[gameprogrammer] Re: Spring model

[grant hallman <unilogic@primus.ca>]

At 05:59 PM 04-04-04 -0500, you wrote:
>At 04:53 PM 2004-04-01, straver@apollo.nu wrote:
>>I'm playing with a spring model but I have run into a trouble.=20
>>The trouble is that I want a dampning effect on the springs so=20
>>they don't go all crazy but I don't want any damping effect on=20
>>the gravity movement for example.=20
>>=20
>>I can't think of any way to seperate them.=20
>
>As a recovering physicist, I can't resist replying. Have you tried just 
>introducing a "drag" force? Basically, just a force that is constant and 
>opposite of velocity. If velocity is zero, there is no drag. For more 
>complexity/realism, make the more proportional to velocity, or velocity to 
>1.5ish power (for air friction), or whatever.

Wouldn't it be more realistic to make the friction force proportional to
speed? Or for computational efficiency, just subtract a small percent of
any velocity, every frame.

grant


>I'll try modifying the code to show what I mean:
><snip>
>>/* apply forces */
>>for (int i = 0; i < (int)points.size(); i++) {
>>         point accel;
>
>// add drag
>point dragForce;
>
>// start out opposite to velocity
>dragForce.x = -points.at(i).veloc.x;
>dragForce.y = -points.at(i).veloc.y;
>
>double v = sqrt( dragForce.x * dragForce.x + dragForce.y * dragForce.y);
>
>if (v > DRAG_TOLERANCE)
>{
>// normalize
>dragForce.x /= v;
>dragForce.y /= v;
>
>// apply coefficient of drag
>const double u = 0.1;
>dragForce.x *= u;
>dragForce.y *= u;
>}
>
>>
>>         // acceleration
>>         accel.x = points.at(i).force.x / weight;
>>         accel.y = points.at(i).force.y / weight;
><snip>
>
>Peter Mikelsons
>plm@snow.org
>http://plm.snow.org/
>Snowbound
>
>
>
>
>
>