Rookie needing some help with GD&T questions

[jlenth]

Hoping I can get some help on some questions that I can’t quite find a clear answer for in either the ASME Y14.5 standard or already on this site (not the most experienced with GD&T yet).

Question 1), cylindricity- I have an application where we are welding up a bearing housing and machining afterwards. With the residual stresses in the bearing housing from the welding process, the bearing housings can slightly go out of round during the machining process. I already know that per the standard, the cylindricity tolerance must be less than the dimension tolerance. From previous application experience I know I need to maintain a tighter than normal control on the average size of the bearing bore. On the other hand, the bearing itself can handle a little more “out of roundness” than what my dimensional tolerance allows per the standard. The only thing I can find in the standard that could possibly allow my cylindricity tolerance to exceed the dimension tolerance is the “average diameter” as described in paragraph 5.5.3 of the 2009 standard. Is my “AVG/cylindricity” callout in the “housing” example legal? I don’t believe I would have to specify the free state condition in the feature control frame??

Question 2), Datum placement on a cylindrical surface-I think I already know the answer to this one…Is the placement of my datum feature A, B, & D legit?Per the Y14.5 2009 standard paragraph 3.3.2 section (a), I believe in my example it may be taken that these datum features are the surfaces themselves as opposed to the axis of the cylinders.However, in section (c) of 3.3.2 paragraph, it does state “placed on the outline of a cylindrical feature surface” in which I think I could argue that datums A, B, & D are on the outline on the cylindrical surface.I just ask this because on my actual drawings space is pretty tight and this is the cleanest way of specifying these datums.

Question 3), true positon of a keyway-Can someone confirm that I’ve properly used GD&T to locate the keyway position in the “shaft” example? I’m only concerned about the keyway being centered on the shaft on the datum C portion, axial location is not critical.I think it’s correct but I still have a little uncertainty….

Question 4), locating of splines with GD&T-I also wanted to apply some GD&T to some internal splines in the shaft and relate it to the bearing seats (datum’s A & B) that this shaft is ultimately located by and spins on. This just to make sure that the mating splined output shaft of hydraulic motor will install into the piloted bearing housing with this splined shaft in. I was just thinking of using true position for a coaxial relationship between the splines and the bearing seat axis and using the ANSI B92.1 callout for the machinist to refer to for the spline information.I can’t find any reference example on something like this, does this look legit?
Question 5), true position of a hole drilled at an angle-I need to locate a hole drilled at an angle in a bore of a part as shown in the “housing” example. Primary datum would be the axis of the bore and located off the face designated as datum E and clocked rotationally in respect to a cross drilled hole (datum F).The location of the hole where it intersects the large bore (1.00 basic dimension) is what’s the most critical.Did I correctly apply this positional tolerance?What leaves me in doubt that this is not correct is how does one apply the cylindrical tolerance boundary?

I appreciate everyone’s input to the question/questions that he or she can help me out with. Thanks

[Kelly_Bramble]

Question #1 – The cylindricity will allow a total size variation of +/- .0007 or .0014 local size and the size variation allowed by AVG is .0005 on diameter. The specification seems ok to me however I am not on board with the functional requirement you seem to have specified – convince me this is what you actually need. BTW, functionally why not invoke the independency (Circle I) on the FOS? Seems this specification may be more applicable for the functional requirements to me.

Question #2 – Looks good to me.

Question #3 – Yes, datum C is a derived axis from the cylinder and will locate the center of the keyway FOS center-ish.

Question #4 – Yes, the Datum axis A-B will center the splines. I assume that both datum’s A and B are bearing locating surfaces in the application assembly? It’s correct to specify which feature of the spline the position tolerance applies to however I don’t think the parentheses are required.

Question #5 – Looks good to me however you might consider some reference information for the manufacturing folks. If this part is manufactured on a vertical mill, not CNC extra data is helpful. Therefore if you extend the centerline of the angled hole to the intersection of the datum axis D and reference dimension it manufacturing might find that data useful – just ask a see what they say. Also, intersecting holes are going to be rough and burr prone - a note specifying what is acceptable should be on the engineering drawing. Also, the horizontal drilled hole you intersect with needs to have a drill point.


