I had another thought while scrolling on SH. Again!

witch_sorrowful

Mmm, Monke.
Joined
Aug 26, 2020
Messages
142
Points
83
BUT two simulations with same initial u(0) will give you the same two "wrong" trajectory for u(k). When it doesn't, it's because of potato code or uncontrolled errors.
Only if it is inside the predictable domain. AFAI remember, predictability breaks down after reynolds numbers become a certain value in fluids at least.
In a mathematical standpoint, that statement is wrong. Chaos theory is the study of the range of outcomes that happen when you vary the initial conditions.
This is fair point. This i precisely what the theory is about. But, it also does say that at certain situations the variations in system given small perturbations is so large it is essentially unpredictable, isn't it?

Anywho, what I wanted to say was that Chaos Theory, Order Theory, and Paradox Theory.... went a little wild by the OP.
 

owotrucked

Chronic lecher masquerading as a writer
Joined
Feb 18, 2021
Messages
1,465
Points
153
Only if it is inside the predictable domain. AFAI remember, predictability breaks down after reynolds numbers become a certain value in fluids at least.

This is fair point. This i precisely what the theory is about. But, it also does say that at certain situations the variations in system given small perturbations is so large it is essentially unpredictable, isn't it?

Anywho, what I wanted to say was that Chaos Theory, Order Theory, and Paradox Theory.... went a little wild by the OP.

You're right about how Reynolds number give an idea about if the flow is laminar or turbulent. It doesn't say anything more than that itself.

"predictability breaks down after reynolds numbers become a certain value in fluids at least" that observation/statement must comes from another source, not the Reynolds number itself because it doesn't claim more than turbulent transition. I'm pretty sure it was an engineering tool for a bunch of scientists to design their pipe sizes and say "oh fuck, the flow is changing shape". I don't have any knowledge to bridge the logic that Re=>unpredictable. Only Re=>turbulent or laminar.

You're right about the fact that laminar flow can be simple enough to have analytical solution. However, engineers are making paintstaking effort to predict turbulent flows thanks to the advance in computer science, saying that turbulent flow just can't be predicted is the same as saying that they should drop their job and do something more useful lol.

"But, it also does say that at certain situations the variations in system given small perturbations is so large it is essentially unpredictable, isn't it?"

I agree for every sensitive systems considered in our physical real world that they are unpredictable from start to finish.

You could write about a fictitious universe where people can control initial condition infinitely perfectly and that doesn't stand true anymore. That world would be pretty easy to predict for the people inside.

Predictability is a spectrum too. It depends on the timeframe and how precise your inputs you are. You could linearize the system and have a close prediction of the real system in the next little timeframe. The more you're accurate in your model and inputs, the longer you can stretch the acceptable predictions.

I'd rephrase your sentence like this: "Chaos theory shows that real physical systems that are highly sensitive to perturbations/initial conditions are unpredictable in an infinite timeframe because those quantities cannot be infinitely controlled/measured."

It becomes compatible with the fact that weather can be forecast and why it's shit.

Functions/systems that can give two outputs with the same inputs aren't deterministic. I guess they exist but they aren't the target application of chaos theory.

Yeah agreed, OP went really wild lol
 
Top