Idled due to an idler

Haven't posted much lately, but I have worked on the Pitts quite a bit. There's a finding that slowed down my reassembly; here's the story:

I was almost going to start to put the airplane back together, but when I looked at the front end of my torque tube, I just couldn't keep it like this. Do you notice what's wrong on this picture? (except from the red dye penetrant residue and the burned paint)

It is subtle on the picture, but if you look closely at the very front end of the torque tube, you will see that the bolt that goes though the front bearing doesn't protrude enough to be compliant with the airplane Bible a.k.a. AC 43-13.

So I decided to bite the bullet and remove the torque tube assembly in order to replace that bolt with a longer one. While doing that, I took the opportunity to inspect each component of the assembly and I was shocked to find the forward idler like this:

As you can see, the part is twisted, so was my facial expression. One thing for sure is that this part was not going to fly again. I compared my part with the original Pitts drawing similar to this one:

The first obvious observation is that the original part uses a plain bushing instead of bearings. After some discussions with my friend Peter, I realized that on the Pitts S1-T, this part is also made with bearings which actually help in minimizing slop in the flight controls.

The second obvious difference is that the original part does not have any lightening hole. I guess the original builder of my idler liked swiss cheese; he made the part just like it. I don't especially like the idea of having lightening holes in a dynamic member of my elevator control system. Looking closely at the second picture, you may see that the top center hole as well as the tiny hole on the right are both elongated. When a part is stressed in torsion, internal shear loads tend to propagate at 45° (see explanation here). The holes being along a 45° line from the part's axis, the part was much weaker in torsion than originally designed for. And it twisted. I was happy to find this out at 0 ft and 0 kt. 

A third difference exists, but this one is not obvious on the pictures: the total height of the part is 3/16" smaller than on the drawings. I have not found the reason why the builder chose to build it like this; perhaps to save weight. When I realized that, I had to make sure that the geometry actually made sense. There is a famous sentence that is often used among the Pitts community: "Never try to second guess Curtis Pitts". 

I made measurements of the torque tube tower as well as the current idler and modelized it to have a clear view on the geometry. Here's what it gave:

Then I modelized the original Curtis Pitts assembly:

My initial fear was that the torque tube tower height was per drawing, but not the idler. However, the builder actually built the torque tube tower also shorter than the drawings, but perhaps not short enough according to the drawings above. The problem is that when the idler is fully aft (full nose down input) the shorter idler pushes the rod ends with more misalignment relative to the torque tube axis than the original design. When combining this misalignment with a roll input, the two pushrods can end up with a lateral misalignment as well. This can be a contributor to the torsional load on the idler.

Bottomline, both the shorter idler and the lightening holes were acting towards exceeding the part's limit loads in torsion.

No way I was going to but back a twisted critical part on the airplane. I had to replace it, which meant making a new one from scratch since it was not per original drawings. I decided to make it a little longer in order to have a better alignment with the torque tube axis while keeping sufficient clearance with the bottom of the torque tube:

In the awesomely well equipped shop of my friend Scott, we built the parts with 4130 steel:

The logical next step was to call back Mark, the welder I hired to repair my fuselage. Again, he did an excellent job. Here's old vs new:

Today I primed my new idler:

Next step is to paint and finally begin the reassembly! Oh I almost forgot to mention I made two new access holes in the fabric to provide better access to the aft idler as well (yes, the fabric shows its age on this one... )

Lots of work to install the reinforcement rings, but this will facilitate maintenance in this hard to reach area.

Thanks for reading!