new suspension

[poesel]

I have to say that I'm not too happy with the new suspension system. To be exact, I'm not too happay with the setup of the paladin.

One, the system cheats. The wheels have way too much spring travel - they punch through the chassis. A paladin is not a monster buggy.

Two, the setup is much too soft. No snappy cornering anymore. Feels like a family station wagon. I wish the suspension would be more 'italian'.

Clay, if you explain what the those do:

<SuspensionERP>0.2</SuspensionERP>

<SuspensionCFM>0.015</SuspensionCFM>

I will get my hands oily.

[Toby Haynes]

I would agree - it does feel a little soft. Maybe that's actually accurate but it does feel like driving an ambulance. However, it does feel more like vehicle suspension than the previous system so I'm going to give it the thumbs up.

[poesel]

Hmm, somehow something was filtered:

SuspensionERP

SuspensionCFM

should have been there.

[Bertram]

Feels like some crazy mechanic set up the suspension, or there is a design flaw in the verhicle (wrong placing of wheels vs centre of gravity). Did you notice that making a corner at speed will flip the verhicle, but making the same corner going in reverse will not flip the verhicle? Try that with your car at home.....

Also, moving at some speed and hitting "x" will flip the verhicle. Not sideways, but nose down, ass over end (very easy with the Strike, quite often with the Paladin). For that to be physically possible the centre of gravity has to be high above the wheels, the wheels close together (or the front wheels near the centre of gravity), the brakes over powered, and the grip of the wheels extraordinary. No real verhicle will ever do that (some bad designed ones will skid and turn over sideways, but that is another story).

[__Yossarian0815[jby]]

Can´t say how much I agree! :mad:

Lets go back to the way it was at least for the Paladin. Pretty Please.

[poesel]

(warning: lengthy post)

For those who don't want to read through all that: theres a link at the bottom to a mod with changed settings for paladin and shrike. Please give them a test drive.

First: I won't debate about beta or not, but I can't remember having so much fun setting up cars since F1GP.

Bertram, vehicles behave VERY differently if you are driving forwards or backwards. Normally theres more weight on the front axis because that makes steering easier. Also only the front wheels are steerable. So the whole kinematics is very different if driving backwards.

I do however agree with the rest of your post.

The wheeled vehicles in DT suffer from heavy understeering (front pushes over front wheels, vehicle does not turn). Normal cars are set up for slightly oversteering (back comes round first) because thats easier to handle and also what people expect. DT behaves different and that irritates me (and probably others).

The reason for that is IMHO too much longitudinal and not enough lateral friction at the wheels (my english fails me here - I meant 'Reibkräfte' and 'Seitenführungskraft'). This leads to the rather hefty reactions when accelerating or braking (longitudinal friction). Flipping over the front wheels is such a reaction.

OTOH the lack of lateral friction lets the vehicle push over the front wheels so you can't really control it. Additionally lateral friction would help stop swerving.

Back side of more latral friction is a higher tendency to roll over the side. But this is more a problem of the center of gravity.

Going XML (the forum filters lesser and greater signs):

- the chassis:

Mass

Density 0.01 /Density

CenterOfMass 0.0,0.0,0.0 /CenterOfMass

Width 16.3 /Width

Length 29.49 /Length

Height 8.9 /Height

/Mass

CenterOfMass seems to have no effect on driving physics. I think this will only be used for collisions.

MaxMotorForce 3400.0 /MaxMotorForce

MaxWheelSpeed 25.0 /MaxWheelSpeed

WheelAcceleration 45 /WheelAcceleration

AccelerationRate 0.006 /AccelerationRate

MaxMotorForce seems to be responsible for the chassis reactions while accelerating. Seems not to have an effect on the actual acceleration.

MaxWheelSpeed times the number of working wheels is the actual max speed of the vehicle.

WheelAcceleration seems to be responsible for the possible real acceleration.

AccelerationRate delays the time until full throttle is in effect. The value is very low in comparison to tracked vehicles (0.1) because else it would flip over backwards. The cause for that is IMHO the high longitudinal friction. Normally the wheels would just spin and kick up dust (or you have ASR which is another german abbrev. I don't know the english equivalent).

- the wheels itself. There are four of these paragraphs. Difference is mainly if IsSteeringWheel is set or not.

