Escalator accident in Stockholm subway

[Skogsgurra]

An escalator went amok a few days ago. Started moving backwards with lots of people on it. Speed way above nrmal. Brakes didn't work.

Coupling bad, was my first thought. Today, there is a picture in the papers showing the coupling.

The press spokesman (yes, a male)said that the accident was caused by rust. I can't see much rust on the part. But fretting in the splines. And I guess that the correspondning shaft is even worse. So, no wonder that E-stop didn't work.

Comments?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[John2025]

I don't have any experience with escalators, but I'm surprised they don't have some kind of emergency brakes that are independent of the drive system , like an elevator.

 

[edison123]

For such a big coupling, the splines don't seem to be deep or wide.

Muthu

http://www.edison.co.in

 

[3DDave]

The escalator mechanisms I've seen are basically large link chain with the stairs carried as attachments. The drive sprockets are the best place to grab the chain. The rate of failure and the typical lack of fatal results when the brakes do fail accounts for the lack of more elaborate stopping systems that elevators necessarily depend on.

For a scare, look at a Paternoster elevator.

https://www.youtube.com/watch?v=0ZXIWgyf6lQ

It is also a chain mechanism.

 

[waross]

About 30 years ago in Toronto. This was a construction skip hoist erected outside the building.
The counterweight balanced the elevator cage with about 50% of rated load.
With only two men in the cage, the elevator fell "UP" instead of down when the driveshaft failed.
After the repairs to the elevator I had occasion to ride the elevator to work for a few months.
Scary.
What do you do if you have engineering training but can't quite cut it? Go into safety. (Not all safety officers, I have met safety advisors who had excelled in management and went into safety as a personal choice.)
I can imagine the discussion:
"But there are safety brakes. Why didn't they work?"
"The brakes only work going down. The cage was going up."
Wrong answer.
Correct answer: The brakes only work if the cables break and the cables didn't break.
The solution? A second set of brakes installed upside down in case the elevator failed and went up again.
Totally useless. If the drive shaft failed again, neither set of brakes would apply.
Twenty years later I rode an almost identical hoist/elevator.
I was glad to see that it had a set of brakes that would stop the cage if the drive shaft broke.

The building's stats alone are impressive: 1,496 corner offices, 44 elevator cars, 28,000 exterior tiles, 5,000 windows, and 18,500 lights, all subtly brightened and dimmed automatically by a light sensing computer. A giant 2.95 million litre water tank in the basement circulates water through each floor, cooling in summer and heating in winter much like the Toronto's Enwave system does today.
toronto scotia plaza

Suddenly, with the building approaching full height, a faulty driveshaft in a construction elevator failed. The knock-on effect sent a car carrying two workers racing at more than 100 km/h toward the top of the shaft. Both were killed on impact and another man caught in the wreck was seriously injured. All work was suspended pending an investigation.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

Gunnar; Possible misalignment in the original installation? Often rust in a spline like that is almost as good as welding, unless there is movement.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[edison123]

Bill - The slenderness ratio of that building is kinda scary.

Muthu

http://www.edison.co.in

 

[itsmoked]

It's all rusty..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

No, Keith. Not rust - dirt. Doesn't influence function the least.
But there is fretting in the splines. And if the splines looks like this, the corresponing shaft is probably more or less round with the splines ground away.

Here is a similar damage:

Fretting produces iron oxide powder. May look like rust. May have a similar chemical composition. But it isnt rust.

These couplings have been in use for a Little more than a year. Indoors.

In the example shown above, there was a very primitive drive with switches going ON/OFF a few times every period - "Full Block" switching - and caused vibrations in the motor shaft. Vibrations that knocked the key to dust.

Similar things happen in PWM drives when you set P too high to get a "crisp" step response. Most guys think that step response should be like the text-book says. That isn't always good. In servo applications, yes. In most drive systems, no.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[itsmoked]

LOL I was referring to the building Bill put up.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

OK, but gave a possibility to show one of my favourite pics. This one started as an ASSumed (Jeff)bearing current problem. Which it was not.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

LOL Keith.
The construction skip or hoist was driven be a pinion gear on an extended shaft driving on a rack gear extending the entire height of the building.
The pinion gear shaft broke.
The brakes were on the motor side of the break.
The hoist that I rode years later had an independent brake installed that worked directly on the tower frame independently of the drive mechanism.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

Gunnar, I am asking you to share your expertise, not challenging you.
I am thinking that the failure of one escalator rather than widespread failure may indicate that there may have been a design change involving the drive parameters, This may have been one of the first of a new design or the failure may have not been related to the drive.
Is it possible that a slight misalignment in the coupling could lead to fretting?
Thank you for sharing your knowledge and valuable opinions Gunnar.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

There are 71 of these escalators. Standard design where not much can co wrong with alignment.

All less than two years old.

The Stockholm Transport (SL) has opened quite a lot of them and found the same problem in many. All will be examined the days to come.

I am betting USD 1000 that this is vibration thing like the one described. Anyone want to lose one grand?

I am following this with great interest now when France won over Croatia...

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[edison123]

Gunnar - I will take that bet. I am going with poor spline coupling design. Who is the judge?

I am too coming off of that Croatia-France humdinger. