Hunter Impala Mast Heel renewal

This is the report of how we replaced Alchemy's mast heel, along with the mast plug and T base. It includes some background, my thought process, some technical details, and some (ever so wonderful) hindsight.


From this
To this

Background: We bought Alchemy in 2022 and that winter we unstepped her mast. It was then when we first noticed that 6mm clevis pin holding the Selden mast heel within the notch of the T-base was severely bent, and the heel was broken around the pin hole. The optimistic me then glued the broken pieces with JB-weld and replaced the pin with a brand new one. Come 2025, the heel was crumbling again and it was time to do something more serious about it.


The old heel after the removal of the plug. The plug sits on the two bosses and is held in place with a M8 bolt. It had to be drilled so the plug could be removed.

The mast heel assembly on old Seldén/Kemp deck-stepped masts has three parts (unlike most modern systems that consists of a mast heel and a T-base). In the old system the mast plug is separate from the heel. The two are connected with two bosses in the casting and an M8 bolt. According to Seldén, for sections used in Impalas the heel is a 502-076, the plug is a 502-052 and the T-base is a 510-113. The part number of our heel and plug matched this, but our T-base was a 510-105. However an old Seldén catalogue that I digged out of the internet says that 510-105 was meant to be used with different heels. Moreover 502-076 seems to have undergone variations in its design and installation. So I was a bit in the dark as to what the "right" setup should look like.

The 502-076 heel hasn't been manufactured for decades. The Kemp mast extrusion profile is obsolete as well, and the new heels don't fit these. My searches in riggers for a second-hand heel in reasonable condition was fruitless. In this case our options were:


Welding might have been the cheapest and quickest option. However the casting was not in great condition overall. It looked like it would start developing new cracks and the heel had a significant "elephant foot". The welders I spoke to were not keen on the job either, and nobody would guarantee a successful, lasting repair. I seriously considered the option of adapting the mast for a modern heel. However I was put off by the idea having to cut the mast to accommodate for the height of the adaptor piece. I also wanted to be able to use the old mast plug in case it proves impossible (for me) to produce a new, working plug. With enough information on the original designs, the decision was made to make (or have made) new parts, or at least attempt it.

The CAD process: The heel is the most complicated part, so I started drawing that in FreeCAD as I am vaguely familiar with this software. It was a good choice until it was time to put the chamfers in the final step. It turns out FreeCAD is notoriously bad at that. It often "says no" and/or crashes. The preceeding workflow is a crucial factor in the possibility to add most of the important chamfers. After days of fiddling, I manage to implement all but one of the chamfers. The last one had to be added manually using a file by the people who CNCd the design. Perhaps I should have used other software that was recommended to me, such as Onshape. However the learning curve made me hesitate, and it wasn't until the late stages that I would find out the drawbacks of using FreeCAD. In any case, despite all the frustration after many hours of work I was able to produce a file that worked. In the meantime FreeCAD has progressed considerably and it might well be that the new version can do these chamfers with no issues.


The heel model in FreeCAD

Prototyping: I rarely build things where I get a single shot the failure of which means expensive consequences. Before placing an order I had to make sure the design was correct and final. I found 3d printing (which I am relatively new to) is very useful for this. I was able to assemble the T-base, mast plug and the heel itself, and check whether the sheaves are in the right place, and the pin holes were aligned. This helped me to catch several issues with my CAD models, the drawings, and my modifications to them. I printed about 6 versions of the heel, 3 T bases and 3 plugs. One of my biggest concerns was whether a custom mast plug would fit into the extrusion or not. (It actually worked in the first attempt! 3D prints using PLA were too fragile to check the fit, and broke with the lightest tapping to insert the model into the mast extrosion. The ABS version responded well to encouragement, and it showed that I got the dimensions right the first attempt!

