The Ship Model Forum

This is a photo of some Hotchkiss 6 pounder barrels machined about ten years ago by Steve Nuttall. Steve also took this picture. Scale is 1/64. The barrels were intended for a 1/64 model of a warship launched in 1890. The prototype guns were thus mounted early in the series of quick firing (QF) guns manufactured by Hotchkiss and various licensees. I have read that these Mark I Hotchkiss QF's had no recoil system. All the subsequent guns had a one or two cylinder recoil system.

Here is a photo showing two Hotchkiss QF's on the fore and afterdeck of gunboat. From a modeling point of view, the noticeable details include the long stemmed shoulder brace used to aim the gun, and the conical crinoline mount.

Here is a QF in a park. Notice the open, U-shaped "jaw" at the back of the breech. This is another very characteristic detail.

It occurred to me to model a Hotchkiss 6 pounder in Rhino3d. Most drawings I have found are a little sketchy. However there are some Hotchkiss patent drawings, done to the patent office standard, that are quite good.

Still sporadically playing with Bongo, which is Rhino3d's animation plug in. There are several features of the QF that might be animated with Bongo, including the breech block mechanism and the shell extractor.

These 'crinoline' mounts were meant to give the mount some elasticity in the absence of hydraulic recoil brakes. Hydraulic recoil brakes were introduced in the early 1870s, but initally only for heavier guns.

_________________
wefalck

Chairman Arbeitskreis historischer Schiffbau e.V. (German Association for Shipbuilding History)

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Thank you for your help wefalck.

Here are basic resources for the CAD model. Most of these drawings were extracted from U.S. patents granted to Benjamin Berkeley Hotchkiss. The most useful patent dates from 1885. The Hotchkiss patents can be found and downloaded using the Google patent search feature.

Here is a summary graphic showing several Hotchkiss QF's. The 6 pounder is also designated as the the 57 mm.

This is a more precise drawing of the 6 pounder.

From the 1885 patent, these are side and cross sectional views of the breech mechanisms.

And a top view I found on the net. This is conceptual, a sketch, but it helps in understanding how the breech block fit into the QF.

First step is to create a primitive, that is, a rough draft shape that can be "machined" in CAD. This can be done for the QF with a quick revolve of a closed curve based on the drawings above. Useful dimensions for scaling the drawings are 57 mm at the muzzle and 2480 mm in length. I set up the file to work in mm. Here, shown in red, is the initial curve:

The revolve axis, which is the barrel's center line is shown in black. The catalog drawing is shown for reference, but most of the curves came from the patent drawings. Here is the solid produced by a 360-degree twirl.

From the left, the first disk is stock that can be milled and cut away to produce the breech area. The middle disk is to be machined to produce the trunnion, and the front disk is raw material for the forward gun sight. I put the front disk on a distinct layer and in a separate file, and will leave it on hold until the rear sight has been put in place.

Here is the QF primitive in a perspective view.

All three disks are filleted. Probably the prototype gun jacket was turned and then milled. This produced subtle curves in these guns that would be difficult to model in any other way. The CAD idea is to mimic the original processes of turning and milling.

With a couple of cutplanes we can establish the shape of the breech. The first is a line drawn from a side view, then extruded in both directions to create a cutplane.
This cuts away most of the cylinder, and the waste can be deleted, as follows:The original 2D cutline is shown in red.
A pair of vertical cutplanes finish up the basic shape.
When the debris is deleted and a vertical hole is cut for the breech block, the breech looks like this.

In a subsequent post I will show how the hole was cut for the breech block. It is basically a Boolean subtraction. It is not quite sufficient to call this boxed part of the gun a breech. In his patent Hotchkiss (or rather, his lawyer) called it out as a "reinforced area." In naming layers in Rhino, I have labeled it as a "breech box." As a convenience I will keep using the term breech box here.

With a few more cuts and extrusions a trunnion emerges: Notice the subtle curves created at the intersections of the fillets and the milled surfaces. In Rhino3D, the jacket and the barrel are a single piece. In the manufactured QF's, they were two distinct pieces. The jacket was shrunk over the barrel.

Michael

From a 1908 card series on the Royal Navy.

The QF required a three-man crew. The 6-pounder version weighed more than 800 lbs, in the range of a Harley road bike. Even though it was balanced and mounted on a bearing, and even allowing for the leverage afforded by tiller arm and shoulder brace, it must have been a project to aim it. The man on the right with his back to the camera raised and lowered, with a double lever, an elevator type breech block. In the photo the block is fully down and the breech is open. The third sailor, facing the camera, loaded shells. After a shot the breech block was lowered very rapidly (it weighed 51 lbs). The falling weight drove a shell extractor, which automatically sent the empty shell straight back toward the gunner. A cusped plate deflected the spent shell so that it fell to the deck. The firing rate with a well coordinated crew was about 25 rounds per minute, hence the terms, Quick Fire, rapid fire or, in French, tir rapide.

Hotchkiss was a Connecticut yankee, but he emigrated to France after the Civil War. His main factory was in Paris.

On this linked site are several photos and images of Hotchkiss guns, including his hand cranked rotating cannons and, if you scroll down, several QF's. The author has done a number of beautiful Rhino3D/Bongo animations of MG's of the late 19th century.
http://www.victorianshipmodels.com/anti ... llery.html

Here is a closer look at the deflector for spent shells. Maybe it is shaped like a bicycle fender. The photo gives us a glimpse of the big double lever used to raise and lower the breech block.

Also notice the U-shaped "jaw" at the back of the breech. This is characteristic of all of the Hotchkiss QF's. Here is closer look at it:In these photos, the elevator type breech block is in or near its up position.

To make the U-shaped jaw in Rhino3D, the first step is to make a curved cutplane that mimics its shape. The curve can be found in the patent drawings SIDE view. I drew the curve, drew the "U", and then did a single rail loft using the U as the rail. Here is the cutplane in isolation:
The cutplane is then positioned in the gun, and used to carve away the back of the breech box:
Finally the cutplane is itself cut. The excess material is tossed. The "jaw" is merged with the structure of the breech.

Actually the term 'quick firing gun' originally did not relate to the rate of fire, but to the fact that it uses cased shells, rather than separate shells and charges as in earlier guns. At the beginning this was only possible for smaller calibres, which naturally had a higher rate of fire, having less moving masses. More correctly such guns should have been called 'quick loading', but the term 'quick firing' came into use and stuck.

_________________
wefalck

Chairman Arbeitskreis historischer Schiffbau e.V. (German Association for Shipbuilding History)

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Thank you for your insight. Makes sense. I greatly admire your amazing S.M.S. Wespe.

The breech block is the centerpiece of this machine. It is necessary to make first, in CAD, a sacrificial part, a faux breech block that is used to cut the hole and create the guides on which the real breech block will slide. The faux block is larger than the real block and extends all the way forward to the rear face of the barrel. The block primitive is made as a two-way extrusion of a line drawn on the SIDE view.

From the inner sides of the breech box, well aft, two slightly inclined guides protrude. We are working toward a Boolean subtraction of the entire faux block from the solid breech box. In order to create the two guides, it will be necessary to cut corresponding slotways in the faux breech block. The cutters are both subtracted simultaneously from the block. This is the result:Next step is to precisely position the faux breech block inside the solid -- but virtual of course -- breech box. Looks like this:In Rhino3D, you can choose to retain the faux breech block or simply sacrifice it. I let it go. Here is the result of the Boolean subtraction.

The guides are inclined slightly so that as the breech block elevators up into the breech box, its vertical front face advances toward the rear of the barrel. In this way a newly loaded shell is snugged into place.

Michael

At this point in the project it helps to have a shell. The idea is to check clearances at the face of the breech block as it ascends, and to get a sense of how the spent shell extractor worked. The shell does not appear in the 1885 patent but I found on the net a cutaway of a shell for a Mark I QF.

