midplane

[#2#]

I am measuring a plane on two oposing sides of a part. I want to create a plane centered between these two planes so I can measure how far from center certain features are located. When I create a midplane it appears correct. Just out of curiosity I dimensioned the distance from each side plane to the midplane perpendicular to the midplane. For one side I got .298 and for the other side I got .299. So, my midplane is actually not centered.

Any ideas what is going on? All I can think of is that the side planes aren't parallel. But, I thought the centroids of the planes are used for calculations.

[Matthew D. Hoedeman]

Rounding error. Also, I think it takes the mid-points of the individual points, then constructs a plane from those, not a mid of the planes. So, if the points are not quite at the same locations on both sides, the center point of each plane will be slightly different, so that could also help with the difference you are seeing.

Try it using single points on each side, construct mid-points from those points and construct a plane from those and see what you get.

[#2#]

These are three point planes. So maybe what it is doing is taking the center point of a line drawn from the centroid of each plane (not perpendicular to either side plane). Then whey I dimension I am getting a perpendicular dimension.

I will try the single point method.

[#2#]

Try it using single points on each side, construct mid-points from those points and construct a plane from those and see what you get.

OK, I did this using the same x and y co-ordinates for the points. When I did it this way it got worse. One distance is now .3035 and the other is .2970!

Any other suggestions?

[Wes Cisco]

Unless your surfaces are tiny, I would try taking 5 or more hits for each plane, then try constructing the midplane. I think the three point planes may be skewing your vectors, and 5 might help reduce that. Also as a check, I would dimension the parallelism of the two planes before constructing the midplane. If they are not parallel, your midplane will be funky. HTH

[Matthew D. Hoedeman]

Can the work-plane be having an effect on the constructed mid-points/planes?

[#2#]

Can the work-plane be having an effect on the constructed mid-points/planes?

Good question. What workplane should I be in? The vectors of my measured points are 0,0,1 and 0,0,-1.

[#2#]

OK, so I have three points on the top surface, three on the bottom surface. Then I created three more at the midpoints of the pairs (top and bottom same x,y) Then I created a plane using the three midpoints.

If I dimension the distance of each point to the plane pependicular to the plane they should pair up (same x,y)and be equal. They do not???

If I create a plane using the top points and a plane using the bottom points then dimension the distance from the top plane and bottom plane to the center plane perpendicular to the center plane I get the same value????

[Jeffrey Lovely]

I have had a tp2 go bad giving me different readings from hit taken on oneside to the other side

[tking]

.02

Level to the midplane and zero on it, create a line perp to the midplane thru 0, intersect this line with the first 2 planes, my guess the 2 intersect point will be the same distance apart from the midplane.

TK

[Scott Appleyard]

.02

Level to the midplane and zero on it, create a line perp to the midplane thru 0, intersect this line with the first 2 planes, my guess the 2 intersect point will be the same distance apart from the midplane.

TK

Along this line of thought....

I think what PC-DMIS might be doing is using the centroids of each plane for the midpoint and averaging the vectors of the input planes for the mid-plane.

[tking]

Along this line of thought....

I think what PC-DMIS might be doing is using the centroids of each plane for the midpoint and averaging the vectors of the input planes for the mid-plane.

Agreed

TK

[cmmguy]

I am measuring a plane on two oposing sides of a part. I want to create a plane centered between these two planes so I can measure how far from center certain features are located. When I create a midplane it appears correct. Just out of curiosity I dimensioned the distance from each side plane to the midplane perpendicular to the midplane. For one side I got .298 and for the other side I got .299. So, my midplane is actually not centered.

Any ideas what is going on? All I can think of is that the side planes aren't parallel. But, I thought the centroids of the planes are used for calculations.

Conclusion:
Dont use the midplane - it is unpredictable in the results. Find a different method. Determining the Midplane of two non-parallel planes is useless.

[Scirocco]

As I understand it the distances to plane goes from the weighted center of the plane. If the planes are not parallel this will cause the diffrence.

Have you tryed taking a single hit on both planes and then look at the distance to the midplane?

Regards,

HH