As Paul states, all of the specifications for the HSA501K0J are given assuming it’s mounted to a 230 x 230 mm (9.1" x 9.1") chassis for a heat sink. If you have no heat sink, then it can only handle 20W. If you are somewhere in-between, then you would have to interpolate, which is not trivial.
Note that the graph showing power overload capabilities are assuming a 20°C ambient air temperature. So, if you have a heat sink of the specified size and you start with an ambient air temperature of 20°C, it will handle 250W for 5 seconds or 200W for about 7 seconds. However, if your ambient temperature is higher, the overload capability will be reduced. If the 195W peak you mention is only a small portion of 5 seconds, you may be OK, but otherwise, you are likely coming very close to its limits.
Another thing to remember is that you should be using either heat sink grease or a thermal interface material between the resistor and the chassis to allow good thermal conductivity. Otherwise, the tiny airgaps between the resistor and the chassis will significantly reduce thermal conductivity between the two.
If your ambient temperature will always be near typical room temperature (around 25°C) and you properly heat sink the resistor, it will likely work, based on the datasheet specifications. However, if you need to assure reliability, you will either need to perform testing, as Paul describes above, or give yourself some more headroom by upsizing your resistor to 100W or more.