Why do railway tracks have crushed stones alongside them? | Hacker News Reader

Why do railway tracks have crushed stones alongside them?

alpharail.co.nz

724 pointsTwixes283 comments

This article doesn't explain one interesting bit:

It is important that these crushed stones stay clean. If dirt gets in them, known as fouling, filling the gaps between the stones, then eventually it will cause a derailment.

Briefly, the stones move slightly as each train passes above them. The tracks 'float' on the denser stone. This effect helps keep the track straight and level - higher regions get pushed down, while lower regions raise themselves up.

Fouling stops this effect, and eventually the track will become so uneven that a derailment happens.

To prevent this, a big part of railway maintenance involves digging out all the stones, washing them to get rid of sand and soil, and putting them back. Usually done every 25 years or so.

> To prevent this, a big part of railway maintenance involves digging out all the stones, washing them to get rid of sand and soil, and putting them back. Usually done every 25 years or so.

Is this that? https://www.youtube.com/watch?v=WqLynLWLd8Y

I think that the title suggests that is repositioning the stones (ballast). This video seems to be of the cleaning process:

https://m.youtube.com/watch?v=XWjUZJtR610

I could be wrong, I had never heard of this fascinating maintenance until this very moment.

I love the response:

>BOURLEZ RENOVATION

9 months ago

Why is it?

Tamping is also of critical importance when tracks are first laid; they’re quite uneven at first (visibly wonky) and the tamper evens out the ballast to achieve the extremely smooth geometry required.

I know nothing about this topic at all but, having browsed Wikipedia for a bit, the machine from the video looks very much like a tamping machine https://en.m.wikipedia.org/wiki/Tamping_machine (The video title also says "tamper")

In the comments they say that "tampering"[1] is done once every year and the wiki page does not mention cleaning as an objective. Tamping is done to readjust the ballast making the track more more level and durable. (One can infer it does some cleaning as well - getting rid of the topmost dust layer)

https://www.youtube.com/watch?v=zNFUgTJR6jw seems to be the full cleaning procedure

[1] https://en.wikipedia.org/wiki/Tamping_machine

According to the description of the video you posted, the ballast cleaning machine is then followed by a track laying machine, https://youtube.com/watch?v=tMXfU8blPMM . What an incredible machine!

> they say that "tampering"[1]

I believe you mean tamping. :-)

Looking at that machine I'm amazed that we as a society are not advancing further. We have incredible engineering feats like that, yet at the same time, we have rampant poverty, homelessness, polluted oceans, smog clouds, and thousands of others.

That thing is a moving synchronized orchestra of a mini-factory, carrying it's own inventory and moving on the same railway track that it's servicing/fixing. And here I am, can barely coral a bunch of juniors to fix a basic CRUD app.

Controlling rocks is easy. Controlling people is hard.

The various cults make it look easy.

They target the most vulnerable people.

I’ve known several people who have been recruited into a particular death cult and they are/were very high performing individuals with very stable backgrounds ie strong social network with plenty of resources of all kinds. Confident upper middle class successful entrepreneur types. The attack vulnerability was simply some level of narcissism and desire to belong. These are universal human qualities.

You haven’t described accurately who is at risk from cults, at all. Think of the idea that every religion could be reasonably described as an established cult.

Austrians have a long history of coming up with efficient engineering projects involving trains.

Worth noting though that until more recently they had some challenges focusing on the right goals.

No, that's a tamping machine: it packs ballast to level tracks. See https://en.wikipedia.org/wiki/Tamping_machine

That was more insightful than the whole article. Another thing I was curious about but not answered from the article is if there are any alternatives.

Yes, there is. High speed rail quite often uses ballastless track [0]

[0] https://en.m.wikipedia.org/wiki/Ballastless_track

TL;DR "ballastless track" is track rigidly set in concrete instead of the aforementioned floating, but it must be done to very tight tolerances and is expensive

If the tracks are rigidly set in concrete, how do they deal with thermal expansion then? Roads made from concrete usually have gaps every couple of meters, but you can't have that many gaps in railroad tracks I suppose.

Either way they buckle: https://www.networkrail.co.uk/stories/why-rails-buckle-in-br...

https://www.youtube.com/watch?v=_LoXgN1QWZM

https://www.youtube.com/watch?v=dqq0543JWzI (wiggle wiggle)

First of all, in the images shown in that article the rail clips are rigidly set in concrete, and the rails are clipped in place. With enough force they can slide longitudinally in the clips.

Second, in most modern rail installations the track is pre-tensioned to deal with thermal expansion. The track is stretched with a hydraulic tool and welded while under tension, so thermal expansion just means reduced tension. This is why trains on modern rail lines don't make the familiar click-clack sound.

(How much tension is applied depends on the country and weather - you can choose between tolerating high temperatures and tolerating low temperatures. This is why in the UK, there are sometimes train stoppages in very hot weather)

>trains on modern rail lines don't make the familiar click-clack sound.

What does the click-clack sound have to do with thermal expansion? Is it because the track is expanding as the train moves over it, if so what’s causing that sound signature?

It comes from the wheels passing over the joints in the track segments. Track that isn't continuously welded is made of individual pieces of rail joined at the ends with a fishplate:

https://en.wikipedia.org/wiki/Fishplate

The expansion gaps.

Theyre not set “in” concrete. The rails are still connected to the sleeper/tie via a fixing plate/spring/shoe, allowing horizontal movement between them. There are still gaps with the concrete bed for that thermal expansion as well.

No. I think your understanding is faulty. The rails are hundreds of meters long: the longitudinal expansion would still have to be accounted for somewhere.

The rails are firmly fixed, longitudinally as well as laterally, with pretensioning in order to minimise stress as they heat up. This is true with ties as much as with a concrete bed.

https://interfacejournal.com/archives/644

"There are a number of ways to increase the resistance of rail to buckling. The first and most common is the use of rail anchors to limit the longitudinal expansion of the rail. Anchors come in a variety of designs, but they all apply several thousand pounds of clamping force on the base of the rail. When applied properly against each side of the crossties, the anchors prevent expansion of the rail in the longitudinal direction."