Type Wheel /Type

ID 4 /ID

Joint

Type Hinge2 /Type

Anchor 8.4,6.4,-3.2 /Anchor

Axis 0.0,0.0,1.0 /Axis

Axis2 1.0,0.0,0.0 /Axis2

LoStop -0.00 /LoStop

HiStop 0.00 /HiStop

SuspensionERP 0.16 /SuspensionERP

SuspensionCFM 0.005 /SuspensionCFM

/Joint

IsSteeringWheel 1 /IsSteeringWheel

ModelFile IFVWheel.cob /ModelFile

Origin 8.4,6.4,-3.2 /Origin

Anchor and Origin should be the same else you get excentric wheels. Very funny but not useful. Although this flexibility is IMO a very good indication of the possibiltys of this game engine!

LoStop & HiStop seem to have no effect whatsoever. I thought that would limit either spring travel or deflection but setting it to small values or zero had no effect.

SuspensionERP seems to be the spring force.

SuspensionCFM seems to be the spring travel.

There is no parameter to change dampening (if that is simulated) and more important: there is no way to change the center of mass/distribution of weight. That would help a lot.

- other remarks:

There is no real brake! Pressing 'x' just sets the throttle to zero, pressing 's' puts the throttle in reverse and will eventually get the vehicle moving backwards. Please implement a brake! Depending on the friction we need an ABS or not.

For everyone that made it down here, heres the result of my tuning efforts for the paladin and the shrike:

Tuning

Its not perfect but IMHO much better. There are some things I can't change (see above).

The new vehicle are at the top of the list. The originals are also still in the list.

The differences:

Paladin:

MaxMotorForce 2400.0 - 3400

WheelAcceleration 40 - 45

AccelerationRate 0.004 - 0.006

Front wheels:

SuspensionERP 0.2 - 0.16

SuspensionCFM 0.015 - 0.005

Back wheels:

SuspensionERP 0.2 - 0.18

SuspensionCFM 0.015 - 0.010

The paladin seems to suffer a lot from the artifical 2x resizing that has been done. It steers much like a bus (which is comparable in size).

The changes make it a bit stiffer and more responsive.

Shrike:

MaxMotorForce 600.0 - 200

Front wheels:

Anchor 6.3,5.6,-2.2 - Anchor 6.3,5.6,-2.8 (and Origin)

SuspensionCFM 0.03 - 0.007

Back wheels:

Anchor 6.3,5.6,-2.2 - Anchor 6.3,5.6,-2.5 (and Origin)

SuspensionERP 0.2 - 0.18

SuspensionCFM 0.03 - 0.018

The flipping over the front has its source in the front of the shrike dipping down to much and so it gets stuck in the ground and thus flips it over. I've raised the front (more) and the back (less) and stiffened the suspension. You can STILL flip it over the front but its much less susceptible now to do so.

The posting ends here. I'm mortally offended if I don't get adequate responses!

[__Yossarian0815[jby]]

Originally posted by poesel71:

(warning: lengthy post)

The posting ends here. I'm mortally offended if I don't get adequate responses!

What, no videos?!!

[Imperial Grunt]

Poesel71, I understood about 10% of what you said, but I absolutely concur!

Having travelled around in AAVs, M-1 tanks, and HMMWVs extensively though, I must admit that the vehicles in DT definetly behave much differently. M-1 tanks (or any others) do not bounce back into the air, even after going airborne off of a berm at 60 mphs...they just land and continue. (I also think that the Thor's and Apollo's top speed is too slow.) How any armored vehicle would bounce and roll over after being dropped straight down on a relatively flat surface is beyond me. It would probably just bog down into the dirt. In my opinion, a more realistic landing sequence would be like the one shown in the movie "Aliens." In that movie the Colonial Marine dropship actually landed and lowered a ramp for the Marine's armored vehicle to drive down and away. Then the dropship quickly took off.

Shrikes and Paladins also flip over way to easily in my opinion. They also seem to keep rolling after the throttle is let up as if there is absolutly no friction in the driveshaft, transmission, etc...and as said before, there are no brakes, just a parking brake.

My Acura TL is a much better vehicle! Just need to add on an ATGM launcher and a 20mm gun onto it somehow.

Bot driving still sucks, especially when you tell them to go somewhere. Putting them in formation on anything more tough than a parking lot or golf course is also a bot death sentance.

Dropships that get shot down and crash have this ability to roll and tumble on the ground...shouldn't they just explode on impact spectacularly? (Also causing damage to whatever they hit and a fire, except in a vacumn environment.)

But with all that, I love playing this game.

[Bertram]

On dropships: stuff (like planes) crashing explodes because they contain something that can explode, the fuell on board. If the dropships use anti-matter (as per background story) they would either not explode, if the anti-matter stayed contained, or go in a spectaculair blaze, if the containment failed. Viewing the energy requirements of lifting from the gravity well repeatedly, the blaze would be of nuke intensity (ending the game abrupt....maybe a good reason to make firing on dropships a no-no?).

