Scanning Reflective parts when you cannot use developer

[[Pa...]]

We have an ATOS Core 200 Scanner and use Zeiss Inspect Pro. I am trying to scan in an opaque ceramic part that has been optically ground to < 1/4 wave RMS. There are scratches in the part and we are trying to measure the depth of the scratches or somehow look at them via a defect analysis. I know this might not be possible with the scanner we have. 

The first step would be to just get a good scan of the part. It is basically a flat plate with some chamfers on it. Currently when I scan the part in I am getting only low quality triple scan points. I have scanning dots on the part and on the rotating platform. It is able to pick up 4 to 7 points for each scan. The room is dark with all the lights turned off (I have tried with the ambient lighting on as well).  I have tried using the shiny sheetmetal scanning template but it I am still getting low quality triple scan points. I cannot use developer on the part. Does anyone have any suggestions?

[[Ja...]]

Hi Paul,

For scratches or similar this small then 3DSL in general is not capable for determining the depth.  You are in Alicona type territory.

The other aspect to consider is that ceramic is a peculiar material, when it seems opaque to your eyes it might not be to the sensor as the light can penetrate it. 

Low quality triple scan points can come up for other reasons , for example if the percentage of the scan area you are interested in is small in relation to the measurement volume itself , it does not signify that your scan is 'bad'

I hope this helps in some way

Best regards

James

 

[[Pa...]]

Please sign in to view this username.

 Thank you for the information. You do bring up a good point about the transparency of the ceramic at different wavelengths. However at the wavelengths this machine puts out, and this particular material,  it should still be opaque to the scanner. 

 

I am scanning the whole part so the scan area is larger than the part itself. I would say the part is fills about 75% of the scan area. The reason I brought up the low quality triple scan points is that after I generate the mesh there are a lot of defects in the mesh that are not there in real life. In addition the surface looks rough as compared to the actual physical surface. We cant use a profilometer on this part either. While the Alicona would be nice I dont think I would get approval to purchase that for these parts lol. Thanks again for your help.

[[Ja...]]

Hi Paul,

It isnt related to the wavelength of the light, it is related to perceived contrast.

The test i used to do is that within the web interface for the sensor it is possible to change the projection so it gives a static image of a fringe pattern, if you then hold a business card/paper across the part with the fringes on you may be able to discern that on the paper the fringes are sharp, but on the material it becomes more blurred , indicating sub surface effects on the fringe pattern.

Given what you say about the part being 75 percent , and this ts pointswarning flagging and with defects in the mesh i would think this sub surface effect is happening.  

Again you can check the triple scan point flag by scanning a plain piece of paper , if ok then nothing wrong with your sensor...but then does indicate that the information back from the fringes ( used to calculate the quality of ts points) on your material is sub optimal.

Given your restrictions i think id try to look for a local lab to see if they have a suitable microscope to attempt to get the information you are after.

Best regards

James

 

[[Pa...]]

Please sign in to view this username.

 Thanks for the info...it turns out that the core is bad. I went to calibrate it and I could it could not find an optimal position. After working with support the core is kaput.

 

-paul