The general approach is described here: https://en.wikipedia.org/wiki/Rail_stressing

Interesting! One argument I see against hyperloop is the need for expensive vacuum bellows to deal with thermal expansion. I wonder if there is a locked in place solution like this that could eliminate the need for them.

I don't think you could just pre-tension rails so much and be done with it somehow, considering... temperature of a rail can probably stretch between -10 to +50C, and coefficient of expansion for steel is 12*10^-6/K, let delta T = 60K and joined rail length = 1km, the expansion can be +/-36cm(~1.2ft). At some point, the error can be more than one rail segment worth, and maybe it's just me but I can't imagine that amount of expansion just solved by clever tensioning alone.

Usually they are joined to a rail with a slant cut or a tongue section at ends, and let ride into the opposing cuts/tongues, IIUC.

The pre-tensioning isn't what prevents expansion.

The rail can only expand if it can move. If the rail is clamped down tightly to the sleepers, and the sleepers are tightly packed in the ballast, then it just doesn't expand. The only thing that happens is that stress increases / decreases as temperature changes.

The pre-tensioning is only to ensure that the average tension in rail in the course of an average year is in a certain range.

wtf.

To add on, a lot of underground mass rapid transit uses ballastless track, too.

And street-level light rail, for the obvious reason of sharing the street (even if not lanes) with other vehicles, as well as aesthetics.

> if there are any alternatives.

There are. There are ballastless tracks that utilize a continuous concrete slab instead of the crushed rock ballast. Their main benefit is that they are lower mamintenance than their counterparts, but their upfront cost tends to be higher and construction takes longer, making them difficult to retrofit onto existing stretches (would require long closures).

However, the low maintenance nature of this style makes it popular at locations where maintenance is very difficult (long tunnels, etc) or where loads are particularly high, either due lots of use or high speeds, like with new HSR construction.

High speed rail track also don't use expansion joints, the rails are continuously welded, which means that instead of expanding/contracting with seasons, they pack quite some tension or compression, which is basically worn by the strength and friction of these concrete slabs.

They do use expansion joints they’re just a little more modern now. Look up “breather switch” for examples. Another common technique is to install the track already heated so it can’t get even larger and buckle.

They only use them where needed for other purposes (i.e. electric isolation). They don't require joints for thermal purposes.

They do (sometimes) preheat continuous welded rail during installation to prevent buckling. The temperature used will vary based on the local environment (Minnesota will use a lower destressing temperature than Florida). Shorter sections of track (like in a switching yard) may not be destressed at all because there isn't enough length to cause a buckle.

> Compressive forces result from stresses induced in a constrained rail by temperature above its "stress free" state, and from mechanical sources such as braking, rolling friction and wheel flanging on curves. The temperature of the rail at the "stress-free" state is known as the rail neutral temperature (i.e. the temperature at which the rail experiences zero longitudinal force). Initially, the rail's installation temperature or "anchoring temperature" is the rail's neutral temperature. Hence, at rail temperatures above the neutral, compressive forces are generated, and at temperatures below the neutral, tensile forces are developed. Track maintenance practices address the high thermal load problem by anchoring the rail at (neutral) temperature of 95 -110 F. This high neutral temperature range prevents the generation of excessively high buckling forces even when the rail temperatures reach 130 -150 F.

https://www.volpe.dot.gov/infrastructure-systems-and-technol...

Some railroads that experience wide summer vs. winter temperature differences may choose to install expansion joints. Like most things, there are trade-offs, like increased maintenance costs and increased chance of derailment at the joint.

Welded rail buckling in hot weather has become a real issue with global warming and wider temperature swings.[1][2] Australia has especially bad problems, due to wide temperature swings and very long runs of straight track.

[1] https://toolkit.climate.gov/topics/transportation-and-supply...

[2] https://thelincolnite.co.uk/2015/07/kink-in-track-freight-tr...

There was a lot of reporting regarding the same issue in the UK last year

https://www.itv.com/news/anglia/2022-07-19/why-cant-uk-roads...

When I was in London in the early 2000's, the Central Line was closed because of the heat damaging the tracks (they never used the word buckle in their announcement, but that was probably what happened). If they have a track problem that far under ground, it's a serious heat wave.

> If they have a track problem that far under ground, it's a serious heat wave.

Unintuitively, parts of the London underground are notorious for being extremely hot. Especially the central line is super hot due to the material the tunnels are dug into being an extremely good insulator/heat store.

https://www.forbes.com/sites/lauriewinkless/2017/06/22/sweat...

It's also that they don't cool the tunnels like they really really ought to (just feed frozen water into the tunnels and drag the meltwater back out if it won't drain by itself, say by fitting some cars with slush sprayers like how manure is sprayed from the back of a tanker on agricultural fields, or dropping ice cubes spread-about).

Use ice-based AC for the passengers, not heat pumps dumping into the tunnels.

Change the breaking resistors to not be on the train, but the rectifier stations, and make them vent outside. Or change them to dump heat into what's a big insulated kettle that gets filled with ice or at least disposes the resulting hot water outside, not hot air into the tunnels.

Just curious: what's the benefit of installing them pre-heated vs installing them at air temperature with a slight expansion space between them? I assume that when they are installed pre-heated, the first thing they do is cool down and contract.

Expansion gaps limit the speed the trains can run at.

Isn’t the main issue that they cause a lot more wear on train wheels, making it a very bad idea to run them at speed?

Must be something like that, we used to run 120 MPH on bolted rail and steam.

Perhaps they just go into tension when they cool? I believe tracks would have much more strength in tension than in compression.

They do indeed. Railway tracks under compression from overheating are prone to bend like spaghetti (this is bad!).

One reason why inductive breaking (Wirbelstrombremse) is only allowed on "feste Fahrbahn" (ballast less track) for Germany's high speed ICE (inter City express) trains, at least as a service brake.

On some ballasted track they're allowed for emergency breaking ("Schnellbremsung"), as they can interfere with equipment like axle counters.

This is done because they have to wait for the track to cool before they are able to safely drive on it after it got heated.

These are btw just a linear induction motor with the field coils locked to standstill.

I have since wondered about using this for traction, as it's likely not too efficient but ought to have potential for a much simpler drive train, using much longer parts of rail under the vehicle to get the desired force with a low slip (and thus low rail heating).

