> Forgings have the added advantage of variable grain direction which generally can be tailored to the stress patterns of a specific design.
This is a super underappreciated fact! It's often repeated that forging is just stronger, but just squishing steel does NOT make it stronger. Forging a part is so much more than just smashing it into a shape.
Steel cable is made of pretty ordinary steel which is stretched 100s of times its original length. That process alone makes it 2-4x stronger in that direction. You stretch steel and it gets stronger in that direction.
Do you see how complicated that optimization process becomes? The process steps are not just trying to take it to the final shape. Your piston rod needs to be strong lengthwise, so you actually want to start with a short fat ingot and stretch it out instead of one that is near-final size.
Think of making an I-beam. You could hammer out the middle, making it thinner. That would give you a bit of strength there but very little on the edges. If you instead pull the edges out, you create a long continuous stretch that will be very strong against bending. Where, how, and in what order you stretch makes all the difference. You may want to leave extra material and cut it off later, so that your grains are all oriented together instead of tapering to a point.
For any moderately complex part, this process is as complicated as modern engineering problems. With poor steel you genuinely need to understand how to foster and bring out those continuous lines or your corkscrew will unwind like playdough. Blacksmiths had a legitimately intellectual job back in the day!
This effect also applies to polymers! Perhaps even more so. Take a polyethylene bag (LDPE) and stretch the material in one direction. You might notice the material becomes thinner but also stronger. This is due to the polymer chains becoming aligned. Eventually you get "drawn fibers" where the molecular strands are aligned with the fibers for optimum tensile strength.
it varies a lot with polymers, and it's a different effect. steel is entirely crystalline; ldpe is mostly amorphous. a big part of what's happening in the strain hardening of ldpe, aside from making it uniaxially oriented, is that it's crystallizing; the crystalline domains become larger, greatly reducing the amorphous volume fraction. (there are also other ways of achieving this effect, such as annealing, which you will notice softens steel rather than hardening it.) ldpe's strength isn't determined by crystal dislocation density in the same way as steel's, and of course steel doesn't have polymer chains to align
HDPE is also what Nalgenes and the like are made of (or used to be anyway), and is very popular for storing chemicals as well. It is very nearly completely inert.
you cunninghammed me: nalgenes are polycarbonate, which is a lot less inert
all the polyethylenes are relatively inert, because polyethylenes are in some sense just heavy paraffins. paraffin is germanized latin for 'relatively inert'
just squishing steel does actually make it stronger, because it increases the number of dislocations in its crystal structure. smaller grains mean higher strength even without the variable grain direction. also, peening, which is not exactly the same as forging but is also just squishing steel, can give you higher strength for a third reason: areas with residual compressive stress can't initiate cracks until you overcome that stress, which increases strength. even more, though, it increases fatigue resistance
> Blacksmiths had a legitimately intellectual job back in the day!
ACOUP noted that blacksmiths might be assisted by unskilled laborers, strikers, who had the actual job of lifting the hammer and hitting the object with it.
> Unskilled feels unfair, it requires a fair bit of skill, and you're also learning how to forge while doing it.
Both of those points are untrue. Strikers are unskilled labor and in general are not learning how to forge. The smith shows where he wants the hammer to fall, and they let it fall there.
> Things were worse for the many strikers and other laborers who were essentially unskilled hired hands or even enslaved laborers (given their depiction in artwork, it seems likely many ancient strikers were slaves) of much lower status and who could not expect to be trained into blacksmiths themselves some day. While some strikers were probably apprentices in training, it is quite clear that not all of them were! These workers would also have been far less richly paid; indeed, the entire point of strikers was to have laborers who could be paid very little but still amplify the production ability of the blacksmith himself.
The reality is that, depending on circumstances, a highly skilled craftsman of yesteryear could have any number of obviously less-skilled assistants. Some would be "career track", some semi-skilled seasonal help, some minimally-skilled (whether due to youth, infrequent day labor, poor talent, or social status), and some in supporting type of skilled work - animal handling, cooking, bookkeeping, etc.
I tried to maintain a certain kind of optimistic humility, that almost anything which employs a person full-time is a problem-area that has fractal layers of complexity one don't have to know from the outside.
The only question is whether someone will pay you for doing the fancy skill/science tricks or not.
Not necessarily a culture war thing. People who aren't familiar with the subject might take "unskilled" in the plain-english sense, as a pejorative. And who can blame them? We're english speakers before we're technical speakers.
Anyone who has worked in a job classified 'unskilled' generally doesn't need a pejorative to feel unloved. It's the nature of the game.
Digging ditches generally sucks, same as I imagine being a striker, but most anyone can do it (until their body gives out, anyway, which in some cases is 'immediately').
One of the most interesting things I learned about blacksmithing (aside from all the metallurgy because I'm not a ME), was how much more valuable "the right hit" was than "the heavy hit."
Sure, it takes a ton of muscle, but you can quickly screw up a piece by repeatedly beating the hell out of it.
I'd really like to see some backing of these claims. I've seen "grain flow" claiming big gains for years in various enthusiast magazines (bike, motorcycles, cars, etc) as to why components are forged.
Then I started working in engineering, and I can't find any support for these claims. For sure when a steel bar is worked down to become wire for a steel rope, it cannot be pulled to an elongation of 100x increasing strength. A36 steel which is a basic structural steel has an elongation at break of 23% in a 2" gauge length [1]. In every rolling mill I've been in, there is a limited amount of reduction per pass through the mill, after which the metal needs to go for thermal treatment to be annealed to remove all the cold work. Every time you anneal the material, you completely resets the elongation (internal plastic strain) and strengthening due to work hardening. If they do too much reduction in one pass or at too low of a temperature, it cracks the material and makes it weaker.
For sheet metal, there is lore about the material being stronger in the rolling direction as that is the direction of grain flow. I have yet to find a source that can point to any large difference. In papers like this [2] there are claims of certain orientations of samples relative to rolling direction have different tensile properties, but when you look at the tensile charts, there is minimal difference. The yield strength in these charts isn't reported, but all three orientations look to yield at the same point. In this test the across the grain (90 degree to rolling direction) orientation had the highest tensile strength which is the opposite of the expectation of the forging "grain flow" promoters. But the magnitude of the difference isn't large, and is small relative to normal factors of safety in a reasonable design.
