I need help in understanding this issue regarding simulating a center plane from a datum feature related to an external width dimension: let's say a width dimention is defined as a primary datum feature. When using a vise-like physical datum feature simulator with two almost parallel faces that close on the part, unless the tangent planes on both sides of the datum feature are perfectly parallel (and in the real world they're not), one of the vise faces will act similar to a primary datum plane - touching on 3 high points, and the opposite face will touch on only one point, similary to a tetriary datum plane. Now, depending on which side of the datum feature will make the more stable contact with the simulator, we might get a different separation width between the vise faces, and therefore the simulated datum plane will also be different. For example, if the measurement set up has the vise faces oriented horizontally, the side of the datum feature facing down will orient the part in the fixture, and if you flip the part upside down for a repeated measurement you might get different results on whatever control called out that datum. Now, I understand that there is only one "actual mating envelope" to the datum feature per ASME and only one of the sides facing down will produce the "minimum separation" condition per fig. 4-13, But that means that you have to mount the part twice in the fixtute and re-check your results, and I somehow doubt that this is the recommended practice... on the other hand, if the vise is oriented vertically, we will have no control over which side we stabilize better in the simulator - which is even worse. Everyone's insight will be much appreciated... Thank you!
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Simulating a datum center plane 2
- Thread starter semiond
- Start date
- Thread starter
- #24
3DDave, if with the current dimensions both inspection methods are legitimate, then i'm afraid i have to insist that the drawing is ambiguous, because as you can see two different center planes can be simulated - one is in the middle of the 4.75 separation width, and the other is in the middle of the 4.71 width. That's two different possible datums. Orientation is only part of the issue: If i want to "center" the smaller width to the datum feature width by a positional tolerance, then the true position of the controlled width can be in two different locations - in other words the location of the tolerance zone is allowed to be either there or there. That's not what supposed to be in a completely defined drawing.
- Thread starter
- #25
pmarc, i read your reply, sorry for not responding earlier. From reading the standard i came to the same conclusion - that only method #2 is correct in this example. But it seems to me that it's way too easy to make a wrong simulation. Like i said in my first post the only way to avoid the mistake when using a physical datum feature simulator seems to be double checking and doing the flipping of the part each time when simulating a center plane. And i really doubt that this is practised, especially in places where mass production is the deal and time consuming activities are not appreciated. I haven't seen this issue directly adressed anywhere in the Y14.5 or Y14.43 standards, and i did expect it to be mentioned somewhere,because the issue with the 3 contact points on one side of the simulator, and one contact point on the opposite, while sides matter, is very obvious. But since i'm relatively new to (correct) GD&T and have little practical experience with measurement equipment, i thought that either there might be a widely accepted datum simulation practice that i'm not aware of, or i don't completely understand the concept of simulating a center plane using a physical datum feature simulator...and that is why i opened this thread.
-
1
- #26
semiond - the condition is clear - the inspector can find a part location and orientation such that all characteristics are met, if there is such a condition available. It is clear that a part can be made to meet those requirements. That there is some other case is immaterial.
An ambiguous requirement is one where it is not possible to determine what parts can be accepted. That is not the case for this feature. In your example you show a particular part with and acceptable parallelism.
There is already guidance in the standard for dealing with nominally symmetric parts and the requirement for identification, which you don't feel the need to follow. It's in the first paragraph on the topic of Datum Features; 4.8 for the '2009 version.
If you are unhappy with this, and it seems you are, you will need to try a scheme involving datum targets. This will produce a different set of problems, but the ones in this case are all dealt with, so new ones are the only way out.
An ambiguous requirement is one where it is not possible to determine what parts can be accepted. That is not the case for this feature. In your example you show a particular part with and acceptable parallelism.
There is already guidance in the standard for dealing with nominally symmetric parts and the requirement for identification, which you don't feel the need to follow. It's in the first paragraph on the topic of Datum Features; 4.8 for the '2009 version.
