1. GPU Fatal Error? Intel HD Graphics 620/630 Crashes? Help us solve them!
    Dismiss Notice

Torque vs. Horsepower

Discussion in 'Content Creation' started by CommandoAir, Feb 8, 2018.

  1. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    What's burn efficiency? There isn't anything about that in the spreadsheet, so I left it at default. Also, thanks for the info about the intakes.

    A 4A-GE 16v makes it's peak HP of 115 at 6,600 RPM, and it requires 148Nm of torque at 6,600 RPM to do so. Looking at a couple of dyno graphs, the 4A-GE has a relatively flat torque curve in the middle, which matches my Automation torque curve. I could probably raise the lower end torque to flatten out the curve a bit.

    The only thing that contains maxRPM is the turbos, which I took off, and the max RPM was still too low.

    I'm certain that I don't have any other files with the same name. The engine name is unique.

    I've set the max Torque to 300, and I'm no longer getting the overtorque warning, but I'm still not making the right power. Perhaps the lack of power is due to the rev limiter being where the peak power should be?
     
  2. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    This point is 556.94, which needs to go to spreadsheet:
    upload_2018-2-11_12-58-12.png

    How I know it is 556.94, 50Nm is 36 pixels, that point in torque curve is 5 pixels higher than 550, so I calculate 50 / 36 * 5 + 550 = 556.94
    RPM is 6166 calculated same way counting pixels, hp is 499.155 and spreadsheet gives hp of 462 and torque of 534 (because friction values being what they are) we completely ignore what graph shows and what values those graphed columns give us:
    upload_2018-2-11_13-27-19.png
    (also ignore all that turbo stuff and efficiency, it is just something I'm trying to get to work)

    W12 claimed values 450hp @ 6200 rpm and 580Nm @ 4000-4700rpm, so either there are more versions than few I found or these curves just are what they are.

    That is torque for maximum HP rpm, there is nothing more to it, that is torque I will input and HP will be what it will be. Sometimes, these graphs are not accurate, one attached here is kind of rough representation, which I call sales brochure type.

    HP is never measured, in every torque curve you will see it is calculated, sometimes it is beautified, made to match some spec which it does not match in reality, never trust HP, sure it is correct most of the time, but it just is not any value.

    Why your max rpm is only 6600rpm is bit odd, if you are absolutely sure there is nothing else limiting your rpm, then dynamic resistance being higher than what you used in spreadsheet can be reason so that instead of hitting limiter engine just does not have the power to get to max rpm, but this should be visible in game torque app.


    So what power will be then in game?

    This will give you different peak power, even you have same peak power:
    upload_2018-2-11_13-30-30.png upload_2018-2-11_13-31-42.png

    You might ask why is that? Well, in BeamNG torque is calculated for each rpm, that is 1rpm resolution, so what you give to BeamNG is guide, which game engine will use to calculate real torque curve it uses. So that is why surrounding data points you input have some effect to peak torque/hp.

    Mostly it affects when changes are great enough in torque, place few more rpm bins around peak power and peak torque and change torque so that it does not change too much between bins, but it is good to know that sometimes you might need to adjust surrounding datapoints.

    So I ended up with this torque curve, it does not look at all the same as one at top of this post, but it is most accurate way to get values from torque curve I know of:
    upload_2018-2-11_13-45-34.png

    This is graph with same data, only size of graph was adjusted and unneeded datapoints removed and it is somewhat similar to graph at top of the post, I just usually ignore the graph completely, only thing I use it is to get smooth fall off and rise of torque curve where I don't have data from:
    upload_2018-2-11_13-49-15.png

    So this is what I do next:
    upload_2018-2-11_14-4-30.png

    That give me torque that should be enough close and hp just is what it is, I know my torque is right so that is how much HP BeamNG engine will have, it might be 20hp more, but I trust BeamNG more than those fancy sales brochure torque curves, so when torque input is correct hp will be what it is:
    upload_2018-2-11_14-9-32.png

    Of course I would need to adjust dynamic friction and friction etc. to get good engine and that changes torque values I need to input to jbeam, but that is how I do engines and I think that gives enough good results.
     
