1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Cars 82-93 BMW 3 Series "E30 Experience"

Discussion in 'Assetto Corsa Modding Discussions' started by Seria17hri11er, Jun 13, 2016.

  1. WIP by
    Modek-- Model
    Serial Thriller-- Physics
    Sounds--Maybe Serial, or Modek, or someone else? We have good samples for 318 and will have 325 samples soon enough.

    Most Recent Model Update:

    April 4

    [​IMG]



    [​IMG]


    Most Recent Physics Update:




    Original Post

    Howdy all.

    Modek and I are working on the

    The E30 Experience.
    "What is that," you say? Well, first let's talk about the E30 M3:


    BMW's designation of the highest performing E30 M3 is simply M3 Sport Evolution. Naming convention went like this:

    M3

    M3 Evolution 1

    M3 Evolution 2

    M3 Sport Evolution -- Named by everyone else as Evolution 3
    Source: BMW M Registry: http://www.bmwmregistry.com/model_faq.php?id=8

    M3 Evo 3 in AC
    ---------------------------------------------------------------
    Engine
    S14B25 2.5 Liter, Inline 4 Cylinder - 238 HP

    Weight
    1200 KG - Weight Distribution 53 (52.8) Front, 47 Rear

    Suspension
    "Increased caster angle. The M3 had specific solid rubber offset control arm bushings.
    Aluminum control arms and the front strut tubes were changed to a design similar (bolt on kingpins and swaybar mounted to strut tube) to the E28 5-series. "

    Brakes
    "The E30 M3 had special front and rear brake calipers and rotors. It also has a special brake master cylinder."

    ---------------------------------------------------------------

    "Ok, so what is the E30 Experience?"

    It will be a collection of several different types of E30s. You see, E30 M3s are very rare. However there are tons of E30s out there and they are quite popular cars. Allow me to show you some variations I am planning:


    Standard 325i (Post facelift - 1990) - M20B25 - 2.5 L Inline 6 Cylinder - 169 HP - 1275 KG. Weight distribution 53/47.

    [​IMG]




    318is


    *Edit 2/11
    320is?
    *

    325i S52 (E36 M3 Engine) Swap - 3.15 L Inline 6 - 240 HP @ 6000 RPM, 240 Lb·Ft @ 3800. Has variable valve timing which provides a higher and flatter torque curve. Somewhere around 100 - 300 pounds more from slightly larger engine + iron block. Plus larger tires/wheels, brakes, suspension upgrades; coilovers front and rear, camber adjustment, anti roll bar upgrades etc.

    325i M20B25 Turbo - 2.5 L Inline 6 Cylinder - XXX HP? - 12xx KG? Plus larger tires/wheels, brakes, suspension upgrades; coilovers front and rear, camber adjustment, anti roll bar upgrades etc.

    325i S52 Turbo? - 3.15 L Inline 6 - XXX HP? Plus larger tires/wheels, brakes, suspension upgrades; coilovers front and rear, camber adjustment, anti roll bar upgrades etc.

    325i V8? o_O:D:cool: Link. Plus larger tires/wheels, brakes, suspension upgrades; coilovers front and rear, camber adjustment, anti roll bar upgrades etc.

    Spec E30? Link. M20B25, Roll cage, gutted interior, Getrag 260 5-speed transmission, 3.73 rear differential final drive ratio, Stock/unmodified ECU, H&R Race spring set, Bilstein Sport shocks, Spec series tire (which is currently the Toyo RR or RA1 in 205/50/15 size), minimum weight with driver at the end of the race is 2700 pounds.

    PBMWC?

    Why do I want to do this?
    Well, first off, an E30 convertible is my daily ride. Unfortunately it is auto--but it still handles marvelously! I very much in the future would like to purchase a standard manual transmission E30, ideally with a serious engine upgrade. I won't be able to do that for a long time. So, I would love to bring these different versions to AC as a substitute!

    Who the hell am I? Why would you want to join me? Do I have physics experience?

    Physics
    This is my first foray into modding. Fret not, I have many good friends here on the forums already providing immense help.

    Thank you @nighteye and @aphidgod for the huge help in physics modding. Also huge thank you to @felipe vaiano for his AC worksheet.


    I have made--I am guessing--20-25% progress in about a day on the standard 325i. I have an intimate understanding of how cars work, and a very hands on technical background.

    Who am I?
    I am a gearhead. I haved loved cars since I was five years old. My dad was a mechanic for most of his life. At 5 I would go into the garage and find an engine torn apart with my dad working on it. I was amazed at the complexity and all the parts. I would ask my dad tons of questions about it. It became an obsession understanding how technical things such as engines work.

    Fast forward 13 years. My dad didn't encourage me to work on cars; his experience as a mechanic wasn't so fruitful. He did get me very interested in aircraft. So much so that I joined the U.S. Air Force at 18 and became an Aircraft Technician specializing in the hydraulic systems that control the Landing Gear, Flight Controls, Cargo Ramp and Door, etc. I was in for 11 years, with 9 years working on aircraft. No, sorry to burst your bubble, I did not work on any fighter aircraft--although I do fly them in the simulator, and I do have a private pilots license :p. I worked on C-130s, C-17s, and KC-135s.

    Why would you want to join me?
    Because the E30 cannot be represented solely by the M3 Evo III. "While the E30 BMW M3 is, of course, an even more iconic performance car, Cammisa notes that the rarity and value of those cars means most are hidden in storage and are rarely put through their paces...

    'The BMWs are impeccably behaved driver's cars and they're impossibly reliable,' Cammisa says as he slides one of the rally cars through mud...

    'That's a nice little piece,' Pobst says after putting the wagon through its paces. 'The level of grip is so high, I could just feel the speed through the corner. Really well balanced.'"

    http://www.automobilemag.com/featur...-e30-bmw-3-series-is-still-so-much-fun-today/



    Do watch this video. It is fantastic.


     
    • Like Like x 7
    • Love Love x 3
  2. For those that haven't been following progress. AC does not support Trailing Arm suspension. So you have to engineer a DWB or McPherson that performs in the same way.

    Engineering of Rear Double Wishbone suspension that behaves like E30's Rear Semi-Trailing 15 degree sweep complete. Big thanks to @mclarenf1papa for his assistance in this effort.

    E30 Rear Suspension:

    [​IMG]





    Suspension points measured by me on my E30, and placed into Performance Trends' Suspension Analyzer program:

    Things to note:
    • Camber Gain
    • Toe-In Gain
    • Roll Center Height (black dot beneath chassis center)
    • Instant Centers (Dots on right and left side of tires)
    • Ignore Spring and Damper positions, correct placement unnecessary for this


    Front View

    [​IMG]





    Side View

    [​IMG]





    Top View

    [​IMG]





    The Double Wishbone build.

    Things to note:
    • Camber Gain
    • Toe-In Gain
    • Roll Center Height (black dot beneath chassis center)
    • Instant Centers (Dots on right and left side of tires)
    • Ignore Spring and Damper positions, correct placement unnecessary for this


    Starting with a a 2001 Corvette rear end:


    [​IMG]






    It's Instant Centers:

    [​IMG]







    Progression of modifications:



    [​IMG]






    [​IMG]






    [​IMG]






    [​IMG]





    Final DWB:


    Things to note:

    • Camber Gain
    • Toe-In Gain
    • Roll Center Height (black dot beneath chassis center)
    • Instant Centers (Dots on right and left side of tires)
    • Ignore Spring and Damper positions, correct placement unnecessary for this


    Real E30, again for comparison

    [​IMG]






    DWB

    Front View


    [​IMG]





    Side View

    [​IMG]





    Top View

    [​IMG]






    E30's Instant Center Height, Instant Center lateral position from center, and Roll Center Height

    [​IMG]


    Anti-dive not shown above, but it is 42%. From what I have read, anti-dive is too strong in AC and the suspension must be modified to make it softer.


    DWB Instant Center Height, Instant Center lateral position from center, Roll Center Height, and Anti-Dive


    [​IMG]



    The progression pics really don't do justice to the amount of time and changes made. Super steep learning curve. I wish I had a time lapse video of the changes. It would look like a constantly shifting Amoeba.



    Finally, and very importantly, graphs.



    E30 Camber Gain

    [​IMG]






    DWB Camber Gain

    [​IMG]






    E30 Toe Gain

    [​IMG]





    DWB Toe Gain

    [​IMG]






    All this took over 16 hours to complete. It was difficult, and tedious. It was also stressful because the Suspension Analyzer program has a demo for 10 days. After that, purchase for $245. Installed on 2nd computer after 10 days ran out to continue using demo.


    I will have some additional information on the Roll Centers "soon."
     
    • Like Like x 5
    • Wow Wow x 2
  3. Alright, if you took them off, put your nerd glasses back on :geek:, or, like we called them in the military; Birth Control Glasses (link).

    So, Roll Center.


    [​IMG]



    So, the further the Roll Center is from the Center of Mass/Center of Gravity, the more the vehicle will roll.


    The roll center is a not a static value. It changes during braking, acceleration and cornering. It can move left and right, up and down. The important thing about it's movement is simply how far it moves away from the CG. The more it moves, the more roll is induced.

    The E30s static CG is approximately 18 inches above the ground---and that's a very big approximation as I haven't found any real good data on the CG height. All I have found from a book on E30 tuning is "Most sedans like the E30 have a CG around 18 inches above ground."

    The static Roll Center is 4.66 inches above ground. 18-4.66 is 13.34. So the Roll Moment Arm is approximately 13.34 inches long.


    That being said, here is how much the E30s Roll Center moves during 2 inches of squat--which would be acceleration--, followed by 2 inches of roll to the right:


    Skip to 10 seconds to clear watermark





    As you can see, it really does not move much at all up and down. 2 inches of squat causes the roll center to move down about .8 inches. That's really good, as the CG also lowers during squat, so the distance between the Roll Center and CG actually decreases, minimizing subsequent (not initial) roll should we begin a turn during this squat.


    During roll, the Roll Center moves to the side about 4 inches. Since this vector is not directly away from the CG, the distance it moves away from the CG is less than 4 inches. Additionally, the CG does not move much at all during a roll.




    Here is what I was able to achieve with the DWB:

    Skip to 10 seconds to clear watermark






    You can see from previous images and this video, that I was able to match Static Instant Center locations and Static Roll Center locations. So the initial Roll Moment Arm is the same as the E30; 13.34 inches long.

    From the video, you can see a 2 inch squat moves the Roll Center down 2.12 inches. Unfortunately, significantly more than the real E30. This was the best I could do with the limited time I had. I am not sure it can be improved from that. The only way I could find to reduce it, was to lengthen the A arms. That is why they are so long and reminiscent of an open wheeler.

    However, during roll, the Roll Center barely moves at all; just .24 inches. This is good as the condition that could move the Roll Center the furthest away from the CG is lift (braking) or squat (acceleration) coupled with aggressive cornering (roll); so turn entry or exit. So these don't couple together to create a huge increase of Roll Moment Arm and disparity between the real E30. However, the real E30 does have a more progressive amount of roll after initial roll.

    These minor--I think--differences between the E30 will be most noticeable during:

    Very first phase of turn entry: more initial roll than should be

    Last phase of turn exit: more roll near completion of turn exit

    Initial hard brake application and steering inputs: more initial roll than should be

    Initial acceleration and steering inputs: more initial roll than should be

    Steady state cornering--no acceleration or braking; maintaining constant speed through corner: a bit less roll than there should be.
     
    Last edited: Jun 14, 2016
    • Like Like x 5
    • Agree Agree x 2
    • Wow Wow x 1
  4. :geek:
    Roll center added to post above.