When designing automotive components, I've only ever seen forging methods selected for efficiency of production. If a part mostly fills the envelope of a bar or plate, it is cut from bar or plate in all cases. If there is a lot of void volume in the part, the calculation will be made to determine if the cost of developing forging tooling and development will get paid back in reduced material and machining cost. I have yet to see the dimensions of the part change with manufacturing method, which would be needed if the non-forged part was significantly weaker.
And finally, a lot of forged parts are subsequently heat treated. When heat treating steel all of the grains in the steel have to be destroyed and recrystallized. That is the mechanism by which heat treatment works. Depending on the exact process and part geometry, this process removes or reduces the grain flow in the finished parts.
Having said that, the claim of superiority of forging persists, and I'd love to see a technical reference that shows the magnitude of the change from someone who has plausibly actually tested the effect.
MIL-HDBK-5 [1] is a good publically-available source for strength allowables for several aerospace alloys, including multiple directions relative to the grain for some of them.
The first relevant example I found was on page 3-86, extruded 2024, 2.250 - 2.499 inch cross-section. For ultimate tensile strength, F_tu, the L (in the direction of extrusion) allowable is 57 ksi, while the LT (perpendicular to the direction of extrusion) allowable is 39 ksi. That's a 30% drop in strength.
Thanks for slogging through that one to find an example. I've come across that handbook before, and went looking through it in the alloys I normally work with. In the alloys I normally see, the L and LT are either identical, or with a couple digits in the least significant digit.
So it looks like the effect is very alloy dependent. I didn't see any of the steels having any notable directionality. Also Aluminum 6061 doesn't show any directionality either. Outside aerospace, I suspect that covers the majority of metal tonnage used.
Thanks for posting that as I was about to! I will note that MIL-HDBK-5 is no longer valid for actual aerospace design, as it has been superseded by Battelle Institute's MMPDS Handbook, which is locked behind a very very tall paywall. The MIL-HDBK is still all perfectly good data.
I think it is unfortunate the US government got rid of all their MIL and STD standards. You used to have a rich resource of technical specifications for materials, fasteners, fittings, etc for free access. Now they are pretty well all cancelled and have been moved to organizations like SAE and ASTM where they are hundreds of dollars per copy. For the metallurgical data, I'm curious if these successor organizations are actually generating any new data. Whenever I'm looking up references like that hand book, it appears they all summaries of investigations that happened back in the 1960's and earlier.
That’s the problem of capitalism : you need it but too much of it becomes a disease for the society abusing it…
Much of the western world is starting to feel the effects…
The best evidence for grain flow on a really atomic scale comes from what is called texture analysis in X-ray or electron-beam crystallography (or related techniques): you get a deviation in the distribution of Bragg peaks due to the fact that you have a non uniform distribution over the orientation of the unit cells within the crystallites in the bulk material. You can fit this in a spherical harmonic basis and quite accurately work out the excess or defect of the distribution, typically quantified in units of 'multiple of a random distribution' or mrd, again either in crystallographic axes or traditionally in three orthonormal axes – parallel to the surface of the workpiece ("rolling direction"), axially transverse to it, and normal to it. The phrase to search for is 'pole plot'. They're rotationally symmetric and an inverse projection over all space, and so usually only a quarter of a hemisphere is shown.
A very good example of the affect of annealing tungsten wire is here [1] – note that (a) there is a very clear orientation dependence that some difficult geometric transformations will undoubtedly show means that they are aligned in the wire drawing dimension; and (b) after annealing at 1600 ºC for an hour the preference is slightly reduced but still about 15 sigma away from random...
I'm not saying grain flow doesn't exist. My claim is that I can't find any support for the idea that grain flow results in superior strength characteristics in the grain direction.
Ahh, I see! This is a common problem with things that are "known" to be true -- often people don't rigourously test them.
This paper [1] has some good data in it:
"The experiments in this study were developed to verify the influence of the grain-flow
orientation on fatigue life and its impact on the anisotropic properties of a mechanical
component. To this end, steel specimens were made, and their fiber was oriented by
machining and hot forging. Subsequently, they were subjected to flexo-rotational fatigue
tests in a piece of specific equipment to determine their fatigue life."
(...)
They then describe three parts: A, properly forged, B, improperly forged, and C, machined.
(...)
"The results showed that specimens of configuration A achieved a much longer fatigue life than configurations B and C, actually doubling it. The results indicated a similar fatigue life behavior between configurations B and C. It is important to emphasize that this similar behavior between these two configurations is valid for this case analyzed (...)"
It's plenty real, and it matters for more than just strength. Cold rolled grain oriented electrical steel has better magnetic properties than non-oriented steel and is used in some applications where the field is in a straight line.
I very much agree with you and also would really like some clear evidence of that.
My opinion was formed from another area that is related : the pretense that “forged” knife are stronger, and hold their edges better, etc. I have seen some mostly nonsensical electronically microscope observations that didn’t prove anything except minor differences in “fiber” patterns that cannot be reliably be shown to be better in experimental protocols (provided you heat treat the metal in the same way and everything else being equal in particular the particular alloy).
I think this is one of those things that people keep repeating without any evidence because very is a large amount of marketing behind it as well as vested interests to sell more expensive supposedly superior “artisanal” stuff.
There are many examples of the likes, being mostly belief/lores repeated ad nauseam that becomes “true” just because everyone is saying it, yet with no hard evidence !
your comment is an extremely valuable contribution!
disclaimer: i don't have a relevant technical reference handy, and i'm far from an expert on the area, which is vast, and i recognize you know things i don't about it. still, i do spend a lot of time reading papers with metallurgical micrographs in them†, and i think i figured out the answer to your question many years ago, so i will explain my understanding
except for the part about grain orientation, anyway
> In every rolling mill I've been in, there is a limited amount of reduction per pass through the mill, after which the metal needs to go for thermal treatment to be annealed to remove all the cold work. Every time you anneal the material, you completely resets the elongation (internal plastic strain) and strengthening due to work hardening. If they do too much reduction in one pass or at too low of a temperature, it cracks the material and makes it weaker.
