Different answers each run halving step size, debugging code correctly

Hello,

I have a code computing a rockets trajectory using Rungekutta. The issue is computing error, I am trying to compute and estimated global error, by comparing two results with different step size. There is a while loop in the code, which halves the step size to compute local truncation error and minimize this. However, I descided to deactivate this, so the issue is not here. I have tried to search trough the code various times, but I can not seem to find an issue anywhere.

Any help is greatly appriciated

[Walter Roberson]

A common error in RK type of calculations is failing to multiply by the step size in all of the appropriate places.