Inaccuracy in obtaining the Absorption coefficient

Hello all,

I have an equation for attenuation of sound underwater. The expected plot of attenuation versus frequency is as shown below

But the obtained result is as below

clc

f= 0:1:1000;

xa = f.^2;
fa = (((0.11*(xa))/(1+(xa))) + ((44*(xa))/(4100+(xa))) + (2.75*(xa)/10000) + 0.003);

plot (f,fa);

The matlab code is as shown above.

For f=0, the plotted result should have been zero ( or 0.03 which is close to zero) as in first image. But the plotted value is somewhat near 2000.

Why am I not getting proper result?

Answer 1

[Roger Stafford]

In your expression for 'fa' you need to use dots in the division operations on vectors:

fa = (((0.11*(xa))./(1+(xa))) + ((44*(xa))./(4100+(xa))) + (2.75*(xa)/10000) + 0.003);

Otherwise it will give you fallacious results.

[Roger Stafford]

The plot results you show are too large by a factor of about fifty from what your formula would produce according to the code you describe. You must have taken a step in the code which you don't show.

However, if you make the "dots" correction, then the plot result from your expression would agree quite well with the "expected" plot.

[Pramod Bharadwaj]

Thank you Roger. It solved my problem.

Answer 2

[Walter Roberson]

What is the point of calculating ga, x, y, and so on, if you are not going to use them? You plot f against fa, and ignore everything else you calculated.

Where is your conversion to decibels?

Why is your x axis showing up to 1000 when you are plotting f on the x axis and f is at most 100?

[Pramod Bharadwaj]

Actually, the code put up here was a part of big program. I have edited the post. Why am I not getting the value 0 or 0.003 (which is also close to zero) for f = 0?