how rotate a number as an vector in quiver plot?

Hi all, I am having a very strange issue with rotating vector along the circle that I cannot seem to debug - I am hoping fresh eyes will help. The problem is best described if I explain it as follows:

r=1; % radius of nodes in cm

n=9; % number of (equally spaced) nodes around the half-circumference

theta=-pi/2:pi/(n-1):+pi/2;% angular position from 12 o'clock

x=r*sin(theta)';

y=r*cos(theta)';

%%Type in modal matrix (i.e. mode shapes are columns)
% Dummy data for you to replace

w=[4.03,-2.75;4.8,-1.17;4.41,-1.5;4.29,-2.42;
3.86,-2.88;3.53,1.31;3.39,1.3;
3.29,1.46;4.56,1.7];

% Set in-plane components of motion to zero since they are unmeasured

u=zeros(size(w));
v=zeros(size(w));

% Type in the natural frequencies here

fn=[171,245];

for ii=1:length(fn)

    figure(ii)

    quiver(y,x,v(:,ii),w(:,ii))

    hold on

    plot(y,x,'-r','Linewidth',3) % overlay geometry

    title(['Mode shape of ', num2str(fn(ii)),' Hz mode'])

    xlabel('x (cm)')

    ylabel('y (cm)')

end

I want the vector be on the circumference of the circle but has an direction that likes passing from the centre of the circle instead of being perpendicular at each point to the circumference of the circle.

[Adam Barber]

Hey Paul,

If I understand correctly, you want arrows that point away from the curve on your plot.

In order to do so, you only need to define the "u" and "v" inputs to "quiver' such that they point in the direction you want. This turns out to be quite easy as you already have those directions as "y" and "x".

For example, if you replaced your call to "quiver" with the following:

quiver(y,x,y,x);

You will see the arrows point away from the center of the circle.

The documentation for "quiver" might be helpful here: http://www.mathworks.com/help/matlab/ref/quiver.html

One thing to note is that this does not change with "fn" or "w" because it is unclear to me what these are doing in the code. The only thing that changes with "fn" is "w", but since "v" is always zero, the vectors are only pointing up as that is the only component they have.

Hope this helps,

-Adam

[Paul]

Hi Adam many thanks for your comment, I have modified my code but still there is one problem on it, I wonder to have a vector on the circumferential of my circle and has a direction from the centre line of my circle. but in case just few of them are passing from the centre of my circle.

clc;clear ;close;

%% Geometry of accelerometer positions (in cm) - edit r and/or n if required

r=1; % radius of nodes in cm

n=9; % number of (equally spaced) nodes around the half-circumference

theta=-pi/2:pi/(n-1):+pi/2;

% angular position from 12 o'clock

x=r*sin(theta)'; y=r*cos(theta)';

%% Type in modal matrix (i.e. mode shapes are columns)

% Dummy data for you to replace

w=[-6.07,-1.74;-6.06,-0.212;-2.67,+3.24;-1.42,1.43; 3.22,0.983;3.96,0.66;6.37,-0.829; 6.64,-2.55;8.4,-3.27];

% Set in-plane components of motion to zero since they are unmeasured

% Type in the natural frequencies here

fn=[50,175];

%% Plot mode shapes as quiver plots - you should not need to edit this part

for ii=1:length(fn)

figure(ii)
quiver(x,y,x,w(:,ii))
hold on
plot(x,y,'-r','Linewidth',3) % overlay geometry
title(['Mode shape of ', num2str(fn(ii)),' Hz mode'])
xlabel('x (cm)')
ylabel('y (cm)')
end

[Paul]

No idea?