What's new
What's new

Gd&t question

Shlomiseg

Plastic
Joined
Jul 10, 2019
Hi there
Say i put an indicator on my lathe turret and put its needle on the face of a turned part’s face While still clamped in the jaws.
Now i rotate the chuck by hand and move the indicator in the x axis also, for the part its in the radial direction.
What am i measuring?
My current assumption is total runout against the jaws axis.

I ask this because in the drawing of the part im required to measure flatness in clamped condition so if im correct, that total runout contains flatness in it
 
Hi there
Say i put an indicator on my lathe turret and put its needle on the face of a turned part’s face While still clamped in the jaws.
Now i rotate the chuck by hand and move the indicator in the x axis also, for the part its in the radial direction.
What am i measuring?
My current assumption is total runout against the jaws axis.

I ask this because in the drawing of the part im required to measure flatness in clamped condition so if im correct, that total runout contains flatness in it

If it is not specifically stated, 'clamped condition' means clamped to your cmm table, or surface plate, not clamped in the machine. If you can share a print (block out customer name or whatever you need to do to make it anonymous) you will get better answers...
 
Why the fuck would someone rotate the work, while jogging an indicator on the Face? Trying to over-complicate a simple thing. TIR generally is called out to another feature, not just Total Indicator Reading on a given feature. If the other Datum is the OD of the part-----maybe?

R
 
If datum is the OD, maybe it's not totally crazy, but I wouldn't call it QA. It sounds like he's measuring a helix, which is not an effective way of measuring anything unless he's combining them in a computer and interpolating.

Inspection shouldn't be performed in the machine you made it in.
 
If datum is the OD, maybe it's not totally crazy, but I wouldn't call it QA. It sounds like he's measuring a helix, which is not an effective way of measuring anything unless he's combining them in a computer and interpolating.

Inspection shouldn't be performed in the machine you made it in.

Not sure which part of this makes less sense.

A. "combining them in a computer and interpolating". I literally don't know how to interpret this...is it a different language?

or

B. "Inspection shouldn't be done in the Machine you made it in". This one is wrong on so many levels it redonculous. I only do FAI in the Machine, unless it's impossible otherwise. ANY Machinist worth a shit knows whether the part is good, before it comes out. Good Machinists don't submit junk for FAI. With that being said; "In" Machine inspection, (probes, lasers, touch setters) is a different case all together. But free form inspection on the Machine is ideal.

R
 
Hi there
Say i put an indicator on my lathe turret and put its needle on the face of a turned part’s face While still clamped in the jaws.
Now i rotate the chuck by hand and move the indicator in the x axis also, for the part its in the radial direction.
What am i measuring?
My current assumption is total runout against the jaws axis.

I ask this because in the drawing of the part im required to measure flatness in clamped condition so if im correct, that total runout contains flatness in it

I'm guessing you have to move the indicator past the centerline in X.

That way you can prove or demonstrate that your X axis is truly perpendicular to the spindle axis.


For example if your X axis was not perfectly orthogonal to the axis of the spindle then you would cut a very shallow taper / dish into the face of your part. [Or if out of square the other way then a very shallow "pointy" cone / convex feature. (No crown -ing ).

Normally if you travel the indicator on the x axis towards the center the reading should be near zero but as soon as you cross the center line and axis of the spindle if your X axis is not square then you will pick up readings that correspond to the mirrored angle difference that the X axis is out of square by.

If you draw it out you'll see what I mean.

Otherwise you are just measuring the angular wobble in the spindle bearings ??? lol. (which on a lathe should be near as damn zero as can be.). [Similar sort of test would be good for 5 axis trunnion to check angular wobble on a C axis rotary (mainly a bearing alignment issue) ].

I guess it also proves that the part was not removed from the jaws at any point or shifted in position (ever so slightly) in cut.

That would basically prove that the turned face is flat AND the X axis is orthogonal to the spindle, as both conditions have to met for a "flat face" … (But obviously you have to cross the centerline as otherwise your indicator is just traveling perfectly parallel to the cut face even if the X axis is out of square to the spindle axis.).

______________________________________________________________________

OP's question sounds like one of those Joe Pieczynski
posed machining related brain teaser's / conundrums on youtube.


[In summary] IF you cross the center line along the X axis... It can show/Prove...

i.) The X axis is square to the axis of the spindle.

ii.) The part is indeed "Flat".

iii.) The part has not moved (in the slightest degree) in cut or from any other operations.

iv.) The part for sure has not been removed from the chuck (most likely) as there is no angular wobble on the face of the part,

e.g. wrong set of soft jaws that got accidentally swiped or swapped etc. (not that one could not spend a long time fiddling with the part to somehow true it up again, but if you have milled features that need to be orthogonal to the face it would take some skill + knowledge to really get it right especially for a longer part.).
 
Last edited:
Sorry i think i didnt explain myself properly.
Flatness was called out on the face, and because the part is thin they also wrote to measure the flatness in clamped condition whatever that means. I ask myself how do i measure flatness on the machine?
 
Sorry i think i didnt explain myself properly.
Flatness was called out on the face, and because the part is thin they also wrote to measure the flatness in clamped condition whatever that means. I ask myself how do i measure flatness on the machine?

When the part is free from the chuck does it taco in half? I'd measure it exactly as you described. Jog in X, get TIR, index the spindle by hand to another position, jog in X etc. Jog the indicator along a chordal path if you want to get fancy. But to measure flatness in a "clamped" position you could say it met the spec "when it was clamped" while handing them a part that looks like a Pringle's potato chip.
 
Ive added a drawing so you can better see the part and the face im talking about
5342300.JPG
 

Attachments

  • 5342300.jpg
    5342300.jpg
    8.1 KB · Views: 72
You are fine, doing what you are doing. There is no callout for Runout from anywhere else, so don't worry about it. Just check Flatness and call it good.
 
Not sure which part of this makes less sense.

A. "combining them in a computer and interpolating". I literally don't know how to interpret this...is it a different language?

I'm not sure what you need explained. All those terms are easily searchable if you don't understand them.

B. "Inspection shouldn't be done in the Machine you made it in". This one is wrong on so many levels it redonculous. I only do FAI in the Machine, unless it's impossible otherwise. ANY Machinist worth a shit knows whether the part is good, before it comes out. Good Machinists don't submit junk for FAI. With that being said; "In" Machine inspection, (probes, lasers, touch setters) is a different case all together. But free form inspection on the Machine is ideal.

R

Your explanation is exactly my point: inspecting in the machine is not FAI. Sure, he can inspect it in the machine, that doesn't mean it's going to pass.
 
Do i have another option for inspecting that?
Its obviously meant for simulate the assembly of the part so..
 
Do i have another option for inspecting that?
Its obviously meant for simulate the assembly of the part so..

What are you actually cutting on this? What material is it? It looks like it might be a spun or stamped, and you're just final machining that one surface? How is that surface supported when you face it?

There are way too many details missing -- including the actual problem! I assume it's that parts are not meeting the flatness spec, is that the case?
 
I'm thinking clamped condition means when this thing is in the "assembly" or "location" it's in when it is assembled to whatever it goes to.
 
Do i have another option for inspecting that?
Its obviously meant for simulate the assembly of the part so..

Seems "They" may be mindful of tri-lobarity(sp) and taco chip effect (as mentioned before) tri-lobed stress relief + (annular / coaxial) undulations etc. and radially distorted profile.

IF you were forced at gun point to map out the geometry of the part on the machine for flatness straightness and warping (radially) (across the radius) and in an annular fashion … you could do it (on the machine) but you'd have to use "Reversal" methods to filter out / subtract the unknowns of the machine (if you had to do this to high degrees of certainty)… It would probably take about a day to perform properly ~ Totally not worth it/ impractical (unless you had bags of time and no real means to inspect.).

I hope you find out what's what and what they want and what they meant.

Cheers.
 
I'm not sure what you need explained. All those terms are easily searchable if you don't understand them.



Your explanation is exactly my point: inspecting in the machine is not FAI. Sure, he can inspect it in the machine, that doesn't mean it's going to pass.

The new guy speaks.

Interpolation is physical motion of a spinning object. Combining it with something in a computer seems dangerous.

An First Article Inspection is not defined by you. It's defined by the inspection department.

R
 
I'll skip responding to the petty personal attack.

I don't believe I've ever heard interpolation described as physical motion, so I'm not sure where you're getting that. Interpolation (mathematically) is fitting an equation to data that's been measured and extrapolating other data from it. In this case, taking points along a helix or spiral and fitting them to a cylinder, circle, or line. I would not attempt to do that by hand (although it's possible), hence suggesting a computer. I would also go straight to software designed for this work to do the interpolation, but it could be done in a spreadsheet program, with way more effort than it's worth.

EDIT: It's occurred to me that you meant physical motion as in interpolated holes, etc. Yes, this is interpolation as I described above. The math is the interpolated part, though. The motion is the outcome of interpolating.

I'm the engineer...I usually both specify and perform FAI for parts I design. Sometimes I make them as well. When I've worked at places with a QC department, they carry it out to my spec. Checking a part to see if it'll pass QC in the machine is not crazy, but it isn't FAI.
 
I'm not attacking you, I'm correcting you. I've had to conform to FAI standards set by Lockheed Martin, Rockwell Collins, Boeing, L3 communications and the DOD. All different requirements.

If I am QC and the Machinist, and I know what will and what will not. Then I do first article in the Machine. If you want to see come visit.

Just because some green Engineer saw it a way, doesn't make it Constitutional. But I do admit being wrong about my definition. Oops.

Robert
 








 
Back
Top