@tomleslie Thank you Tom for walking me thourgh this. It saved me a lot of confusion!

Haley

@Carl Love 

Hi Carl,

The new title sort of fits my question, but I'm mostly interested in finding the parametric expression of the surface area in terms of spatial derivatives of MF. So, my question is simply, when I have x = f(z,y), given MF (x,y,z) = constant, how can I define the surface area of x over a rectangle projection on (y0,z0) to (y1,z1), in terms of spatial derivatives of MF or grad(MF).

P.S.: MF = constant(s) form surfaces with some perturbations on them.  You can think of them as multi-mode sinusoidal perturbed planes

@Carl Love 

Thanks a lot Carl for all the inforamtion and help.  I tried to fit another model in the form of  a*(((x/b)-1)/c+1)^c

I givr this equation to matlab curvefit, but because of the imaginary parts generated in the begining, it retens an error. Is it possible to model a regression method in this form as well?

@Carl Love

@Markiyan Hirnyk

The reason for getting the first few data points as outliers is physical, we are not expecting the flow to behave as a power law model initially.

@Markiyan Hirnyk 

As I am using Maple on our univeristy's network, I don't have access to install new applications, that needs to go though a formal request procedure. Therefore I haven't tried this nethod yet. I will do that and let you know about the results. I appreciate you gave me a script for download.

Haley

@Markiyan Hirnyk 

Thanks a lot Markyan for the script. Apparently you have written your code in a newr version than 18.

@Carl Love 

Thanks a lot Carl. It works perfectly and I'm able to play around with the script to get my desired results. I omited more data points in the begining of the array, so your model fits exactly to the data at the end point.

Haley

@Carl Love 

As I am very new to Maple, (I used to use matlab), I try to run the script but it does not return any answers. I have attached an image of my worksheet for your consideration. I'd appreciate it if you could please help me to resolve this issue. I am using Maple 18.

Thanks,

Haley

@Carl Love 

Thank you very much Carl, I have solved this problem from another method, by taking time derivatives of W (I had to manually remove outliers at each step .), so I know what range of values to expect for each parameter. The values that I have got are A  =2.48, b = 1.1e-4 and c = 3.5. The model doesn't need to fit the data in the beginning.  I have uploaded my matrix as the file WT_1.

Thanks,

WT_1.xlsx