I literally just saw you in our GD&T training that we have in our learning portal at work. You are the best instructor I have ever had out of the 5X classes I have taken.
Thanks for another great video. So, you're basically saying that plus-minus size tolerance and position are appropriate to those portions of the shaft where relatively accurate fit takes place to the mating parts, so that size deviation is important, and on the other hand - where clearance is larger and location/coaxiality and size matter just the same, profile is better. Another scheme that is also recommended often where there is fit with relatively large clearance is a loose size tolerance and position at MMC to protect a VC boundary that will prevent interference. That combination also connects in a way the effects of size and location on the outcome of the position evaluation. But I think there's a good reason you chose profile in this case rather than loose size and position at MMC - and that's because it's a rotating shaft and profile also limits the runout - that's important for balance. The loose size and position at MMC scheme would be poor at preventing balance issues and the designer would probably need to add runout requirements - resulting in an overall too many requirements that can simply be replaced by one profile of a surface specification relative to the applicable datum. Is my interpretation correct?
Yes I think that makes sense. The profile tolerance gives two hard boundaries: MMB and LMB. Position at MMC gives only one hard boundary of MMB (yes, also called VC). The LMB (also called resultant condition) gets bad with the position at MMC. If clearance is all that matters, then position at MMC works great (who cares about the LMB).
1st of all, I’d like to thank you for your excellent videos with details. 2nd, I’m happy to see you are still looking healthy so hopefully you will continue to provide more useful videos. ❤
It is a great video. I do have the following comments: At 1:55, for a better comparison (same Inner and Outer Boundaries) between the Position and the Profile controls, I think the Profile tolerance should have been 0.35, instead of 0.2. In addition, I have not heard how fast this shaft spins in the assembly. If the spin speed is relatively high, then I would recommend that even the "clearance" features should be controlled for dynamic balancing, with Concentricity (◎) before 2018 (1994 or 2009 standard), or with Profile of a Surface + the new Dynamic Modifier (2018 standard). I hope this makes sense. I did not want to write an essay explaining.
Good comments. The position and profile example were not meant to have the same boundaries. Generally, I've found the best applications for position are a tight size tolerance (for fit) and a looser position tolerance (for coaxial). I agree that if this is spinning at a higher speed, a control that keeps the surface centered may be needed. I would then keep the loose profile tolerance and add a circular runout. Each cross-section is evaluated separately because 2D surface evaluation is all that is needed for the balance. Dynamic profile is tighter than needed because it all has to be evaluated at the same time. Concentricity has too many interpretations and misconceptions around it to be useful.
I like this video. It shows the basic points that a designer should pay attention to. I'm just wondering: Are bonus tolerances possible now? ("M" for fits).
For the profile tolerance, what if we use size tol + position tol ? I think it should get better control than only using profile tolerance when you need unequal tolearnce for the size and positon
Whether you want to group the size and location with a profile tolerance or separate the tolerances into size and position depends on the application. That is the point of the video. You do not always want to do one or the other.
You want all features located to the same datum reference frame. Even the datum features A and B are located to the common datum A-B. Datum features A and B do not establish the datum by themselves so they need relationships to the A-B also.
Although a great and functionally accurate definition based on the function of the part, I would think that only very high end manufactures with high end inspection equipment and knowledge, plus a significant order quantity would have the capability and time to forensically appreciate all these features. I have problems with the shaft datums A-B ,C as these are not directly referenced in the part actual manufacture. From an inspection point of view, exactly how is A-B established over these surfaces allowing the other features to be measured from during a 360 deg part revolution? I think they're not in both manufacturing and Inspection but rather achieved using standard long shaft machining/ Inspection techniques where more careful approaches need be made on accurate faces / features. So I understand GDT theory is required to define the parts function, but from this parts manufacturing and inspection perspective, there needs to be a different approach to suit how it's to be made and Inspected. I can imagine this shaft to require machining centers to be necessary either end with those being also used in Inspection with a series of electronic probes monitoring face deviations ( perhaps with clever relative outputs) during a single revolution. Of course, highly experienced shaft manufacturers would see similar definitions and know how adapt their manufacture and Inspection techniques to meet the drawing spec. My view of the thread precision is that it's needs to be very accurate ( and naturally it will be anyway) as the nut when locked, has a firm radial contact on both the shaft thread and the bearing race due to friction from the pressure of locking the nut. Designers deciding tolerances, especially GDT need much manufacturing and assembly experience in my view. What's the point of specifying a feature can be very loose, when not specifying it to be highly controlled actually means the same thing but with less fuss and notes on the details that need skill / knowledge to decipher. The only singular advantage that I can see is that the specifier really 'Thinks' about what is required for proper part functionality, which is imperative. It seems to me that GDT is necessary, but unnecessary GDT may induce additional but unwarranted manufacturing thought and inspection increasing the cost especially in early quoting. Despite these comments, I appreciate these video's and I'm learning from them.