    #22 fufsgfen, Feb 11, 2018
    Last edited: Feb 12, 2018
  3. BombBoy4

    BombBoy4
    Expand Collapse

    Joined:
    May 16, 2015
    Messages:
    2,048
    Possible, I thought it was the burn efficiency on extracting torque from the stroke.
     
  4. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    I'm not sure what you're trying to tell me in the first part.

    As I said, I set the dynamicFriction in the spreadsheet to be the same as the jbeam. It's definitely hitting fuel/spark cut.

    As for your point about surrounding data points, this is part of the reason as to why I added 6,600 RPM with peak power, just after 6,500. To test further, though, I set from 6,000 all the way to 7,000 RPM as the peak power, and it still didn't do anything.

    I tried another test, and found something interesting. I set the rev limiter to 10,000, and I set the entire RPM range to 500 Nm of torque, and... Not a single thing changed (And yes, I did swap out the engine and swap it back in again to refresh it). So the game isn't reading the torque figures. But it is definitely reading the Jbeam, because I tried setting the max torque strength to 5, and sure enough, the engine would blow when trying to pull away even in first. Not only that, but the name of the engine is exactly what I set it as in the Jbeam. But it's just not reading the torque or rev limiter parameters.
     
  5. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    First part shows how you get torque values from dyno graph, shape and hp has nothing to do with it really, only torque numbers matter.

    If you don't get changes in game you make in file, then you have some other part that is overriding your rev limiter and torque curve, it really can't be anything else. ctrl-r twice after every change, if still nothing, you will have some other engine in game with exact same name, it can be other mod too.

    Add B or something to engine name and name that UI showsm if that does show up and still torque is not changing, then it is some other part overriding your engine torque settings.

    Also with Notepad++ there is on catch that gets you very easily, you have two of same name files open, they are at different paths, sometime you just edit and nothing changes, sometimes you edit correct file and something changes, check the path of the file you are editing.
     
  6. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    Update:

    I've partially fixed the issues. There was some kind of conflict with one of my other engines for the Covet, but I'm not sure why since it had an entirely different name. The 4A-GE based engine I am making now is my first mod under my new initials, so I've never used them before. The other engine wouldn't work when selected in the selector either, but it would work if I loaded the configuration from the vehicle creator I had already made. Perhaps it has something to do with them both starting as "hatch_engine_"?

    The game now reads the RPM limit and torque values, but I'm still down on power. I'm making about 86 PS (even less than what it was making before the fix). I'm putting in the values from the Jbeam Input section of the spreadsheet still, and I have the Friction and Dynamic Friction variables set the same as in the Jbeam.

    Something doesn't seem right. Using the equation mentioned earlier -
    148 Nm (Peak torque) x 5,800 RPM (Where peak torque is achieved) = 858,400 ÷ 9565 = 89 Kw = 119 HP.
    But the spreadsheet claims that it would result in 102 HP. To make 115 HP @ 6,600 RPM like a real 4A-GE, you need to make 148 Nm at 6,600 RPM.
     
    #24 CommandoAir, Feb 12, 2018
    Last edited: Feb 12, 2018
  7. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    Do you want to send file to me, I could have a look at it too?

    Spreadsheet calculates power same way as BeamNG, but it will not show you power you get in game, because that power is calculated from torque where frictional losses are taken out, but correction has not yet done to make up for losses.

    You need to change formula of spreadsheet a little and also to make graph use new data (torque you write in to spreadsheet) to see HP you would get in game.

    To get HP that you should see in game, you need to calculate hp from gross jbeam torque instead of how spreadsheet currently calculates (that is why say ignore hp, because of hp in spreadsheet has no practical use):
    upload_2018-2-12_22-26-42.png

    That will affect only to how graph in spreadsheet shows curves and HP column though, it has nothing to do with what you get in game.

