Project F82 M4: Part 1 – M for Midlife Crisis?

After nearly a decade of responsible financial choices and reliable cars, I did what any car guy would do … I bought a used, modded BMW to replace my perfectly reliable Honda Accord. In true midlife crisis fashion, I opted for a 2016 BMW M4 with the “slower” 6-speed manual gearbox, low options, carbon slicktop roof, and some tasteful modifications.

F82 M4 Specifications

The F82 M4 is a platform that is in the Goldilocks zone of stock capability, tunability, and affordability. It produces strong power from the 3.0L twin turbo Inline-6 S55B30 engine backed by a rugged drivetrain, substantial cooling capacity, a reasonable suspension design, and brakes suitable for occasional track days.

Source: BMW Technical Data and Empirical Measurements

Compared to the 4-door F80 M3 the M4 coupe is 10 lbs lighter, 5% more aerodynamic, and 40mm shorter. It’s even lighter and more aerodynamic than the M2 Competition!

The F80 / F82 both underwent substantial amounts of weight reduction from the E90s. BMW engineers went through the car extensively, looking for every gram to reduce, including hydroformed, hollow camshafts and utilizing carbon fiber throughout the vehicle, such as continuing the carbon roof from the E46 M3 CSL and E92 M3.

Lightweight components for the F82 M4 include:

• Carbon single-piece driveshaft
• Carbon roof reinforcement (M4 exclusive)
• Carbon fiber roof panel
• Carbon composite trunk (M4 exclusive)
• Aluminum hood
• Aluminum fenders
• Forged OEM wheels
• 2-piece brake rotors
• Tubular rear subframe
• Aluminum front subframe
• Aluminum control arms and knuckles
• Aluminum / carbon chassis bracing
• Lithium-ion battery

So as you can see, many of the common areas to cut weight on other platforms have been covered by the OEM design. This is a big reason why special badge versions like the GTS and CS are often heavier than the base trim, which led to automotive writers and M fans claiming that BMW had “gone soft” and lost their willingness to trim the fat since the E46 CSL and other cars had cut so much weight.

A reason those cars could cut so much was that they weren’t nearly as weight-optimized in base trim as they were in CSL/GTS trim.

F82 M4 Baseline

The twin-turbo S55 3.0L Inline-6 engine is rated at 425 HP / 406 lb-ft of torque and was BMW’s replacement for the naturally-aspirated S65 4.0L V8 in the previous generation E90 M3. At 452 lbs (dry) the S55 weighs 3% less than the S65 V8, which seems lackluster until you realize a stock 2JZ-GTE or RB26DETT is over 600+ lbs (272+ kg) and a comparable GM 6.2L LT-1 is 465 lbs (dry – GM data).

The F8X S55 is based on the previous N55 engine that powered the 335is, 535i, and many other BMWs starting in 2009 – 2010, with 75% of parts being shared with the N55 and the remaining 25% being S55-specific. Major improvements to the S55 include:

• Closed-deck block (5 lbs lighter)
• Twin monoscroll turbos with e-wastegates (N55 is single, twin-scroll, pneumatic)
• Full-counterweighted (12cw) nitrided forged steel crankshaft (N55 std is 8cw, iron)
• Mahle M174+ (AlSi12Cu4Ni2Mg) cast alloy pistons (N55 std are Kolbenschmidt)
• Nitrided piston rings – ES profile oil control ring
• Strengthened piston wrist pin in hardened steel (16MnCr5)
• Flow optimized primary oil pump (18% more flow vs N55)
• Additional oil pump (suction pump)
• Lower temp oil cooler thermostat
• Cylinder head coolant passages for direct injectors
• Improved pressure balance in crankcase (larger windows above mains)
• Magnesium oil pan with extra baffle plate (2 lbs lighter, earlier motors only)
• Water-to-air intercooler (N55 uses Air-to-Air) with its own low-temp cooling circuit
• Twin HPFPs (N55 is single)

It’s worth noting that the F87 M2 has an upgraded N55 that has a few of these S55-spec updates as well, and the M2 Competition is equipped with the S55B30 engine.

Extremely German changes to the S55 include:

• The Most Complicated Cooling System in the World
  “Das ist sehr gut und natürlich optimal”
  • 5 Radiators
  • 2 Separate cooling circuits
  • Water-to-air charge cooling
  • Large oil cooler
  • Mechanical and electric cooling pumps

Baseline Dyno

Initially, the car was on the Stage 3 Dinan map, but while driving it back to Austin, the midrange felt inconsistent, particularly as boost ramped. A few days later, I messaged Juan Vargas at Bardabe Performance to arrange some runs on their Dynapak hub dyno to get a baseline and see what the torque was doing in the midrange. Specifically, was it consistent and smooth, or did it show a roller coaster with wide swings?

