Page 1 of 2 12 LastLast
Results 1 to 10 of 15

Thread: Suspected Bug - Altitude impact on Horsepower

  1. #1
    GT5 Pilot F1_Racer68's Avatar
    Join Date
    Aug 2012
    Location
    Canadian Tire Motorsports Park (Mosport)
    Posts
    1,905
    Platform
    PC

    Suspected Bug - Altitude impact on Horsepower

    In the course of running our league, I have received a lot of complaints about certain cars being well off the pace of others. Of course, it is always pointed out that the cars that seem off pace are Normally Aspirated and that they are affected by the altitude, but the complaints persisted and seemed to carry some weight. So I did some investigation myself and found something quite concerning.......

    In order to keep things easy on myself I performed the test using the 5 GTE class cars (these are the ones the concern was raised over), and also performed the test at 2 super speedway ovals (easier to watch on screen telemetry there). The tracks chosen were:



    The general rule of thumb that I learned for horsepower loss from altitude is as follows:
    • NA cars - 3% per 1000 feet of elevation
    • Turbo cars - 1.8% - 2% per 1000 feet of elevation


    The mathematical formula that I have been taught to calculate this is:
    • HP = (Elevation x 0.03 x HP@sea level)/1000 (replace 0.03 with 0.018 - 0.02 for turbo cars)


    As you can see in the screen capture below, the ACTUAL horsepower loss by the NA cars is significantly higher than either the mathematical formula or even the general rule of thumb estimate.

    Indianapolis Motor Speedway is less than 1000 feet above sea level, so the calculated loss of HP should have been ~12 HP. Even using just the general rule of thumb of 3% per 1000 ft would give a rough loss of 15 HP. In the game however, the Corvette lost 40 HP and the Aston lost 38 HP. Also, the Ford and the BMW both lost less than 1.8% (but an acceptable loss), while the Ferrari lost 0 HP and actually GAINED some torque.

    Click image for larger version. 

Name:	2017-11-07_9-16-30.jpg 
Views:	210 
Size:	188.2 KB 
ID:	244692

    I made every effort to keep the test as consistent as possible, so both tracks had the same date (May 24th, 2017) and time (14:00), and all cars where tested with their SMS Default "Loose" setup.

    HP and Torque numbers are the peak numbers observed using the on screen telemetry HUD.

    I am fully aware the temperatures and humidity levels also play a factor, however there is no way that the Corvette and Aston should have lost 8% Horsepower over an elevation increase of a mere 706 feet.

    So far my testing has been limited to the GTE class, but I suspect this issue exists across all classes. Could someone please tell me why this happened? Did I look at something the wrong way? Did I use an incorrect formula? Or is there truly an issue with the in game calculation?

    I thank you in advance for your insights and for looking at this further.
    CanAm Virtual Racing League - League President Twitter: @canamvrl @F1_Racer68
    System Specs
    CPU - AMD FX-8370 w/Wraith Cooler MoBo - GB-78LMT-USB3 (Rev 6.0) RAM - ADATA 12GB 1600 DDR3, Video - Zotac GTX1060 AMP! 6GB DDR5, OS - Win10 Pro 64bit, SSD - Corsair 240GB Force LE, Controller - Fanatec CSW v2.5, CSL Elite LC Pedals, DIY Button Box, G27 Shifter (USB) Monitor - Samsung UN28H4000AF 28" LED x3

  2. #2
    GT5 Pilot F1_Racer68's Avatar
    Join Date
    Aug 2012
    Location
    Canadian Tire Motorsports Park (Mosport)
    Posts
    1,905
    Platform
    PC
    So, I have now gone through and done the same tests using our other classes for this season (LMP2 and GT3). The Normally Aspirated cars are all losing just over 9% horsepower over less than a 1000ft elevation change.

    Again, from my admittedly limited knowledge, this seems excessive. It makes it very difficult from a league management perspective, because we end up with everyone wanting to be in the same small number of turbo cars in order to not be handicapped.

    With the above data, I have 2 requests for anyone from SMS that can answer them:

    1. Is there a glaring flaw in my calculations or expectations?
      1. If my understanding of the horsepower loss calculation is flawed, then I am very happy to be corrected.

    2. Is it possible to provide the tools whereby league leadership can implement BoP on the Turbo cars?
      1. Allow us to have a way to police either a Turbo Boost pressure limit or Air Intake restrictor limit
      2. We would need a way to be able to "see" what the boost pressure or air intake openings are so that we can ensure rules are followed.