    Net ingame columns show you values that would be ingame if you would input values you put in Gross jbeam torque to your engine jbeam file, but as you should put jbeam input values to your engine jbeam (like you do) then those Net ingame columns have no meaning, just ignore them or change formulas so that they show meaningful values.


    In game dynamometer app shows wheel power, while torque curve app shows flywheel power, that is what I think, can't be sure, but there is that difference in apps too.

    I did change formulas so that HP that is shown is now ingame HP and I used values you posted earlier in screenshot of your spreadsheet, this is what hp game should have if you use those values in Jbeam Input column, see how much hp there really is :
    upload_2018-2-12_22-54-21.png

    148Nm @ 5800rpm would give 120hp, which perhaps is within tolerances. But I don't get it why would you get only 86PS in game, I would like to test what my game shows up, so we can see if it is yet another conflict or if it is some other issue that can be spotted.
     
  8. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    File is attached.

    That explains some. I thought the sheet showed HP after losses. Wouldn't it be as simple as making another column that uses the Jbeam input torque for the formula rather than the Gross torque?

    Also, there's a dynamometer and a torque curve app? I thought there was just the engine debug app that displayed PS, torque etc. and the graph app that would constantly update, showing HP, Torque and RPM.

    When you say "148Nm @ 5800rpm would give 120hp", is that at the wheels, after losses?

    Thanks for all the help by the way.
     

    Attached Files:

  9. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    New column with adjusted formula should do it fine, then you would need to make a new graph or modify old graph so that it shows correct data.

    If you check ingame apps, there is torque curve app and AFAIK it is only one that shows flywheel power and torque, that is what should match your inputs of spreadsheet.

    120hp figure is flywheel power.

    This is how your file looks in my game, in torque curve app, which I resized from corner (little arrow) to be more of the size what those dyno curves usually are, it has 136hp. Spreadsheet shows with 148Nm input at 6600rpm: 137hp as I have changed the formula to show flywheel power ingame (change I described earlier) so I think it is decently close enough which spreadsheet calculates and what you get in game.
    upload_2018-2-13_0-36-54.png

    Your inputs are bit high though, so just input values in torque curve without changing them at all and Jbeam input column values should give you exactly correct torque curve with decently close enough hp readings as mentioned previously.

    I think that greatest confusion you had was with looking those hp numbers in spreadsheet and using other apps than torque curve app to check ingame power.
     
  10. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    So the torque curve app shows wheel power, then?

    I've tweaked some stuff on the spreadsheet. From a bit more research, it appears that your standard AE86 will make around 80-90 HP at the wheels, and the 4A-GE will make 115-128 HP at the crank.

    If I understand the spreadsheet correctly, the G column (Net (Ingame) Power) is HP at the wheels, correct?
    And if I were to take the Gross (Jbeam) Torque, multiply that by RPMs, and put that through the Torque to Power Constant in the top left, that would give me crank horsepower, correct?

    If that is all correct, then I have tweaked the figures as so:
    • Increased Friction to 12
    • Increased Dynamic Friction to 0.04
    • Tweaked torque curve to make 128 HP at the crank, and 92 at the wheels.
    The torque curve app shows about 92 HP, so that must be power at the wheels. But what does the engine debug app show? Because that is displaying about 53 PS peak at around 4,500 RPM, which is below both the peak torque and peak power. The car does also feel a little sluggish. I know 90 WHP isn't much, but the car only weighs about 990 KG.
     
  11. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    No, Torque curve app shows power from flywheel.

    G column (net ingame power) is at the flywheel IF you use your gross Nm values as jbeam input, I'm telling you just ignore those two net ingame power columns completely, you can even hide them as they have no meaning really, it is just what you would get ingame as flywheel power without using the spreadsheet.

    Idea of the spreadheet is that you don't need to tweak anything, you input values from dyno graph to spreadsheet and to get those values in game you use jbeam input values for jbeam and no tweaking is needed, it will be correct.

    In game torque curve app numbers will then match to what Nm you did input to spreadsheet, but never ever tweak that makes use of spreadsheet quite meaningless as purpose of whole spreadsheet is to save trouble of tweaking.

    Anyway I made new tool for you, try that as it is bit simpler and I put also export function to it (inspired by atv_123 and based on 440cid's spreadsheet).

    Update: I'm sorry, there was an error in spreadsheet I attached, export did show data from wrong column, so I fixed error and made small updates to spreadsheet, delete old one.

    I did try to make it more simple to use, export works so that you select values from export page, then copy and paste to your jbeam. There should be less confusion with this one:
    upload_2018-2-13_14-53-33.png upload_2018-2-13_17-46-36.png
     

    Attached Files:

    #29 fufsgfen, Feb 13, 2018
    Last edited: Feb 13, 2018
  12. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    Surely it can't be crank HP. It only hits about 111 Nm torque on that, rather than the 148 Nm I set. It also shows about 90 HP, which is what I set the wheel HP to be. And what does the engine debug app show?

    I thought the spreadsheet was just so that you didn't have to calculate for drive train losses. Either way, I think this method is more suitable in this case, since I know both crank and wheel HP, and I've tried to hit both. When I said tweak, I meant I was tweaking the friction and torque curve in order to achieve both crank and wheel HP.
     
  13. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    I haven't really ever needed those other apps, so I had to check what they show, this is my guesses of what those are:
    upload_2018-2-13_23-17-17.png
    I thought dynamometer app would of shown wheel power, must be my memory playing tricks on me.

    Torque curve app is only one I use and when I input real data from dyno graph to spreadsheet I get same numbers showing in torque curve app, that is purpose of the spreadsheet.

    Whatever those other apps show, I would say ignore them.

    Now when you have set crank power, so that torque curve app's curves show correct numbers, after that you go to powertrain differential and gearbox friction values are ones that you need to tune to get wheel power right, but that is only after you have set engine torque correctly so that you see correct flywheel power in torque curve app.

    That is because you have to have source correct before you can tune result. Then, but only then, you can use values in engine debug app and you will adjust only friction of differential and gearbox, touching engine torque at that point will cause whole process to be started all over again.
     
  14. Diamondback

    Diamondback
    Expand Collapse
    Moderator
    Staff Member

    Joined:
    Apr 8, 2014
    Messages:
    1,318
    Hm seems like I need to clean up a bit of this app mess :D
     
  15. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    If you want, I let you use my arrows, they are really useful :D
     
  16. Diamondback

    Diamondback
    Expand Collapse
    Moderator
    Staff Member

    Joined:
    Apr 8, 2014
    Messages:
    1,318
    Haha thanks, I think I'll just need to fully redo and adjust what they actually display. And make it more clear.
     
  17. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    I got actually now same issue, I have engine that has unique name as well as unique UI name, but while it displays torque curve from engine file I made, it won't update no matter what I do.

    Reason for this was that I first accidentally saved my engine file with extension .txt, then I changed extension in folder, but I had .txt file open in Notepad++ so when I edited engine file in Notepad++ and hit save, it created file with .txt extension again. Removing that .txt file and opening correct .jbeam file in Notepad++ fixed the issue.

    Did you get your engine adjusted like you wanted at the end?
     
  18. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    I don't think that was the issue I was having, as when I fixed mine, I didn't touch the files. I just disabled my other engine mod.

    I'm pretty sure the engine is performing as intended, as the torque curve app shows what I worked out to be the power at the wheels in the spreadsheet. I was just confused as to what the engine debug app was displaying, since it showed around 50 HP, which didn't match with either the crank or wheel horsepower.
     
  19. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    7,078
    That is great that you got it to work as you wanted!

    Did you notice this:
    https://www.beamng.com/threads/development-screenshots-2.24313/page-2#post-799723

    There is going to improvements to these apps in future updates so hopefully less confusion :)
     
  20. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    112
    I didn't see that specific post, but I saw he/she mentioned it earlier in this thread. It's great that they're improving it, should clear up a whole bunch of confusion.
     
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice