We have a DataFrame with Transaction Date column which is timestamp.
When we write the DF as ORC files we applied the partition logic on Transaction Date value ( not timestamp only date value), we created a separate field only for applying partition on that field.
If we read the ORC files again with where condition as Transaction Date(timestamp) value, will it prune the partitions?
No. You need to reference the "separate" field appropriately. It stands to reason and is a fundamental DB rule wrt partition pruning.