/ck/ - Food & Cooking

>Not wanting crispy baked mac with a bread crumb top

>>11591105
No OP, this will not do. The soup forms a compact set in R^3, since it is closed (due to gravity) and bounded (since it fits in the bowl). Also, the function determining soup temperature is clearly continuous in the spatial dimensions. Therefore some spoon of the soup attains the infimum of the range of the temperature function, and the so-called "coldest part of the soup," i.e. a value at which this minimum is attained, is well defined.
Now, to locate this point, we can employ the classical heat partial differential equation which simulates diffusion in a set. Here, we'll use the boundary condition that the bowl begins at a constant room temperature, as does the air. We'll also need an initial condition - here, the temperature of the soup at each point at time t=0 will be determined by its position in the pot pre-pouring.
Propagating the system forward by convolving the heat kernel with the initial condition and normalizing for boundary conditions, we then have, for each time, a function of soup temperature parameterized by distance. Now we can apply a convex optimization method to locate a suitable local minimum. We'd expect it to be near the surface, since that's where the soup gets the most exposure to the air, so we'll pick k points on the surface using a poisson disc-sampling stochastic algorithm (don't wanna pick points too close to one another!) and find local minima near each sample.
This is a much more rigorous and developed plan of action than you suggest. I recommend it because you indicated that you'd like to eat from the coldest point of the soup first, but I'm not sure how you're able to find that to any degree of accuracy without a mathematical proof that the point you've found is indeed the coldest.