If the ships would not explode they would crumple on inpact. If the impact point was a hill side or some such, and the impact at a shallow angle (or on hard terrain at a shallow angle) several pieces would shatter around (like tail pieces and/of wings of airplanes in accidents where they break up). In soft ground they would bury pretty deep (in the soft clay here WW2 airplanes or airplane parts are still found occasionally, the heavier parts like engines usually at about 2 mtrs below ground level).

As to driving: Poesel, I understnad driving backwards gives a different effect - in fact I was trying to say that in the real world you turn over faster driving backwards, in contrast to the experience in the game. (There was a Dutch made car, the Daf 66, that was one of the first variomatic users. They could go as fast backward as forward. A TV programm organised a backward driving race on the Dutch grandprix circuit. Of the 66 entries only a few finished, and even those had rolled over repeatedly. All others crashed to hard to make the line).

I am surprised that the centre of gravity isnt used in the calculations. Makes you wonder how the caracteristics are calculated... just each wheel (corner?) independent? In that case it would not be strange that the back wheels keep going, once they are airborne...no friction to slow them down.

[ October 30, 2006, 10:49 AM: Message edited by: Bertram ]

[Bertram]

Post didnt work, I'll try a new one..

[ October 30, 2006, 11:19 AM: Message edited by: Bertram ]

[poesel]

Which file do you mean? Unpack Tuning.zip inside your /Mods directory then you can choose that mod from within DT.

Yes, the -10 was a test. Forgot to revert that back. Makes no difference however (AFAIK).

The center of gravity is most probably used - you just cannot change it from the XML files (or so it seems).

[Bertram]

Poesel: where did you get that file from? I seem to be unable to find it...

In the CentreOfMass for the chassis there is a -10.0 value...is this original or is this you trying some things out?

Edit: found it, I was looking fopr ShrikePhysics...

There are actually 2 files for the Shrike out there, one old one (19 august 2006, called ShirkeA) with

{Type}WheeledChassis{/Type}

{ID}1{/ID}

{ModelFile}ShrikeChassis0.cob{/ModelFile}

{Origin}0.0,0.0,0.0{/Origin}

{Mass}

{Density}0.035{/Density}

{CenterOfMass}0.0,2.0,-10.0{/CenterOfMass}

{Width}7{/Width}

{Length}7{/Length}

{Height}2{/Height}

{Type}Wheel{/Type}

{ID}7{/ID}

{Joint}

{Type}Hinge2{/Type}

{Anchor}-6.3,-5.6,-2.2{/Anchor}

{Axis}0.0,0.0,1.0{/Axis}

{Axis2}1.0,0.0,0.0{/Axis2}

{LoStop}0.0{/LoStop}

{HiStop}0.0{/HiStop}

{SuspensionERP}0.2{/SuspensionERP}

{SuspensionCFM}0.1{/SuspensionCFM}

The new one (from 25 oktober 2006, called Shrike) has your values. Stands to reason that with the very low Density, changing the CentreOfGravity has not much impact.... Might be interesting to see what happens when you upp the density again, and change the centre of gravity around.

[poesel]

Good idea - will try. ShrikeA is used nowhere in the game. Dunno where it came from.

[Jack Carr]

Yeah, it's strange alright. I used to have a bit of confidence with the Paladin because of its speed and maneuverability, but not anymore. It's still fast, but try zig-zagging through a patch of trees with the new suspension...WHAM! You'll run into something and stop dead right quick. And then, as everyone well knows, once you become motionless in a hot zone, you're just a sitting duck.

[ClaytoniousRex]

front pushes over front wheels, vehicle does not turn

Poesel, can you explain that again? I will probably agree with what you're saying but I don't understand it yet.

The reason for that is IMHO too much longitudinal and not enough lateral friction at the wheels

The lateral friction is indeed lower, mostly to prevent rollover. What we really need to do is stop hardwiring it and give it to you in the XML so you can play with it yourself. Will do that.

CenterOfMass seems to have no effect on driving physics. I think this will only be used for collisions.

No. COM is definitely used for driving (and all other) physics. To unambiguously see its effect, try some huge numbers like -1,000 on one of the axes or somefink.

MaxMotorForce seems to be responsible for the chassis reactions while accelerating. Seems not to have an effect on the actual acceleration.

This is the torque force applies to the wheels. It definitely has an effect (try really small values, for example) but remember, as you noted in your post, the THROTTLE is clamped over time by a different tag, so only raising this number won't make the wheels spin faster. But, for example, if you raised the wheel friction to a very high number and then tried to drive up a steep slope, you might need to raise the MaxMotorForce in order to get enough torque to make the wheels actually spin.

