Why is CSIRO, a well-regarded Australian Government research organization, doing this? For some rich collector?
Making duplicates of broken or worn out parts is a standard industrial operation. Here's a laser scanning service that measures out parts.[1] Here's a mechanical coordinate measuring machine measuring out an engine block.[2]
Here's a Kuka industrial robot machining a sand mold for a one-off job.[3] That's somewhat unusual, because you only get one part per sand mold.
If you're making more than one, you first make a pattern, usually from wood, that's then packed with sand. Then you remove the pattern and pour in hot metal. The pattern has to be adjusted in size a bit to compensate for shrinkage as the metal cools. Standard foundry practice, centuries old. After casting, you do some finish machining on the surfaces that matter, and leave the rest as cast.
There's commercial software for all this. Autodesk Moldflow, for example. This is not an R&D project. It's the sort of thing someone would do in a good manufacturing technology 2-year degree program. I've seen people do this kind of thing at TechShop, although in a smaller scale than an engine block.
Jay Leno (serious car buff) has been doing this for years. I think he and many other car collectors realized the potential of 3D printing of obsolete parts when the cost came down enough to be affordable.
Making duplicates of broken or worn out parts is a standard industrial operation. Here's a laser scanning service that measures out parts.[1] Here's a mechanical coordinate measuring machine measuring out an engine block.[2]
Here's a Kuka industrial robot machining a sand mold for a one-off job.[3] That's somewhat unusual, because you only get one part per sand mold.
If you're making more than one, you first make a pattern, usually from wood, that's then packed with sand. Then you remove the pattern and pour in hot metal. The pattern has to be adjusted in size a bit to compensate for shrinkage as the metal cools. Standard foundry practice, centuries old. After casting, you do some finish machining on the surfaces that matter, and leave the rest as cast.
There's commercial software for all this. Autodesk Moldflow, for example. This is not an R&D project. It's the sort of thing someone would do in a good manufacturing technology 2-year degree program. I've seen people do this kind of thing at TechShop, although in a smaller scale than an engine block.
[1] http://www.dirdim.com/serv_replication.htm [2] https://www.youtube.com/watch?v=8oiDK864Agk [3] https://www.youtube.com/watch?v=sjMNKYAYdFE