machines which I have made, am making, or intend to make, and some other stuff. If you find this site interesting, please leave a comment. I read every comment and respond to most. n.b. There is a list of my first 800 posts in my post of 17 June 2021, titled "800 Posts"
This is the lathe which was used to turn the 18.1″ guns of the IJN Yamato. It was purchased by Japan in 1937 from Germany. Japan did not have the capability to make such a huge lathe. The only countries which did have that capability were Britain, USA, and Germany, and Japan did not want the British or Americans to know how big a battleship was being built. The weight of the Yamato (>70,000 tons) was not known to the allies until after WW2! It was the heaviest battleship ever constructed, and carried the biggest guns ever installed on a battleship.
I assembled a plastic model of the Yamato earlier this year.
….and showed my model engineering club.
Somehow the conversation involved the guns, the Yamato Museum, and the huge lathe which is being moved from its original factory, where it was in use until very recently, to the Museum.
A senior member of our group, who lives near Yallourn, Victoria, stated that the lathe which made the guns for the Bismarck was currently located in Yallourn, Victoria, Australia!!!
There was a collective dropping of members jaws. WHAT???? HOW??? It cannot be true!!
To cut the story short, further investigation by the member, confirmed that a very large Wagner lathe had been shipped to Australia, as reparations after WW2. The lathe had been installed in the State Electricity Commission workshop to turn large generating armatures. The SEC workshop was now a private manufacturing factory, and the lathe was still there!!!
The managing director of the company was contacted, and a few weeks later 24 of our club members visited the factory with the tour being conducted by the owner-manager. The factory was fascinating, and the tour lasted over 2 hours, but I shall concentrate on the lathe.
The distance between centres for the lathe is 70’/20m, certainly big enough to have turned the Bismarck guns. The lathe OAL is 90′. The original tailstock and a section of bed was lost at sea in a storm during transportation to Oz. Replacements were found and installed.
Currently the lathe has a 50hp electric motor. It turns between 1 and 20rpm.
So. That was a great day.
The factory owner manager is searching for documentation about the lathe, which is essential if the Bismarck association (and Tirpitz, and more than a few shore defense guns around France, Denmark, Norway) provenance is to be firmly established. The locals certainly believe that it IS the lathe which made the Bismarck guns. It is possibly true. The Yallourn lathe looks very similar to the Yamato lathe.
In the above photograph, taken I think of a 110pr Armstrong breech loader in Canada, of a Garrison mounted gun, there are several very interesting features. The Smith’s elevating screw for instance, and the remnants of the left hand breech tangent sight. But I am particularly looking at the flat surfaced item which is attached to the top of the breech. It took me some time to work out the function of the rather complex shaped item.
The breech block, which weighed 130lbs, had to be lifted out of the breech by two strong gunners to permit swabbing of the bore from the breech aperture (also visible in the photo), then loading of the next projectile and gunpowder bag, after which the breech block was lifted back into position and screwed tightly closed prior to the next firing.
Ah….. the flat topped attachment is where the breech block was placed while the swabbing and loading took place!
So I set about making the breech block rest (as I called it) for my model.
The rest looked complex and difficult to model. The inner surface had to fit the external surface of the breech, including two convex fillets. The external surface has to fit the breech block, without denting or otherwise damaging it despite its considerable weight and frequent manhandling. And there are holes for 6 attaching screws.
Now, dear readers, I must inform you that I have only enough WordPress memory for another one or two posts and a few photographs.
The Armstrong 110pr breech loader cannon model project is almost finished.
The remaining parts, including the Smith’s elevating screw, carriage wheels, rope eye bolts and capstain were all described in the build of the Armstrong 80pr rifled muzzle loading model cannon, so I will not repeat those details for the 110 pr.
I will leave the remaining small memory for the assembled model of the 110 breech loader, in a few weeks time.
And since I will not delete any more old posts, that will be my final post. (unless WordPress changes their policy of not increasing memory limits. And I do not expect that to happen.)
In the final post I will notify you, my readers, of the site where I will post photos of future projects. Not quite yet decided, but it will NOT be WordPress.
My workshop uses a 6kw Phase Changer machine to convert the 2 phase supply to 3 phases, which is required by my CNC mill, DRO mill and big lathe.
When I turned it on 2 days ago, I was startled by a very loud “bang” and a puff of black smoke from the phase changer. And no power output, unsurprisingly. On opening up the machine I was pleased to note that the (very expensive) transformer looked intact. But one of the large capacitors looked a bit ragged.
So I have been limited to single phase equipment for the past 2 days. It did force me to finish some outstanding tasks….
and yesterday my colleague and friend Stuart T used his 30w fibre laser to engrave the manufacturer name, number, barrel weight, date, and Queen Victoria’s cypher.
30mm diameter, 11mm thick, a rail groove on the edge, circular divots on the faces. Simple!
I decided to make them from stainless steel. And tonight I have multiple small cuts on my fingers to prove it.
Stainless steel is a bugger to machine. It requires slow feeds, deepish cuts, COOLANT, and sharp tools. Carbide is OK for roughing, but for accurate final surfaces, sharp high speed steel is required. And it produces razor wire. Copious amounts of it.
First I machined some 40mm stainless rod down to 31mm. Too late I realised that was too big.
Then I used a HSS form tool bit to cut the edge groove. But got too much chatter. So spent some time getting the coolant pump and nozzle working. Some improvement, but still some chatter. So I switched to a HSS parting tool 3mm wide, and that seemed to work well. The DRO was handy to achieve the final groove depth of 2.5mm, and 6mm wide. And then to take 0.5mm off each face to produce a boss 12.7mm wide and 0.5mm deep.
Then completed the parting off. Oh. Forgot to mention the 5mm shaft hole which was drilled.
But when I tried to install the wheels in the wheel brackets I realised that I need to remove about 0.1mm from each boss. This is the setup which I used.
So, the first pic is the current situation, . Next steps are to make the washers for the axles, trim the axles to length, and drill/install retaining pins. These steps always seem to require at least double the predicted workshop time.
The problem was that the 4 wheel brackets needed the wheel recess deepened by 2.5mm, and the brass shape had few clampable surfaces.
So, I tried option 1.
I machined a wooden jig to hold the bracket in the milling machine vice. The wood is Australian desert ironwood, which is unbelievably hard, but would not mark the brass. The pocket was machined to the diameter of the circular base of the bracket, and then sawn in half.
The bracket was clamped in the jig and squeezed tightly in the vice. Then machined with the slot cutter, the required 2.5mm deeper. The workpiece showed NO tendency to move.
Previously I have made model wheel brackets using 2 different methods…. 1. casting 2.turning/milling
So, these are the wheel brackets which I have made for the current model Armstrong 110pr…
I think that when these are finished they will look closer to the original than either of the previous examples, and they certainly look more robust and fit for purpose IMO. So, what do you think?
The above wheel brackets were milled from 38mm brass rod….
On reviewing this post I noticed that the slot for the wheel looked a bit shallow, and when I measured it I found that it is 2.5mm too shallow. A simple mistake, but must be fixed.
The problem is how to hold the workpiece while cutting the slot the extra 2.5mm deeper.
Possibilities. 1. make a circular jig to clamp the bracket in the milling vice. 2. just hold the bracket in the milling vice and hope for the best. 3.solder a 38mm cylinder to the top of the bracket, and hold the extension in the vice. 4. make new brackets.
At this moment I am thinking that I will try 1. and if unsuccessful move to 4.
The Armstrong 110pr model cannon trunnions sit in semicircular cutouts in the carriage cheeks. In the model the cutouts are 20mm diameter, and they are slightly deeper than semicircular. Making the “slightly bigger than semicircular” cutouts is complicated by the fact that the cheeks toe in towards the front, by 2.65º.
When I originally cut out the cheeks I made the cutouts 18mm diameter, allowing 2mm to be removed at the assembly time, and to then remove some extra material to cope with the toe-in angle. I did not know in advance how that would be done, but I figured that I would use a drill or reamer at the correct angle to remove the extra material.
Today was the day.
But when I was actually confronted with the task, I realised how difficult the job was actually going to be. I also realised that a drill or reamer was NOT going to do the job accurately or neatly enough.
Then I had a brainwave. And I am really proud of this one. I made a round file, exactly 20mm diameter, and long enough do file the cut-outs together, exactly in line.
How to make a file?
And how to make the teeth small enough so they leave a smooth finish with no edge tear-outs?
Ah! But I forgot something. When I measured the diameter of the “file” the 20mm shaft now measured 20.25mm. I had forgotten that knurling INCREASES the effective diameter. So I turned off the knurls and machined the shaft down to 19.75mm, and repeated the knurling. The diameter was now 20.05mm which I considered acceptable.
Since I was only intending to file wood, I did not bother hardening the silver steel.
Garrison based cannons did not use trunnion caps, unlike the naval versions, relying on the weight of the barrel to keep it in place. The centre of the trunnion is just below the top surface of the carriage cheek.
The file worked well in hardwood. I would have hardened the steel if it was to be used on brass or other metal.
The carriage cheeks for the Armstrong 110pr cannon have 10 roughly vertical bolts which attach the wooden slides, and also bolt together the thick planks which make up the cheeks.
The nuts for the bolts are all at the bottom end, and are buried in the slides. In previous model cannons I have milled pockets for the nuts, and tightened the nuts with a socket spanner, but I was not happy with the large diameter of the pocket which was needed to accomodate the socket spanner.
So, this time I decided to tighten the nuts using a screw driver, having cut a slot in the surface of the nuts.
With the barrel almost finished (except for sights and engraving), I have returned to woodworking. The carriage was made of wood in the 1860’s.
I had previously cut out the carriage sides and the slide blocks, but now the parts need to be bolted together. Today I marked out the bolt holes, and drilled some. The holes were 4mm diameter, and up to 90mm long. Definitely “deep drilling”, despite being in wood. Due to the figuring in the wood it can sometimes be difficult to keep long series drill bits from wandering off to the sides.
Next job is to continue the bolt holes through the slide blocks. And to make the transom. And then to enlarge the trunnion cutout to the correct size and angle.
I am glad that I had no visitors to my workshop in the past 2 days.
The language in the workshop has been a touch foul.
Because I have been making handles for the previously made bronze breech blocks.
The handles are very small, very exacting, and difficult. In a word, I struggled.
End result photo….
I was not enthusiastic about this job. I had a feeling that it would be a bugger. And so it was.
So why all of the bad language?
Well, I needed 2 of these, so 4 handles. I made 2 spare.
I dropped one. Could not find it, despite hours of searching, including using a fibre optic 5mm diameter device to look under the milling machine and sweeping the floor. (no snakes in this cold weather. I hope). But no luck, so I made another.
Another handle jammed in the Dremel drill. I hear it hit the tin wall 7 meters away. I did look for a minute or 2, then succumbed to common sense and made another. The language really was foul.
Anyway. You have seen the final result. not too bad. Another bit of brass bling.
The original trunnion rings of the Armstrong 110pr breech load cannons were made with a smaller internal diameter than the barrel. Then the ring was heated, installed, and it shrunk firmly into its permanent position. Heat shrinking multiple coil cylinders to build up the cannon was shown to be a very strong method, albeit expensive.
I decided to try the same method with my 1:10 model, and discussed the method at the recent GSMEE meeting. I listened carefully to the advice from members, read Machinery’s Handbook on the subject, and was ready to proceed.
The ring internal diameter was turned to 0.05-0.06 mm smaller than the diameter of the barrel where it would be positioned.
I let the ring soak up the heat for an hour or more.
Lifted it out with pliers, carefully lowered it down the barrel (having earlier has a couple of practice runs), and felt it slide easily into position.
Another ring followed later and it also dropped easily into position. I had to rotate it, and noted that it locked up after only 10-15 seconds, so the working time is indeed very brief!
I will clean up the blackened heat affected trunnion ring later. This was a very satisfying day in the workshop.
Unfortunately the laser cut trunnion ring blanks were unusable because they were undersize. Rather than wait for another run of laser cut parts, with 3-5% increase in size to cope with the problem, I decided to mill the shapes from some old 27.5mm thick mild steel. In my last post I showed the preparation of the stock.
I made 2 of the ring blanks today. They have a 45mm hole, and to speed up the milling process I chose to use my magnet drill and a 35mm annular cutter to get the hole started…
As explained in the previous post, it was crucial that the breech screw was tightened securely.
Watch the following YouTube video of a demonstration firing of a 110pr at Fort Nelson, video’d by Nick Cafferata, and used here with his permission. Note how the weighted handle was swung by 2 gunners and repeatedly knocked to ensure secure closure. Also note the volume of smoke from the firing, and this was a charge of only 2lb, compared with the 10-11 lb used in 1861.
The central 18mm diameter shaft was also heated to break its Loctite bond and the shaft came free.
Next job is to make and attach the blocks to the handle which knock the octagon/breech screw, and the heavy weights to the ends of the handle which enhance the momentum of the action.
Watch this space to see it being CNC’d into a trunnion ring…….
If the screw which held the breech block in place was not tightened, when the gun was fired, explosive corrosive gases would escape backwards rather than propelling the projectile. An inefficient and destructive result.
If the screw was not not tightened at all, the breech block, which weighed 130lb, could be ejected with great force, and devastating, potentially fatal results to the gun crew.
So it was important that a gas tight seal was achieved when the block was inserted and tightened. That required a seat like a valve seat in an internal combustion engine, and a corresponding 45º angle on the breech block.
First I made the breech block. The plug was turned from LG2 bronze. This will seat against the steel barrel bore. In the original the block was made of steel or iron, and it seated against a copper insert seat. I decided that it would be too fiddly and difficult to reproduce the original steel/copper system, so I substituted the bronze block which fitted against the steel end of bore in a 45º seat.
I used 2 tools to make the seat. A commercial carbide seat cutter, and a shaped stone to finish.
The seat was cut with the carbide cutter, by hand and using cutting fluid. When it was 1-2mm wide, some chatter marks were just visible, so they were polished out using the stone, also by hand.
And today I picked up some laser cut parts from the cutter. (JR Laser, Geelong)
A few photos of painting the model Yamato. I used Tamiya paints. Spray cans for the large areas- and hand brushes for the small ones.
Then removed the masking tape and applied some more tape over the red. Painted the top half of the hull, and the other modules, “battleship grey”. Then glued the modules together.
The end result….
The model is complete, except for the flags and aerial cables. Took me a week to make and paint. I really enjoyed the build. And I really like the model. It was not an easy build, but the real credit goes to the people who designed and made the kit. It is truly impressive how well everything fitted together.
Now. Where to put it? And how to keep it dust free?
The 1:350 Yamato model is made of plastic. Mostly Polystyrene, but also a small amount of ABS. Different glues required for each type of plastic. Both types will hold the parts in a minute or so, but several hours are required for rigid holding.
