Dimensioning Organic Forms? 1

[fenixfire]

I have a product family that I need to create a set of QC drawings for using GD&T. Does anyone know of or have an example of a drawing for a product built entirely by lofting 3D splines. Due to the molded fit of the product this was the only method possible to give the surface control and continuity required for proper function of the product and proper fit to the anatomy.

I really just need to see a similar layout to get a starting point.

Thanks,

MJ

 

[PaulJackson]

Spline specifications are typically given in a table like the one at the end of the Hexigon illustration from the link below.

http://www.hexagon.de/wn5_e.htm

There are references to ISO and ANSI standards where the specification criteria is explained further.

Paul

 

[ewh]

For more organic forms, an approach used by the aircraft industry was to cut a series of sections, then define however many spline definition points you require in each section.

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[PaulJackson]

I agree with ewh... if by splines you mean NURBS... define each of the sections surface contours with a reasonable set of basic coordinates and specify profile to tolerance the functional limits of the fabricated contour.

We did the same with reactor, impeller, and turbine blade contours in automotive torque converters.

Paul

 

[fenixfire]

Thanks,

Yes they are basically Nurbs, but I would not consider what Pro or Solidworks can do Nurbs. I researve that for Alias.

Basically how would someone take a 3d scan of a human ear and then dimension it so that "someone who has never sean an ear can understand it". More than just a basic table for a spline gear. I have already provided the basics, as well as 3d part files, items are being produced and the automated QC inspections are fine, but client wants complete GD&T documentation as he wants to manually check the tolerences of every part once a week.

 

[PaulJackson]

The only GD&T documentation necessary to define a form like that is a solid model and an "all over" surface profile specification. The model can be analytically compared to the measured "point cloud" data gathered by an optical CMM scan, physical CMM trace scans, photogrammetry, or other methods of retriving point data from the produced form. A software product called "verisurf" among others I am sure can show conformance/deviation from the solid model graphically as well as tabular to monitor process performance and provide conformance summations. If location and orientation of the forms are required to functional attachment features then define them on the model and tolerance the form's orientation and location from them.

Paul

 

[fenixfire]

Paul,

Thank you, this is what I have been trying to convey to the client. However dispite is age he is very oldschool in thinking.

 

[KENAT]

Fenixfire, as Paul said, perhaps the best way to 'Completely' document it is with model and 'all over' profile tolerance. That said you can apply different tolerances to certain surfaces if need be.

ASME Y14.14 is the standard you want to take a look at for digital data product definition but it probably wont give all your answers.

Here's a sample of notes I've used for similar items.

1. DIGITAL PRODUCT DEFINITION PREPARED IN ACCORDANCE WITH ASME Y 14.41-2003.
2. THIS DRAWING SHALL BE USED WITH MODEL (INSERT PART NUMBER) (REVISION PER THIS DRAWING) FOR COMPLETE PRODUCT DEFINITION. MODEL GEOMETRY IS BASIC.

I then have default surface profile as required.


KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[ajack1]

I totally agree with Paul Jackson, that is certainly how it is done in the automotive industry for any truly organic forms, body panels, facia, etc.

Why would anyone want to try and cut a whole load of sections and then try to dimension things that cannot be dimensioned when the only way to measure the part is on a CMM against a model?

 

[KENAT]

Oops, 14.41 is the US standard.

I will say this, unless you are in a suitably integrated supply chain where you can all share CAD Data without translation issues or losing control of the version etc. I'd suggest a hybrid drawing/model approach not pure MBD. You have a simplified drawing, essentially to accomodate notes and put any annotation that's required such as identifiying datums or different tolerances for individual features etc, some reference overall dims can be usefull too.

However, this has been discussed at length elsewhere.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[fenixfire]

Thank you to all who offered their insight.

Luckily the entire system is digital using Solid Works native files with .step backup files. The only analog aspect is a micromanaging client, but that can be mitigated fairly easily with the right input and direction.

Thanks

MJ

 

[KENAT]

Make sure you know how to do annotations etc. in SW and that the machine shop knows how to access them and that the files are 'secure' so they don't get changed accidently or intentionally.

The last I'd heard Solid Works couldn't really do full MBD but maybe that's changed.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484