AI for Aerospace

I want to ask about application of timeseries in flutter phenomena, i have a project about it and i think i am stuck. My project has this problem statement,
" given data about flutter properties in several mach and flutter speed , build a model to be able to predict the properties of flutter in all mach and flutter speed region the input of the model should only mach and flutter speed and the output can be all the aerodinamics properties" i have worked with it several weeks, and my result is like this
image
, it seems i can only predict a data with the same mach number and flutter speed i have in training data so if i used mach number 0.6 in training i can predict the data in those mach number but i cant use it to predict a data in mach 0.61 or 062 because i dont have it in training data. i am so confused how to make my model able to predict with the input just a scalar(?)" my dataset is like this,
image
image
thank you very much for responding my question

i am using LSTM for the current result