SuspensionERP seems to be the spring force.

SuspensionCFM seems to be the spring travel.

Correct, more or less. At least, that's as well put as I could have done. It's actually "Error Reduction Parameter" and "Constraint force mixing", but your definition is good for their actual effects.

There is no real brake! Pressing 'x' just sets the throttle to zero, pressing 's' puts the throttle in reverse and will eventually get the vehicle moving backwards. Please implement a brake! Depending on the friction we need an ABS or not.

Yes, there is a real brake. Holding X actually applies torque to the wheels to make them stop. It does not only set throttle to zero. However, with the increased mass of the vehicles, the amount of force applied by the brake is now too low and needs to be raised.

heres the result of my tuning efforts for the paladin and the shrike:

Off for a test drive now. Wish me luck!

[ClaytoniousRex]

but try zig-zagging through a patch of trees with the new suspension...WHAM!

One thing that's very important is to do what you do when driving in Real Life, which is to take your foot OFF of the gas when turning. Especially in an all-wheel drive vehicle, it's very important to not have those rear wheels pushing you forward while trying to turn with the front wheels.

Anyway, try being more delicate with that W key. It can have a tremendous impact on turning.

[aittam]

Originally posted by ClaytoniousRex:

</font><blockquote>quote:</font><hr />but try zig-zagging through a patch of trees with the new suspension...WHAM!

One thing that's very important is to do what you do when driving in Real Life, which is to take your foot OFF of the gas when turning. Especially in an all-wheel drive vehicle, it's very important to not have those rear wheels pushing you forward while trying to turn with the front wheels.

Anyway, try being more delicate with that W key. It can have a tremendous impact on turning. </font>

[poesel]

Originally posted by ClaytoniousRex:

[QB] </font><blockquote>quote:</font><hr />front pushes over front wheels, vehicle does not turn

Poesel, can you explain that again? I will probably agree with what you're saying but I don't understand it yet.

</font>

[Toby Haynes]

Originally posted by poesel71:

AFAIK all cars are set up to be slightly oversteering since that is easier to handle: when that happens, the back moves first but the steering wheels do not (yet) slide and you keep control of the car.

I'd say that most front-wheel drive cars understeer, rather than oversteer. Oversteer in a car quickly leads to the car spinning - most people can't correct for oversteer safetly. Certainly my car (Acura RSX, front wheel drive) understeers when I apply power in a sharp corner. The result is that the car drifts wide of the intended line. Oversteer would step the back of the car out and would require less steering lock or reverse lock to correct.

The behaviour of the paladin is certainly understeer. As Clay points out, using the throttle in the corners causes significant understeer. Given the power of the engine on these vehicles (which is, as far as I can tell, much greater than domestic vehicles) that is entirely correct.

The reason for understeer when accelerating through a corner is complex. Think of a tyre having longditudinal friction (i.e. in the direction of travel) and transverse friction (sideways). Longditudinal > transverse for most (all?) tyres. However, the amount of friction the tyre can deliver sideways is dependent on the amount of friction being delivered longditudinally. As the longditudinal friction is increased, less lateral friction is available, leading to sliding and understeer.

For a mathematical description of the physics of tyre friction, try looking up the Pacejka Magic Formula on Google. For example:

http://www.racer.nl/reference/pacejka.htm

Cheers!

[poesel]

I have to say that I was wrong with that cars do generally oversteer. I looked that up and its really understeering. Sorry, had that the other way round in memory.

[..]Certainly my car (Acura RSX, front wheel drive) understeers when I apply power in a sharp corner.[..]

Putting power into the equation is another thing. All front wheel drives tend to understeer when you accelerate in a curve (the added acceleration 'eats up' the available friction on the tire).

Other way round for back wheel driven cars: they tend to oversteer for the same reasons.

But DT has 4 wheel drives. Still, the front wheels have the added burden of steering the vehicle (lateral forces) which lowers longitudinal friction. That means understeer.

What a manufacturer usually does now is to shift the weight balance to the front wheels. The added force on the front creates more friction and compensates for the steering.

And lets brake in curves: deccelarating or braking creates a momentum around the front wheels -> the back gets 'lighter' -> oversteer. Physics is fun.

Back to DT: I can of course only speak for myself, but I would want my vehicle to be oversteering (like a rally car is set up (yes I looked it up this time )) to get the thing faster around corners. That is very useful especially with keyboard control.

[Bertram]

I would like to add some wishes: as it is you get the "engine" sound when you hit the <w> (forward)key, when you stop depresiing it your engine (sound) stops. This gives the impression that you need to keep that key depressed for driving along. According to the devs this is wrong, each time you depres the key you actually add more power, and leaving it alone makes you move at a constant speed.

Those two things are not campatible and tend to confuse. I would like a constant engine sound when driving, and a heavier sound when driving faster/uphill (thus when adding power).

Second, the behaviour of tanks and cars seem to be different in this regard. Driving a tank, and leaving the power alone will make it coast to a stop. A car will keep going. This makes it extra cnfusing, as you need to keep hitting that key in a tank, or it will stop, while doing the same thing in a car will keep you acelerating.

[Jack Carr]

Originally posted by ClaytoniousRex:

</font><blockquote>quote:</font><hr />but try zig-zagging through a patch of trees with the new suspension...WHAM!

One thing that's very important is to do what you do when driving in Real Life, which is to take your foot OFF of the gas when turning. Especially in an all-wheel drive vehicle, it's very important to not have those rear wheels pushing you forward while trying to turn with the front wheels.

Anyway, try being more delicate with that W key. It can have a tremendous impact on turning. </font>

[Smeltz]

I kinda skimmed over the long posts so sorry if I walk over someones previous thought.

Clay-- You mention about leaving off the gas while turning. This is true in the OLD 4X4's without proper differential set-up. Now days, with limited slip, when in a slide, floor it and pull out. Am I also to understand in a tech filled universe with space travel they never heard of yaw control or traction control? Okay, enough of picking on Clay, now to the heart of the problem...

I fully agree that the wheeled vehicles act goofy with the new suspension set up. I believe the true problem lies in perception. Since the DT world doesn't reflect ground speed well in the third person view and you can't FEEL anything it makes it very hard to drive. Driving with a keyboard almost impossible as it is in a driving sim, let alone a battle sim. The input you get from just your eyes, as your also looking for targets, isn't enough to help you feel the attitude of the vehicle. Tracked tanks are easier because they are inherintly more stable and slower.

I considered trying with a joystick but haven't ponied up for a new one. I certinally won't be getting a driving wheel, though that would be cool, just because of cost vs. use ratio. So what is the solution?

I hate to say it but the vehicles need to be made more "gamey" Arrgggg, I hate that. Since the inputs for controlling the vehicle are so limited, the vehicle needs to be dumbed down to compensate. Right now it's like play Grand Prix Legends (one of the best racing sims ever made) with a keyboard; it just doesn't work.

All in all I'll end my slap at Clay with a compliment; The suspension and physics of the wheeled vehicles are well done and I would guess acurate. Problem is, too much so for playablity. A new learning curve is fine but I fear the wheeled vehicles will not be used since they get smoked by the bigger guns easily, and you just can't steer the darn things! They have lost their reason for being...speed.

As for the brakes...they need an upgrade cause if I had them on my car I'd be taking them back!

Schmeltz

[Tarquelne]

Tracked tanks are easier because they are inherintly more stable and slower.

I keep getting spooked when a sudden stop has the rear end of an Apollo lift like the thing's going to flip end-over-end. Even the tracked vehicles are less stable than I expected.

Are these AFVs supposed to be a lot lighter than the "real world" modern versions?

The Paladin - and especially the Shrike - strike me as remarkably unforgiving for fighting vehicles. I dunno... maybe at the high end of the learning curve you can really use their advantages.

I hate to say it but the vehicles need to be made more "gamey" Arrgggg, I hate that. Since the inputs for controlling the vehicle are so limited, the vehicle needs to be dumbed down to compensate.

Could be realistic - in the fits-with-the-sci-fi-world sense - if the vehicles have some sort of drive-by-wire system that condenses a complicated operation to a few simple controls. An excuse to make the vehicles more forgiving. Make it a damage-able component (Knocked out with "yellow" on the dirver?) and I wouldn't think it at all gamey.

After spending much of the day rolling Paladins - attempts to use that potential speed for armor - I'd love to see the wheeled-ones able to survive a roll intact and keep going. Either they strongly tend to land right-side-up, or auto-jacks put them upright. And with the turret intact.

That way you could have a hoot zipping around the battlefield, but without hitting control-K or \ every 45 seconds.

Or at least _I_ could.

[poesel]

I don't think the vehicles need to be more gamey. The 'old' paladins were more 'crisp' and hard to drive but after a while you could drive on two wheels with these things - on a keyboard!

The new suspension was a first try to get more realism and I think it was a very good try! It was only a bit on the safe soft side. Lets see what the next iteration brings.

What is definatly missing is a speedometer in the normal view. And the one in the gunners view should be readable. That would give you a feeling how fast you really are.