    I look forward to being shown the light in regards to the horsepower drops (or confirmation that my knowledge is correct?).
    CanAm Virtual Racing League - League President Twitter: @canamvrl @F1_Racer68
    System Specs
    CPU - AMD FX-8370 w/Wraith Cooler MoBo - GB-78LMT-USB3 (Rev 6.0) RAM - ADATA 12GB 1600 DDR3, Video - Zotac GTX1060 AMP! 6GB DDR5, OS - Win10 Pro 64bit, SSD - Corsair 240GB Force LE, Controller - Fanatec CSW v2.5, CSL Elite LC Pedals, DIY Button Box, G27 Shifter (USB) Monitor - Samsung UN28H4000AF 28" LED x3
    The following 4 users likes this Post: AbeWoz, ermo, Fight-Test, wesker6664


  3. #3
    Moderator Mahjik's Avatar
    Join Date
    Oct 2011
    Location
    Kansas City, MO
    Posts
    5,843
    Platform
    PC
    We have noted it for SMS. Whether or not they will comment though is out of our hands.
    Intel i7-4790K @4.4GHz | 16gb ram | 2x EVGA GTX 970 | 5040x1050 | Windows 8.1 64-bit | Thrustmaster TX Servo Base | Thrustmaster TH8A Shifter
    ProtoSimTech PT-1 Pedals | rSeat Evo V2 | 2x Buttkickers | SimVibe | Black Max II Button Controller | Sennheiser PC 360 Headset


    Forum Rules
    PC2 FAQ & Tips
    The following 5 users likes this Post: cluck, ermo, F1_Racer68, GenBrien, Konan


  4. #4
    GT5 Pilot F1_Racer68's Avatar
    Join Date
    Aug 2012
    Location
    Canadian Tire Motorsports Park (Mosport)
    Posts
    1,905
    Platform
    PC
    Quote Originally Posted by Mahjik View Post
    We have noted it for SMS. Whether or not they will comment though is out of our hands.
    Thanks Mahjik. I understand.

    I do have to say that, generally speaking, the cars all seem relatively well BoP'd to each other (in GTE class at least) for most tracks. In GT3 the 488 seems to be a clear run away for us, but that COULD also be down to the driver (he's got alien potential). Where the concern comes from is as we get to tracks at much higher elevations (Bathurst, Spa, etc.), with the level of competition in our league, it APPEARS to us that it starts to become necessary to be able to do a "per track" BoP.

    Unfortunately, while we can mandate certain rules verbally/in writing, there is no way in game to verify or police if the drivers actually abide by it. If there is a way for this to be added in the future, it would be very much appreciated.

    While we don't race for prizes (other than bragging rights), we do strive for maximum authenticity and maximum "fairness" in the racing. Having cars overly hampered by an excessive loss of horsepower (if that is in fact the case) is an immediate handicap to those drivers.

    Again, I do want to fully stress the fact that I am only raising a question based on our findings, and not making any claims of an actual "bug" or "flaw".
    CanAm Virtual Racing League - League President Twitter: @canamvrl @F1_Racer68
    System Specs
    CPU - AMD FX-8370 w/Wraith Cooler MoBo - GB-78LMT-USB3 (Rev 6.0) RAM - ADATA 12GB 1600 DDR3, Video - Zotac GTX1060 AMP! 6GB DDR5, OS - Win10 Pro 64bit, SSD - Corsair 240GB Force LE, Controller - Fanatec CSW v2.5, CSL Elite LC Pedals, DIY Button Box, G27 Shifter (USB) Monitor - Samsung UN28H4000AF 28" LED x3
    The following 3 users likes this Post: ermo, GenBrien, Konan


  5. #5
    WMD Member Franco Ferrari's Avatar
    Join Date
    May 2012
    Location
    Piacenza, Italy
    Posts
    310
    Platform
    PC
    Kudos for the data collecting so far.

    You can check the same figures at Mount Panorama: with an elevation of 650m, or about 2100ft, its about exactly 3 times the Indianapolis elevation.
    If the power loss is linear with the altitude, you should get 3 times the power loss of Indy.


    EDIT: Anyway, you should keep in mind that in your measurements the ambient temp at Indy was 15°F lower than the one at Daytona. Lower temp means lower pressure, even at the same altitude.
    Last edited by Franco Ferrari; 14-11-2017 at 15:55.
    "I was playing pCARS on PC before it was mainstream" - cit.

