spark advance, then the advance produces great advantage up to a point, then it doesn’t really help or hurt much until it reaches another point where it kills the engine with detonation.

Side note: One missing dynamic is reliability vs. fuel mixture–lean, high compression engines make so much power they put holes in pistons, reforge valves, and crack heads, which is another very good reason to run richer, and all manufacturers do it–adding more fuel to burning fuel cools the cylinder and head.


Interesting, so if I understand correctly, one is better off running lower compression for fuel efficient engines (so that the AFR can still be 15:1 and efficient)?

ICE thermodynamics are very complicated. A higher compression ratio should yield more efficiency in general. But that means your engine wasn’t running the best setup in the first place.


I posted a very similar test years ago, you might want to include your ignition settings for clarity. It’s been a long time, but I remember on my particular test engine there seemed to be an ideal compression/ignition compromise in the middle, with performance tapering off in either direction. That was a lower revving engine, however, and I’ve noticed that high revving engines like yours usually prefer lots of timing. Maybe try building a low RPM motor and try the test again to see if the results change.


Will do some more tests soon…

1 Like

Not precisely. It’s a huge string of balances, compromises, and assumptions. There is no perfect compression ratio. You might be surprised at the economy that a low-compression engine with a lean-of-peak mixture and low spark advance can get. It won’t be powerful, but it will be relatively fuel efficient, strictly in terms of g/kwh or %efficiency. Its power to weight will suck though. To produce a car with high MPG, you have to consider engine weight.

You also need to line up, as best you can, your peak efficiency RPM with the vehicle’s intended cruise rpm. Since the latter is a variable, you’ve got another string of compromises to make. A long, flat efficiency curve with a peak of 0.6#/hph is better than a sharp trough with a 0.5#/hph in a teeny sweet spot, although you can mitigate that with a lot of gears. Just ask Chrysler–they spend more time hunting for the right gear than they do IN the right gear. :-p

Auto Test engineer & QA since 2002. I’m not at liberty to name my employer, but I can say that we all make fun of Chrysler. The stuff above is crude, but If you want some gory detail, feel free to ask.

Happy motoring.


Engine weight is a very minor thing in fuel efficiency, at least in Automation - for similar power possible differences in engine weight are about 1-3% of the car weight. And in Automation AFR is quite directly linked with efficiency - leaner mixture, higher efficiency. Considering cruise RPM doesn’t help, because they are determined by the game in a very unclear way, probably tailored to engine’s efficiency graph, so there’s no need to worry about that, at least in my experience. One thing I’ll 100% agree with is the flat curve vs low sweet spot.

I’m not questioning your knowledge, but this game is not a perfect simulation, so real life experience =/= Automation experience. For example, many benefits of the undersquare designs aren’t really present at all or are very minor in the game. There are no problems with going leaner than 14.7:1 AFR, although AFAIK IRL that would be harmful for the catalytic converter and valves.

1 Like

IRL, anything lean-of-peak is going to run hot except at very low power. A carbureted engine (read: not actively managed) will not last running lean, and that is a big important metric that is completely absent from the game.

One thing that fuel mixture absolutely CANNOT be in an automobile is a fixed number. 13:1 while idling fouls plugs and lets gasoline drip past the rings into the oil. Under power, 13:1 is on the lean side of reliable. Most carb’d engines had secondaries that ran rich of 12:1 to keep the temp under control.

Both rich and lean are problematic IRL for catalytic converters–high EGT from being lean or unburned fuel entering the cat both heat the thing up. The mounts for the ceramic catalyst fail, the catalyst rattles around in the housing for a while, then it shattters a little at a time and blows chunks out the tailpipe. Or it burns through the housing and falls out. Either way it’s bad.

On engine weight–IRL it’s not just the weight of the block for an OEM, the frame has to get heavier to stay in one piece. In-game, vehicles don’t lose reliability from being overpowered to the point that a real frame would twist. But you are correct–swap block A and block B, the weight difference doesn’t matter much IRL or in-game.

I wish the in-game fuel economy was less cryptic. The EPA test isn’t much better–about 50mph is it for the EPA highway estimate. It gave the industry uniformity, but has not been altered since the repeal of the 55mph national speed limit. Automation might use a European standard for all I know. It produces poorer fuel economy numbers, but anybody who buys a car today usually finds the EPA Hwy estimate hard to beat. I can’t drive 55, not even on my old Honda rebel. Maybe on a windjammer…


Vast majority of economy gains in Automation are in cruising and, as far as I can tell, the cruising tests are executed in the highest number gear (lowest RPM) possible that has enough torque to maintain speed and not stall (unclear what throttle position). Unless your car has very poor power to weight ratio or inappropriate gearing (or a turbo, which effectively reduces the power-to-weight ratio dramatically before spooling, making the test avoid using pre-spool RPMs unless the engine is huge) then the better the efficiency near idle, the better the car’s economy. This is why using 0 setting on the low cam with VVL gives the best economy in most non-turbo applications. Unfortunately, a nice wide ‘economy band’ is relatively useless.


enough torque

Ahhh, make the pain stop! xD It’s power you need to keep a certain speed, torque is completely irrelevant unless you also say at which RPM it is, i.e. how much power you’re making.

There are a mistakes in there I’d like to correct, not blaming you for those (only for torque ;D), as we have not explained that at all in the game yet. The economy calculations drive the car in every gear possible (if enough power is available) in every cruising speed test and then select the most efficient variant for each of the tests. The tests themselves do take into account efficiency at that RPM, throttle position and pumping losses associated with the throttle position, drivetrain losses, and of course drag, etc.


I understand the criticism of saying ‘torque’ but I defend my use in the context of the rest of my statement, which mentioned gearing and RPM–it’s redundant and unspecific if more technically precise to say ‘power’. Power is an abstract value in reality, so I prefer to refer to RPMs and torque separately, especially in the context of ICEs mated to finite gear sets, devices physically characterised by torque curves and fixed ratios.

1 Like

Yeah, the fuel mixture thing was originally a representation of fuel mixture under power… or average fuel mixture… or something like that, I can’t quite recall.

Really we should probably change it to be an arbitary number, like Ignition timing is, that just generally represents how rich you tune things on average (which is all the current slider is anyhow) but people seem to hate it when we do things like that :wink:

The calculations assume that it’s running leaner than the slider at cruise, and the more advanced the fuel system, the leaner it’ll run at cruise. So that mostly works.


(Just to elaborate on what I am trying to get at, I would stop talking about torque if the engine was mated to a CVT because then the power curve becomes much more representative of the physical performance of the system. It’s along the same line of thought as how air and space drive systems, like jets and rockets, tend to be discussed in terms of thrust (analogous to torque) because the power output isn’t directly relevant.)

I believe the slider has the leanest mixture possible listed when you point to it… or it used to anyway.

Oh, and arbitrary numbers are indeed a sin if you want my opinion. :stuck_out_tongue:

As for the original topic… I have noticed that the ratio of compression to timing also seems to depend somewhat on the cam profile used. It usually tends to favor more advanced timing over higher compression in most cases until compression falls under 8 to 1, and then the opposite becomes true.

I tend to use certain specific cam profiles in the engines I build; with 33 being a common one I use, and I’ve noticed that 9 times out of 10 I’ll use a timing of 51 along with it.

Ahh yes, but what is the worse sin, arbitrary numbers with fake real numbers attached to them (like AFR now), or straight out arbitrary numbers that admit what they are :stuck_out_tongue: (I’m honestly not sure!)