Radius gauge and reverse engineering

**GKK** · Dec 28, 2018, 08:36 PM

Thanks. A tool I do not have, but sure could use.

**scrdmgl** · Dec 29, 2018, 05:09 AM

Hi Tony:
Could you please describe the principle behind this radius gauge? I have uses for it but I need to understand its application.

Thanks

Jorge

**tonyfoale** · Dec 29, 2018, 08:08 AM

Originally Posted by scrdmgl

Hi Tony:
Could you please describe the principle behind this radius gauge? I have uses for it but I need to understand its application.
Jorge

Did you read the PDF file? I think that it is explained well in that. If that doesn't answer your needs then please ask a specific question and I'll try to help.

**Toolmaker51** · Dec 29, 2018, 04:00 PM

Originally Posted by scrdmgl

Hi Tony:
Could you please describe the principle behind this radius gauge? I have uses for it but I need to understand its application.

Thanks

Jorge

Not to intercept Tony, with such a well developed tool and programming:
The dial indicator reads contact, plus or minus of plane established by the outboard 'feet'. Knowing distance between outboard feet [chord], and midpoint distance found by indicator [arc] boils down to representative geometry, or calculative trigonometry. It's all around us...roadways, bridges, ballistics, architecture, navigation, astronomy, surveying, even ocean towing by chain/ wire rope.
And recreating castings from machined billets.

**Phantom89** · Dec 30, 2018, 05:08 PM

Awesome👍👍 such a simple idea but very useful. And the cherry on top is the wooden case. Nice touch.

**Mr.Pete** · Dec 30, 2018, 07:35 PM

Thanks for this nice tool. I have to make one, its soo cool !

**prm** · Dec 31, 2018, 06:14 AM

Originally Posted by tonyfoale

I have just enhanced the software to allow calculation of the curve radius and centre coordinates when the XY coordinates of three points on a curve are known.
This is useful if you use the DRO on a milling machine to determine the coordinates of three points. This does not need the radius gauge described above.
The link to the software is the same as that in the PDF RadiusGauge.pdf

Tony,

I always read your articles with interest. As I had not used my brain over Christmas, I decided to work out the formula for the radius before reading the appendix.
I ended up with a different and, perhaps, a simpler formula: R=(s^2 + h^2) / 2*h
(where s is half the distance between the probes, h is the value on the indicator and, ^2 means 'squared')

I hope you won't mind if I also note that there are some typos in the pdf file : in the sections were you discuss errors in the indicator reading and errors in measuring the probe spacing, the radius value are not correct.

Regards,
Paul

**tonyfoale** · Dec 31, 2018, 08:52 AM

Originally Posted by prm

Tony,

I always read your articles with interest. As I had not used my brain over Christmas, I decided to work out the formula for the radius before reading the appendix.
I ended up with a different and, perhaps, a simpler formula: R=(s^2 + h^2) / 2*h
(where s is half the distance between the probes, h is the value on the indicator and, ^2 means 'squared')

Paul,

I always like to find an easier way regards of where it comes from but I just don't follow your formula. Please show the full derivation. I assume that brackets could enclose the 2*h as (2*h) for clarity and that it is not ((s^2 + h^2) / 2)*h, to remove any ambiguity. In any case I am just not getting it, sometimes we get tunnel vision.

Originally Posted by prm

I hope you won't mind if I also note that there are some typos in the pdf file : in the sections were you discuss errors in the indicator reading and errors in measuring the probe spacing, the radius value are not correct.

I never mind it when errors are pointed out. Thanks, I'll check it out later, I am on the wrong computer at the moment.

**prm** · Jan 1, 2019, 03:50 AM

Originally Posted by tonyfoale

Paul,

I always like to find an easier way regards of where it comes from but I just don't follow your formula. Please show the full derivation. I assume that brackets could enclose the 2*h as (2*h) for clarity and that it is not ((s^2 + h^2) / 2)*h, to remove any ambiguity. In any case I am just not getting it, sometimes we get tunnel vision.

Tony,

I have produced a document with the derivation. It is very convenient that the r^2 terms cancels out and leaves a remarkably simple formula.

Cheers,
Paul

**tonyfoale** · Jan 1, 2019, 06:27 AM

Originally Posted by prm

Tony,

I have produced a document with the derivation. It is very convenient that the r^2 terms cancels out and leaves a remarkably simple formula.

Cheers,
Paul

That's much better. At first I did think about going the (r-h) route but I thought that it would end up being more complex so I did it the way that I did.
I'll amend my pdf to show your derivation.