aasikolo
asked on
Gates Gaudin Schumann Plot in Excel
Good AFternoon,
I have been trying to figure out how to perform the Gates-Gaudin-Schumann Plot using Excel, but am missing something, as usual. The articles on the subject use different formulas for obtaining the slope m for , y=mx+b, as well as the constant b, and I can't reconcile the results.
The distribution modulus should be around 200-ish, but am obtaining decimal results for this value?
The Excel file is attached for your convenience.
I appreciate any guidance you can provide,
Thank you,
aasikolo
GGS-Test.xls
I have been trying to figure out how to perform the Gates-Gaudin-Schumann Plot using Excel, but am missing something, as usual. The articles on the subject use different formulas for obtaining the slope m for , y=mx+b, as well as the constant b, and I can't reconcile the results.
The distribution modulus should be around 200-ish, but am obtaining decimal results for this value?
The Excel file is attached for your convenience.
I appreciate any guidance you can provide,
Thank you,
aasikolo
GGS-Test.xls
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
aasikolo,
I believe the approach I used is equivalent to the GGS plot.
Whether you use log base 10 or log base e doesn't matter because log base 10 of x = ln(x)/ln(10). When you calculate the slope, the ln(10) term appears in numerator and denominator and cancels out.
The power law correlation results in a straight line on log-log paper. It is exactly the same as the linear fit of log y versus log x.
I posted the power law approach because it uses native Excel features rather than the more laborious method I found on the web for GGS plots.
My first step was to plot the data on a log log chart. Neglecting the point at 0 micron mesh opening (because log of 0 is undefined), I noticed that the last two deviated from the straight line--so I deleted them from the first series and put them in a second series. This is a judgement call, and you could have alternatively decided to eliminate the last three points.
My day job is that of a process engineer, so sieve analysis is something I have actually done a few times.
Brad
I believe the approach I used is equivalent to the GGS plot.
Whether you use log base 10 or log base e doesn't matter because log base 10 of x = ln(x)/ln(10). When you calculate the slope, the ln(10) term appears in numerator and denominator and cancels out.
The power law correlation results in a straight line on log-log paper. It is exactly the same as the linear fit of log y versus log x.
I posted the power law approach because it uses native Excel features rather than the more laborious method I found on the web for GGS plots.
My first step was to plot the data on a log log chart. Neglecting the point at 0 micron mesh opening (because log of 0 is undefined), I noticed that the last two deviated from the straight line--so I deleted them from the first series and put them in a second series. This is a judgement call, and you could have alternatively decided to eliminate the last three points.
My day job is that of a process engineer, so sieve analysis is something I have actually done a few times.
Brad
ASKER
Thanks for your insight. This makes the job a whole lot easier.
Thanks again,
aasikolo
Thanks again,
aasikolo
ASKER
Your answer raises some interesting questions. The GGS plot is the least squares linear regression of the log base 10 of the cumulative percentages over the log base 10 of the mesh sizes, used for determining particle size distributions. You have a better fit using your method, however.
I'm still looking to reconcile the linear slope calculations and the distribution moodulus k if you have any ideas.
Thanks again.