Destruction Derby

[IchI]

I was thinking about the Havok Physics engine inside of hl2 and it got me thinking. Could you make a vehicle from over 100 peaces and put them alltogether. Just like a real vehicle (Obviasly not as complex).

What I mean is split the vehicle up into sections first of all you have a metal frame like any car. They you stick all the body parts on the sides. What I mean by stick these on is program them to have areas in which they tie too. Then if this area is damaged the seem breaks and it will proberly fall off or hang on by another thread. This way you pritty much have a 50% realistic car. Most other car damage games have premade damage animations and other types of things. Could you not just get rid of this completely and run of pritty much 100% physics.

Also if I am correct the Havok Source Physics engine can handle the bend and minipulation of geomtery. This means if the front of your car was slashed from the left hand side it would bend your axil and maybe even bend your whole car.

With all these physical objects how would the CPU cope. I was thinking if you have about 20/32 of these vehicles all in a round arena (is not complex and would take about 30minutes to map, proberly consist of about 10,000 triangles , i think u get my picture)

All of this obviasly has to be done. How much programming do you think it would seriously take? I mean some people might say the vehicle and damage system would take ages. But if you think about it. Once you have that done the mod would be pritty much finished. If you wanted a race system it wouldn't exactly be the hardest programming in the world. So how advanced would you say this programming would be?

But then you also get a problem with the LOD system, mainly because you have objects which are been bent and buckled. How would you LOD these. since all LOD is prerendered? Maybe LOD would have to be taken out of vehicles and other objects.

I was kind of thinking a mod like this could be pritty much made by 6 people.

Mapping/Modeller
Programmer
Skinner
Audio
Concept
Graphics

If you wondering why I am thinking about this its because The makers of the original Destruction Derby have failed ever since the 2nd game. Alltogether I think they have pritty much failed 4 times now. The games just don't have the style and class of Destruction Derby 2 or 1. Also I have never seen a mod pull of a Destruction mod. Not once...

Why has this area not been filled? Hopfully my questions above my put me in the right place (It been too difficult etc...)

 

[theGreenBunny]

AFAIK: The Source physics are based on rigid bodies, so stuff can't bend. That's one of the limitations of the physics system they used. As a large part of real car crash damage is the bending, it probably won't look that good, unless you model in a whole lot of different models with different levels of bending or something.

I'm not 100% sure on this though. If it can be done it'd be fun.

 

[Computer5k]

I think it could bend. You'd just have to give it a skeleton with joints at all the spots you want to be bendable and ragdoll it with a lot of resistance.

 

[James Isaac]

That sounds like a cool idea. If you went ahead with it, I would like to help with this project, as destuction derby was one of my favourite games. But there's not much i can do.

I am not good enough at programming to do it on the HL2 engine.
I can only make really basic maps/models
I haven't had any experience in skinning, audo or graphics.

I could do concept, but i don't think you need that

 

[IchI]

Hmmmm, I am not planning on starting this project right now. Especially with no SDK and no real idea of if its actually possible. If I did make it then it would only be small. Most of it would be program and it would only start off with 1 car and maybe 3 maps. I would do all the modelling and mapping my self and could proberly get away with no concept artist and just all have Shelby Mustangs GT 500's I can't do sound my self and no graphics. But if I could find a programmer/graphics person it could be handy.

Together it could be made with 3 people. When it comes to beta if we actually get it working and it seems all good. Maybe get a better team and push it forward. It would only be a small project though.

I have always had plans for my own small project I actually wanted to do a simple deathmatch mod by my self and get more experiance in more areas. But anyway, I need a little more information and capabilites before I start to do anything serious. I supose I could model a car in 3dsmax6 and use reactor to see if even 3dsmax can handle it. If it can't I know for a fact hl2 will not

The first thing I need to do is see if hl2 can handle at least 32 cars all with full on 16,000 triangles (LOD would be vertually impossible, or maybe not whatever ). So thats already 512,000 triangles along with the lvl which would at max be 30,000 60,000 triangles. So lets say you have 600,000 triangles on your screen at one time. Along with the CPU load. It will need alot more testing before it is more than an idea

If I know for a fact hl2 can handle 600,000 triangles at max at one time it wouldn't bother me that much, as long as the engine didn't crash I couldn't care You need to remember not everything will be on the screen at the same time. But its a good idea to see how high it could go before u start to plan seriously.

Anyway, i will hopfully get some more techniqual advice of some people. Keep it coming and thanks so far.

 

[ankalar]

Let's make an engine with finite element analysis.

 

[IchI]

Let's make an engine with finite element analysis.

what do you mean?

 

[Majestic XII]

Seen the game Flatout? Looks really sweet.

 

[IchI]

Ye I saw Flatout about 3/4 months ago with its exclusive in edge. It seems pritty dam kewl but I think its main pitfall will be the 8 players max. It also seems to lack the real overwelming destruction of Destruction Derby 1/2 mainly because the lack of 24 cars

If your wondering why iw as thinking about this idea when Flatout is just aroudn the corner its because I want to test the physics of hl2 and see how far the engine can be pushed on it. This is why I am saying if I did do anything it would still be a very small project. Only consisiting of 1 car to choose from. It would be more of a fun source test than a mod to beat CS. It just gives me the chance along with some other people to see how far we can go. Who cares if u get 2fps At least we can say it semi works...