Concentricy Issues

[Mr_Cris]

Hi everyone, I hope you can help me understand why I am having some measuring inaccuracies. Sorry if I am long winded.

I have a large part that has been run on a lathe. There is a big bore at the top of the part followed by a bunch of smaller and smaller bores. The cross section would look like a bunch of steps. The big bore at the top is about 5” and the smallest bore at the bottom is about 1”. The smallest bore starts at a depth of about 6”, so I have to use a long probe to get to it. The only long probe I have is 5”. It reaches fine, but when I go to report the concentricity of the smaller bores to the largest bore, I get more and more inaccuracy as the bores get smaller and deeper. The deepest bore is non-concentric by .003”. Why is that?

--I understand that long probes are not as accurate as short ones. With the long probe that I have, I measured a bunch of different sized ring gages (1- 5”) to see how far off I am on the diameters; about .0002”.

--I know that those bores should be both round and concentric. The part I am measuring is made of strong solid steel bar stock. All of the machining is done in one lathe chucking.

--I also know that my alignment is sound, using a DCC alignment on top of a manual alignment.

-- I am also certain that the part is not moving/bending/flexing/etc.

--I am also certain that I have the right probe module/stylus configuration. A SP25M to a SM25-3 to a SH25-3 to a 5X21 stylus. I have tried both to touch trigger the features and to scan the features, thinking that scanning might compensate for some of the stylus deflection- the readings are the about the same either way.

I should also add that I have measured all the bores using the same long probe. I’m not sure if that is the best thing to do.

Any insight would be greatly appreciated as to why I would have such good readings on diameter, but not on location, when using my long probe.

[cmmguy]

You dont say what you are leveling on. Is it the 5" diameter??

[craiger_ny]

Do you level to the face that is located at the end of the part that the large bore is? I'm thinking that the part is chucked against the other face and that the bores are generated normal to it. If the machining operation that creates those bores does not also face the part where you are leveling then that could be your problem. To make a long story short, is the face you are leveling to parallel to the face that is banked against in the machine. if the face you are leveling to is not prependicular to your bores your error will increase as you move away from that face.

I still think that is a lot of styli length though.

Craig

[Mr_Cris]

Good point, I am leveling the part to it's face. The face of the part was machined in the same chucking with the bores. The O.D. of the part is 6" so I have a good 0.5" wide face to take points off of. And, I should note that I used the auto feature method with a circular disbursment to take the points of the plane after my manual alignment.

[Underspec]

Have you checked to see if your face is perpendicular to your bores?

I would also dimension your X & Y values to see where it's going out at. Maybe even a graphical analysis.

[John Kingston]

I'll throw out my old standby...., Have you tried verifying the CMM results using other measurement methods? Can you check concentricity on centers? Can you check concentricity in the lathe? I realize it might change when you take it out, but maybe you really do have a concentricity problem in the part. I think you might have a runout issue too, sounds like it is a very long part hanging out of the lathe chuck.

[Mr_Cris]

Interestingly all of the bores are non concentric in the same direction. If it were simply probe deflection wouldn't the non-concentric nature be all over the place?

Oddly, the first 4 bores are out by just about the sam amount .001"

Here is a graphical.

[Mr_Cris]

I have been thinking about the too John K. I do run into that problem around here. I am going to try to set up something to check that with. The setup for that part has already been broken down, so I will have to get creative.

[#2#]

I haven't seen this mentioned yet so...

Is this concentricity called out on a drawing? Concentricity ALWAYS requires a datum reference. Try aligning to the datum and then measure the concentricity. Make sure when you dimension that you select the feature first and the datum second. Also, are you measuring the bores as circles or as cylinders?

[Mr_Cris]

Well Mr. Goodluck, you have a good point. My print does have a datum reference. It is just a centerline drawn down the center of the part. I talked with someone here and here is how they told me to set this up. They said to use the top bore and the bottom bore to create my datum. I used a best fit line between the top and the second to last bore and called them my datum (I didn't want to use the bottom bore because it is a really Ugly looking drill hole). Here is the results difference. Datum A was my original, it considered the top bore as my datum. Datum B is the revised one.

I think I would get even better results if I used a best fit line between all the bores as my datum. Does this method make sense, or am I just fabricating good readings at this point?

[#2#]

Is the outside of this part cylindrical as well? Is it symetrical about the centerline? If so, I would interpret the drawing such that the centerline of the outside is the datum.

I think your drawing needs to be clarfied.

[Mr_Cris]

I have been told that the bores should all be concentric to each other. A mating component is supposed to spin in those bores freely.

I'm pretty certin at this point that my problem has been how I have defined a datum. If all the bores are to be concentric to each other, then how do I define that datum. Is my idea of using a best fit line between all the bores as my datum appropriate? Is there a better way?

[bob mappes]

try using the center of one bore ( tightest tolereranced bore maybe )
as your "datum" and check concentricity of the other bores to it

[Tony.Hirotec.Aus]

I would interpret that to mean the core's concentricity is relative to themselves, so using the best fit through the diameters would be correct

[Mr_Cris]

Thank you everybody. This was my first time using this forum. My experience has been very welcoming.

I feel much more confident about my setup here. Using a best fit line between all the bores gave me very sensible results that match what I have found with traditional techniques (a test indicator). I do have about .0002" of error, but that is to be expected with a real word example and a long probe.

I look forward to visiting the forum often.