If you are unhappy with this, and it seems you are, you will need to try a scheme involving datum targets. This will produce a different set of problems, but the ones in this case are all dealt with, so new ones are the only way out.
- Thread starter
- #27
3DDave, i'm already familier with paragraph 4.8:
"Datum features must be
readily discernible on the part. Therefore, in the case
of symmetrical parts or parts with identical features,
physical identification of the datum feature on the
part may be necessary."
That is not the case in my example, because even though the part is symmetrical, the datum is the center plane of the 5mm width feature. There is only one such feature on the part and it doesn't need marking or identification, only a correct practise of simulating a datum out of it is what's needed.
Maybe you meant that there are several options for selecting the surface to control for parallelism - i agree it is not "readily discrenible". If the part's geometry was exactly like in the sketch it would be up to the inspector to make a choice what face to check and find one that passes, which is OK by me. But as i already said it is my bad for not making a small chamfer at the end of that face to identify it better and not letting this matter to distract from the true problem. Notice how in both methods the same face is being checked. See it's location relative to "SIDE A" and "SIDE B" in both measurements. The part was only flipped around X axis, and the same face was checked for parallelism. Think of it as if it was carrying an identification mark saying: "check parallelism here".
So, the symmetrical geometry is not the cause for the difference in results. It is only the fact that there are two (at least) options to simulate the datum from the datum feature using a vise as physical simulator. When you don't have the time to deeply analyze everything (and in the inspection department of a large plant you don't) both options seem correct, but only one truly is. See how it is described in fig. 4-13 in ASME Y14.5 -"parallel planes at minimum separation". Like pmarc rightly noted: In the example, only when the distance between the simulator faces is 4.71, this condition is met. But there seems to be no way to know ahead how to simulate, and the inspector has to rely on trial and error. The chance to make a wrong datum plane simulation seems to be too great (unless simulating the datum at least twice and comparing results is always required - but it is not mentioned anywhere). This can be an open door for trouble and disagreements between manufacturers and customers.
Thank you for the suggestion about datum targets, they can indeed solve some of the ambiguity. I guess it is possible to define a center plane using datum targets in several ways. But as you say that might introduce other problems, and it is still the subject of a center plane defined by a width primary datum feature that i'm trying to understand thoroughly. I thought that since it is commonly practised, there might be common solutions to the problem i described, other than alternative definitions.
"Datum features must be
readily discernible on the part. Therefore, in the case
of symmetrical parts or parts with identical features,
physical identification of the datum feature on the
part may be necessary."
That is not the case in my example, because even though the part is symmetrical, the datum is the center plane of the 5mm width feature. There is only one such feature on the part and it doesn't need marking or identification, only a correct practise of simulating a datum out of it is what's needed.
Maybe you meant that there are several options for selecting the surface to control for parallelism - i agree it is not "readily discrenible". If the part's geometry was exactly like in the sketch it would be up to the inspector to make a choice what face to check and find one that passes, which is OK by me. But as i already said it is my bad for not making a small chamfer at the end of that face to identify it better and not letting this matter to distract from the true problem. Notice how in both methods the same face is being checked. See it's location relative to "SIDE A" and "SIDE B" in both measurements. The part was only flipped around X axis, and the same face was checked for parallelism. Think of it as if it was carrying an identification mark saying: "check parallelism here".
So, the symmetrical geometry is not the cause for the difference in results. It is only the fact that there are two (at least) options to simulate the datum from the datum feature using a vise as physical simulator. When you don't have the time to deeply analyze everything (and in the inspection department of a large plant you don't) both options seem correct, but only one truly is. See how it is described in fig. 4-13 in ASME Y14.5 -"parallel planes at minimum separation". Like pmarc rightly noted: In the example, only when the distance between the simulator faces is 4.71, this condition is met. But there seems to be no way to know ahead how to simulate, and the inspector has to rely on trial and error. The chance to make a wrong datum plane simulation seems to be too great (unless simulating the datum at least twice and comparing results is always required - but it is not mentioned anywhere). This can be an open door for trouble and disagreements between manufacturers and customers.
Thank you for the suggestion about datum targets, they can indeed solve some of the ambiguity. I guess it is possible to define a center plane using datum targets in several ways. But as you say that might introduce other problems, and it is still the subject of a center plane defined by a width primary datum feature that i'm trying to understand thoroughly. I thought that since it is commonly practised, there might be common solutions to the problem i described, other than alternative definitions.
I'm sorry they don't have a paragraph with your exact case. Since you are familiar with that paragraph you are also familiar with the reason and expected outcome and therefore will be marking the parts in the spirit of the suggestion.
Also, try not to edit every single post you make, particularly without any clue as to why you changed the text. It makes your involvement look unreliable. Like your inspectors, I don't have time to go back and see what the intention was.
Also, try not to edit every single post you make, particularly without any clue as to why you changed the text. It makes your involvement look unreliable. Like your inspectors, I don't have time to go back and see what the intention was.
- Thread starter
- #29
3DDave, don't bother to go back looking at unedited versions of my posts. If i edited a post, what i wanted to say is best expressed in the last version. If it matters to you, i go over my posts and correct things mostly because English is not my native language, and sometimes i add something to the content if find it useful for coneying my thoughts. I accept your advice not to do it, if it makes my posts look suspicious.
Anyway if i'm still allowed to comment on the sunject matter, i don't think that the problem i presented is a private case relevant to a specific part. I believe it is relevant to every part drawing where a center plane is called out as the primary datum feature, and might be of interest to many other people as well. My question was solely about simulation of a primary datum center plane. But i also took notice of your suggestion to mark parts, thank you.
Anyway if i'm still allowed to comment on the sunject matter, i don't think that the problem i presented is a private case relevant to a specific part. I believe it is relevant to every part drawing where a center plane is called out as the primary datum feature, and might be of interest to many other people as well. My question was solely about simulation of a primary datum center plane. But i also took notice of your suggestion to mark parts, thank you.
pmarc
Mechanical
- Sep 2, 2008
- 3,247
semiond,
If part's symmetry dillemma is taken out of the equation (as your last post seems to imply), then I do not really understand your issue with simulation of primary datum center plane. Does not Y14.5 standard give all necessary information on how to deal with primary datum widths at RMB (also with those that can rock when brought in contact with its datum feature simulator) then?
First, there is the para. 4.11.4(b) and figs. 4-13 and 4-14 that say/show that the parallel planes of the datum feature simulator must be at minimum seperation. This basically means the inspector is obligated to make sure that he did everything that was possible to bring the two planes as close as possible to each other - in your example this would result in rejection of inspection method #1.
But even then, as 3DDave suggested, there could be a situation where for minimum possible distance between planes (4.71 in your example) different orientations of datum feature relative to the simulator could be possible. For those cases there is the para. 4.11.2, mentioned at the beginning of the discussion by J-P, essentially saying that inspector is allowed to make adjustments of the datum feature in the simulator, and as long as there is at least one datum feature-to-simulator configuration (aka canditate datum set) that leads to positive assesment of a geometric control (parallelism in you example), everything is according to the rules.
Could you help me understand what I am missing? Because at the moment it looks to me that your problem has more to do with lack of inspector skills and knowledge than with lacks in the standard.
Side note: It seems like in the future version of Y14.5, the default adjustment procedure will not be candidate datum set any more. Para. 7.11.2 in the draft states:
"If irregularities on a datum feature are such that the part is unstable (that is, it rocks) when brought into contact with the corresponding true geometric counterpart, the default requirement is that the part be adjusted to a single solution that minimizes the separation between the feature and the simulator per ASME Y14.5.1. If a different procedure is desired (candidate datum set, Chebychev, least squares, translational least squares, etc.), it must be specified"
If part's symmetry dillemma is taken out of the equation (as your last post seems to imply), then I do not really understand your issue with simulation of primary datum center plane. Does not Y14.5 standard give all necessary information on how to deal with primary datum widths at RMB (also with those that can rock when brought in contact with its datum feature simulator) then?
