Correct Angle Measurement Technique

[Vladimir]

Greetings dear forum participants!
It is necessary to help in the method of measuring angles, before that I measured some parts of their angles, but here the tolerance is too strict and the measured surface is too small, how will this size be correctly measured?
I tried to build lines, tried to build planes, measured between them and the results vary greatly from 60.026 to 59.873, but I’m sure that the angle is there in the tolerance and the part is in order.
1.jpg2.jpg

[brian.freeman]

Are you sure the tolerance is correct? Because this is only 36" (seconds) and 18" (seconds) of a degree?

[Schlag]

Depending on the size of that surface.... you really could have NO tolerance. Use profile as reference ? What are you seeing for deviation ?

[Vladimir]

I did a scan on one side, and when measuring the profile formand location - MAX 0.118 / MIN 108

[Mike Ruff]

Yeah, I've never seen a tolerance this tight. I'd be surprised if the CMM could even repeat within .005°

[Vladimir]

I am also very surprised by such a tight tolerance, but the engineers want to see how much the part corresponds to the drawing, because it must be in perfect condition for work, and I have too much difference from section to section

[brian.freeman]

Don't know what type of material, but if it is steel or something to that effect, the thermal gradient (expansion) will cause deviations greater that your tolerance.

[Vladimir]

yes it's steel
yes, actually it is, I will inform the engineer about it
the problem is explaining this to them because they think that smm can measure everything

[brian.freeman]

@Vladimir
The best way to measure this would be to have it 'soak' in the room for about 2 hrs, depending on the size.

[Schlag]

show them the specs on the CMM

[JEFMAN]

If they think that it's ok, showing them a tol as ±a+ n*L/1000 won't be efficient

[Schlag]

I am pretty much the go to for anything measurement / tolerance related so I guess I get lucky and don't typically have to explain those things that deep.

[JEFMAN]

I would measure both sides as plane, ttp (not scan !) with 4 rows mini and a lot of columns.
Then I would dimension the flatness, and the angle.
0.005° is around 85µrad, so if you level on a plane and look at the vector of the other, the tolerance would affect the number at the 4th decimal place... (0.1µm/mm)
Depends on the size of the part, but I believe the flatness will be too important to give an accurate result on the angle.
That's how I would explain it, without explaining that the classical cmm geometry can't do it (too many peoples think that a cmm is valid for everything - and it's too hard to explain them that they're wrong !)

[Vladimir]

Yes, I also applied the results for the last measurement of 60.071 along the entire length of the plane (50 mm).
I will leave it as it is
That's the whole problem, to explain to them that the cmm is not omnipotent, but unfortunately they believe that if the ball touches the surface and makes a "beep", then there is a 100% result, and sometimes they try to explain it to me...

[fatso666]

Over 100mm (I don't know the size of the surface you are measuring), 0.005-degrees amounts to just 0.0087mm. If your 'V' is half that deep, then the tolerance is half that. The CMM is probably close to its ability to measure that AND the surfaces would need to maintain flatness to less than that amount, too. Sounds to me like the engineers are asking for too much. (Big surprise)

What is the approximate depth of the 'V'?

[Vladimir]

Depth of V - 3.20 mm