GD&T vs. Direct tolerance comparison

[BiPolarMoment]

Cross posted from Solidworks:

[ref: Solidworks 2006]

I have a tooth profile that was previously directly toleranced 100%, I'm trying to relieve the manufacturer trying to inspect it by applying a profile tolerance instead but I need to ensure (for multiple reason) that the overall geometry is held to the same or tighter tolerance overall.

I am not expressly trained in GD&T but have a basic grasp of it and I'm trying to evaluate what profile tolerance will satisfy my needs.

Here is an example of the profile I'm working with:

The profile is currently controlled by radii on the upper surface with a 0.5mm offset to the base of the teeth, all "valleys" (flats) are positioned perpendicular to an angle from the center of the same radii +/-0.5 degrees [All linear dimensions are +/- 0.1mm and angular are +/-0.5 deg]. A quality engineer at our manufacturer suggested a profile tolerance (with no datums) of 0.2mm. I'm nearly certain this will control the angular position of the teeth more accurately than the current dimensioning scheme but I'm concerned with how it affects the tooth shape and overall profile contour.

Finally to my question: Is there a fairly simple way of generating what would be the acceptable profile silhouette using solidworks? Alternatively, any way to evaluate my concerns on this profile?

 

[KENAT]

.2 is almost certainly tighter than what you already have, without running the calcs to check.

You've got a significant tolerance build up from all the dimensions, some of which you don't show in your shot.

You've got to remember that radially all your angular tolerances add up on so just from your picture, your last 'notch' could be off by 8*+-.5°. so +-4°

Let me reverse the question. Based on function, especially how it interfaces with it's mating toothed (or whatever) part what tolerance do you need? This should be your starting point, and then work back to make sure it can be inspected and manufactured.

There are tolerancing standards directly related to gears etc, would it be appropriate to look at and maybe invoke one of these?

KENAT, probably the least qualified checker you'll ever meet...

 

[BiPolarMoment]

I appreciate your input and I definitely agree with you on the angular tolerances.

The problem is that this is currently being manufactured to this tolerance set in Europe--it is necessary to make "no" changes to the geometry. So it's not so much what tolerance we need as what tolerance will get us within the original spec and allowing for an easier inspection procedure (profile silhouette).

So it's an apple to pear comparison, they're allowed to taste a little different but the apple has to fit within same pear shape at the end.

As for what tolerance is appropriate? These tolerances are grossly over-tight in some respects but I can't do anything about this because it would throw regulatory complications into the mix.

Additionally, the time line is likely too short to attempt and re-do the entire set of drawings with more appropriate dimensioning schema.

I'm trying to think of what an 0.2mm profile tolerance (with no reference datum) would look like--if you assume those teeth are theoretically sharp (they aren't), is it a curve that is offset "vertically" 0.1 in each direction, each with perfectly sharp teeth? I apologize for my lack of knowledge in this area--it will hopefully be changing soon.

 

[ctopher]

"...this tolerance set in Europe"
What tolerance is that?
What standard are you following?

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 08
ctopher's home (updated Aug 5, 2008)
ctopher's blog
SolidWorks Legion

 

[BiPolarMoment]

Chris, er.. I didn't mean to imply they are following any standard-- I just meant the "set" of tolerances they are currently using (i.e. the same as the picture posted).

 

[KENAT]

Bipolar, by tightening tolerance (if that's truely what you're doing) you will at most be more closely matching the nominal shape. So I don't follow your apples & pears concerns.

The .2 surface profile is simplistically the equivalent of +-.1 for the individual surface. Without a datum reference (sorry overlooked this in my initial response) your profile tolerance will only give you form control. This is probably only slightly tightening the tolerance as the angle is over a fairly short distance and I'll assume the radius of the bottom of the notch is +-.1. So if you're only concerned about the individual notches/teeth, may work for you.

I'm almost tempted to say that the solution might be a composite profile. This would allow you to 'roughly' locate each notch/tooth but then have a closer tolerance on the individual tooth profile. However, given what you've said I'm not sure you have enough GD&T experience to successfully apply this.

I don't know the application of this part but if built to the current maximum tolerances I wouldn't be surprised if its non funtional. If you are changing it, and people will be machining/inspecting to the new tolerances/dimension scheme I would think you'd be best to address this issue.

Also don't dismiss Chris's question so quickly. There are differences between ASME & ISO standards which might impact your answer.

KENAT, probably the least qualified checker you'll ever meet...

 

[BiPolarMoment]

I wasn't really dismissing chris' question so much as I'm fairly certain they (our European counterparts) don't follow any standard.

For this particular part I imagine they intended for each tooth to have a noncumulative angular position w/tolerance but there's nothing on the print to indicate this explicitly--so as you've said, at maximum tolerances the part is likely non functional. A lot of the drawings they have made seem to have implicit tolerances that are not stated. For example this same part has a relief specified at the end of a blind bore (under-cut) but the dimensions they chose would allow the feature to be optional per the general tolerancing scheme--I am almost certain this is not their intent, but there are no notes on the print specifying otherwise.

Hopefully we will have a discussion with the engineers across the pond to address issues like this, but I hypothesize they can only provide insight and not a solution.

 

[ewh]

If no standards are being followed, you have no guarantee that invoking a profile tolerance will be followed as you intended.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[BiPolarMoment]

This is true--and is the subject of a much more in depth discussion of how backwards our company is currently. Unfortunately that one I can't tackle myself.

 

[MechNorth]

ewh has made a critical point, a standard must be identified (on the drawing) to which the drawing is to adhere, otherwise there's no definition of Surface Profile to use. Also, the definition / applicability of Surface Profile is different for ISO & ASME, so you have to be specific. You can use profile of a surface without a datum reference to control the size & form for an enclosed boundary (such as a hex cutout for a wrench), but if you're not controlling an enclosed boundary, you MUST have datums in order to locate & orient those surfaces.

I hate to say it, but you are better off not applying any GD&T until you get some training & backing from the company. Sorry, I know that's not what you want to hear.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[BiPolarMoment]

I know that's the answer that is correct, but the problem is that it doesn't change my situation any.

Unfortunately, I don't think it's an option to leave it as is--this was intended only as a bandaid on what is a much larger problem.

On the flip side, I believe some training is potentially in my near future since there has been some reorganization in the company --I've been asking for nigh on 2 years with the answer that it's not useful for our industry, and that it would mean training people above me to be able to read it .

Anyway, thanks for the consideration folks, I apologize if I wasted any of it

 

[ctopher]

I agree with the others about following standards. I can't really give an answer without knowing which standard you follow. Since you don't follow a standard, you may do whatever you like to the drawing...to some degree.

Looking at your drawing, it would be very difficult to machine or inspect the part. Do a Google search, look at the Machinery Handbook, and purchase your standards documents for gear design.

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 08
ctopher's home (updated Aug 5, 2008)
ctopher's blog
SolidWorks Legion

 

[KENAT]

Yeah, without the standards you're in iffy ground trying to apply surface profile, even if it's not the answer you want (not to mention other implications to the entire drawing).

How's this as a very poor band aid, apply whatever standard your manufacturer is familiar with, possibly just to the profile call out. I'm not sure how valid it would truely be but is maybe better than calling out profile with no standard invoked.

Once again, the .2 vaule is probably pretty close to what you currently get for each 'notch'.

KENAT, probably the least qualified checker you'll ever meet...