Manufacturing: Of course committing to building custom fittings required a plan for their actual manufacture. My contingency plan was using the 3 axis CNC at the Makerspace So Make It where I am a member. Again, in hindsight, I'm glad that I didn't have to do that. I was also enquiring about getting some time on a metal 3d printer. In the end, an engineer friend's suggestion was a no-brainer: He pointed me to a website called "Craftcloud". The customer uploads a CAD model, and enters specs such as material, tolerances and finish. Almost instantly manufactureres worldwide provide the customer with quotations and lead times. It turned out I could get all three parts machined for less than the price of a new heel, adaptor piece and a new T base.


3D printing prototypes helped me check the designs. The first iteration of the plug did not cover the mast extrusion well, so it was modified before manufacturing.

Once I was happy with the designs I submitted the designs to a manufacturer on Craftcloud. I had the option to use 5xxx-series aluminium or 6xxx series. I opted for 5xxx, bead blasting and black anodisation as it seemed to be a closer match to the original. The lead time from China was 13 days which was incredible but turned out to be correct.


Delivery from Shenzhen

Design modifications

I made a number of modifications to original Seldén designs.

The mast heel: The T base
Delivery from Shenzhen

Other hardware

The mast heel also needs sheaves for the halyards and pins to hold them in place. I have never been happy with the friction on the jibs and wanted to improve the solid sheaves of the old heel with better ones. In the process I learned that ball-bearing sheaves are not great for static loads like halyards. In the end I found that Antal make 50mm Mekton sheaves that have bearings on their side. I thought these would be a good option. The original sheaves are 45mm so I had to modify the positions of the pin hole slightly. The two pairs of aft-facing sheaves need standard clevis pins. However I couldn't find a suitable pair so made them myself on the metal lathe at So Make It. The pins of the 90º sheaves are a little more intricate. The wall aft of the sheave is not very thick and can only 2-3 threads of 1.25mm M8. In fact tapping this was one of the most stressful parts of the job. I also wanted eyes for the halyards on the forward face of the heel. So I made two stainless pins with M8 threads on both ends. The forward end takes the lifting-eye nut shown in the photo. I did not use Duralac here as I probably want these to corrode and become a part of the heel. I am not good on the lathe and can't machine stainless with a very food finish. I was also too impatient to polish it. This isn't good as some traces of rust has alreay appeared on these pins.
Making the clevis pins for the sheaves



Installation

To install the mast plug, I first cleaned the inside of the mast with fine sandpaper. Then I put some duralac on the mast heel. I don't know if this is necessary as both the plug and the mast are made out of aluminium but if I ever need to remove the mast heel, I'd like it to be easier than last time. It didn't take much encouragement to tap the mast heel into the extrusion. Light tapping with a mallet was enough and the heel was in place, with its shoulder flush with the mast. That was a huge relief, even if I had checked its dimensions with plastic models. The mast heel then screws into the mast plug, and I again used Duralac between the A4 stainless bolt and the aluminium plug.

I was also replacing the T base with a different design. So the old base had to be removed, its screwholes had to be sealed and the new T base had to be installed. The trickiest part for me was to make sure that the new T base is perfectly parallel with the centreline. I couldn't think of a reliable way of achieving this, nor was I certain what to reference. I marked the new screwholes in reference to the old ones, and verified the position by measuring the distance of its front and back to chainplates (over the curved coachroof). I tried to get it all right within 1mm precision. I also eyeballed using a sail batten which I hoped would exaggerate any unevenness. I sealed the new screws and the base using CT1. I'm sure it's a few degrees off but the heel itself has more play than that.
Making sure that the new T base is staight

The stepping of the mast was straightforward. I used the pivot pin to scure the heel on the T base, only because I lost the pin that goes through the notch and that's what I have seen on two other Impalas. One rigger advised me that using the notch on the T base would be better, and another said it does not matter.

Conclusion

The mast base is now installed and has sailed for a good few days in various conditions 🧿. As can be seen the coating is transforming from coal black to a golden colour - perhaps that's Alchemy!

Hindsight

And some hindsight - If I was tackling this project for the second time, I would do some things differently: