The visual damage in LFS is pretty decent as-is, but there's one thing you could add that would really make it a lot better: Hood bending.

Whenever a car hits something, the hood will always bend upwards, seperately from the rest of the car.



However, the way it is in LFS now, the hood just stays flat, and actually just gets SHORTER instead of bending when the car hits something.


I'm not sure how this would be implemented, but it would be a huge difference. The engine bay wouldn't need to be modeled, it could just be filled in with black. I know there are a lot of games that don't have this and the damage just doesn't look right. In the few games that do, such as DiRT, it makes a massive difference.

It's a well-known fact that LFS's damage model is unfinished. In fact much of LFS is unfinished. That's probably why it's still in development.

Besides, what you're asking for is pure eye candy, and we tend to frown upon such things around here.

Quote from Forbin :

Besides, what you're asking for is pure eye candy, and we tend to frown upon such things around here.

No, its not.

It affects downforce in a serious way, and just as importantly, it affects visibility for the driver...neither are trivial..

Yeah, and it would also make it much easier to tell how much frontal damage you have without going to an external view and looking at the front.

I think it's a good idea. When your car hit's something and the hood just bends in, it's like the hood isn't a hood; it's just welded on to the rest of the car.

1) The downforce isn't modeled due to damage yet
2) I see your point, and it'd be useful, but why the **** are you crashing into shit?
3) Nurburgring, just to bring up something equally as fun to argue about.

Quote from rjm1982 :

No, its not.

It affects downforce in a serious way, and just as importantly, it affects visibility for the driver...neither are trivial..

In real life maybe but as far as I know the aero model is not affected by damage at the moment.

Modelling this would also require modelling the engine underneath as it would become visible once bent up. I am sure once Scawen turns his attention to the damage model it will be considered.

Trouble with a good visible damage model is it tends to encourage the wreckers so, although I would like to see it sometime, I don't mind it being at the bottom of the S3 updates.

Maxim

Edit: Too late - dawesdust_12 types faster

Better believe it, and I had to wait for the spam timer to not tell me to go away too.

Fast fingers come from practice spamming, and out-witting Kev.

if your bonnet looks like that i doubt downforce would be the thing you worry about...

i think eye candy for crashes can wait till the actual damages are modeled

For the road cars at least they all have the same drag - hopefully each one will have different amounts of drag in the future, then this might be possible.

I think there are some threads like this. It's planed for S3, search next time

As people say the samage system and everything related to it (how it affects downforce etc etc) are beeing worked on. It probally will take some time, so I suggest we just wait and see when the DEV`s will surprise us

Quote from dude56 :

Whenever a car hits something, the hood will always bend upwards, seperately from the rest of the car.

LFS actually does do that, it is just very slight. But if you crash head on at just the right speed, and look at the hood, then you can see how it is slightly bent away from the rest of the body. If not, then I'm just crazy (I am anyway though).

Personally I'd rather see LFS model damage properly than with cheap effects and TBH the mesh deformation we've got now is already a million times better than a system that uses pre-made damaged 3d models.

Quote from Jakg :

For the road cars at least they all have the same drag - hopefully each one will have different amounts of drag in the future, then this might be possible.

Hey got a source for that?

Bob said it a while ago - i think he also said it was set to 0.35...?

Quote from Jakg :

Bob said it a while ago - i think he also said it was set to 0.35...?

I think it was about all the cars have the same drag coefficient (comes from the body shape) and drag area (the frontal area of the car going through the air) which form the aerodynamic drag that slows the the car down. What it means in practise (if I'm right) the LX4 and the FZ50, for example, have equal aerodynamic drag at sane speeds.

In reality it isn't like this: http://en.wikipedia.org/wiki/Automobile_drag_coefficients

Also notice that a Caterham Seven has drag coefficient of 0.7 to 1.1 while more normal cars like Peugeot 106 ('91) has only 0.34. This will have quite an effect on the high speeds of the cars in LFS...

Of course it gets harder when we add things like yaw, roll and pitch (and other magnificent things like ground effects, turbulence affecting downforce when following other cars, heat effects etc...)

Quote from Jakg :

Bob said it a while ago - i think he also said it was set to 0.35...?

Well I hope not because as has been shown in the Wikipedia link the range should be huge, the FZ50 probably lower than that the TBO cars around that mark and the LX should be double that. I know LFS's aero model is pretty basic but I can't believe the cars are all assumed to have the same frontal area and drag co-efficient or at least have some kind of overall drag value based on those factors which would be different for each car.

Sadly, that is the current limitation. Hence the LXs go crazy fast for a car of that style. It's one major limitation I'm hoping will be removed for S2 final. The value stored is 0.4 for all of the road cars, but does vary with the racecars (0.3 in the MRT5 for example, that's why the MRT5 slipstreamed like crazy back in 0.3H), it refers to 0.5 * Cd * A * rho (air density), so the value can be simply multipied by the speed squared to get the drag force.

For completeness, here's the list, including converted CdA values:

Scawen has said the system would be improved when the models are finished, he can create a front on image, pixel count it to get the frontal area, and use a sensible estimate for coefficient of drag.

He needs to change his thinking and base it on the model itself, not static numbers.

X-Plane (flight sim) determines all of the aerodynamic properties of the aircraft based on it's model. It works on a point weighting system to determine sharpness (the physics of a wing is smooth even though the model is less smooth due to LOD issues...but you can tell points to be weighted more and maintain more of their squareness)

This system works live in X-Plane, and on top of that, X-Plain is a single man/small team of support project as well, similar to LFS. So it could definitely be done.

The affect this would have would be great, as it would open up alot more features automatically. Damage modeling affecting the aero would be automatic... allowing for custom cars would be easier, it could figure out the properties on the fly...etc.

Apparently, its not too cpu intensive because X-Plane is very very low on teh cpu hit... i run it on my laptop all the time... and it makes things so much better, like if you over stress the airframe and lose a control surface, no only does your control go away, but the effect of extra drag and reduced lift is felt as well, and its all automatic...the developer doesnt have to think about what happens when a piece falls of, it just works

LFS could re-calculate aerodynamics based on damage without being CPU intensive. Maybe if it was only recalculated when there is major damage inflicted on the body. For example, if you tap someone's bumper, and it bends in a little bit, LFS would not recalculate the aerodynamic properties. But, if there is major damage, then the aero dynamic properties would be caluclated but at a lower priority then all the other calculations that are more nessasary. I'd rather have some other things added before damage affects aerodynamic properties.

CPU limitations could be helped by making it run on a 2nd thread, although it would be a pain for single core users, and it would be fairly hard to implement.

Quote from rjm1982 :

Apparently, its not too cpu intensive because X-Plane is very very low on teh cpu hit... i run it on my laptop all the time...

Xplane doesn't have to calculate the complicated physics of tires/suspension/drivetrain/engine for 20 cars and having AI drivers drive those cars while it does damage calculations. I just don't think LFS has the spare CPU cycles to do something like that. Not with the current CPU requirements and userbase anyway.

And like others have said, of you bend your bodywork like that, aerodynamics are likely the least of your worries.

Generally speaking, physics calculations are usually pretty easy for a CPU to eat through, according to Scawen they are not the main reason why LFS is CPU intensive. It is the ideal route to proceed along at some point, and I think would be a first for a race-sim. The quesion is, is it time yet?

I know this is off topic, but just need to ask - why is LFS only depending, or at least it seems, to be depending on CPU?
Why doesn`t it use the graphic card or something?
(Mainly because that`s the only good thing I got in my computer )

Quote from The Very End :

Why doesn`t it use the graphic card or something?

The graphics card is for graphics calculations. That's why.