The extractor is essentially a hook that engages the shell base and first tugs and then jerks the shell out of the barrel. The hook mechanism was originally developed for the Hotchkiss Mountain Gun of 1875. There is a picture of it here:The hook is shown in some detail below the image of the breech. This early mountain gun was in many respects the design model for the QF. The main difference was that the mountain gun's breech block traversed the breech box from side to side, rather than shifting up and down as in the QF. The extractor mechanisms were nearly identical, but positioned differently. In the QF, the extractor guide moved in a slotway along the side of the breech box.Here is the extractor slotway cut into the 6-pounder QF. The adjacent photo shows a 3-pounder QF preserved in British Columbia.

Here is the breech of a 6-pounder QF in the Gosport museum. It shows pretty well how the extractor fit behind the rim of the shell. Also note the barely visible stud. In operation this stud is confined within a carefully shaped sigmoidal groove carved into the side of breech block. As the 51 lb breech block falls, the stud is driven sharply aft by the cam effect of the sigmoidal groove in the breech block. This propels the spent shell out of the gun.

The sharp edges and corners of this model have not been filleted. At this early stage, it is useful to leave sharp intersections that can be used as parking places for marker points, and sharp edges that can be extracted as helpful lines. For example, I used the lower right and left corners at the back of the breech box to help align these cross sectional drawings from the patent.

The main project now is the breech block.On the right side of the gun a crankshaft is sandwiched between the inner surface of the breech box (hatched) and the right side of the breech block. The crankshaft is essentially invisible -- it can be viewed only if the breech block is removed.

The right side of the breech block is grooved and relieved in two tiers. The deepest groove is traveled by the stud at the end of the crankshaft as the shaft is rotated.This shows the crankshaft with block in its full DOWN position. The breech is open. The image below shows the crank with the breech block in its full UP position, with the breech closed.The elaborate sculpting work on this right side of the breech block is necessary to create enough clearance for the crankshaft to swing. This is a photo of the right side of an antique Hotchkiss breech block from a smaller gun. The relieved patterns for crankshaft clearance are similar to those for the 6-pounder. The lever at the bottom of the block is a cocking lever.

It looks like a segment of the guide protruding from the inner surface of the breech box will have to be carved away to prevent interference with the crankshaft.

The crankshaft is turned by a double lever. The finished levers are shown here. The levers are inclined outward by about 77 degrees, so they trace a conical path as they are moved -- not a flat disk. Neither of the two sectional views show us a literal view of these levers. Both views are foreshortened because of the angled mounting or positioning of the levers. On a first try, I used the view shown in the longitudinal section, and set up a centerline and a curve for a rotation. These are shown in red.

At about this point I figured out that the club-like lever I was constructing would be too short. And not to scale. In an effort to get the handle sized correctly I drew its cone of rotation:In Rhino you can project a line onto a surface, but I have never quite mastered this feature. At this point I needed to project both the centerline and the curve of the handle onto the cone. To be sure of the result, I extruded the lines and then ran an Intersect.The Intersect command automatically creates a line at the intersection of two planes.From this point, it remained to Revolve the handle curve around its centerline curve 360 degrees.This is the finished handle modeled at its full length. Probably took the long way around to get it, but I think it is pretty close to scale. A simple inclined plane might be a better intersection surface than the cone, however, and I may try this as well.

The amount of information packed into these small, 130 year old patent drawings is amazing. But getting it from 2D to 3D is sometimes a puzzle. Michael

This mechanism for raising and lowering the breech block is similar to a Scotch yoke. It uses a crank and a slide to convert rotary motion to reciprocating motion. The system is used, for example, to raise and lower the needle of a sewing machine.

