This is how I 'renovated' my Ty37 chainplate knees about 12 years ago:
I removed the chainplate and cut into the knee along side of where the plates were attached into the knee to remove the internal 'E-shaped contraption' that holds the non-removeable studs, and most importantly then removed all the wet internal 'softwood' that was apparently used as a form over which the knee was constructed. As stated before, this OEM knee was a damn good idea ..... if the knee was kept sealed from water intrusion. However when the interior of the knee does become wetted and then rots the studs SEQUENTIALLY loosen and the chainplate and studs start to rust AND you get 'saddle stress' cracks (sequentially) across the face of the plate emanating horizontally from each stud borehole in the plate - such saddle stress failures look like mini-happy faces. All this is due to rot and moisture under the plate in the knee. The only real problem with the OEM 'design' is that the studs are permanent and the wood core retains moisture.
Remove that wood, put nuts on that inner plate, put ferrules through a FRG core so the bolts are removable, beef up that knee because you disturbed/cut it --- all is well.
So, cutting through the knee next to the OEM plate, I removed the internal strap with the stud heads welded, all the wood, and tapered the cut to a 12:1 face angle
After chiseling, prying and cussing to remove all the sodden softwood, I then ground all the interior faces to 'clean' and added approx. 1/4-3/8" of roving/cloth to all the internal faces - this to add stiffness to the remaining parts of the knee. I built 2 new 'webs' of approx 1/4-3/8" FRG cloth - these 'webs' are perpendicular to the hull face ... the new 'box' fits 'sloppily' between the these 'webs'. My rebuilt chainplate knees are mostly 'hollow' (a fabricated box beam).
The 'box': I constructed an inner 316 SS plate to mimic the internal 'OEM strap with the studs attached', this time with 316 nuts seal welded to the new inner 316 plate. (Imagine the 'box' as having one side/face being the internal plate, the opposite side being where the the chainplate attaches, and flat sides between the two.) I then measured all the 'in place' dimensions, took the interior plate + welded-on nuts
made a form (a 'box'): Internal Plate with nuts, stainless steel 316 ferrules though which the new removable
3/8" 316 shoulder
BOLTS would pass through to the weled-on nuts, then liberally WAXED the internal bore of the ferrules and nut threads to prevent epoxy adhesion, filled the 'box' with cut and chopped FRG roving and filled the 'box' to a smooth
top face (when on a table). When cured, I simply drilled into to internal ferrules to clear any wax and errant epoxy and ran a 'long' tap into the nuts to clear them. This all created a solid (box) of strand filled epoxy + 3/8+" ID ss ferrules + 316 nuts on the internal plate, which I then inserted and epoxied in place and flat/flush to the area where the new chainplates would attach - exactly to where the OEM plates attached; I left the 'far' side of the plate/'box' with some extra room/space for the new bolts to penetrate through.
With the new 'internal structure' epoxied in place, I then overlayed the remnants of the OEM knee (overlapping to well beyond the sides of the OEM knee to the interior hull sides) and applied ~3/8"+ of new heavy FRG cloth/epoxy over the entirety of the old knee. The ferrules were easily seen through the new overlay, so it was easy to bore into the new (not fully cured FRG) to locate those ferrules.
Before the new epoxy/FRG overlay full cured, I covered the new chainplates with plastic wrap (so they didnt stick to the curing epoxy/FRG) and torqued them in place to the internal 'box' and knee structure ... in order to have a 'perfect
' flat face (friction surface) on the the new FRG overlay between the new chainplate and the new internal structure *** (see below).
*** When I was the assoc. chief engineer in a 'heavy/mega-lift' / hazardous design bureau I would NEVER EVER allow any structure to be constructed with ANY bolts primarily subject to SHEAR!!!! all such non-welded joints were to be full compression FRICTION JOINTS in which the tensile strength of properly torqued bolts supplied the FORCE to drive the two mating faces together; therefore, friction carried all the load; and if and only if, the bolts stretched the bolts then 'could' support the shear load on the shoulders of any bolt ... but that would be if and only if .... AFTER the frictional force joint failed. You need perfectly FLAT faces to affect a good friction joint. Such a friction joint also 'prestresses' the underlayment and makes it MUCH stronger (in shear, in bending/buckling, etc.). Bob Perry apparently used this very same design concept of a proper 'friction joint' in those Ty37 knees - admirable and quite remarkable.
My new plates were increased by ~25% thickness, that OEM 'kink' where the plates pass through the deck was made into a 'long radius bend' to lessen the obvious and vulnerable stress riser of the old plate (press brake) 'kinked' design. The plates are mirror polished to 'brilliant' - and 'slathered' with Boeshield™. I retorque (to 80%) the bolts yearly, replace them every 5-6 years.
I guess what Im stating here is that the OEM design was excellent stresswise; the only fundamental problem was moisture intrusion + wood core, and the studs werent 'removable'. My calcs of the OEM design indicated that the intrinsic Safety Factor well exceeded 4:1. Since I was 'disturbing' the old structure, I decided to beef the new replacement structure ... 'up a few notches' for 'just in case'.
I didnt want to go to the external chainplate route, as external plates require one to constantly re-reeve jibsheets every time one changes the fairlead car position; plus, external plates require a larger 'kink' (stress riser) in that chainplate. Plus, when I was PHRF racing this boat, I was sheeting and barberhauling the headsail 'inside' the shrouds, to make it 'point' better and external plates would be a 'REAL' PITA when doing so and without adding additional inside jib tracks.
If I were doing this all over today, Id consider to rip out the entirety of those knees and put in a system equivalent to what Beneteau uses - Rods instead of plates with kinks, those rods affixed to trunnions
affixed to the internal hull instead of 'bolts', and plate through-decks with leak-proof membranes that accommodate rigging movement (and no side loads on the through-decks). ---- Pure
and elegantly simple
straight-through tensile stress accommodation - a nice evolution. (Someone at Beneteau probably designed heavy lift cranes, etc. ??)
Sorry but I was too busy and covered head to toe with epoxy to take any pics of these knee refurbs. I still have one left to do (totally hidden behind all that 'cabinetry, shelving and veneer, etc.' in the shower stall) so maybe Ill take some pics then.
FWIW - Assuming that you have replaced those disastrous screwed together Grand Deer rigging toggle bolts, that welded together bowsprit cranse collar is probably more vulnerable to failure than any other part of the rigging. I have a 'better' design ... in the shape of the greek letter 'phi' (Ø)where all the principal stress is carried by a continuous plate between the headstay and bobstay connection. The welding on the OEM cranse collar is very vulnerable to cracking - check it often.
The infamous Grand Deer toggle bolts: