Tank turns forever

[scotg 204]

My tank turns left or right forever (in neutral steer) whenever I press "A" or "D" and let go. Sometimes, I can cancel it by driving forward. Any ideas what parameters can eliminate this ridiculousness? I've been trying to adjust some of the tankX wheel params, such as latStiffX, latStiffY, dampingRateInAir, dampingRate, and MOI. Tweaking these might reduce the problem (like slow it down or it'll only turn around twice before stopping on its own), but I have yet to find a solution that doesn't mess up something else entirely. Any ideas?

[wld427 1705]

your MOI or sprung mass is borked. Start by messing with those two values. 

[scotg 204]

Thank you for the reply, but I'm not sure how much to adjust since those settings have such different value ranges. Looking at the the Tank Guidelines gives a small hint, but I am having troubles figuring out the rest. I borrowed the tankX settings for RHS's M2A2 and got it to work for my tank, but in doing so I had to increase my tank's mass. It's supposed to be a light tank based on the RDF/LT platform, so it needs to maintain a low mass for sling-ability. That's half the point of me doing this tank. Also, after testing the borrowed "working" settings, I now find that sometimes it pauses while attempting to do neutral steer, and then suddenly jumps out in an uncontrollable flying spin. Incidentally, it also does that with the M2A2, but since I haven't played with the M2A2 much I am only just now finding that out. I wonder if RHS are aware of this. In both vehicles this weirdness happens only about 1 out of 6 times.

So basically, messing with the MOI, sprung mass (not so much), and dampingRateInAir have proven to affect its stability. However: fixing the spin makes it stuck; fixing the stuck makes it fly off ballistic. Fixing the flying off makes it spin forever, and I'm back to the original problem. I avoid messing with the sprungMass too much because the math in the Tank Guidelines suggests sprungMass = mass/12 (the number of wheels it has), then tweak a little.

[wld427 1705]

the guidelines are junk.... and essentially just that.. guidelines

 

Typically distribution of weight through the GEO lod, as well as sprung mass will have drastic changes. 

 

I fought very hard with this on my Type63 APC. 

 

Sprung mass in my findings is much more key than that calculation. 

 

MOI can also have drastic effects such as your shooting off into space.

 

If you are already"borrowing" the RHS physX configs skip trying to use the Abrams... try the settings for the BMP or something with a closer weight to what you are constructing.. 

 

 

 

 

[scotg 204]

Roger that, but the M2A2 is the Bradley. In the RHS line-up I think the mass of the M113, at about 13,000 kg, is the closest to an original RDF/LT, which is 14,800 kg. Although the one I'm doing is fictional, I want it to be believable, and I imagine an up-armored RDF/LT to be approximately in the range of 17,000 to 22,000 kg. I'm not familiar with many other mods to be inspired from, but if you have any suggestions then I'm all ears.

Side note: In doing research on real life helis and their payload capacity, few could potentially carry an up-armored RDF/LT. Fortunately, I am also working on a fictional upgrade to the CH53E, which is one of those few that might do it. Although it lists a payload of about 14,000 kg, it also boasts is can carry an M115, its crew, and its ammunition. I estimate that to be about 20,000 kg with 14 men and 14 rounds, based on the M115 specs at wikipedia.org. Therefore, a 20,000 kg tank and three crewmen should be feasible, right?

[x3kj 1247]

thrustDelay config parameter should be close to 0. (0.1 is ok for example). Wrong rpm and gearbox settings can cause permanent application of "force". RHS has very simplified physx settings on their tracked vehicles (last time i checked anyway, which has been quite a while ago) - that could cause some other unique problems if you use their configs as base.

[scotg 204]

The replies you guys give are great! Even if someone suggests something that doesn't work or has already been tried, it's so nice to have responses because they motivate me to carry on. When no one replies it can be a little discouraging, and then I start hearing that Genesis song in my head, "No Reply at All." 

Quote

Is anybody listening? Whoa-oh-oh-oh.

So, a hearty "Thank you!" to you guys imparting your experience on guys like me, and thanks to others who keep motivating us by just contributing to the conversations. There may be a general lack of interest for my mods, but at least you guys don't disparage me from trying to accomplish my goals.

 

@x3kj: I did notice RHS' values seem to be significantly down-scaled compared to the BIS sample tank, if that's what you mean. I pushed the numbers of the gearbox tighter at low gears and wider at high gears, which allows better acceleration. Very helpful!

Of the two tanks I'm working on simultaneously, one is tall and the other is squat (the previously mentioned, modified RDF/LT). Taking on two at a time might seem more challenging, but will hopefully give me a better understanding for my future tracked vehicle projects. After some hours of minuscule, tedious value tampering, I managed to get them to not radically fly off most of the time. Although, they both sometimes hesitate when turning, and then suddenly leap a little as if released from a snag. Initial testing in the Virtual Garage shows they can neutral steer, accelerate, turn, and stop with much better control, and adequately within the response time I want for them. They still have some individual issues to overcome:

 

Tall Tank (fictional High Speed Sentry tank)
It has 10 wheels and a mass of 20,000 kg. Top speed is 91 kph in Virtual Garage. It will easily spin out and stop when turning at speeds of 20 kph or more, but it's hard to roll over on flat ground. Testing on a terrain map (I tried Stratis and Tanoa) shows that the spin-out effect seems greater going downhill, and it makes cruising down a curvy path very difficult with WASD keys. It is sluggish going uphill, but maybe that's just expected for heavy vehicles.

 

Squat Tank (modified RDF/LT)
It has 12 wheels and a mass of 18,500 kg. It is much easier to control on curvy paths. Top speed is about 72 kph. However, instead of spinning out it rolls over while turning near its top speeds or in sharp turns on downhill paths. Ok... that seems fair, as I imagine that could happen IRL, and even some of the RHS tanks do that. It's just a little weird-looking when my squat tank rolls while my tall tank doesn't. Squat tank is also very sluggish going uphill.

 

[wld427 1705]

try upping your gear box values for more force going up hill in the lower range of the gear set. remember bumping these up will have an effect on the vehicle on level ground so don't get too crazy or the bitch will launch like a top fuel dragster. 

 

for rolling take a look at your weight distribution. you may need to add some boxes in the GEO lod and lower your center of gravity and disperse the weight more

over the tacks instead of dead center

[scotg 204]

7 minutes ago, wld427 said:

try upping your gear box values for more force going up hill in the lower range of the gear set. remember bumping these up will have an effect on the vehicle on level ground so don't get too crazy or the bitch will launch like a top fuel dragster. 

 

for rolling take a look at your weight distribution. you may need to add some boxes in the GEO lod and lower your center of gravity and disperse the weight more

over the tacks instead of dead center

Good idea! I will keep tweaking that gearbox slightly.

I'll have another look at my mass cubes. From a top-down view, perhaps I'll arrange them in a rectangle instead of a diamond, and then lower them near the tread base.

[wld427 1705]

a diamond might not be a good idea....

 

Think of the geometry as a toy.... when you slide it across the floor what is likely to happen. 

 

a low 3D rectangle like a book will slide.... a ball will roll... other shapes do other things....

[scotg 204]

5 minutes ago, wld427 said:

a diamond might not be a good idea....

 

Think of the geometry as a toy.... when you slide it across the floor what is likely to happen. 

 

a low 3D rectangle like a book will slide.... a ball will roll... other shapes do other things....

Well it's flat on the ground, no matter what orientation the book is. If you rotate it 45 degrees, it will still slide the same. That's the principle I see happening, but I want to try it book-style on my tank anyway in case there are any other benefits to having a quad/book layout. The cubes will have to all be the same weight, though. The benefit of the diamond shape is that the side cubes could be lighter or heavier than the front and back cubes, and that the front and back cubes could even be a little different mass from each other.

[scotg 204]

Messing with the gearbox in tall tank is only having a moderate affect on acceleration, so far, and I haven't accidentally come up with a graph that makes the tank launch with a drag wheelie. It's a good thing, but the flip side is I haven't come up with a solution that helps it climb, either. There are some more details I have noticed while watching the performance more closely, thus I think the skid-outs and poor climbing might be related. Skid-out steering is less drastic (but still present) when not applying the gas. In fact, I can gain some degree of control when coast-steering. Any amount of steering, however, causes more speed and power loss than it should, but it was not obvious to me at first. Again, it seems to be more noticeable when also giving it gas. Lastly, trying to roll over the tank is really difficult. All of this combined makes me think maybe there's too much grip, but I'm not exactly sure what controls that. I've used latStiffX and dampingRateInAir to control unwanted continuous turning (the original problem), but maybe now one or both of those settings are causing the severe oversteer. Doesn't seem logical in writing, though.

I haven't messed with the squat tank yet, so I'll get to its rolling over issue at that time.

[scotg 204]

I had previously tried borrowing tank wheel and engine parameters from the RHS M2A2 Bradley, but my adjustments gradually got so far off that I couldn't get back to its base performance. Several iterations throughout the trials encountered the "tanks turns forever" problem, or something akin to it.  Every iteration caused high-speed skid outs and other steering difficulties, no matter what. I decided to look at all the tracked vehicles available to me once again, and settled on basing tall tank's parameters on the BIS Tigris AA. I was able to increase the speed and drop the braking power to where it suits my tank very well. There's no problem steering, unless it is sharp turning at very high speeds (65 kph or more). Being a tall medium tank, it's actually good that there's a speed/turn breaking point. The only issues it has now are visual wheel spin. Because of the variety of wheel sizes, some of the wheels are invisible functioning wheels, and some are just visible rotational pieces that don't actually interact with the ground. Visible wheels are attached to the same podkolo as their corresponding invisible wheels, thus creating the illusion that it's doing the work.  Since the visible wheels are not influenced by the ground it's a noticeable over-spin, but it's also somewhat ignore-able for now.

 

After all this, I'm not sure I have a real grasp on tank editing, but at least I found an acceptable set of values for this one!

 

Many, many thanks to wld427 and x3kj!