There are 17 different colours specified, which explains why the paints were so (unexpectedly) costly. Mostly IJN grey, and dull red, for the hull exterior, and wooden deck tan. I bought Tamiya spray cans for the dull red, IJN grey, and primer. The wooden deck tan was unavailable, so I bought some laser cut sheets of impossibly thin wood, already in the correct colour, and made for this particular model. Pictures later.
So I sprayed the primer coat.
The question was whether to make the entire model, then paint; or paint the individual parts on the sprue frames before assembly ; or something in between.
I thought that painting the entire model would be simplest, but some small parts and areas would be inaccessible, and the result would be messy.
Painting every component on the sprues would leave a bare cut area on every part which would need to be touched up later, so that did not appeal. Plus it would be very time consuming.
So I decided to make the ship in modules, and paint each module separately.
Tomorrow I hope to start applying the final colours.
Painting is really NOT my thing. So to finish the day I spent some time restoring an old small Westcott adjusting wrench which had been given to me by a friend.
So far, glueing up the model has been interesting and a lot of fun. Look at the progress after 2 days….
The tools which I have found useful are lined up.
Alligator clip on a chop stick, rubber bands, Extra Thin Tamiya Glue for polystyrene plastic, flat non serrated small pliers, needle nose small pliers, safety razor blade, sharp side cutters which I have modified so the cutters are thin and very pointy, steel ruler used as a scraper, small fine file, fine sand paper, fine tweezers (actually from my microsurgery kit of 30-40 years ago), coarse strong tweezers, and utility knife. And of course an A2 cutting board, and Tamiya Instruction book which I have found to be accurate and very helpful.
The Extra Thin Tamiya Glue is very good. It sets in a couple of minutes so parts can be finger held in position. It is so thin that it tracks into small cracks by capillary action. And it is transparent. Time will tell how paint adheres to the glue.
The Tamiya parts are also very impressive. Beautiful smooth finish, minimal flashing which can be scraped off with a finger nail. And the parts fit together very accurately, for the most part. Rarely I had to use the razor blade to make parts fit together, and that was usually because I had missed a bit of the sprue when separating the parts from the sprue.
It was quite exciting to see the hull coming together.
I am close to painting the components. I will use Tamiya spray cans, brush applied paints for tiny parts and fine lines, and possibly an air brush. I have been watching YouTube videos to pick up hints on the painting process. It was surprising to me just how many YT videos exist on the subject of painting model Yamatos.
Then the major components are glued together.
Then the smaller guns and other surface equipment will be glued on to the painted surfaces.
P.S. Another 1/2 day gluing up these tiny planes. One more to go.
In common with many other males, (whoops. Possibly females as well, although I know of none.), I have long had a fascination with battleships. Of all eras from the ancient Greeks and Romans, Nelson’s, dreadnoughts, WW1 and WW2. Read the novels, made models from kits and from scratch. I have quite a library of books.
Recently, I purchased this book…
336 pages. 350 colour views, including some original photographs, and lots of details. 1020 scale drawings of excellent quality. 43 pages of history and specifications. The bulk of the book is superb quality pictures and drawings.
These battleships, at 72,000 tons, were the largest ever constructed. And they mounted the biggest guns ever used on a battleship at 18.1″. Each of the 3 gun turrets weighed as much as a heavy destroyer, 2500 tons. They were 250 meters long, and 50 meters from keel to the top of the superstructure. Their 4 turbine engines drove the ships at 30knots/50kph. Each ship had 25,000 tons of armour, up to 560mm thick!!
“Awesome”, seems insufficient.
The Imperial Japanese Navy had them built to outgun the most powerful battleships of the US Navy and western powers. However they were dinosaurs, and both were sunk by aircraft. Neither fulfilled their intended role of fighting other battleships.
The book is divided into 4 sections….
Section 1: Introduction, Superbattleships and Summary of Service. 43pp.
Section 2: Primary Views. 25pp.
Section 3: The Drawings. Subdivded into general arrangements, Hull structure, Superstructure, Rig, Armaments, Fire Control, Fittings, Aircraft, Boats, Author’s Model. 252pp
Section 4: Yamato and Musashi at sea, Remains of Yamato and Musashi 12pp. The pictures “at sea” are computer constructions, using the author’s model, and incredibly convincing. Initially I took the pictures to be of the originals.
Both ships were sunk by massive US air power, with the loss in Yamato’s case of 90% of its crew of 3,300 sailors. Almost as sad, almost all of the original construction plans and details were destroyed by the IJN after the Japanese surrender.
So, if you have any interest in battleships, massive marine engineering, WW2 naval history, or ship modelling, this book is an absolute must.
After reading the text, and going through the pictures multiple times, and being captivated by the wonderful lines of the ships, I decided to make a model of Yamato. Kits vary from 1:1000, to 1:100, with the larger scales being in the thousands of dollars.
The following is the only surviving plastic model of mine. Another ship with wonderful lines.
A question to my readers….. would the progress of making the model Yamato be of any interest?
For reasons which I will not detail here, I am spending more time at home, and much less in my workshop. Work on the Armstrong 110 pr breech loader is progressing, slowly. However, the rifling is complete.
I detailed the rifling setup in a previous model build, but in case you missed it……
The setup took several sessions to complete. I had previously drilled and D bit finished the bore, and drilled and cut a large thread to accept the breech screw. Then I turned the exterior of the barrel so it would fit the jig. It will be turned to its final shape in a future session.
I could not find actual specs for the twist, so I randomly decided on 90º. The cut started in the powder chamber and finished just beyond the muzzle. The rifling in the original started distal to the projectile chamber, but I had to ignore that due to limitations of my setup in accessing the adjusting screw. The powder chamber and projectile chamber were slightly bigger than the bore in the original, so I might be able to machine away the unwanted rifling in those areas in my model.
So. Having made the decision to make a model rifled breech loader, Armstrong gun, on a wooden sliding carriage and wooden traversing platform, I gathered my references. A lot of these guns were made, 959 in use in 1878. Many on wooden carriages, some on iron carriages. They were used in several wars, and I will be delving into the history. Examples of the guns exist in quite a few countries including UK, USA, Canada, and Australia. There are references in Wikipedia, and several artillery books of the era (1860-1890). Various models have been made and documented, including good descriptions, particularly by jefenry.com.
I have several reasonable scale drawings, including some kindly sent by jefenry. (Thanks again Jeff!)
In the drawing above, the traversing platform is identical to the ones under the 80pr Armstrong RML’s which I recently modelled, so my previous experience will be useful for the current build. The carriage for the 110pr RBL is similar, but not identical. The barrel itself will be quite different, and will be the main challenge in the current build. Apart from the breech block, and breech seal, there are 76 (!) rifling grooves, compared to 3 rifling grooves in the RML. I am already thinking that I will be reducing the number of grooves, to maybe 28.
Another handy resource which I found during my Internet searches of Armstrong 110pr’s, is ETSY.com, a Canadian site, where the Armstrong 110pr has been CAD drawn in very fine detail, and available for $AUD34. The drawings are not perfect in every detail, but even so I rate them as very good. Only available as Fusion 360 files, but Fusion 360 is available free of charge for hobbyists, with some restrictions relating to file numbers and some features.
Yesterday I purchased a lump of 1020 shaft, 1270mm long. I only required 305mm, but the supplier was unable to cut it for 3 days, so I took the whole piece. A burly worker picked it up as if it was made of balsa wood, and put it in my car. I struggled to unload it at the other end. 40+kg/ 90lb.
…and weighed the 306mm piece…
The next step for the barrel is to rough drill the bore. I have an extended 16mm drill bit from the previous model, but will have to modify a 17.75mm bit and extend an 18mm reamer or make a long 18mm D bit, before I can proceed. So instead, today, I made a start on the traversing platform.
Having made a few errors in the machining sequences last time, hopefully I can avoid the mistakes this time. Also, with multiples of some components, such as wheel brackets, and rope rings, I will be casting some of these in bronze, and getting laser cut parts for others such as the metal slides.
The original Armstrong barrels were constructed in multiple pieces which were shrunk together, using the “coil” method to construct the pieces. The trunnions were on a separate ring which was forged, then machined to final shape, then shrunk into position. I am considering machining the model trunnion ring separately, and shrinking it into position, but the rest of the model barrel will be turned from a solid piece of 1020 steel.
I had thought that the 1:10 scale model Armstrong 80pr rifled muzzle loader would be the last cannon which I would make. It is currently being given finishing coatings to the woodwork. Later this year it will be given as a gift to a family member.
To be honest, having made five 1:10 scale model blackpowder cannons, I am ready to move back to my first modelling passion, which is steam engines. I had no real interest in weapons or guns or artillery, except as a means of increasing my understanding of history, specifically military history. I have no interest in firing guns, although I must admit to an illicit satisfaction in watching You Tube videos from USA of cannon modellers who can actually fire their creations.
My interest in cannons started when, as a newbie in CNC machining, and looking around for a project to use my newly acquired CNC lathe in 2015, I made a model long gun.
And the most recent Rifled Muzzle loader, the same 80pr Armstrong Barrel, on a Dwarf carriage, and wooden traversing platform.
I truly thought that this would be the final cannon which I would model. So I could get back to my model steam engines.
Like this one from 2-3 years ago, now gracing our kitchen, with decorations by SWMBO.
BUT….then my eldest daughter, who has absolutely NO interest in cannons, asked ” are you going to make a cannon for me?” I must point out that this daughter rescues injured animals and takes them to her vet, is vegan, the most pacifistic and socially conscious person that I know. I questioned why she would want a model cannon. “I just do” she replied.
Oh well. I guess that I will be making one final model cannon.
I spent a day searching my books, Google Images, Wikipedia for a cannon which would look interesting as a model, be interesting for me to make, and for which some plans or drawings are available. I offered my daughter the choice of my existing models, but no, she wanted one built just for her.
Then I thought of jefenry, my reader from the USA, who has made several model cannons, including one which intrigued me when I first saw his pictures and videos several years ago. It is a 1:9 scale Armstrong rifled breech loader, 110pr, of 1861. One of the first breech loaders of relatively modern times. (Breech loading cannons have been around since medieval times, but they were less reliable than muzzle loaders, more inclined to explode and kill their own gunners.). The Armstrong 110 pr RBL saw action in several wars, including against Japan, the NZ Maoris. It was the largest cannon on HMS Warrior, but was replaced by the more reliable muzzle loaders.
So that is what I will model for my daughter. An Armstrong 110pr, rifled breech loader, on a dwarf carriage and wooden traversing carriage. Here are some pictures.
So, my plan is to make a 1:10 model of the barrel, on a carriage and traversing platform like the Fort Henry example above. Not sure how much of the build will be featured on this blog. I am again very close to my WordPress.com memory limit.
Today I milled the rebates which the wheel brackets fit into. Only 1mm deep and at an angle of 15º to the base line. It went fairly well, but when I reversed the milling pattern for the reverse sides, It went a bit askew by about 0.5mm. Not much, but enough to be noticeable, so I filled the defect with wood putty.
Then I milled the 3º chamfer in the wheel brackets. Straight forward process.
Finally, with the brackets sitting correctly in their rebates I wondered how to make the bracket retaining bolts, and the wheel axle shaft.
I prefer to use stainless or brass bolts, but none come with dome heads, so I considered various options. I chose to use a method which I have used previously.
I selected some 3mm stainless cap screws, and filled the head with 50% silver solder.
…And used a rounding over milling bit, held in the toolpost, to round over the capscrew head and its silver solder filling. The first screw bent during the form turning, so I placed them deeper in the ER collet chuck. A later one broke, so I slowed my feed rate. I ended up with 4 bolts.
I did the same with some bolts for the axles, but they are fully threaded, so this will be a temporary solution until I can make more suitable axles.
You will notice the filled hole in the carriage cheek. That was a mistake, but rather than start the cheeks from scratch again, I chose to fill the holes. They will be almost invisible when the cheeks are finished, I hope.
The Armstrong 80pr cannon on the dwarf carriage and wooden traversing platform, slides wood on metal slides. But, when the carriage and its heavy barrel (4+ tons) are returned to the firing position, there are two small bronze wheels to make the return easier.
Two strong gunners lever the rear of the carriage and barrel slightly, so the two small wheels at the front of the carriage take some of the weight, and the carriage runs forward. In fact, the return was a bit uncontrolled, so a rope was added to the rear of the carriage, thrown around the bollard at the rear of the slide, and a third gunner added some control to the return.
Today I made the 1:10 scale wheels. They are 20mm diameter, and 10mm wide. 13mm wide if the hubs are included. I spent a couple of hours with the design. And another couple experimenting with various CNC processes. Not many photos of all of this I am afraid. I learned some new V Carve Pro commands, including nesting commands using the same milling cutter, but there was some trial and error. The first two wheels took a couple of hours. The final two took only 30″.
NB. these parts are not finished. Sharp edges remain. I will probably put them in the gemstone tumbler to smooth the edges.
Today I CNC milled the cams. And silver soldered them to the bearings.
PS. A few days later. In a fit of perfectionistic idiocy I removed the bronze cams, and replaced them with steel ones. The originals were iron. The pins which pushed on the cams were also steel. That took about 3 hours, but now I can sleep easy.
And by the way, the compressor was working perfectly when finished. But a few days later, with a change in the weather, it is not applying enough pressure to the slides. That is the problem with articles made from wood….. they expand in humid weather, and shrink in dry weather. Dimensions changes of 3% are common, across the grain. It was probably one reason the wooden compressors were abandoned in favour of Elsworth iron compressors, and hydraulic mechanisms.
A very pleasant drive to Warrnambool yesterday, and re-inspection of the very rare compressor which was the recoil arrestor for the LowMoor 68pr cannon. And probably for all guns on the same carriage and platform, including the Armstrong 80pr RML’s at Elsternwick, Queenscliff, etc which I am currently modelling.
I wanted to closely examine the iron riveted pieces closely to check my theory that the short straight sections are the parts which acted as the cams to close the gap between wooden leaves and release the friction from the braking action. Unfortunately the rust concealed any such evidence. But I still believe that was the purpose of these iron pieces.
So, today, I commenced making a 1:10 scale model of the compressor to fit to my miniature cannon.
At 1:10 scale the bronze bearings would be less than 1mm thick. How to make them?
I milled the steel elliptical post from silver steel. Yes, CNC’d.
Another workshop session require to make the iron cams and the handle with pins.
After milling the rebates in the wood, I attached the bracket with the brass screws, and sanded them flat with the surfaces. Most of the strain will be on the steel screws. The brass screws are screwed and Loctited into place.
Then drilled and tapped the wood for the BA10 stainless steel bolts. It is fairly close to the original fastening method.
It took 4+ hours.
A short post. Tomorrow I am visiting the Flagstaff Hill Museum at Warrnambool, 2.5 hrs each way, to get some final details about the wooden recoil brake, the “compressor”. There is a problem with my CAD drawing of the compressor, and I am hoping that close inspection and measurements will answer my query. I will be accompanied by my expert friend Stuart for some extra perspective. The compressor will be the final substantial component to make for this model.
This is one of the few parts required to finish the model Armstrong 80pr RML cannon on a wooden carriage and traversing platform.
The issue in silver soldering the pieces together was that they are quite small, about the size of my little fingernail, joined at an 95º angle, with the ring also soldered in place in the same heating session. And I did not want solder getting into those 1.6mm diameter holes.
So I screwed the angle pieces to a block of hardwood which had a 95º angle, having fluxed the edges carefully to exclude the flux from the tiny holes. I would have added typists white-out if I could have found it.
I knew that the wood would catch on fire with the soldering torch, but hoped that it would retain its basic shape until the solder solidified. The steel on top was to hold the ring in position during soldering. If the method did not work I figured that I could make an aluminium shape to replace the wood.
After soldering, I put out the fire by dunking the assembly in a bucket of water.
And it cleaned up quite well. Now to carve rebates in the transom so the bracket sits flush with the wood surfaces.
The circular cutout is to allow the end of the Smith’s Screw to protrude under the transom.
Not much to show for several hours in the workshop, but it’s better than working. And best of all the method was successful.
Another hot summer day today, so I arrived at my workshop early, before the heat set in.
First I drilled a 1.5mm hole through the Smith’s screw yoke and bracket, for the pin which completes the hinge mechanism which engages and disengages the screw handle. Sounds simple? Well, actually, my intention was insert a 1.0 mm pin, but the first drill bit broke. Now why didn’t I make that sensitive drill press when I first considered it?
So I had to disassemble the parts, and grub and poke around with a fine tungsten probe until all of the bits were out. Then set it up and drill it again. Used a 10BA bolt and nut as the hinge pin.
Then silver soldered some 1mm old drill bits into the previously drilled pin holes as the driving pins for the screw gear.
By this time the day was really heating up.
The threaded post length might need to be adjusted, because I made it slightly longer than thought necessary. I have some spare length at both ends if necessary to adjust.
Unless you have one of these resin printers I suggest that you close this post and look at something of greater interest.
One problem which I encountered with my AnyCubic Mono X 3D printer, was that it was often difficult to separate the print from the base without damaging the print due to excessive adhesion. I have changed the print settings to reduce the initial layer UV exposures to 16 seconds which has helped somewhat, but I decided to try using a magnetic plate. A magnetic plate worked really well on my filament printer, and I was hoping for a similar result on the resin printer.
So I purchased another aluminium base, and a 3M stick on magnetic surface. I could have used the original base, but that would have committed me to using only the magnetic plate surface. Having a second base leaves my options open.
The base which I purchased looked similar to the original, but I noticed that it was not flat. In fact it had a concavity of approximately 0.25mm over its length. Also, it was missing the rather distinctive AnyCubic patterning in the aluminium surface which I think was a reason for the high adhesiveness of the original plate.
So I spent about an hour sanding the base with 200g sandpaper on a surface plate, and finishing with 600g emery paper, also on the surface plate. After that I could not pass a 0.003″ feeler gauge under the edges of the plate. Not dead flat, but should be close enough.
I had watched YouTube accounts of other AnyCubic Mono X owners using these magnetic plates, and finding that the extra thickness caused by the magnetic plate (2.6mm) was too great for the levelling mechanism to function. Various work arounds have been used, including moving the position sensor, and 3D printing a spacer for the sensor, to gain the extra 2.6mm.
My solution? With a milling machine waiting to be used?
I milled the screw slots 3mm longer, and milled 3mm deep rebates along the edges as shown above. Admittedly, the same result could have been achieved with drilling and filing.
The cost? $AUD40 for the new printer plate, and about the same for the magnetic surface.
When the Armstrong 80pr barrel was mounted on a wooden carriage, the angle of elevation was fixed by the weight of the breech resting on a wooden wedge shaped item called a quoin. The quoin was marked with graduations to correspond with degrees of elevation.
To change the elevation, the breech of the barrel was levered using the steps of the carriage cheeks as a fulcrum and the quoin position was adjusted. The trunnions of the barrel were placed forward of the centre of gravity, and the weight that gunners had to lever was considerable.
The angle of the wedge of the quoin was important. Too great and it could shoot out backwards when the gun was fired, and risk injury to the gunners. Too shallow would make it too long or restrict the range of elevations.
Fine adjustment of the angle of elevation was managed with a screw mechanism called a Smith’s Screw, introduced ~1860.
I am currently making a Smith’s Screw for my 1:10 model. I must rely on old drawings of the Smith’s screw, because I have been unable to find a single example of a museum specimen anywhere. And the Smith’s Screws have been removed from all of the existing original wooden carriages. When not in use for actual firing, the screw and handle and gears were removed and placed in storage, along with the gun sights. Who knows what happened to the Smith’s screws when the guns became obsolete.
Some dimensions can be inferred from the base, which sometimes does remain in the original carriage, and from the rounded cavity in the iron pivoting slab which the screw supported. There are very few original wooden carriages, and I have been fortunate to find a handful in Victoria. I am told that they are exceptionally rare in UK, having been broken up when the guns became obsolete. Unfortunately, the drawings which I have found are of poor reproductive quality, and have no dimensions apart from the diameter of the screw (2.25″).
One design feature of which I am reasonably certain is that the screw itself would have been a square thread. Acme threads were introduced in 1894, and replaced square threads in most applications because they were easier and cheaper to manufacture, stronger, and when the nut became worn it could be adjusted to take up the wear. Square thread nuts had to replaced when they became worn. The only downside to the Acme threads was that there was more lateral pressure on the nut, and greater friction and resistance to movement.
So, I have been on a learning exercise to make a square thread. So far I have had about 6 failures. Maybe more. I can see why the square threads are more expensive than the Acme threads.
I had decided to make a 5mm diameter screw. A bit smaller than the 1:10 scale of the 2.25″/57mm original. Actually, 6mm would have been closer. (thinking). It needed to be 1.5″/38mm long. The pitch is unknown, but I had a tungsten cutter which appeared to have been ground for just such a purpose, with a width of 0.8mm, and therefore a pitch of 1.6mm. So the cutter determined the pitch. I have a CNC lathe, so I could decide on any pitch without changing gears. For example I could choose a pitch of 1.6mm, or 1.61mm. Whatever. But to be a square thread the thread depth should equal half of the pitch.
The next problem was with my CNC threading software. Mach 3 has a simple threading “wizard”, and I tried it on my CNC self converted Chinese lathe, which works fine for most applications, but the lathe’s shortcomings (lack of toolpost rigidity mainly), and use of stainless steel rod, gave poor results, then caused the cutter to snap.
So I switched to Ezilathe. Several problems due to my inexperience with square threads vs. conventional 60º threads and a software bug, prompted several calls to the software author, who resolved all software issues without much ado. (thanks Stuart)
But, I was still not getting good results, so I tried my Boxford CNC lathe. It is a beautiful little lathe, but with one serious fault. The tailstock is horrible to use. It is a real fiddle to install, limits the movements of the cross slide/toolpost, and worst of all I did not have a suitable morse 2 centre. I suppose that I should have taken time out and made a dead centre. But I didn’t. Wanting to see some results I pressed on.
With Ezilathe now working well, I decided to practice the square threading using 5mm brass rod. Without a tailstock the 40mm protrusion from the chuck was too much, and the rod bent. Sharpened the cutter, used minute depth of cut (0.02mm), and reduced the protrusion to 22mm, to make a 20mm long thread. Ahhhh. Looking better.
Now to try it with the steel.
That also worked well! A very nice square thread 20mm long, and the rod barely deflected at all. Copious lubricant being brushed on at every pass. 300 rpm. 0.02mm DOC. Sharp cutter.
Now, the rod duly square threaded is required for the screw, but 20mm was a bit short. It really needs to be a minimum of 30mm of thread. 38mm would have been ideal. And I need a length for the screw itself, and another length to make a tap to thread the nut. So I tried a 30mm protrusion. And heard a “click” as the cutter snapped. I think that the deflection causing chatter was the cause. Or maybe the discolouration of that end of the steel indicated that I had used it previously during silver soldering. Maybe I had hardened it.
So I stopped there to lick my wounds, went home and slept on the problem.
Next session I will: 1. make a dead centre for the Boxford, to support longer stick out. 2. Use silver steel instead of stainless steel. It will harden better for the tap, and might turn a bit easier. 3. Use 6mm rod instead of 5mm. For extra rigidity. 4. Make the thread 5mm longer than essential, to keep the cutter clear of the tailstock. I will turn the diameter of the extra 5mm length, down to 5mm diameter, to minimise the impact of the cutter plunge.
Oh, and by the way, I have been making left hand threads. The Boxford has a rear toolpost, and I forgot to invert the cutter which is required to reverse the direction of the chuck to make a right hand thread. I do not know what handedness the original thread had. But right hand is more common generally.
And if all that still fails I will make Acme threads. They will be easier, and at the scale I doubt that most observers will pick the difference.
Next day, next workshop session.
I decided that tailstock support was essential, so I went to my Colchester 2500 Master lathe, and plugged in the 2mm pitch settings. Easy. The tailstock was introduced. I made some right hand threads, on 6mm silver steel, no problems. Just time consuming. Had to regrind the 1mm width cutters several times, but eventually I had 2 reasonable lengths of square thread. One for the Smith’s Screw on the cannon, and one to make a tapping tool.
I machined a taper on the tapping tool, then used a Dremel with grinding wheel to produce the reliefs. Heated the tool to dull red heat and plunged it in cold water. Then gave it some slow heat to anneal it. It was still able to be filed, so the hardening process had not worked well. But it was to be used for only one tapped brass nut, so I accepted it, and proceeded.
Cutting the thread in the brass nut was not easy. I needed several revisions of the thread cutter, using the Dremel with a small grinding wheel.
Gumption is a kitchen cleanser which has a mild grinding action using rotten stone. It lasts only a few strokes, then disappears. But it worked brilliantly, and the nut now fits perfectly in the base. The excess Gumption just wipes or washes off.
So that was a day in the workshop. Not much to show. Maybe I should have spent the day with wine, women and song. It’s OK. SWMBO does not read these posts.
Next session to finish the threaded post with a hemispherical head. (just fantasising about the W, W, and S). Then the cog, handle and corresponding hole in the “iron” support.
I started this mechanism for the Armstrong 80pr gun model today, by making the bronze base.
There were 4 components of the base, which were joined with silver solder. I could have printed the whole base and cast it in bronze, but I had nothing else to cast so decided to fabricate it with basic machining.
Then to machine a recess in the posterior transom.
Next session to make the cylindrical nut with a 5mm acme thread, and the matching threaded post.
I have made an appointment to see the original compressor unit at Warrnambool in a week, so I am deferring making that final component until I have checked some dimensions.
The 80 pr muzzle loading cannon was supplied to the colonial government of Victoria on a wooden traversing platform with a 5º slope.
I assumed that the slope was the means of absorbing the recoil.
The later iron platforms (from about 1875) had a 4º slope and hydraulic recoil control.
But, I was recently informed that there was a wooden “compressor”, which acted as a primitive brake, to reduce the distance of the barrel and carriage recoil. And that there was a compressor at the Flagstaff Hill Museum, Warrnambool, Victoria.
In fact I had previously seen the compressor, but neither I, nor I suspect the museum staff, really understood then how the compressor functioned.
Using Victorian Collections photographs published on the web, my own photographs, information from “The Artillerest” Peter Webster, some old drawings of wooden carriages and platforms, and a Google book “British Smooth Bore Artillery” by David McConnell, and a fair bit of deduction, I think that I have finally worked it out.
Firstly, the Victorian Collections photographs…
Now to make one at 1:10 scale.
P.s. reader Jeff sent me some photos of a recoil control system used in 19th century USA, where a large metal screw clamp was utilised in these rifled muzzle loaders
I had a phone conversation with Peter Webster, “The Artillerist”, yesterday, after I emailed him about the recoil control compressor which I had photographed at Flagstaff Hill, Warrnambool.
I could not see how it could fit into the carriage or slide of the LowMoor cannon, or how it functioned.
Peter, who has a passion for Australian garrison artillery, 1788-1950, and has encyclopaedic knowledge on the subject, had seen this object at Warrnambool almost 20 years ago, realised what it was, and subsequently wrote a report for the museum. The compressor is classified as being extremely rare, most having been removed from the guns, probably to remove the gun metal components for scrap.
Peter explained to me that the flat, rectangular compressor sat between the platform slides with the metal corner tabs resting on top of the slides and the centre join of the compressor located along the centre line between the slides. The front and rear surfaces fitted between the cross members of the carriage.
The central hole was almost vertical. The hole is elliptical, not round. Sitting in the hole was a neat fitting elliptical post, which had a handle which protruded out to the right hand side between the carriage and the slide. When the handle was pulled, the post rotated and increased the separation of the 2 halves of the compressor, pushing them against the sides of the slides, as a brake.
Peter was sure that all carriage/platforms of this type would have been fitted with these compressors, until the wooden structures were replaced with the iron types a decade or so later.
So clearly I will have to make a scale model of the compressor for my current model.
I needed to add some substantially strong rings to the slide of the Armstrong 80pr on the wooden chassis. These rings are the attachment points of the blocks and tackle which are used to point the cannon in the direction of fire. i.e. the traversing mechanism.
Scaling off photographs and drawings I determined that the rings had an o.d. of 100mm, and an i.d. of 50mm. i.e the material was about 25mm diameter.
I had made some rings for a previous project, and had some of the material left over…
So I annealed some thicker rod which was 2.5mm brass…
Actually, the wooden slides were used on other British garrison cannons as well as Armstrongs. For example, at Flagstaff Hill, Warrnambool there is a 68pr LowMoor mounted on a wooden slide, which is identical to the slides used for the Elsternwick Armstrong 80pr’s. And I have a drawing of a breech loading 110pr which was also mounted on an almost identical slide. The only differences were in the carriages, and those differences were minor, depending on the diameter and weight of the various barrels.
So I have used measurements from several slides, located at Port Fairy, Warrnambool, and Elsternwick. The Warrnambool slide is unrestored and badly rotted in some places, allowing inspection of the interiors of the big longitudinal beams. The Elsternwick slides have been restored, painted, and have metal protective covers, which conceal details of the metal strips on the tops of the slides. The Port Fairy slides have been extensively and expertly restored.
And there are always compromises to be made when scaling down structures by a factor of 10. Fasteners for example are only approximately the scale dimensions.
Here are some pics of progress to date on the slide…
To shape the stainless steel strips, on Xmas Eve, I roughly bandsawed them to shape, then milled the edges to end up with 23mm wide strips, 480mm long. The steel is only 1mm thick, so holding it for milling required some planning. Guillotine or laser cutting would have been preferred, but not wanting to wait until mid January for a pro shop to cut it, I did it myself, using 2 bits of straight hardwood to hold the thin stock in 2 identical vices on the milling machine.
For my previous model Armstrong 80pr cannons I made the iron carriage and slides using metal casting of 3D printed PLA filament for the complex castings. The results were OK, but I was not satisfied with the surface finish.
So, I bought a resin printer, and I have been very impressed with the results of the resin prints.
But, to date, I have been unable to get any castable wax resin suitable for the resin printer, with which to make the bronze castings.
So, I decided to revert to traditional machining methods, using reductive technology. Milling, lathe, etc, removing brass chips from bar stock to end up with useable parts.
This is what I am trying to make at 1:10 scale.
Then, I pondered long about how to remove the 20mm of stock which was allowed for the chuck jaws. I realised, too late, that I should have allowed another 10mm or so, because the parting line leaves too little to be held in the lathe chuck while parting.
So, I came up with this work holding solution…..
Actually, 5mm allthread is not much to hold a 36mm diameter piece for parting. So the thread was nutted and lock-nutted at each end. And torqued as tightly as I dared.
After parting the first part by hand winding the cross slide, I became more adventurous with the next three. Made sure that the gibs were tight, the carriage locked, and setting the spindle at 500rpm, used the power feed to do the parting automatically. With plenty of coolant-lubricant (my home made mixture of olive oil and kerosene.). But still finishing with a hacksaw.
With end result shown in photo 2. All good.
Next to make the wheels and axles from steel. Those brass bar offcuts will go into the “might be useful oneday” container.
So, I got a container of basic grey printer resin with my new Anycubic Mono X resin printer, and I have been learning the basics of resin printing. Lots to learn. Not like filament printing at all. Lots of failures, but getting there.
Almost at the point where I would like to make a metal casting, using the lost PLA/resin/wax method.
1 litre of basic grey resin costs about $AUD40.
On YouTube, the experts seem to be using special resins suitable for casting. For example Sirayatech Cast Resin. Costs about 3 times as much as the basic grey resin when postage from US, and taxes are added in. And about 6 times as much as filament on a weight basis.
But, I wondered, can basic grey resin be used for casting? It is MUCH cheaper.
So I performed a little experiment.
I placed two small PLA filament printed objects in the burn out oven, with a resin printed object of about the same size. And progressively turned up the temperatures.
At the end of this simple test, I hesitate to title it an “experiment”, I have to conclude that basic grey printing resin is totally unsuitable for using as a “lost plastic” in metal casting. It leaves too much carbonised ash which would be incorporated into the melted bronze/aluminium.
OK. so I have ordered a litre of the expensive Sirayatech Cast resin.
Actually, I bought it myself. 71 years of experience has taught me that Santa has little clue what I really like. And although it was justified on the basis of being an Xmas present, it did not arrive until New Year’s Eve, thanks to Australia Post. It sat in a clearing facility for 10 days, about 5km from from my house. They were too busy to bring it the 5km. Maybe APO executives are still really pissed off at missing out on their Rolex watch bonuses this year, or whatever.
Anyway, it did finally arrive, and I enjoyed unboxing the bits, and reading the instructions.
IT is a resin printer. An Anycubic Mono X, which converts liquid into plastic objects, with an incredible degree of accuracy and surface detail. 0.05mm layers, which are invisible to my eyes.
On the left is a semi automatic alcohol washer, and an ultaviolet hardening light, which was strongly recommended by various users. After 2 days of use, I am SO glad that I paid the extra $$ for it.
So, why have I moved from a filament 3D printer, to a resin based printer? And paid over $AUD1000 for the gear? (if I had waited until after Xmas I would have got the gear for $100-150 less).
Well, the promise of greater surface detail, absence of visible printing lines, waiting hours rather than days for prints to finish for starters. And it is newer technology, which usually means better. But not always. And the fact that several johnsmachines.com readers have recommended the technology for my cannon parts was quite influential. (thanks guys! You were right.)
There are a few downsides, compared with filament printing.
The liquid resin does have a chemical odour, a bit like rotting fruit, but frankly, it is not too bad. Even SWMBO has not objected to my initial prints being conducted on our breakfast table.
And the resin is said to be toxic. Masks, gloves etc recommended. But I wear neither. I do wash my hands frequently, and I wipe any drops/spills quickly. If I start twitching or talking rubbish or scratching a rash, you will know why.
And the maximum print size is smaller than possible from my filament printer. That had a maximum print size of 300x300x400mm. The MonoX resin printer has a maximum print size of 200x125x245mm. That means that any larger models will need to be split into 2 or pieces, and the parts joined later. But the parts are so accurate, that joining them to make larger models is a real possibility. Bigger resin printers are available, but not at this entry level price.
Resin printing is a bit messy. And cleanliness is essential to get good results and to prevent damage to the machine components. So there are a lot of paper towels, tissues, and alcohol. And I mean 99% Isopropyl Alcohol. I bought 1 litre from Bunnings which cost $AUD29, only to discover that the cleaning machine requires 8 litres. I quickly discovered a firm which sells 20 litres for $AUD100 posted, and bought a container (at $5 per litre).
And what do I have to show you so far?
Well, this is the standard test print. It worked at the first attempt. It is quite small, and I used the default settings. Note: no visible printing lines, no lumps or bumps or support marks. Pretty good!
Then, I had a few frustrating failures. Parts which I had designed, refused to print properly. So I went online to the MonoX users group on Facebook, and I got immediate helpful advice, which did not appear in the official operators manual.
For example, my prints were so strongly attached to the build plate, that I had to destroy them to get them off.
The advice? 1. freeze the build plate and attached parts in the freezer for 30″, then heat them under hot water. They separated easily. 2. reduce the intensity and duration of the UV light to 80%, and 20 seconds (rather than the default 40 seconds). Problem solved.
I am currently printing a cannon wheel bracket, as a test. With 1mm wall thickness. I am not interested in making plastic parts, except to use them to cast bronze or aluminium parts from them. The next test is to burn them in the potters oven to see how much ash remains.
Photo to be added…..
First actual part! 36mm diameter, 1mm thick walls. Drain holes added will be plugged with wax before burn out and casting. How perfect is that surface? (it is a wheel bracket for a wooden slide under an 80pr RML Armstrong cannon.)
This is what the casting looked like after I had removed most of the investment, and turned a flat surface on the top of the casting.
I was surprised that the levering pins, and the big thread came out much better than the simpler flat surfaces. That might be because I had concentrated on those areas with the painted on investment slurry. And also because that end was at the bottom of the pour. That end got the first, most liquid melt, and the pressure of the melt above.
Having decided that my casting equipment is inadequate for this this size and weight object, I do not intend to have another attempt at making a 1:10 bronze Ottoman bombard. Plus, even this sad 3/4 complete component is VERY heavy. I would not enjoy carrying the full size 1:10 model.
The only question remains, what will I do with the above failure? It could join my gallery of failed parts (like the crankshaft of the triple expansion engine, which had a single incorrect dimension). It is useful to occasionally survey this gallery. It does motivate me to measure twice, cut once. Or it could become a very heavy and expensive door stop. Or I could drill out the bore and use it as a specimen flower vase. Or I could cut it up, and use the bronze in future projects. Maybe I will just sit on the decision for a while, unless any readers have any persuasive suggestions.
I am reminded of one of my late father’s aphorisms. “He (or she) who never made a mistake, has never made anything”.
This project was put aside when I broke some ribs unloading the melting furnace which I had borrowed. Each of the 2 halves of the bombard weighed about 8 tonnes in the original, and in my model will weigh about 8 kg each.
These 8 kg parts will be the biggest which I have attempted to cast.
I am using the lost PLA method, having 3D printed the parts in PLA.
Today I attached the PLA breech to a PLA pouring funnel (also 3D printed), and poured the investment medium around the part in a 5″ steel cylinder.
In order to minimise the possibility of air bubbles sticking to surfaces and corners, I painted the entire model with investment, before positioning it in the casting cylinder, and filling it with investment slurry. It will set overnight, and I will commence the burnout in the morning.
Fingers crossed for the pour late tomorrow afternoon.
It is now the next evening. I am despondent.
I woke early, and when I arrived at the workshop at 7:30am turned on the potters oven, and placed the cylinder containing the PLA model and investment medium inside. (Problem #1.) The cylinder was too big to sit vertically or horizontally, so I placed it diagonally. It was awkward, and I was concerned that the bore piece, being supported only at one end, might break free. It did. (Problem #2.) Started the burnout cycle at 250ºc, slowly increasing to 750ºc over 8 hours.
While that was happening I set up the melting furnace, gas cylinders (3 of them), tongs, bucket of water, face masks, gloves, aluminised apron, etc outside. It was going to be a warm day. Unfortunately it was also windy. Not ideal.
Stuart arrived, and he checked his furnace. We lit it to pre warm the furnace and crucible. (Problem #3.) The crucible fitted in the furnace, with little space to spare. Just enough for the crucible with its tongs to fit. Stuart commented that it looked very big. It was, I answered “a 14kg crucible”. When the 12kg of bronze eventually melted it only half filled the crucible. It was not the size which I had ordered. It was too big, and restricted the gas flame, reducing its effectiveness. The melting phase required 3 hours! Much too long. (Problem #4.) (PPS. note added 23 Dec. I checked the dimensions of the crucible. It is a 30kg crucible!!! No wonder it was too big for the furnace! I had ordered and paid for a 14kg crucible. No wonder it was too big for the furnace. I should have checked before using it.)
Then it appeared that the flame was not as fierce as Stuart expected. The gas was piped from 2 cylinders, and one was not icing up as expected. It was close to full. Why was the gas not coming through? Could there be a ball valve somewhere in the system? Later we discovered that the pipe from that cylinder worked in only one direction because there was indeed a hidden one way valve. There was no direction arrow. (Problem #5.)
So, when we did get to the pour, and discovered the central core broken free (#1),
I inverted the now red hot cylinder to shake the core free. I calculated that the bore would fill with bronze and need drilling later. But would there be enough molten bronze to fill the cavity? I had allowed 1.5-2kg extra bronze to cope with unexpected contingencies but this would be cutting things fine.
So, we did the pour. There was a LOT of slag, possibly due to the slow melt. The molten bronze seemed to pour OK, and it filled the mold and the central bore. But it stopped about 3 cm from the top. Bummer!. Not enough bronze. Oh well. A learning experience.
And worst of all….
It is only half the weight of the cannon, and it is just too bloody heavy!
I could fix the mistakes, reprint the part, and recast it.
But, you know what? I am not going to. The biggest issue is that even if I am able to fix all of the problems, and get a good result, it will be too heavy to move around. It will be too heavy to use even as a door stop. Hmm. Maybe I will clean up the failure and use it as a heavy door stop. Either that, or cut it up and reuse the bronze in the next casting projects, which will be much smaller!
This will be another failed, abandoned project to add to the list. (Chess pieces, etc). Oh well. Live and learn.
(it does cause me to appreciate the Ottoman cannon makers of 1465 who cast these parts with wood fires, where each component weighed over 8 tonnes!)
“The Artillerist”, Peter Webster is a Sydney based expert on historic Australian artillery. So I contacted Peter to see if he could explain how the 4 ton barrel of the Armstrong 80pr was elevated when it was mounted on the wooden carriage and slide.
Peter explained in detail that there was a screw sitting in a gunmetal nut which raised an iron bar on which the breech of the barrel rested. If more depression of the barrel was required, a wooden wedge (quoin) was inserted between the barrel and the iron bar. Peter had seen this arrangement on a cannon at Fort Queenscliff.
Several other readers have sent me diagrams from old publications of the setup, and I sincerely thank those readers for their help. Here is one of the diagrams.
I could have made the model screw and quoin from these details, but I decided to visit the Queenscliff Fort to see them for myself. Queenscliff is only a 30″ drive away. It has been Covid closed to visitors for almost 2 years, but had reopened very recently. So off I went today.
The 1.5 hr tour included the cells, the magazines, the remaining guns, the lighthouses, the museum.
Another interesting story which I had never previously heard, was from WW2, 1942. An aeroplane was launched from a Japanese submarine in Bass Strait. The plane flew around Port Phillip Bay, taking aerial reconnaisance photographs. It was spotted from Fort Queenscliff, but by the time it was realised to be the enemy, it had gone. Telephone calls to the Laverton airforce base were similarly unsuccessful in raising a response in time. The plane completed its mission and was picked up by the submarine. The pilot visited Australia after the war and related the story, and showed photographs. Needless to say, the Australian population was not informed until many years later. Google showed this article…https://www.ozatwar.com/japrecce/recce02.htm.
After the tour had finished I was quite disappointed not to have seen the gun and wooden carriage indicated by Peter Webster. So I asked the volunteer guide about it. She kindly introduced me to the gun expert at the museum. He took me to the only gun which matched the description, away from the tourist areas.
Bummer! The elevating mechanism is missing, replaced by a wooden prop which was used when the gun was not in use.
Working with wood. It is quite nice to get back into the woodworking. And slightly daunting. Those saws can remove a finger or a limb in an instant of inattention. I use a 12″ radial arm saw, and an 18″ bandsaw. Somehow, the woodworking tools seem more dangerous than the mill or lathe. However, having seen videos and pictures of metal working lathe accidents, where an arm was ripped off at the shoulder, and similar, I know that they are ALL dangerous. At the time of writing I still have all of my bits.
At 1:10 scale, the wooden beams which form the base for the slides are 488mm long, and 30x30mm square section. They have a 5º slope back down to front.
I am using Victorian mountain ash, a pale, tight grained hardwood, and I happen to have some offcuts in my hoardings.
Oh. And some really useful woodworking tools which I bought from Banggood last year, and used for the first time on this project. They are laser cut spring steel, with holes and slots at 1mm and 0.25mm intervals, and a propelling pencil for marking. Accurate by woodworking standards, and they work really well, and were not overly expensive ($15-20 from memory).
And another bit of technology which I find useful with this project….
By fiddling with the magnification settings on our printer, I was able to print the plan on A3 paper, at a scale of 1:2 of my 1:10 model. The plan is quite accurate, allowing me to measure off dimensions of the components, angles and so on. This has been really useful.
Note that the wooden assembly is held together with large nutted bolts. And mortise/tenon joints as revealed by the Warrnambool LowMoor cannon. I will use bolts, and brass dowels, because MT joints are fiddly, difficult to make accurately, and will not be visible.
This is the video which I shot at the Flagstaff Hill Maritime Museum of the LowMoor 68pr cannon on an original teak wooden slide and carriage. It focusses on structural aspects, which I can review when I am building the model. It will probably be boring for most viewers, but I am posting it in case it is useful or interesting to some.
The cannon is actually located outside the front entrance of the museum. On this occasion I did not go through the museum, but can highly recommend it for many fascinating exhibits, including the Armstrong 80pr RML cannons, artefacts from the tragic wreck of the “Loch Ard” especially the beautiful ceramic peacock, and the recreated colonial village. It is well worth visiting.
When I sat down at the computer to draw up plans for the wooden chassis using the dimensions and photos I had obtained at Elsternwick, I realised that I needed some extra details. Some measurements I had just forgotten to take. And some details were not visible due to the protective covers on the Elsternwick cannons.
But, I remembered that there was a wooden chassis at the Maritime Museum, Warrnambool, and that it has not been restored. In fact it was an original teak slide and carriage, supporting an older smooth bore 68lb muzzle loader. I seemed to recall that the slides had rotted away to some extent, and that might reveal how the transverse beams were joined to the long slides, details that I had not been able to determine at Elsternwick. Being an older cannon, the slide and carriage might have been different from those at Elsternwick, but I decided to make the 2.5hr drive and check it out. 30 minutes further on were the restored wooden chasses at Port Fairy, so I decided to make a day of it.
So I was able to obtain the missing measurements, and to see that the transverse beams were joined to the long slide beams with large mortise and tenon joints.
However I was still puzzled by the barrel elevation mechanism. Was it a quoin (wooden wedge), or a screw mechanism? Or possibly both? And if both, why?
But, when I checked my blog at home that night, several readers from Australia and USA/Canada had provided references which described the mechanism. Thanks Jefenry, David and Richard. (and Australian expert, “the Artillerist” Peter Webster).
The barrel elevating mechanism is a large screw with the nut in the cross beam (the rear transom), which supports the hinged iron beam, and above that is a wooden wedge (quoin). Apparently the screw was for fine adjustments and the quoin for larger adjustments. I am reasonably convinced that was the arrangement of the Elsternwick 80pr Armstrong RML’s too.
In the diagram above note the roller/lever. That was used to lever up the rear of the carriage, to transfer the weight of the carriage and the barrel to the front wheels, permitting it to be rolled to the firing position at the front of the slide. Sometimes that process was bit uncontrolled, so the rope and bollard were added to control the rate of forward motion/descent.
Incidentally, that barrel is the one which was made in miniature by Jefenry, and featured on You Tube, firing at a range and off a canoe! Worth a search on You Tube. Just do a search on videos by Jefenry, or try these links.
The videos are from Jefenry, who is located in the USA.
So, today I battled 1.5 hours of post covid lockdown Melbourne traffic to take a closer look at the wooden slide and carriage of this 1866 Armstrong RML cannon which I am intending to model. There are 2 of them in the Hopetoun Gardens, Elsternwick. They are more complex than I had imagined.
The barrel is identical to the barrels which I had modelled on iron slides. The iron slides were a later improvement – modification.
It was a beautiful spring 20ºc day. I spent almost 3 hours photographing and measuring the wooden components. Some parts have been restored, and it was lucky that there are 2 examples to check and compare.
I was climbing over and under the cannon, and groundsman came over to check. Then a pair of grandparents came over with their 5 yo grand-daughter, and a further pleasant conversation followed.
Some examples of the photos…..
And an example of many pages of measurements and sketches…
There are 11 pages filled with details like this, representing my 3 hours.
And I still do not understand how the barrel elevation mechanism functioned. It could have been a wooden wedge called a quoin, but there appears to be a metallic disk set into the wooden bearer. Could there have been a screw mechanism which has since been removed/stolen/lost? Pictures on Google Images do not help. Does anyone know?
You might have seen the above photo in johnsmachines.com earlier this year. 2 Armstrong 80pr muzzle loading, rifled cannons are sitting on their ORIGINAL wooden slides, in Hopetoun Gardens, Elsternwick, Victoria. These were originally installed as garrison guns at Fort Gellibrand, Williamstown, Victoria and were never upgraded with the more modern, accurate iron and geared slides such as at Warrnambool, Port Fairy and Portland, and which were the inspiration for my 2020-21 modelling efforts.
I have decided that I will make another 1:10 scale model of the Armstrong 80pr RML, this time sitting on a wooden slide.
You might wonder why I am so obsessed with this particular cannon? Well, I wonder too. Perhaps it is the ready availability of an original in good condition, which I can visit, photograph and measure.
Anyway, I have made a start on this next model.
The next step was to centre the 10kg rod in a 4 jaw chuck, install a 3 jaw steady, and drill the 16mm bore. Sorry, no photos, forgot. I had made a long series 16 mm drill bit by turning a shoulder on the shank of a good 16mm bit, and boring an accurate hole in the end of some 5/8″ (15.875mm) drill rod, and silver soldering them together. Then honing the bore to an accurate 16mm diameter, along its 275mm length. It worked well. So well, that I can insert a 16mm “projectile” in the bore, and watch it slowly drop through.
Then, continuing to hold the blank rod with its 16mm bore in the 4 jaw, and using the tailstock to hold the other end I manually turned the exterior of the barrel.
My tandem trailer, was desperate for repairs and repainting. I bought the trailer about 23 years ago. It is 10′ x 5″, very solid construction, and a hydraulic tipper. I used it originally to transport animal manure to my olive trees. In recent years it gets more use on SWMBO’s building sites to remove builders rubbish.
Later, I increased the size of the ram to a multistage 5 ton unit, and changed the geometry to provide more lifting power. Also contracted a professional trailer maker to install Landcruiser hubs and wheels and heavier duty springs.
But the floor finally rusted through. So I installed a new 2mm thick steel floor over the top of the old rusted one. I should have removed the old rusted floor, but time was short, so I took the “lazy man’s” option.
Now, 23 years after the original trailer purchase, and about 18 years after the temporary floor fix, the floor needed to be replaced again. This time I did the job properly. I bought 2 sheets of checker-plate steel. Paid the supplier to fold the long edges. And started to remove both layers of the old floor.
Then attached the new floor to the trailer frame with galvanised hex head screws. You might wonder why I did not weld it in place? Well, removing the previous floor which had been welded in position was job which I never want to repeat. Plus, whether the new floor is galvanised or painted, welding destroys the zinc or paint, including in areas which cannot be touched up. The technique which I used allowed all surfaces to be thoroughly painted. So I removed the newly joined floor, and primed and top coated all surfaces, including the frame underneath.
Then re-attached the new painted floor permanently to the frame with the gal screws, and cut off the protruding points. I decided to not weld the floor to the frame, because that would destroy the rust inhibiting paint. I used silicone roof and spout sealant in the screw holes, and between the trailer sides and the new floor side lips. Then applied more silicone sealant into any cracks between the side lips and trailer sides.
Oh, I forgot. The cross members were U sections with the opening at the top. No wonder they rusted. Any accumulated water could not escape. A really dumb design decision by the maker. So I drilled drain holes in every cross member, removed the rust with a needle gun, and painted the insides of the U sections. I will finish the painting when the steel repairs have been completed.
Judging by the coats of existing paint, and alterations to the trailer construction, I reckon that this is the 4th major change to this trailer’s construction. It has done a lot of work. And lots more to come.
I have a 35 year old JCB back hoe, a left over from when I grew olive trees and made olive oil. These days it is used only as a yard crane, and other small jobs on a 5 acre property. But since I am between model engine and cannon projects I decided to expend some TLC on the rather neglected JCB.
One feature which always made me grimace was the broken slew lock plate.
A new plate is $AUD 1300 + GST I could not find a machine for wrecking, and was told that wrecked JCB 3CX’s are almost unknown, most owners, mostly farmers, keep them going for ever.
So I measured up the plate, drew it up on CAD, and had it laser cut from 20mm plate.
My JCB is a 1986 model, 35 years old. I have owned it for 10-12 years, using it for manure handling, as a yard crane, tree transplanting, and twice to lift a 50,000 litre water tank onto a tray truck. (in combination with a front end loader). Despite its age and hours of work (about 7200), it has been very reliable.
When I first bought it I engaged Enzed P/L to inspect and change any suspect hydraulic hoses, because many of them showed signs of cracking and delayering of the exteriors. And one had burst. I think that the onsite Enzed engineer remade and installed about 8-10 hoses.
Then recently, the hose to the rear bucket and boom controller burst. Well, actually it was the metal fitting which was crimped to the hose which split and allowed the hose to separate from the fitting, spilling quite a bit of hydraulic oil onto the ground. Enzed were on the job within a day, and an hour later the new large diameter hose and fittings were installed.
The hydraulic oil level was well down, and topped up with 20 litres.
I had been aware that one of the hydraulic cylinders had been leaking for many months, so not all of that 20 litre deficit was due to the burst hose. It had been leaking at the rate of one drop every 6 seconds for a long time, and increased to one drop per second recently.
I had recently purchased a kit of new seals for the leaking cylinder. After watching some YouTube videos, I decided to have a go at replacing the seals myself. No big deal really, except that I had not worked on a cylinder this large before.
The large gland nut was loosened before I removed the end pins. I had been warned that the nut would be extremely tight, and that it would be easier to loosen the nut while still attached to the JCB. The only spanner/wrench large enough was a Stillson wrench. I did not enjoy using it because it bit into the nut and marred the smooth surface. I had priced a new open ender spanner, but the cost was high so I used the Stillson. The Stillson was about a meter long, and the nut would not budge, despite using all of my strength and weight. Using a 1.5 meter pipe extension, the nut finally moved, and I loosened it until it stopped fighting. I caught much of the hydraulic oil which spilled out in a bucket, not to be reused of course.
I carried the ram into my workshop. At that moment I decided that any future resealing jobs on larger rams would be done professionally. It was quite heavy.
In the workshop I completed the removal of the large gland nut, discovered that the gland O ring and the main seal were in pieces.
Then removed the piston and its rod from the cylinder by pulling carefully.
The piston seals looked OK to my inexpert eye. But I had purchased a whole new kit of seals, including the piston seals and guides, so I replaced them all.
By the way, I had obtained a JCB service manual for my machine online, from Best Manuals, (USA) for $US19.99 and downloaded electronically. It had some very useful information. Including that replacement seals might be different from the originals. And they were. They looked different.
The seals went onto the piston fairly easily, with the assistance of a large cable tie, which was used to pull the large central rubber seal into its groove. The remainder of the piston seals were split, and positioned easily. The gland seals fitted easily, except for the large internal rubber seal which required considerable pushing and effort.
The gland was then pushed onto the piston rod, then the piston was screwed on, after cleaning the threads, treating them with Loctite 7471, and thread retainer 720. Then tightened with wrench and 1.5m extension.
Then I attempted to insert the piston into the cylinder. But it would not go. So I pushed harder. Still no go. Tried wiggling. twisting, pushing harder. No Go. Bummer. 4pm on a Friday afternoon. Long holiday weekend imminent. Threw the parts into the boot and drove to Enzed. About 15″ away. Maybe there is a special tool or press to push the parts together?
They helped me immediately. I guess that my previous business a few days earlier and quick payment of their bill helped. They mounted the cylinder in a chain- pipe vice, and pushed and strained, but no better result than I had. The piston with its new seals just would not fit into the cylinder.
So, he compared the old and the new seals. The new ones were 0.3mm thicker than the old ones. And the diameters seemed larger. I had bought the seals from a JCB dealer, who had said that they were “after market”. Originals no longer made for such an old machine.
Discussion. Decision. Use the old piston seals. The new gland seals seemed fine, so keep them. The old piston seals appeared to be in good condition, and they were installed. And guess what? The piston slid into the cylinder with some pushing. The gland nut was tightened. (with a Stillson I noted).
The charge for 30 minutes of heavy, dirty, specialist time??? $AUD 23.00. I said, “that seems too cheap”. “That’s OK” he said. They will continue to get my business.
Today I re-installed the ram on the JCB.
Fired up the diesel. No leaks!
Operated the bucket control……no leaks, but no movement! Bummer! What now.
Hmm. Could I have put the hoses on back to front? No. They looked correct.
But hang on, the cylinder is facing the wrong way! The paint scuff marks which were on the outside are now facing inside!
I HAD INSTALLED THE RAM 180º ROTATED!
So, left the ram as was, switched the hoses, and tried again.
WOO HOO! No Leaks! (including when I tried different ram positions off camera).
I am still waiting to pour bronze to make a 1:10 scale Ottoman bombard. I have all the equipment and materials ready to go. But, held up by 1. needing some dry, wind free days, 2. ability to have a friend on hand to assist with the pour. I need 2 successive days for making each mould and doing the pour. Waiting, waiting. Meanwhile, SWMBO has had me breaking up a concrete drive, manually loading the broken up concrete, and transporting it to the recycler. It took 3 x 2 ton loads, so far, and still more to go.
Then in lockdown, we decided to fix a leaking balcony at home, and replace some rotting, ceiling boards under the leaks. At 71, I dislike working on ladders, above my head. Almost finished, thank goodness. And no more broken bones.
The lockdown restrictions eased a bit last weekend, permitting me to visit my workshop. I have been waiting for some suitable weather to burn some rubbish, and it was not too windy a couple of days ago, so this was the first task…
I have been considering my next modelling project. Nothing really is grabbing my attention. But I had to use the JCB backhoe to load the concrete onto the tipping trailer and I was aware that the JCB is looking really tired. Not surprising, considering that it is 36 years old, and has 7200 hours on the clock. I bought it third hand, more than a decade ago, and used it for general farm jobs, including manure handling, transplanting mature olive trees, as a yard crane, moving machinery, digging trenches. As a general farm machine it was incredibly useful. When the farm was sold, it was just about the only big machine which I retained, because even on 5 acres it is still used occasionally.
So, until some new model engineering project takes control of my life, I will spend some time and TLC on the JCB.
I watch YouTube videos about megalithic sites around the world, including Peru, Cambodia, Russia, Bolivia, and especially, Egypt. I have been fascinated in the subject for over 50 years, since reading an article in National Geographic as a teenager, about the almost unbelievable stone work in Peru which was then ascribed to the Incas, (but that Inca origin theory now has many serious doubters).
One of those YT sites, “UnchartedX” , (to which I subscribe and support), frequently refers to the book “Lost Technologies of Ancient Egypt”, and recently did a 2 hour interview with the author, Christopher Dunn. The book was published in 2010, based on many visits to Egypt by the author. The interview led me to purchasing and reading the book. Although now 9 years (oops 11 years) since publication, his work is respected by Egyptologists, academics, and more free thinking enthusiasts such as YouTubers like Ben of UnchartedX, quite an accomplishment considering the degree of hostility between the opposing views.
Christopher Dunn is/was a toolmaker, engineer, and manager in the US aerospace industry, and expert user of CAD, CAM, lasers, metrology, and photogrammetry. He is also into ancient history. So when he visited Egypt he looked at the pyramids and other buildings and monuments, with the eyes of an engineer, and wondered how they “did it”. Over the course of many visits, he took increasingly sophisticated metrology devices and started to measure and take detailed photographs of monuments, temples, statues and quarries. He was staggered to discover the degree of precision to which many of these huge objects were made, in many cases of granite or basalt some of the hardest of all stones.
And he examined magnified views of the surfaces, to see the marks which remained, which might give clues about the tools which were used to create the objects, which in some cases are at least 4500 years old.
He carefully analyses the Egyptologists’ views that the tools were simple and primitive. Like copper chisels, and stone pounding rocks, and while not dismissing those views out of hand, leaves us with the impression that such results would be almost impossible in this CAD CAM era, and much less with copper tools and stone pounders. He does not mention aliens or pre-dynastic civilisations, but just states that the tools which made the pyramids, obelisks, huge precise statues, and stone boxes, those tools, unlike the copper chisels and stone pounders, have never been found.
He does point out evidence of large circular saws with a 38 feet diameter blade, hole saws up to 6″ diameter, and straight saws which have left tell tale marks in stone objects and quarries in many places.
In many cases, such as the huge, incredibly precise stone boxes in the Serapeum, and the absolutely identical pair of 40′ statues of Rameses 2, he just states “we have no idea how this was done”. The precision is not just in linear measurements, but in complex curves, and surface shapes and areas.
Dunn’s analysis is principally about the tools and engineering of Ancient Egypt. Equally fascinating, but not covered in this book, are the mathematics associated with the pyramids, but that is another story. Also, he does not believe that the Great Pyramid was designed as a tomb, but as a machine. But that also is the subject of another of his books, which I have not, as yet, read.
“Lost Technologies” is 360 pages, paper bound, illustrated with many black and white photos of variable quality, many excellent diagrams, and 16 pages of good colour plates.
The text is technical, but quite readable cover to cover. I found it difficult to put down. If you enjoyed Simon Winchester’s “Exactly” you will probably like this one.
This is not an in-depth examination of the question. It is rather my experience with a particular 3D printer. But I believe that my experience has been experienced by many other 3D printer owners, so I have decided to make this record.
The printer which I purchased was a Creality CR 10s. I bought it in January 2020. I had been considering such a purchase for a year or more, and finally took the plunge when faced by making components for the model Armstrong cannon which featured in this blog over the past 18 months, I realised that I would have to metal cast quite a few complex parts. And the “lost PLA” method seemed like the best option to cast those parts.
So I accumulated the equipment for printing the PLA, making the molds from jeweller’s investment medium, a potter’s oven for burning out the PLA and baking the moulds, a furnace for melting the aluminium and bronze, and assorted other necessary paraphernalia.
Choosing the 3D printer was difficult, coming from a knowledge base of close to zero. I asked members of my model engineering society which printers they chose, and why. I watched numerous YouTube videos, and read reviews. As usual, I discovered that the more information you absorb, the more confusing it all becomes. I have the same feelings when researching which car or camera to buy. Eventually, I decided to buy a lowish cost printer, with the idea that I would eventually replace it with a better unit for long term use.
For this first 3D printer I wanted a well known brand with a good reputation, a build volume which would allow me to print the biggest component of the model cannon (the barrel, which is 300mm long x 65mm x 100mm).
I was still mulling this choice when Amazon advertised a special deal for the CR10s, and I made a snap decision to take the plunge. Would I make the same choice today? Quite possibly (actually, no. see below). Although technology has advanced. I am now considering whether to add an Elegoo Saturn resin 3D printer to my ever growing list of machines.
So this is what I bought. It is an open frame 3D printer, with a separate box for the motherboard and controls, and a side mounted spool. Single extruder. Advertised as an auto self levelling base (but it is not. It is manual, time consuming and fiddly). Filament end detection (hence the “s” after the 10), which works well. The build volume is 300x300x400(h), which is at the high end of low cost 3D printers, and bigger than any of the low/medium cost resin printers. I used close to the full extents of the volume on several occasions.
Actually, the very first modification was the slicer software. The printer came with a free version of “Cura”, but I accepted some expert advice to use “Simplify 3D”, which I purchased ($AUD 175) and used exclusively until recently. More about that later.
The instructions for using the printer were in an illustrated 10 page booklet, and a pdf file. As instructions for assembling the printer, they were just OK (do manufacturers EVER test their instruction booklets on novices??). As instructions for fault diagnosis they are hopelessly inadequate. In my previous post I showed a paper back book which would have been immensely useful when I started this 3D printing journey, and HAS been immensely useful after almost 2 years of wallowing about in ignorance. (“3D Printing Failures” by Sean Aranda).
By trial and error (mostly error), I printed the parts for my model cannon, and also came to grips with the casting processes.
BUT. When I started printing components for my next project I experienced failures and frustrations which I could not overcome. The Ottoman bombard has only 2 components, the breech and the barrel, which I intend to cast in bronze, using the lost PLA technique. These will be the biggest castings, and biggest 3D prints which I have attempted. Not surprisingly, I had problems with the 3D prints. Some of my attempts at fixing the problems caused further problems. After reading Sean Aranda’s book, discovered that my problems had ALL been described, catalogued, and fixes known.
The problems were:
1. Poor plate adhesion causing models to break free during printing.
2. Poor adhesion between layers causing gaps and structural failures.
3. XY shifting between layers
3. Gaps between filaments which would cause casting holes.
4. Excessive stringing.
5. Lumps on surfaces.
And this was typical of the failures….
Then I decided on some upgrades….
An enclosure to prevents drafts, and keep the printing environment warm during cold nights. $AUD155. A temperature and humidity logger kept a record of overnight temperatures. The heated printer bed provided the heat. I noted that temperatures remained between 22 and 26ºc inside the enclosure.
An all-metal fully geared extruder. $AUD25
A filament dehydrator, and warmer. “Sunlu” brand. $AUD60. Old filament can be reconditioned by warming at 50ºc for 4-8 hours. The same machine can hold the filament during printing to keep it warm and dry.
Around this time I experienced a serious filament leak and blockage which bent the hot end enclosure, broke the wire to the thermostat, and broke some insulating material. The leak was caused by a loose extruder nozzle, and an imperfectly seated Bowden tube. It was probably repairable, but when I saw that the cost of a replacement unit, including the wiring loom, hot end, 2 fans, silicon boot, etc was only $AUD35. So I bought one, had it installed easily in a few minutes. It came very well packaged, and quickly.
Also about this time I read Sean Aranda’s book. It has been a game changer.
Aranda uses “Cura”, and although his fixes can been used by any other slicer, one of my problems was holes between walls and internal surfaces. He says that this is a problem which is worse with “Simplify 3D” than with “Cura”, and he also thinks that “Cura is a better program overall, mainly due to the quality and number of online updates. The fact that “Cura” is free is an added bonus.
4. Sean Aranda’s book. “3D Printing Failures”
5. Changed slicer to “Cura”. Although it is called “Ultimater Cura” it works on most if not all 3D filament printers.
6. Changed the stick-on printing surface to a new 3M cover. This was after I read the product information which stated that these surfaces last for only 10 prints! I am quite sure that mine lasted for at least 50 prints before becoming unusable. Now I print on different areas of the surface, and keep a record of the number of prints at each location. Since then I have bought a magnetic cover which I will use when the current 3M cover starts to fail. (p.s. I have now installed the magnetic base, and so far, it has been wonderful!). No break aways despite not using brims or platforms, and easy to remove prints
After all of those changes my prints have been excellent. No break aways, no X-Y layer shifts, good adhesion between layers, better surfaces, and no holes/gaps between filaments. The only problem is that I am not sure exactly which changes were effective and which ones were not. Probably they have all helped to a degree.
Postscript. I have been considering buying an Elegoo Saturn resin printer, or maybe even substituting the Saturn for my Creality CR-10s. From the reviews the Saturn produces much smoother surfaces, and more precise dimensions. And the prints are much faster. My 3 and 4 day barrel prints could be printed 5-10 times faster. The known down side is the is the cost of the machine (on special at Amazon at present for $AUD639), the smell, the need to avoid skin contact with the uncured resin, the desirability of a print washer/UV curer, and the need for extra space. The other major consideration for me is the smaller maximum print size. 200x192x125mm. The bombard parts would need to printed in halves and glued together in order to make the molds. So, while the CR 10s is working well, I will hold off buying the resin printer. There is some advantage in waiting because 3D printing is a constantly evolving and improving technology.
So, were the upgraded components worth it? A resounding yes, as far as I am concerned. Still pondering the Elegoo Saturn
P.S. a month or 2 later. Since I wrote the original article I have taken delivery of, and installed some 1 meter cable extensions, which were made for this machine. Obviously I am not the only Cr10s owner to have decided that the cables are a bit too short. The 6 or 7 cables were fitted with connectors and installed quickly and easily. The printer functions perfectly, and I can now place the control box a comfortable distance from the printer enclosure without worrying whether the too short cables will cause a print failure. Actually, the one meter extensions are a bit too long, and need to be carefully positioned to avoid snagging. Half a meter extensions would have been ideal. And the cost? $AUD35 with free postage. Considering the labour involved in making and packaging these items, the price was cheap.
I guess that title should read “2 Person Tongs” but I doubt that SWMBO will be volunteering.
I am still planning to pour a model bronze Ottoman bombard.
The plastic model has been 3D printed, the flasks for the investment powder mould are ready, and I have the potter’s oven ready to dry, burnout, bake, and heat the moulds.
I have borrowed a melting furnace from Stuart Tankard, which is large enough to fit the crucible. The crucible has 14kg capacity. The crucible itself weighs 4kg. Unloading the furnace from my Toyota Landcruiser cost me a couple of broken ribs, which set back the project a few weeks.
Then I wondered about tongs to insert the crucible into the furnace, and, more importantly, how to lift the crucible full of molten bronze out of the furnace and pour the bronze into the moulds. The weight to lift and pour I estimate to be: bronze 10kg, crucible 4kg, plus tongs say 4kg = 18kg. The crucible with its bronze load will be at approx. 1100ºc / 2000ºf so some distance will be required for the gloved hands from the red-hot load.
I have several pairs of tongs for smaller crucibles, but nothing approaching a 14kg crucible. So I asked Stuart T for his thoughts on the matter. He recalled seeing a video by an MSMEE member and suggested that I check it out.
John M’s tongs looked like they had been designed by an engineer, which was actually the case. I contacted him (by email because Melbourne is in Covid lockdown), and he generously offered to send photos, a video and a drawing.
I copied his design, with a few modifications based on the materials which I had on hand, and also to enable a 2 man lift and pour. In retrospect, I could have fabricated a one man pouring apparatus, using a swivel on a frame, but to be honest I would prefer someone else present for safety reasons.
The remainder of the tongs construction was basic cutting, welding, and drilling.
I used to be a half reasonable amateur welder, but lack of ongoing practice lately, and dodgy eyesight is my excuse for the lumpy welds and essential use of an angle grinder.
Next steps…. I need some dry, non windy weather, and availability of assistance for the pour. I will make the first mould, of the breech since it is shorter than the barrel, dry it, burn out the PLA, and bake it at 750ºc. That will take most of a day. While the baking is in progress (about 4 hours), I will start the melting of the bronze ingot. Stuart says that I will require 2 full 20kg cylinders of propane.
Then the pour. Then after some cooling with fingers crossed. Camera running…..
The 6th Victorian Covid lockdown was the shortest, but seemed to hit me the hardest. It was unexpectedly relaxed after only 5 days in regional Victoria, where I live. With escalating numbers in Melbourne, and Sydney, and NSW reacting by putting its collective heads in the sand we expected the be in lockdown for weeks or months, and frankly it was quite depressing. For the first 3-4 days I did a bit of garden tidying (with a chainsaw, much to SWMBO’s horror), and time on YouTube, Ebay, Banggood and Amazon. A fair bit of impulse buying, as follows.
Paragraph deleted. My political and religious views have predictably caused offense to some of my readers. While I do not resile from any of those views I accept that others have different views, and me having a rant is unlikely to be at all persuasive. So I have removed the paragraph. For those who agree with my views, my apology. Any further conversation about Trump, Liberal and labor politics in Oz, and religion, will have to be in private. (I still consider Trump to be a lying, ignorant, con man, and a disaster for USA and western democracies.)
So, having offended and lost 3/4 of my readers I will get back to my little buying spree…..
Firstly, a book.
I read some good reviews about this book, and since I have had considerable frustration with my 3D printing of late, I decided to buy it. 3D PRINTING FAILURES by SEAN ARANDA.
Paper back, 298 pages, large format, large print font size, 2020 edition. Under $AUD30 including postage from Amazon.Australia.
And it looks excellent. Clearly laid out and written, lots of pictures and diagrams, and the author even gives his email address and offers expert help if there should be a problem not covered in the book.
Some of the pictures admittedly are not great quality, but the author has a service which astounded me. If proof of purchase is emailed to him he will send high definition colour photographs for download. He sent me all of the photographs within 24 hours of my request. AND, a pdf version of the entire book. AND, a promise to send me free of charge a PDF version of the 2022 edition which will be published near the end of this year!
I have cherry picked some of the chapters and I am VERY impressed. They are VERY helpful. Some random pages follow….
This book should be included with every 3D filament printer purchased. Note that it does not cover liquid 3D printers.
In my previous post I showed a photograph of the enclosure which I cobbled together from cartons and a blanket to try and avoid printing problems arising from overnight temperature drops, and draughts. I intended to make an enclosure from MDF and perspex, but while browsing Ebay came across this one.
As you can see it was not a trivial cost. But when I factored in the difficulty in obtaining the materials during the lockdown, and the fact that the commercial one claimed some fire resistance, I bought it.
It came today, and with some levering of the cover on the frame, it assembled quite neatly, tightly, and well. Here it is with my printer.
As you can see the electronic control box is outside the enclosure. There is a flap on top for those who prefer the filament reel on top.
The front and top zip open. And there is some spare room for bits and pieces. It does look slightly neater than the previously used cardboard boxes. The printer is fully enclosed, even with a build in floor. The price seems to have risen a bit since I paid for this one. Time will tell if the print quality improves. I am predicting that the print quality will improve. After reading the chapter on fire safety and 3D printing in the book above I will feel more comfortable about leaving the printer unattended with this “Fire proof” enclosure. I suggest interpret that as “fire resistant”. I will be watching temperatures closely for the first few runs.
Still on the subject of retail therapy, a couple more purchases….
This is a woodworker’s gauge from Banggood. I bought it after watching a YouTube video about its uses. Nicely made, and reasonably accurate by wood working standards. I will do a separate post about it when I have more fully explored its applications. (it is for making perfect grooves and lap joints on a table saw).