First, there is the para. 4.11.4(b) and figs. 4-13 and 4-14 that say/show that the parallel planes of the datum feature simulator must be at minimum seperation. This basically means the inspector is obligated to make sure that he did everything that was possible to bring the two planes as close as possible to each other - in your example this would result in rejection of inspection method #1.
But even then, as 3DDave suggested, there could be a situation where for minimum possible distance between planes (4.71 in your example) different orientations of datum feature relative to the simulator could be possible. For those cases there is the para. 4.11.2, mentioned at the beginning of the discussion by J-P, essentially saying that inspector is allowed to make adjustments of the datum feature in the simulator, and as long as there is at least one datum feature-to-simulator configuration (aka canditate datum set) that leads to positive assesment of a geometric control (parallelism in you example), everything is according to the rules.
Could you help me understand what I am missing? Because at the moment it looks to me that your problem has more to do with lack of inspector skills and knowledge than with lacks in the standard.
Side note: It seems like in the future version of Y14.5, the default adjustment procedure will not be candidate datum set any more. Para. 7.11.2 in the draft states:
"If irregularities on a datum feature are such that the part is unstable (that is, it rocks) when brought into contact with the corresponding true geometric counterpart, the default requirement is that the part be adjusted to a single solution that minimizes the separation between the feature and the simulator per ASME Y14.5.1. If a different procedure is desired (candidate datum set, Chebychev, least squares, translational least squares, etc.), it must be specified"
- Thread starter
- #31
pmarc, first of all you are right, if i was a very skillful and knowledgeable at inspection techniques i would probably not be asking this question on the forum.
But since the company i work at does not have a great knowledge base in application of GD&T "by the rules", i will probably have to be the one explaining the "how to measure" issue to inspection people if i decide to follow the standard in my drawings - and this is why i asked for guidance.
"My problem" as you put it, is with the technical difficulty to simulate the "minimum separation" condition per paragraph 4.11.4 (b), and fig. 4-13. Seems like practicing it with a physical datum feature simulator requires trial and error. Just like you said: " This basically means the inspector is obligated to make sure that he did everything that was possible to bring the two planes as close as possible to each other - in your example this would result in rejection of inspection method #1." From my experience in the manufacturing industry, this can not work in practice - unless there is some easy and clear verification methodology that i was hoping to learn about when opening the thread. Without that, if the inspector finds a condition where the feature seems to be in tolerance like shown in method #1 in the sketch, he will immediately approve the part. He will not be looking in the standard for figures and paragraphs to make sure he has covered all requirements. As you have seen from the replies in this thread, even knowledgable people in GD&T would approve the part according to method #1 and support the inspector's position. Isn't it a clear indicator that there truely is something to discuss here?
As for para. 4.11.2 - it talks about parts that rock when brought in contact with the datum feature simulator. I was not talking about instability situations or parts that rock. I was talking about a situation where a part seems to sit stable in the inspection fixture, but the datum simulation is not correct, and the error is too easy to miss. There is difference.
I do realize that what i'm talking about is not neceserally an issue with the standard (which is more about theory then practice) and i'm not here to critisize the standard or anything like that. My sole intention was to learn - either to be pointed at something i missed in the standard or learn from other people's experience, i'm sure someone encountered the same problem.
Regarding the quote in your side note, that seems like good news - a more practical approach to unstable situations. Thank you for bringing this up. I would hate to have to explain anyone about a "candidate datum set" when requested to address issues in production and inspection
I surely hope i brought my point across better this time and sorry for being stubborn.
But since the company i work at does not have a great knowledge base in application of GD&T "by the rules", i will probably have to be the one explaining the "how to measure" issue to inspection people if i decide to follow the standard in my drawings - and this is why i asked for guidance.
