Duplicating curves with new criteria - Forecasting/Trending

Hello all,

I've hit a wall and looking for solutions/path forward with trending/forecasting analysis - a weak spot for myself.

So I have a cash flow over a period of 20 months that reaches a total of $15M.  I want to create a second set of data that has a similar curve/slope, however, this second curve will be stretched to 48 months and reach a total of $20M instead.  I have the monthly distribution of the first curve, just not the second. 

I have a large set of data containg approximately 7,500 different curves that I will want to duplicate this process on a mass scale.. so simplicity is key (if that's possible).  To add in a second (potential) challenge, many of these curves I anticipate to be different in shape -  i.e. linear, bell, and s curves.  I assume that would be a major factor in how this is all forecasted.

 Do any of y'all have any ideas of how to do this?  Any kind of feedback is helpful.

Jason