User login

You are here

Tragic train accident in Milano for a broken rail: 3 killed and 46 injured

Mike Ciavarella's picture

There has been a tragic accident in Milano.

See here.  Or here.

2 / 3

Slide Show



Witnesses said the train trembled for a few minutes before the accident.

Rescue at crash scene, 25 Jan 18Image copyrightEPAImage captionFirefighters are battling to free passengers from the debrisThis handout picture released by the Italian Vigili del Fuoco shows firemen working on the site of a train derailmentImage copyrightAFPImage captionThe train derailed at Pioltello Limito station

It looks very similar to the famous Hatfield accident, a rail had to be replaced, but maintainance arrived was late.

The rail that broke on the Cremona-Milan line was about to be replaced. In that stretch, maintenance work was under way. This is proved by a photograph taken exactly at the point where the breakdown occurred, a couple of kilometers before the one in which the train has broken down. At that point a piece of rail about twenty centimeters is missing. Next to the track, the section of new track that was supposed to replace the old one and for which maintenance work was under way. According to a first reconstruction, the dynamics should be this: the rail yields to the passage of the first carriages, but those that leave the tracks are only the central ones. The train continues its journey, but the central carriages now have their wheels off the track. Two kilometers later they cross and it is the real derailment.

Fatigue and maintanance of railways are still a big unsolved problem.   It is all very empirical, since the times of Ken Johnson attempts of using "ratchetting" plastic model, there has not been too much progress --- in fact the early very good agreement with experiments of models of Merwin and Johnson have later been discussed by Ponter et al to be oversimplistic, and later models by Kapoor based on Tyfour experiments are also very empirical and based on not enough data to be generalized. 

All the classical multiaxial fatigue criteria (DangVan etc) fail to show a good criterion under high hydrostatic compression (see Ciavarella and Monno), also because they do not take into account of wear. and those who have attempted to model also wear require a number of empirical parameters (see Kapoor , Donzella, etc.).

Maintanance strategies are safe?  The current measuring techniques of corrugation are really precise, and predictive of wear and fatigue?

We really need more investigation, and while the academic is looking mostly on graphene and other fancy topics, these old problems remain unsolved, and regularly a new accident kills people.

Ponter, A. R. S., Afferrante, L., & Ciavarella, M. (2004). A note on Merwin’s measurements of forward flow in rolling contact. Wear256(3), 321-328.

Ciavarella, M., & Monno, F. (2010). A comparison of multiaxial fatigue criteria as applied to rolling contact fatigue. Tribology International43(11), 2139-2144.

Ciavarella, M., Monno, F., & Demelio, G. (2006). On the Dang Van fatigue limit in rolling contact fatigue. International journal of fatigue28(8), 852-863.

Afferrante, L., Ciavarella, M., & Demelio, G. (2004). A re-examination of rolling contact fatigue experiments by Clayton and Su with suggestions for surface durability calculations. Wear256(3), 329-334.

Donzella, G., Faccoli, M., Ghidini, A., Mazzu, A., & Roberti, R. (2005). The competitive role of wear and RCF in a rail steel. Engineering Fracture Mechanics72(2), 287-308.

Tyfour, W. R., Beynon, J. H., & Kapoor, A. (1996). Deterioration of rolling contact fatigue life of pearlitic rail steel due to dry-wet rolling-sliding line contact. Wear197(1-2), 255-265.

Afferrante, L., & Ciavarella, M. (2009). Corrugation models and the roaring rails enigma: a simple analytical contact mechanics model based on a perturbation of Carter’s solution. Journal of Mechanics of Materials and Structures4(2), 191-209.

Afferrante, L., & Ciavarella, M. (2010). Short pitch corrugation of railway tracks with wooden or concrete sleepers: An enigma solved?. Tribology International43(3), 610-622.

