IJK values not changing with alignment

[DaveR]

Hi All,

I'm making some measurements on a cubic block that I move around to center it on an optical beam. The basic process is to align a set of axes to the block so that I can make measurements in the frame of the block. I then create a generic point with XYZIJK of 0,0,0,0,1,0, revert back to the original (machine default) alignment and output the XYZIJK of my generic point to a file. This should be the position of the corner of the block in the default alignment.
Thus when I move the block around and repeat the process I should get different XYZIJK values depending on where the block is and at what angle it is to the machine axes. The first part of this works fine, and the XYZ values seem sensible, but the IJK values are always virtually the same at ~(0,0,1)* even when the block is at ~25 degrees to the machine axes. I feel I must be misunderstanding something but at the moment I can't see what.

Can anyone explain where I am going wrong in all this please?

Dave


* Yes there is an axis rotation involved so going from IJK of 0,1,0 to 0,0,1 is OK provided the block's axes are parallel to the machine's axes

[Paul Barriage]

Chances of doing measurements without SOME kind of alignment are not really good.

Machine alignment is always in the temporary mode, looking for something to align to. Its function to actually measure something is nowhere close to anything with alignment.

[bw_bob]

From my understanding of generic points your IJK will never change. You will have to change them IE generic ( you created point not machine)

[DaveR]

Paul, I should perhaps have given more detail.

What I am doing is positioning the block with respect to a laser beam, measuring where the block is, then moving it along the laser beam and repeating the process. Thus the generic points I generate will all be at 0,0,0 in the alignment of the block, but will be at different positions wrt the default machine axes. If I change to the default alignment and then output the XYZIJK coordinates of the generic point to a file I get numbers like
390.5209,174.3784,-425.367,0.003,-0.0021,1
I can then reposition the block, repeat the process, and get
386.8862,174.294,-425.3847,0.003,-0.0021,1
etc.

I can then read these points into a separate program and e.g. generate a best fit straight line through the points and see how well they all fit the line.
So I don't really care what alignment I am using to define the XYZs for the points, as long as it is the same alignment each time.

What surprised me was that when I looked at the output file, the XYZ coordinates changed as I expected when I moved the block, but the IJK values did not, even when I deliberately rotated the block wrt the machine axes.

[cmmguy]

I dont think points per say have vectors. I dont mean points on a surface, but just points in general. I dont know why there is an IJK value associated with a point. In geometry, a point is only X,Y,Z