If you crash the schedule, what you want to see is your duration to reduce. That's the objective of crashing. The threat created is against both cost and quality, mostly cost, and for cost it is really imminent that costs will climb.
Also, in your chart, weekly crash cost can never be negative unless someone else paid your cost. The formula would be: crash cost / week or $9,000 / 5 = $1,800.
Of course, it should not be five but something less than four.
In reality, crashing your schedule can either reduce your duration, not touch it, or cause it to climb. Different tasks having different resource elasticity where adding people to do a task may do nothing or make things worse.
For example, moving boxes from point A to B should have a ton of resource elasticity and you should experience a reduction in duration unless the path where people are walking creates a bottleneck. Doing surgery on someone's brain would have no resource elasticity and likely make things worse. Adding surgeons inside the head would be catastrophic.
So to answer the question, why can't crashing time be greater or equal than normal time: It cannot because that's the objective of crashing is to reduce your duration. It can, however, due to many other drivers that affect duration but when that occurs your crashing intervention failed.