Artisanal Gunmaking, No compromises, No corners cut, EVER

Wednesday, April 15, 2020

A Hoffman Arms Shotgun

Here's a gun that definitely doesn't come across the bench every day.  Many are familiar with Hoffman Arms Company for their custom bolt-action rifles built on Springfield '03 and Mauser 98 actions but they did also offer shotguns.  The most commonly known Hoffman shotguns were their side-by-sides (in both boxlock and sidelock form) and trap singles, which were most likely bought-in from England in unfinished form.  This one however proves that Frank Hoffman wasn't afraid to try something different.  It's a 20 gauge over/under and, having had it apart, it seems to be one hundred percent American.  The action is unlike anything I've seen before, nothing revolutionary or even avant garde but definitely not a copy of anything else.  Its monobloc barrels use split lumps that (actually) bear in the frame sides, a mid-level bolt, A&D style cocking, Southgate style ejectors and a Parker style doll's head rib extension.  The frame is quite heavy, with the massive width being necessary to house the cocking levers on either side of the lower barrel.  Imagine a Westley Richards Ovundo with fixed locks and a Parker doll's head and you'd be fairly close.  The workmanship is exactly what one would expect from a pre-war American custom gunmaker, which is neither complimentary nor derogatory, just an accurate assessment.  The serial number is a low single digit and if their are any others of its kind, I would doubt that they are identical in very many respects.

"That's all well and good but why is it here?" you say.  Well that's a tale of a manufacturing defect and a "local gunsmith's" over-estimation of his ability.  I'll cover the manufacturing defect first, then deal with Bubba McHack's "repair" work.

The ejectors work on the Southgate over-center principle which involves two moving parts, the tumbler and the spring, and a more elegantly simple system for heaving empty shells over the shooter's shoulder has never been devised.  As simple as the Southgate mechanism is in "firing", it still needs a way to recock the tumblers when the gun is closed.  Some guns use the ejector segments themselves to perform this task, pushing the tumblers back into position as the ejector segments are forced forward by the breechface as the gun is closed.  Others use a separate cocking lever, employing mechanical advantage through the miracle of leverage, to rotate the tumblers as the gun is closed.

Hoffman's design is, at least conceptually, a combination of the two.  There is a plunger situated in the forend iron (one on each side) that pushes the tumbler into the cocked position as the action is closed.  The plunger is situated well above the tumbler pivot and works against the frame just above the action knuckle, so that when it makes contact with the frame as the gun is closed it travels axially in its bore, pushing the tumbler into the cocked position.  It's an ingenious design that combines the simplicity (almost) of no cocking system, with the mechanical advantage of the cocking lever system.  It is a design that works very well, so long as the cocking plungers fit their bores, and that's where the problems begin.  In this gun the plungers were .020" to .025" smaller than their bores in the forend iron, which were .187" diameter.  This amount of play allowed the plungers to cock in their bores and bind, rather than smoothly travel for and aft.  In fact they would bind so badly that even with the ejector tumblers and springs removed, they would hinder the action's closing.  Since the leverage that would normally be employed to rotate the tumblers was now directed against the unsupported vertical portions of the forend iron, something was going to give way.  That something was the right side of the forend iron.

Enter Bubba, the local gunsmiff, who clearly figgered "no problem, I'll just weld 'er up".  Now, having seen this weld (and so will you, shortly), two things become readily apparent:  The first is that he did not anneal the case-hardened forend iron, thus dooming the weld to failure even if he knew how to weld.  The second is that he clearly doesn't know how to weld. At all.

Apparently, showcasing his talents with a TIG torch wasn't enough, because he also exercised his woodworking skillz in repairing the cracked forend wood.  For this trick, he used what is hands down, the absolute worst wood-repair product ever hawked to the trade.  I speak of course of Micro-Bed, a single-component, air-drying "bedding compound" that when dry has the consistency of rubber.  Not old-school Colt handgun grip "hard rubber" mind you, more like bicycle tire soft rubber.  The only thing positive that can be said about it is that it sticks really well, but that's not such a positive when you have to remove it.

