Jet Trails
For missile trails, the fx was first set up with a rough particle simulation to place an estimated pyro source for the primary simulation. The particles in the source simulation are assigned relevant attributes such as temperature, density, and velocity. Next, the particles are converted to volume fields, which are inputs for the pyro simulation. In this case, the pyro source is located at the back of the missile and has a velocity going in the opposite direction of the missile’s motion. The result is shown above where the initial stream is straight, but the trail becomes more defined after a while due to the various forces in the air at work.
Muzzle Flash
The base setup for the muzzle flash is an explosion setup in Houdini. I orient the explosion’s source volume to the cannon’s barrel and simulate the smoke out of the AA guns. I turned up the smoke’s dissipation to make real-life examples of a tank firing or an artillery gun.
Projectiles
Here, the projectiles of the AA guns are just fast-moving particles with a line inserted into the points. The line is then given thickness. I did not sharpen the two ends of the tube because the motion blur would do it for me during rendering.
Missile Animation
I would say this is one of the most complicated shots in this project because I have to layer many different simulations, and the different layers of FX also have to work together.
Firstly, I simulated the fall of the main missile. Using this animation, I input the missile animation into a rigid body simulation to get the fins breaking apart.
Next would be the animation of the smaller missiles hidden within the main missile. They are done by instancing the geometry of the smaller missiles to a custom particle simulation, which I configured to emit out of the main missile after the fins break apart.
Lastly, using the animated smaller missile geometry, I can use their animation path as a source for the smoke trails.
Explosion
The explosion for the last two shots was also relatively more complicated. It consists of two parts: the falling of the missiles and the explosion at the same time the missile touches the ground. The sourcing of the explosion needs to be consistent with the end of the missile dropping onto the ground. This would require me to code in a specific start frame attribute to the particles so that I can split the output into two groups. This problem stalled my progress for a few days before I could figure out the solution. Ultimately, the setup now works.
Debris
I originally intended for the last two shots to have debris flying off from the explosion centres. This would be a particle simulation with small, medium, and large rocks instanced on them. However, I decided not to add them in due to the tight deadline, as the shots were very fast. Adding them will give the overall scene more detail and realism but would add to the render times.