Say, for vehicles that don't normally need (this much) acceleration, and would thus prefer few/no driven axles.

We are at a point where power electronics are cheaper than some traditional transformer technology, and being able to deliver high accelerations (on the order of 3~6 m/s²) near zero speed to regional trains could vastly improve their average speed.

Just make the people have a wall/backrest and those forces are harmless. Turn them around for the next stop if you want to brake that quickly or use it to catch up a small train car to a full trai;, have two stories, walkway backwards on top, forwards bottom, and drop the tail off from the through train after it has exchanged people. Don't need seats there, just walls at, like, shoulder width pitch, to let people stand with their back in the right direction.

If only track pairs were more cleanly parallel, you could just have bridges and let people transfer to and from the fast through train using a local access facilitator train. That in particular would also let you catch a connecting train without either of them slowing down (a lot) to provide this opportunity, which might even be the bigger benefit.

Sadly trains take ages to board lengthwise as European track clearance gauge doesn't allow for two proper unidirectional lanes of people traffic with decent cross section left for seating, so you'd have to spread the arrival/departure shuttles between front and back for a single dominant people traffic direction (from arrival seat to their transit seat and onwards to their departure seat).

I guess you could work around by using distinct small shuttles that maybe briefly combine for aerodynamics, in relative breaking distance w.r.t. service brakes, and respond to a switch that doesn't reach a safe locked end position by unfolding wave breaker barricades into the gangway (articulated so that they naturally support people leaning against them due to deceleration forces, so they get kinda pushed into place by people using them) and alerting passengers about the imminent rapid coordinated deceleration. A couple seconds notice should suffice to let people move to hold onto their stuff that'd otherwise go flying from a G of deceleration.

Coming to a stop from 300km/h takes 2.3km (27.8s of travel) at 1.5m/s² (the AFAIK limit for normal standard train carriages to match how the passengers behave), but at 10m/s² it only takes 247m (4.2s of travel).

This would be the minimum dead space/time to stop safely in case a switch fails to transition. Regroup convoys between branching points a bit, and you could run at current-day capacity levels despite most carriages not stopping at any given station (the others just skip past as they didn't branch off).

Honestly I still wonder if suspension rail system Eugen-Langen isn't better for high speed people transport due to the severe passive tilt capability (+-15° in production for a century; +-30° tested (the production deployment wasn't authorized to anywhere near it's tilt angle speed limit for many years, and the frequent stops/stations didn't make that angle look restrictive for the technology at the time: the track isn't even banked as far as I know!)), especially because you don't typically want to share track between high speed people transport and generic mainline rail traffic.

Not needing the extensive tunneling/wide-span bridging ought to make the track suspension requirement worthwhile... Particularly with how cheap we can make steel truss sections in automated factories, and e.g. sling it under the track for transport to installation site. Drop it for the irregular junction areas, and just put cheap normal rail to wheel the segments across the junction.

Roller coaster technology has made the needed fast track switches a proven technology, too. (It's really similar to a suspension rollercoaster just with motors and an electric "3rd" rail.)

I for one can't wait for full ETCS Level 3 to become a thing used in production.

Expansion gaps, generates the "classic" train noise.

I believe it's also louder, with a worse ride.

At least anecdotally I can say worse ride performance is not true. Lower noise emission is hard to tell as a passenger, but seems to to be the consensus in the literature I found. Thats probably among the reasons these tracks are usually only installed on stretches not running through any towns/cities.

Nah, just use noise absorber instead of bare flat concrete; e.g. https://en.wikipedia.org/wiki/Escher_Stra%C3%9Fe_station shows some tactics for silencing ballast less track (not in the wiki article itself, sadly).

I have run a locomotive in a tunnel on concrete slab. Low temperature range for the air. The air had high humidity due to persistent water leakage. We didn't do much track maintenance and simply went slowly on bad sections of rail.

I see this type of ballastless tracks on the big train stations on Madrid (Spain)

The alternative is generally some form of ballastless or slab track – https://en.m.wikipedia.org/wiki/Ballastless_track - essentially rail fixed directly to a concrete substrate using a variety of different approaches. Lower maintainance costs at the expense of higher installation costs, and most often seen on high-speed systems, tunnels etc.

There are also various forms of baulk track – https://en.m.wikipedia.org/wiki/Baulk_road - in which the rails are supported longitudinally with regular tie-bars to maintain gauge. I don’t think this is used much outside of some specialist applications.

> To prevent this, a big part of railway maintenance involves digging out all the stones, washing them to get rid of sand and soil, and putting them back.

This is called "ballast cleaning" if anyone wants to do a search:

* https://www.youtube.com/watch?v=zNFUgTJR6jw

* https://en.wikipedia.org/wiki/Ballast_cleaner

Cool video. That machine is incredible.

Wild.

The workers seem to be wearing the amount of PPE that I'd want if I was just walking along downwind, but nowhere near the amount I'd want if I was working in the thick of things 8 hours a day. I wonder what their lung disease rates are like.

The ones I see are wearing N-95 masks or equivalent. I'm no expert, but based on what I've learned since early 2020 I'd say that's plenty, provided they only use them for one day and toss them out.

Steam locomotives, at least, used total loss lubrication. Their brakes were made of asbestos. And a good preservative for railroad ties was creosote. Never mind the stuff they spray to keep vegetation at bay. So even I would probably be wanting to wear an N95 mask if exposed to dust from that stuff all day.

> And a good preservative for railroad ties was creosote.

Still is. There is even an EU exception to the ban because no replacement was good enough.

Pro tip: never use railroad ties or any part of a railroad tie in your campfire.

Don't use any kind of treated wood.

In this case it's a high speed line which will never have seen a steam locomotive, and where most trains will use dynamic braking (returning the power to the electricity grid, or else dumping it into resistors).

I think I made a Factorio like this once.

Prior to this machine existing, was ballast just not cleaned and there were more frequent derailments?

According to the Wikipedia article:

> In the 19th century and early 20th century, ballast was shoveled or forked, then screened by hand using portable devices.

"Here is 1000 miles of railway. Each yard of that railway has 1000 lbs of crushed rock under it. Your job is to pick it all up, put it through this sieve, and put it back. I will employ you for $1 a month for 50 years to get the job done. Do we have a deal?"

Throw in room and board and you've got yourself a deal!

Mostly the railroads employed thousands of immigrants from China, treating them pretty much how you'd expect Americans to treat migrant labor. https://americanhistory.si.edu/american-enterprise/new-persp...

"Americans"

You have a valid point, they were treated like shit but talk about broad-strokes...

As an American, I am not afraid to acknowledge the true history of my country.

Did any country at that time period treat migrant laborers better?

Ha. That takes me back to a job I had once ...

No-one asked, but ... I worked at a power station one Summer, my job was to clean the filters at the bottom of the cooling tower. The filter was a stack of plastic bricks; each brick about 1m x 1m x 2m that were a mesh of orthogonal perforated plastic panels with a spacing of maybe 8cm or so. The filter was stacked about 5 "bricks" deep. We washed the bricks under a running hose of about 20cm diameter, standing in water all day; shaking out the mud. The bricks were heavy because they were full of what seemed like river mud. The detritus from the bricks fell down into the pond below the cooling tower, about 20m beneath us. Small skid-steer cat loaders dug out the pond.

I had a runner's build and was being paid £2.70 an hour (late 1990s); the other guys had weight-lifter builds and were on >£10 (they travelled around the country doing the same job, there wage was good money for low-skilled work at the time). I found out that the agency were being paid about £10 for me ... stood in dirty water all day, with the fall risk, the low wage, ... wasn't great. Then I had to quit because of injury after a few weeks. Knee was never the same.

That's not a filter, that's structured packing to give the water a large surface area to evaporate from. Or possibly used as a drop catcher with spray nozzles underneath; either way the gas stream hits the surface without excess wasteful turbulence.

I’m guessing this machine ate hundreds/thousands of jobs

And took hundreds to design and build

The railroad tie/“sleeper” page is equally interesting:

https://en.wikipedia.org/wiki/Railroad_tie

That's probably like one of the coolest purpose built machine I've ever seen. In my head I originally thought they did it by hand, lol.

You'll love the railway track laying machine: https://www.youtube.com/watch?v=tMXfU8blPMM

Yikes, that guy just sitting underneath the moving gantry loaded with tons of sleepers... And then the other fellow crawling in between the wheels of the whole machine _while it is moving_... Do they not have something like OSHA wherever this was filmed??

Also, disappointing they didn't actually show the new railstock being laid, just the ballast and sleepers :/

That was filmed in Germany, I would think safety rules are similarly to the US. Maybe it's not as dangerous as it looks at first glance.

Definitely! A train on the train, lovely real yo dawg (:

Great video. Most neat parts for me:

- Removing the ballast from a section at a time to do the cleaning means that the rails and sleepers are floating in the air (~1:30 to 2:35)

- They're doing all this work, including transferring waste to another train, underneath catenary wires. low clearance!! (3:40 to 5:20)

Also important to note, they aren't so much washing the stones (though that may be part of the process), they're sieving (grading) the stones to remove everything that's under a specific size (looks like they're removing everything that passes a ~25mm/1" sieve, but scale in a video can be deceiving). Rail ballast is very uniformly graded, i.e. every piece of aggregate is about the same size. Vibrations over time will cause edges of the stones to chip and wear down, with these chips contaminating the ballast. It also leads to rounding of the aggregate, so there's less interlock between the particles.

reminds me of how specific they were with crushed stone sizes while developing the first roadways for cars

I don't know much about the earliest car roadways, do you have a link with some info?

Granular gradation is still a major part of modern roadway construction, there are just certain standards that are very typical so they get used without too much thought required. For some applications a visual inspection is enough, but there's generally some standard proportion of sieve analysis required (eg. 1 test per 10000tonne produced), along with proctor tests (to determine maximum dry density) and frequent nuclear densitometry. I've even seen loads of aggregate get rejected due to the large aggregate being too oblongated.

When you get into concrete or asphalt mix design, the aggregate standards can be even more strict - chemical makeup of the stone can be important to avoid alkali-silica reaction; minimizing fines increases the strength of asphalt mix but creates a tougher riding surface; etc etc etc.

See https://en.wikipedia.org/wiki/Macadam#McAdam's_methods, but it's the first roadways for carriages, not cars.

> A workman could check the stone size himself by seeing if the stone would fit into his mouth.

The sequencing was fascinating!

I'd never considered how much "it helps to have rail, to replace rail" impacted the process -- but it makes sense as everything is incredibly heavy! (Rails, sleepers, aggregate)

From that perspective, of course you'd excavate and clean the track ballast while you're also on the same track. And same with sleeper and track replacement.

We don't see any tank wagons, and there would be no other possible water supply, so I don't think they are washing the stones. Given it gets rained on anyway, I doubt there would much benefit.

Here's a bit closer look at the mechanics: https://www.plasseramerican.com/en/machines-systems/ballast-...

SSL cert expired 1 day ago on this linked page FYI

The next video in my feed explained a more complete rehab, including replacement of the sleepers (ties) and rails while allowing the adjacent track to stay in service. F’n amazing.

https://youtu.be/7zsyMxRnQ68

Awesome. I didn't expect my country to appear as an example for such an engineering masterpiece. Thanks for sharing.

Unreal. I was stood on the platform two days ago wondering why tracks have ballast and how it was kept so clean, and today I'm watching the most incredible process and machinery that answers my question. I have a totally new respect for train tracks now.

I worked for different parts of the national railway company of my country (France) for four years (circulations management and scheduling mostly - briefly on a CBTC system and the first studies on autonomous driving). It’s hard to understand how complicated it is to keep trains running if you haven’t experienced it.

I wonder how that will compare to keeping Starships running to/from Mars.

Aviation and Space has a lot of challenges, but also a lot of flexibility.

Planes, for example, can share the same geographic location -- just at different altitudes; e.g. 25000 ft vs 33000 ft vs 45000 ft., etc. After a certain altitude there isn't anything. No clouds, no birds, just air; set your auto-pilot and go. Takeoff and Landing can be hairy, and you gotta worry about mountains, but that's about it. A basic radar makes it easy to see, and not hard to broadcast by radio.

