Houdini - Hard Impact

This is my second dive into rigids, this time working more with creating breaking and bending metal. It was also a great opportunity to learn how to use and implement simple VEX coding.


My third and final project for my portfolio period at TGA; Hard Impact was the most education of the three. Evolving from being a dream of something different, a yearning to not work in Houdini. Instead i became the most Houdini focused project of them all. From the start plagued by a lack of time, nature felt the need to do to me, what Peter Gibbons does to the office printer in Office Space. With three weeks of the flu to ponder and plan, I returned with a vengeance. Opting to blaze down the path I had already began, and dive deeper into Houdini.

0065.png

Building upon the lessons learnt from the previous project, I set out with efficiency as my goal. It burning like a guiding start as nodes landed like bricks in a puzzle, each bringing me closer to my goal. The road to the final product was anything but straight. More than often I had to dive into Vex to solve, what in hindsight, were simple problems.

This video is a compilation of the flipbooks I made from start to near finish of the project. It shows some of the sometimes funny oddities that occurred while setting up both the fracture and simulation. And when troubleshooting things that did not look like they should.

fractureloop.png
soft_deform.png
constraints.png

Breakdown

The setup I made can both create animated destruction sequences, and static destroyed meshes with only the smallest of adjustments. All the while retaining a small memory footprint and being easy to adapt to various uses.

At its core this project was about two specific things; getting a basic understanding of Vex and broadening my knowledge of rigid-body simulations. This meant relying less of the many useful nodes in Houdini and more on what I could make myself.

Because of how Alembic stores paths, it was easy to split the mesh up and define specific properties for specific parts before sending them through the fracture loop. This also made it much easier to iterate and fine tune the results of the simulation to match the desired outcome.

Compared to the Derelict Impact project where I made use of four separate meshes, this method of splitting the mesh apart was far more effective. It also meant that instead of four different fracture loops. I only needed one, as all fracture parameters were set beforehand, and not in the loop.

The next step was creating the many constraints needed to hold the now shattered mesh together. Because of what I wanted to achieve this required more than simple glue constraints. My goal was to have various forms of breakup. That required me to have constraints that could deform and break in various ways. To do this I again made use of some rather simple VEX Expressions to allow me to set properties in an accessible and efficient way.

The final step to creating the effect I wanted, of a bending metal-like deformation demanded one last step. With the hard constraints set up so they did not shatter, I could bind the parts of the original mesh that should only deform, to the fragments I marked as deformable.

The current iteration video was rendered in Blender Cycles. Currently this project is on hold as I cannot bake out updated alembics on the apprentice licence of Houdini.