INTERPOLATE GRAPHIC DISPLACEMENT

[cobayatron]

Alright, let's see:

Suppose we use 4 accelerometers (locations x1,y1;x2,y2;x3,y3;x4,y4) where we read deformations (z1,z2,z3,z4). The wing is defined with a 50x50 nodes mesh. When I use the interpolation function in this VI I have:

Inputs:

Z vector: z1,z2,z3,z4,0....0 (the deflections at the accelerometers + 50 "0" reflecting the deflection of the x=0)

X vector: x1*,x2*,x3*,x4*,0....0 (the x coordinate of the nodes closer to the accelerometers + 50 "0" of the x=0 nodes used)

Y vector: y1*,y2*,y3*,y4*, y11...y501 ( the y coordinate of the nodes closer to the accelerometers + 50 y coordinates of the x=0 nodes used)

X Matrix: All the X coordinate of the 50x50 mesh nodes

Y Matrix: All the Y coordinate of the 50x50 mesh nodes

So I was expecting the interpolation function calculates the deflection in the 50x50 mesh nodes but if a node has a specified deflection defined in the X,Y,Z vectors it would be respected. If you run the VI, you'll see that the x=0 nodes have no z=0 deflection. Only when the method is set to "nearest" that line is correct. So my question is, how can I obtain a better representation than "nearest" but still respect the deformations I input?

Hope this clarifies a little.

[LocalDSP]

We are missing Wing Geometry parameter values on your VI front panel. Any chance you could add these and re-post the VI? I made a wild guess using the following data

Then I can see as you mention that the specific nodes reproduce correctly. 

Using 'nearest' the specific accelerometer position X=2.5, Y=0.5 gives you deformation z =1

Likewise the corner point 0,0 has zero deflection

The values are the same when using Cubic Spline, though you get significant overshoot (cubic spline may not be the best choice when you only have few measurement points). So I am not sure what you you mean when you write "you'll see that the x=0 nodes have no z=0 deflection".

Bicubic gives you a smoother plot (here shown without the mesh) but is has limitations near boundaries. For the wing length of 3 it doesn't even cover the accelerometer positions at X =2.5

[cobayatron]

Thanks for your response. Well I guess, it works then with that combination of parameters.. I'm trying with the actual values  (2.5 length, 0.5 chord root, 1 degree TE; 1 degree LE; different mesh sizes) and the x=0 condition is not respected unless I set the method to "nearer". I find this quite weird to be honest..