And finally, this one was a splurge, impulse buy. But something that I had wished I had on quite a few occasions when making models.
As you can see it is a pin gauge set. It is Imperial because it was a fraction of the cost of a metric set. 190 pieces of ground and laser labelled cylinders, up to 1/4″. They seem to be as accurate as my Mitutoyo micrometer can assess. It does mean that I will be committed to a moment of calculation to metric when in use. Cost? About 50 cents per piece, including the case and postage.
Fortunately for my credit card, the lockdown ended 2 days ago. I have spent a couple of short sessions in the workshop, tidying up and doing some machine repairs and maintenance. Nothing really to show. But it is nice to be back.
It has been cold here during the current lockdown. And I mean temperatures. Not by American midwest standards by any means, but since we are confined to our homes except for limited predefined purposes, some days and nights are chilly. Down to 5-8ºc here.
I have been spending a lot of lockdown time doing 3D prints. And really struggling to get decent results.
Some of my GSMEE colleagues have been urging me to make an enclosure for my 3D printer. To be honest, Stuart T had urged me originally to buy a printer with an enclosure, but I pressed ahead and purchased an open structure model because I wanted the extra print size it offered. The Creality CR10s can print up to 300x300x400mm which I have fully used for my Ottoman bombard prints.
But in recent weeks, with the onset of the cold weather, I have noticed a distinct deterioration in print quality, particularly with poor layer adhesion when printing overnight, when the house heating is turned down or off.
So I decided to make an enclosure!
But, I did not have the materials on hand, and visiting hardware stores is verbotten with lockdown rules.
So, don’t laugh. This is what I cobbled together……
A couple of cardboard cartons, an artist’s A0 paper case (SWMBO hasn’t noticed it missing yet), and a blanket.
The heated printer bed is the heat source, at 50ºc. And I was surprised at the temperatures reached inside the rickety construction.
The steep temperature rise on the left is inside the enclosure after printing started. As you can see, the temperature rose from about 18ºc (room temp), fairly quickly to over 30ºc. After midnight, when the house heating was turned off there was a slow drop to 25ºc, and then a further drop to 18ºc when the printing finished and the bed self turned off.
The temperatures were measured with this gadget. A temperature/humidity logger.
And the printing result??
This is the best quality print which I have had since the onset of winter weather. It is solid, water tight, and a reasonable finish. 0.2mm layer height. It is a molten metal pouring funnel, so I was not trying to get a super smooth finish, just an intact water tight object.
As soon as I can get access to Bunnings, I will make a more purposeful enclosure. Meanwhile, the cartons and blankets can remain in use.
I have given some thought to how to manage the bronze pour for the barrel of the Ottoman bombard. It will be at the size limit of my potter’s oven for the PLA burnout.
The red 3D printed PLA is the barrel. The breech, although significantly shorter, weighs almost exactly the same, but being shorter, should be less problematic. The wall thickness of the breech is greater than the barrel.
I had thought that the steel cylinder would be adequately long to cast the barrel, but it is about 50mm too short when I take into account the bronze feeder reservoir which will be required. So I will add a 50mm length, probably by arc welding another bit of tube to one end. It wont matter if it is not a perfect join. I will make it waterproof with duct tape. The tape will burn off during the investment melting/burnout.
I will cast the barrel with the threaded end downmost. The molten bronze feeder reservoir will be 60mm deep which I hope will provide adequate pressure and extra molten bronze if required during cooling contraction.
I decided that the usual rubber pouring funnel (pictured above at the bottom of the steel cylinder) would not have an adequately large bronze opening or reservoir depth, so have 3D printed one in PLA. The PLA will disappear during burnout, but will leave its shape in the investment medium and become the funnel and reservoir.
Hard to get your head around that one? It certainly was for me when designing it.
You (and I) need to remember that any space around the PLA will become investment medium. The PLA will disappear and become a void which will be filled with bronze.
The 3D print took over 8 hours. 0.2mm layers, 210ºc extruder temperature, 3000mm/min. I will need to do a similar 3D print for the breech. If either or both pours fail the whole process will need to be repeated.
Still in lockdown. Cannot visit my workshop due to the 5km travel limit. So 3D designing and printing at home is fairly good use of my time.
I have been attempting to print a 1:10 scale barrel of the Ottoman bombard, in PLA, so I can make a cast in jeweller’s investment, and use that to pour a bronze version of the cannon.
I borrowed a big furnace to melt the bronze, and broke 2 ribs unloading it from my vehicle. That was about a month ago. They still ache a bit, but apart from careful positioning in bed, are steadily mending. I have to sleep on my back, which would normally make my snoring unbearable, but the CPAP machine is working quite well. SWMBO absolutely insists that it is in constant use.
And I have purchased a length of 5″ stainless steel pipe to make the mold.
I have featured the Ottoman bombard in previous posts, having made a wooden version some years ago. It is over 500mm long, and 107mm diameter. In 2 pieces with a big thread joining the pieces.
I can’t really justify a bronze version. It will weigh close to 20kg. But it is a challenge. And I think that it will look more authentic in unpainted bronze.
I printed the breech part a few months ago.
3D PRINTING THE BARREL.
The barrel is 315mm long. My printer has a maximum print size of 300x300x400 mm so I was not anticipating any problems. I knew from the slicer program that it would take 2/3 of a 1kg reel of PLA, so I bought some new transparent PLA, thinking that it might melt/vapourise more completely in the burn out cycle of the production than the coloured PLA.
So I tried to print it. I have lost count of the number of unsuccessful attempts. Each time the print would start well, but at some point, sometimes after a whole day or 2 of printing, the print would come loose from the printer base and I would have clean up the mess of PLA spaghetti, and start again. I cleaned the printer base thoroughly. Scraped it. Wiped with acetone. Re-levelled it multiple times. But every time the print would break free.
I also noticed that I was getting a lot of stringing, and lumps of PLA would form on the printed surface, cool and harden, and sometimes the print nozzle would hit the hard lumps. That is when the print would loosen from the base and eventually break free.
I have been using a 3M printing cover over the aluminium printer base, quite successfully for over a year. Maybe the cover had worn out. I looked up the P.I. for the cover, and yes, it has a stated expected life of 10 uses! So that was likely the cause of the adherence problem because I must have used that cover at least 50 times!
I had no replacement 3M cover, so I reverted to the original cover supplied with the machine, which was boro-silicate glass. Initially it worked well, with good print adherence, but the hard lumps were still forming, and when the nozzle hit them, there was enough force to break the glass plate free.
What could be causing the hard lumps?
I watched multiple YouTube videos. Re-levelled the bed again. Checked every nut and bolt on the printer for tightness with no problem found. Checked the Z axis for level.
By this stage I was contemplating buying a new printer. Maybe one of those liquid + UV light jobs. But one of those big enough to make my barrel would cost thousands. So I got a quote from a professional printing service to print the barrel…. almost $AUD600. I would have done that, but the print is destroyed in the making of the cast, and it is possible that more than one attempt of bronze casting will be required. I was considering abandoning the entire project.
One last try at a print. I replaced the 3M cover with a new cover, and started a new print with a new reel of red PLA.
All seemed to be going well.
The print was adhering solidly to the new 3m cover. The hard lumps were still appearing, but the print head ploughed through them or knocked them off completely. The problem was, that after 3 days of printing, with 10% of the barrel still to go, the multiple jarrings were producing axis shifts. The appearance was pretty bad, but I figured that I could fix it with some extensive post printing hand finishing.
By this stage the print was almost 300mm high, and I could watch the laying of the PLA extrusion from the print head directly. In retrospect I should have used a mirror to do this at a much earlier stage.
What I saw explained the issue of the hard lumps appearing.
PLA was slowly oozing from around the base of the extruder nozzle. It was gradually building up into a pea size lump, and eventually falling off onto the print face!
So, I paused the print, picked off the accumulating lump, and watched some more. The same thing happened.
Why was the base of the nozzle leaking? Another pause. Checked the tightness of the nozzle. It was totally loose. About a full turn!
Tightened it up. Resumed printing.
The next layer did not adhere at all to the previous one, because tightening the nozzle had lifted it at least one mm.
I thought that I could start a new print of the final 10% of the barrel, and glue it to the part pictured, but when I examined it, the layers were poorly adherent, and falling apart. It went into the plastics bin. I expect that the loose nozzle caused multiple print faults in x, y, and z axes.
So, I am now 32 hours into the next attempt, with 47% completed.
Now that the Model Armstrong cannon is finished, I feel able to move onto some smaller projects which have been hanging around on my to-do list.
4-5 months ago I had these parts laser cut from 3mm and 4mm plate. GSMEE members have been making the Kant Twist clamps, and over the past 3 workshop sessions I have made a pair of the larger clamps.
The machining of pins and jaws was very basic so I did not record those processes. The laser cuts holes were accurate enough to be reamed to size. I found 2 annoying mistakes, neither of which was fatal. There is an extra hole in the small side arm. And the position of the pivot hole in one of the arms is about 1mm out. Not sure how that mistake crept in. Neither of these mistakes will affect the functionality of the clamps.
I would suggest one design change to the clamps, which I will apply to the small ones when I make them. I would add a small extension to the handle boss, say about 12mm, and knurl it. That would facilitate speedy changes to opening settings, before tightening with the handles.
Time and use will tell whether my choice of brass was sensible.
I know that these clamps can be purchased on Amazon and other sites, but this was a very satisfying project, and I have no regrets about deciding to invest the time to make them.
Small drill bits (up to 3.2mm diameter) are almost impossible to sharpen. Most of us just just buy new ones when our bits get dull. They are not expensive even in packs of 10.
But, sometimes we have parts which require sharp bits. (see recent post on installing model cannon sights). Even new bits are not necessarily correctly sharpened. I use Sutton drill bits which in small sizes cost ~$AUD18-20 for packs of 10 bits, but for crucial jobs I would like to touch up even those quality bits. The Sutton bits which I used for the cannon sights worked well, but the tense job made me very aware that in future I NEED to make sure that the bits are sharp.
So, I made a jig for sharpening small drill bits. The plans were published in Model Engineer 29 Dec 2000 and 26 Jan 2001.
It was a simple build. Took me 2 workshop sessions of about 2 hours each. I had the jig plate laser cut, very inexpensively from 2mm mild steel plate. Distributed to interested GSMEE members. The drill bit holder is an Asian copy of the English “Eclipse” pin chuck original. Also inexpensive. Came with 3 collets, to hold sizes up to 3.2 mm diameter. Cost $AUD10 inc postage. Note that the 1/4″ shaft actually measures 6.25mm diameter.
I used M2 and M3 metric fasteners in preference to the specified BA10 and BA8 fasteners.
The pin chuck should be through drilled in order to accomodate longer drill bits. The Asian pin chuck was not difficult to drill with a 3.5mm cobalt bit. It appeared to be case hardened only. And I used 2mm mild steel plate instead of the specified 16g brass plate for the jig. The wheel mounts were modified to cope with the different plate thickness.
The plans and instructions for use are in the articles in Model Engineer listed above.
p.s. GSMEE members who intend to use the laser cut plates. The drilling positions marked by the full thickness crosses can be successfully drilled to 1.6mm and 2.5mm by using sharp drill bits at high speeds- 3000rpm, slow feed rates, and cutting liquid. I used TapMagic.
And, it works! Here I am testing a 1.5mm bit which has been sharpened with the jig. Drilling through 4mm mild steel.
The sights were the final parts to be made for the model Armstrong RML.
There were reasons for delaying these items. They are tiny, easily dropped and lost, have tiny almost invisible details (to my eyes), and involve fine and very deep drilling into the barrel, on which many hours have previously been expended.
First I looked up every reference I could find about the full size originals. I could find no picture of the sights on the 80pr Armstrong, but I did find some diagrams of the sights on the Armstrong 64pr, on which the 80pr was based. Another problem was that there were rapid developments in sight technology, and I had to decide which period I would choose. The later periods (after 1880) had complexities which did not exist in 1860. In the end I just made decisions, knowing that they might not be exactly correct, but thinking that if further information surfaces I could make and install new sights.
The next step was to drill some 2mm and 3mm holes into the barrel.
First I milled 3mm flats. The first milling bit, solid carbide, just snapped as it bit into the barrel from the side. A HSS bit was more long lived.
Next, a 2mm hole was drilled right through the barrel, missing the bore, and exiting through the bronze bracket which supports the elevation quadrant gear. At 40mm deep that hole qualifies as deep drilling. Tension drilling.
Most hobby modellers who use a CNC lathe, run the lathe with Mach3 software. Some have progressed to Mach4, but apparently 4 is not straight forward, and my expert advisor says that it is best to stick with Mach3 for the time being.
Mach3/4 requires instructions to control movements of the lathe carriage (X and Z movements), spindle on and off, spindle speed, coolant on and off etc, and those instructions are in the form of G codes.
G coding is not difficult to learn. There are excellent YouTube tutorial videos on the subject, Udemy courses, books, and so on.
The problem is that mistakes in G coding can be easily made, costly, and sometimes scary. The biggest problem is that G coding is time consuming.
Fortunately, software is available to make G coding automatic, fast, and reliable. This post is about the program which I use for G coding. It is called Ezilathe. It is available at no cost as a download from CNCZone. (see details of downloading later in this post).
Ezilathe was written by Stuart Tankard, who I met a decade or so ago at a meeting of GSMEE. Stuart gradually persuaded me to get into CNC, and frankly, the decision to do so has resulted in a quantum leap in the scope and standards of my model engineering efforts. Now I use a CNC lathe, CNC mill, CNC rotary table, CNC 3D printing.
But I digress. This post is about G coding for CNC lathe. Until Ezilathe came along, I was programming G codes manually, making mistakes, and consuming lots of time. And experiencing crashes. Exciting, damaging, and sometimes dangerous.
To CNC turn an object, the profile must first be drawn with a CAD program such as AutoCAD. There is a very basic CAD drawing facility in Ezilathe, but it is so basic that I prefer AutoCAD. In AutoCAD I then rotate and move the object as it would be held in the lathe chuck, and move the tailstock end of the profile to the X=0, Y=0 position. Only half of the original drawing is required, so I delete everything below the X 0 line. I delete or hide all lines which are not part of the profile. The profile can be saved as a polyline, or as separate lines, arcs, splines etc. It is named and saved as a dxf file.
Then Ezilathe is opened, and the dxf file is imported.
