Good wording to state fabrication based on model geometry. 4

[KirbyWan]

Howdy all,

What's some good wording to put in the notes of a drawing to state that the fabrication is controlled by the model so drawing details are incomplete.

My thoughts are:

"Fabrication of part is controlled by model geometry contained in file "PARTXXXX.STL" derived from 3D CAD model. Drawing dimensions are for reference only and may be incomplete."

So I'm printing a complex shaped HX header that I plan on 3D printing. I want to have a drawing to go through our regular signoff process and document approval, but since this will be fabricated directly from an STL derived from the model, there isn't a lot of value in creating a complex drawing. that will not be used. Some of the obvious external dimensions may be used to check part fitup.

Has this been handled before? Are there other suggestions for this note?

Thanks,

-Kirby



Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[WiscoBro]

We frequently use something along the same lines as the following for complex parts where not all geometry will be measured by QC or used to actually manufacture the part:

"REFER TO PART MODEL FOR DIMENSIONAL DATA NOT SHOWN ON THIS DRAWING. TOLERANCE TO MODEL SURFACES NOT SHOWN ARE +/-.XXX"

Dimensional tolerances can either be boiler plate from normal title block, or tolerance of your machines if you know them.

*EDIT: Dwg would include KC's/overall dims; note applies to all other features not dimensioned on dwg

Thanks,
-Wiscobro

 

[rb1957]

"CAD CONTROLLED DRWG" ? This means that the CAD model is the approved Engineering, not the drwg, and so there are a lot of issues around controlling the CAD model ("how do I know this model file I'm looking at is the approved one, unchanged ??") ... digital signatures and such. A bit of a minefield (or dungeon ?).

A possible way around this is not to declare "CAD controlled" ... send the CAD file to the manufacturer, make sure the resulting part doe the job, then when inspection say "how do I check this ?" either give them some dim'ns that'll pass or have them raise an MRB/RNC (whatever you use) and have Engineering buy off "acceptable as is". Now this means you lose repeatability (the next part may not work without some "remedial" action), but it gets the problem solved today.

Of course the proper way is to go "CAD controlled" and to create the data assurance procedures ...

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[WKTaylor]

KW... most companies have 'canned' statements for their CAD designs... like this. NOPE... can't let you see my company's exact words... but close to statement by WiscoBro's statement.

Also... isn't this addressed in one or more of the following???

ASME Y14.31 Undimensioned Drawings Engineering Drawing and Related Documentation Practices

ASME Y14.38 Abbreviations and Acronyms for Use in Product Definition and Related Documents

ASME Y14.41 Digital Product Definition Data Practices

ASME Y14.47 Model Organization Practices

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

 

[ewh]

TOLERANCE TO MODEL SURFACES NOT SHOWN ARE +/-.XXX

I would recommend NOT using +/- in the default note, instead using a profile tolerance without any datum reference and include the statement that the model is considered basic.

"Know the rules well, so you can break them effectively."
-Dalai Lama XIV

 

[rb1957]

all our drwgs have a standard note ... dimns 0.0" (1 deciminal place) tolerance +- 0.10", 0.00" +-0.03", 0.000" +-0.01"

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[KirbyWan]

I like the idea of stating the model is basic and providing a profile tolerance. Since this is 3D printed, that really is the controlling reference. I was avoiding the Digital Product Definition Data Practices since that is more focused on getting away from paper drawings and this is more using a kludge to continue using drawings as a signoff tool. This is just for a research project to compare costs of 3D printing the header vs. the multi-step process of machining, brazing, welding etc. I'm sure I'll need to get a more formal buy-in for how to do this if we move forward with this.

Thanks for your suggestions.

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[mfgenggear]

Sorry my personal opinion is using cad model as just plain lazy. And not creating drawings that the floor can review while manufacturing the product. So now the sub contractor has to draw the blue print for their customer. So people on the floor who don't have access to a tube. Can view a drawing. Notconly as stated controlling revision changes. Making sure the model is the correct latest. Please note this not what you wanted to
Notified. But I thought I would give con of using a model.

 

[GregLocock]

Once upon a time we had the slogan that the "3D CAD was the master".

To me as a one time metrologist, never mind the weather, that was silly. I need to be able to pick up a part, put it on the table, and measure critical dimensions. So we then had 3D CAD is the master except for the inspection drawings which are also the master at least so far as QC were concerned. Fun.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

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

 

[mfgenggear]

9 out of 10 times a cnc programmer has to re redraw the model to make it cad cam freindly, and to break down the inprocess manufacturing.
The model does or is helpfull in viewing how the finished part. Because some people have a hard time visualizing what the part looks like.
A manufacturing engineer will disect a part and break it down in steps. The design grouo will save money up front but I garanty if any special process is required it will be more expensive to purchase.

 

[rb1957]

greg, these days we have CMM to check the finished part. Or the inspector can access the CAD model, pick a dimension, and verify on the part.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[mfgenggear]

Rb
True, but it really depends on the type of product. Rotating shafts and gears have extremely tight tolerances and runout not conducive to CMM.
Extremely large assemblies like aircraft, and rockets are to large for CMM, need lasers or optics.

 

[GregLocock]

Yes, but CMM isn't accurate enough for engines (eg main journal diameter), and it is not up to QC to identify which dimensions are critical to function.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

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

 

[WKTaylor]

FYI SIDE-NOTE, ONLY...

