Predicting drying time (similulating physical systems)?

Question

ACM Question:

Simulations involving service times and queues are fairly routine despite using statistical distributions to generate data of interest. Most physical systems are far more complex.

Given

• Weight of a Towel
• Surface Area of Towel
• Dryer Volume
• Dryer Max Temperature
• Dryer Revolutions/Minute
• Boiling Point of Water
• Weight of the Wet Towel
• Specific Heat of Water
• Enthalpy of Vaporization
• Towel Only Item Dried

Predict

• Minimum seconds s, where Towel Weight(s) <= Weight of the Towel + 2%

Is this information sufficient? At a minimum the relative humidity, outside air temperature, temperature of the towel, heat loss of the system, and air flow would be needed. Physical systems are more accurately predicted using empirical observations that simulating their behavior. I could see the surface area and use of all of the other variables being important. How would you incorporate the Revolutions/Minute factor? If you were to design such a simulation, what would it look like?

Answer 1

In the past, I used neural networks to solve problems which cannot be solved by purely emperical evidence. Anytime you see some nonlinearity, it is best to model it using neural networks.

This could be a good starting point for you: http://documents.wolfram.com/applications/neuralnetworks/NeuralNetworkTheory/2.6.0.html http://en.wikipedia.org/wiki/Neural_network

But you can do some more research in wikipedia and get to the right resources.

Answer 2

This information is insufficient. I would suggest that you need some information about the rate of evaporation of water (below boiling point). Additionally, I would suggest that you would need information about the power of the dryer (relating the the speed which it takes for it to heat up to maximum temperature).

Answer 3

This problem immediately suggests a differential equation. It's a lot like the water drop evaporation problem:

http://www.math.lsa.umich.edu/~rauch/256/07HW1sols.pdf

Instead of a sphere you have the surface area of the towel. And for the rate of evaporation use change of the weight of the water, not dV/dt as shown. Then your simulation could be a time plot of water weight, and you plug in different settings for estimated conditions.