    MB: Asus Strix Z270F Gaming
    CPU: i5-7600K @4.63GHz
    RAM: 16GB (2x 8GB) Corsair Vengeance LED 2666mhz
    GPU: Asus Strix GTX 1080 Gaming A8G
    HD: Corsair MP500 240Gb M.2 PCIe + Crucial MX300 1Tb M.2 SATA
    Audio: Onboard Realtech sound chip + Sennheiser headphones / stereo speakers
    Wheel: Thrustmaster T500 RS + Fanatec Clubsport Pedals V3 + Fanatec Clubsport Shifter SQ + Playseat racing seat
    Monitor: Asus PA 238Q - 23" 1920x1080 P-IPS
    Case & PSU: NZXT H440 V.2, Thermaltake Water 3.0 Ultimate, Corsair RM850i
    Operating System: Windows 10 Pro N, nVidia WHQL Drivers 388.31

  6. #6
    GT5 Pilot F1_Racer68's Avatar
    Join Date
    Aug 2012
    Location
    Canadian Tire Motorsports Park (Mosport)
    Posts
    1,905
    Platform
    PC
    Quote Originally Posted by Franco Ferrari View Post
    Kudos for the data collecting so far.

    You can check the same figures at Mount Panorama: with an elevation of 650m, or about 2100ft, its about exactly 3 times the Indianapolis elevation.
    If the power loss is linear with the altitude, you should get 3 times the power loss of Indy.


    EDIT: Anyway, you should keep in mind that in your measurements the ambient temp at Indy was 15°F lower than the one at Daytona. Lower temp means lower pressure, even at the same altitude.
    I am fully aware that both temperature and humidity play a factor. Unfortunately, those are both factors that are difficult (if not impossible) for us to control in game and therefore difficult to ensure they are exactly the same in all tests. Also, I have not yet found a formula for real world calculations that incorporate the impact of humidity and temperature. Even so, I suspect the total impact will still be lower than 8 - 9% per 1000ft of elevation.

    I know SMS will probably not want to share their math for how the game calculates it, and I fully understand that. I was hoping this thread might start a (civilized) discussion about what would be the proper and complete method for calculating this in the real world, and how we should apply that in game.

    I am planning to run the test at Bathurst as well to see if the increased elevation continues to drop power at the same rate, or if there is some threshold in the game. In theory, based on my results, the Corvette C7.R should lose 80 - 100 HP at Bathurst vs. Daytona, dropping it to somewhere around 425HP at Bathurst. Considering the Ferrari, BMW and Ford GT will all remain relatively close to their base HP, that would be a significant handicap to the Corvette. The Aston would be in the same situation as the Corvette.
    CanAm Virtual Racing League - League President Twitter: @canamvrl @F1_Racer68
    System Specs
    CPU - AMD FX-8370 w/Wraith Cooler MoBo - GB-78LMT-USB3 (Rev 6.0) RAM - ADATA 12GB 1600 DDR3, Video - Zotac GTX1060 AMP! 6GB DDR5, OS - Win10 Pro 64bit, SSD - Corsair 240GB Force LE, Controller - Fanatec CSW v2.5, CSL Elite LC Pedals, DIY Button Box, G27 Shifter (USB) Monitor - Samsung UN28H4000AF 28" LED x3
    The following user likes this Post: Reiche


  7. #7
    WMD Member RomKnight's Avatar
    Join Date
    Oct 2011
    Location
    Lisbon
    Posts
    2,297
    Platform
    PC
    But surely there's a "margin of error"?
    What is (not) acceptable?
    ---
    My feedback is always on default setup unless stated otherwise. Auto-Clutch is OFF
    Still hoping for Estoril one day

    | CPU: i5 4670K@4.2GHz | GPU: MSI GTX1080 Gaming X 8GB (387.92) | Mobo: AsRock Z87 Extreme6
    | RAM: 4x4GB G.Skill TridentX 2400Mhz CL10 | PSU: Corsair TX850M | OS: Win10 x64 | Sound: Onboard Realtek ACL1150 Asus Xonar DX | C:\>: Samsung 840 EVO 250GB
    [B]| Monitor:[/B] 3x IPS LG23EA63 @5760x1080 (Bezel Corrected) -- HTC VIVE

