This has gotta be the most effort someone has put into their mod in the history of BeamNG. Its almost like you have an entire Facebook group on your side helping to create what is bound to be the best mod (when it's released)! And for this, I do not beg, I do not plead, I thank you for going this far to make a mod.
Hey, I appreciate your mod (Ibishu Blackfoot) and looking forward to your Jeep mod, but there's something I want to clarify. It's a very common misconception that a Locked transfercase/differential is 50/50 split. Think about it, if the locked is always 50/50, then how does driveline binding occurs? In fact, the opposite is true, an open diff(front or rear) is always 50/50 split. To make things easier to understand. Let's make a simple modeling, a differential or transfercase, has one input and two outputs, the mechanism splits the torque in half, and according to diff types, there's also a cross torque transferred between the two outputs. In an open diff, no cross torque, thus the torque split is always 50/50. If the input torque is 1000Nm then both the output torque L and the output torque R is 500Nm regardless of output speed L/R. In the case of planetary center diff, the initial torque split is different depending on design, such as 40/60 and if the diff is open, the split will be always 40/60 since an open diff has no cross torque. In an LSD, the cross torque is always from the faster output to the slower output. If the input torque is 1000Nm, the cross torque is 200Nm, then the slower output torque is 700Nm and the faster output torque is 300Nm. The cross torque depends on LSD mechanism, in a clutch or gear LSD the cross torque is preload + proportional to the input torque. If an LSD has a 100Nm preload and the torque bias is 0.1, when the input torque is 1000Nm, then the slower output torque is 500(initial torque split 50/50) + 100(preload) + 1000(input torque) * 0.1(torque bias) = 700 and the faster output torque is 500(initial torque split 50/50) - 100(preload) - 1000(input torque) * 0.1(torque bias) = 300. If the input torque is 0 then the slower torque is 100Nm, and the faster torque is -100Nm, since the clutch is allowed to slip, axle binding won't occur. A clutch LSD has high(such as 100) preload and low(such as 0.1) torque bias, while a gear LSD has low(such as 5) preload and high torque bias(such as 0.4). Viscous LSD cross torque depends on speed difference, if the viscous coefficient is 5 and the faster output is spinning 10rad/s faster than the slower output, then the cross torque is 50Nm regardless of input torque. In a locked differential or part-time transfercase, both of the output is forced to be in the very same speed, the differential locker or the transfercase will do whatever it takes to make the two outputs rotate at the same speed, input torque/speed doesn't matter and the maximum cross torque is the transmission strength. If you are traveling in a straight line, then the split is 50/50, if one wheel is in the are and the other wheel is on the ground, then the split is 0/100. If you are turning on tarmac, the wheel torque caused by tire friction is 1000Nm, the input torque is 100Nm, then the slower output torque is 1000Nm and the faster output torque is -900Nm, the torque split is -900%/1000%, this is too much for your drivetrain to handle and it may damage your vehicle. And that's the reason why you should never leave the locker(s) on or using part-time 4WD on tarmac. About the '4 FULL TIME' of your Jeep, the initial torque split is 40/60, and the cross torque is caused by viscosity, then if the rear wheels lose traction, then the split might be 50/50 or 60/40(higher/lower) depends on how much the wheelspin, if the front wheels lose traction, then the split might be 30/70 or 20/80(lower/higher) depends on how much the wheelspin.
All this may be true, but I've watched, read and been taught by many different people who own XJs and work on them. NP242 and NP231 Selec-Trac and Command-Trac transfer cases are fundametally identical. The NP242 is a 2.72:1 low range box, as is the NP231, the major difference being that the NP242 incorporates a planetary differential with a locking facility, and the NP231 does not. It has no differential. The NP231 is simply geared directly, and locks the front and rear final drives in the transfer case together with a collar. NP242 Selec-Trac has this same facility. In '4 PART TIME', a collar locks the front and rear driveshafts together. This WILL cause axle binding because Cherokees past 1986 or so do not have vaccum actuated front hubs, they are simply unit bearings, which is why the front driveshaft on all of these Jeeps will spin regardless of whether you are in 2WD or 4WD, in any selection, while moving - because the pinion will still be spun in the front differential, and the driveshaft will still be rotated. All that will happen at the NP242's end, is a layshaft will let that output run free in 2WD, with a collar engaging only the rear drive gear. In the part time mode, this shaft is locked with the rear one and the differential is bypassed. In '4 FULL TIME', which is puroposefully beyond '4 PART TIME' in the selection pattern, the axles are instead bound by the planetary differential, which is open, and works just like a regular axle-housed unit. Power is sent through it, but power will have little or no route to return back - so the axles will NOT bind. Upon my findings over time, I've discovered that the full time mode does not split in 40/60 F-R, but instead 48/52, with a torque bias depending on load scenarios. Having experienced the use of this selection in the real car, it's true that in this mode, the front axle is still capable of providing most of the pulling power on flat ground, and the variable torque split can be felt depending on if the engine power is ramped or reduced. As you say, it is technically possible in edge case scenarios for up to around 95% of power to be sent to the front or rear in this mode, since there is no fixed bias limit (no clutches or hydraulic actuators) with the open nature of the differential - this will happen where severe tractive differences are experienced between the front and rear axles. All of this is why Jeep encourage use of this mode year-round, because it not only eliminates axle binding, but also any undesirable driving characteristics which would result from a fixed torque delivery bias, such as that in '4 PART TIME'. N.B: '4H' on an NP231 is identical to '4 PART TIME' on an NP242. 4L is also indifferent between the two, the driveshafts are also locked together. Everything above agrees with what Jeep says on the subject, both in original brochures (which I now have) and original owner's manuals for the vehicle. The owner's manual in particular talks about the certainty of axle binding in 4H/4L on Command-Trac and 4-PART TIME/4L on Selec-Trac. Moreover, after experimenting with 4-PART TIME in my father's XJ on a gravel track, we experienced severe axle tramp and binding when turning off onto a tarmac section - not between the front wheels, but between the front and rear drivelines; it struggled to deliver any power until it was put back into 2WD. I hope that clears up a few things. This system has been extensively developed on the Archer Phoenix test truck, and has been tuned to perform correctly - I don't really fancy starting a debate about such matters. Edit: There is no cross-torque binding in full-time, because it is a 1-way differential. In part time modes, the drive is fixed, which is why cross torque scenarios (ie axle binding) exist.
i know this one still aint done, but any thought about doing the ZJ AKA Grand Cherokee? i have an XJ and a ZJ so this mod is like a dream come true lol.
Wow, this mod is insane! The attention to detail is just impressive. I don't think their are any other mods that have this much detail
Sorry to keep people waiting on here, but stuff is still being done. More scans are being collated, and although no images are to be shown yet, most of the floorpan is actually complete in a kind of alpha stage, at the general vanilla car poly density. I was going to just post an unannounced Streamable video (and probably still will) of a body-in-white driving without panels, so that's probably what you'll see next. However I can concede that the bare bodyshell is not finished yet. That being said, it isn't too far off. Keep your fingers crossed and you might spot something by December. In the meanwhile, here are some more photos of some of the new scans being used: A very generous part of the team actually owns one of these. I can't say too much but it is certainly something we are considering later down the line. If we get to making the ZJ, I'll be sure to let you know, it might come in handy. That's definitely a goal of mine. Not to be the wildest thing out there, but just to make it properly. Things will work that wouldn't on a base game car, although it might not be as fully featured as some other mods. This isn't being made to flaunt the limits, nor be the best mod ever, it's being made to demonstrate how close to dev-quality you can get while still adding your own touches. Although I have to admit, I do like my attention to detail, it's part of the reason this is taking a fair old while!
how do you go about the interior scans? can you just pan the camera left and right and have a decent enough scan (without seats and consoles blocking the floor)
I’m pretty sure he mentioned once he had to take hundreds of different photos. Because the image needs to reproduce the car fro, every angle to get proportions right.
Over 20 people have now contributed said hundreds of photos per scan. Small scans take about 30-60, larger ones (such as full body scans) can take up to 200. It depends on the scenario but really, the pattern is more important, since the program has to be able to build a 3D structure from your 2D photos. Hence varying angles are essential. You have to also make sure to get a good path between photos, instead of just jumping from one perspective to another. There is, but it's private. The server the car will use as its 'community hub' has no more updates than this thread does. The private server exists only for people who either own an XJ/MJ/SJ/ZJ or WJ, plus personal connections who are helping out. In short, be patient please. By the way, if any of you reading this own any of the Jeeps mentioned above, and would like to get involved, please get in touch via PM. As mentioned, the next update will be when the bare bodyshell is completed in its first stage. Modelling a car in this accuracy takes time and concentration, moreso than if I was just guesstimating what things look like, rather than making them to accurate shapes and dimensions. You'll have to take my word for it when I say that the floorpan is almost complete, and the trunk mostly in, but if you can't be bothered to wait, then this mod is not for you - go and find another WIP to get hyped about.
Happy new year everyone. I think the image above speaks for itself. The first bodyshell JBeam is done and is now undergoing further developement. All beams were done 'by hand' and the node arrangement is unique to the car; the file was written manually from start to finish. It features some special n/b structures for accurate physics in certain areas, but you'll have to wait until the car comes out to find out what they are. I know that this is probably the first evidence you've all had in a while that the car is still WIP; as mentioned it's because most progress is private. The final mod will of course be public, though. Apologies again for keeping you waiting for updates on it.
Really glad there is some progress man! Don't think we are disappointed, we get it if you need a break or can't finish the mod. JBeam looks great, however I am not educated in that too much. Keep it up
