Generating New Simulated Sedimentation Data
There are a number of different ways in SEDFIT to simulate the sedimentation process and save the result in a sequence of files with the XLA format. The first two take utilize the function Generate that allows generate from scratch a series of simulated scans for specific models. The last uses existing data as a template with regard to rotor speed, data range, and time interval. For simulating data of interacting systems, you can use SEDFIT to generate a template from scratch, and then simulate the distributions in SEDPHAT.
Alternative Simulation using Existing Data as a Template
Simulating Interacting Systems
This is probably the quickest way to simulate sedimentation of a single species, or if you have a mixture of a few (< 4) discrete species. For example, one can use this to predict the minimum time to reach equilibrium and the shape of the equilibrium profiles at a certain rotor speed.
Make sure you select the Non-Interacting Discrete Species model. This works also with the inhomogeneous solvent options switched on (but you will need to familiarize yourself with these features first, and have the necessary parameters already set; I suggest to explore these options with existing data before attempting to simulate them).
It is best to set the parameters already to the values you want to simulate. Note that the first concentration is the total, the following are fractions. For example, with the concentration entry in the first component being 1 and the second concentration entry 0.05, will make the total loading signal 1, the first species 0.95, and the second species 0.05 loading concentration. Be careful also to set meniscus and bottom to describe the proper solution column.
For studying the approach to equilibrium, I would suggest to request scans in a few h intervals, over a time period of a couple of days. That will help you to explore the shape of equilibrium, and by counting the number of scans to attain equilibrium, you can figure out the minimum time (e.g. assuming no chemical reaction) to attain equilibrium.
Another frequent application is the prediction of c(s) resolution and faithful recovery of quantities of trace components. In this case, simply simulate the distribution of discrete species (optionally with dynamic density gradients, if that model is selected), specifying the mixture of interest.
Follow the instructions for the Generate function.
This is more complicated than the simulation of discrete species, and should be used only if either the number of species of interest exceeds 4, or if you are truly interested in continuous distributions. There are several steps.
1) Select the c(s) distribution model.
2) Use any existing data, load the data, set up the c(s) parameters as you wish your simulated distribution to look like, with regard to: s-min, smax, resolution, f/f0, vbar, viscosity, density), and fit a c(s) distribution (never mind if it is a good fit or not). Use the function Data-> Save Continuous Distribution to save the c(s) distribution into a file. As a filename, use c:\temp\distrib.dat. This will be a template to modify. Note: You can generate such a file from scratch, using the notepad. But it cannot be an empty file.
3) Use the Windows Explorer to locate this file, open it with notepad, and edit. You will see a 2-column matrix, with the first column the s-values, and the second column the distribution values. You can delete as many lines as you like, or enter new ones. This will be your sedimentation coefficient distribution.
4) Use the Generate function. (Make sure the c(S) model is still switched on.) In the parameter box, use the desired settings for f/f0, vbar, viscosity, density. You can ignore the s-min, s-max, and f/f0 entries, since it'll take this from the file c:\temp\distrib.dat.
5) A question will appear if the distribution is normalized (a true differential sedimentation coefficient distribution, where the integral corresponds to concentrations), or not. If not, the second column of the distrib.dat file will be interpreted directly as loading concentration values.
Note: Even though you use the distrib.dat entries directly as loading concentrations, the distribution displayed after simulating will be a differential distribution, which will appear distorted. However, the data loaded and saved are perfectly fine.
Alternative Simulation using Existing Data as a Template
You can easily simulate sedimentation data for which you have already a template regarding time-intervals, rotor speed, and radial data range. Simply load these data as a template, select your model, and do a RUN. Be sure to switch of the baseline parameters and the concentration parameters. Obviously, the "fitted" traces may have nothing to do with the loaded data template. However, they describe the distribution from the model you've specified.
Use the function Data->Save Fit Data to store the simulated distributions on the harddisk in XLA format.
The disadvantage of this approach is that it doesn't have flexibility with regard to the rotor speed and time-interval of scans. Also, the radial range of the saved data is the fitting range specified. Although you can make sure that the "fitting range" is as wide as the template data permit, it cannot exceed the radial range of the template.
Simulating Interacting Systems
This is also a three-step process.
1) Use the Generate function with a discrete species model to generate a template with regard to rotor speed, time-interval of scans, and radial range. Save the data somewhere in a temporary directory. Export the to SEDPHAT, and launch SEDPHAT.
2) Within SEDPHAT, use the experiment parameter box to set the extinction coefficients, meniscus, bottom as desired, and switch off baseline parameters. Then select the interaction model, set the global parameters, set the loading concentrations, and do a RUN.
3) Use the function Data->Save Fit Data within SEDPHAT to save the fitted curves containing the simulated distributions to the harddisk.