PDF icon pioltello-gazzetta-30-1-2018.pdf556.12 KB


Mike Ciavarella's picture

In Hatfield, it rapidly became apparent that the derailment was caused by a fractured rail on the outer line of the curve. Of particular concern was that beyond the first fracture, the next 35 metres of rail had broken into 300 pieces, and some 44 metres further on, another length of about 54 metres was similarly fragmented. It was clear that the original and subsequent fractures had largely been triggered from fatigue cracks existing in the rail: although the term “gauge corner cracking” was used as the first description, “head checking” and the more generic “rolling contact fatigue” (RCF) were used later. These events caused detailed inspections to be made of tracks throughout Britain in the days that followed. Many sites were located where cracks were visible on the surface of rails. Speed restrictions, some as low as 8 km/hour, were quickly introduced. Many trains were cancelled; schedules were revised to double or treble normal journey times and even then were unreliable. Motorways became choked as people switched their journey to roads and internal domestic flights became overbooked. When a journalist wrote, “no other railway accident in British history, or, I would guess, any other country’s history – has led to the degree of public anger, managerial panic, political confusion, blame and counterblame that came in the wake of the Hatfield crash. In fact, outside wars and nuclear accidents, it is hard to think of any technological failure which has had such lasting and widespread effects” [2], he was not exaggerating. 



Here, there is no report of many fragments, but just one.  However, I do see large signs of wear, meaning the rail may have been really quite old and overfatigued.   We will see what the investigation brings.


Thank you for highlighting these problems with rail (fretting?) fatigue.  Sometimes I wonder if there can be technological solution to such problems that is cost effective.  Perhaps what's needed is an attitude fix?

-- Biswajit

Mike Ciavarella's picture


particularly in this case, there was a fishplate joint which clearly failed, but was set to be replaced within few hours?  Why? The authorities do not say there was any sign of a problem, even just 2 weeks before inspections were done. So replacement was based on some preventive maintenance?   But these strategies are very empirical, and  perhaps arrived too late.


On the other hand, inspections are based on specialized perssonel + some "diagnostic trains" ------ the former could be lazy, and not do their job.  On the second, I have many things to say, do you really beleive these diagnostic trains, which sell for millions, can really diagnostic failures of small cracks by means of vibrations?   They can perhaps, if we are lucky, measure some wear and geometrical status of the rails, but not really cracks.

But my golden rule is to see things first person to make a really final assessment.  I am not officially in charge of the investigation nor in the technical commettee -- as for yet.


The "preventive maintenance" conjecture is interesting, partly because a lot of AI and machine learning companies (cough)IBM(cough) say that they can predict when a component is going to fail.  If inspections did not see anything but machine learning did, that's a massive advance!

Regarding vibrations and crack location, I don't have aa clear idea about the latest state-of-the-art but don't they do that sort of inspection on airplanes routinely?

Keep up posted on what is found if you find out more.

-- Biswajit

Mike Ciavarella's picture

Biswajit, your comment about airplanes is very pertinent.

What I can say is that in airplanes, things are done very properly in general (although human error is always possible), and "damage tolerance" is a very refined technique. I have described elsewhere how it works.

In Train, there is nothing remotely so sophisticated, and the Hatfield accident in 2001 showed that it was all based on a mass of people who admitted candidly they were "ignorant" on safety issues!

In the Italian accident, it is early to say what happened, clearly I have some ideas, not all of them I can share in public.  Do you believe in these "diagnostic trains" which run fast on tracks and pretend to diagnostic?  I do beleive they can approximately, measure rail corrugation, but defects?  What I can find in scientific literature is very scarse, I only find commercial presentations of systems that in fact are even Italian, and sold all over the world, so reported as a success story: here and here.   They also say that an inspection with the diagnostic train DIAMANTE had been done, and nothing had emerged. 

However, in July on a very close line, a derailment did occur.  What do you think? What is your impression?


I had a quick read through the articles and advertisement and couldn't find any claims of being able to find cracks using the trains (though they say something about being able to find defects but don't specify what that means).  I agree with you that at the operating speeds of the trains it will be next to impossible to separate signal from noise if one wants to detect cracks.