    I'm sure :rolleyes: everyone has enjoyed these physics posts :cautious: :coffee::laugh:
     
    • Agree Agree x 3
  5. LilSki

    LilSki
    Premium Member

    :sleep::coffee::geek::geek::geek::sleep::coffee::coffee::coffee::O_o::D:p:laugh::sleep::sleep::sleep::sleep:
     
    • Haha Haha x 3
    • Agree Agree x 1
  6. jlnprssnr

    jlnprssnr
    Premium Member

    Hmm, really makes you wonder how many mods have realistic physics. Can't imagine most people to go this far into detail! Incredible work
     
    • Like Like x 1
  7. LilSki

    LilSki
    Premium Member

    It would depend if AC fully supports whatever suspension setup a car has. In this case AC does not support this particular suspension so it had to be re-engineered to act the same way.
     
    • Agree Agree x 2
  8. There are a lot of half baked quick mods out there using ripped models and such. When learning some of these things from a physics modding guru, he basically said if they never talked about physics in the WIP thread, then the physics of the car are suspect. Makes sense.

    But also, like LilSki said. I had to go through all this because AC doesn't do Semi Trailing arm--at least at this time.

    And 3rd, if someone is doing a car with a Double Wishbone suspension or Mcpherson suspension like AC supports, but they do not have or cannot obtain the actual suspension measurements, then they will have to go through some of this.

    For example, some people have done racing cars for actual racing teams--these were never released publicly. The team gave the modders almost all the data on the car; almost. But they didn't give them the suspension pickup (mount) points. They did however give them the graphs of the car like you see above. So they had to work backwards from the graphs like I did.
     
    • Like Like x 3
  9. aphidgod

    aphidgod
    Premium Member

    The graphs are by far the more useful data, unless you have data from an engineering document and do not have to go through the process of measuring a real car. Some of these pickup points are sensitive to the tenth of a millimeter and that's a lot to ask.

    Having done this with my WRX, the data i had was invaluable in correcting my measurements to achieve accurate behavior. (I got very close with the measured geometry, but some points were very difficult or impossible to measure - like the inner tie rod location, which is hidden underneath a thick rubber boot and is an extremely sensitive point.)
     
    • Agree Agree x 1
  10. For sure, it's much better to more or less match full travel graphs than to perfectly match a single point (since all of the others could still be incorrect). Onto the second point; pickup points may be sensitive to the tenth of a millimeter, but no car on earth is manufactured to that level of precision. Even most wheelbase measurements are on a +-5 mm basis depending on the individual car.
     
    • Agree Agree x 2
    • Like Like x 1
  11. aphidgod

    aphidgod
    Premium Member

    Absolutely... heck, I gained 0.6" (15mm) in wheelbase by exchanging my front control arms with an 04 STi, which achieves an extra 1.5* of caster solely by locating the ball joint farther forward. That's a 10mm increase in wheelbase per degree of caster on my chassis, which is specced at 3.5* +/- 0.75* in the factory service manual. Techincally that means two 07 WRXs, with wheelbases divergent by up to 15mm, could both be within factory spec. They'd be statistical outliers at the extremes of the range, of course, but still.

    (I know this is all stuff you're well aware of, but for everyone else who's reading...)

    I guess what I'm getting at is that small errors in certain spots can make a big difference in how the car feels, and unless your measurement procedure achieves that kind of accuracy you have to understand that your measurements probably won't get you exactly what you expect (or what you need) when translated to the game. It's very helpful then to have some empirical data to verify your results - individual specimens of that car will differ somewhat, but I feel you're likely to be off by a wider margin with your in-situ measurements than the manufacturer is with their assembly processes. (And thus the chart is more likely to be correct than your geometry data, if you're relying on measurements and find a discrepancy.)

    [edited for grammar, not content.]
     
    Last edited: Jun 25, 2016
    • Agree Agree x 1
  12. Up

    [​IMG]



    [​IMG]
     
    • Like Like x 9
    • Love Love x 4
    • Wow Wow x 1
  13. Love the way you go into the details (on physics and 3d side)!
     
    • Like Like x 1