1. Trouble with the game?
    Try the troubleshooter!

    Dismiss Notice
  2. Issues with the game?
    Check the Known Issues list before reporting!

    Dismiss Notice
  3. Before reporting issues or bugs, please check the up-to-date Bug Reporting Thread for the current version.
    0.30 Bug Reporting thread
    Solutions and more information may already be available.

Maluch JBeam rebuild

Discussion in 'Troubleshooting: Bugs, Questions and Support' started by AdamPLGames, May 17, 2021.

  1. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    I always thought it was just an issue with extremely slow acceleration off the line rather than brake delay. I don't think it's a problem in the new version, as I couldn't even reuse any of the old brake values because I had to separate the brakes from the wheels to implement universal wheels support, but actually brake delay like this is a problem irl, just not so big, and for the normal brakes instead of parking brake. It takes like a second to start and stop braking for some reason. Idk if it's a problem related to age or if they had it from production already, so I haven't implemented that yet, if someone knows something about this then please tell me.
     
    • Like Like x 1
  2. SKB

    SKB
    Expand Collapse

    Joined:
    Apr 22, 2017
    Messages:
    1,964
    i thought it was to simulate idk, jammed up drum brakes, as they usually just stick sometimes.
     
  3. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    I doubt the intention of the original mod was to simulate anything, seeing how it was made without any research or attention to detail. Probably the original author did that by accident and had no idea how to fix it so just left it in the mod. The original Jbeam was a mess, but I really respect the fact that it was done completely from scratch instead of being based on a Covet for example, the Piccolina is the only suitable candidate to base the Jbeam on (but still needs a ton if changes as this thread shows) and back then it didn't exist.
     
    • Agree Agree x 1
  4. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    Dashboard and gauges. I had to manually move and rescale all the needles, it took a lot of time. FL and EL use the same dashboard but with different gauges, and there are 3 selectable tachometers now: with 5000RPM limit (for stock version), 6000RPM limit (TAN version, which I assume is supposed to have the Z-1/30 tuning kit) and unlimited (for Abarth rally version).
    dwawda.png
    Also, there's an interior color slot now instead of having separate parts with 1 changed texture for TAN body, dashboard, doors, seats, etc. That makes stuff a bit cleaner. It also means you can have this interior color on newer trims.
    ddawad.png

    Edit: Added all the vanilla and custom steering wheels
    wdsaad.png
     
    #244 Agent_Y, Aug 2, 2021
    Last edited: Aug 2, 2021
    • Like Like x 7
  5. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    Having problems with weight again... 60kg is way too much for the interior as it turns out.
    Front seats I found out that they weight around 10kg each with headrests, so without them standard ones could be 9kg. Seems reasonable compared to other vanilla cars, a bit heavier maybe but they have steel bases so makes sense. Rear bench seat could be 9kg as well, the Covet one is 7.4kg but I've seen many people complain that the Maluch one is heavy so... makes sense? The interior mirror could weight the same as an exterior one at most. All pedals, levers, etc. are not heavy enough to make a difference, all together could be 1kg at most. (Keep in mind that the inner sides of them count as exterior, for example if you remove clutch pedal the clutch itself stays, when you remove the brake pedal and lever all the braking mechanism stays, etc.) Now the dashboard along with everything on it... I assumed it would be the heaviest thing, but no, it turns out that even on the D-Series the whole thing weights only 3kg, including the radio, steering wheel and all, and it's twice as big, this doesn't really make sense to me... I made mine 3.5kg so it makes a bit more sense. Door panels are quite heavy as it turns out, 4kg each on the Covet, the dashboard being lighter than them makes no sense to me. All of this added to the base weight of the car gives only 582kg, still about 20kg has to go somewhere and I have no idea where, any help?? I really suspect the dashboard because on the Vivace it's as much as 60kg, so 20 times more than the D-Series, but the D-series got remastered after the Vivace came out so I thought it would have more accurate weight...
    Also kinda important info I found is that it's impossible to strip the car down to 550kg while keeping it street legal, many people attempted it back in the day as a loophole to be able to drive it below 18 years old but nobody succeeded. Even with a fiberglass decklid, no bumpers and 1 seat it's over 550.
     
    #245 Agent_Y, Aug 2, 2021
    Last edited: Aug 2, 2021
    • Like Like x 3
  6. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    Id also expect the Dashboard to be a bit heavier. Although the Maluch’s dashboard is really, basic 3.5 kg is simply wait too light.

    Interesting info about the loophole, don’t know why nobody tried the Fiat 500, they were sold in the eastern bloc (At least in Czechoslovakia they were) and were much lighter.

