What's new
What's new

CMM Data Filtering

allloutmx

Titanium
Joined
Mar 6, 2013
Location
Rochester, NY
Good Day All-
There comes a point where probe stylus/point spacing/data filtering will interpret your dimensional result. Excluding surface finish defects/waviness from your dimensional measurement is something worth considering. Im am not an expert in this field and wondering if some one out there would care to elaborate on this specific subject. Data filtering specifically throws me for a loop. I dont understand how to apply a filter appropriately. How does this relate to probe stylus diameter? obviously the size of your probe tip will filter some of your surface imperfections so maybe there is a point where considering the appropriate stylus diameter is important.

thanks in advance
 
I filter out data on my Zeiss CMM with active scanning head and on my Mitutoyo roundness measuring machine. Filtering is needed when you have a lot of data, let's say: more than 250 points per revolution.

You filter out the high frequencies, which is roughness. A larger probe filters out more roughness; a smaller probe will semse more roughness. There are standards to what filters you should use. For measuring diameters, it depends on the ratio between probe diameter and workpiece diameter. Under a 1:50 ratio, you filter out everything higher than 150 undulations per revolution (upr). Under a 1:15 ratio you filter out everything higher than 50 upr. The exact numbers depend on which standard you choose to follow. The same applies to measuring lines and planes. In that case you don't have upr, but wavelengths.

Here's an interesting link about the topic:
Tips for Choosing Filter Settings When Measuring Round Shapes |


Modern Machine Shop
 
Its a tough subject. The article notes 'standards'. I wish they would list the standards. I am not aware of any official standards. You are correct that the diameter of the probe can have a self filtering or 'Mechanical' filtering effect. Also, Zeiss in their instructions list different UPR for the size of the hole. So, smaller hole, lower UPR. For the degree mentioned in the article, there are 7 points. So, you need say 7 X 50 UPR points or 350 points to use a 50 UPR. So, I use 360 points. I also take a 3 sigma filter, to take off the fliers. Not mentioned is Tchevychev and Least Squares. Then there is Inscribed, Circumscribed and Outside Tangent. I use a 3 Sigma filter and Outside Tangent on planes. Other filters, I think you can filter down to much, so you always get parts in spec.
 








 
Back
Top