However, before heading out, I thought I’d ask everyone at the shop (SPL Parts) to guess the power numbers for a chance to win free lunch.

A few of the guys were a bit optimistic for 100°F dyno numbers on 93 pump gas, haha.

The very first pass on the Stage 3 Dinan tune resulted in 419whp and 449wtq with a short peak that quickly declined to 390wtq with a wavy plot resembling stair steps.

This is important to know because the BMW calibration reacts to knock by immediately dropping ignition timing on that cylinder (individual cylinder knock correction) by roughly 1.5° each count until the knock stops. This happens nearly instantaneously, as the ECU (DME) must take enough timing out to get ahead of a vicious cycle of one knock event feeding the next. The ECU then raises the ignition timing in slow steps back up to the target value, resulting in the torque output having a wavy, stair-step shape.

You can see this in the green plot, where torque diminishes faster after 4000 RPM than the other plots due to a slight heat-soak pushing the engine into knock.

After the first run, I switched the car over to the Stage 2 Dinan tune. On 93 octane (98 RON) pump gas, the M4 produced a baseline of 443whp and 462wtq on the Dinan Stage 2 93 map. This was a much happier torque plot, and afterwards the car felt smoother, presumably as a result of less knock correction in the warm weather.

Unfortunately, I had no way to confirm this, as the Dinan piggyback doesn’t have datalogging, and the OBDII data would be false due to the signal being intercepted by the Dinan controller. In fact, I had no way to even know how much boost the car was running.

If there’s one thing that really, really grinds my gears, it’s not having any data or insight into what is happening, especially when it comes to metrics critical for keeping the expensive BMW engine intact.

Current Mods (Baseline Dyno)

• Dinantronics Piggyback (S2)
• Dinan Axle-back Exhaust
• Active Autowerke Single Midpipe
• Dinan Carbon Intakes
• Mishimoto Charge Pipes

“What’s the Plan? … When’s the First Track Day?”

I have given this some thought and have come to 3 conclusions.

1. M4 CSL
2. Analytics Driven Design
3. Pushing My Abilities

#1 M4 CSL

The CSL badge stands for “Coupé Sport Leichtbau” and is a factory-built lightweight (“leicht” light / “bau” build) created by BMW for a limited number of chassis since the 1970s.

For the F82, BMW M decided to give us the M4 GTS instead, which is even more track-focused than the CSL and features the same “leichtbau” base with a 4-pt roll bar, carbon splitter, carbon hood, GT4-esque rear wing, water injection system, fire extinguisher, and lightweight seats with harnesses.

I intend to build something similar to the GTS, but rather than call it a GTS, I’ll borrow the CSL nameplate that BMW didn’t bother to make. I suppose in typical BMW parlance, what I want to build is a “GTS Evo.”

A dual-purpose M4 that is as optimal as possible for both purposes: street and track.

#2 Analytics Driven Design

I intend to design, develop, and implement a datalogging system that combines traditional systems with a comprehensive suite of sensors designed for discovery, development, and education.

This will include datastreams that are implemented at the OEM level in test mules and frankly, never seen outside of that environment. This is to leverage my analytical data skills and push me past the comfort zone of niche, small-scale logging setups and into a comprehensive, unified system. I intend to self-source anything that is beyond the typical AiM Solo level of attainability, as having a turn-key Motec or similar aftermarket logging suite, while convenient, is well past the budget of most enthusiasts.

Ultimately, I want this data to inform active elements on the car. Aero, cooling, power output, and more.

#3 Pushing My Abilities

Over the last decades, I’ve learned a substantial amount of theory, been exposed to all manner of projects, surrounded myself with people who know more than me, and used that to help niche projects for strangers and friends alike as a means of applying that information. Engine calibration, shock valving, datalogging, data analysis, aerodynamics, tire behavior, suspension setup, and more.

It’s time to act on all of the crazy ideas, do something different, learn something new, prove myself wrong, reinvent the wheel, learn how to drive a car again, and probably fail a lot.

After all, it seems the car community has devolved too far into cookie-cutter “builds” (my F82 included) that shy away from deeper refinement for a multitude of reasons (cost, risk, lack of skills?, perhaps a lack of confidence?).

At the risk of being egotistical, I want to push myself out of that rut and help others along the way.

Project F82 Sponsors

SPL Parts

Bardabe Performance