Thanks for the comments. The tolerancing scheme, including the datum feature selection, doesn't require a method of manufacturing. It only defines the functional requirements. The requirements of this part are: when the part mounts on the bearing journals, all features must be sized and coaxial within XXX, some tighter than others. As you said, manufacturing will probably cut all features at the same time using machining centers. Quality has to ensure the functional requirements are met, which could be done in many ways: One method is a set of vee blocks, micrometers, and a dial indicator. Another method is a CMM. One plan may measure only tight tolerances and leave the looser ones unmeasured but to be met by monitoring the process control data (don't measure any tolerances over .010). Another quality plan may measure everything during first article but then only monitor a few critical tolerances during production. The point is: the tolerancing doesn't tell you how to manufacture or inspect, it only defines the functional limitations of the product. This gives both accountability and traceability when the product has variation. Don't have a design engineer define manufacturing and inspection methods. Design should let manufacturing and quality use their engineering expertise to develop their own plans that meet the clear functional intent.
@@GeoTolPro Thanks for the in-depth reply. As an apprentice over forty years ago now, I did stints in the methods and QA departments. We had a 'route card' dictating the the method of manufacture and other tolerances not on the drawing and then what inspection had to check or not etc. So everything very similar to what you have said and thank you again for the extensive reply. My concern is really for the smaller job shop that just hasn't got the expertise to interpret such drawings. Having said that the methodology certainly makes sense and the specifying of GDT seems quite straightforward and entirely logical and right now but what manufacturing needs in many cases can be entirely different to what's on the drawing as you have said in a few video's now. It probably comes down to difference between common ordinary job shops verses professional manufacturing. I've always taken great care in detailing my designs from the makers perspective as I can't afford untimely delays or misinterpretations...I will change a little.
I literally just saw you in our GD&T training that we have in our learning portal at work. You are the best instructor I have ever had out of the 5X classes I have taken.
That's awesome. Thanks for the compliment!
This is the first time I see someone explain a seal surface control! Thanks!
Great example and explanation. Please keep posting new videos.
Thanks for another great video.
So, you're basically saying that plus-minus size tolerance and position are appropriate to those portions of the shaft where relatively accurate fit takes place to the mating parts, so that size deviation is important, and on the other hand - where clearance is larger and location/coaxiality and size matter just the same, profile is better. Another scheme that is also recommended often where there is fit with relatively large clearance is a loose size tolerance and position at MMC to protect a VC boundary that will prevent interference. That combination also connects in a way the effects of size and location on the outcome of the position evaluation. But I think there's a good reason you chose profile in this case rather than loose size and position at MMC - and that's because it's a rotating shaft and profile also limits the runout - that's important for balance. The loose size and position at MMC scheme would be poor at preventing balance issues and the designer would probably need to add runout requirements - resulting in an overall too many requirements that can simply be replaced by one profile of a surface specification relative to the applicable datum. Is my interpretation correct?
Yes I think that makes sense. The profile tolerance gives two hard boundaries: MMB and LMB. Position at MMC gives only one hard boundary of MMB (yes, also called VC). The LMB (also called resultant condition) gets bad with the position at MMC. If clearance is all that matters, then position at MMC works great (who cares about the LMB).
1st of all, I’d like to thank you for your excellent videos with details. 2nd, I’m happy to see you are still looking healthy so hopefully you will continue to provide more useful videos. ❤
It is a great video. I do have the following comments:
At 1:55, for a better comparison (same Inner and Outer Boundaries) between the Position and the Profile controls, I think the Profile tolerance should have been 0.35, instead of 0.2.
In addition, I have not heard how fast this shaft spins in the assembly. If the spin speed is relatively high, then I would recommend that even the "clearance" features should be controlled for dynamic balancing, with Concentricity (◎) before 2018 (1994 or 2009 standard), or with Profile of a Surface + the new Dynamic Modifier (2018 standard). I hope this makes sense. I did not want to write an essay explaining.
Good comments. The position and profile example were not meant to have the same boundaries. Generally, I've found the best applications for position are a tight size tolerance (for fit) and a looser position tolerance (for coaxial).
I agree that if this is spinning at a higher speed, a control that keeps the surface centered may be needed. I would then keep the loose profile tolerance and add a circular runout. Each cross-section is evaluated separately because 2D surface evaluation is all that is needed for the balance. Dynamic profile is tighter than needed because it all has to be evaluated at the same time. Concentricity has too many interpretations and misconceptions around it to be useful.
