[He...] Posted May 14, 2018 Share Posted May 14, 2018 Hello, I have a few questions about radial runout. I have a dimension of a runout with a tolerance of .0005 inches. I am scanning 2 cylinders and comparing the two. I have done a gage R&R and my repeatability over 10 parts is over 100%. Are there any tricks to improve the repeatability of runout? Thank you Link to comment Share on other sites More sharing options...
[De...] Posted May 14, 2018 Share Posted May 14, 2018 First make sure to take enough points so you have a good data set to begin with. Then use Minimum feature evaluation on your feature and use Least Squares evaluation on your Datum. Unfortunately any taper in a cylindrical datum makes the runout evaluation very erratic when using Outer Tangential. When it evaluates the datum using outer tangential it determines the centerline of the axis and the taper can cause it to skew the axis based on out of round / taper in the cylindrical datum. I feel I should mention before someone jumps down my throat that this is technically the wrong way to check runout as outer tangential 'should' be used. That said, feel free to change the settings I mention and then check your part on the surface plate and compare results, in virtually every real world case other than parts that are ground to a gage quality, results using the above method are more likely to match what is found using a 'plate check'. Link to comment Share on other sites More sharing options...
[Da...] Posted May 14, 2018 Share Posted May 14, 2018 Can you give some more info on the part ? How big is the cylinder, is it longer than the diameter by at least 2-3 times? Is there a large offset to the other cylinder ? How much ? Did you look at cylindricity of the two cylinders ? All of these things can influence the repeatability. Link to comment Share on other sites More sharing options...
[He...] Posted May 16, 2018 Author Share Posted May 16, 2018 Thank you for the responses. The first cylinder that is the datum is a diameter of .528 in by .220 long, the second cylinder is .2501 by .150 long. The cylinders are 1.5 in apart. Tall order with a tolerance of .0005!! Thanks again. Link to comment Share on other sites More sharing options...
[Da...] Posted May 16, 2018 Share Posted May 16, 2018 Is there a datum FACE somewhere on the part ? Hopefully a very large face. You don't stand a chance using cylinder. The cylinders are far too short to use as a spatial rotation. A face that could provide spatial rotation , then the cylinder to establish X,Y origins would help. Can you make a sketch of the part ? Link to comment Share on other sites More sharing options...
[To...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. 🤣 🤣 🤣 Link to comment Share on other sites More sharing options...
[To...] Posted May 17, 2018 Share Posted May 17, 2018 Just thinking out loud here, but to help stabilize the measurement of the 2 cylinders, would it make sense to create a secondary alignment using a 3d Line created from 2 circles on the 2 cylinders, using the 3d Line for the Spatial and X/Y Origin. Then, measure the cylinders referencing the secondary alignment? Link to comment Share on other sites More sharing options...
[Da...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. While I agree with that procedure, isn't that technically "cheating" ??? 😕 Link to comment Share on other sites More sharing options...
[He...] Posted May 17, 2018 Author Share Posted May 17, 2018 I tried to make a crude sketch, but it is too big to attach. There are some internal flats that are a datum, but they are parallel with the two cylinders. I don't have a datum that is perpendicular to the cylinders. If a line is constructed between the two cylinders, wouldn't that make the runout zero? Link to comment Share on other sites More sharing options...
[To...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. Not really. The use of the secondary alignment is only for creating a consistent coordinate system for measuring the cylinders, instead of relying on one of the cylinders to establish the alignment. In other words, the short cylinder is not creating a stable alignment, which means you're not going to measure the other cylinder with any consistency. You're still going to measure each cylinder and process your characteristics per normal "legal" guidelines but the secondary alignment, or even your base alignment, will create a more stable, repeatable coordinate system to help you measure the cylinders. Link to comment Share on other sites More sharing options...
[To...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. We don't use the line for the runout evaluation. We're only using it for navigation. See my reply to Dave Scott below. Link to comment Share on other sites More sharing options...
[De...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. If you were to use that constructed line for a measurement the runout would not necessarily be zero. You would be establishing an axis through the circles, however that does not mean the two component cylinders could not runout to the axis that was created. Just imagine you have a shaft with a cylinder at either end and the two end cylinders are each angled in relation to the main shaft by 5 degrees. You could now create a circle around each cylinder as close to the ends of the shaft as possible and create a shaft axis, the cylinders would runout to the axis created because of the angles of the cylinders. Circles are 2D features but cylinders are 3D so runout wont always be zero because you are essentially checking the perpendicularity of the cylinder to the axis of the two circles. Link to comment Share on other sites More sharing options...
[Da...] Posted May 17, 2018 Share Posted May 17, 2018 Please sign in to view this quote. You will use one CIRCLE from each cylinder to construct the 3D line so the runout will still have some variation when you ask for runout of the CYLINDER . Link to comment Share on other sites More sharing options...
[He...] Posted May 17, 2018 Author Share Posted May 17, 2018 I think I'm understanding now. Runout is a new one for me to work with. Really appreciate the information. I'll give the suggestion a try. Thanks again. Link to comment Share on other sites More sharing options...
[Ma...] Posted February 20, 2020 Share Posted February 20, 2020 Please sign in to view this quote. Out of the blue here, but I'm having a similar issue with getting my measurements to line up with those from a gage/indicator measurement that I think I should be able to rely on. Do you mean creating the 3d line from only two circles, one from each cylinder? or two from each to incorporate any misalignment of axes? Link to comment Share on other sites More sharing options...
[Ke...] Posted February 20, 2020 Share Posted February 20, 2020 As it has been mentioned in this thread, the form of the Datums that are called out is of critical importance (when not "cheating" 🤣 )... is there a callout on the print that controls the Cylindricity? I would collect a lot of data on each of the cylinders to evaluate the cylindricity, to eliminate or confirm that as a root cause (comparing to the Derek's LSQ method). If the Cylindricity exceeds specification, then (obviously) the parts are not "good." If the Cylindricity is within spec, but the form *is* the root cause, then Engineering needs feedback, to take this into consideration. IF (for some end of story reason) Engineering is unwilling/unable to refine the form tolerances to an acceptable level to evaluate the runout using "legal" evaluation methods, and the only option is to use a "modified technique", this should be communicated and approved by all parties. The print can be modified to communicate the approved evaluation technique OR an internal work instruction can be developed to cover under what circumstances the "modified technique" can be used. Maybe I'm a stickler for adhering to "legal" evaluations, or maybe I'm a huge fan of CYOA (Cover Your Own A**), or probably a combination of both... but that's my two cents. Link to comment Share on other sites More sharing options...
[To...] Posted February 20, 2020 Share Posted February 20, 2020 Please sign in to view this quote. Link to comment Share on other sites More sharing options...
[An...] Posted February 20, 2020 Share Posted February 20, 2020 Please sign in to view this quote. Try to watch the landing zone of the Apollo 11 mission on the moon by a very large telescope. This would be very difficult because of the earth rotation and the very large distance. See attached.Contribution_20_02_2020_1.pdf Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in