"My problem" as you put it, is with the technical difficulty to simulate the "minimum separation" condition per paragraph 4.11.4 (b), and fig. 4-13. Seems like practicing it with a physical datum feature simulator requires trial and error. Just like you said: " This basically means the inspector is obligated to make sure that he did everything that was possible to bring the two planes as close as possible to each other - in your example this would result in rejection of inspection method #1." From my experience in the manufacturing industry, this can not work in practice - unless there is some easy and clear verification methodology that i was hoping to learn about when opening the thread. Without that, if the inspector finds a condition where the feature seems to be in tolerance like shown in method #1 in the sketch, he will immediately approve the part. He will not be looking in the standard for figures and paragraphs to make sure he has covered all requirements. As you have seen from the replies in this thread, even knowledgable people in GD&T would approve the part according to method #1 and support the inspector's position. Isn't it a clear indicator that there truely is something to discuss here?
As for para. 4.11.2 - it talks about parts that rock when brought in contact with the datum feature simulator. I was not talking about instability situations or parts that rock. I was talking about a situation where a part seems to sit stable in the inspection fixture, but the datum simulation is not correct, and the error is too easy to miss. There is difference.
I do realize that what i'm talking about is not neceserally an issue with the standard (which is more about theory then practice) and i'm not here to critisize the standard or anything like that. My sole intention was to learn - either to be pointed at something i missed in the standard or learn from other people's experience, i'm sure someone encountered the same problem.
Regarding the quote in your side note, that seems like good news - a more practical approach to unstable situations. Thank you for bringing this up. I would hate to have to explain anyone about a "candidate datum set" when requested to address issues in production and inspection
I surely hope i brought my point across better this time and sorry for being stubborn.
pmarc
Mechanical
- Sep 2, 2008
- 3,247
semiond,
Let me start from the end.
. You ask good, thought-provoking questions and based on your answers I would not say that you are "relatively new to (correct) GD&T".
In my opinion the concern you are having applies to many other GD&T concepts too. This is classic example of theory vs. reality dillemma. There is theoretical definition and there is reality where inspection tries (or at least should try) to get as close to the theory as possible. Factors like time pressure, cost limitations, equipment availability, lack of knowledge definitely do not help to achieve that goal, but to me this should not be a reason to say that "this issue is what stops me from defining a center plane as a primary datum in a part, where functionally it appears correct".
Let me start from the end.
You do not have to apologize and you are not stubborn... well, you are but in a good waysemiond said:
. You ask good, thought-provoking questions and based on your answers I would not say that you are "relatively new to (correct) GD&T". I am not sure that this is more practical approach to unstable situations. Most likely it will make some (most?) people's life easier (provided that the single solution method will be well explained in future version of Y14.5.1), but I am not sure it will mimic the reality better than the candidate datum set method. After all, if a part can rock in an assembly and still meet design intent, why narrow everything down to a single, not necessarily the only functional, solution.semiond said:Regarding the quote in your side note, that seems like good news - a more practical approach to unstable situations. Thank you for bringing this up. I would hate to have to explain anyone about a "candidate datum set" when requested to address issues in production and inspection
I did not mean to discourage you to criticize the standard. A lot of us on this forum do that simply because not everything in the standard is correct. I just feel that this particular topic is quite decently covered in the standard - just maybe not as straightforward as it could be.semiond said:
Yes, there is difference. I mentioned para. 4.11.2 just to say that in case when the minimum possible separation between two planes of the simulator has been found and yet the part can still rock, there is a certain standardized procedure to follow.semiond said:
In my opinion the concern you are having applies to many other GD&T concepts too. This is classic example of theory vs. reality dillemma. There is theoretical definition and there is reality where inspection tries (or at least should try) to get as close to the theory as possible. Factors like time pressure, cost limitations, equipment availability, lack of knowledge definitely do not help to achieve that goal, but to me this should not be a reason to say that "this issue is what stops me from defining a center plane as a primary datum in a part, where functionally it appears correct".
- Thread starter
- #33
pmarc,
Thank you for that reply.
I suppose that i simply encountered an example of the fact that there is no perfect tool to communicate design intent, and GD&T is not an exception. If it was perfect we wouldn't see revisions to the various standards every now and then. And, it wouldn't be that interesting without the dilemmas
Thank you for that reply.
I suppose that i simply encountered an example of the fact that there is no perfect tool to communicate design intent, and GD&T is not an exception. If it was perfect we wouldn't see revisions to the various standards every now and then. And, it wouldn't be that interesting without the dilemmas
semiond,
What is the design intent in this case?
Is the parallelism tolerance on your example part intended to ensure the surface meets some angularity requirement relative to a feature on another part within the assembly? If so, what is that requirement? If not, what is the purpose of the parallelism tolerance?
pylfrm
What is the design intent in this case?
Is the parallelism tolerance on your example part intended to ensure the surface meets some angularity requirement relative to a feature on another part within the assembly? If so, what is that requirement? If not, what is the purpose of the parallelism tolerance?
pylfrm
It's clearly an academic exercise - the example basis for measurement has 10X the variability allowed vs the feature variation under consideration and the process is so poorly controlled it uses all of it. Perhaps there is a real-world application that is intended, but if so then the use would be subject to the same failure.
The real answer is to close the loop hole by adding a note to the standard that indicates that:
Use of RFS is not a guarantee of repeatability or innate stability. RFS references may allow some motion and resultant orientations within their boundary. Acceptance of the feature in one of those orientations may allow for unacceptable variations for some or most of the alternative orientations.
I would clean up the suggested example diagram and not bother with the wiggly lines - just flat surfaces with a large taper so that two positions with identical minimum separation are shown - one position accepts the parallelism, the other position rejects it.
The real answer is to close the loop hole by adding a note to the standard that indicates that:
Use of RFS is not a guarantee of repeatability or innate stability. RFS references may allow some motion and resultant orientations within their boundary. Acceptance of the feature in one of those orientations may allow for unacceptable variations for some or most of the alternative orientations.
I would clean up the suggested example diagram and not bother with the wiggly lines - just flat surfaces with a large taper so that two positions with identical minimum separation are shown - one position accepts the parallelism, the other position rejects it.
- Thread starter
- #36
pylfrm, 3DDave,
The sketch is just something i did on CAD to emphasize a point i was trying to understand. The tolerances are exaggerated on purpose. The real width tolerance is +-0.03 and the the parallelism requirement is within 0.02. I think that it still may produce an error, even though maybe not such a big one (the measurement result in the example was only a little bigger than one third of the actual value!). 3DDave, i agree with you that the bad flatness was unecessary. After starting to think about it i realized that it is just the parallelism between the tangent planes that really matters.
The real part i that i have in mind is a spinning part that mainly needs to have good balancing. Why parallelism and not runout then, you ask? The first engineer who was dimesnioning these parts many years ago used that control and since then everyone are copy pasting on the new items ("if it ain't broke..."). Also, for now the datum feature of width is actually controlled for parallelism between it's two faces. We have this very disgusting method of defining it - when a FCF with a parallelism control is placed beneath the size dimension of width WITHOUT referencing any datum, it means that side 1 of that feature of fize has to be checked in reference to side 2, with no difference which one is chosen as the planar datum. Similar approach is to datums derived from an FOS. If a datum feature symbol is placed beneath a width dimension, it may mean "you have two options for a planar datum, choose one" (for parallelism control). If a symmetry control (we rarely use position) is referencing that datum, measurements are made from the two actual faces of the datum width FOS to the two parallel (or in some cases, angeled to form a symmetrical taper) faces of the controlled feature, and then the difference between the distances at the two sides is calculated and reported as the symmetry value. Establishment of center planes is not practiced. Don't ask me where these twisted work methods came from...
I haves thoughts of droping all that and starting making my drawings proper.
Functionaly the primary datum should be a center plane, because the part is clamped from both sides when it works. I was considering to use that center plane as primary in a 3 datums DRF, but i also need to get rid of thst datumless paralellism control between the two faces of the datum FOS. This may cause some cosequences like i described, and hence my dilemma.
