Although most of my novice experiments in 3D-printing have been just that, experiments intended to familiarize myself with the technology and its benefits and limitations, I have already had the opportunity to design and print something actually useful. Like most refrigerators, mine has storage areas on the door where items are secured by a removable or adjustable rail, keeping the mustard from flying when you yank the door open. These rails are hollow aluminum sections that accept plastic friction-fitted end caps, which in turn have hooks that engage slots on the door. One of the hooks broke off and naturally fell inside the door, so fixing it with glue was out of the question.
I modeled a replacement clip in Solidworks, taking measurements with calipers off of the broken end-cap and the intact one from the other end of the rail. The new model differs from the old in only one respect: I chose not to include the outside bevel, since it's only cosmetic and I intended that face of the part to be on the bottom of the print—including the bevel would require a support structure there that I didn't want to bother cleaning up. A raft was still necessary under the two hooks however.
After printing and removing the support material, the part was ready to be friction-fitted into the original rail. It works well; this is a good example of the kind of tricky little plastic part that's ideally suited to household 3D printing.