Here are a couple references for front and rear crashes, both are at 35 MPH. Not sure how fast you were going, but the A and C pillars shouldn't deform at speeds that low. Keep up the good work! https://www.youtube.com/watch?v=vjuougjSuEg https://www.youtube.com/watch?v=hE_uVIIbcP8
Those will help out aloooooot. Thanks bro. Its not exactly there yet but its getting there. Sent the latest to cbr so he can put his magical touches on it.
Thanks! that's awesome, that'll help a lot. Time for a crash test skin then! Btw, that looks like it's a very safe car, specially for that era where safety wasn't on the top of the list for car manufactures. I wouldn't want to try it in real life, lucky for me I have it in 3D to do so
Doing some progress to the rear. Still need to strengthen the B and A pillars so the roof won't collapse like that. (that's a 35 mph crash)
One thing I noticed. U need to separate the quarter panel beams closer to the cabin and make them stronger. I was gonna do that before I sent it off to you. That's why we are having a hard to making just the trunk smash most of the beams for the rear end are all together as one setting for their strength. Dang good texture work on those wheels btw! Can't wait to see the interior get some color.
Looks far more accurate, keep it up! I wasn't able to find any references for a side impact, as side impact testing wasn't very common back then. But I was able to find a side crash test of a '81 Civic at 30 MPH. Hope this helps a bit. http://youtu.be/mmQ2p626y7g
Now it's getting somewhere: That's a full front crash, and 3/4, both at 60 km/h or 35 mph. Did a LOT of work to the front, I think the main deformations are done. I'll just need to add the missing parts and that's it. Need to work the rear and roof still. I've improved the handling and front suspension, but now when I hit the brakes the front struts break, and the front wheels are a bit unstable. The crash test skin is just a joke, didn't bother much with it, but might include it, why not.
Sounds like the strut towers were made too weak to assist in the crumpling in a wreck. Need to separate the beams some more and make different sections weaker. Whenever I get the next version I'll see what I can do. Looks amazing btw
I think your front crumple zone needs to... well... crumple a lot more. This is the NHTSA 35 mph crash test compared to yours: On the real car it's crushed almost all the way to the front wheel and the front wheel is pushed back into the firewall. Your car's about a foot less crumpled and the front wheel hasn't moved at all. Also, I'd recommend some more support beams from the fender to the door (or door frame) to stop it clipping and instead have it crumple like the real car. I think the passenger cell ought to be weaker, too - 35 mph offset in a 35 year old car should fold it up considerably, especially since a 1992 Honda Civic does this poorly at just 5 mph faster:
Thanks Gabe, I'll keep what you said in mind. Now I'm a couple of steps back because I've had to fix a lot of nodes from the chassis that were giving me problems with the suspension and other parts. Right now I have a steady suspension setup except for the rear wheels wich I still couldn't figure out how the axis on the pressurewheels work exactly, and need to set up the anti-roll bar properly. Getting this to work like it should is being a pain in the rear without proper documentation, but it's getting there. I'll post more progress as soon as I get something worth showing. - - - Updated - - - Ok guys, I've got to a point where I'm almost giving up with this jbeam stuff. I just don't want to spend more time on this getting nowhere, I want to move ahead, and it's getting really frustrating taking one step forward to step back two. Once I fix something, another thing breaks, once I think I've figured something out, turns out I'm wrong, and it's a full circle. Here's a video to see what I'm talking about, this happened when I started to reduce weights: I can't even bring the damn thing down to it's proper weight (should be about 920-950kg, not almost 1.100kg), every time I start reducing weight on the nodes they just break or do crazy stuff. Even by using the scalenodeweight factor, I just can't seem to find the right way to do it. I know I can delete some nodes to keep the weight down, but is that really a solution? I'm not interested in having a 1200kg car when it should be less than 1 ton, neither I'm interested in just taking another car jbeams and fit them to mine just to make it work. Anyway, I'm getting reeeeeaaaally tired of this and needed to say it... but I'm not an easy quitter.
That happens when the beamspring or beamdamp are too high relative to the node weight. Usually when you get to <1.5kg per node this starts to happen. Some solutions include: Lower the beamdamp on metallic parts to around 10, high damping values often cause the jbeam to implode. Are these light nodes concentrated in a specific part of the jbeam? if so, try to even out the nodeweights over the part that's causing the problem. keep in mind that you must also take into consideration the front/rear weight distribution and centre of gravity. High beamspring values always cause problems with light nodes. gradually lower or raise the beamspring value until the jbeam is stable. Too little spring will give you jelly, and too much will divide by zero Also, if you can't make a part rigid enough without it exploding, try adding extra support beams, which increases your rigidity mostly without any troubles. Give these beams separate values, so they don't affect the deformation of the vehicle. lastly, you should not really use the scaling factors (I chose to remove them from my jbeams altogether), as they affect large parts of the jbeam, and may compromise the realism of your jbeam. Anyway, you just have to keep trying different things, and you will solve the problem eventually. it seems the error is in the chassis, so try vastly lowering the beamspring values until the frame stops imploding. This way you can pinpoint the section of jbeam that is not cooperating
That's just the beamspring and beamdamp being too high for the weight of that node. If you want, flick me a PM with the unstable version and I can take a look, but it shouldn't be too hard to fix yourself.
To the contrary, doing that may actually send him to the nutter. Really enjoying watching this project unfold though (despite current frustrations), especially with the devs helping out. It's a sign of a healthy community in the peak of it's form.
The hood buckling is selfcollision related. if the nodes have self collision enabled, they need to be at least 2x collision radius apart (usually 0.04m total) At the edge of the hood your nodes are too close to the body or fender or what have you. Edit: Making a car is really really tough the first time around. There is a LOT to figure out, and it would be difficult even with documentation because a lot of it comes down to mechanical intuition. But I think you are making excellent progress.
