Statistical analysis of collected points

Discuss PolyWorks software here.
Post Reply
Ethan_Kay
I have made <0 posts
I have made <0 posts
Posts: 2
Joined: Tue Apr 21, 2020 3:11 pm
3
Full Name: Ethan Kay
Company Details: RL Best Company
Company Position Title: Sr Engineer
Country: United States
Linkedin Profile: Yes

Statistical analysis of collected points

Post by Ethan_Kay »

We are using Polyworks with a Faro Laser Tracker to inspect customer equipment and reverse engineer components. These are large machines and are often old, 20-70 years. There are rarely paper drawings available, much less 3D models and many are modified from the original configuration anyway.

What I am trying to do is rather than create a deviation from nominal is gather a group of points taken on a surface and do a quick calculate Min, max, nominal, and deviation from one of those values. The problem I ran into yesterday is the requirement was a total deviation of 0.010" The low value was 20.7657 and the high value was 20.7757 so I had to manually set the nominal dimension at 20.7657 with ta tolerance of +.010/-.000 in order to get a pass which as an engineer makes it seem shady in the report since the drawing shows 20.750, but we were looking for flatness.

What we are often looking for is the alignment of components of the equipment with relation to each other, and overall levelness of the equipment because the building foundations have a tendency to settle slightly so while everything is aligned the entire piece of equipment may be tilting meaning alignment can no longer be performed with a spirit level.
User avatar
smacl
Global Moderator
Global Moderator
Posts: 1409
Joined: Tue Jan 25, 2011 5:12 pm
13
Full Name: Shane MacLaughlin
Company Details: Atlas Computers Ltd
Company Position Title: Managing Director
Country: Ireland
Linkedin Profile: Yes
Location: Ireland
Has thanked: 627 times
Been thanked: 657 times
Contact:

Re: Statistical analysis of collected points

Post by smacl »

I haven't used the Faro Tracker, but for this type of application you might be better off with a good total station than a scanner. I'd tend to approach this by having a fixed series of points for the total station (or scanner) surveyed and leveled, mark out four or more points on each rigid section of your machinery and also your floor and maybe walls for control purposes and re-survey these as needed to check for movement. You'll be more accurate directly observing marked discrete points with a total station than picking points from a cloud. Having four or more points per rigid section will allow you to do a six parameter rigid body transformation to see how that section has moved over time in terms of x,y,z shift and roll, tilt, yaw rotation. Doing this with points on the floor and walls as well will give you some idea if other changes are due to changes in the building. Basically what you're doing is a mix of deformation monitoring and dimensional control.
Shane MacLaughlin
Atlas Computers Ltd
www.atlascomputers.ie

SCC Point Cloud module
User avatar
southamd
V.I.P Member
V.I.P Member
Posts: 129
Joined: Thu Nov 18, 2010 8:42 am
13
Full Name: David Southam
Company Details: Faro UK
Company Position Title: Regional Manager EU North
Country: UK
Linkedin Profile: Yes
Location: Worcester
Been thanked: 1 time
Contact:

Re: Statistical analysis of collected points

Post by southamd »

Hi Ethan I will get our Faro contact Michael Fisher to give you a call.
Many thanks Dave
David Southam| 3D Documentation Regional Manager Europe North
Faro Technologies UK Ltd.
Mobile: +44 (0)7794512714| Fax: +44 (0)2476 23 61 50
Nasdaq: FARO | [email protected] |
mike_fisher
I have made <0 posts
I have made <0 posts
Posts: 1
Joined: Thu Apr 23, 2020 4:21 pm
3
Full Name: Michael James Fisher
Company Details: FARO
Company Position Title: Account Manager
Country: United States

Re: Statistical analysis of collected points

Post by mike_fisher »

Thanks Dave!

@Ethan I just talked to Jeff and got your contact through him. There are some easy ways to include min/max, standard deviation, etc. You can report this info for the form/flatness of the feature itself or in a more general sense back to another feature/datum. Multiple ways to approach what I think you're going after.

Shoot me an email and we can do a zoom call/screenshare and I'll walk you through what I'm talking about.

Mike
Post Reply

Return to “PolyWorks”