Forming .300 Blackout in Quantity — Necks Made From Body Brass

Cutting a .223 case at about 1.37 inch means the new neck is formed from brass that used to be case body, and body walls are thicker than neck walls. Everything else in this volume follows from that sentence. It is the mechanical fact underneath the neck-thickness argument, the donut, the neck-turning dispute, and most of the quality difference between a formed case and a purpose-made one — and it is almost never stated plainly in the published workflows, which tend to give the step and not the consequence.

This is the production-detail companion to Volume 16. That volume owns the concept: what forming changes and what it cannot, the nine-step sequence, the difference between a form die and a sizing die, the contested finished-length figures and why neither is printed, the missing yield rate, and the .300 Blackout-in-a-5.56-chamber hazard in full. None of that is repeated here. What follows is the layer below it — parent-brass selection as a quality decision rather than a sourcing one, what the forming operation does to neck wall thickness and where it puts the surplus, the consumable that actually drives the per-case cost, and the one tool in the sequence that exists specifically to make length a fixture dimension instead of a measured one.

25.1 Where the New Neck Comes From

A .223 Remington case is 1.760 inch at maximum. Cutting at roughly 1.37 inch removes about four tenths of an inch from the front, which is more than the parent’s neck and shoulder together. Every part of the original neck is scrap. The new mouth is cut in what was, a moment earlier, case body.

That matters because a drawn case is not a tube of constant wall. It is thick at the head and web and thins progressively forward, because that is what the drawing sequence produces and what the cartridge needs: strength where the pressure acts on unsupported brass, thinness where the neck has to spring open and release a bullet. The taper is deliberate. Cutting the case in half and calling the remaining front end a neck inherits a wall thickness that was engineered for a different job.

The form die then necks that brass down to take a .308-inch bullet and pushes a new shoulder into it. Forming does not remove the surplus thickness. It moves it. A die that works from the outside — which is what a conventional form or sizing die does — sets the outside diameter and drives every thousandth of excess and every thousandth of variation inward, onto the bullet. Volume 9 makes that point about bushing dies in general and Volume 12 about the mandrel workflow that reverses it; here it is not a refinement but the central mechanism, because the excess being driven inward is not a few tenths of a thousandth of manufacturing scatter. It is the difference between a body wall and a neck wall.

Two quantities come out of this, and they are worth separating because the sources that disagree about neck turning are not always disagreeing about the same one.

The first is absolute thickness. A thicker neck makes a fatter loaded round. Loaded neck outside diameter is bullet diameter plus twice the wall, and the chamber neck has to clear it with enough room for the brass to spring open and let go of the bullet. If it does not, the case cannot release, and the pressure does not go where the load data said it would. This is the failure mode that matters, and it is a safety item rather than an accuracy one.

The second is variation, both between cases and around the circumference of one case. Variation between cases gives inconsistent bullet grip and therefore inconsistent start pressure. Variation around one case pushes the bullet off the axis of the bore, because the die works the outside of the neck while the bullet sits against the inside.

The chamber-side figure that settles the first question is the chamber neck diameter, and it is worth noting where that lives: the copy of the industry standard held in this collection carries chamber and test-barrel drawings, which is exactly the kind of drawing that would carry it — unlike case length, which Volume 16 could not source there and therefore does not print. Anyone forming brass for a specific rifle has a better reference still, which is a chamber print or a fired case from that rifle.

25.2 Parent Brass as a Quality Decision

Volume 10 covers sourcing and sorting brass, and Volume 16 states that sorting by headstamp precedes forming because forming amplifies every inconsistency in the input. The sharper version of that rule, and the one specific to this job, is that the headstamp being selected is a proxy for a wall dimension that will become a neck.

The most fully specified production process located for this dive — one loader’s documented workflow, run in batches of three thousand — selects Lake City 5.56 for exactly that reason, on the grounds that it “has the proper dimensions, especially regarding the wall thickness.” The same account reports measuring commercial brass at .012 to .0135 inch neck thickness and rejecting it as too variable. That is a spread of a thousandth and a half across commercial headstamps, which after forming shows up as three thousandths on loaded neck diameter and as a corresponding spread in bullet grip.

Two things about that figure deserve care before it is used as a buying rule.

It is a consistency claim, not a thinness claim. The argument is not that Lake City brass has thin walls; military 5.56 brass is generally reported as having a thicker web and less capacity than commercial .223 of the same cartridge, which Volume 10 records at around one to one and a half grains of water. The argument is that a single military lot is dimensionally uniform, and uniformity is what a production process can be set up against. A batch of one headstamp from one plant can be measured once and then trusted; mixed commercial brass has to be measured case by case, which defeats the point of batching.

And it is a single loader’s measurement, reported as practice rather than as a published specification. No manufacturer publishes a neck wall thickness for .223 brass, and nothing located for this dive independently confirms the .012 to .0135 range. It is used here as what it is: a measured observation from a working process, consistent with the mechanism, and worth repeating at the bench rather than taking on faith.

The practical form of the rule is stronger than “sort by headstamp,” which Volume 10 already requires for other reasons. It is: select one headstamp from one lot, measure its wall thickness at the point that will become the neck before cutting anything, and treat a change of lot as a change of component. The measurement is a ball or tubing micrometer — an anvil small enough to reach inside the case — taken at several points around the circumference and at a stated distance from where the cut will fall. The absolute number matters less than the spread, and the spread is what decides whether the next section applies.

25.3 The Donut

The same mechanism has a second consequence, further back on the case, and it is the one that catches loaders who checked neck thickness and thought they were finished.

A donut is a thickened ring at the neck-shoulder junction. In ordinary handloading it develops over a case’s life as brass flows forward from the shoulder into the base of the neck through repeated sizing, and it is a slow-onset problem of high-mileage brass. On a formed .300 Blackout case it is present from the first operation, for a different reason: the form die pushes a new shoulder into brass of body thickness, and the transition between the shoulder cone and the neck ends up carrying more wall than the neck above it. It is not wear. It is geometry.

What makes it worse in this cartridge than in most is where the bullet sits. The case is short and the projectiles that justify the cartridge are long. A heavy subsonic .308 bullet seated to a workable overall length puts its base at or below the neck-shoulder junction, which is precisely where the donut is. The result is a bullet gripped hard at one narrow band near its base and normally along the rest of the bearing surface — inconsistent seating force, inconsistent release, and, if the donut is pronounced, a bullet being squeezed at the point where it has nowhere to go.

The diagnostic is mostly tactile. Seating force that rises sharply at the end of the stroke rather than staying steady, or an expander mandrel that runs smoothly through the neck and then meets a distinct step, both point at a donut. A pin gauge will find it directly.

Three responses exist, and they are not equivalent.

Seat shallower, so the bullet base stays above the junction. Free and effective when the load allows it, and the heavy subsonic loads that most need it are exactly the loads that do not.

Set neck tension from the inside. An expander mandrel sets inside diameter directly and drives thickness variation outward, where it does nothing. Volume 9 and Volume 12 cover the mandrel workflow; its relevance here is that it does not remove the donut but it does stop the donut from being transmitted to the bullet along the rest of the neck. It addresses the symptom in the right place.

Remove the material — an inside neck reamer, or turning the neck with the cutter run down past the junction and slightly onto the shoulder. This is the only response that eliminates the donut, and it is another operation on every case, which is the theme of the section that follows.

25.4 Neck Turning, and a Contradiction Worth Leaving Open

The record contains a direct disagreement here, and the honest treatment is to present it as one.

One account reports that after forming and trimming, “you may need to neck turn about 90% of the cases to keep the neck wall thickness within limits,” and that military 5.56 brass is more likely to need turning than commercial .223. That is a single source, reported as practice, and it is mechanically exactly what the first section of this volume predicts.

The most systematic process source turns no necks at all. It reaches a finished case in nine steps, none of which is a turning operation, having selected Lake City 5.56 specifically for its wall thickness at the outset.

These are not compatible as stated. One says military brass is the problem; the other selects military brass as the solution. Two reconciliations are available and neither is confirmed.

The first is that the two accounts are optimising different quantities. The turning account is about absolute thickness — brass too thick for the chamber neck, which military brass is more prone to. The process account is about variation — a lot uniform enough that one die setting serves every case in it. Both can be true simultaneously, in which case the right reading is that selecting for uniformity does not by itself guarantee the necks clear the chamber, and a loader who did the first check still has to do the second.

The second, and the one the research file favours, is intended use. A plinking supersonic AR load tolerates neck variation that a precision subsonic load does not. A round that chambers, fires and hits a steel plate at a hundred yards is not troubled by half a thousandth of neck-wall spread or by a donut under a short bullet’s base. A subsonic load built for group size, with a long heavy bullet seated deep into the neck, is troubled by both. On that reading the disagreement is not about brass at all; it is two people answering different questions with the same word.

What would settle it is a measurement that does not appear to exist in the public record: loaded neck outside diameter, and neck wall thickness with its circumferential spread, for a run of formed cases sorted by parent headstamp, against the chamber neck of a specific rifle. That is a morning’s work with a ball micrometer and a notebook, and it would turn a contradiction between two accounts into a number. Until someone does it, both accounts stand, and the defensible position for a loader is the conservative one: measure the formed necks before assuming either source describes the brass in hand.

The practical consequence for the sequence is significant, because neck turning is not a step that can be added cheaply. It is another handling of every case, another tool, another setup that must be checked, and — since a turner references off an inside mandrel — it needs the neck expanded to the mandrel first, which is a further operation again. A nine-step process becomes an eleven-step process. That is the point at which the economics of Volume 16 stop being a question about tooling cost and start being a question about labour, and it is the point at which purpose-made brass, which has never been anything but a neck, becomes very hard to argue against for the precision project.

25.5 The Form and Trim Die, and What It Does That a Trimmer Cannot

Volume 16 draws the distinction between a form die, which reshapes a case, and a sizing die for the finished cartridge, which only sizes a case that has already been formed — and records the specific warning that at least one maker’s .300 Blackout set does the latter while carrying the cartridge name on the box. That distinction is not repeated here. What is worth adding is the third member of the family, because it occupies a position in the sequence that neither of the other two does.

A form and trim die is a die whose top face is a hardened length reference. The case is pushed up into it by the shellholder, the die necks it down from the parent and establishes the intermediate shape, and whatever length stands proud of the die’s top face is removed — filed or cut off flush against the hardened face itself. Redding’s own description of the family puts it exactly where the sequence needs it: for case forming and necking brass down from another caliber, the trim dies are the intermediate step before full-length resizing. Redding lists one for .300 AAC Blackout.

The interesting property is not that it trims. It is that it makes length a fixture dimension rather than a measured one. A trimmer, of any architecture in Volume 13, produces a length by referencing a surface and cutting to a stop that someone set. The die produces a length because the case is held on the die’s own geometry and material above the face is simply gone. Nothing is set, nothing is read, and the operator cannot drift. On a production run of thousands that is a different kind of process control from a well-adjusted trimmer, and it is the argument for the tool.

Three qualifications belong with it.

The cut is made with the case fully supported inside the die, which matters more here than it would on ordinary brass. A freshly cut case has an unsupported, work-hardened, out-of-round mouth, and a cut taken on it in free air is a cut taken on the least rigid part of the case. Inside the die there is steel all the way around it.

Its length reference is the die-to-shellholder relationship, which means it indexes from the case head. Volume 13 sets out why head-referenced tools drift when rim thickness varies between headstamps and lots. That consequence carries over: a form and trim die is superbly repeatable within a sorted lot and no better than any other head-referenced tool across mixed brass. Since this whole volume argues for running one headstamp from one lot, the two requirements are the same requirement, but a loader who ignores the first should not expect the second to rescue them. This is reasoning from how the tool is held rather than a manufacturer’s statement, and it is offered as such.

And it is not the same tool as an on-press size and trim die. Volume 13 covers Dillon’s architecture, where a motor-driven carbide cutter trims through a hole in the top of a sizing die on the same stroke that sizes the case. That is a powered station in a loading sequence. A form and trim die is an unpowered forming step in a manufacturing sequence, and the case going into it is not the shape the case going into the other one is. Sharing three words in the name, the two answer different questions.

25.6 The Blade Is the Consumable

The economics in Volume 16 are built on tooling cost against brass saving, and they are missing a term. Forming has a marginal cost per case, not only a fixed cost, and the biggest single component of it is the cutting blade.

The documented process specifies a jeweller’s slotting saw — two inches in outside diameter, .020 inch thick, 24 teeth per inch — and reports that a quality blade lasts 10,000 or more cuts against 3,000 to 4,000 for a generic one. Treat those as the working figures they are rather than as a manufacturer’s rating; they come from one production account. The ratio is what matters, and the ratio is roughly three to one.

The arithmetic follows directly. A program that forms ten thousand cases burns one good blade or three to four cheap ones. Whether that swing is large in money depends on the price difference, which is not quoted here because no dated vendor figure was confirmed for this dive. What is not in doubt is the shape of the cost: it is per-case and it does not amortise. Tooling gets cheaper per case forever as volume rises, which is the whole of the standard argument for forming at scale. Blades do not. Past the point where the form die and the cut-off jig have paid for themselves, the consumable floor is still there, and it is the term that keeps the break-even from running away to zero.

The more interesting consequence is that blade life is a quality variable and not only a cost one, and that it fails quietly.

A saw blade does not stop cutting. It stops cutting cleanly. As the teeth dull, the blade begins to burnish and push rather than shear: cutting forces rise, the thin blade deflects further under them, the cut wanders, the burr gets heavier and the cut edge is more heavily worked. Every one of those degrades cut-length consistency, and cut-length consistency is the input to a trimming step with a two-sided tolerance — Volume 13’s point that .300 Blackout is the one cartridge on this bench where the live risk is finishing too short, and Volume 16’s point that cutting long is what makes an error recoverable. The cases most likely to fall outside the window are the last few hundred off a worn blade, and nothing announces which those are.

That gives a bench rule that no source states outright but that follows from the two facts together: count cuts, change the blade on the count rather than on feel, and retire it early rather than late. A blade is the cheapest item in the whole process and the one whose failure mode is a batch of cases cut short, which is the one defect in Volume 16’s failure list that is unrecoverable.

Two smaller points on the same tool. The blade’s .020 inch thickness is not about the kerf — the material in the kerf is scrap either way, since everything forward of the cut is discarded — it is about rigidity, which is what holds the cut square and repeatable. And a fine-toothed blade in brass has very little chip clearance per tooth, so feed rate and heat are what load the gullets and end a blade early; the cut is a slow one by nature, which is part of why the cut station, not the trim station, sets the pace of the whole program.

25.7 Throughput, and What Actually Sets It

Volume 13 records the only concrete rate figure located anywhere in this equipment class — a user-reported figure for a drill-driven trimmer in .300 Blackout, of roughly five hundred cases an hour — and notes that no maker in the field publishes a rate at all. That figure is not restated as new here, and the more useful observation for a forming program is that it describes the fastest station in the sequence, which is not the one that governs.

A forming run is a batch process with nine or eleven operations, two of them length operations, at least two of them cleaning passes, one of them an anneal, and one of them a swage of the primer crimp on military brass. The rate is set by the slowest station, which is the cut, and by the number of times each case is picked up, which the single-pass tools in Volume 13 reduce and the turning operation in this volume increases. A batch of three thousand cases is days of bench time, not an evening, and the documented process runs exactly that batch size for the obvious reason: setups are amortised over the batch, so small batches are disproportionately expensive in setup.

That has a consequence for the hazard that Volume 16 states in its session form, and it is worth drawing out because the batch nature of the work extends it.

Volume 16’s rule is: never process both calibers in the same bench session, because .300 Blackout is made from 5.56 and the two are at their most confusable when both are clean, deprimed and waiting. A batch of three thousand cases makes that a rule about weeks, not about an afternoon. For the duration of the run there is parent brass in the shop, cut-but-unformed brass in the shop, formed-but-unannealed brass in the shop, and finished brass in the shop — four populations, three of them intermediate states that no gauge in anyone’s normal workflow is set up to identify, and all of them carrying a 5.56 headstamp. The formed ones are the hazard Volume 16 describes. The cut ones are worse in one narrow sense, because they are neither cartridge and belong in no loading sequence at all.

The control is the one Volume 4 specifies for components, applied to work in progress: a labelled container per stage, labelled with the stage and the date and not only with the caliber, and an empty bench between stages rather than at the end of the run. The label that reads “5.56” on a tote of cut cases is not merely unhelpful. It is wrong, and it is wrong in the direction that puts a formed case into a 5.56 magazine.

25.8 What the Production Detail Changes About Form-or-Buy

Volume 16 reaches a verdict: for .300 Blackout at the quantities a private bench consumes, buying formed brass is usually the better decision once tooling is costed honestly, with the exception of a loader who already owns the trimmer and the annealer for other work. Nothing here overturns it. Three things sharpen it.

There is a marginal cost as well as a fixed one. Blades are consumed at a rate of one per three to ten thousand cuts depending on quality, and no volume makes that per-case cost go away. The break-even against bought brass is therefore not a simple crossing point after which forming is free; it is a crossing point after which forming is cheaper by the difference between the brass price and the consumable floor.

The quality tier of the intended load decides how many of these steps apply, and the economics get worse exactly where the brass matters most. For a supersonic plinking load, thick necks that still chamber and a donut under a short bullet’s base are tolerable, the process is the nine steps Volume 16 lists, and forming is a genuine saving at volume. For a precision subsonic load — the application that justifies the cartridge to most of the people who load it — the necks have to be measured and probably turned, the donut has to be dealt with, and the process grows two operations on every case. Purpose-made brass has never been anything but a neck. It has no thickness step, no donut from forming, and no parent-lot dependency, and it sidesteps the entire content of this volume.

The yield question named in Volume 16 now has more defect classes in it, not fewer. That volume records, correctly, that no located source gives a scrap or yield rate for this process. The failure list it draws on includes cases cut short, inconsistent cut length by headstamp, split necks from forming without annealing, neck damage from inadequate lube and missed primer crimps. Neck wall thickness outside what the chamber will clear, and a donut severe enough to interfere with a seated bullet, belong on that list too — and they are both measurable rather than merely observable, which means a real yield measurement on a real batch could count them. That measurement remains the single most valuable unpublished number in this subject, and it is still unpublished.

The verdict that follows is narrower than Volume 16’s and points the same way. Forming .300 Blackout is a manufacturing process whose output quality is set at the first step, by which headstamp was selected and what its walls measure — not at the form die. A loader who will run one lot of uniform military brass, count blade cuts, batch in the thousands and load supersonic is doing something economically sensible with a predictable product. A loader who wants a few hundred cases for a precision subsonic load is buying a neck-turning program they did not intend to buy, and should buy brass instead. Volume 16 owns annealing’s place in the sequence and Volume 14 owns annealing itself, including the reason a temperature quoted without a dwell time is not a specification at all; that caution applies with particular force to formed brass, where the neck was worked harder in one operation than most cases are worked in their lives, and where the case is short enough that heat reaches the head easily.