07-03-2013, 06:48 AM
07-03-2013, 08:54 AM
I've sent Marv an email with the same information as in the thread. We'll wait and see where my thinking has gone astray.
07-03-2013, 09:15 AM
Thank you Arvid! After reading the article yesterday, I stared at Marv's code and knew there was something wrong with it but could not put my finger on it. No wonder I keep getting a negative number for the mandrel diameter when I select phosphorus bronze with my program. It's curious that it's not the case when I run Marv's version of the program. I get the same results whether I select music or bronze wire.
Excellent analysis!
With your information I can now finish up the program.
Ed
Excellent analysis!
With your information I can now finish up the program.
Ed
07-03-2013, 09:33 AM
"k = mx + b
where m = -0.0124727272727273 and b = 0.980818181818182 for music and stainless steel wire
and m = -0.0109727272727273 and b = 0.984909090909091 for phosphor bronze wire.
and x is the Spring Index."
Arvidj, are thes values for m and b constants? Or are they just for the example in the article?
Chuck
where m = -0.0124727272727273 and b = 0.980818181818182 for music and stainless steel wire
and m = -0.0109727272727273 and b = 0.984909090909091 for phosphor bronze wire.
and x is the Spring Index."
Arvidj, are thes values for m and b constants? Or are they just for the example in the article?
Chuck
07-03-2013, 10:34 AM
(07-03-2013, 09:33 AM)chucketn Wrote: [ -> ]"k = mx + b
where m = -0.0124727272727273 and b = 0.980818181818182 for music and stainless steel wire
and m = -0.0109727272727273 and b = 0.984909090909091 for phosphor bronze wire.
and x is the Spring Index."
Arvidj, are thes values for m and b constants? Or are they just for the example in the article?
Chuck
They are constants derived from the graph, not from the example.
I looked at the graph on page 20 and entered the data points into Excel ...
Music or Stainless Steel Wire
Code:
5 0.922
6 0.907
7 0.893
8 0.880
9 0.867
10 0.854
11 0.842
12 0.829
13 0.818
14 0.807
15 0.798
Phosphor Bronze Wire
Code:
5 0.933
6 0.920
7 0.908
8 0.896
9 0.884
10 0.874
11 0.863
12 0.851
13 0.842
14 0.833
15 0.823
I then ran a linest function on the data to determine the "best fit" straight line that fit the data ... i.e. a line that is defined by y = mx + b. The results are:
Music or Stainless Steel Wire
Code:
-0.0124727273 0.9808181818
0.0002209421 0.0023172607
0.9971838746 0.0023172607
3186.8803611737 9.0000000000
0.0171125818 0.0000483273
Phosphor Bronze Wire
Code:
-0.0109727273 0.9849090909
0.0001841018 0.0019308762
0.9974728525 0.0019308762
3552.3275535085 9.0000000000
0.0132440818 0.0000335545
The first row, first column is m and first row, second column is b.
The third row, first column the coefficient of determination ... a value between 0 and 1 that indicates how accurately the mx + b formula predicts the given y values for the data that was provided. A value of 1 means 'perfect fit' and 0 means 'not even close'.
In this case the values are so close to 1 that we are safe in assuming that "for any value of x [the spring index], the value of y [the elusive k] is ((-0.0124727273 * [spring index]) + 0.9808181818) for music and stainless wire and ((-0.0109727273 * [spring index]) + 0.9849090909) for phosphor bronze wire.
07-03-2013, 10:39 AM
(07-03-2013, 09:15 AM)EdK Wrote: [ -> ]...
It's curious that it's not the case when I run Marv's version of the program. I get the same results whether I select music or bronze wire. ...
Ed
Very possibly a case where the code and the exe do not match. Developers love those types of challenges ...
07-03-2013, 11:12 AM
(07-03-2013, 08:54 AM)arvidj Wrote: [ -> ]I've sent Marv an email with the same information as in the thread. We'll wait and see where my thinking has gone astray.
I just got this from Marv ...
From: Marvin W. Klotz
Sent: Wednesday, July 03, 2013 11:04 AM
To: Arvid Jedlicka
Subject: Re: Possibly a bug in the mandrel application ...
Arvid,
Good catch. I don't think I've ever seen such a thorough analysis of a
suspected error. I'm convinced. I've incorporated all your constants
into the program and prepared a new version to go up on the website soon.
Thank you so much for taking the time to work through this and advise me
of the problem. I really appreciate your efforts.
Regards, Marv
Home Shop Freeware
http://www.myvirtualnetwork.com/mklotz
07-03-2013, 02:10 PM
I got the Windose program running now with Arvid's new numbers. I'll upload it once I've done some more testing.
Ed
[attachment=5590]
Ed
[attachment=5590]
07-03-2013, 02:12 PM
Ed,
When you get all the details worked out, I'll give it a real world test drive.
Tom
When you get all the details worked out, I'll give it a real world test drive.
Tom
07-03-2013, 02:13 PM
(07-03-2013, 02:12 PM)TomG Wrote: [ -> ]Ed,
When you get all the details worked out, I'll give it a real world test drive.
Tom
That would be wonderful Tom! I'll have it uploaded this evening.
Thanks,
Ed