Which angle am I measuring?

[kv802]

I am measuring the angle between 2 lines in the ZPLUS workplane. The nominal is 90 +/- 0.5deg. Line1 is pointing in YMINUS direction, line2 is pointing in XPLUS direction. When I dimension it as line1 (1) then line2 (2), I get 89.28840. When I dimension it as line2 (1) then line1 (2), I get -90.71160.

My question is - which one is the correct result?

Any clarification would be very much appreciated. PC-Dmis v4.2

Thanks,
Keith

[#2#]

Well, yes.

If you add both of those together you get 180º! One is measuring from one side and the other is measuring from the other side. Which one is correct depends on how the print is drawn.

[#2#]

Here is a sketch which hopefully clarifies what I said.

[Matthew D. Hoedeman]

I think he wants to know WHICH angle it SHOULD be giving, with one line pointing "down" and one line pointing to the "right". Answer is, WHO KNOWS? If you are doing these lines as lines, it will be hard to give a dimensional 'positive' answer, if you are doign POINTS and constructing the lines, you could dimension the 2 points used for the line (the SECOND line in the dimension is the 'datum or master' line) and that will tell you if it is OPEN or CLOSED. I would NOT trust Pcdmis to come from the 'same side' every time, put some points in there just to help you know which way the angle is 'leaning'.

[#2#]

I think he wants to know WHICH angle it SHOULD be giving, with one line pointing "down" and one line pointing to the "right". Answer is, WHO KNOWS? If you are doing these lines as lines, it will be hard to give a dimensional 'positive' answer, if you are doign POINTS and constructing the lines, you could dimension the 2 points used for the line (the SECOND line in the dimension is the 'datum or master' line) and that will tell you if it is OPEN or CLOSED. I would NOT trust Pcdmis to come from the 'same side' every time, put some points in there just to help you know which way the angle is 'leaning'.

Oh, in that case, I have no idea. It is pretty rare that I run into 90º called out on a drawing. It is usually easy to tell with angles such as 45º and 135º. I guess only at 90º and 180º would it be hard to tell.

[kv802]

I constructed both lines from 2 points (the center of a circle and a mid-point of a slot).
With line1 in YMINUS direction and line2 in XPLUS direction, I want the simple angle between, not the complement - the lower right quadrant when looking from ZPLUS direction.

When looking at ZPLUS, my first instict was to pick the lines of counter clockwise direction giving the 89.28840 result.

Is there some rule that PC-Dmis gives a positive result in CCW direction and the complementary angle in CW direction?

[Wingman]

Until I read this I was sure that when PC DMIS calculates angles between features the 1st feature is the reported feature and the 2nd feature is the DATUM feature. If that were the case both of these angles should have been the same absolute value but different sign (+/-). Your example blows that out of the water.

As a test, do like Matts says. Intersect the lines and create a zero point at the intersection. Align to one of the lines and report the location of the point used to construct the end of the other line. This will tell you whether the actual angle is acute or obtuse. Please, if you find this out post the details of your findings. I am interesting in seeing what you get.

Also, can't you turn graphic reports on and get a visual representation of the angle being checked?

[Alexx]

On PC-DMIS, every feature has a direction, even circles and cylinders. If you could highlight the 2 lines, you will see which directions they are pointing. The angle is the one between the lines going in the same 'general' direction.

[kv802]

I set my origin to the interesection of the 2 lines and aligned my Y-axis to one of the lines. Instead of measuring the angle between the 2 lines, I measured the angle between the line and the Y-axis. The result was the 90.7 rather than the 89.3.

On this part, I have a 26deg angle between the feature and the Y-axis of the part. Measuring the line to the Y-axis gave me -25.1deg. I flipped the vector of the line and again measured to Y-axis, getting 154.9deg.

I turned on the graphics and got these funny blue lines, which did not make help much.

I went back and measured the angle between the 2 lines and did not repeat my orininal issue. Instead, alternating the order of the lines simply alternated the sign of the resulting angle. It seems that I was doing something different to what I first described.

At any rate, thanks heaps for the advice. I am now satisfied that my 90 degree angle is 90.7, not 89.3.

Keith

[#2#]

I went back and measured the angle between the 2 lines and did not repeat my orininal issue. Instead, alternating the order of the lines simply alternated the sign of the resulting angle. It seems that I was doing something different to what I first described.

When you tried it again, were you back in the original alignment or were you in the alignment you used to figure it out?

If you have a nominal input in the window where you create the dimensions it will give you the angle closest to that. So for example, if your angle is 60º depending on how you choose the lines you could get 60º or 120º. If you input 60º as the nominal before you create the dimension, you will get 60º every time.

I know that doesn't help with your current problem but something to keep in mind for the future.

See two posts below this one for further explanation.

[Wingman]

When you tried it again, were you back in the original alignment or were you in the alignment you used to figure it out?

If you have a nominal input in the window where you create the dimensions it will give you the angle closest to that. So for example, if your angle is 60º depending on how you choose the lines you could get 60º or 120º. If you input 60º as the nominal before you create the dimension, you will get 60º every time.

I know that doesn't help with your current problem but something to keep in mind for the future.

I have a real hard time with this. Are you SURE that is how it works? It seems to me that this would leave ambiguity. For example lets say I have a 90 deg angle and I want to know what it is. I out in a 90 niominal and I get the one closest to 90? How can that work? Let's say the correct angle was 89.7 as in this case. The other one is 90.3. Both the same distance from nominal so your results would be ambiguous. Tell me it isn't so.

[#2#]

I have a real hard time with this. Are you SURE that is how it works? It seems to me that this would leave ambiguity. For example lets say I have a 90 deg angle and I want to know what it is. I out in a 90 niominal and I get the one closest to 90? How can that work? Let's say the correct angle was 89.7 as in this case. The other one is 90.3. Both the same distance from nominal so your results would be ambiguous. Tell me it isn't so.

Sorry, I should have explained that that trick works for angles other than ones close to 0, 90, 180 and 270. Because it either gives you the angle or the compliment. So, at 90 and 180, the compliments are 90 and 180 respectively. But at say 85, the compliment is 95, 80-100, etc. As you depart from 90 or 180 until you approach another 'problem area' the angle and its compliment diverge and it becomes obvious which one is reported.

My tip was a workaround (at angles that don't fall in 'problem areas') to ensure that the desired angle rather than the compliment is reported regardless of the order of selection or vectors of the features.

[Wingman]

Sorry, I should have explained that that trick works for angles other than ones close to 0, 90, 180 and 270. Because it either gives you the angle or the compliment. So, at 90 and 180, the compliments are 90 and 180 respectively. But at say 85, the compliment is 95, 80-100, etc. As you depart from 90 or 180 until you approach another 'problem area' the angle and its compliment diverge and it becomes obvious which one is reported.

My tip was a workaround (at angles that don't fall in 'problem areas') to ensure that the desired angle rather than the compliment is reported regardless of the order of selection or vectors of the features.

OK, I got you. I just double checked and PC DMIS does have a set rule. Feature 1 is the FROM Feature. Feature 2 is the TO featuer. The direction vector of the features count. [email protected] is positive. With those we can always KNOW what angle PC DMIS will report.

I did see what you wrote about so you can override the rule and force PC DMIS to report the angle you want.

[#2#]

OK, I got you. I just double checked and PC DMIS does have a set rule. Feature 1 is the FROM Feature. Feature 2 is the TO featuer. The direction vector of the features count. [email protected] is positive. With those we can always KNOW what angle PC DMIS will report.

I did see what you wrote about so you can override the rule and force PC DMIS to report the angle you want.

It's easier for me to input the nominal than to try to remember the rule but it is good to know there is a rule. Did you find that in the help files?

[Wingman]

It's easier for me to input the nominal than to try to remember the rule but it is good to know there is a rule. Did you find that in the help files?

Help files hinted at it. So I did some testing to prove it out. It does work as stated. Like you I have always keyed in the Nominal and got what I wanted. However, I do believe 4.2 is different yet. I seem to recall trying that with 4.2 and it didn't work the same as 3.7.