These are not “Rookie“ questions nor applications, just saying.

[jlenth]

Kelly,


Thanks for the response and your input, I appreciate it. Responses to your comments below and I have another question that popped up in my mind as I get closer to finishing with detailing my actual drawings.

Question #1)-In the real life application, specified cylindricity is actually .00035 for .0007 local and the dimensional tolerance remains at .0005, so it's not as extreme as my original post.I arbitrarily used .0007 cylindricity in my example just to make sure I drew attention to the fact that I was violating rule #1 of the ASME Y14.5 standard. As far as functional requirement and convincing you, very long design history behind this part and mating pieces that are well beyond the scope of this forum and nor could I state the specific details publicly.But ultimately this specification is for manufacturing purposes.Due to the size, shape, and stresses in the actual part from welding (we are even annealing these after welding), this would be an extremely difficult part to manufacture and hold the boundary conditions per rule #1 or paragraph 2.7.1 in the standard.Scrape rate would be very high.We've been making this part for years; we're in the process of turning this into a casting and so I'm updating/adding additional GD&T while in the process.Our experience in manufacturing of this part overall roundness on average runs about 0.0004-.0007" while occasionally worse resulting in rejection.We know from over two years of designing and testing this particular part that we need to maintain a much tighter than normal control on the bore size and thus this is where the .0005 dimensional tolerance comes.Bearing manufacture is comfortable with the bearing going into a bore with up to .0007" overall out of roundness, so we are accepting this out of roundness to keep scrap rate down.Additionally, the bearing manufacture's requirement for an acceptable bearing seat/bearing bore does not follow rule #1 either.Their requirement is such that the average of four measurements evenly spaced on a single plane to be within dimensional tolerance, with three planes of measurements taken across the bearing seat/bore (12 pt. measurement system).Roundness of measurements on each plane cannot exceed .0007" (for our particular application) nor there be any more than .0007" of difference between the average diameters on the three planes.This is the criteria we’ve been using to pass/fail these bores since the beginning of manufacturing with the exception of being a little stricter on the cylindricity.

Independency-I was considering this route as well, but I thought the AVG was the better route due to the bearing manufacture’s criteria for an acceptable bearing seat/bore as described above??? I also have the measurement process stated on the drawing to achieve the “average” diameter.I’m not sure if I have a true 100% understanding of the independency function.If I use the independency function, this releases the feature of size from the boundaries at both the MMC and LMC correct?Or does it only release the boundary condition at MMC?Please do let me know your thoughts on AVG vs. Independency in this application after this additional background info.

Question #4)-Yes, datum’s A & B are the bearing seats and you’re correct on parenthesis not being required

Question #5)-Everything is done on CNC and will definitely ask manufacturing about the extended centerline of the angled hole.Everything else has already been handled on my actual part/drawing.Please keep in my mind examples are just quick models to help convey my questions and not drawn for complete accuracy; in real life my colleagues actually accuse me of being overly detailed….

Question #6), Simultaneous Requirement-per my “Example 2” attachment, I have a multiple single segment positional tolerance on the 1/4-20 tapped holes.As a pattern, relationship of these tapped holes is important to the tertiary datum C, but doesn’t require a tight tolerance.However, a much tighter positional tolerance is required in relationship to datum plane A and datum axis B.Relationship of the Ø.250 and Ø.125 hole is also important in relationship to the ¼-20 tapped hole pattern, datum plane A, datum axis B, and I would like all these holes to be treated as a single pattern.Since the datums in the feature control frames for the Ø.250 and Ø.125 holes are the same, in same order of precedence, and of the same boundary condition (4.19 & 7.5.4.1 in 2009 standard) to the lower feature control frame of the tapped holes, does this invoke simultaneous requirement?Or do I need add the upper feature control frame of the tapped holes to the Ø.250 and Ø.125 holes to invoke this requirement?

Again, thanks for the help and input.

