When you provision a machine with multiple disks from a vSphere template, the disks will not show up on the canvas or the grid because they are part of the template, and vRA is aware of it.
When you onboard a discovered machine with multiple attached disks, vRA does not know if the disks came from a vSphere template or have been attached later (somehow). Thus, vRA will onboard them as individual components, and they will be visible as Disk X in the grid and Disk_X_Machine_ABC on the topology view.
The cause for this is how vRA presents the data to users:
Disk X (vSphere) <- Disk X (vRA Provisioning Database) <- Disk_X_Machine_ABC (Catalog Service). In the grid, you see object names from the Provisioning service database (provisioning resources), and in the topology, you see names from the catalog service database (catalog resources). Catalog resources are wrappers around provisioning resources.