Clear concept thanks sir
I like this video. It shows the basic points that a designer should pay attention to.
I'm just wondering: Are bonus tolerances possible now? ("M" for fits).
Since all the fitting features are "press fit" or "need to be centered", the MMC modifier isn't applicable.
this is amazing
For the profile tolerance, what if we use size tol + position tol ? I think it should get better control than only using profile tolerance when you need unequal tolearnce for the size and positon
Whether you want to group the size and location with a profile tolerance or separate the tolerances into size and position depends on the application. That is the point of the video. You do not always want to do one or the other.
Hi, just want to know that can we call datum A-B to feature having datum A.
In your example datum A feature dimensions FCF have A-B.
Please.... guide.
You want all features located to the same datum reference frame. Even the datum features A and B are located to the common datum A-B. Datum features A and B do not establish the datum by themselves so they need relationships to the A-B also.
Tanks a lot for your support
Amazing
Although a great and functionally accurate definition based on the function of the part, I would think that only very high end manufactures with high end inspection equipment and knowledge, plus a significant order quantity would have the capability and time to forensically appreciate all these features. I have problems with the shaft datums A-B ,C as these are not directly referenced in the part actual manufacture. From an inspection point of view, exactly how is A-B established over these surfaces allowing the other features to be measured from during a 360 deg part revolution? I think they're not in both manufacturing and Inspection but rather achieved using standard long shaft machining/ Inspection techniques where more careful approaches need be made on accurate faces / features.
So I understand GDT theory is required to define the parts function, but from this parts manufacturing and inspection perspective, there needs to be a different approach to suit how it's to be made and Inspected. I can imagine this shaft to require machining centers to be necessary either end with those being also used in Inspection with a series of electronic probes monitoring face deviations ( perhaps with clever relative outputs) during a single revolution. Of course, highly experienced shaft manufacturers would see similar definitions and know how adapt their manufacture and Inspection techniques to meet the drawing spec.
My view of the thread precision is that it's needs to be very accurate ( and naturally it will be anyway) as the nut when locked, has a firm radial contact on both the shaft thread and the bearing race due to friction from the pressure of locking the nut.
Designers deciding tolerances, especially GDT need much manufacturing and assembly experience in my view.
What's the point of specifying a feature can be very loose, when not specifying it to be highly controlled actually means the same thing but with less fuss and notes on the details that need skill / knowledge to decipher.
The only singular advantage that I can see is that the specifier really 'Thinks' about what is required for proper part functionality, which is imperative.
It seems to me that GDT is necessary, but unnecessary GDT may induce additional but unwarranted manufacturing thought and inspection increasing the cost especially in early quoting.
Despite these comments, I appreciate these video's and I'm learning from them.
Thanks for the comments. The tolerancing scheme, including the datum feature selection, doesn't require a method of manufacturing. It only defines the functional requirements. The requirements of this part are: when the part mounts on the bearing journals, all features must be sized and coaxial within XXX, some tighter than others. As you said, manufacturing will probably cut all features at the same time using machining centers.
Quality has to ensure the functional requirements are met, which could be done in many ways: One method is a set of vee blocks, micrometers, and a dial indicator. Another method is a CMM. One plan may measure only tight tolerances and leave the looser ones unmeasured but to be met by monitoring the process control data (don't measure any tolerances over .010). Another quality plan may measure everything during first article but then only monitor a few critical tolerances during production.
The point is: the tolerancing doesn't tell you how to manufacture or inspect, it only defines the functional limitations of the product. This gives both accountability and traceability when the product has variation. Don't have a design engineer define manufacturing and inspection methods. Design should let manufacturing and quality use their engineering expertise to develop their own plans that meet the clear functional intent.
@@GeoTolPro Thanks for the in-depth reply. As an apprentice over forty years ago now, I did stints in the methods and QA departments. We had a 'route card' dictating the the method of manufacture and other tolerances not on the drawing and then what inspection had to check or not etc. So everything very similar to what you have said and thank you again for the extensive reply. My concern is really for the smaller job shop that just hasn't got the expertise to interpret such drawings. Having said that the methodology certainly makes sense and the specifying of GDT seems quite straightforward and entirely logical and right now but what manufacturing needs in many cases can be entirely different to what's on the drawing as you have said in a few video's now. It probably comes down to difference between common ordinary job shops verses professional manufacturing. I've always taken great care in detailing my designs from the makers perspective as I can't afford untimely delays or misinterpretations...I will change a little.
Coaxial.