    | Wheel: CSWv2 + CSPv3(usb) + CSS SQv1.5 - (driver v289 - fw 299/130) | Damping: 100 | 900š
    | In game: Deadzones: 0 |Sensitivity: 50 | Damping: 0
    | Wheel Profile: Sens: Aut | FF: 100 | SHo: 0 | ABS: 0 | Dri: OFF | For: 100 | Spr: 100 | Dpr: 100

    | FFB: Custom |Gain: 100 | Volume: 50 | Tone: 50 | FX: 50 |Frequency: 600

    | Visual FX: Default except Interion Flare: Off
    | Performance - Aniso: 16x | MSAA: Medium| SS: 1.4| Vsync: Off | Grass: Low | Envmap/Reflections/Shadows: Low/medium/low| Car/Track: High | Blur: Off | E. Mirrors: Yes | Particles: High/High
    | Driving assists: ABS/TC, maybe - SC always OFF
    | CmdLine: -ffb600 -pthreads 4 -skipcrowds

  8. #8
    WMD Member ermo's Avatar
    Join Date
    Oct 2011
    Location
    Aarhus.dk
    Posts
    400
    Platform
    PC
    @F1_Racer68:

    I like this thread.

    I'm personally guilty of having touted the altitude (and ambient temp) corrected torque numbers*, but I haven't actually done the math and checked the calculations myself.

    *: Assuming the inlet restrictor size, engine bore+stroke and atmospheric conditions are all kept constant between two altitudes, torque is mostly a function of peak combustion chamber pressure (I don't think this is simulated -- just a look-up table with torque as a function of RPM which is then scaled to atmospheric conditions) and as you probably know, power is merely a function of torque and RPM. From the numbers you've recorded, I'd say the turbo torque numbers look to be within margin of error and should be considered constant between the two altitudes (perhaps with slightly more lag as the turbo needs to spin a little bit faster to generate the extra boost dP = wastegate absolute pressure - ambient pressure).
    A.K.A. ermo | CPU i7 3770K@4 GHz | RAM 32GB DDR3-2400 | GFX Vega 64 AC @ 1280x1024\1920x1080/1280x1024 | Controller G27 w/NIXIM mod | OS Win10 Pro x64 | UTF-8 characters: ✓ ✔ ✗ ✘ Đ ™ Ū ° ← ↑ → ↓

    "We respond to well constructed criticism, good ideas which are properly described and thought through, and ideas that will benefit the whole community, not a vocal minority.

    Do not repeat requests, be patient and polite, and have some confidence in our ongoing desire to improve the game for all users, which we have clearly and ably demonstrated in the past." -- Richard Steer, Slightly Mad Studios
    The following 2 users likes this Post: F1_Racer68, Reiche


  9. #9
    Superkart Pilot
    Join Date
    May 2015
    Location
    Deep South
    Posts
    705
    Platform
    PC
    F1, awesome info. I love reading about these things. One other thing to bring into the equation is at altitude the air is thinner which makes it much harder to cool the turbo (all of the car but the turbo is what is relevant to the conversation). Thinner air also means the turbo is spinning faster. The number one concern this brings is reliability for the turbo cars at altitude if they don't monitor the heat. Right now we have no higher level failure for the turbo cars using full potential of the turbo at altitude. No risk, all Reward.
    Last edited by Fight-Test; 14-11-2017 at 20:43.

  10. #10
    Kart Driver
    Join Date
    May 2015
    Posts
    79
    Platform
    PC
    Depends what altitude they're using for the tracks, the loss at Indy seems excessive. FWIW I'm seeing ~80BHP loss for the Vette at Bathurst along Skyline which according to the website is 862M (~2825ft) above sea level and 63BHP loss along the pit straight which is 688M (~2250 ft), which would put the loss at ~5.5%/1000ft which is at least reasonably consistent.
    The following user likes this Post: F1_Racer68


Similar Threads

  1. Altitude of each tracks ?
    By ApexSight in forum General Discussion
    Replies: 20
    Last Post: 03-10-2017, 17:53
  2. Question - horsepower in VR and mirrors
    By Javaniceday in forum General Discussion
    Replies: 2
    Last Post: 30-05-2016, 13:20
  3. Impact glitching
    By leithnow in forum General Discussion
    Replies: 7
    Last Post: 15-09-2015, 03:26
  4. Impact calculations, how do they work?
    By Mattias in forum Multiplayer
    Replies: 5
    Last Post: 22-07-2015, 07:32
  5. Horsepower RUF RGT-8 GT3
    By 987Paule in forum PS4 - Technical Help & Support
    Replies: 3
    Last Post: 17-06-2015, 13:57

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •