I wanted to create a computer simulation on Matlab, using a model for a pendulum from this study (A double pendulum model of tennis strokes. Rod Cross. Uni of Sydney, 2006) - Link I wanted to use the model from that study specifically, as it uniquely includes the motion of the elbow for various tennis shots, so it's effectively a triple pendulum. I've tried to contact the author and the author of similar studies but to no avail.
Before I created a simulation, I wanted to check I could replicate the results of the study. I inputted the formulas into excel, which I'm more familiar with then Matlab. Link for spreadsheet. I tried to replicate the diagrams in Figs 2 and 4, using the equations in the appendix of the study. The shape of my charts matches that of the study, which is encouraging, but the values are too high.
I've read all the references but I'm none the wiser, as they layout the same/similar formulas but don't show a result in the form of a number, that I can use to compare against my own. Changing the numerical method from Forward Euler to RK2/4, or to a much smaller timestep makes very, very little impact. So, I feel that can be ruled out.
The constants such as mass, moment of inertia and gravitational acceleration are all as shown in the study. Icm1/2 and I1/I2 are common formulas for Moment of inertia at different axis, that I know well, so they're definitely right.
I've checked all the formulas on MathPapa and Mathematica. They're all inputted properly. So I think the problem must be with the metric I've used. I used SI, so I've used kg, m, and rads, including for cos and sin. Excel calculates Cos and Sin as an input of rads and an output of rads, but I don't think thats an issue as rads are the natural units for angular rotation in the SI system.
It's a lot to ask but if anyone is willing to read the study and check my spreadsheet. Or it could be solved using other avenues, like Mathematica etc.l I'll keep checking on here regularly and try and work through it all. The spreadsheet is not far off being right. Thanks.