Alignment residual difference was exceeded.

[[De...]]

I've been fighting a problem trying to get blades for turbines. The program use to run just fine and other the months it has gotten worse. It pops up with the error, "Alignment residual difference was exceeded", and now it's constantly coming back. I replaced some of the bad reference points but didn't help at all. What is the best path to recertify the reference points in the program? Any other pointers would be appreciated. Been fighting the problem the last two weeks. 

Thanks

[[Ja...]]

Hi Derek ,

This warning most commonly is related to part movement in the fixture ( or fixture movement relative to part) 

The software is trying to align images by reference points , but also factor in the surface data of the scans.  If the surface moves relative to the points , then the residual value is too high - resultant mesh would return wrong result.

You did a reasonable first step of checking ref point quality , but next is checking the stability of everything in your measurement setup

Please sign in to view this username.

 i think your project stability tools could be helpful here.

[[De...]]

Please sign in to view this username.

Would that also relate to getting an outdated data error sometimes as well? Do you think that there is any chance of overtightening the part? We have 2 blades right now that won't run at all, but the co-worker that changes out parts doesn't put very much strength behind it. I know one of the fixtures was dropped and a guy fixed it by bending it back with pliers. Definitely could explain the issues with that one. 

[[Ch...]]

Hi Derek,

I've reviewed the project supplied by Mike Davies (Capture 3D).  I can see that your ScanBox includes an ATOS Plus 45M sensor with MV500 but the ATOS project does not contain a Photogrammetry Measurement Series. 

In the simplest workflow, the ATOS Project Template would contain an ATOS Measurement Series and a Photogrammetry Measurement Series.  In the case of the project supplied by Mike, the Reference Points are instead imported.  In order to give you the most relevant advice, please can you explain your workflow.

• Are you using Kiosk Interface?
• Do you use the same reference frame to measure multiple different parts (different compressor blades)?
• Do you apply a coating to your compressor blades before measurement?
• How were the imported Reference Points measured? 
  • Maybe you have a separate template which includes a Photogrammetry Measurement Series for measuring your reference frame?

As James mentioned, there is a process, and soon there will be an App available, for checking the stability of a ScanBox setup.  Using this process, it is possible to check the stability of the reference frame as well as the stability of the part within (in relation to) the reference frame.  

In the first instance, it would be great if you could share some details of your workflow by answering the above questions.

Cheers for now,

Chris

[[De...]]

Please sign in to view this username.

Correct, Mike Davis had been helping us out and we haven't been using photogrammetry measurement series for quite a few months, and it ran just fine. Not sure why all of the sudden it's been becoming an issue other than the fact some of the non coded and coded points needed replaced. We were told quite a while back we didn't need to have the scale bars on top going across either. 

The really tightening the bolts down to prevent movement helped get rid of the movement error, but now we are at outdated data error.

 

We are using the Kiosk Interface.

We use the same frame for our LPC blades but some of the holders are different but we do not have reference points on them.

We use Helling 3D Laser Scanning Spray.

I'm not quite sure where the imported Reference Points came into the file, but I assume Smart Teach. 

[[De...]]

I got where the movement I believe is under control but I'm getting outdated data and I'm certain it's for the A5 point. I have like 6 shots getting taken in that area and none are picking up on it? Some shots have 4+ reference points and some don't; the issue is I can't move the camera in a good position to get the shot without it having a collision on the platform itself.