    Those 20 kg will need to be put somewhere, but where? Maybe into the body to balance the weight between front and rear.

    According to the original Fiat Service Manual it should be 230 kg front and 350 kg rear for the original 600 (manufacture states that it weighs 580 kg).
     
    • Like Like x 1
  7. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    580 is the weight of the 600cc version I think, because that engine is way lighter for some reason. I saw a weight comparison between 600 and 650 versions, the only difference was the engine. I'll check the weight distribution, this is good to know and I should get it to be accurate. I can balance it with the weights of the front panel and rear fenders + engine mount bar, these are the parts I made from scratch so they are the most likely to be inaccurate in terms of weight. As for the dashboard, I made a quick test and increased its weight so that the whole thing reaches 600kg, and it turned out to behave way more realistically, more similar to vanilla cars. I suspect that since most vanilla vehicles don't have removable dashboards (it's probably easier to model them like that), the weight of the dashboard itself is treated as one combined with the firewall and stuff around it, so maybe the overal weight in that area is accurate. The Vivace might have more accurate weight distribution there because it also features a lightweight interior variant so the difference in weight between that and normal interior must be realistic.
    Also that loophole was after communism, in more recent times, when the Fiat 500 was beyond discontinued already lol.
     
    • Like Like x 3
  8. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    I see, I can remember 500 being used in regular traffic around here till about late-00s, typically driven by some old ladies who bought them new for their high price (during the early 70s Fiat 500 cost 59 000 Czechoslovak crowns compared to the 42 000 for the Maluch), for such an amount of money for such a small car I’d drive it for so long too.

    About the weight, I’d keep it the way it matches the weight balance and behaves realistic, 20 kg is maybe too much for the dashboard, but it must be fine-tuned in some way and if that means that the dashboard would be a bit too heavy, I don’t think it would be such a problem.
     
    • Like Like x 1
  9. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    I know it's too much for a car of this size, as I said, I'll either increase the weight of the front or the rear, depending on which one is wrong. But remember that the weight is counting the steering wheel with the steering column and that can be heavy probably, would 15kg be good in this case or do I have to go down to 10?
     
    • Like Like x 2
  10. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    With the steering wheel and steering column included I’d say 15 kg is about right.
     
    • Like Like x 2
  11. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    Perfectly ballanced, as all things should be.
    wdwddas.png
    After the last weight fix (I really hope it will be the last one in the mod), the dashboard weights 14kg, the whole 600cc version exactly 580 and 650cc exactly 600. Weight distribution of 600cc version is 230kg F and 350kg R as you can calculate from the data on the image. The issue was with the rear fenders, they were too light, missing 5kg added there and all weight redistributed itself properly. But it took me quite some time where the problem was lol, of course it had to be the largest area I made from scratch. Also, it turns out that the fully stripped chassis weights closer to 350kg than 300, so... my first estimation was actually correct, and 250kg would have been way too light. That's the weight of a fully stripped Piccolina, which with added fenders is similar to the Maluch chassis, and also to a fully stripped Covet with no fenders, no fascias, no bumper support, etc. So without the stuff that is integrated in the Maluch.

    I know this chassis weight may not seem that reasonable, but think about it: all the body panels I have gotten accurate compared to official cars, engine and transmission weight accurate to real life, and the weight distribution was still off, as well as the whole weight was not big enough. It could only mean that something in the back of the chassis is "disproportiontely" heavy compared to other cars, that must be the rear fenders. They are quite bulky, way wider than in other vanilla cars proportionately, carry LeGran sized taillights, have wheel wells integrated in them, and have to carry all of the engine bay accessories, we can say this weight also adds to them because this model doesn't have any visual engine bay stuff, additionally the engine cooling pipes are inside of them and they are probably also quite heavy... So they add quite a lot of weight to the chassis and it's explainable, can this all make the car chassis 50kg heavier than it "should be"? Maybe not, but this extra 10kg weight had to go somewhere in the rear, so I'll just leave it as it is now.