as i understand it, this is exactly right, but you say it as if it's contradictory. strain hardening increases the yield strength of metal (by making it yield). it can also change the tensile strength, but to a much smaller degree. when the metal can no longer handle stress by yielding, in particular by yielding in a way that produces further work hardening, so that the yield is distributed over the metal rather than being concentrated wherever it starts, it cracks. that's why strain hardening metal makes it more prone to cracking. in general, a given metal is more prone to cracking when you harden it, whether you harden it by cold forging, case hardening, or quenching. (peening is the exception; it inhibits crack initiation by a different method.)
the change in yield strength from cold working can be quite large, a factor of 4 or so. it doesn't change the ultimate tensile strength much (or at all in the case of your wire rope), but there are a lot of cases where what you care about is the yield strength, not the uts, because if the part yields by more than a tiny amount, it is out of tolerance and has therefore failed
(with respect to a36 steel, elongation at break, and wire rope, this is a minor detail, but it's possible to elongate it somewhat more through rolling than you can through wire-drawing. but you are certainly correct that you cannot elongate it 100×, and wire rope is mostly made by drawing, not by rolling.)
there are different kinds of heat treatment, but the most common kind for steel involves a phase transition to austenite and back, which does indeed destroy the entire grain structure of the steel, losing any potential advantage of forging, precisely as you say. i'd think this would also be mostly true for hot-forging, where steel is forged while still austenitic; the relevant grain structure for strength will be the one that the steel acquires when it leaves the austenite phase. there are other kinds of heat treatment (more commonly used with things like aluminum) that don't involve fully recrystallizing the metal, and i would expect some grain structure to survive those
probably none of that is telling you anything you don't already know, but perhaps it's a different way of thinking about the things you know that explains the apparent contradictions
as for which direction i would expect grain orientation to make things strongest in, i really have no idea at all
I'm not sure what you thought was contradictory within that quote. I thought it was reinforcing a single idea? I was pushing back on the idea that very high reduction ratios keep causing higher and higher strength. There is a pretty low limit to the amount of deformation you can make in steel and aluminum before you wreck the metal. You need to keep resetting the cold work via annealing to be able to keep forming the metal. Cold work is just done as the final pass with a very limited final dimensional reduction in rolling.
I'm not saying cold working doesn't happen, or doesn't affect strength. It certainly does. I'm pushing back on the idea that forging creates superior strength via grain flow. One of the sibling comments pointed out the MIL spec materials handbook[1] where he found some materials that do exhibit a strength dependency on grain direction. That is interesting.
That seems to be the exception rather than the rule. If you go to page 3-220 in that spec, they show 5052 Aluminum in varying degrees of cold work (H32, H34, H36, and H38), where higher degrees of cold work have higher ultimate and yield strengths, but the L vs LT directions are identical in many cases, or 1 different. That goes against the general idea that forging grain flow creates superior strength in general.
So basically what you say is agreeing with his observations ?!
If you cold forge there are some benefits (but the question would be what can actually be reliably be cold forged and be a useful object in our precise world ?).
If you hot forge and/or heat treat, most of the benefits are lost pretty fast, so it doesn’t make much difference.
As the OP seems to intuit there is probably not much real strength benefits to forging for useful objects in real use cases scenarios, the reason they are forged have to do with manufacturing processes more than anything else.
true, conventional welding or sintering would be a bad idea. but you can connect them together with brazing, laser-welding, explosive welding, ultrasonic welding, self-propagating high-temperature synthesis of an intermetallic like nickel aluminide, electrodeposition, lashing, or globs of glue
Doesn't annealing take hours, particularly at the lower range of temperatures? Perhaps the additive process can keep the metal hot for a much shorter time. Granted, this also means stresses from the manufacturing process will not be removed.
Most forms of metal AM require melting, which gives solidification microstructures. There are solid state (no melting) forms of metal AM though. Look up AFSD. MELD just got (part of) a billon dollar contract from the air force for it.
>"Can you forge metals in a highly controlled and directed magnetic field where you can orient the grain/alignment of atoms/fields in whatever direction you want. Further, if true, what happens when you make damascus from varying plated that have particular alignments/grains - and what are the features of this material?
> Can you forge metals in a highly controlled and directed magnetic field where you can orient the grain/alignment of atoms/fields in whatever direction you want.
This is how you make magnets. "Soft" ferromagnets have small, round grains that rotate to reinforce outside fields. "Hard" ferromagnets have permanent fields of their own and long grains that can't reorient.
Forging with a field has a very low impact on the material properties because of how weak a magnetic field is compared to the forces moving atoms- same reason steel loses its magnetic properties when it gets hot.
> what happens when you make damascus from varying plated that have particular alignments/grains
"Damascene" is the layered look most often made from acid etching sandwiched and forge-welded layers of different steels. Damascus is a single alloy for which the pattern is named.
Since in both cases the material is melted together, it's far too hot for any magnetic properties to have any impact.
If you're reading this article, you may wish to know that arguably a "counterpart" to heavy press forging is explosive forming [1] in which a chemical high explosive is used to force a template material against a template. The overpressure generated by the explosive can be equivalent or maybe even greater than heavy press forgings (a 50,000 US short ton force press exerts ≈500 MN force; peak overpressure close to detonating TNT or PETN explosives can be MPa or higher [2] so depending on the geometry of the part they may be comparable) and it has the added "fun" fact that complex cylindrical or spherical shapes can be made very easily and accurately.
The only people I know who have worked with this have used it to make superconducting magnets, explosively forming either titanium or high grades of nonmagnetic stainless (A4, which has µr ≈ 1) without causing marsenite formation due to machining. This includes a major international MRI scanner manufacturer, for one relatively niche product. It's like the "extreme" version of metal spinning [3] – forcing a rotating chunk of metal against a rotationally symmetric mandrel.
I used to work as a design engineer for a pressure vessel manufacturer - we used explosion bonding all the time to bond more expensive corrosion resistant layers to carbon steel backing parts.
Example: In a shell and tube heat exchanger, the tubes might have some really reactive stuff in it so you might make the tube side out of indium , titanium, nickel, or even an expensive stainless steel like S32205/S31803. The shell side might just have river water for cooling, and can just be painted and have a sacrificial anode somewhere inside.
The bulkhead where all the tubes penetrate (the "tubesheet") might be 6' (180cm) in diameter and 4-8 inches (10-20cm) thick - an extraordinarily expensive hunk of material (or possibly not even available in the thickness needed) when made 100% of the more exotic materials; easily in the 6 figure range.
Sometimes this problem is solved by having a welder coat the entire surface with weld metal that's good enough to withstand the corrosion characteristics of the process stream, but with larger parts this can take _days_; with some metallurgies (e.g. brass) it's not even possible.
Instead, the practice was to explosion bond a "thin" layer (1/4" (6mm) or so) of the expensive stuff to a more standard carbon steel forging. The tubes are usually very thin walled and welded/brazed to the cladding.
What's cool is the interface layer between the two metals looks like when two liquids meet with swirls and whorls of the two materials interleaving, but frozen solid.
That sounds amazingly fun -- thank you for sharing, and for including si units! Do you have any pictures of the whorls? (I'm imagining something like damask steel)
There are some popular videos making spheres using explosive hydroforming, which is quite fun, and much lower tech than explosively forming magnets to avoid the formation of marsenite (sp?).
I heard it from my brother who visited one such company, where they use explosive forming to shape metal parts used oil refinery; if I remember correctly for the corrugation on the metal that separate the hot and cold liquid in an heat exchanger, with the idea behind using explosive forging was that it allowed them to shape a big sheet in one go.
These massive presses are/were of strategic importance. Something politicians, most of whom seem to be lawyers these days, completely fail to grasp. I am not in general a fan of government subsidy, but I was very disappointed when the government of the UK declined to fund a large press in Sheffield, which would have been used to build the next generation of nuclear reactors. June 2010 timeframe. Luckily natural gas is cheap and has no geopolitical supply chain problems eh?
We're not going to build the nuclear reactors either. The problem will finally be solved by re-permitting onshore wind and a lot of batteries imported from China.
(The "local supply chain is vulnerable to political uncertainty over long term project funding" problem is much worse in regard to trains, and has resulted in losing most of our train building capacity. See HS2 fiasco.)
This is basically a pipe dream. But as long as the subsides are going to Big Green scammers, we will have numerically ignorant journalists peddling this bullshit.
I don’t understand how anyone can push that “solution” with a straight face. Surely they have worked out the numbers and figured out it’s not realistic at all, right,…,right ?
> By the early 2000s, parts from the heavy presses were in every U.S. military aircraft in service, and every airplane built by Airbus and Boeing.
>The savings on a heavy bomber was estimated to be even greater, around 5-10% of its total cost; savings on the B-52 alone were estimated to be greater than the entire cost of the Heavy Press Program.
These are wild stats.
Great article! I was fascinated to learn about the Heavy Press program for the first time, here on HN[1] a month ago, and am glad more about it is being posted.
It makes me think: what other processes could redefine an industry or way of thinking/designing if taken a step further? We had forging and extrusion presses … but huge, high pressure ones changed the game entirely.
Silica aerogels are dermally abrasive.
Applications for non-silica aerogels - for example hemp aerogels - include thermal insulation, packaging, maybe upholstery fill.
There's a new method to remove oxygen from Titanium: "Cheap yet ultrapure titanium metal might enable widespread use in industry" (2024) https://news.ycombinator.com/item?id=40768549
> The largest, the 50,000-ton forging presses, were behemoths: each was the size of a ten-story building, and could exert enough force to lift an entire battleship. The 35,000-ton forging presses weren’t much smaller.
and then
> Following Germany’s surrender, the U.S. and the Soviet Union divided up its large press capabilities as well as its rocket scientists. The U.S. dismantled four German presses and had them shipped back to the states
I wonder what the logistics for moving something like that across the ocean was. I know Soviets dismantled a bunch of factories during the war and moved them far behind the front lines...wonder what that was like.
For context, I work on my car a bit, and the shop manual for it is hundreds of pages (thousands?). Now scale it up to a factory, maybe without manuals. Disassemble, crate it, assemble it 5000 miles away.
Based on my software experience, I can sort of go in blind and figure out how a system functions. I suppose that translates to real world too..
There’s a UX school that says that if the user needs the manual you fucked up. I mostly subscribe to that school. Mostly.
It drives my family nuts that I will assemble a piece of furniture without reading the instructions. But the thing is with a little mechanical sympathy, and a well designed product, there’s only one sensible way for the parts to go together, and if you organize them right while you disassemble it (granted, harder to do when shipping overseas) then you’re good.
Imagine you had a device where four hardened steel bolts held the critical parts together. It would be stupid if the handles used the same bolt sizes in mild steel, right? Someone will fuck that up and use the wrong spare parts or do deep maintenance wrong. You’d use a different size bolt so they can’t get mixed up.
Your viewpoint is rather naive in the mechanical world. OK it's enough to assemble a coffee table, now try an engine with 2,000 parts. Without a manual how can you set bearing clearances? How do you know how much thrust clearance there should be? How do you know which way up a piston ring goes, or how much torque to apply to a head bolt?
Machines are extremely complex, and that's before you even touch electronics and hydraulics, both of which are highly complex systems. Simply moving large machine parts safely requires documented procedures, let alone order of assembly.
there's a lot of shade-tree mechanics who have successfully rebuilt engines with thousands of parts. essentially figuring out how to rebuild an engine is similar to figuring out how to build one from scratch, which is within human capacity, particularly with background knowledge. but the guy rebuilding the engine has a lot of hints
granted, he'll probably fuck up his first two or three pretty good without haynes or chilton
> granted, he'll probably fuck up his first two or three pretty good without haynes or chilton
Given the assumptions, inaccuracies, and mistakes I've seen in some Haynes and Chilton manuals they'll probably fuck up with them. Factory manuals are usually worth the price (Honda's are, KTM's not so much).
There is a lot of stuff you can do when it’s pass fail. As a pro you have a time limit and you’re bad if you can’t rebuild in X hours.
My dad will tell you I helped him rebuild a bike coaster brake at 14. But the truth is the only decision he made was to buy the repair kit. I got rags and laid all the parts out like an exploded diagram, we cleaned them or swapped them and they went back in the way they came out.
I worked as a bike mechanic for two summers in college. Cars have manuals and maybe the mechanic you work for has them. Bicycles do not. You’re all shade-tree until you’ve seen everything a couple times.
I've rebuilt several motors, transmissions, various other mechanical contrivances. Sometimes with decent documentation, sometimes not so much. Also done a bit of amateur machining, and worked as an engineer on physical products.
Under no circumstances would I claim that rebuilding a motor was essentially figuring out how to build one from scratch. In software, maybe that's like claiming that figuring out how to configure a new Linux box is essentially the same thing as figuring out how to write an OS.
yes, i agree. what i was trying to express is that rebuilding a motor is generally strictly easier than building one from scratch, because it's building a motor from something more than scratch. i don't think i expressed it very well
(i mean, if all the parts of your engine are trashed, you are going to have to machine replacements for them, and that might actually take you longer. but it's clearly achievable given that people have built internal combustion engines without a working example to take measurements from)
It's a bit academic, but set theory doesn't really apply to such fuzzy human things as knowledge and experience. Repairing and designing are different pursuits which might have a lot of similarities, but I wouldn't presume that a design engineer could competently do the work of a technician.
Just consider that any particular field of engineering as might be described by a lay person, can be far too broad and deep for an individual to be competent in all facets of it. I'm reminded of my neighbour asking for some help configuring email for her new iPhone, because she knows I do computer work. Mainly firmware.
there's something to that, for sure; there are plenty of design engineers who don't know nearly as much as they think they do, and who depend heavily on the expertise of their technicians to get anything done in the real world. they could never build an engine on their own! but there are also others who are eminently capable at the technical level, and i think their designs benefit from that
repair and design have in common that they require a lot of hard thought about the causal relationships involved in making the artifact work, tracing the causal chains through until they break, then patching them up. but they both also certainly involve other skills that the other does not; design also requires figuring out how to make new things happen, which involves imagining things that have never happened, while repair also requires knowing how not to bust your knuckles or spill the gasoline
Does every car mechanic have the entire Chilton’s catalog or do the mechanics just have to memorize things or look them up on the Internet? My understanding is it used to be a little A and a lot of B, and now it’s a mix of B and C
I was using the engine example as an analogy in an apparently failed attempt to help you appreciate the complexity involved. Mechanics in general will feel their way around an issue, but almost universally have access to paid repair databases when non-intuitive and complex issues come about.
You might recall that we're all talking about manufacturing equipment that was exfiltrated as war reparations.
I'm much less convinced than you are about the availability of accurate and detailed manuals. Which is why I keep steering the conversation to more murky engineering projects.
But I do want to circle back to say that I did at some point higher up gloss over the importance of things like torque and clearances. I'm not trying to say that those are things you can just intuit. Even if we could both probably dig up an old mechanic who tightens things by feel.
I don't understand the problem. You take the equipment, the manual, the guy who used to read the manual and maintain the equipment and the guy who wrote the manual and designed the equipment.
I think everyone has access to full shop manuals in soft copy these days. I imagine it's a lot like writing code. You remember how to write a for loop (oil change? a brakes?) but have to look up API docs for more esoteric functions (a clutch job?).
Yeah I laughed my ass off reading that. Something as basic as assembling a modern carbon bike can have you make a hundreds dollars mistake but there will always be smartass like those who always know better than the people who actually engineered the stuff.
But I guess IKEA furniture is a pretty low bar to clear so there is that..
Better hope they done have one set of 50mm bolts which are mild steel, and another set that are hardened high tensive steel that aren't clearly marked and the same length!
Or you're going to have quite an adventure when you go to do your thing and a random selection of bolt heads come pinging off at you at mach 5.
This should be a lesson for free-market advocates, especially those who see the US economic boom as a result of laissez-faire economics. In reality there are opportunities the free market doesn't take, and wise government intervention can yield enormous benefit to the public.
That is a very simplistic perspective of 'capitalism', it is a bit more nuance than that. It typically requires intervention to prevent it from destroying the economy because of insider trading, liquidity, and monopolist practices. The only criticism you say about it mirror aspect of our-selves that we don't like to admit. It isn't perfect but people are not perfect how-ever based on how fast we're pull people out poverty, you have to admit its pretty good!
I don't believe the free market would ever have produced the heavy presses. The coordination problem between making designs that use the heavy press and the investment to make the presses is essentially the prisoner's dilemma - each actor reaps a large reward if they both commit but loses massively if only they commit. The outcome is that they don't commit, especially if you add in real-world factors like alternate uses of the investment that competition tends to force the businesses involved into. The only hope would be something vertically integrated that can commit both, but there's no particular reason to believe that such an enterprise could enter into the market.
The free market probably could and would have optimized the situation at hand. Machining would have become cheaper, solutions to the fastener issues mentioned would be found and so on. This might even end up being better than the heavy presses - that's a technical question not an economic one - although the article makes it sound like the forging solution really is inherently superior.
Was reading this NYT op ed - https://www.nytimes.com/2024/08/19/opinion/chris-murphy-demo... - and reflecting that the people talking don’t get that the secret sauce for meaningful labor is not just that it pays, but also the sense that it is “special” somehow, in the sense that manufacturing parts with 50000 ton press is manual labor but also remarkable because of the uniqueness of the machine being used.
Most discussions about trying to build industrial capacity in the US seem to focus on either our high labor costs or on the disinterest in capital to invest in low margin places. I would love to understand what time frame of guaranteed business the government provided these companies to convince them to participate, and also what other industrial processes the government invested in which failed to take off. Specifically, why didn’t this sort of thing work for the solar industry a few years back?
That's pretty ungenerous. The press release says "The press is the world’s strongest hydraulic pull-down die forging press in pit-mounted design" so it's easy for a layman to read "world's strongest ... press" and take that at face value.
i admit i don't know what 'pull-down' and 'in pit-mounted design' mean, and i'm not sure my understanding of 'die forging' is correct, but that seems likely
on the other hand, the press release might be written by the same sort of people who say things like 'the world series', which is a baseball tournament between teams from the usa (and canada)
> The United States leadership only lasted two years: in 1957 the Ukrainian company Novokramatorsky Mashinostroitelny Zavod (NKMZ), specialized in steelworks equipment, built two 75,000-ton presses. The first one, destined for a plant in Samara, is now owned by Alcoa’s Russian branch. The second was installed in Verkhniaïa Salda and is used by VSMPO-AVISMA, the world’s leading producer of titanium and other specialty alloys.
> Outside the two superpowers, France was the third country to equip itself with a hydraulic press of this size: also built by the Ukrainian NKMZ, this 65,000 ton presse hydraulique* was installed in Issoire between 1974 and 1976. Owned by Interforge, the machine is 36 metres high and manufactures components for Airbus, Boeing, the space and transport industries.*
...
> After 60 years, the USA has added a new 60,000-ton hydraulic forging press. Built by SMS Group and managed by Weber Metals in California, it started operations in October 2018.
> The heavyweight champion, of course, is Chinese: a machine with the incredible power [sic] of 80,000 tons is in operation since 2013 for the giant Erzhong Group in the province of Sichuan. As tall as a 10-storey building, its use is very confidential: it seems to be used to build parts for military aircraft, like its titanic sisters. To give an idea of the power of this machine, with its 780,000 kN it could easily lift an entire cruise ship. As often happens, larger does not mean better: it is not the most technologically advanced press in the world. It was built by adapting old USSR projects from the 1980s, and is currently underused due to competition from the other giants we mentioned.
Nobody in North America has a press that can make a vessel for a nuclear reactor without making it in pieces and welding them together (probably adding a year to the schedule if you don’t use a new welding technique just developed in the UK not to mention people being anxious over the welds)
There is a Canadian company that is gearing up to make small reactor vessels like the BWRX-300 but so far I haven’t seen a sign they aren’t Nuscale 2.0
I work in an EB shop doing everything from welding, engineering assistance, machine repairs, and complete machine rebuilds. It's quite an impressive process and Ive had parts in my hands that come from some of the biggest names you can think of in aerospace, semiconductor, military, medical and more.
Are they for prototypes? Or how do you service production runs for parts?
If for arguments sake it were a part for an AWACS or an aircraft carrier you might only need to make eight or a dozen. But even military aircraft tend to run into the hundreds.
We do production runs with the rare service run. Most parts are one time use as you might imagine. Service work is the rare mold repair.
We are part of a few companies just in time manufacturing so they pay for expedite processing on orders as small a one piece to a few dozen. And we can get production run orders in the tens of thousands.
That is, a vessel for a pressurized water reactor. A reactor operating at lower pressure (like one using molten salt as a coolant) wouldn't need such a vessel, as the salt would not become highly pressurized, even in accident conditions.
I saw a link from some government minister in Canada named Don Morgan who stated a BWRX-300 would cost $5B (CAN), which comes to about $3.6B (US). $12/W (US) is not the worst, but it isn't great. Not clear if that was all-in cost or just overnight cost.
I understand that this press, and the Cleveland site, is now owned by Howmet Aerospace, which is a "daughter company" of Alcoa (along with Arconic) after its 2020 separation.
Howmet is an interesting company. My dad worked for them his entire life (it was called MISCO, then Howmet bought them, then Alcoa). He worked specifically in titanium injection molding--they would heat titanium until it was liquid and then force it under pressure into intricate molds they made onsite.
Most of the time they made turbine blades for jet engines, but if it was slow they would make golf club heads for PING and companies like that. In all the years he worked there, the golf club heads were the only tangible product of his work that I saw because everything else was tightly controlled in the facility.
The blades were actually perfect crystals, with either all the grain boundaries aligned, or no grain boundaries. This lets them run at higher temperatures without melting, increasing efficiency.
Pratt & Whitney appears to have developed most of the technology.
I got curious as what happened to the other 4. Wikipedia lists them as extrusion presses being scrapped in the 90's and 2021 in Maryland and Torrence CA. https://en.wikipedia.org/wiki/Heavy_Press_Program
I wonder if they wore out?, something better came along?, or just no demand?
They've been maintained, wear parts such as bearings and seals can be replaced. You can keep well-made machinery running almost indefinitely if you take care of it.
I'm not sure it's rose-tinted glasses or some sense of nostalgia for the "good old days", but it does feel that our manufacturing processes have taken a step backwards since then.
The second moment of area is quadratic with thickness (t^2), so 250/8" plate. But in reality 2-3x more than that, because at that scale steel gets a bit... goopy.
I saw a guy weld 1.5 or 2 inch plate once to repair the bucket on a giant bulldozer. That was… interesting. I was trying to figure out why he left such a big gap until he started welding. He had to fit the head in to lay down layer after layer of welding bead to join the pieces at full depth.
For a butt joint V it out on both sides and do many, many passes. Maybe have a few guys with rosebud torches working heat into it as you go. Big rods make it go faster (https://youtu.be/j61ezBX-EyA). For a lap joint it's the same idea, you're going for a great big fillet. But if you have a gigantic press at your disposal there are other options. For inspiration: https://youtu.be/k_LA_R4ifYk
The idea behind forge welding is you get both parts nearly molten (e.g. "welding heat") then your hammer blow (or the pressure from a huge press) puts enough energy into the weld area to briefly melt it.
Also, hot rivets might be a better option than welding if you can get away with it.
> Partly this broad range of uses for the heavy presses came from expanding the range of materials used in them. Originally the presses were designed to make parts from aluminum and magnesium, but by the 1960s they were not only pressing aluminum and magnesium but steel, titanium, nickel, copper, columbium, beryllium, and a variety of other metals.
Columbium is apparently an old name for niobium, and one perhaps still in use by American metallurgists.
There's no way anyone was making huge niobium forgings, though. Or nickel? Surely this is a reference to the use of those elements in superalloys.
They talk about machining, casting, and pressing. Mostly with casting and pressing in a good light, machining not so good.
As somebody who knows absolutely nothing about this stuff, I wonder—casting, I thought, was generally a lower quality option (like cast iron doesn’t have fantastic high-performance material qualities, and I had some crappy cast pewter toys as a kid). Are there different, higher quality casting processes, and I’ve only seen the bargain bin results? Is there a general ranking of the quality of the result or is it all very complicated and material specific?
> like cast iron doesn’t have fantastic high-performance material qualities
Cast iron is a material, not a process. It's an unfortunate legacy term for very high carbon steel (>2% by weight, or <11 iron atoms per carbon atom). For reference "standard" steel is ~.08-.18% carbon, and high-carbon steel is ~.8% carbon.
The >1% carbon precipitates out into graphite within the steel, causing it to behave totally differently. Less rusting, but weaker and much more brittle when solid. Less viscous when liquid, so you can cast long and thin parts.
> Are there different, higher quality casting processes, and I’ve only seen the bargain bin results?
There are, but it mostly is independent of the material. Some turbine blades are cast as single crystals for heat stability; you can't really cut a single crystal without introducing cracks and issues. There's also vacuum casting and spin casting (using a centrifuge to force liquid into the mold), which lets you cast metals that react with air or cool too quickly for normal methods.
Most of the variation in process is about the final form you cast into, though. Engine blocks are sometimes cast into a one-off ceramic shell that is sprayed onto a sand form. It's an expensive process but it lets you do the whole thing in one step.
> is it all very complicated and material specific?
It is very material specific. Fundamentally its all about shaping the grains. In many steels you can physically alter grains. In others, like precipitation grains (aluminum alloys, some steels) the structure is determined by the cooling and you can't physically shape them. In that case you may often get a better structure by casting since you can choose how to cool parts down, while a billet will have a homogenous structure that is usually worse towards the center.
Monocrystalline turbine blades are one of those things that make me appreciate how deep the state of the art is for a lot of "simple" things when you look closely.
> The >1% carbon precipitates out into graphite within the steel, causing it to behave totally differently. Less rusting, but weaker and much more brittle when solid.
If one adds some magnesium or cerium to the alloy, the graphite precipitates out as spherical nodules rather than feathery dendrites. The resulting material, called ductile iron, is much less brittle than traditional cast iron.
An advantage of the higher carbon content is a reduction in the melting point (by > 300 C), so the material is easier to cast than low carbon steel.
Castings can achieve shapes that are impractical to machine. A classic example is a spoked wheel. The spokes are really hard to make precisely with machining, but fairly straightforward when casting. Bandsaw wheels are virtually always cast, for example.
Also, the processes are not really independent. It can be much cheaper to do a rough casting, and then machine just the critical faces of it, instead of using an “off the shelf” hunk of metal and machining it all into shape. So it’s not really “casting is superior to machining” or vice versa. More that machining is high precision but expensive. Casting has some up-front cost but once the patterns are made, each item will use material quite efficiently.
Spoked wheels aren't cast because it's hard to machine them. You can easily CNC them if you really want. But that would be insanely expensive and wasteful and not any better.
I've watched enough Chip Foose on TV to know that in the custom car world it is commonplace to start with a solid blank wheel and CNC away 90% of the face to achieve any particular design. It takes only minutes, and unlike casting or forging you do not need to purchase hard tooling for each different design.
Cast iron is a material, it's good for what it's good for but its material properties don't really come from the casting process, just the ratio of iron to carbon.
Casting has problems with thermal expansion, plenty of materials shrink significantly as they cool and complex parts cool unevenly which can cause them to break or deform.
Casting has problems with microstructure, plenty of materials, especially steels develop complex crystal structures with multiple phases of materials as they cool from liquids and even extensively in hot solid phases. It's hard to control this in a cast part.
Casting has problems with precision. The molds just can't be all that precise when in machining, a thousandth of an inch can be a relatively large distance.
However casting gets a bad reputation because most of the time you see it it's because it's actually very cheap, cheap materials, cheap process, minimal post processing. Higher cost things don't necessarily realize the savings from casting as much so they don't use it. And also a lot of higher quality materials have higher melting points which require more advanced tools to melt and handle.
Plenty of things though are cast and then machined, you notice this if you look.
They also talk about forging. Forging can produce higher quality parts because metal has a grain structure. Just like wood, it's easier to break between grains than across the grain. Forging is deforming the metal without melting it. If the temperature isn't too high, it will deform the grain structure along with it. The grain can be aligned to make the metal stronger along the axis where it needs to resist the most force.
Wikipedia has a image of an connecting rod that has been etched to show the grain:
You can see the grain has been stretched along the length of the narrow parts. Wrenches are another example of something that's commonly forged for this reason.
Casting isn't necessarily "lower quality," it just has different properties. It's also much older (by thousands of years) than machining or pressing, but you can't get to the pressing or machining step without having an ingot, which comes from the casting process.
The main drawbacks of casting are you get a hard, but brittle product with (generally) uneven quality. There are processes (like annealing, though I don't know how you anneal a massive component) that can solve these problems, but all iron/steel is "cast" at some point.
depends on what you mean by 'machining'. grinding is at least 7000 years old https://en.wikipedia.org/wiki/Shoe-last_celt and roughly contemporary with pottery. casting presumably began with pottery but was definitely in full swing by the bronze age, a mere few thousand years later
Which cast iron? There are many different grades/alloys with different properties. Sometimes cast iron (some specific grade) is better sometimes it is worse. And of course you can cast things other than iron, cast steel does exist (rare outside the feed to a press, roll, or some other process, but you can cast steel into a specific shape if you want to). There are trade offs. The topic is so broad we cannot even start to talk about it. Instead we start with what your application needs and then look at the options to get that.
Cast iron is fantastic for building machine out of - while it isn't as strong, it is stable against vibration. There is a reason engine cylinder selves are often cast iron.
> I had some crappy cast pewter toys as a kid
Those were pot metal, not pewter (pester has many definitions but implies some qualty control). They are made out of whatever melts in a pot - often whatever is cheap at the recycle yard (without trying to identify what is in the metal - including lead which shouldn't be used in toys). Typically no control of the alloy was made and often they start with several different things that are great in isolation but when mixed result in bad behavior. Then the next time the make the toy they use different mix and get different properties. If you spend a little extra to get a known alloy pot metal is a high quality castings with great properties.
> I’ve only seen the bargain bin results
You have likely seen a lot of non bargain bin results. However since the parts are invisible you never thought about it or the alloy used. You see the failures and so casting gets a bad reputation because it is obviously used in the cheapest things with low quality control. (the door knobs in your house are likely cast pot metal plated with brass, but they last for decades)
>Cast iron is fantastic for building machine out of - while it isn't as strong, it is stable against vibration. There is a reason engine cylinder selves are often cast iron.
I thought it was because cast iron had very high resistance to wear.
Machining is fine, but in the context of the article the problem is that it usually starts off from plates or otherwise block-shaped materials. That means that if you have a weirdly shaped part full of holes and/or with many protrusions in weird angles (as vehicle parts tend to have), then you either have to combine many smaller parts into a final part or have to machine away a huge amount of material from a solid block to come to the final shape. It would be much better to cast or forge the part into roughly the right shape straight away, so you'd need only minimal finishing work. That is what these presses enabled.
I first learned about these presses several years ago from the YouTube channel Machine Thinking which is excellent and unfortunately infrequent. The linked video has some excellent footage of these beasts in action.
> Experience developing the SR-71 led Kelly Johnson to believe that the U.S. needed even larger, 250,000-ton presses, and in the 1980s the Army studied whether presses of up to 200,000 tons might be useful. But no such presses were ever built, either in the U.S. or elsewhere.
I imagine extremely large construction and digging. Kilometer-tall buildings and deep large tunnels.
We didn't have any of these presses in WWII. We have several of the article's presses still operating today.
Furthermore, all presses mentioned in the article have been surpassed by a 60,000 ton press that opened in Los Angeles, in 2018.
People with agendas will happily feed others narratives about the US not investing in manufacturing anymore, but it isn't true.
US manufacturing output has been steadily increasing since always, with the occasional 1-3 year dip during recessions. However, manufacturing does represent a lower percentage of our GDP with each passing year, despite the absolute value increasing.
>People with agendas will happily feed others narratives about the US not investing in manufacturing anymore, but it isn't true.
It's a matter of perception. Most people don't directly see (or buy) the stuff manufactured in the US these days. Normal people don't buy nuclear power plants, aircraft engines, commercial aircraft, etc., and certainly not military hardware which the US makes a lot of. They do buy clothes and various consumer electronics, and they see "Made in China" (or for some clothes, places like Bangladesh or Vietnam or Cambodia) printed on all those, when 50 years ago all that stuff had "Made in USA" printed on it, or for the nicer consumer electronics 30-40 years ago, "Made in Japan". People still might be getting a CPU for their laptop computer manufactured in the USA, but the chip will probably say "Made in Malaysia" because only the silicon was made in the US, and was then shipped somewhere else for packaging.
>However, manufacturing does represent a lower percentage of our GDP with each passing year, despite the absolute value increasing.
I'd say that's probably a bad sign: what other sectors are increasing? Likely they're things that aren't actually productive, such as healthcare (the value received does not represent the price paid in the US by a long shot, compared to other advanced economies; most of the money goes to insurance companies and waste), legal services, ever-increasing real estate valuations, etc.
> Does the US invest in manufacturing like this anymore?
It might be easy to argue over the exact degree of similiarity, but I'd argue that the US has repeatedly made manufacturing investments since the 1950s. Buried in bills signed into law, you'll find such investments.
Recent examples include the Recovery Act of 2009 or the American CHIPS act of 2022.
However, this might not matter as much now as it did in the past due to nuclear weapons being the primary deterrent in war these days, and the fact that our standing fleet of aircraft, aircraft carriers, nuke subs, tanks, etc... is essentially second to none. Additionally what we do have is highly capable and extremely specialized, in my opinion, leading to not really needing as many (quality over quantity). Take for example, an F35, which doesn't really have an equal in the skies, we have over 630 of them, with the goal of having around 2500. China only has 300 J-20s which are basically a copy of the older F22. Russia only has 22 non-test Su-57s. Would we have a realistic need to build 1000 of them within a year?
Due to many factors, but primarily free trade and globalization, it's unlikely that we ever see that non-automated manufacturing capacity return, though if needed we could probably mobilize the economy via the defense act to force more manufacturing capacity, though it's hard to imagine we would currently need to.
> Does the US invest in manufacturing like this anymore?
No, but that is different from not investing. Today the US invests more in automation and engineering and less in manual labor.
> if there were a war today would the US be able to produce as much as it did in WW2?
It took several years to ramp up to WW2 level production. We would see the same, a couple years of trouble on the fronts while building industry at home, then when the industry is built up massive production.
Historians (amateur so I'm not sure if they are right) tell me Hitler was ready for WW2 first and Italy begged him to not start the war as their industry wasn't ready. However France and Briton saw the war coming and were building their industry and so waiting might have made things worse.
I wonder how history would be different if Hitler had told France and UK to f** off with the WWI reparations stuff, threatening a war if they tried military action to force payment, but then didn't invade anyone and concentrated on developing their economy and becoming a technological and manufacturing power.
Amazing achievement. But I can't help but notice how much we celebrate these efficiency gains in the past, while being scared of them in the present.
At heart, this is the story of using less labor to product a unit of output, and possibly also improving the quality of those units of output. This led to much better outcomes for society. Was anybody then scared of what would happen to the riveters? Maybe, but those voices would have been drowned out.
These days, production is demonized, particularly at scale. Is there any successful producer being celebrated for being so? Are there any new, innovative, labor-reducing technologies being celebrated?
Does it take a war to snap us out of our fear of being better?
Yes, it is mentioned in the article, and it specifically says: "We can see a modern, obvious parallel between the Air Force Heavy Press Program and Tesla, which has created a similar revolution in car manufacturing by using large aluminum castings to replace dozens or hundreds of smaller parts."
So I didn't think what I said was out of line. Obviously, I was mistaken.
This is a super underappreciated fact! It's often repeated that forging is just stronger, but just squishing steel does NOT make it stronger. Forging a part is so much more than just smashing it into a shape.
Steel cable is made of pretty ordinary steel which is stretched 100s of times its original length. That process alone makes it 2-4x stronger in that direction. You stretch steel and it gets stronger in that direction.
Do you see how complicated that optimization process becomes? The process steps are not just trying to take it to the final shape. Your piston rod needs to be strong lengthwise, so you actually want to start with a short fat ingot and stretch it out instead of one that is near-final size.
Think of making an I-beam. You could hammer out the middle, making it thinner. That would give you a bit of strength there but very little on the edges. If you instead pull the edges out, you create a long continuous stretch that will be very strong against bending. Where, how, and in what order you stretch makes all the difference. You may want to leave extra material and cut it off later, so that your grains are all oriented together instead of tapering to a point.
For any moderately complex part, this process is as complicated as modern engineering problems. With poor steel you genuinely need to understand how to foster and bring out those continuous lines or your corkscrew will unwind like playdough. Blacksmiths had a legitimately intellectual job back in the day!