The sketch is just something i did on CAD to emphasize a point i was trying to understand. The tolerances are exaggerated on purpose. The real width tolerance is +-0.03 and the the parallelism requirement is within 0.02. I think that it still may produce an error, even though maybe not such a big one (the measurement result in the example was only a little bigger than one third of the actual value!). 3DDave, i agree with you that the bad flatness was unecessary. After starting to think about it i realized that it is just the parallelism between the tangent planes that really matters.
The real part i that i have in mind is a spinning part that mainly needs to have good balancing. Why parallelism and not runout then, you ask? The first engineer who was dimesnioning these parts many years ago used that control and since then everyone are copy pasting on the new items ("if it ain't broke..."). Also, for now the datum feature of width is actually controlled for parallelism between it's two faces. We have this very disgusting method of defining it - when a FCF with a parallelism control is placed beneath the size dimension of width WITHOUT referencing any datum, it means that side 1 of that feature of fize has to be checked in reference to side 2, with no difference which one is chosen as the planar datum. Similar approach is to datums derived from an FOS. If a datum feature symbol is placed beneath a width dimension, it may mean "you have two options for a planar datum, choose one" (for parallelism control). If a symmetry control (we rarely use position) is referencing that datum, measurements are made from the two actual faces of the datum width FOS to the two parallel (or in some cases, angeled to form a symmetrical taper) faces of the controlled feature, and then the difference between the distances at the two sides is calculated and reported as the symmetry value. Establishment of center planes is not practiced. Don't ask me where these twisted work methods came from...
I haves thoughts of droping all that and starting making my drawings proper.
Functionaly the primary datum should be a center plane, because the part is clamped from both sides when it works. I was considering to use that center plane as primary in a 3 datums DRF, but i also need to get rid of thst datumless paralellism control between the two faces of the datum FOS. This may cause some cosequences like i described, and hence my dilemma.
-
1
- #37
- Thread starter
- #38
Thank you for this advice pmarc, good idea.
How about the practical measurement?
Will this require a special inspection device in order to make the side faces accessible for a probe of a dial indicator while the same faces are contacted by the datum feature simulator? Are there any rules defining what portion of the datum feature surface area has to in contact with the simulator?
Thanks again.
How about the practical measurement?
Will this require a special inspection device in order to make the side faces accessible for a probe of a dial indicator while the same faces are contacted by the datum feature simulator? Are there any rules defining what portion of the datum feature surface area has to in contact with the simulator?
Thanks again.
Functionally, the center plane doesn't contact anything so it is not a functional interface.
I would use the tab face as the primary, giving it a tight flatness tolerance and then specify the contacting surfaces as parallel to that one face. Problem solved.
I would use the tab face as the primary, giving it a tight flatness tolerance and then specify the contacting surfaces as parallel to that one face. Problem solved.
- Thread starter
- #40
3DDave,
Datums are theoretical and not functional by definition.
Datum features from which they are derived are functional and contact things in an assembly.
Consider a part with an external cylindrical feature by which it is held in an assembly. If the cylindrical face is the main contact area, it makes sense to define it as the primary datum feature, and the primary datum will be the axis, although the axis doesn't contact anything.
By the same logic if a part is clamped in an assembly from both sides on flat surfaces, the functional datum will be a center plane rather than a planar datum derived from one of the sides.
Datums are theoretical and not functional by definition.
Datum features from which they are derived are functional and contact things in an assembly.
Consider a part with an external cylindrical feature by which it is held in an assembly. If the cylindrical face is the main contact area, it makes sense to define it as the primary datum feature, and the primary datum will be the axis, although the axis doesn't contact anything.
By the same logic if a part is clamped in an assembly from both sides on flat surfaces, the functional datum will be a center plane rather than a planar datum derived from one of the sides.
- Status
Similar threads
- Locked
- Question
- Locked
- Question