There may be occasions where profile parameters will be refined but should have the same values for several histograms. A good example of this might be where multiple samples are used for characterization of the instrumental broadening, for example to extend the Q range of the fitting. While each histogram has a separate set of instrumental parameters, constraints can be created to force the parameters to have the same values.
Likewise, if the same specimen is used with multiple diffraction measurements, it can be reasonable to constrain a phase’s sample broadening term(s) to be the same for all the histograms. Note that some thought should be made first to know that the measurements really do probe the exact same material. For example, a multigram neutron specimen and microgram synchrotron specimen might come from different synthesis batches and might have different crystallite sizes or microstrain values. Even when taken from the same bottle, the sampling may not be homogeneous. Use constraints with care and it is a very good idea to try a refinement where the constraint is removed. Without the constraint, parameters may refine to inappropriate values, but if the constraint is valid the fit without the constraint will not be significantly better than the constrained fit.