Nested for loop problem

I am modelling seismic rays, (I am very new to Matlab, this is my first attempt at solving a problem), and I have to plot a series of seismic rays at various angles (any number I choose up to 20 degrees) that penetrate through 5 layers, and reflect at the boundary of the 4th and 5th layer. Through each layer, the seismic ray needs to refract. I have managed to achieve this with the code below.

HOWEVER, I need to plot a SERIES of rays, I have only managed one. I wanted a ray to propagate at 1 degree, increments of 5 degrees up to i=20 (the initial angle you see plotted). I want to use a nested for loop to achieve this, but I'm not sure where i'm going. Ultimately, I'm trying to achieve a single plot of several seismic rays with their reflections such as below, at various angles, with the maximum angle being 20.

A plot similar to this picture is what I'm trying to achieve; multiple rays and their reflections. The code for the reflections is

reflectionx=h(end); 
plot(2*reflectionx-h,depth) 
axis ij

%COMPUTE DISTANCE TRAVELLED BY THE RAY IN EACH LAYER  
clear  
%DEFINE VARIABLES  
z1 = 0;                      %Depth at y=0   
v = [3:7];                   %Velocities at each subsequent layer  
i = 20;                      %Takeoff Angle in degrees  
%COMPUTE DIAGONAL DISTANCE BY THE RAY IN EACH LAYER  
 for j = 1:5:i
for n = 1:5;                        %n = number of layers      
    v = v(n);
    i;  
    d=z1/cosd(i);                   %calculate diagonal distance
    h(n)=z1*sind(i)                 %calculate horizontal distance
    t(n) = d/v;
    v = [3:7];
    depth(n)=z1
    sini = ((sind(i)*(v+1))/v);
    i= asind(sini)
    z1=z1+2;               
    end
 end

HorizontalDistance=[cumsum(h)];
disp(HorizontalDistance);  

plot(h,depth)                  %Plot horizontal distance vs depth  
hold on  
reflectionx=h(end);  
plot(2*reflectionx-h,depth)    %Plot reflected ray   
axis ij                        %Places origin at upper left corner of plot

[JDilla]

Im still struggling with it

[Guillaume]

One piece of advice which won't help in solving your problem but will make your code much easier to comprehend: give some meaningful names to your variables. This is so much easier to comprehend:

depthy0 = 0;            %Depth at y=0
layervelocity = 3:7;  %Velocities at each subsequent layer
takeoffangle = 20;    %Takeoff Angle in degrees.
for angle = 1:5:takeoffangle
   for layer = 1:5
      velocity = layervelocity(layer);
      diagonaldistance = depthy0 / cosd(takeoffangle);
      %etc..

In particular, your variable named i could be confused for the imaginary unit.

[JDilla]

Thank you, I will make changes soon!

[Guillaume]

In addition, on the line

sini = ((sind(i)*(v+1))/v)

Did you really mean to do

 (v+1) / v

with v a vector? / (or mrdivive) in this case finds the least-square solution x to x*v = v+1 with v = [3 4 5 6 7].