Hi Robert

We have considered your observations and we have modified  our proposal.

The new drawing are summarized in the attached sletches.

Now the insert in the grid has an external 1/2" thread . The tolerance

between the insert thread and the threaded hole in the grid (about

10mils) will be filled with epoxy. The epoxy will work in compression

for the orthogonal forces, the thread will  work against on axis forces.

Inside we have a cylindrical part 3/8" diameter, 3/8" lenght + 3/8"

theraded hole, 1/2" lenght. The inside thread can be 3/8". This is a

small piece, it can be done with a good alignment between cyilinder and

thread axis. In the flexure head we think a  3/8" hole. The tolerance

between the shoulder of the screw and the holes can be

holes = 3/8"-0+0.0005"

screw=3/8"-0.0005"+0

With the new design:

1. The dimensions  fit with the grid constrains now.

2. The joint is exentially a glue thread joint

 

The concept: DECOUPLE ALIGNMENT FROM FIXATION, AVOID ANY REQUEST OF

TIGHT TOLERANCE ON THE FLEXURES

with a simple tool, using the reference holes in the bottom tray, we can

refer the tower to the grid (better than 4mils). In flight units, the

tower will be held up by some device (lifting device). The pins and the

alignment tools keep the tower in place without  a real  g force on them

After a dry test, the tower is removed, the inserts are put in place

with the epoxy, the tower is put back, the screws are inserted. We do

not need a strong tolerance between flexures and inserts, because the

1/2" thread have about 10mils tolerance. If needed this tolerance can be

increased a bit

Advantages:

NO ASIMMETRY, All the 12 joints are identical

Delicate works on small piceces only; the required tolerance to be able

to put the tower in place is easy NO RISK TO MISS ANY SCREW ON ANY OF

THE 18 TOWERS

If there is a problem (p.e. a hole in one flexure head is  too large or

too small) it is easy to machine a dedicated insert  and screw

If we have to dismount the tower, the holes in the flexures  remain

clean; to restore the grid we have to machine away the insert (no

special tolernce is needed)

When we dismount the tower from the simulation grid, the tower flexures

are clean.

Sandro and Riccardo