How baffles model rack leakage at two levels of detail in a hot aisle containment pod, and how cloning and grouping lay out a large room quickly.
This video builds a hot aisle containment pod and models rack leakage with baffles at two levels of detail. It then shows how cloning and grouping replicate the pod across a large room.
What the video shows
- Lay out two rows of ten racks with a small gap at the end and underneath, and assume no leakage between racks.
- Level one: add in-room baffles around the pod as containment, set to a percent leakage, and clone them to form the hot aisle. Leakage is turned off on the end baffles, since leakage tends to occur along the rack rails and where servers meet the containment.
- Level two: lay a second, smaller baffle over the first and name it rack rail leakage. Size it to leave a gap above, below and beside the servers, which gives a closer picture of perimeter leakage. Merged baffles need different colors to stay visible.
- Add ceiling grills with a high percent open area, group the pod, and clone it with an offset to build many pods. The presenter reports 12 pods, 240 racks and 2.4 megawatts in a 10,000 square foot room, built in a few minutes.
- Level three: split each rack row into individual racks the width of a server, with gaps between them, then clone those the same way.
- Edit parameters, such as rack rail leakage or rack loads, from the hyperlinks in the pre-simulation report.
Tips
- Name objects such as baffles so they are easy to find in the pre-simulation report.
- Use whole numbers or one decimal place where possible, and clone the first pod rather than rebuilding it, so everything stays aligned and the mesh does not fill with cells from small offsets.