We have young Design Engineers re-drafting parts/assemblies from paper/vellum to CATIA. They often get frustrated when most sheet metal elements are a 'bit off and cannot be made to mesh' perfectly. Us old guys point to nominal sheet metal drafting tolerances... quick and dirty and good-enough WWII era design... and explain that any 'small miss-fit' will simply 'strain/settle/relax' all-together into one 'working assembly'... NO sweat.

And sometimes greater 'weirdness' also settle in' with ancient parts machined from die forgings... where the there are [as many-as-possible] 'as-forged surfaces' blending into precision machined surfaces. Often the Die Forging drawings less-than detailed with notes like... blend this surface to this surface... relying on the 1950-to 1970s era hand-made forging dies to define the ACTUAL/TRUE 'surface contours' that were 'made to work'.

Now of course, with modern plate and forged block alloys... and extremely accurate high speed machining processes... all of the old practices from faded old drawings... are anachronistic.

Making spare parts for these uniquely hand-crafted jets usually starts with the wording "make per drawing and suitable 'sample-part' from 'that-aircraft'" [unless the part is AFU]. That part is digitized and made exclusively as a 1-for-1 replacement part for the very Acft it came off-of. This is oddly necessary... since old jets were hand assembled with 'theme and variations' thru many production years... and were close-enough dimensionally... and have often seen one or more flight-over-loads that become 'permanently strain-set' into the airframe structure making each one a 'unique entity'. Sigh.

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

 

[mfgenggear]

Will
Lol yet durning WW2 the US built thousands of jeeps, trucks, tanks. Bomber aircraft and fighter aircraft. And all interchangeable
How every single detal , jigs, drill plates, assembly fixtures EG was coordinated back to a master plaster. All details fit exact. Now any body whos work withh sheet metal knows thats why we match drill with pilot holes then clico in place and rematch drill to size.
And it was all calculated and designed on paoer with out computers. Slide rule and notes.

 

[mfgenggear]

Kirby
An auditor could be a quality engineer, will want to view how the drawing and the software
Are been revision control. A drawing starts out with a no change release, so does the 3 d model.
If your company wants the 3d model then driving factorial be Quality must sign off on it. to meet what ever meets or exceeds their a quality requirements.
Then I would specify as stated above.
Are there any critical close toleraces.

 

[RoarkS]

Agreed. Model is basic. Profile tolerance for non-critical parts... just make sure the profile dimension is appropriate... don't just slap .005 in there... I'll add zero to the price tag... actually that happens for any GD&T callouts really.

100% it is not up to the machinist or QC to determine critical dims. I find what usually works out is calling out the interface features at minimum and really allow freedom on what doesn't need to be tightly controlled which is honestly most... especially near net castings and forgings.

Keep your QC goons out of the drafting offices. Released files are dumb models like *.igs, *.stp, *.x_t. that are isolated in a PDM vault... often times embedded into encrypted PDF's. If you haven't tried that it's pretty amazing.

Yes machinist like to redraw models...I should know... But for me that's really starting to go away past few years with cad/cam getting to be better friends recently with "direct editing" and Boolean commands... not to mention the Renishaw plug in for on machine inspection... I mean that's so cool. That and the level of automation for 5 axis stuff is getting to the point where it's almost drag and drop.

I do run Hexagon PCDMIS... and almost as a rule import the "released" cad model for inspection programs.

Please don't forget to ease up on surface finish where 16 micro inch or better isn't needed... Don't be afraid of that 125 or even better "Stock" or "as cast" acceptable.

Yes I will go pull out an air gage if I have to, but you better let me know on the print or procurement if you expect special tooling for inspection because there isn't much anymore my CMM isn't gonna handle. That PH20 head can do some wild stuff... and I'm calibrated under 20 millionths on the near side of the machine.

 

[jaceb]

"FOR MENUFACTURING REFER TO CAD DATA"
SIMPLIEST
"MANUFACTURE PER CATIA MODEL"

GD&T SYMBOLS AND NUMBERS CAN BE ADDED IN 3D MODEL - AT LEAST IN PRIME TYPE SOFTWARE LIKE CATIA AND SIEMENS NX
OF COURSE IF YOU CAN AFFORD ON RESULTANT PRICE TAG AS MENTIONED BEFORE...

 

[KirbyWan]

This turned into a great discussion. Thanks everyone for responding.

I agree having a drawing/model for quality to inspect to would be an important piece of the puzzle. For this application I'm not concerned with remodeling for CNC since this is going to be 3D printed, the model really is controlling the profile of all the surfaces directly. One of the real benefits of 3D printing it the ability to model complex internal geometry which if machined would be in multiple pieces that are inspected before assembly through brazing/welding. So I'm creating an internal geometry that could not be inspected, except destructively. That destructive inspection may be the requirement for a first article validation, but a functional test may be all that is needed for production. For this I went with "REFER TO PART MODEL FOR DIMENSIONAL DATA. TOLERANCE TO ALL MODEL SURFACES ARE +/-.005. DIMENSIONS SHOWN ARE REFERENCE ONLY." If this gets past the research project stage I'll probably remove "DIMENSIONS SHOWN ARE REFERENCE ONLY" and change it to "INSPECT TO DIMENSIONS LISTED" and have a destructive inspection managed separately as part of the FAI process.

Thanks all,

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.