Same basic idea for space. Gotta get out of orbit and not get cooked by radiation, but otherwise there is a whole lotta empty, and it's not hard to see what's going on with things like radar.

On the ground it ain't so simple. Someone stuck on a rail, avalances, bridge collapses, hills vs inclines, snow / ice / rain / high-winds, and no ability to multiplex, so scheduling gets even harder since there are only so many ways into or out of a location. Trains full of high explosives or toxic chems drive right through downtown areas, often multiple times a day. From a crime and security perspective, airports are locked down pretty hard, while most people could probably just drive to a siding near them and fuck around; you can derail a train with a $400 block of steel you can order online. Communication and signal loss is also way more complicated, esp. if you're going through a mountainous or remote area.

What a great rabbit hole. Thanks for sharing!

That's a quality job. I've been in a station when a Caltrain ballast replacement train came through. Caltrain just had a hopper car full of ballast rock dumping on new rock. A railroad tie was being pushed by the wheels of the hopper car to smooth the ballast. I was amazed that didn't result in derailing the hopper car. It was really low rent. I assume they followed up with a ballast tamper at some point, but it wasn't part of the ballast dump train.

Caltrain outsources much of their maintenance of way work, and it shows.

They'll[1] usually follow[2] that with a machine that picks the rails and ties/sleepers up and lays it back down on top of the newly-dropped ballast.

[1] For some value of "they". I don't know if Caltrain does it that way; most mainline US railroads do.

[2] That is, within a few days. Not necessarily within the time of someone hanging around to watch.

Here's a video of that operation.[1] Yes, they really did level the ballast by putting a spare railroad tie ahead of the wheels of the hopper car. This was before the upgrade to concrete ties.

[1] https://www.youtube.com/watch?v=UWvuYdVV5-Q

That was 23 years ago, but still looks low-rent. Here's how Brightline & the Florida East Coast Railroad are doing it in Florida for their double-track (mostly) expansion to Orlando.

https://www.youtube.com/watch?v=vINzBkK1d-k

Note that the ballast hoppers have a solar panel on them to charge a battery, which runs a motor to actuate the hopper doors. Much better than a couple of guys with a wrench.

Thanks for that video link, curious how they do this on single track. Perhaps pick up the waste further down with the dump cars on a siding?

looks like dna transcription

How the heck did we figure this stuff out? Honest question.

And this is just the kids stuff! Imagine getting people to walk on the moon.

Left foot, right foot. Repeat.

Stupid question from a software person: Why don't they just pour cement, epoxy, or tar over all the stones to fix them in place, fill in the gaps, and prevent dirt from getting in them?

How does the water drain through them?

Again here's my software mind:

(a) If you fill all the gaps, water can't get in in the first place, and it doesn't need to drain. Abstract out the problem.

(b) If water can still get in, it can also get out.

Except sometimes water falls out of the sky and it needs to drain somewhere.

The rocks also need to be able to move a bit for elasticity, and have sharp edges. For that you also replace a part of the ballast every few years. That's hard to do when everything is sealed. Even with such "glue", the lifetime couldn't be eternal, after all.

water snow and ice will readily destroy anything solid, better to drain it away - this is partially why roads need constant repaving

Three is this thing called slab track, and that's essentially what you are talking about, just without any stones.

https://grn.ams3.digitaloceanspaces.com/wp-content/uploads/D...

Truly almost all new lines should be built this way, its simply a far, far better method. Slightly higher capital cost but far less maintenance.

Apparently, trains do not derail that easily.

Army Experiments In Train Derailment (1944) https://youtu.be/agznZBiK_Bs?t=15

There was a famous train accident in England thanks to dirty ballast stones, bad drainage, and rainy day.

Thank you! I saw one of these yesterday and I couldn't work out why it was such a large machine given that the only thing it seemed to be doing was moving ballast slightly away from the tracks.

This gets its purpose regarding plants a bit wrong, the only time roots would be any issue is on tracks which have not been used in quite awhile and at that point the tracks will almost certainly have other issues which need to be addressed before they get used again. The metal on metal of the wheels on tracks often causes sparks and these do occasionally cause fires in dried vegetation alongside/on the tracks which the ballast helps to keep in control, which is why the ballast generally extends further from the edges of the tracks than is needed for stability reasons. If you work track maintenance part of your job will be torching any plants growing too close to the tracks especially on grades where wheels often slip and brakes applied which can produce a surprising amount of sparks. They also have cars which spray weed killers alongside the tracks which run the line once or twice a year depending on conditions.

In the Netherlands there is an ongoing debate on the use of ballast stones giving off quartz dust that can cause lung disease and cancer.

https://www.dutchnews.nl/2021/04/prorail-is-jeopardising-wor...

This concerns quartz dust that comes with the stones from the groove, as well as new dust from putting the stones into place.

Well literally everything causes cancer to some degree and any particles one breathes in that aren't air can cause lung problems.

The real question is probably "do enough people get sick from it that it's more expensive for the rail company to pay out lawsuits than the track to be replaced".

Maybe you should mention that these concerns are about the workers who "work with the stones on a daily basis, often standing in great clouds of dust without wearing masks" (maybe they should simply wear masks then?), not the general public.

Whenever there is a lot of dust, you should wear a mask. One good thing the pandemic taught us: how to use masks. And now the n95/ffp2 are cheap again.

I never figured out how to get an FFP2 mask to be tight. Air always passed on the sides somewhere.

The only mask that I've tried that is actually air-tight is a 3M silicone mask with replaceable filters. It works even with a short beard. I think only proper masks like that actually prevent the bad stuff from getting into your lungs (assuming you buy the correct filters).

That’s tricky for some people, especially for skinny people in combination with big cheekbones. Did you try the 3M masks too, in tests they seem to perform in average much better than the Asian style masks just out of two pieces laminated together. They seem to work best on round faces, which are very common in Asia.

A half-face P100 respirator is a much better choice for any serious dust. Also, surprisingly, P100 respirators are effortless to breathe in while offering superior protection.

..while simultaneously 100x your chances of being mistook for a Subzero or Scorpion cosplay. YMMV on whether that is preferred

Is this from the stones, or because the stones are polluted? There has been a lot of talk about where to put the old stones here in Denmark (another densely populated region), because if you “put” it too close to sources of drinking water all that pollution they’ve soaked up over the years will eventually get into the ground water.

If it’s that dust, and not “just” quarts dust from the stones themselves, then I wonder if switching will help. I mean, the trains are still going to drop whatever it is they drop as they ride along.

It's from the rock dust itself. Quartz is silica, and silica dust causes silicosis:

https://en.wikipedia.org/wiki/Silicosis

I wonder why railway ballast is polluted...? Has someone been running trains for years which leak loads of oil or something?

it's not; it's the dust from the actual stones themselves not a contamination.

In Denmark it’s because we still have diesel trains from the 80-90ies because their replacements didn’t work. (Don’t ask).

I’m not sure the electrics they plan on implementing will have the same issue.

Other than suburban and light rail lines, it's only the lines from Germany to Sweden that are electrified: https://openrailwaymap.org/?style=electrified&lat=56.2860592...

Also, steel wheels rolling on steel rails results in lots of fine metal powder everywhere. Also, particularly in cities, everything gets covered in small particles from a lot of sources.

Living in europe where ~once a year storms bring massive amounts of super fine quartz sand from the sahara desert, you'd imagine the amount from railway ballast wouldn't be a concern...

People campaign about these weird issues and then ignore the plastic in the drinking water and the hormones in the food…

If you run a high speed train over a ballasted track the under carriage aerodynamics suck the ballast up and they can hit the bogeys and axles and cause cracks in the train axle. If this happens in a bad way it can derail the train.

This abstract has a good summary https://research.birmingham.ac.uk/en/publications/a-full-sca...

So the question becomes can we replace ballast with some other mechanism where this doesn’t happen?

We can and do for many new high-speed rail developments: https://en.m.wikipedia.org/wiki/Ballastless_track

It's more expensive to build and fix, but it's also more reliable and needs less maintenance.

I'm pretty sure TGV tracks do use ballast though, don't they? And they are among the fastest trains in operation, so it seems to work fine for them.

Yeah the occurrence is fairly rare which is why we don’t do much about it.

And we trust that NDT(non destructive testing) like die and ultrasonic etc, periodic testing will catch any cracks.

Here’s a TGV speed test on a ballasted track. There’s an axle cam and it doesn’t look like there’s big chunks flying about but it does stir up a lot of dust at 575km/h (~360mph)

https://m.youtube.com/watch?v=EOdATLzRGHc&t=11s&pp=ygURdGd2I...

As time goes by the ballast settles and under repeated heavy loads and the underlying land becomes a little uneven. In comes the Tamper, a ballast tamping machine. This is a long heavy machine that runs on the track.Quite strong it lifts a length of track under load a few inches and tamping arms reach down and force fresh ballast under with dynamic level sensors and moves on down the track. Tracks are surveyed with sensors on trains and ballast tamping/washing/etc are scheduled. https://www.youtube.com/watch?v=WqLynLWLd8Y

I was riding the train to work about a year ago and randomly googled this exact issue. I was amazed at how much thought actually goes into designing the ballast, as I had always assumed it was just some random fine gravel.

Almost everything always has much more thought given to it than it might first appear. Civil engineering is quite a fascinating subject, not least because how old it is, by far one of the earliest branches of engineering.

> Almost everything always has much more thought given to it than it might first appear.

This statement ought to be in the guidelines for HN.

Oh, definitely.

> Almost everything always has much more thought given to it than it might first appear.

A credo to live by in work and in general - one of the most terrifying phrases in the English language is "it's just...".

Maybe I missed it but is there a type of stone that is used, or just whatever is locally available? Around where I live that would be limestone but I am guessing that's too soft for this sort of use.

During a boy scouts summer camp, we were on a separated camping area near a railroad track. We created a fireplace and used some stones from the track to build the fire on.

A pretty big mistake as it turned out. Because the stones were in fact vulcanic, they exploded when heated - so we were bombarded with stone fragments.

I watched a video just yesterday about wood ties vs. concrete ties. One of the arguments in favor of wooden ties in the US is that they are less brittle and thus more able to handle the frequent derailments we get. The nightmare scenario with concrete ties is that a loose wheel goes unnoticed for miles and cracks every single tie.

They mentioned a previous test experience with older concrete ties held together with naked angle-irons. A derailment severed all of the irons.

https://www.youtube.com/watch?v=UtyTTijswwg

Fascinating how such a simple solution is so timelessly effective. A "framework" going two centuries strong.

For the majority of track, this is fine.

High speed track uses ballastless track because at high speeds, making gravel fly is very damaging. (It also tends to be noisier and less smooth.) https://en.wikipedia.org/wiki/Ballastless_track

Appreciate the link, could you edit the `.m` out of the URL so it doesn’t point us all to the mobile version? Ta.

> Therefore, ballastless tracks must be concreted within a tolerance of 0.5 millimetres

Wow.

It would be even better if Wikipedia themselves fixed the problem. Here's how I would like to see it: There is only one URL per article, not a desktop URL versus m-dot URL. When you land on the page, if you have no cookie, then Wikipedia autodetects based on your size/orientation/device and gives you a desktop or mobile layout. If you choose to override the presentation, a cookie will be set, and all future visits will use your preferred layout.

Personally, I'm on desktop all the time and I am peeved by people who post "m." links because the mobile version hides some useful features.

In France they use ballasted track for their high speed trains. Even the newest tracks. With daily operational speeds up to 320 km/h, and with some tracks rated for speeds up to 350 km/h.

It's not impossible just has different tradeoffs. IIRC lower upfront cost but higher long-term maintenance. Also idk I just lived next to a tgv line for a few years but it's loud as a fucking jet even dug into the channels they use around residential areas.

Noise reduction techniques sometimes channel or reflect noise instead of absorbing it, so it is only quieter for some people in some areas and louder for others.

The Romans were using crushed stone in (non-rail-) road construction more then two millennia ago, and it was an old technology then: https://acoup.blog/2023/06/02/collections-roman-roads/

Rails on rubies

> Without the ballast in place, railway lines would simply not be able to cope with the stress from the weight of the trains passing over them, not to mention the impacts of expansion and contraction.

How does this relate to tram lines? They don't seem to have much ballast at all, and they are often just inset into asphalt, so there's no draining, and the asphalt presumably limits expansion and contraction?

Trams are much lighter and have way fewer carriages/cars, so I expect the need is much reduced. The foundation of the road is probably even more sturdy than the ballast, and there are gaps between rail sections for expansion.

The speeds you can reach on traditional tracks embedded into streets is much lower.

Outside of cities where trams run at higher speed ballast is preferred.

Modern tram tracks have a steel-enforced concrete base under the street, similar to ballast-free highspeed tracks.

Long-haul rail lines are engineered to support freight loads, which can be extreme. Local passenger lines have lighter cars that are mostly enclosed air, and commensurately lighter engines as well.

There is a longer article https://www.ejrcf.or.jp/jrtr/jrtr15/pdf/f38_tec.pdf from a series on Railway Technology from the Japan Rail and Technology Review https://www.ejrcf.or.jp/jrtr/technology/index_technology.htm... . It contrasts ballasted track with slab track and purports a 9 year breakeven on the higher construction cost but reduced maintenance cost. It notes that the main Shinkansen from Tokyo to Osaka is ballasted track and I understand it still is. Mind you how you could ever changeover on this super busy line is going to be somewhat difficult.

With time i've come to realize any piece of engineering can be a source of inspiration, even to a software dev / architect like me.

I feel like it took me a lot of time to figure this out, as i've got 20 years of experience. Anyone else ?

Wait until you get into car engines or mechanical watches, then you are truly in the rabbit hole!

Or spend a few coffee breaks with the team leader of a particle accelerator during commissioning...

I am working for a railroad company at the moment, we learned cool stuff during onboarding.

Modern train have an electromagnetic emergency brake system that uses a large magnet lowered onto the rail to stop.

Also the standard braking system using fluids has a clever indirection where it does not activate the brake when pressure rises but when it drops. This is because when a waggon gets loose and the braking line rips and fluids get's out, the waggon should immediately brake as soon as pressure drops.

I'm curious why all brakes aren't designed like this.

It's an exaggeration to claim that ballast would have been used for railway tracks for centuries as the article claims. Railways haven't even existed for 2 centuries. Although there are only 2 years left. Still in the 1970s, only 50 years ago, Finland had railways build with weaker material, small round stones.

> It's an exaggeration to claim that ballast would have been used for railway tracks for centuries

"At Whitley in 1704, when the word first occurs in a railway context, clinker from the salt-pans was used 'for the ballast of the waggonway'"[1]

By my reckoning, 2023 - 1704 = 319 which I think comfortably fits into "centuries"?

[1] https://www.railscot.co.uk/articles/A_brief_history_of_railw...

Of course it always depends what borders you draw.

I started with steam railways. Of course horse driven railways existed before, but they were of a completely different load class.

I don't know what clinker from salt mines looks like. But I guess it's significantly different from sharp, crushed rocks the NZ article describes. Also the Scottish article says that just some sentences later that ballast had completely different properties those days.

Some ballast has been used for centuries. But not the kind consisting of sharp, crushed relatively coarse rock pieces we know today.

> Of course it always depends what borders you draw.

You seem to be rapidly redrawing yours in a different place all of a sudden.

"It's an exaggeration to claim that ballast would have been used for railway tracks for centuries"

You may well have meant "ballast (specifically sharp crushed rocks as they use nowadays)" but that's a very different statement.

> Some [X] has been used for centuries. But not the kind [...] we know today.

I mean, yeah, truism.

Ah sorry, I thought I had written ballast of the current form. That was my intention, but I failed to type what I had thought.

Otherwise there wouldn't be a point to mention that 50 years ago other, less stable ballast was still in use in some places.

That's a wagonway, the wagons would have been pulled by horses.

Steam powered railways were only widely deployed from the 1830's

Right. I referred to Stockton - Darlington 1825.

> Steam powered railways were only widely deployed from the 1830's

Sure but GP didn't specify steam powered / iron-railed / commercial usage / whatever kind of railways in https://news.ycombinator.com/item?id=36184243.

GP suggested that "ballast hasn't been used on railway tracks for centuries" (incorrect) because "railways haven't even existed for 2 centuries" (incorrect).

Relevant machine that is called Gleisschotterbettungsreinigungsmaschine in German. Nice video for children about it (in German, there's probably an inofficial YT video where you can switch on automatic subtitles): https://kinder.wdr.de/tv/die-sendung-mit-der-maus/av/video-s...

> Gleisschotterbettungsreinigungsmaschine

39 letters..

It looks funky, but it's just railway track gravel cleaning machine, without the spaces.

Modern high-speed tracks don't use ballast stones anymore though

Depends on the high speed rail system. Not all use slab track.

The low pressure air below passing trains travelling over 300km/h can cause "ballast lifting". i.e. Where ballast can be lifted up and become mobile causing safety issues, as well as accelerated deterioration of the track.

High-speed trains therefore use concrete track beds.

That’s terrifying.

Travelling across France I soon came to dislike going under bridges and tunnels and the compressive effect it had on air as you hit them.

That it’s also capable of sucking up rocks came as news to me, but it isn’t really surprising.

> Travelling across France I soon came to dislike going under bridges and tunnels and the compressive effect it had on air as you hit them.

modern high-speed trainsets are pressure sealed for precisely that reason

There are a lot of parallels with modern hardscape construction. Nowadays many builders grade the soil base, add a woven stabilization fabric, then add clean angular stone, and finish with a finer angular stone on top. I wonder if railroads are using any geotextile fabric under their stone these days.

They do, or at least I've seen it in some railway renovations here in Argentina (through posts in SkyscraperCity forum). The funny thing is, when the state runs out of money, they start to cut corners and the geotextile fabric is one of the first things to go.

That’s sad and hilarious. Geotextile fabric is so cheap and adds so much longevity to these type of projects.

The stones must be good reflectors of noise, which makes the ballast rather undesirable under trams in cities. I wonder why is it still used in such circumstances instead of having noise-absorbing soil and grass under and around the tracks as is sometimes done.

Ballasted track is cheaper to build, so somebody needs to be willing to fund the difference in costs. This especially applies to existing infrastructure, where it's more likely than an upgrade to green track might have to be funded by the municipality alone.

Several comments mention track buckling in extreme heat.

When this happens, how is it resolved ? Does the track usually revert to its original positioning at all points, or does every stretch of recently-buckled track have to be inspected and possibly realigned ?

It depends on the severity of the problem. The track may satisfactorily straighten itself out as it cools. It might mostly straighten itself out, but there could be lasting geometry problems from it settling differently than it was before. This will need to be corrected by a tamper, which basically vibrates the ballast under the tracks so it "flows" as the track is pulled into the correct position. In especially severe cases, the track may need to be inspected and have some hardware replaced---things like joint bars and track fasteners.

Tampers have a reference system based on axles far apart whose relative positions are measured by a simple optical setup. So especially with modern computer controls, the operator can program the correct track profile into the tamper and then it slowly advances along the track pulling it back into position. Since tampers are slow and there are only so many around, a lot of railroads have work trains or even hirail trucks with optical equipment to measure the track geometry so that they can inspect for tolerances more frequently. The exact tolerances depend mostly on running speed... a train moving very slowly can often handle even badly out-of-alignment track as long as it is well attached to the ties (so the distance between the rails stays correct). So you will often see industrial sidings in really bad shape since they're only used occasionally at 15mph. But once trains get up to 50mph and faster they become a lot more sensitive to the geometry and you start worrying about damage and derailments if things are out of whack. So the mainline track gets inspected a lot more frequently, often a simple visual inspection by MOW workers in a hirail is done daily besides of course train crews reporting areas with a rough ride.

Speaking of it being more of an issue at high speeds, the freight trains through our area go through far more slowly when we are having a hot spell.

I've been told it's to be more safe on the expanding possibly-not-perfectly-straight rails.

Thank you for elucidating !

A nice video with good visualizations of the mechanics of the ballest from ABC Science: https://www.youtube.com/watch?v=TlSOMfDX-yY

My first thought, on reading the title, is maybe the sparks sometimes thrown off by the trains would catch dry weeds on fire if they were near the tracks, so the crushed stones give the sparks somewhere safe to land.

Does anyone know about that device on electric engines that touches the wires above?

Why is that it doesn't spark and do the overhead cables need frequent replacement because of being constantly rubbed against?

It's called a pantograph. There is a carbon strip on the top of the pantograph that is typically replaced every couple of months as it wears. The overhead wire isn't wire like you normally see, it's more like a thick copper bar that is unwound and has a flat surface on the bottom that the carbon strip runs on. The overhead does wear out over time, but is much longer lived than the carbon strips.

Sparking I think is when there might be damage to the overhead wire that temporarily interferes with current as the strip passes over, as you see sparks happening at the same place as trains pass

Also noteworthy, the overhead wires usually zig-zag from left to right, so they touch different parts of the pantograph surface as the train moves.

Oh, but it sparks...

https://youtu.be/p5u_5iQ3Ego?t=342

(By the way, the part where it goes over a turnout at 500 km/h as if it is nothing is also pretty amazing.. https://youtu.be/p5u_5iQ3Ego?t=291)

I entered the title question of the article into ChatGPT, as I was curious whether it could factually synthesize information on the Internet similar to how this article did. Notably, it missed the detail about keeping vegetation away

<chatgpt> Railway tracks have crushed stones, commonly known as ballast, alongside them for several important reasons. Here are the main reasons for using ballast on railway tracks:

Stability and Load Distribution: Ballast provides stability to the railway tracks and distributes the load of the trains evenly. The crushed stones interlock with each other and form a stable base for the tracks. This helps prevent the tracks from shifting or sinking under the weight of the trains.

Drainage: Ballast allows for effective drainage of water from the tracks. The open spaces between the crushed stones allow water to flow away from the tracks, preventing the accumulation of water. Proper drainage is essential to maintain the stability of the tracks and prevent damage caused by water-related issues like erosion or track bed saturation.

Track Alignment and Adjustment: Ballast allows for easy alignment and adjustment of the tracks. The crushed stones can be added or removed as needed to maintain the proper track alignment and ensure a smooth ride for trains. This flexibility is especially important during track maintenance or when new tracks are being laid.

Noise and Vibration Damping: Ballast helps in reducing noise and vibrations generated by passing trains. The layer of crushed stones absorbs and dampens the vibrations, minimizing the impact on the surrounding environment and nearby structures.

Track Maintenance: Ballast provides a protective layer for the tracks. It acts as a barrier between the track infrastructure and the underlying soil, preventing soil movement or erosion that could damage the tracks. Additionally, ballast makes it easier to detect and repair any issues with the track, such as loose fastenings or track bed settlement.

Overall, the use of ballast alongside railway tracks is crucial for maintaining the stability, alignment, drainage, and longevity of the tracks, ensuring safe and efficient train operations. </chatgpt>

Aren’t roads built in a similar way, just with a different surface?

tl;dr

Crushed stones, or ballast, along railway tracks provide stability, aid drainage, control vegetation, and distribute train load.

drainage so the rails don't float off somewhere else due to erosion. not alongside them either. under them.

...so many words and just at the end they comment on 'flooding' - whereas everyone knows that a top function of ballast is to prevent water buildup / facilitate drainage.

Waverly, TN had bad floods a couple years ago because their long span of train tracks around the bottom of the hill hadn't had its drainage maintained. The ballasted, built up tracks trapped a large volume of water which blew out finally and caused catastrophic flooding.

The sheriff said he'd been complaining to the railroad for 2 years about the water building up, trying to get them to clean out the drains. That's an incredibly busy line tho.

After the floods they had the train tracks back up and running in weeks, before some the roads through town. I'm told they built exactly the same plan that had been there before, inadequate drainage and all.

"drainage" as in primarly preventing fouling of ties, not as in "preventing water buildup behind a trackbed". I would call that obvious, but maybe i have too diverse interests as to nit know this...

For permitting regular water flow tru a trackbed: what culverts etc. are for.

Culverts, aqueducts, etc, yes. It all has to be maintained. Even when it's designed well (and the current design is minimal); all those measures have to be kept clear of obstructions. In this case they were not, and very likely still are not.

It's purely aesthetic. The sharp edged stone contrasts nicely with the polished rail and the dark woody tie.

Reply on news.ycombinator.com