    Yesterday I forgot to metion what exactly I did in the interior, so I'll say it now. In the original mod, the seats float above the ground, I added race seat bases below them and they fit pretty well:
    dsasds.png
    They are the standard bases that race seats have, so you can normally put race seats on them. The seats are also affected by the interior skin. I'll add optional headrests later.
    dasdasa.png
    Seats can fall off like before (race ones too), but it takes a much higher force, because previously it was unrealistically easy. The standard seats will deform a lot before they fall off, but it's realistic deformation - all vanilla cars have the same physics for seats, and these too. Vanilla ones also get janky when you try to rip them off like this.
    wdwdwasdasd.png
    Last thing, the whole interior Jbeam is set up in a way that I can put crash test dummies in it later and they will collide with stuff. Steering wheel collision will be added with the dummy driver as without it it doesn't serve any purpose if the wheel itself is just a prop.
    wddsadw.png wdwdwdwdwd.png
    What I have to do now is testing how the complete base version of the car crashes, and comparing it to real life version. I need as much reference material as possible, preferably crash test videos or pictures with descriptions of what kind of test it was and at what speed. Remember that I have not touched the deformation values at all since I made the old version with wrong weight, so a lot might be off right now, some parts may deform too much and others not enough. The more reference material you send, the more accurate the final mod will be.
     
    • Like Like x 7
  12. Doge_guy

    Doge_guy
    Expand Collapse

    Joined:
    Jul 13, 2021
    Messages:
    196
    Make PBR Material Compatible, if you can, THanks!
     
  13. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    Weirdly all the Fiat 126 crash tests appear to be done at 40 km/h.

    This is an inserting crash test, mainly because the front axle is different from the rear one. The front one uses newer hubs whilst the rear on still uses the old hubs.









    With this crash test conducted by Auto, Moto und Sport we at least know that the static deformation of the car was 297 mm.

    Originally, I expected results similar to the Fiat 500, but the front end of the 126 seems to be much stronger. Especially the A pillars and roof which is a point were basically every 500 folded.



    Sadly I was unable to find any crash tests from the rear although I’d expect that the results wouldn’t be that bad considering how much steel is there.
     

    Attached Files:

    • Maluchcrash1.png
    • Maluchcrash2.png
    • Like Like x 2
  14. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    This will really help, thank you! I assume the 40 km/h speed is because it struggles to reach any higher lol.
    About the weird axle difference, it wasn't actually that uncommon to see a Maluch like this. It was called the "transitionary model" and it was a real option to buy in Poland from 1984 to 1987. Basically due to communism being weird, most factories weren't informed about the planned facelift on time, and had already produced tons of old parts when they recieved the orders to start selling the new ones. They had to use the piles of leftovers for something, so FSM (the company responsible for production at the time) decided to introduce them in transitionary models. They would slap whatever parts they had onto it, sometimes the result had different front and rear axles, but there was some standarization about available versions, which was also a mess: Base transitionary model had the old carburetor that officially had been deprecated, old dashboard with old gauges and P4 bumpers (either Polish or Italian version of them depending on factory, sometimes mixed, sometimes missing rear lights), and richer trim had new carburetor, facelifted dashboard and gauges, and FL bumpers. As for the interior, it was not standarized at all, they would throw whatever parts they had in it, sometimes even seats with mismatched colors lol. I want a config like this in the mod.

    Also, for crash testing I need dummies, so I added some things:
    dwwassdads.png
    dasadsadsasd.png
    sadasas.png
     
    • Like Like x 2
    • Agree Agree x 1
  15. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    Only thing similar to that would be the transitional era with the Škoda 100/110 when they switched from a white with a reddish tint to one with a bluish tint. Some Škodas from that era had sometimes mismatched panels in the two colours but I’ve never heard that they would actually mix parts like that. Although it must be said that there were tons of parts made long after the production of the original car ended.

    Can't wait to see how the Maluch behaves fully loaded, the manufactures states that fully loaded there are 360 kg front and 540 kg rear and that just way too big of a difference to work well. Fully loaded driving to the sea there must’ve been well over 600 kg on the rear axle, those rear tires were probably going through hell at that moment.
     
  16. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    This is why I need good suspension, to simulate stuff like this. People would commonly overload this car with cargo space 100% filled, + piles of stuff on the roof or/and many people inside. Maybe when I manage to make it behave realistically at the limit, the normal behavior will become more realistic too, I have to try.
     
    • Like Like x 2
  17. EozCompanyCZ

    EozCompanyCZ
    Expand Collapse

    Joined:
    Dec 11, 2016
    Messages:
    196
    We’ll see, this is one advantage of the rear engine set up, when the car is fully loaded (or overloaded), under breaking both front and rear brakes are under load evenly whilst on a car with a heavy front end bias the front brakes are much more stressed.

    Something that was again typical for the Škoda 105/120 and even more for the later models with four piston brake callipers (Yes, Škoda really used four-piston callipers on the rear engine Škodas), when unloaded the front end locked up way too soon, almost dangerously so for those that didn’t expect it, but when fully loaded (preferably with the heavy stuff in the front boot), they were well balanced.