This Hotchkiss drive is novel, however, in that the slotway for the crank stud does not follow a straight line. Note in particular the curved part of the slot where the crank is centered in this illustration. The curve is actually concentric to the crank, so that when the crank starts moving, the breech block does not budge. The breech block does not move until the crank stud is turned far enough to the left to exit the concentric part of the arc. The block is then nudged free of the shell, and drops very quickly.

In the patent Hotchkiss calls this first part of the counterclockwise turn of the double lever "lost motion." To the man operating the double lever, it may have felt like slack in a steering wheel. Hotchkiss used cams to capture energy delivered during this small arc of idle motion -- to cock the gun. This was a patented feature. In other words, the firing pin was already cocked for the next shot before the spent shell from the last shot was ejected.The left side of the breech block is less complicated than the right. There are three slots milled into the surface of the block. These are cut with the familiar technique of Boolean cookie cutters.Here is result. The leftmost, sharply "Z" cut slot is the drive for the shell extractor. The rightmost slot is the slideway the block rides up and down in the breech box. The middle slot is fitted with a stop screw. Its purpose is to put limits on the up-and-down travel of the breech block. To remove the breech block, the stop screw would be backed out, and the block dropped out the bottom of the gun.I have not drawn the threads on the stop screw because threads tend to balloon the size of a CAD file. Threads are not visible when the breech block is mounted.From a modeling point of view, the stop screw is very helpful. Points set in the slot center at the upper and lower limits of travel give us a way to position the breech block quite precisely at the limits of its travel -- and to quickly and repeatably shift the block up and down. This is useful because the block can now be further machined "in place". This helps assure the perfect alignment of curves and surfaces of the breech box with those of the breech block.

With the block pinned in its full DOWN position, a U-shaped cutplane is used to machine a trough into the top of the block.The trough has tangency with the jaw. Some additional work is necessary at the front of the trough.This cutplane cuts a ramp into the front of the trough.Here is a Hotchkiss breech block manufactured in 1934.It still follows the same basic design, with a spooned area at the front of the trough. Why is it there?

As the breech block descends to eject the spent shell, the shell extractor is, slowly at first, prising the shell out of the barrel. The spooned area provides clearance -- room to move -- as the shell backs out of the barrel. This patent drawing from the 1880s shows the extractor mechanism from the top. This particular extractor uses a diamond shaped tenon (No. 17) to borrow energy from the falling breech block. Michael

The "spooned" area would also serve as a wedge to push a shell all the way into the chamber as the breech block rises.

Phil

_________________
A collision at sea will ruin your entire day. Aristotle

Phil, you are right, the ramp should work in both directions. There appear to be wear or brass scuffed patterns on the old breech block shown in the photo above. If we assume these were repetitive use patterns, they were probably made on the upstroke, at the conclusion of the loading sequence.
https://books.google.com/books?id=_zZOA ... ne&f=false
Here is a link to a contemporary encyclopedia entry that describes and compares various rival rapid fire guns. The Driggs Schroeder paragraph suggests that because its block rotated, it was less likely to "guillotine" a shell being loaded.

Less likely than what? Maybe the Hotchkiss, since the two were head-to-head rival designs. However, in Driggs' patents, he doesn't claim easier loading as an advantage, so it could just be a Driggs Schroeder marketing gambit, a talking point.

Michael, just a question to the practical modelling in Rhino3D: did you import scans of the patent drawings into Rhino3D as a basis for developing the 3D-models of the various parts ? I have done this in 2D, where I constructed parts over scans.

And, are you planning to have the gun arranged for 3D-printing ?

I personally don't have any experience with Rhino3D, just some basic AutoCad training and a lot of work with 2D-CAD programmes. However, I have been thinking of getting into 3D with a view of producing some printed parts, namely for a Hotchkiss 37 mm Revolving Cannon. The frame and the casing for the mechanism are complex castings that are not so easy to e.g. mill from the solid. I thought of having these pieces 'printed' in brass (i.e. printed in wax for lost-wax casting as offered by e.g. ShapeWays).

