Runout - Total and Circular - How is it calculated?

Post Reply
randomguy
Frequent User
Frequent User
Posts: 53
Joined: Thu Jan 30, 2014 8:43 pm

Runout - Total and Circular - How is it calculated?

Post by randomguy » Wed Feb 25, 2015 10:20 am

Hello all:

We've run into a situation where a manual check of some features shows circular runout being in tolerance, yet the full DCC program of the same part, shows circular runout being out of tolerance.

So we've had several conversations around here about the theory of runout and how it would be calculated, and obviously, why we're getting different reasons. Here is my thinking and/or guess:

Circular Runout (feature = circle, datum = cylinder) should be something along the lines of 2X (Radial Offset of feature to Datum) + Circularity(Feature Circle). Seems like this would simulate the FIM of an indicator placed on the feature while rotating about the datum.


Total Runout, (feature = cylinder, datum = cylinder) however is more complicated since there's not a single radial offset from the feature to the datum. Would this be something along the lines of 2X (Max(Each Section's Radial Offset)) + Cylindricity(Feature)?

Thoughts on this? Am I close?

User avatar
Ryan Christopher
Nikon
Nikon
Posts: 245
Joined: Mon Jun 06, 2011 3:00 pm
Location: Sumner Washington
Contact:

Re: Runout - Total and Circular - How is it calculated?

Post by Ryan Christopher » Wed Feb 25, 2015 12:45 pm

Although runout "may be used to control the cumulative variations of circularity and coaxiality" (Y14.5-2009 9.4.1) you couldn't simply use some calculation that includes the form of the feature to find the runout. Think of if you measure the feature (not the datum) with the minimum number of points. In that case the form would (for calculation purposes) be perfect, but the runout could still be really, really ugly.

Whenever I am trying to find an analog to using a particular software feature, I just think of how I would measure on a surface plate. In this case, a vee block (or 2) and a dial indicator is all you need.
Last edited by Ryan Christopher on Wed Feb 25, 2015 4:34 pm, edited 1 time in total.
Ryan Tackes
Applications Manager, VMS Inc.
http://www.vantagemeasure.com/

JohnK
Frequent User
Frequent User
Posts: 48
Joined: Thu Jun 27, 2013 10:28 am

Re: Runout - Total and Circular - How is it calculated?

Post by JohnK » Wed Feb 25, 2015 4:08 pm

In layman's terms I explain Runout as circularity + position. A perfect circle out of position has runout equal to position. A perfectly located circle will have runout equal to its circularity. It will never be better than either of these aspects of the feature.

But in response to your differences, I would look into how you are controlling the datum's outside the CMM versus how you are measuring/constructing them in the CMM. I.E. on a set of centers/rollers you are contacting a cylinder only at two locations while measuring more surface area on the CMM will make your datum affected by the form of the cylinder.

randomguy
Frequent User
Frequent User
Posts: 53
Joined: Thu Jan 30, 2014 8:43 pm

Re: Runout - Total and Circular - How is it calculated?

Post by randomguy » Thu Feb 26, 2015 5:52 pm

JohnK - I look at this the same way, and I hoped to have some validation that CMM Manager calculated in the same way that I "thought"..... I didn't get there.

However, I did speak with a helpful tech from Nikon about how to proceed. Thru some trials and recommendations, we found the issue. This part basically has a plane for Datum A, a turn (positive cylinder) for Datum B, which is the origin, and a bolt hole for Datum C.

By adjusting the # of points on the cylinder, we found that there was a "low" spot on the cylinder that by coincidence, the DCC program was hitting every time, but the operator wasn't hitting when he was checking with the joystick. This spot wasn't visibile to the eye, but the diameter tolerance is +/- 0.05 with a runout of 0.05 on another feature called back to that cylinder.

We performed a little rework and we'll be able to ship the parts I believe.

One thing that I picked up from the tech I spoke with though.... For this type of reference frame, we'll likely now be following this alignment procedure:
1 - Measure Datum A
2 - Level to Datum A
3 - Measure Cylinder, Datum B
4 - Create Point, Intersection of Cylinder and Datum A
5 - Set Origin to Construction Point
6 - Measure Datum C
7 - Align to Datum C
8 - Sync CAD

Previously, we had checked a circle on the cylinder and used that feature for X/Y origin. However, doing this subjects your origin point to the form of the cylinder at whatever height it was checked. Creating the intersection point ensures that the reference frame origin is on the Datum A plane. This will take out any form imperfections in the cylinder out of play.

Just some thoughts.... Thanks everyone for your help!

Post Reply