Injecting material into the jet, and then turning off the flow is representative of any density increase in the jet flow. As the material
moves to greater radii the flat spectrum builds up, the infrared flux remains the same, and the millimetre flux increases as the flat spectrum moves down towards it. Then when the complete range has been reached the jet flow is turned off, and the material left to expand freely without
further injection. We instantly lose the infrared flux, but the millimetre flux gradually decays.
This is not really the sort of behaviour observed from GRS 1915+105.
A better light-curve profile can be produced by gradually
increasing the density of the injected material, and then gradually
decreasing it (say a Gaussian density injection function), but there will still be a time-lag between infrared and millimetre fluxes.
However, the major shortcoming of this model is that the flux is much too small, to get that calibrated correctly requires an increase in the electron density to the point where it is unphysical.