how the operator realizes dimensions by providing tolerances

[opua]

hi to everyone, I have a curiosity that will certainly seem trivial to you but it has attracted me for a while, I step to the question: when a project foresees that two mechanical organs should be coupled with minimum play, of course the nominal size of both will be the same but, in order to allow the correct assembly it will be necessary to define the right tolerances. I make an example, suppose that a diameter shaft of 45 ¬+0,1/0 should be coupled with a diameter hole of 45 -0,1/0, the operator who will perform the processing on both organs which measures will have to take to obtain the above dimensions? I mean, will he try in every way to make it at a nominal price of 45, or because there are some shocks at stake will do the work not exactly accurate? and if for both we obtain nominal values without tolerances, i.e. if some pieces were processed in size 45, would this prejudge the coupling or not having no more game but interference?

 

[zeigs]

except if you put 0,+0.1/-0.1,0 risk that you don't get the pieces.

the operator will try to make so that the dimensions of the pieces respect the tolerances. I better explain: the nominal size is not real, there will always be a break-up, and this depends on many factors (manuality of the operator, precision of the machine, checks done, etc.). the definition of a tolerance in the first place defines how well done a processing must be done: the more tolerance is narrow, the better the processing, the higher the cost. and this defines the difference between the maximum and the minimum within which the actual measure must remain.
then the position of the field is defined, and this depends on what to do the coupling. should you turn? I assure you that the mating is lasco. must be stuck? I want interference.
so if you need the operator to approach as much as possible to a certain size (you need a well defined force, the maximum game you can afford is minimal, etc.) you will have to define a tight tolerance, if you leave it free he will try to approach us trying to lose us as little as possible :biggrin:

 

[opua]

thanks zeiga for the answer, you say that the nominal fee is not real, but if at the test the two components come out at a nominal fee? if for a case of the operator the two pieces were both at 43mm, the planned project game that would end?

 

[zeigs]

is well there: in that case the game would be 0, so if you need the component to move you wrong to put the tolerance (the possibility exists that you can't mount the two pieces). in this case you could use 43 (0.-0.1) out of 43 (0.2,0.1) that gives you a game between 1 and 3 tenths (the second is like saying 43.1 (0.1.0)). then depends on what game you thought about your mating

imho, then obviously depends on what you have to do, h7 on f6 (relatively wide) or h7 on g6 (straight) are recommended choices to have a precise movement. the first gives you a game 25-66 um, the second 9-50 um. and use standard tolerances

 

[Fulvio Romano]

thanks zeiga for the answer, you say that the nominal fee is not real, but if at the test the two components come out at a nominal fee? if for a case of the operator the two pieces were both at 43mm, the planned project game that would end?

that questions...if the nominal is understood in tolerance, the piece is accepted, otherwise it is discarded, reworked or in the worst hypotheses, it is given fire! :biggrin:

 

[zeigs]

I only do the dots on the i: if the real (measured) share is included in tolerance then the piece is acceptable. It may happen that the nominal share is not included in tolerance, e.g. misuro 43.0 while the share is 43 (0.2,0.1): in that case even if the actual measure corresponds to the nominal the piece is however to be discarded:

I make the pygnolo, ok, but I have had to say for several thousand euros on the acceptability or not of details in similar conditions, and we had to clarify well on what are the nominal quota, the real and limits of the real quota regarding the nominal, which obviously did not correspond but was a design requirement that was to be respected

 

[stefanobruno]

and we had to clarify well on what are the nominal, the real and limits of the real quota regarding the nominal

What is there to clarify?
the workshop must respect the quotas in the field of tolerance indicated in the drawing, point.
It is clear that if the designer/particularist writes an incorrect tolerance, his ears will be caked and donated a de fero port.

Bye.

 

[Fulvio Romano]

There is not much to clarify, norms are already clear enough. less of particular things written in drawing that can be interpreted, but they are usually things to avoid.

My boss, in the part time work I was doing when I was at university, wrote the phrase "to be realized in tight tolerance". I used this phrase often as a joke, both at university and now at work, no one believes that it is true!

 

[opua]

I thank all again for your interventions, now returning to my perplexity I was taking a look at just now for an example taken from a mechanical design text, where a series of coupling cases are proposed including a defined *high precision slide*. the selected tolerances are h6 for a tree and h6 for a hole. if for example we consider two pieces of diameter 30 and therefore 30 0/ -0,013 (for a tree) and 30 0/ +0,013 (for a hole), examining the values of the scaffolding it is clear that both pieces can have a size of 30mm, had, if this happened (shaft and hole both at 30mm), the high precision coupling would be guaranteed equally? I think I'm not, I think I'm more in interference! and yet the project tolerances ensure a precision game?
Thank you.

 

[Er Presidente]

I thank all again for your interventions, now returning to my perplexity I was taking a look at just now for an example taken from a mechanical design text, where a series of coupling cases are proposed including a defined *high precision slide*. the selected tolerances are h6 for a tree and h6 for a hole. if for example we consider two pieces of diameter 30 and therefore 30 0/ -0,013 (for a tree) and 30 0/ +0,013 (for a hole), examining the values of the scaffolding it is clear that both pieces can have a size of 30mm, had, if this happened (shaft and hole both at 30mm), the high precision coupling would be guaranteed equally? I think I'm not, I think I'm more in interference! and yet the project tolerances ensure a precision game?
Thank you.

There is no interference, the upper limit of the shaft coincides with the lower one of the hole.

