How to dimension trapezoid part? 1

[floide]

Hi guys,
New here, this is my first post/thread. I just stumbled on a part and I just can't figure out how to dimension it properly in accordance with ASME Y14.5.

(See attachment for drawing) My doubt lies in the 3.625 dimension: everybody knows that symmetry is implied in this part but I really don't know how to dimension its location properly. Should I use a half dim, a profile tolerance or a position tolerance like the 2.200 pocket in the side view? Can I even use the 3.750 hole axis as my symmetry plane?

I asked a couple of engineers here in the company, but none of them had the same answer and GD&T wasn't even close to being addressed.

Thank you for your answers.

(By the way, I'm french speaking so I apologize for any language ambiguities or strange formulations/translations)

Fred

 

[drawoh]

floide,

A couple of remarks here.

I can see your symmetrical feature, but it is not clear. I might not have spotted it if you had not told us. I would dimension everything, and not force your fabricators to guess.

This looks like a casting that will be machined out a bit. In such a scenario, your very accurate holes are not good datums. They will not be present at your foundry. I strongly recommend that you specify datum targets. Consider modelling in the actual datum features. Now, the foundry and the machinists will be jigging to the same features.

Your bottom datum face looks machined. If so, it is not a good datum either, for the reason noted above.

If you create separate casting and machining drawings, this problem will be obvious to you. This is one of quite a few reasons why you should make separate drawings.

This is not a GD&T issue, but the exising GD&T on your drawing highlights the problem. Can your slot be fabricated (easily)? Your positional tolerance is very tight for (sand?) casting. Making the slot penetrate to the hole is not impossible, but it is not simple, either. Applying good GD&T to fabrication drawings often show up manufacturing issues, which is one reason why you should do it.

--
JHG

 

[MechNorth]

I don't see any issue with the symmetry on the right-side of the part. You have a centerline drawn in and the features appear centered on it. I wouldn't add the half-dimensionm even as a basic dimension.

I would support that cast datums are needed, but on the final part, these datums are fine. The important thing to remember is that you need to tie the machined datums back to the cast ones, as well as controlling the machined datums to higher-precedent machined datums. The tolerances on the machined holes are tight, but achievable.

As for the GD&T, a general surface profile is needed to start with. Datum Feature-B needs to be controlled back to Datum-A. Datum Feature-C needs to be controlled back to Datums A and B. The groove needs to be controlled wrt Datum-C as well. Other than the two holes and the slot, none of these features should have +/- tolerances, and should all be controlled by geometric tolerances.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[pmarc]

I am with Jim on that.
- No issue with symmetry;
- This approach to tolerancing is nicely shown in fig. 4-9 of Y14.8-2009 standard on "Castings, Forgings, and Molded Parts".

http://files.engineering.com/getfil...-ab68-02c71c45e3be&file=casting_machining.png

 

[floide]

Thanks everyone for your answers.

Maybe I should have pointed this detail earlier. It is true that the part looks like a casting: because it was one. The part actually broke on the machine and we are going to have one machined entirely; no casting involved here. So my datums will actually be available for use.

Fred
Montréal, Qc, Canada

 

[pmarc]

In this case it is a good idea to follow Jim's (MechNorth's) suggestions except the part of his reply about cast datums.

 

[KENAT]

Sorry, but unless there are some FCF's for the 3.625 dimension and for the various diameters (other than 3.75 hole) then this part is not fully dimensioned.

So I would say you need to add a position FCF to the 3.625 dimension similar to that for the 2.200 dimension to give some control on just how symmetrical the 3.625 dimension is.

Posting guidelines faq731-376

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

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

 

[MechNorth]

The 3.625 isn't a FOS either, Kenat. No clear directly opposed points, so position isn't appropriate. A general surface profile would cover all "non-critical" features, with tighter controls applied as needed.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[KENAT]

Ah, the old caliper rule argument. I'd argue there are 'opposed elements' but if concerned then use use of profile could alleviate concern.

Posting guidelines faq731-376

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

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

 

[MechNorth]

Hard to verify that the points are actually opposed otherwise, Kenat.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[floide]

The groove needs to be controlled wrt Datum-C as well.

Jim, I don't get why datum C would be required in this case.

Fred
Montréal, Qc, Canada

 

[fsincox]

Jim,
For those of us still trying to beat this idea in their heads, why does this method actually make finding the opposed points easier?

 

[pmarc]

floide,
Will this part be anyhow rested against datum feature surface C when mounted in an assembly? If yes, then having C in positional callout for the groove will better reflect function of the component and constrain last degree of freedom left.

Frank,
For profile of surface check, suggested by Jim, you do not have to measure any distance between the two points. You do not have to bother whether they are opposed or not. All you have to do is to verify whether the corners fall within boundaries of profile tolerance zone. Much more practicable and logical than verifying the distance between something that may even not exist in reality due to break edges or fillets (which probably be there due to technological reasons).

 

[floide]

pmarc,
No it will not rest against C. I've attached a picture of the assembly for your understanding. The T-slot bolt is pulled upward towards the bottom of the guides on each side and it doesn't touch the bottom of the 2.200 pocket.

For those who are interested, I've also attached the completed drawing to the best of my GD&T understanding. Feel free to comment/ suggest corrections for I am just beginning my, how should I say, "learning of GD&T in applications".

Again, I want to thank everyone for their helpful answers.

Fred
Montréal, Qc, Canada

 

[MechNorth]

Frank; using the long (external) jaws of the caliper (with the flat surfaces) makes it easier to find directly opposed points. Not so much luck on the internal (short) jaws.

Fred; without reference to datum C, the orientation (perpendicularity) of the center plane of the slot is not controlled. This is an assumption that face of datum feature-C is riding on something; if the arm is free-floating on a shaft / pin without contacting on that face, then reference to datum-C is irrelevant.



Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com