    It won’t behave perfectly, after all when loaded on such a trip (let’s say 4 people 83 kg each, 50 kg luggage and about 7 kg for the roof rack that’s 389 kg) that is 67 % of the weight of the 600 and 65% of the 650 and those are some optimistic numbers. Such a massive difference just cannot be offset, even by a well-designed suspension. It’s the equivalent of getting a VW Passat B5 and loading it with 900 kg of luggage, it just won’t end well.

    There is only thing that is and that’s fun ;)

    Even the manufacture’s stated maximal load of 320 kg is 55%/53% (600/650) of the cars weight.

    Must say, I love the mirror fitted to the 126 on the last picture.



     

    Attached Files:

    • MaluchCaravan1.png
    • MaluchCaravan2.png
    • Like Like x 1
  18. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    Maybe I should try to work on the suspension before doing the deformation stuff, because otherwise I could make some changes to the front or rear subframe that would ruin the suspension behavior and would be hard to reverse later. Deformation values should always be the last ones to tweak, and I haven't even tested how well the roof can resist weight of the roof rack and the stuff on it, and if the cargo area can even store any heavy stuff inside without issues. I'll add the roof rack and some loads for it and the cargo area, support for vanilla loads, try to model some suitcases and bags (shouldn't be too hard I hope) and put them on, just to see how the chassis and suspension behaves and tweak what needed. If I didn't add the dummy today, I wouldn't have discovered that the connection of the floor to the front fenders was too stiff and couldn't hold a human, I had to fix that instantly, maybe there are more problematic things like this, I need to find out before I do the deformation.
     
    • Like Like x 3
  19. Xupaun

    Xupaun
    Expand Collapse

    Joined:
    Aug 18, 2020
    Messages:
    761
    What do you mean by "good suspension"? Maybe I understand less about JBEAM that I think I understand, but starting with the arms that controls the wheels, even if they are too strong, or the front ones doesn't deform like they should on longitidinal accelerations, and then start evening out is the way to go.
    For example, I work on a team that designs a Formula Student car, so we have to design its suspension from the ground up. First of all, we define the geometry (something that you already have) and then we start designing the components, (control arms, steering knuckes, hubs, steering arms and so on). Start with something not animated, that makes the car behave somewhat realistically (as I said, get the pigeon front suspension, move the nodes to fit the geometry that you have and for the beams make all of them adjustable, so you can fine tune them ingame). Even the vanilla cars doesn't have suspension JBEAMS that fit perfectly on their models, some are WAY off or are simply weird looking. You are way better than me on playng with jbeams and stuff, so I believe that you can easilly make a suspension work (it does not have to work exactly like the real one).

    Long story short, try to make the suspension geometry work (even with almost
    indestructible arms, and start tuning the spring rates. Everything else on suspension will happen almost automatically.
     
  20. Agent_Y

    Agent_Y
    Expand Collapse
    Jbeam/QA support
    BeamNG Team

    Joined:
    Jul 10, 2020
    Messages:
    10,056
    This is not as easy as it seems, because on this suspension the springs are integral parts of the mechanism. On the front the transverse leaf spring acts as the lower arms, and on the back the coil springs and shocks stabilize the movement as if they were the upper arms. It's not possible to do it from the ground up, because everything has to function properly at the same time and you never really know what is wrong. This sure would have worked on a more modern car, but not on this. Also, because of the lack of real lower arms on the front suspension, the upper arms Jbeam had to be made more complex than on other cars to remove the wheels flopping effect, the leaf spring also had to be designed in a really weird way unseen in other cars to simulate its side mounting points and reduce the effect of the wheels moving/wobbling back and forth under heavy braking, which is a thing in older cars but not this one. And on the rear suspension it's really hard to get the values right due to the springs being really short (which is due to the fuel tank being really close to the suspension and mounted really low, they have to be short to prevent the car from bottoming out too much). Using Pigeon suspension would make zero sense because it's fundamentally different from the one I have to make. I have used bits and pieces from other cars (front upper arms from Covet, rear arms from Pessima, coil springs from Miramar), but modified them so much it may as well be called scratch made, some parts ended up being redone completely from scratch several times. I know the spring rates are the most important, but due to them also directly affecting the way everything moves, when I do some changes to them, I often have to also change other bits in the suspension. I think the shocks should come the last to tune, and these are really hard to get right on such a lightweight car, and they are so uneven, front ones are 2 times longer but also the weight is much lower there... I think now that the weight distribution is accurate to real life, it will be easier to tune the suspension properly, but still it will be a nightmare.
     
    • Like Like x 2
  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