Gear center distance question

[Robi]

I have a print from a customer when he only has the straight line distance from one hole (a datum) to the other holes

I know he doesn't care about the gear position per se, just the center distance so that the gear meshes correctly.

I have a CAD model so I can get the XY distance from hole to hole, but can you use TP with just one dimension?

If not what would be the best way to GD&T it without over constraining it?

Thanks,
Robi

 

[KENAT]

I'd say you are missing 2 dimensions.

You either need the angles for the 2 holes, or the 'vertical' or 'horizontal' dimensions from B.

It doesn't have any note about MBD or other digital definition to allow measuring off the CAD model/drawing. In fact it explicitly says 'DO NOT SCALE DRAWING'.

I'd go back to customer and get the missing dimensions, or worst case put in your best guess and tell him that's what you'll make it to unless told otherwise.

Also, are you sure you wanted to post something 'confidential'.

Posting guidelines faq731-376

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

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

 

[Robi]

Hi Kenat

The problem with just adding another basic dimension is it would lock down the position radially as well and my customer does not want that. He's ok with the radial position moving, but the center distance of the gears needs to be controlled.

As far as the drawing goes I changed the dims and put it on my drawing sheet, so I'm not giving anything away.

Robi

 

[KENAT]

As currently dimensioned the holes could be outside of the outline of the part, or interfere with the other features (holes?) is that OK?

Adding an angle for each hole, or equivalent vertical or horizontal dims, is not over constraining it, it's adequately constraining it.

Unless I'm missing something.

Perhaps do either you or your customer not understand position tolerance?

I notice you don't have dia symbol which may be correct.

Do you actually want unequal bilateral tolerance?

You'd still need the extra dimension but you'd also add FCF to control movement in the other direction.

Posting guidelines faq731-376

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

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

 

[pmarc]

I am with Kenat in saying that additional basic angle dimension or two 'vertical', 'horizontal' basic dimensions are needed for defining TP of each holes precisely, otherwise holes can be anywhere, theoretically even outside the part.

If your customer is interested more in controlling distance between datum B and the holes rather than their angular relationship with B, I would suggest method shown on fig. 7-29 in Y14.5-2009 std. or fig. 5-42 in 1994 version.

 

[PaulJackson]

Robi,

Assuming has nothing to do with the axis to axis distance between the functional gear centers... attached is sketch of how I guess the refinement would be detailed if my functional assumptions are correct.

The lower position tolerances to [A|D] control the center distances exclusively with cylindrical zones perpendicular to [A].

I would sanitize and possibly alter a customer's drawing before posting it to the internet.

Paul

 

[fsincox]

The specification of datums "C" & "A" are a little unconventional.
I would agree with Paul's modifications if his assumptions are correct, it is the same assumption I made looking at it.
The standard states "the location of each feature is given by basic dimensions, or such" that establishes the true position, the ideal location, you let it float with the applied tolerances and if those gears float you will want them to float together
Frank

 

[KENAT]

Robi, take a look at section 5.8 of asme Y14.5m-1994 where it talkes about bidirectional positional tolerancing of features.

I think this might be what you're asking about, especially 5.9.2 Polar Coordinate Method, & figure 5-42 especially.

I've attached a VERY crude sketch (I don't have much time), hopefully enough to make you go look at the standard.

http://files.engineering.com/downlo...a-5300d4563bf4&file=polar_coord_bilateral.tif

Posting guidelines faq731-376

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

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

 

[Robi]

Hi guys,

Thanks for the input on this. I actually know that the customer has not specified this correctly. My question was how to keep the center distance controlled tighter than the radial location. Which the customer did not even have on his drawing.

Kenat's last post I think is the way to go on this. I don't actually have the standard and all the books I have on GD&T don't mention bidirectional positional tolerancing. So I wasn't aware of this.

I have asked my QA manager to buy the standard. Aftermarket books seem to leave things out and if you don't know that you don't know something......

fsincox, the callout for datum C is for a locating pin on the mating piece which is flush to datum A. I interpreted that to be the center of the notch since they give a (non-basic) dimension to datum B from the center of that feature. is this correct?

They have agreed to let me redo their print and I want to do it correctly.

Thanks,
Robi

PS the drawing was cleaned up and put on my title block. If this is considered inappropriate I will edit it.

PPS how do you edit posts? I couldn't find an edit button or a FAQ on how to edit a post.

 

[KENAT]

To edit you redflag and ask management to make a change, however if in fact there is no confidential info on there it may not be an issue.

I think the issue of polar bi directional tolerancing has come up before, maybe you can find it. This is the better sketch I did at that time

http://files.engineering.com/downlo...b9f3-9077d109649e&file=5-42-APPROXIMATION.tif

Posting guidelines faq731-376

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

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

 

[fsincox]

Robi,
I mean the leader for "A" should terminate with a dot, when the intent is the surface normal to the view, and the box should not be attached directly to the leaderline without the "flag".
Frank

 

[Robi]

Thanks for the info on editing. I made an even more generic print and will probably replace it just because.

My QA manager said the standard will be here today. I will go over those sections you mentioned to make sure I understand them correctly.

It's kinda funny, we had a 12 week GD&T class for our machinists that ended just a few weeks ago and he didn't mention bi-directional tolerancing at all.

Thanks
Robi

 

[KENAT]

If it was 12 weeks full time then you were ripped off.

If it was one hour a week for 12 weeks, then it's perhaps understandable. It is only a couple of paragraphs and 2 figures.

The most difficult thing with bidirectional polar can just be getting it to format correctly with most CAD systems!

Posting guidelines faq731-376

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

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

 

[Robi]

It was 3 hours each Monday, from the local community college. that probably explains it.

I use Inventor 2011, I havn't tried it yet. I hope it will let me format it right!

Robi

 

[TheTick]

Use a composite frame. If you have a center-to-center distance and one of the centers is a datum (i.e. datum "B"), then you can control the distance with a tight tolerance position relative to "B", and then a larger tolerance applied to a full set of datums.

 

[PaulJackson]

Robi, Kenat,
Unless the specified bi-directional tolerance references one of the gear centers as the datum feature that establishes X0, Y0 coordinate reference… the bi-directional detailing will only orient the planar tolerance zone to [A|B|C] and if each has a tolerance to [A|B|C] (no matter whether cylindrical or slab-like) their deviation would be compared to the DRF [A|B|C]… not directly to each other.

Thetick,
If one of the gear centers was declared as the datum… a composite control on the other would be confusing if not erroneous. Composite controls are designed to apply to patterns and if one of the 2X pattern was declared the datum feature… maybe you can see the dilemma. It could work if there were leader lines pointing to each of the separate size callouts of the (assumed gear bearing counter bores) ending in a composite FCF (making them a 2X pattern by dimensional detailing) that would refer to [A|B|C] in the upper segment and just [A] in the lower. The tolerance for the allowable deviation in the center distance would have to be split between both diameters since there would be two cylindrical tolerance zones, one at each gear center perpendicular to [A], that would be controlling the axes of the 2X pattern… This control method would be in my opinion “overkill and/or convoluted.” You would still have to tolerance the other bores coaxial with the gear centers.

I do agree with you though that one of the gear centers should be established as a datum feature to tolerance the axis to axis distance to the other… I suggested that also.

Paul

 

[TheTick]

If not oposte, then a pair of position callouts: one with only primary (or add seondary) with tight tolerance and seond for angular position with looser tolerance.

 

[KENAT]

Paul, I assumed datum B was already one of the gear centers, and the other 2 holes were the other gear centers that were related to gear 'B'.

Posting guidelines faq731-376

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

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

 

[PaulJackson]

Kenat, Understood.
Maybe I am wrong in assuming that B is not a gear center but rather an assembly locator.
Paul

 

[KENAT]

Based on my understanding of the OP & sketch it seemed he was trying to control distance from 'B' more tightly than orientation to C, hence my assumption.

Either way one of us is a donkey;-).

Posting guidelines faq731-376

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

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