As the title states, I have on the way a collection of 10,30,50 ohm resistors that look similar to TO-220, would I need a heat sink for them?

They are rated at 35Watts.

 

Yes.

Think of it this way. You can melt solder with a 15W soldering iron.

 

Yes.

Think of it this way. You can melt solder with a 15W soldering iron.

Roger that, thanks. Good thing I also included Heat-sinks in the order.

 

In this case, it is clear that a TO-220 package can not get rid of 35 watts of heat without a heat sink.
But there is a lesson. All resistors can remain cooler with a heat sink attached. They can also be made worse if you add shrink tubing or confine them in a box with no air flow. Many high watt resistors arrive with a flat side and some have ears with bolt holes. You can get much better performance and/or reliability by adding a heat sink to most resistors. Sometimes it is nothing more than a larger area of copper on the circuit board. Sometimes it can be liquid cooling! In between those is air flow. It can be surprising how much heat a little bit of air flow can remove.

 

OK, I'll play too.

No, they do not need a heatsink, as long as the maximum dissipation times the thermal impedance from element to ambient is below the maximum element temperature.

 

IOW, keep the current below 0.2A and you don't need heatsinks.

 

When in doubt, check the datasheet.
It will tell you maximum operating temperature and its thermal resistance, case to ambient.

With those parameters, and the maximum expected ambient temperature, you can calculate the maximum power dissipation that won't damage the resistor.
You can exceed that power dissipation briefly. But for long term reliability, the recommendation is to give the circuit some margin, and derate the maximum power dissipation.
A 75% derating is a good rule of thumb.

 

Roger that, thanks. Good thing I also included Heat-sinks in the order.

From the datasheet of a 35W and 50W resistor in TO-220 package

http://riedon.com/media/pdf/PF2200.pdf