I must agree with "truth is not a compromise". Your best option is compound the mixture, or a set of mixtures, and catalog the results of the density determinations.
As far as I know, any attempts to estimate the bulk density from the known densities of the materials will result in a higher number, since you will undoubtedly incorporate some air bubbles with the iron filings when you perform the experiment (unless you de-gas the mixture by pulling a vacuum). Or you could use controlled mixing/agitation for extended periods of time on a larger blend (say 2-3L) while removing small samples of known volume for weighings, and plot them. The line could then be extrapolated to some asymptotic value that represents the maximum density of the controlled mixture (for each proportion). Once completed, you will have a data set that can be used for process control?
__________________
If it ain't broke, don't fix it. Just build a better one.
Good Answers: