I could have modelled this just using piece-wise functions with restricted domains. When it was loaded into Geogebra though it was clear that the bottom left section was doing more than just being horizontal. There is a definite negative (but increasing towards zero) gradient… sounds a bit like an exponential decay function to me.
My Geogebra model – feel free to use it and modify it to see if you can make it a better fit to ‘the data’, AKA the wall.
Pro-Tip: The thicker the line, the better the fit. This is a joke.