Adding features just to use as datums

[cjccmc]

I have another cone shaped part in work and cannot find a set of features that make clear practical datums. Can't even get a set of points on this curvey thing that I think will be repeatable with good accuracy. I'm thinking of adding three small bosses and a hole to function as datum features. I have not seen this done before in my dept and know it will get funny looks at peer review. Just wondering if any of you have used this approach before.

 

[MikeHalloran]

A hole in a flat face would also work for two datums.
A flat or a hole could tie down the rest.

I have, on a couple of occasions, added features that were not functionally necessary in parts, solely to serve as datums.

I remember one coupling for a cantilevered mixer shaft that comprised a cone divided on a diametral plane through its axis. At the fat end of the cone was a tooth, or key and keyseat, set up to couple the two halves for thrust. The halves had to be assembled radially, then a big-ass nut with a chamfer inside one end was run up to lock the shafts together.

The coupling halves were normally made in sets in Germany, but I had to have one made in the USA to mate with a German part that I didn't yet have. (I did have the German drawing.) I added a couple of center-drill chamfers to the ends of the USA part's central hole, so it could be swung on centers while the cone OD was ground. I even showed centers in phantom so that I wouldn't get questions about what the chamfers were for. I have no idea how the Germans manufactured the pieces, but there was no seat for a center in their parts. It didn't affect the fit or function either way. The parts fit OK on the first try, and everybody was happy.

I'd say that center drill holes are a decent example of features that are used as datums and usually are not essential to function. Okay, they are also used in manufacture, not just as inspection datums.

ISTR adding simple datum features to odd-shaped plastic parts, just because the parts were very difficult to measure otherwise. ... this was before GD&T became common, so I didn't call the features datums, because that would have cost extra. ;-)







Mike Halloran
Pembroke Pines, FL, USA

 

[cjccmc]

"...so I didn't call the features datums, because that would have cost extra." ha!
We used to joke in the early days of GD&T that the machine shops would see all our callouts and just say "I don't know what any of this means, but we better just make it tight and double the price".

It's becoming increasingly apparent to me that Y14.5 datum principles are only practical on parts with flat or cylindrical mounting interfaces. Once you get away from these it's like fish out of water. I believe the people on this forum have a much higher understanding of GD&T than the average user out in the working world, but I can tell from the lack of response to my previous posts on complex parts that 14.5 is unweildly in that arena.

 

[MikeHalloran]

In a way, it's still early days for GD&T.

With difficult and unstable business conditions, shop are constantly appearing.
... and disappearing.

Nobody wants to pay for training, certainly not more than once. High turnover negates the investment.

The direct workforce seems to have more thieves, incompetents, illiterates, innumerates, dopers, and dopes in it than ever before. Sure, they get weeded out, eventually, but because of privacy laws and a surplus of attack lawyers, just move on to become someone else's undiscovered problem. In the meantime, work gets screwed up and is delivered wrong or late.

In light of all that, I grit my teeth when people here get arguing over obscure GD&T terminology, like C language programmers argue about the ways in which the C preprocessor will (mis)behave, fed an odd fragment of source code.


My point is that a lot of the effort put into GD&T is wasted on people who can't read a drawing, or a simple sentence.




Mike Halloran
Pembroke Pines, FL, USA

 

[fsincox]

This is a very common practice, how many outside edges were really that important, other than for manufacturing purposes?
Frank

 

[CheckerHater]

Aren’t “features just to use as datums” called “datum targets” (at least when it comes to casting)?

 

[cjccmc]

"Aren’t “features just to use as datums” called “datum targets” (at least when it comes to casting)?"
I had something different in mind. Actually adding bosses and a hole that have no purpose other than to serve as datum features.

A part as shown in Fig 4-39 in the 1994 std is closest to what I am dealing with in this application. Maybe it's just me, but it seems that part has more than one stable position on that set of six points. It might in theory have just one but I wonder how repeatable it would be in actual practice.

 

[ewh]

I have also seen shallow flat bottomed holes used to establish an intial datum structure in castings. The hole bottoms were used, and it worked well.

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

 

[CheckerHater]

Same with small flat dimples. Couldn't find good picture though.

 

[TheTick]

I've done this in various ways. Usually, it's putting a flat edge on sheet metal part. Often it's to avoid using faces or edges with contours that are hard to repeat as datums.

Most of the automotive sheet metal parts I worked on had locating features separate from fastener holes.

Simple thing to justify: cost of an additional hole or changed contour versus cost of checking and rejecting otherwise good parts.

 

[KENAT]

I've added bosses to castings to support datum targets as suggested above.

Posting guidelines faq731-376

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

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[cjccmc]

Thanks to all for the inputs. Seems like fairly common practice when typical datum features are just not there.

 

[MintJulep]

Nobody thinks twice when you say you added a "tooling feature" to a part that can't otherwise be held for a necessary manufacturing process.

Inspection is just another manufacturing process.

So call them tooling features - because really that's what they are.

 

[fsincox]

From my years of experiance, it was the dimensioning by function that was more unusual, not dimensioning for manufacturing.
Frank

 

[Tarator]

It is very hard to talk about a part's GD&T scheme without knowing the function of the part and how it mates with other components.

Back to this cone shaped part, why can't you use the cone portion as a datum?

 

[cjccmc]

"Back to this cone shaped part, why can't you use the cone portion as a datum?" Maybe it could be, but this part is just a small portion (10 deg sweep) of a very large (8 ft radius) cone.