UltraScan III Frequently Asked Questions:
LIMS fitting error messages:
- My 2DSA fits fail with the error: ***ABORTED***: Implied Grid Size is
Too Large!
That error message is a bit misleading. The grid
size this error is complaining about is not the s/f grid, it is the number
of radial discretization points in the adaptive finite element solution
describing the point at the bottom of the cell where you normally find
back diffusion. You will get this error if either the selected s or f are too
large for the speed you are simulating, and there is no back-diffusion,
which makes this transition point too sharp (i.e., where the solution
goes from horizontal to vertical at the bottom). Then the program
tries to put additional points into the corner edge to try to make it
a continuous solution. When the transition is too sudden or too steep,
it can't fit enough points in there, so you effectively end up with a
discontinuity or so many points that the adaptive grid is too large.
The solution is to reduce the the s-value maximum and/or the f/f0 maximum and retry.
- My 2DSA fits fail with the error: ***ABORTED***: Meniscus value
extends into data
This error occurs when your meniscus fit
region extends into the left data range limit. Example: Let's say your meniscus
is at 6 cm, and your data range is defined to start on the left side at 6.01 cm,
then the following problem will happen: When you do a meniscus fit, the
program will take the meniscus position you assigned during editing and move it
to the left and to the right of the meniscus and re-fit to see which position
close to the originally chosen meniscus will give a lower RMSD. The amount the
program will move the meniscus around is defined in the LIMS under "Meniscus Range".
By default, this will be 0.015 cm. With this range, the program will move the
meniscus starting at 5.985 cm all the way to 6.015 cm (using 10 points by default)
and re-simulate the solution. If the meniscus crosses the point that was
defined to be the left data range (6.01 cm) then you have this problem, and some
of the points you will be simulating will have meniscus positions that are to
right of the left data range. This of course violates the boundary
conditions of our Lamm equation and hence the program will abort. The solution is
to either make the meniscus range smaller so this doesn't happen, or to
change the left data range limit and move it a little bit further to the right.
To check what your actual definitions are, open up the editor, load your dataset
and apply prior edits. Then you can read off the starting meniscus position and
the left data range position.