_________________
wefalck

Chairman Arbeitskreis historischer Schiffbau e.V. (German Association for Shipbuilding History)

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

U.S. Patents can be downloaded as .pdf files from Google's patent search service. I use the "snapshot" feature of the Adobe Reader to select and extract the drawings from the patents as .jpg files. The .jpg images are then loaded directly into Rhino using a Rhino command, PictureFrame.

The 6-pounder is destined for installation on a CAD model of an 1898 warship from the Spanish-American war. I need four 6-pounders. The beauty of CAD modeling is that having created one intricate assembly, such as a gun or a lifeboat, you can simply copy it multiple times.

The 37 mm Hotchkiss sounds like a fascinating project. 3D printing I have not tried. I have used FirstCut for CNC machining of (commercial) components designed in Rhino3D. I have also used Rhino to produce files for laser cutting wooden parts for models.

Rhino3D was originally conceived and marketed as a plug-in for AutoCAD. Later it became a standalone program. I was certified in AutoCAD before I tried Rhino for the first time, around 2008, and the transition was not difficult. I still use both programs but I have always been more comfortable working in 3D.

These days, I suspect the best way to try out 3D or to experiment with it is SketchUP. There is a free version available, and lots of online training support.

On the McNeel site I see there is classroom training available for Rhino3d in Paris (CADAtlantique). Once registered as a student I imagine you could get an interesting price on the Rhino software package. In the US the student price is a fraction of the commercial price.

Finally, you might wish to look into AutoDesk's Fusion 360. It offers CAM as well as 3D CAD. Some discussion here:
http://www.rcgroups.com/forums/showthre ... &t=2512531

Hotchkiss used a ribbon type mainspring. The hammer was mounted inside the breech block. This screen grab shows the mechanism in the context of his patent drawing:Below is the device in isolation, shown cocked:And shown again with the spring expanded.His patent argues that because the spring simultaneously pushes up and pulls down, it produces less or perhaps more symmetrical wear on the mechanism. Another possibiity might be that he was "improving" his way around some rival patent. In any event, the ribbon spring is typical of the Hotchkiss series of QF's. Here is a photo taken long ago of a dismounted spring.Here is another view in a photo shot from underneath the breech block.Immediately to the right of the mainspring is a ratchet and pawl that operates on the cocking shaft and securely cocks the mainspring.Here is a view from below showing the centered hammer and mainspring and, offset to the right, the sear. A linear spring bears upward on the sear. The left end of the spring is solidly anchored in the wall of the breech block.

In this view the cocking shaft is rotated counter-clockwise until the pawl on the spring-loaded sear catches the ratchet. The mechanism is shown cocked.

Finally, shown below is an X-ray view of the hammer and sear mechanism positioned inside the breech block. At this point they are simply embedded in the block. Next step will be to machine away a cavity and mounts for this mechanism. Benjamin Hotchkiss's obituary noted that he owned the Empire Spring company in Saratoga, New York. Thinking this might be a spring manufacturing company -- that he had perhaps bought in one of his suppliers -- I googled it. It turns out the Empire Spring Company was in fact a hugely successful producer of spring water, like Evian or San Pellegrino. Full of surprises, his story.

The two main remaining cuts to the breech block are indicated here.An angled, U-shaped cutplane was used to make the 46-degree slanted cut. The Boolean cookie cutter in the middle of the block is designed to open up a slot for the hammer. The cutter, which is in effect a mould to the inside of the slot, is here: Shown below as a test fit is the hammer in its slot.After some additional detailing, the completed breech block looks like this. Notice the cutouts at the rear of the block. These secure the springs and provide an exit port for the sear.
The assembled breech block is shown below. Notice the firing pin projecting through the front surface.When this assembly is turned around it is possible to notice the wide slot that secures the mainspring, and the sear projecting from a tiny window at the back of the block. When the sear is pressed down by the trigger arm, the cocked hammer is released. The mainspring was left free to move back and forth a few millimeters in its slot as the hammer was cocked and released.