Needless to say, this repair was not long-lived, not only due to the "quality" of the weld repair but because the root cause of the break was never addressed.  Since the existing forend iron was beyond salvage, the only course of action was to make a new one, which sounds really easy when you say it fast but there are a number of factors that conspire to complicate the job.  The first is that the new iron must be made to fit the already existing forend wood, left and right ejector mechanisms and frame contours.  The second is that the original forend iron was made in two pieces (shoe and leg) welded together.  The original weld is plainly evident and is one of many clues that this gun is, if not one-of-a kind, still entirely hand-made.  With only one option available to me, I forged ahead.  The first thing was to determine the radius of the action knuckle and the lower barrel channel.  With those numbers and a chunk of 1018, I started blanking out the new forend iron shoe.

Another clue that the gun is hand-made is the fact that NOTHING is symmetrical on either side of the gun's centerline.  Nothing, not the position (or height) of the cocking lever slots, not the external contours, not even the height of the forend iron (one side of the frame is lower than the other).  The fun never ends.

Once the forend iron was blanked out it is welded to the original leg.  The weld was filed flush with a generous radius at the transition from shoe to leg (to minimize stress concentrations).  At that point I could actually fit the iron to the knuckle and the forend lug on the barrel.  Once that was done, I had to adjust each cocking slot so that both hammers reached full-cock at the same time and a few degrees before the barrels reach their stop because those last degrees of travel are reserved for the ejection cycle.

Now that we're back to square one (it should be clear by now that a "cheap" gunsmith really isn't cheap) I can solve the root cause of all of this grief, the cocking plungers.  They are machined and filed from O1 and each side is individually fitted before heat treating.

Both cocking plungers are fitted, so now it's time to fit and time the individual ejector mechanisms so that they trip only after the hammers are cocked and just before the barrels reach their mechanical stop.

The forend iron is off to Geoffroy Gournet for engraving and when it returns, I will color caseharden it and then artificially age the external surfaces to match the frame.  In the interim, I'll work on correcting the poor repairs to the forend wood. 


Forend iron back from Geoffroy Gournet, engraved exactly as the original.

I've case hardened the forend iron and will artificially "wear" the externally visible portions to match the frame.  Next is the inevitable refitting (hard fitting) of the iron to the action and final assembly.
The finished forend assembly.

Saturday, April 11, 2020

A Somewhat Unusual Stephen Grant

Here we have one that doesn't show up every day.  It's a Stephen Grant side pedal gun with lockwork designed and made to the Grant and Adams Patent #2101 of 1883.  In operation, both tumblers cock on opening of the barrels but what is different is that only the right mainspring compresses.  The left mainspring compresses upon closing the barrels, thus making the opening and closing effort exactly the same.  It is commonly believed that this is a self-opening or an easy opening design.  It is most assuredly NOT a self-opener and, since the opening effort is the same as the closing effort, it's not an "easy-opener".  I won't even get into the fact that there is no such thing as an "easy-opener".  This is not the only gun to employ the concept of equal effort for opening and closing, the first Holland & Holland Royal (Holland and Robertson's patent #23, of the same year) was designed to operate in such a way, though it was mechanically quite different and cocked one tumbler on opening, while the opposite tumbler cocked on closing the barrels.  The effect was the same.

Those pads in the action bar that look like the "lifters" in a Beesley-type gun are not actually lifters.  The engage the eccentrics at the front of the mainspring housings on each lock.  Also note that there are no interceptors.  The locks were sourced from Stanton's.

Friday, March 27, 2020

Why ???

Here's what happens when you mix laziness, lack of imagination and a Dremel tool.  Rather than properly polish and spot (it is NOT jeweling or engine turning) the bolt body, our "big name" "gunmaker" did this.  Perhaps the randomness of this "decoration" is supposed to be the Yin counterpoint to the Yang of a bolt action's precision?  Nah, it's just lazy hackery that didn't take long to do, compared to just about any possible alternative.  Did he actually think that this looked good?  Perhaps.  Did he think that the customer would think it looked good?  I don't know, but our "custom gunmaker" was proud enough of it to put his name on the barrel.  Think about that for a minute.  What a hideous mess.

Even the locking lugs got the decorative treatment.

Tuesday, March 24, 2020

Spring Has Sprung

So it's as good a time as any to discuss those flexible, boingey bits, without which guns couldn't function.  I'll cover each of the different types of spring most commonly found in various gun designs, the materials that they're made from, their design and application.  I will not get into any of the non-ferrous alloys such as stainless steels, copper alloys or brass alloys since they, for the most part, are uncommon in gun mechanisms.  For the same reason, there will be no discussion of the more "exotic" materials from which springs have been and are made such as wood and fiberglass.  Belleville and air springs will also not be discussed, since their appearance in firearms is quite rare.  The explanations herein will be very simplified for the sake of brevity and to prevent any deaths due to terminal boredom.

Springs in gun mechanisms serve to return a part or parts to their original position by releasing the potential energy that is stored within the spring, when it is deflected by the movement of the part that it drives.  In simple terms, this happens because the amount of deflection in the spring remains well within the elastic range of the spring's material.  This range will vary based on such variables as the material from which the spring is made, the heat treatment of the material and the actual design of the spring.  Depending upon those variables, if the spring were loaded beyond its elastic range, it may simply deform or it may actually fracture.  Most leaf and "V" springs in gun-specific applications are made from high-carbon plain steels with a carbon content of 0.70 to 0.95 percent.  Coil springs in gun applications are generally made from music wire, which has a carbon content of 0.80 to 0.95 percent and very high tensile strength, which comes from the cold drawing process by which music wire is formed.

Generally, the springs found in firearms will experience either of two types of stress:

bending (as a beam)

or torsional (twisting).

Which type is not always intuitively obvious which you will soon see.

The term "spring rate" will also come up.  What this means is that for X amount of force applied, the spring will deflect X amount.  For example, assume a straight-rate coil spring of 10 pounds per inch.  What this means is that 10 pounds of force applied to the spring will cause it to deflect 1 inch, 20 pounds will be 2 inches, 30 pounds will be 3 inches and so on in that fashion.  That would be what is referred to as a "straight-rate" spring, easily recognized by its equally spaced coils.  There are also what are referred to as "variable-rate" springs, easily identified by their more closely spaced coils at one end.  In a variable-rate spring, the more closely spaced coils make up the softer portion of the spring and when those closely spaced coils stack solid, then the rest of the spring comes into play.  The reason that the area of more closely spaced coils is "softer" (it has a lower rate) is because that area of the spring is actually longer, which is why there are more coils and why they must be more closely spaced.  Variable-rate springs are rarely found factory installed.  They are common in the aftermarket as recoil springs for semiauto pistols.  Their value in these applications is dubious.

You may have noticed that dashed line in the center of the beam in the first drawing.  It is what is known as the neutral axis, which is the cross-sectional point where the opposing forces of bending (or twisting, in a torsion bar) change direction and cancel each other.  In the case of the beam in the following drawing, the upper surface is experiencing compression and the lower surface is experiencing tension.  Since these forces cancel each other out at the NA, it should be obvious that they are greatest at the outer surfaces and diminish toward the NA.  The NA is not always at the centerline of the stressed portion of a beam or spring.  It will coincide with the centerline only in a beam of symmetrical cross-section.  Calculating the NA for a non-symmetrical part is a bit beyond the scope of a blog post but it generally moves toward the more massive side.  It should also be obvious, now that you know this, why tool marks (especially across the surface) would shorten the lifespan of a leaf spring.

Now, on to different types of springs:

The torsion bar, not to be confused with a torsion spring:  A torsion bar is a spring that acts by being twisted about its long axis when a torque force is applied.  For a given material, diameter and load, the longer the torsion bar, the lower its spring rate will be.  This means that the longer it is, the more it will flex under a given load.  Torsion bars are not commonly encountered in firearms but understanding its function is essential because it is identical to its first cousin, the coil spring.

The coil spring:  The single most commonly used type of spring in all firearms is the coil spring.  They will be found in both compression and extension types.  It is commonly assumed that when a coil spring is compressed (or extended), that the coils bend, this is not the case.  When a coil spring is deflected, the entire spring (assuming a straight-rate) experiences torsional strain through the cross-section of the wire diameter, throughout its entire length.  This is why a coil spring gets shorter when compressed (or longer when extended).  Take a torsion bar and wind it perpendicular to its long axis into a spiral, you now have a coil spring.  As with the torsion bar, for a given diameter and material, the more coils, the lower the spring rate (it will be softer).  The coil spring has the possible advantage of still functioning even when broken, assuming it is sufficiently guided, either internally (with a guide rod) or externally (in a tunnel), and the broken ends do not "thread together" and induce coil bind sufficient to restrict movement of the mechanism.  Coil springs that have a uniform coil pitch are straight-rate springs while coil springs that have an unequal coil pitch are of variable rate.

The leaf spring:  This type of spring is commonly used as the mainspring (hammer spring) in many repeating rifles, single shot rifles and shotguns and a number of revolvers.  It is basically a deflecting beam.  It is invariably tapered as well, in either plan view or elevation, sometimes both, with the thickest point being at the anchor.  The reason for the taper is to load as much of the free length as uniformly as possible, without stress concentrations that would affect its rate and service life.

The "V" spring:  Maybe the second most common type of spring in the gun world and the single most common where my work is concerned.  In most (non-American) doubles, the hammers, ejectors and toplevers are powered by springs of this type.  It's basically two opposing leaf springs, joined at the fat end.  The limbs always work opposite each other.

The final commonly encountered spring type that I'll touch on here is the torsion spring.  Contrary to its name (and non-intuitively), it does not work in torsion at all.  Its name is derived from the way in which it drives whatever part it is driving.  That is, around a shaft or pivot pin.  This type of spring experiences beam-type bending stresses, from one end, around all of its coils and all the way to its other end.  For a given wire diameter and number of coils, the longer the "legs", the lower the rate.  Similarly, for a given wire diameter and leg length, the more coils, the lower the rate.  This is because in both cases the spring is longer.

Now that the boring stuff is out of the way, let us debunk some more "gun-world lore" because, like it or not, the gun world is not exempt from the laws of physics.

Cutting springs to make them softer:  Shortening the length of any spring INCREASES its rate.  That means it is stiffer, NOT softer.  "Bullshit!" cries the kitchen table trigger jobber, "I cut the rebound spring in all of my Smith action jobs and it works."  No Cleetus, it doesn't.  What happens is that, by cutting the spring, the installed preload is less because of the spring's now-shorter free length.  It therefore exerts less pressure when installed and because the spring is now shorter, it also does not compress as far during its cycle.  The only thing you've lessened is the reliability of the trigger return.

If the logic of cutting springs were sound, the short piece of spring that was cut off would be much softer that the rest of the spring.  This is easily proven to be false:  Take an old 1911 recoil spring (or any coil spring) and cut 4 coils off of it.  Now try to compress those 4 coils.  Let me know how that works out for you.

Leaf vs. Coil Mainsprings:  If I had a buck for every word of flowery prose written by gunwriters who've never competently operated a screwdriver about the virtues of V springs over coils, I'd be pretty well off.  Apparently, V springs are "snappier" (whatever the hell that means), faster, stronger, less filling, taste better and are generally superior in every way to coil springs.  It is a bit hard to argue while being intellectually honest, since (with one exception) there are no guns that are identical in design except for their form of mainspring.  With so many differing designs and lockwork geometry encountered, an apples-to-apples comparison isn't really possible.  It's a bit like arguing that purple is better than Ferrari.  The above referenced exception to this case is The Perazzi MX-series over/under.
These guns utilize a detachable trigger group that is available in both V and coil spring form.  So in that case a direct comparison is possible. 

The mainspring's job is to accelerate the hammer with sufficient speed to deliver an impulse to the firing pin, sufficient to detonate the primer.  Ideally, this will be accomplished in the shortest possible span of time.

Let us assume that the mass of the hammers is the same for both spring types, that the energy stored in the deflected springs is the same and the hammer travel is also identical.  At that point the most important variable is the weight of the springs.  I don't mean the weight, as in the "pressure" that the springs exert, we've already assumed that that is identical.  I mean the actual physical weight of the parts themselves.  This is important because the spring, regardless of type, must not only accelerate the hammer but its own mass as well.

Obviously the spring that has more mass will be slower to accelerate, so the lighter one wins.  Is the coil spring and its guide rod (which must also be taken into account) lighter in weight than the v spring and its swivel link?  I don't know because I've never compared the weights of these parts but it seems reasonable to think so.  Let's assume just for the sake of argument that it is slightly lighter.  Can most people discern the microseconds of difference in lock time?  That seems unlikely since most shooters can't discern a 4 pound trigger pull from a 5 pound pull.  What about the much-touted "snappiness" of the v spring?  If the hammers in both cases are accelerated at a rate that differs in microseconds, how can the claim that one is "snappier" than the other stand?  The reasonable conclusion seems to be that the vast majority of shooters would never know which trigger group that they were shooting, if they didn't already know beforehand.  Now there are some firearms whose lock time is so glacially slow (Coltt SAA) that it is readily apparent to the shooter, Perazzi MX guns are not among them.

Then there are those who fetishize the v spring because they imagine Luigi the white-haired gunsmith, looking over his wire-rimmed glasses, hand filing and polishing their beloved v springs to perfection before heat treating them in the last remaining cup of whale oil on the planet and finally hand polishing them again before installation.  The reality is a bit different.  Perazzi wire EDMs the v springs by the hundreds, they are mechanically polished and heat treated in huge batches and then polished again, mechanically.  Each one is identical to the last and the next.   This is how Perazzi's v springs can be so supremely reliable.  If it's as reliable as a coil spring and offers little to no tangible advantage, then why would they offer it at all?  Well, for the same reason that they still make things like barbecue-flavored potato chips and tarted-up Toyotas called "Lexus": because people will buy them. 

My verdict:  Find something more important to argue about, like how many angels can dance on the head of a pin.

Friday, March 20, 2020

Further Adventures in WTF?-Land

It's bad enough to see the criminal mechanical mischief that passes for gunsmithing on an almost daily basis but when one sees something like this, especially considering that the replacement part is actually readily available, it makes me think about taking up another line of work.  Maybe something less stressful, like Komodo Dragon dentistry, or training Tiger sharks to eat out of my hand like my neighbor's Koi do.

This Parker had a failed sear spring.  The correct (and easy) method of repair would be to remove the remaining locating post from the frame,  fit a new spring, verify no interference of the new spring's free leg with the shoulder on the sear (and correct if needed), reassemble and go about the rest of your day.  That method was clearly too easy for this guy.  He drilled and tapped the frame to accept a screw and washer that held not one, but two pieces of leaf spring to drive the sear.  It's also readily apparent that these two pieces of spring are exactly that, pieces broken from already-existing springs.  Quite an interesting arrangement, with the first very light leaf stacking against a much stiffer "helper" spring and then both coming into contact with the remains of the original spring.  Brilliant.  Why didn't the factory think of this?  To add insult to injury, he also literally knife-edged the sear noses, which I can only surmise was an attempt at a "trigger job".