Page 1 of 2
I am new and this is my scan
Posted: 05 Apr 2018, 20:06
by erol
Hello I am from turkey. I am a design specialist in a private company (mold design and cnc programming). This is my scan. I scanned it using david sls3 + david 4.5 + single camera + turn table. The part sizes are 290mm x 250mm x 70mm. The scanning accuracy is 0.05 mm on average. I replaced the camera lens with a wide-angle Cosmicar 8.5mm lens. Have a nice day
another scan
Posted: 05 Apr 2018, 20:11
by erol
david sls3 + single camera + david 4.5
another scan
Posted: 05 Apr 2018, 20:28
by erol
david sls3 + single camera + david 4.5 + turn table
Re: I am new and this is my scan
Posted: 06 Apr 2018, 12:55
by Micr0
Thank you for sharing. Excellent work!
Why did you change the lens?
Are you coating your parts for scanning?
Re: I am new and this is my scan
Posted: 06 Apr 2018, 17:14
by erol
Thank you micro.
I use it (cosmicar) to scan a wider scan area at a time. How do you scan large parts? Do you have any suggestions for me.
Yes i am coating the parts.
at the same distance from the scanner.
1st screen view: computar 12mm
2nd screen view: cosmicar 8mm
Re: I am new and this is my scan
Posted: 06 Apr 2018, 18:50
by Micr0
erol wrote: ↑06 Apr 2018, 17:14
Thank you micro.
I use it (cosmicar) to scan a wider scan area at a time. How do you scan large parts? Do you have any suggestions for me.
Yes i am coating the parts.
at the same distance from the scanner.
1st screen view: computar 12mm
2nd screen view: cosmicar 8mm
Any additional area that the camera sees that the projector can't cover is useless. More importantly the camera needs to see the distortion of the projected patterns on the subject. The better is sees that the more detail you will get in your scan. Hence the advantage of HD cameras over SD. Ideally the camera field of view would exactly match the projected image. As for scanning larger objects.... I have literally done pieces that are the size of a small truck (sculptures). I calibrate to the largest panels I have (I made 1m panels for this job), then scan the object in sections and assemble in fusion. The engine I just posted was done with a setup calibrated to 280mm. It's wouldn't completely fit in to the cameras field of view so I did it as upper and lower sections.
Re: I am new and this is my scan
Posted: 06 Apr 2018, 20:55
by erol
Thank you for your advices Micro.
I upgraded my david sls3 scanner with my second camera and david 5.4 software.
I did a few tests today with two cameras and a turntable.
1st test (morning): 70mm piston. The calibration panel angle is 90.05 and the calibration is nice but the accuracy is bad (0,15mm-0,5mm).
(Optimize multi camera calibration: on)
2nd test: I made it with 50mm piston. 1st calibration ok but calibration panel angle 90.5 (optimize multi camera calibration: on)
3rd test: Same setup as 2nd test 50mm pistons and optimizing multi-camera calibration: off, calibration okay.
calibration panel angle 90,035
accuracy: 0,02mm-0,05mm
Do you have problems with double camera and turntable? Accuracy is important to me because I do 3d modeling with scans and designing molds
Re: I am new and this is my scan
Posted: 06 Apr 2018, 21:50
by erapip
Nice scan erol.
Can you explain how you measure the accuracy? And what about the panel, did you produce them? Did you compare your scan linear acuracy with any other system or even with clasical metrology? It will be with great interes to make some datta comparision. I think that also scale on standart HP panel is not ok, for RE process for mechanical part. They shoul be wroten in the range of micron precision.
Versio 5.4 for me is OK. Attache some work on reliefs.
Thanks
Re: I am new and this is my scan
Posted: 08 Apr 2018, 13:07
by Micr0
Are you using David/HP calibration panels?
Have you opened the debug console and looked at the error values for the calibration procedure?
Re: I am new and this is my scan
Posted: 08 Apr 2018, 21:11
by bigbomber
erol wrote: ↑06 Apr 2018, 20:55
Thank you for your advices Micro.
I upgraded my david sls3 scanner with my second camera and david 5.4 software.
I did a few tests today with two cameras and a turntable.
1st test (morning): 70mm piston. The calibration panel angle is 90.05 and the calibration is nice but the accuracy is bad (0,15mm-0,5mm).
(Optimize multi camera calibration: on)
2nd test: I made it with 50mm piston. 1st calibration ok but calibration panel angle 90.5 (optimize multi camera calibration: on)
3rd test: Same setup as 2nd test 50mm pistons and optimizing multi-camera calibration: off, calibration okay.
calibration panel angle 90,035
accuracy: 0,02mm-0,05mm
Do you have problems with double camera and turntable? Accuracy is important to me because I do 3d modeling with scans and designing molds
Hi erol and other friends. As I think you are try to judge about your system accuracy regards to scand piston, I can mention that a piston is not a real circle. Actually pistons are made like a ellips to optimize their performance. so if you scaned a piston and for measurment force the CAD software to create perfect circle on the mesh, there will be a cosiderable error on result.