Once a size of spring had been ascertained in a basic model, with slightly unrealistic constraints, it was modelled again in a more representative simulation. The portion of the instrument where it sits was taken from CAD. This allowed us to check the effect of adjusting the position of the u-spring in the throat (hence adjusting the moment on the beam end of the spring).

This gave us confidence that were we to design a spring of suitable deflection in the FEA software, we could replicate this result in real life.


It was decided to keep the thickness of the material the same, as this suited supply and fit to instrument, but to increase the width of the flange, and thus the moment of inertia and resistance to deflection iteratively until the correct result was achieved.

A prototype part was created and allocated a number. A Works Order was raised with Operations to review the product after every stage of manufacture to test deflection pre and post heat treatment (hardening). Hardness does not affect deflection (technically it does, but in a highly localised way), but we were worried about the part’s changing shape or springing back during stress relieving. To this effect we tack welded some laser cut braces into place.

Prior to this however we performed a deflection test.

The results of the deflection test correlated very closely with the Finite Element Analysis (within a few percent).

We performed the same test after stress relieving and found the same results. The part was then tested on the instrument and achieved the desired results.


The U-spring part could now be produced in-house ensuring guaranteed quality, continuity of supply and suitability to instrument build.

This part also proved our ability to solve problems quickly using theory and the FEA software available to us and to put the theory into practice in a quick and efficient manner.