The stock diameter is entered. The z position of the tailstock end of the profile is entered (usually Z=0). Feeds and speeds are entered using the tables in Ezilathe or another source. I use an app named “FS Pro”. Also select which lathe tool is to be used (the lathe tools are all entered in advance) or select a “point tool” which has an infinitely sharp point, and no width. In the example shown in the pictures below the turning was achieved with one tool. If multiple tools are required on one job, each tool will have to been set in the tool editor. So far I have not used the tool editor, but I can vouch that it works well, having seen the superb results which have been achieved by Stuart.(see the photos below)
Then the polyline of the profile, OR, each line and arc etc is selected in the machining sequence. Important not to miss any small lines or arcs as unintended results can occur. That is why it is sensible to save the original drawing as a single polyline. (Stuart tells me that any gaps will be automatically filled by Ezilathe as straight lines.)
Then go to the simulator, and see how Ezilathe will manage the sequence of cuts. In the picture below, the cuts are the straight lines, and there are 2 finishing cuts along the profile.
If all looks well, generate the G code. I usually visually scan the G code, looking for obvious errors. Usually there are no errors. Save the G code. Then ready to use the G code to make the part.
AN air cut with no work stock or cutting tool is sensible for beginners.
There is a lot more to Ezilathe. This has been a very brief introductory summary. Creating a lathe tool library, and entering startup strings for different lathe setups for example.
To download Ezilathe, you need to register at CNCZone. It is quite a good, useful site. In CNCZone go to downloads, Postfiles, Page 2. Select Ezilathe, and Ezilathe.pdf. The pdf is a comprehensive manual. Save them to a directory named C:\Ezilathe and unzip them. The program has been fine tuned, and updated. The latest version is 1.7.3. Then, on page 1 of Postfiles there is a small bug fix, Version 18.104.22.168. 22.214.171.124 is an executable which should be run after 1.7.3 is installed. If there are problems or questions Stuart can be contacted via CNCZone.
I have no hesitation in recommending this excellent program. It is just amazing that it is free.
Well, that heading is bit misleading. I had no pre-existing plans. I drew plans up myself, using the original cannon to measure, scale down and get my model dimensions.
In my posts over the past 18 months you have seen plenty of pictures of the original cannons, and the models.
I decided to model the particular cannon because I liked the shape, the degree of mechanical complexity with gear trains, riveting, etc, and because of the association of the town (Port Fairy) with my son in law. Also, because Port Fairy is within driving distance, if further dimensions and details should be required, which they were.
I thought that a model would look interesting and impressive. And it could be built in 3 distinct stages. Chassis-slide, carriage, and barrel.
So, in the order in which things happened……. (this info is all in older posts, so turn off now if the repetition is boring. I am writing this post for an overseas reader who wants to know how it was all done.)
I used a digital camera and iPhone to take photos of the entire cannon and its components. Close up, and from a distance showing the relations. The straight lateral, front and rear photos were particularly useful. In total, about 200 pictures. The photos were printed A4, and catalogued. As the model building progressed I realised that I had missed some components, and 2 further photo sessions were required. That was complicated by the Covid restrictions, but managed with some delay. The best photos were in overcast weather conditions, with and without flash.
I measured as many dimensions as possible, using a builder’s tape measure. The dimensions were recorded on hand drawn sketches in a note book, and a phone app (<My Measures Pro>. Very useful). On a later trip I used digital callipers for some small details. The barrel was measured using circumferences, and calculating diameters from those.
Using AutoCAD, I drew up a simple side profile plan. In that process I realised that many more measurements were required, and repeat visits were made. Actual components were later modelled using 3D CAD and saving as STL files. The STL’s were used to make 3d prints for casting, and in some cases just to see how the part would fit into the cannon assembly.
I searched the Internet for any references to the particular cannon. In that process I discovered that the particular cannon was unusual, and it appeared that all 26 examples from the Royal Gun Factory had been sent to colonial Victoria. Identical cannons were to be found in 2 other towns close to Port Fairy, and in 2-3 locations elsewhere in the state of Victoria. The design (of the 80pr original) was very close to a smaller capacity cannon (64pr) of which examples are still located in Hong Kong, Singapore, UK and probably elsewhere. It gradually dawned on me that the differences between the 64pr and 80pr related only to the construction method of the barrel, and not its appearance. Google, Google Images, Wikipedia all had information. I even discovered some images of simple plans with dimensions of the 64pr cannon, which corresponded very closely to the dimensions which I had obtained from the Port Fairy original.
As I posted progress of the model build on this blog, various readers offered further information, some of which was very useful. I have had some very satisfying subsequent correspondence with model cannon builders. In fact, the blog correspondence has led to some very useful experts who have been of immense help. One even supplied me with a copy of the original operating instructions for the Warrnambool 80pr cannons.
Searching for books on the subject of artillery in the 19th century has produced two which were published in 1877 and 1897. These are available free of charge as downloads, and bound reprints were purchased in addition. The cost of the bound books was very reasonable ($AUD20, and $AUD50). These have been thoroughly read, thumbed, dog eared, and used. So much so, that one of them is due for replacement.
Discussion with several experts has provided further great information, and resolved some of my questions. I had been mislead by some inaccurate restorations of the cannons at Warrnambool and Portland, and my suspicions about the restored details were confirmed in those discussions. The cannons at Port Fairy have had no restoration, which has resulted in severe rusting, but at least the remains are original and genuine.
Steps 1,2 and 3 occupied me on and off for about 5 days. 4,5,6 and 7 occurred progressively over the following 12-18 months during the model construction.
Regarding the accuracy of my scale models compared with the originals, I make the following observations.
As far as possible the linear dimensions are at a scale of 1:10. That was the scale of the first cannon I modelled, and I have stuck with that 1:10 for consistency, comparison between different cannon models, and simplicity of conversion. Of course that means that surface areas are 1:100, and volumes and weights are 1:1000. I have made some compromises to enable use of standard fasteners, shafts, and bevel gears, but have kept the compromises as small as possible consistent with the visual appearance.
Materials. The original cannon was constructed of wrought iron, cast iron, bronze/brass and wood. Steel was used only for the top surface of the slide. I used mild steel for the slide, the chassis, the carriage and the barrel. I used bronze or brass where the original had bronze. The wheel assemblies were cast iron on the original. I was restricted to cast aluminium and later cast bronze, but since these components were painted, the exact metal is not detectable, and probably irrelevant. The original wooden chassis/slide was made in the UK. The later iron chassis/slide was ,made in the UK but modified in Melbourne. That is the slide which I modelled. I imagine the original wood components to have been oak, but with no certainty. Scaling wood is difficult, due to the problems of scaling the figuring. I have some European oak offcuts, but decided to use Australian jarrah, because of its fine grain and attractive reddish colour.
Colours. I could find no reference anywhere regarding original colours. The experts had no hard information about colours and neither of the books mentioned colours. Some opinions were that black was most likely, and for cooler climates that seems reasonable. But in a hot Australian summer, black iron objects can be too hot to touch. I therefore suspect that white or at least a lighter colour would have been most likely used. But, you know what? SWMBO offered the most sensible advice. She said that if the model is to be on display it has to look attractive. She recommended painting the boring bits flat black (carriage sides, recoil tube, wheel assemblies), and leave the remaining metal uncoloured. Even the copper rivets, which would have been iron in the original. And to polish the bronze and brass components. So that is what I have done.
The base. A lot of thought went into this. The original ran on circular iron tracks which were set into concrete slabs. I rejected a circular track on the grounds that the model would have been too wide for convenient storage and display. I also rejected concrete or model concrete in favour of the eventual choice. That decision, (incidentally, not “set in concrete”…. might be changed at a later stage), was for a single piece of 12mm thick gloss black cast acrylic Lexan, sandwiched to a piece of thick plywood which is painted flat black. It looks good IMO, is hard wearing, and reflects the internal components so they can be seen. The first model Armstrong which I made was set on a wooden base with oil polished top and black painted sides. That also looks quite good. (IMO).
I might add to this list if further memories surface of the various processes used in the planning stages.
Just one picture for those readers who never read the text….
I made them promise to not make me laugh or cough or do any heavy lifting. And it was a pleasure to attend another face to face GSMEE meeting after a few weeks of Zooming.
And Swen Pettig brought along his Grasshopper beam engine for a progress assessment.
Swen has also recently made this sphere making lathe attachment. The cutter is a 6mm diameter TC disk. He showed us an example of a job, which has a lovely finish and excellent shape, but a rather thick neck. Swen plans to make another cutting tool with a sharper point to make thinner necks on the spheres. Chatter has not been a problem with this tool.
I am not quite ready to show the Armstrong RML gun, because it is still waiting for me to get back into the workshop for some finishing touches, so I brought some goodies to show which I recently bought from Banggood.
And this one is a woodworking tool, which I have wanted for some time, but the US version is about x4 the cost , and postage is an extra deal killer. The costs mentioned all included postage.
Today I picked up the laser cut metal plate which will become the small drill sharpening jig. For drills bits up to 3mm. You might ask why bother? Well, I can see the jig being used where I want a really sharp tiny drill, on a critical job. I would touch up the cutting edges on even a new bit.
The drill bits to be resharpened will be held in a small collet chuck, purchased commercially for about $10, including 3 collets. The lasering cost about $5 each.
Several weeks ago one of my grandsons broke his wrist. In a discussion about the accident with SWMBO I mentioned that I had never experienced a bone fracture. (mistake #1).
Some days after that I collected a metal melting furnace from Stuart Tankard. A loan, so I could melt enough bronze to cast the Ottoman bombard. My own furnace has a capacity of 3kg of bronze, but I needed to melt about 10kg for each of the two parts of the cannon. Stuart’s furnace is gas fired and will take much larger crucibles than mine. He readily agreed to the loan of the furnace, but I noted that he hesitated when I rang to arrange the pickup. I discovered why when we loaded it.
It is bloody heavy!! The two of us could not lift it more than a few centimeters. It must weigh around 200kg.
So, we manoeuvred it onto a hydraulic lift trolley, and used that that to roll it to my Landcruiser, jacked it up, and slowly and progressively moved it into the back compartment of the cruiser. I assured Stuart that I would be OK unloading it. (error #2) I have an identical hydraulic lift trolley. And I was sure that I could carefully manoeuvre the furnace onto the trolley.
So, at my workshop I backed the cruiser to the concrete pad where the trolley was positioned, jacked up the trolley platform to the level of my cruiser and slowly pulled the furnace towards the trolley. But it would not budge.
The furnace has a handle for rolling it around. The handle slips into position quite neatly. So I climbed onto the cruiser floor, and started pulling and manoeuvring (error #3). Suddenly, the handle came free, and I was flying through the air tumbling backwards. Fortunately the furnace stayed where it sat.
But I hit the concrete pad.
With my chest.
Initially I was winded. And my chest hurt like hell. And my head hurt. When I stood, I realised that I was conscious, bit giddy, and in PAIN from my chest. Oh shit. I knew that I had probably broken a rib(s). It hurt to breath. I took shallow breaths. No bleeding, but I was sure that there would be some impressive bruises.
As was alone, with no-one nearby, I rang my wife. By then I had regained my breath and I reassured her that I was not seriously hurt, and did not need an ambulance. But if I did not turn up at home that evening maybe she should come and find me. (sensible#)
As time passed, the pain did not seem too bad, so I had a cup of coffee, some Aspirin, finished unloading the furnace, and a couple of hours later drove home. I took some more aspirin, and started cooking dinner as usual. My wife said that I should see a doctor, but hey, I used to be a doctor, and I knew that nothing serious had happened. (error #3). Also it was a public holiday. My doctor would have either been on call, or arranged suitable cover. But I figured that it could wait until the next day.
After an uncomfortable, restless night, the next morning I felt sore, but not in severe pain, and it settled with some more aspirin. We were child minding that day, so seeing the doctor would have been very inconvenient. So I did the child minding with my wife, with no great problems. Then another restless night. I planned to see my GP the next day. Or at least to talk to him with a phone consultation.
Next morning, I sneezed. Hay fever. The chest pain was suddenly much worse. Aspirin hardly touched it. Some movements were very painful. I was terrified that I would sneeze again. Deep breaths were out of the question. I rang for an urgent appointment with my GP. It was made for late that afternoon. I knew that I needed an X-ray of my ribs, so I sent an email to my doctor. It is almost impossible to get to talk directly to a doctor without a prior appointment but I thought that an email from an ex-doctor might might get through. And it did. I had a phone call from him within minutes, and the X-ray was arranged. He also arranged an electronic prescription for a stronger pain killer, which my wife collected. It helped.
I had the X-ray, driven there by my wife, who by this stage was very concerned.
At the pre-arranged consultation the GP examined me. Clinically he agreed that at least one rib was fractured. However the X-ray report said that there was no fracture. He invited me to view the X-rays, and I was certain that I could see a fracture. I was no radiologist but I had seen a lot of X-rays in my medical career. A second radiologist opinion would be obtained. (the second opinion confirmed the fracture). My lungs were intact at that time.
I did not require hospital admission. I was given a prescription for more pain killers, instructions about keeping active, taking laxatives (to avoid the analgesic induced constipation), and to stay in touch for the 6 or so weeks that healing would require.
The next day the situation changed. Despite the strong pain killers, the pain progressively worsened. By evening, when SWMBO arrived home I was having frequent episodes of agony in the area of the fracture(s). I discovered the meaning of the phrase “pain bringing someone (me) to his knees”. It was worse than the root canal pain, the ski accident multiple ruptured ligaments and crushed bone ends pain, you get the picture. I was almost fainting with the pain. SWBMO wanted to get an ambulance. I did not want the delay. So she drove me to the hospital casualty. Every bump in the road provoked the agony.
At the hospital, I staggered into the casualty department while SWMBO parked the car. There was a queue at the triage assessment. I had put on a mask as per Covid regulations. I was doubled over with pain, and wondering if I would faint if a severe wave occurred. I was aware that I was groaning with the pain. Waiting people were staring. But the triage nurse was not looking. So I walked to the window and said “I am in severe pain”. She said “sit there”. I said “I cannot sit down”. She pointed to an adjacent room and said “wait in there”. So I did, while she continued getting details from the two people ahead of me in the queue. I was bothered by the attitude, but in no condition to argue. I guess that she was following protocol. I wonder what protocol would have required if I had collapsed.
Eventually, after an interminable wait (probably about 5 minutes) I was motioned to window, gave my details, and I was walked with assistance from 2 orderlies, into a cubicle. Things moved quickly then. An IV line was inserted, and I was given IV Morphine. A few seconds later the pain was almost gone. It was heaven.
A CT scan confirmed the fractures. And partial lung collapse. My liver was intact. There was no internal bleeding. Hospital admission was recommended, and SWMBO insisted that I accept, although after the Morphine I was tempted to go home. Anyway, I stayed overnight, with strong oral painkillers every 3 hours.
I was discharged home the next day with a supply of really strong painkillers, and arrangements for follow up. Since then, I have been slowly improving. I have a strange “clicking” sensation at the site of the fracture. Something is moving. Yesterday I visited the workshop, but not really up to anything interesting. Doing a lot of dozing.