[Kelly_Bramble]

In the real life application, specified cylindricity is actually .00035 for .0007 local and the dimensional tolerance remains at .0005, so it's not as extreme as my original post.

.0002” is not what I would call generous but I’m sure manufacturing is appreciative.

Please do let me know your thoughts on AVG vs. Independency in this application after this additional background info.

The functional and E.O.L. requirements of the installation/bearing is the stop-gate specification. I think that the tolerance structures required regardless of the design approach are achievable with an optimal manufacturing sequence. I'm sure you have considered a post weldment machining operation to create the desired size/form as well as surface roughness. One could argue this component requires post-welding turning then grinding to achieve the tolerance specified --> + more $$$.


As I have indicated above and in other words the outer boundary size and form of the cylinder need be tolerance matched to the bearing - If your sole purpose is to let a little off the death grip of Rule #1 then both approaches (AVG and Circle I) will likely get you there.


Question #6 – Since you need a position (location) tolerance control to Datum B for the .25 threaded holes, that leaves out composite positional control (orientation only)… The way that I view simultaneous requirement is that the concept allows for a separate requirement concept. Now let me explain… If the Datums and order of precedence are identical and applied with position and profile tolerances – then the features are dimensionally verified in the same DRF setup. As shown in Y14,5-2009, Figure 4-40, the keyways are measured/verified with the same DRF setup even though there is not a rotational control datum between these features. This fundamental requirement ensures that the features will be aligned within the tolerance variations specified - concurrently.

If we specify “SEP REQT” then the alignment between the keyseats is not required. We would not be required to measure the feature tolerances concurrently. Therefore, the keyseats could be not aligned – even at 180 degrees opposite of each other regardless of the engineering drawing illustrations.


So… the Dia. 125 hole is implicitly oriented back to datum C by way of the lower FCF specified on the .25-20 threaded holes. This is because the tolerance verification for the lower FCF and the .125 hole is done “simultaneous” and the upper FCF includes control for the threads back to Datum C.

Or do I need add the upper feature control frame of the tapped holes to the Ø.250 and Ø.125 holes to invoke this requirement?

No… Lot of words to get to this though.

The “SEP REQT” concept is not universally understood BTW..

I’ll be back to re-read this and see if I messed up later…

[jlenth]

Kelly,


Hopefully you can provide your insight on one last question, total runout. I’ve seen other posts on this same topic, but haven’t seen anything that came out and specifically said what I want to confirm.In the event that the allowable runout tolerance is greater than the dimensional tolerance, is it a waste to specify total runout as per my attached example?

I understand that circular runout controls circularity and coaxiality while total runout controls circularity, coaxiality, straightness, and taper. Referring back to my example, the dimensional tolerance of ±.001 is going to be the limiting control in this case for circularity and taper; thus the only thing the total runout is going to control is the coaxiality/straightness of the specified feature relative to datum axis A-B correct?And therefore I could just as well put a circular runout in place of the total runout as the circular runout would be doing the exact same in this particular case correct?

You’ve been a big help in help clarifying some of these grey areas for me. Thanks

[Kelly_Bramble]

Since there is a static relationship between Datum A-B and the form (Circularity, Cylindricity, Straightness) is controlled by the limits of size, the runout specification is simply a location control.

A circular runout would control the location however in this case I would recommend you analyze and consider a position tolerance.

I do agree that a Circular Runout is likely adequate though I don't know the tolerance/alignment/fit requirements of the end-item.

[jlenth]

Kelly,

It’s “implied” that when circular runout is called out on a drawing that the circular runout is theoretically applied to an infinite number of 2-D cross sectional planes along the length of the surface correct? This is the fancier way of ASME saying “each circular element” as they did in figures 9-2 & 9-5 of the 2009 standard?



I’m guessing you are suggesting the positional tolerance to avoid any misinterpretation/confusion that a machinist may have on circular runout callout such as I am?



Thanks

[jlenth]

Kelly,

Please disregard my above questions, due to lack of sleep my brain hasn't been firing on all cylinders. Thanks for all help