-- Biswajit

Mike Ciavarella's picture


a more scientific paper is for example this

you can see that trains can run even at 285kmh, while performing the diagnostics.  But these are mostly significant geometrical irregularities able to lead to significant accelerations.

Now, it is too early to say what happened in Milano (the newspapers say that a problem had been identified if a rudimental reparation had been introduced in the form of a wooden pad, but the authorities say that such wooden pads are NOT permitted by their standards).  

On the other hand, the diagnostic train had been used, and if there was a significant problem in the rail, it would have been noticed.

So obviously someone is lying here.  But if the problem were a crack (most likely, horizontal and deep under the rail, not a surface one, from the images), then it would not have been noticed most likely by diagnostic trains, nor by visual inspection.  It is all very confused and, maybe this is different from what would happen in UK or US, no academic is speaking to the press.  Only some anoninous, and the infrastructure administration managers.  So journalists are using very rudimental ways to launch articles.

What do you think?


Mike Ciavarella's picture

An update from italian newspapers is that some workers were found secretely into the closed area, running some NDT inspections (ultrasounds).  They were kept by police, and the mistery is large on what they were doing and for whom.

The impression is that the rail management is worried that there may be more cracks on the rail, nobody says the truth.

By the way, there is an interesting other parallel with another accident in UK, that which caused most of the UK rails to be replaced from jointed rails to welded ones.  Apparently on this italian line, we were still using bolted joints, since this is not the high speed train lines, but the "poor" line, which is quite neglected.

While for high speed trains, we spent 35 billions euros, for the poor lines, we were using techology similar to the third world.

Any comment?

Dear Mikey,

[Nice to connect with you after a long while :)]

But why ultrasound?

As might be expected, the suspected species of the cracks is the fatigue cracks, right? I mean, the surface-breaking cracks, the finer ones. And those precisely are the ones that would get lost in "boundary layer"/"edge effect" of the UT. [Sorry, but my NDE is some 20 years old, i.e., when it used to be NDT.]

If you were to say ET (I esp. mean eddy current), or MPI (magnetic particle inspections), and it would be OK by me [going by my good old knowlege]. But UT?

Guess what they showed (if they at all did) is a more good-natured reaction to help things out, even, to figure out things for themselves, rather than knowledge as such. Appreciated. Curiosity, when not interfering with a proper set of reactions in sad/bad situations like these, is perfectly OK, at least by me. ... Well... At least they weren't found busy taking selfies with the tracks and their equipment, I say! (The way it often happens... err... elsewhere and with others!)

...Anyway, I appreciate your trying to connect the knowledge with the practice.

Best wishes and regards,


[PS: BTW, as you can see, I've by now dropped my famous "E&OE" signature. I have resigned myself to whatever English that I do possess!]


By NDT is 30 years old, not 20.



I am certain I typed "My" and not "By".



Mike Ciavarella's picture

Anyway, I gave an interview to a newspaper today, but it will be in italian, so I would need to google translate it.

But nothing specific of course, there is a big trial now, but a serious investigation. The journalists wanted to have a "scoop", whereas I was more inclined to complain about the present maintanance strategy, the present investments in high speed trains (35 billions) whereas we need more safety on poor mid speed trains, and other things.

Mike Ciavarella's picture



Non mi piace autocelebrarmi, ma ho voluto fare un invervento sulla questione del disastro di Pioltello perchè sarebbe auspicabile, parallelamente alle indagini della Magistratura, aprire un dibattito serio tra tecnici politici e popolazione, sulla necessità di mettere in sicurezza TUTTE le reti, e non solo spendere 35 miliardi in Alta Velocità. Non è possibile che molte reti siano trattate in questa maniera, con giunti non saldati di vecchissima concezione che già dagli anni '60 hanno mostrato la loro vulnerabilità, nei disastri inglesi, cui si aggiunge una incredibile catena di imperizia e negligenza su cui non voglio ulteriormente commentare. Infine, in puglia avendo un'eccellenza nella produzione dei Treni Diagnostici, MERMEC, sarebbe anche auspicabile che dialogassero maggiormente con l'eccellenza del Dipartimento di Meccanica del Politecnico per fare questa seria indagine su tutta la Rete Italiana e sulle procedure di manutenzione. MC

Mike Ciavarella's picture

«Those joints among the main causes of derailments»

Professor Ciavarella (Polytechnic of Bari) opens the debate: politics chooses if it is the case to bet on the High Speed or to register and secure the network





BARI. For Professor Michele Ciavarella, full professor of machine design of the Department of Mechanics, Management and Mathematics of the Polytechnic of Bari, the railway disaster on the outskirts of Milan should push the country, and politics, a make priority choices between High Speed and security of the rest of the network. The expert, premising that the judiciary will clarify on the dynamics, it is skeptical that the derailment it was caused by those 23 centimeters of binary jumped. «In a situation of rectilinear motion there are examples of evidence made with trains, where they are put small bombs below, and do not derail, "he says. Thinking to «accidents of this type, strictly due - apparently - to defects on the tracks », Ciavarella cites the case of Rometta Messinese, in Sicily. In 2002, he ceded a joint. There were 8 dead. "Then - he says - there was a very long trail, many were released of any failure for time limits. And basically none remembers any more. Here, I would like to raise the debate on security in Italy. I would not like that this case is closed immediately. Yes, we must open the debate on High Speed (Av) vs low speed. " And he explains: "For the Av we are investing billions and welded type tracks, not with the joints of this last derailment. The first cost more to purchase but they cost less maintenance. And then the doubt arises that these figures are not or are not expenses good. A political question: how to spend the money? Replace the tracks on these lines a little neglected or focus on the Av? ».

The professor explains that "the joint is put for issues of thermal expansion, to connect two tracks that must be in place distance". The old model was used «up which was not developed in the 1950s the technology of welded joints. Those not welded - it clarifies - they have always given problems. On the contrary, it has been considered in the literature one of the main causes of derailment.

They are very critical elements if continuous maintenance is not ensured, both with diagnostic trains both with the use of ultrasound, that are those machines that were using Rfi technicians surprised in the area seized by the magistracy ".

As for the wooden plate placed under the joint, he says it's illegal, "it's outside of everyone prescription". He doubts that he who has put it has made a report, put it in writing: "There they are very serious prescriptions and operations they are all registered. But in this case it does not they could do it, because they could not put the piece. Writing it would have been a self-condemnation. If they did, they did it verbally.

But I wonder: how many problems like that there are they all the way? After the derailment, for a broken rail at Hatfield (October 17th 2000 died 4 people; ed), in England they checked the network and found themselves 2,000 problems of the same type. Now I believe that we should do serious verification work Also in Italy. And maybe from the third part, why those who have done it so far should not do it.

We can not wait for another accident similar, perhaps due to negligence, and the death of travelers ".


It is not clear if the Milan commuters have activated the emergency brake. Yet there is who has had time to call their loved ones. Also the professor shares that the passengers they should be trained, a bit like what happens on them planes. But this also requires funds.

Mike Ciavarella's picture

In China, see

See in general this link here.   There are many pros and cons and debate is high on political agendas.  If in Italy we talk of 35 billion euros, this is an amount of money which makes a difference in any country.

If people here are interested, we could debate.

Mike Ciavarella's picture

Nice debate! It has been a very long topic here in Germany, e.g. in the context of the new track from Munich to Berlin, or also re the new Stuttgart main station. A few links attached.


People from the railway industry say: the future are the local trains:


Stuttgart’s new station:


The new track from M to B:


And you may find lots and lots about the topic… At present it is not that hot, since there is enough money floating around for investments, so both local tracks and main lines are expanded.


Cheers, Norbert

Subscribe to Comments for "Tragic train accident in Milano for a broken rail: 3 killed and 46 injured"

Recent comments

More comments


Subscribe to Syndicate