 

[opua]

There is no interference, the upper limit of the shaft coincides with the lower one of the hole.

Let me understand, but why then does zeigs say that *the nominal size is not real* if there can be the possibility that both components are at a nominal fee or zero shock?

 

[Mocca]

many have already told you. for tolerances there are tons of tables and norms and everything has already been foreseen. googla “recommended couplings”.
from your example (shaft 30 0/ -0,013) (hole 30 0/ +0,013) means you can make a tree “tendent to the nominal” 29.999999999999 and in this case you are in tolerance but if you realize it to 30,000001 means you are out of tolerance as well as for the hole 29.999999999 you are out of tolerance while 30,0000001 you are in tolerance then taking into consideration the maximum limit
tolerable you will have tree 29.9999999 hole 30,000001 i.e., as the president says, there is no interference.
regarding your original question; the designer must put the tolerances with knowledge of cause i.e. he must know how to interact the 2 coupled organs while regarding who realizes the pieces must respect the inclusive quotas of tolerance. if on a design there are no quotas with tolerance means that the operator on the machine can do what he wants and no one can contest it.
Good things cumpà.

 

[stefanobruno]

if on a design there are no quotas with tolerance means that the operator on the machine can do what he wants and no one can contest it.

a well-made paper mill will however have a reference to the general rules that the company claims and to that, at least, must stick to the operator in the workshop.

Bye.


Moucca, how do you get such an avatars?

 

[zeigs]

the nominal share is not real because it is 43.0000000 (infinite accuracy!), while the real measure can be one of the infinity between 43.0000000 and 43.1000000. tolerance is put precisely to define the maximum acceptable error in reality. because reality, unlike the cad, is not perfect. the nominal share is only one of the infinite possible real measures, if it is included in the tolerance: there are deviations (almost all those regulated apart from h, j, js and k), in which the nominal quota is not precisely understood in the field of tolerance, for which the nominal measure cannot be real, but this is wanted.
then if by hypothesis it was possible to create details with tolerance 0, and it is not, unless you produce an infinite number of pieces and choose only those that respect the nominal quota, and I guarantee that you find one (maybe1), it would be possible to create pairs 43.0000 on 43.0000 with absolute precision and furniture. but should be perfectly smooth (wrinkle 0!). and I am not, so I have (I don't want to be president) also a mobile coupling h/h is wrong, because it doesn't guarantee the movement, because (oh, lose!) if I double two pieces to 0, only for roughness they block.
oi, we talk about a thousandths, but it is the concept that makes the difference.

as to what the operator does, the above mentioned above: if the piece is in tolerance, the operator is right and if it does not act, it is the designer's fault

 

[Mocca]

stefanobruno. instead of inserting a simple image "jpg" insert an image "gif" as long as it is obviously animated.

 

[zeffiro62]

besides the roughness there is then to consider the cylindricality of the two processes that can be the tolerance of quota, but out of geometric tolerance.
However if I do not remember badly (I am not a designer) there are already tables with recommended couplings for the various types of couplings.
I believe in the industrial expert manual

 

[Fulvio Romano]

stefanobruno. instead of inserting a simple image "jpg" insert an image "gif" as long as it is obviously animated.

[OT mode ON]I tried with an animated gif, but the image in the avatar then was static...who knows why.
I'm so fond of my dairy. .[OT mode OFF]

 

[vittori]

hello to all, I put myself in this discussion of "man from workshop". In addition to the drawing requirements, a newly decent operator will always try to stay at the center of the tolerance to be performed for the simple reason that, due to the inevitable wear of the tools, there is a margin between the moment when the quota begins to vary and the construction of the 1st piece does not conform. tolerances serve those who design, but especially those who have to make the trucile. . .

 

[zeffiro62]

Is it true, however, if you have a piece at the limit of tolerance and not outside that you do the waste?
a question to all
we put an example my piece has 30 nominal and tolerance +0,000 -0,020
I'm 30,0005 in this case, how does it behave? Tolerance says to measure within the micron the measuring machine detects a + 0.0005 as it behaves?

 

[zeigs]

Is it true, however, if you have a piece at the limit of tolerance and not outside that you do the waste?
a question to all
we put an example my piece has 30 nominal and tolerance +0,000 -0,020
I'm 30,0005 in this case, how does it behave? Tolerance says to measure within the micron the measuring machine detects a + 0.0005 as it behaves?

imho is acceptable.
even if you've quoted a very blurred measure. I explain: the rounding of 5 is always at the top, so it would come 30.001, therefore out of tolerance. but also the machine that makes the round measure, so the measure is actually between 30,00045 and 30,00054. under 30,0005 roasting at 30,000 (acceptable), over roasting at 30.001 (not acceptable). this to make the egg counterpart.
more realistically the piece must be coupled to another also in tolerance. As much as I said before, I will never make a couple that predicts the possibility of a 0/0, indi for which it is not the thousandth, more in this case with a 50% probability of having the rounding at the top thousandth, which will make me the difference. If I define 2/100 tolerance, I imagine I have at least as many on the other piece.
in addition to the fact that the simple surface roughness is of the order of magnitude of the micron, so if the machine is able to appreciate the tenth of micron I would like to see the surface of the measured piece.
In practice, unless the micron makes the difference, and in these cases I use a much more precise control machine, realistically the piece is acceptable. to the maximum very, but very, in relaxation I ask the supplier to tarnish to stay in the center of tolerance to avoid any future discussions with testing and living serene