I'm aware of this, but asking costs (next to) nothing.
Yes, of course you are right. And if we DON'T ask, it could be they won't tell us... ever.
A few years ago the surface engineering laboratory in which I worked was applying a micrometer-thin film of a proprietary coating to the core pins of injection molding dies used to manufacture automobile engine aluminum cylinder heads... the big metal thingy on the top of an internal combustion engine that seals the tops of the piston cylinders and contains the spark-plug and intake and exhaust valves. Cylinder heads were traditionally made of cast steel, but aluminum has its advantages, being lighter and having better heat transfer characteristics than steel, among other things. Core pins are used to create small cylindrical holes in the cylinder heads, holes that are typically used to pass bolts through when the head is assembled to the engine block. The core pins are made of a highly polished steel alloy and pass through both halves of the die to create the holes that will appear in the injection-molded aluminum.
Problem is, steel and aluminum don't play well together. The aluminum tends to dissolve and adhere to the surface of the steel. This can cause a big problem when it is time to separate the molding die into its two component parts. If the core pin is coated with aluminum after the mold cools enough to be opened, it can prevent the molding die from separating. This shuts down the production line until the stuck die pin can be removed from the die and replaced. The proprietary coating is designed to prevent the molten aluminum from adhering to the steel core pin. Naturally, the customer didn't want to just take our word for it that our (moderately) expensive proprietary coating would increase the up-time of their aluminum injection molding process and literally pay for itself. Their "bean counters" wanted some experimental evidence, and our surface engineering laboratory wanted to know how well certain variations of our proprietary coating lasted. So... I was tasked with the design and construction of a machine to perform a test.
A sample steel rod of the same general diameter and exactly the same alloy and polished surface finish as a real core pin was inserted into a collet mounted to a vertical linear slide, which in turn was mounted to a longer horizontal linear slide. Both slides were connected to pneumatically actuated pistons. At one end of the horizontal slide was an electrically heated oven with a crucible of molten aluminum. At the other end of the horizontal slide was a large open container of water. The vertical slide was used to dip the sample steel rod into the molten aluminum for a certain period of time and then withdraw the rod. The horizontal slide then positioned the sample steel rod over the container of water. The vertical slide was then used to dip the now hot sample steel rod into the container of water for another certain period of time, to quickly cool it, and then to withdraw the rod. The horizontal slide then positioned the sample steel rod over the aluminum crucible and the process just described was repeated.
All of the parts for this test apparatus were "commercial off-the-shelf" (COTS) components. I designed and built a cylindrical furnace to heat the crucible that held the molten aluminum. I tried using a steel crucible initially, but the molten aluminum dissolved the bottom out of it, creating somewhat of a mess. So I switched to a ceramic crucible and everything worked fine after that. The pneumatically actuated linear slides used Hall-effect sensors to detect the limit positions, left/right and up/down. An inexpensive, French, programmable logic controller (PLC) implemented the test algorithm, and a PC tied it all together by acting as the data acquisition system and programming device for the PLC. I built a "spiffy" control panel with a COTS temperature controller for the oven and manual switches to allow the linear slides to be tested. The switches didn't actually control any hardware: they served as binary inputs to the PLC whose program decided how to respond to switch actuation.
I didn't have any experience building electrically heated furnaces, but I did know two things: the furnace needed to be well-insulated if there was to be any hope of reaching molten aluminum temperatures, and the outside of the furnace needed to be cool to the touch for safe operation. To solve the insulation problem I ordered a high-temperature insulating material that could be cut and shaped with a saw. We made annular rings and stacked them up around the outside of the oven heating elements. Then I made a coaxial cylinder around the outside of these rings, attached to the aluminum plate bottom of the apparatus. Slots machined in the bottom plate allowed air to flow into a plenum mounted below the bottom plate, exhausted by a squirrel-cage fan. Thus cool air would flow into the top of the coaxial cylinder (which formed the outer wall of the furnace) and be drawn into the plenum by the fan and exhausted underneath the apparatus to the room. All the furnace components, including the Type K thermocouple and temperature controller, were COTS items available from Omega Engineering. Not exactly cheap, but readily available and reliable.
So here's hoping that
@tnnelectro has as much success and fun building their stainless steel "welding pit" or "welding crucible" or whatever it is he or she is trying to do as I did building my little molten aluminum test rig. As for how much power is needed to reach any given temperature... that depends on how much power is wasted in radiation, conductive, and convective heat losses. It would be a good idea to learn some thermodynamics, or as
@(*steve*) says: it's not just a good idea, it's the law!
