Monday 19 October 2015


In the fall of 1825, Sir Edward Parry was court martialed at Sheerness for the loss of one of his arctic exploration ships, HMS Fury, which he had abandoned at Somerset Island.  He and his officers were easily acquitted, but during the trial he admitted that things might have been far worse, if not for a new piece of equipment installed on his ships. He claimed that a new capstan, Captain Phillips’ Patent, had saved the expedition from at least another year trapped in ice.

Before the 19th century, capstans were largely direct drive machinery, meaning that all of the capstan's power came from the men turning it. However, with the introduction of geared capstans, the power of these fittings was greatly increased.  Inventors had experimented with these “improved”, or “power”, capstans since the mid-18th century, but poor manufacture of the gears appears to have hampered their reliability. (1)

Charles Phillips, then a commander in the Royal Navy, believed these powerful new capstans had the potential to revolutionize the way ships were moved using lines and anchors (critical for saving grounded vessels). In 1817, he submitted a plan to the Admiralty outlining a method for using them, but without a reliable power capstan, the plan could not be implemented. In 1819, he again returned to the Admiralty, but this time with a patent for his own improved design (Patent No. 4394).

An 1854 update of Phillip's 1827 capstan patent. 

His patent used planet gearing within an encircling gear ring to increase the power of the capstan by a factor of three. The gears could be rapidly engaged or disengaged using four drop pins. With the upper pins engaged and the lower pins removed, the capstan operated in direct-drive mode. When the lower pins were engaged and the upper pins were removed, the gear drive was enabled and the power was increased (and the speed decreased).

It seems the Admiralty began to adopt the new design, albeit in a limited manner, after 1819. Phillips was given a command as a result of his invention, and he continued to command vessels while filing new patents for nautical inventions. (1) John Richardson named an island after Phillips in 1826 and in 1829 he was nominated a Fellow of Royal Society, both largely on the popularity of his improved capstan.

Seeing the potential for using them in ice conditions, Sir Edward Parry installed Phillips’ Capstans on both Hecla and Fury for his third voyage to the Arctic in 1824.  He gave the highest praise to the devices, which he used to literally haul the ships through the ice.

Parry wrote:

“The strain we constantly had occasion to heave on the hawsers, as springs to force the ships through the ice, was such as, perhaps, no ships ever before attempted; and by means of Phillips's invaluable capstan we often separated floes of such magnitude as must otherwise have baffled every effort.” (2)

He continued, in a footnote:
“I cannot omit this opportunity of expressing my admiration of this ingenious contrivance in every trial to which we put it in the course of this voyage.” (3)

It seems Parry’s endorsement carried great weight with “…the Lords of the Admiralty, who, with a laudable zeal for the service, gave instructions that all the vessels of his Majesty's navy are in future to be fitted out Capt. Phillips's new capstans.” (4) It is no surprise then that Phillips’ Capstans were installed in Terror and Erebus in 1839 for Ross’ Antarctic voyage. 

The 1839 profile of Terror and Erebus clearly shows a Phillips style capstan installed on the upper deck, and the 1848 profile of HMS Investigator provides a virtually identical illustration. Unfortunately, the draughts are missing many crucial details, and therefore I was forced to conduct further research to create accurate plans. Like nearly every part of my HMS Terror project, this proved to be far more difficult that I originally assumed.

Because of the popularity of Phillips’ design, myriad updated versions of these capstans were constructed over the years. Besides the original patent drawings, several plans for different versions can be found in the historic literature. Additionally, several demonstration models have survived in museums, all with slightly different designs. Finally, a Phillips designed capstan still exists on HMS Unicorn (1824), and it is also slightly different from the contemporary models and plans.  

A merchant vessel version of Phillips' capstan, dated to 1837. This
configuration was very similar to that used on HMS Terror

I started my reconstruction with an assumption that the basic profile, shape, and features of the capstan were accurately depicted on the 1839 profile plans (their consistency with the Investigator plans, as well as their overall accuracy, provides good support for this). Using the 1839 drawings as a base, I determined to add details from sources dated as closely as possible to 1839.

A primary source of information comes from the 1854 Encyclopaedia Britannica (5), which contains a detailed plan of a two-capstan (upper and lower deck) version of Phillips’ 1827 patent (No. 5505), specifically for Royal Navy ships. The plan differs little from Phillips’ original 1827 patent, though it shows a gear mechanism placed below the pall rim and a less robust gear plate, both common traits in later designs. Fortunately, an 1837 plan of a single capstan for a merchant vessel (recall that Terror and Erebus were based on merchant designs) exists (6), and I drew crucial information from this plan. It accords very well with the 1854 and 1827 plans, although the gear mechanism and other details would not fit the plank and beam arrangement shown on the 1839 Terror and Erebus plans. The notes accompanying the plan indicate that Royal Navy models were slightly different than this plan, and hence I believe the 1854 plans, which are consistent with the planking arrangement on Terror, are likely to be most representative of the overall gear design.

Further valuable information about the gear design and gear plate design, as well as information on the pawl configuration and rim, comes from contemporary models. A beautiful 1827 model single capstan exists in the Science Museum collection. Two additional models of the 1819 versions of Phillips’ capstan can be found in the Rijksmuseum in Amsterdam.

The capstan on HMS Unicorn has been heavily modified (the upper even has table lamps built into it), but it provided valuable construction information, especially on the appearance and construction of the barrel and whelps on the upper capstan, and the gears and gear plates on the lower capstan. Believing that additional robustness would have been necessary, I added full thickness chocks on the upper part of the capstan to my plans and model, matching those on HMS Unicorn’s upper capstan. This is the only detail of my plan not consistent with the 1839 Erebus and Terror sheets.

An 1839 era Phillips capstan, as I believe it may have been 
configured for use on HMS Erebus and Terror.

Constructing the model of Terror’s capstan began by transforming my technical drawings into construction plans. Given how intricate the resulting construction plans were, I decided to begin construction by using a laser cutter at my local library.

Cutting the capstan components on an Epilogue laser cutter. 

The completed pieces. 

Vellum was added to enhance the joints of the capstan.  

The assembled capstan before sanding. 

Sanded to shape. 

Drilling the bolt locations. 

The completed drumhead. Some Phillips' capstan models show lined sockets, so I added boxwood liners.
I admit that it was primarily an aesthetic choice.

I cut the drumhead plate from an unused pipe fitting which I flared to the right size.  

I filed a lip into the plate by hand. 

I cut the pawl rim (ring) out of brass plate using a jeweler's coping saw.

I filed each stop by hand, after carefully scoring the brass. 

The pawl rims were each made from pipe fittings flared to the precise diameter. 

The pawl rim prior to soldering and sanding. 

The piece following soldering. I used  copper solder for the first time on this piece - despite being very dirty, it worked well. 

To maintain the proper curvature, I cut the pawls from a copper fitting. 

The pawls cut roughly to length. 

Bolt holes were drilled before shaping. 

Each pawl was filed and shaped by hand. 

The completed pawl rim (ring) and pawls. Two traits unique to Phillips' capstans can
be seen here. First, the pawl rim was bolted through each stop, rather
than in the spaces. Second, Phillips' capstans had between six and
eight pawls, while earlier models typically had four. 
The pawlhead. 

The completed pawlhead with the top plate soldered in place. Contemporary models show that the pawlheads on
Phillips' capstans were made entirely of iron. 

Dry fitting the metal pieces. The pawls need some thinning here. 

Pieces prior to finishing and assembly. 

Blackening the metal parts. 

The metal pieces after blackening, buffing, and sealing
(I use Krylon matte coat as a sealer). 

The completed capstan. The wooden pieces have been treated with Minwax wipe-on poly. 

Detail of the pawl rim and pawls. 

The drumhead (the drop pins indicate it is in direct-drive mode). 

Detail of the lower drop pins (I couldn't find scale chain small enough to model that feature). 

Mini-Crozier inspects his capstan, recalling his good times with  Parry. 

Of course, one further Phillips' patent capstan is known to exist – on the wreck of HMS Erebus. Parks Canada’s recent video tours show that it is heavily damaged and that it is very obscured by the growth of marine life. How it compares to my plans and model is impossible to tell from the video, but hopefully my model isn't too inaccurate. If the historic plans are any indication, any differences should be quite minor.

1 Harland, John. 1999 Improved Capstans. Nautical Research Journal 44(4): 214-220.

2: Parry, Sir William Edward. 1826. Journal of a Third Voyage for the Discovery of a Northwest Passage- from the Atlantic to the Pacific, Performed in the Years 1824-25, in His Majesty's Ships Hecla and Fury, Under the Orders of Captain William Edward Parry, R.N., F.R.S., and Commander of the Expedition. John Murray, publisher to the Admiralty, and Board of Longitude. Page: 14.

3. (ibid, Page14)

4.  The Mirror of Literary, Amusement, and Instruction. 1825.  Page 451.

5. Encyclopaedia Britannica. 1854. Capstan. Pages 128-129, Plate CXLVI.

6. American Magazine of Useful and Entertaining Knowledge, Volume 3.Nathaniel Hawthorne, Elizabeth Manning Hawthorne, Boston Berwick Company, 1837. Page 203.

Tuesday 8 September 2015


I build very slowly. 

This has a lot to do with the dual nature of my HMS Terror project, which aims to create both an accurate model and a complete set of 1845 era plans. The latter has consumed hundreds of hours of my spare time and has led to some interesting side projects. It’s necessary and rewarding work, but it keeps me out of my workshop.  

Another reason I build so slowly is that I enjoy the process of building; figuring out the steps and making the templates and jigs to build accurately. My most recent sub-project was constructing a “building board,” essential for accurately aligning the bulkheads with the false keel of my model. The board needs to keep the false keel perfectly straight and perpendicular, while providing a flat surface for gluing the bulkheads at right angles to the keel.

I started the board by cutting a 1/4“ groove into a 32" long piece of MDF sheet. The groove is just slightly deeper than the false keel on the model, providing a full view of the rabbet, and fits tightly so that the keel remains perfectly straight.

Marking the groove on the MDF sheet. The board is 12" x 32 ". 

I cut the groove into the board with a table saw. MDF is brittle, so care is needed at this stage. 

The stern and stem of the model also needed to be kept perpendicular to the board. To achieve this, I designed two supports, one for the stern and one for the stem, which would slot into the building board groove.  I used a laser cutter to make these from 1/4” MDF sheet; they fit together somewhat like a piece of Ikea furniture.

My custom plans for the stem and stern supports. The idea for the interlocking pieces came from children's toys.  

Cutting the pieces from 1/4 " MDF. My local library provides a 60 watt laser cutter for public use. 

The pieces after cutting. 

The finished supports (stem on left, stern on right). 

It's hard to see, but the support has a rail designed to fit into the groove for proper alignment. 

The supports were glued into place, and their alignment was double-checked with a square. 

I lined the groove with painter's masking tape to prevent damage to the model's keel. 

I lined the supports with masking tape as well. 

The finished building board. 

Instead of using set screws, which might damage the model, I used rubber bands to secure the model in place. In practice
the fit was tight enough that rubber bands weren't really necessary. 

The bulkheads fit tightly and relatively squarely, but it was necessary to use a carpenter's square to
ensure that they were at perfect right angles to the keel. These stainless steel squares from Lee Valley
were perfect for the task.  

The second-to-last last bulkhead is put in place. 

The completed superstructure. It's hard to see here, but two of the forward stations have
alignment issues at the area of the chock channels and bulwarks and will need to be sanded and/or modified with
 trim.  I double checked the plans, and the errors, around 1.5 mm off, seem to have crept in from my
original measurements. These errors weren't large enough to force me to re-cut the stations. 

I'm quite happy with the alignment (the stern looks slightly twisted here, but it's just a trick caused
by the camera angle). 

Midships bulkheads at the position of the rabbet. 

A view of the bottom of the ship. Again, I'm quite pleased with the alignment
(the stem and stern are still little wobbly and will need some filling blocks). 

Mini-Crozer stands on his deck. It's been a long time.  

Mini-Crozier inspects the stern, noting that work is needed on the bearding line.
Though Terror was small by Royal Navy standards, a 102 foot ship was still a
substantial vessel, as this image indicates.  

Mini-Crozier stands at the Captain's sacred spot on the quarterdeck.  

Now that the bulkhead of the model is finally assembled, I can move towards planking it. A significant amount of fairing (sanding to ensure the fore and aft edges of the bulkheads match the ships lines) is necessary, and I hope to finish that shortly. Meanwhile, I've almost completed the "Phillips' Patent Capstan", and I hope to reveal my plans and model here soon.