Home | Wiki | Discord | Dev Stream | YouTube | Archived Forums | Contact

How To Train Your Turbo By Dragawn

A temporary Turbo Tuning guide

So, I’ve been putting off this guide for years, since I never knew when the turbo revamp is coming, but as it still is not done, I’ll be finally writing the guide. Now I’m by no means an irl turbocharger expert, but hopefully I’ll be able to set new users on a good track.

Contents:

  • Turbo fundamentals
  • Current turbo limitations in Automation
  • How to train your turbo (Good automation turbo design practices)
  • Time, Place and Occasion for a turbo

Turbo Fundamentals

In order to tame the turbo, one must understand the turbo.

The principle:
An engine cycle essentially exists out of “suck-bang-blow”. Without help the only “breathing” an engine got is from expanding its capacity, causing air to flow in to fill up the new void. This is what is called “naturally aspirated”.
If you can stuff more air into the engine you have more oxygen to combust fuel with, causing a bigger bang. You can stuff more air in using a type of air compressor, but these air compressors need to be powered by something. A supercharger draws its power directly from the crank, while a turbocharger draws its power from the velocity of the exhaust fumes the engine blows out. A supercharger will directly reduce engine power, while with a turbocharger you reduce the engine’s capability to breathe out.

Tidbit: Now as usually the kinetic energy that’s in the velocity of exhaust fumes goes to complete waste after leaving the exhaust, a turbocharger drawing its power from there is generally more efficient than a supercharger. Also why turbocharged engines are quieter.

Turbocharger components:
image
The compressor
The compressor is the part that, with its turbine wheel, sucks in air. If it sucks in air faster than the engine can digest you’ll raise your air pressure. The more air pressure, the harder it is to stuff in more air. Without a blow-off valve these two effects will counteract eachother and determine the air pressure.

The blow-off valve
With a too high pressure in the engine, the fuel will combust before it should, and cause engine knock. To prevent this the maximum pressure of the air waiting to enter the engine is regulated by the blow-off valve.

The intercooler
After compressing air to a higher pressure, it has the nasty tendency to also increase in temperature quite drastically. Engines like cool air. These two effects don’t mix, so to cool the air back down after having been sucked in by the compressor modern engines use an intercooler.

The turbine
This is what makes a turbo a “turbo”. It basically does the reverse of the compressor, in order to power the compressor. It takes the high pressure from exhaust fumes being pushed out of the engine by the pistons, makes it flow through its turbine wheel, driving the compressor’s turbine wheel.
“But, what’s the gain then?” The gain is that naturally aspirated engines suck at sucking. And also that you can cram more oxygen and fuel into the engine than you would be able to naturally due to the higher pressures.

Turbine A/R Ratio:
image
The compressor generally doesn’t really care about its A/R ratio, but for a turbine it should decide its characteristic. A low A/R ratio will leave less area for the air to flow, but driving the turbine relatively efficiently at a relatively high velocity for the turbine size. A high A/R ratio will leave more area for the air to flow, at a relatively low velocity for the turbine size. big size low A/R ratio turbines will generally have more mass to them, causing more turbo lag due to inertia, while small size high A/R ratio turbines will generally have a low velocity going through them, causing turbo lag due to them struggling to be powered by small flows. Generally this means a low A/R is more efficient at low rpms, while a high A/R is more efficient at high rpms.

For automation? Ignore all this, smallest turbine with the biggest A/R ratio.

Turbo spool RPM:

When there’s not enough power from the engine available, turbine blades of the turbocharger simply don’t spin, as simple as that.

Turbo lag:

As you probably have guessed now, the engine and the turbocharger have a symbiotic relationship, providing power to eachother, the delivery of this power, through air flows, however can have delay. Better known as turbo lag. The length of the tubing decides how long it takes for air to reach the turbo from the engine or visa versa, and the size of the turbo decides how hard it is / how long it takes for it to get spooled up. The length between the engine and the turbine is the most important.

Though, as long you got enough power from the engine the turbo will kick in eventually, maybe not at max boost yet, but it could definitely reach max boost earlier than a dyno graph would indicate. A dyno graph is a “recording” of your car accelerating at a pace decided by the dyno after all. It also means that yes, even at high rpms you have turbo lag, although the turbine gets its food significantly faster at those high engine speeds.

Also more cylinders feeding a single turbo helps with turbo lag, as the flow of exhaust fumes is more constant and responsive to engine load.

Current turbo limitations in Automation

1 Turbo for inline engines, 2 for V or boxer engines:

Yes, this even counts for Boxer 4’s and the V16. This means that you can either have a hilarious amount of cylinders feeding the turbocharger, or a 4 cylinder that has 2 turbos. Which, thanks the to limitation listed below is limiting on itself. Also means bi-turbo or sequential turbo setups are not an option right now.

Turbo dimensions:

A big one, automation uses one standard exhaust manifold for all turbochargers, which keeps the turbo piping minimal. Sadly this also means that in old builds the turbocharger physically was close to hitting the block with a too small turbine or too large compressor. In the current build (as of time of writing) this is no longer the case, but they kept the same dimensional limits (Goddamnit Killrob). This means that in automation engines with more cylinders in a row can run relatively smaller turbos, while engines with less cylinders in a row could run relatively bigger turbos.

Turbo technology:

No technology here altering the turbo size (as perceived by the air flow) here. This tech rose in the 90’s, so this means that modern turbocharger engines in automation cannot use fancy tech to make themselves more flexible. But hey, even if that option was available in Automation it would cost budget and make the car less reliable, so we just got cheap turbos now.

How to train your turbo

This section is solely for turbo taming, I assume that you know the basics to making a naturally aspirated engine: setting your redline according to your components, engine size, dealing with knock, etc.

The initial/basic turbo set-up

  1. To begin with, start off making sure you got low enough compression, properly low so you don’t have to return to it when a change makes the engine require 0.1 ron/aki better fuel and start to knock. This is as compression doesn’t have much influence on engine characteristics other than flat power/efficiency increase, and we’ll want to nail that characteristic here.

  2. Choose ball bearings for the turbo when available, the small price hike definitely is worth it every time.

  3. Estimate how much power you want your new turbo engine to make. Now multiply that number by 2-2.5x. That’s your initial intercooler size. Later on you’ll want to fiddle with its size to find maximum power, or a compromise in power if weight and economy are huge concerns.

  4. Initial Compressor size, well, take a shot in the dark here for now, it is very important but it’s hard to know how much you need straight away in automation asides from experience, put it somewhere in the middle.

  5. Initial Turbine size / AR Ratio: Minimum turbine size, maximum AR ratio. This is the most efficient way in automation to get the most turbine power at the least lag.

  6. Initial Max boost: Want economy / low lag? Go low, around 0.55 bar / 8 psi should be fine to start off. Power? 0.8-1.5 bar. All depends on how much lag you’re willing to accept. You can go for 2.5 bar if you’re really building a racing engine for a competition with long straights and smooth turns. But keep in mind that even in F1 they sometimes dialled back their boost to below 2 bar for driveability.

  7. Intial Exhaust diameter: Make it big enough to allow for about 2x the power you want. You can go more if loudness and some weight is of lesser concern, or less if loudness is a big concern.

  8. Activate flow bench, which can be done on the right graph window:

  9. Make sure your turbine and Compressor are not 1.00 already in this flow window, this probably means they’re already too large for your engine and purpose. Reduce them in size (or A/R ratio for the turbine) until you start losing power.

  10. Now it is time to assess your turbo lag - power.
    For some reference, the turbo kicking in fully at 1800-2200 is non-laggy / family car territory, 2200-2500 is modest lag for modest performance (think Golf GTI), 2500-3000 is getting quite laggy, but still liveable ('18 Honda Type R), 3000-3500 is your classic laggy car, 3500+ is race engine territory.
    Reduce your compressor and turbine a/r if you feel you’re too laggy for what you’re aiming for, otherwise you can ramp up the sizes, but ramp them up in balance so the airflow restriction in the flow bench is about equal.

  11. Time to evaluate your boost pressure. a higher boost pressure often will net you more power than increasing your compression ratio, especially your torque will increase drastically. This is however at a trade-off: your fuel economy will be worse, worse reliability, worse driveability and worse aero due to increased cooling needs. Not to mention that carburators cannot handle high boost pressures if you’re using a turbo this early into the game.

  12. The icing on the initial cake: optimize your intercooler size for optimal power, or a bit less for less weight. Optimize your exhaust size to get more power, balance it out with your loudness (comfort) and weight. Last but not least, increase your compression ratio and/or engine timing to fully use all of your fuel’s octane.

Congratulations, you now got a turbo roughly suiting your needs. Your torque graph probably could use more refining though, to make it a better feeling engine, or a better engine in general. For this there’s…

Advanced turbo tuning / making the engine refined / flat torque guide

Now, there won’t be any step-by-step hand holding for this one, as it’s heavily engine dependant. I will explain here though how to alter your engine torque characteristic and what to look for.

What to look for? Well, the ideal torque graph is often considered as “flat” as possible, meaning your torque remains constant through the graph. This makes the engine feel neutral when accelerating, giving you linearly more power as you ramp up the RPMs. When your torque decreases the engine will feel “choked”, as if it’s running out of breath the higher in RPMs you go, still giving more power, but struggling to do so. When your torque increases the engine will feel “pushy”, like it wants you to go higher in rpms because the power increases exponentionally. So yeah, linear torque feels the most predictable and satisfying to the driver. You always should have your torque decreasing near the redline though, so the power flattens out, and the rev limiter doesn’t feel like a surprise, so the engine doesn’t want to go past the rev limit, but so that the rev limit also is the engine’s power limit. Generally your rev limit should be 5-10% past your peak power to optimize your engine, but this may vary and is very much just a rule of thumb. Also a neat little tidbit: a short sharp torque increase is that “VTEC KICKED IN YO” sensation.

How to tame your torque graph then? Well, come along for this ride, it’ll be a wild one. You can control your torque mostly by 5 different settings: your camshaft profile, ignition timing, compressor size, turbine size + turbine AR Ratio and Max. boost.

Camshaft profile generally sets your engine preferred RPM to operate at, putting your peak torque there. With VVL in automation you can basically choose 2 preferred RPM, which may come in handy, but also at the cost of not being able to use 5 valves, being heavy and costly (especially in engineering).

Ignition timing, well, there’s usually an optimal ignition timing in trade-off for compression ratio for optimal power, but it can shift some torque bumps in your graph and help flatten them out, I like to use it for that. Also makes your engine a little more/less responsive.

Compressor size + turbine size/AR. These two are closely related, but there’s a neat difference in a less effective compressor vs a less effective turbine. A lesser compressor will quite linearly decrease your torque up top as it struggles to feed the engine. While a lesser turbine will cause a sharp drop-off of engine torque, as it struggles to feed the compressor to feed the engine, which in turn feeds the turbine less. Using this you can shape how your top end torque looks quite effectively together with your camshaft profile.

Max. boost: With turbocharged engines, more boost generally means a more decreasing torque line, also ofcourse more torque in general.

If you got a “pit”/hollow in the middle of your torque graph, you can increase your camshaft profile if it’s quite low (but remember, low camshaft profile gives good fuel economy), and/or decrease your upper camshaft profile if you got VVL, or decrease your compressor size. If your torque graph is convex you can sort this by doing the opposite.

Peaky torque at high rpms? Time to murder that compressor size and/or camshaft profile if it’s high (VVL).

This won’t give you optimal power, but a satisfying engine to drive, I gave small pointers here, but the best thing to do is start fiddling a lot and see how the engine reacts to all changes yourself. If you want more power, you can always put a small “V-tec kicked in” bump in your profile between two flat torque lines, but this can be hard to pull off well, and disturbing to the driver if they aren’t driving full attack.

The “Turbo float” technique

WIP

Small disclaimer to end with:

This may not give you the ultimate best turbocharger result in every case, but right now in automation it’s great especially if you have some form of realism / actually useable powerband in mind.

Time, Place and Occasion for a turbo

WIP

Changelog:

21/07/2019: Release version with "Turbo float" and TPO in WIP
21 Likes

We’ve been waiting for your turbo tuning guide for the past few years, and having just read it, I have so far found it to definitely be worth the wait as it is right now.

4 Likes

I’ve been saying this for a very long time, but…

Boost threshold is NOT equal to turbo lag.

Where the boost comes in the rev range has basically nothing to do in reality with how much turbo lag an engine has. Turbo lag, by definition, is the difference in time between when the driver commands a power increase, to when that increase actually happens. Boost threshold plays a part, sure, but it’s far from the only thing that affects it; gearing, for example, plays a huge part in determining how laggy the powertrain turns out to be.

6 Likes

Cough
Boost threshold =/= lag, but boost threshold however IS an indication of lag, as it is when the turbo kicks in on the dyno. Always use the same dyno in the same way and state (as Automation is a simulation, this is true) and you’ll have a good reference how laggy the engine is by itself.
Gearing is an influence as it affects both engine load (the pace at which the revs will increase) and the engine speed (rpms); So in my opinion when comparing engines to engines, this should not be in consideration other than an engine revving to 9000 rpm having a boost threshold at 3000 rpm, obviously is better than an engine going to 6000 with a boost threshold at 3000 rpm. But yeah, I should include that.
Guess I’ll clarify that section more that these are merely an indicator.

4 Likes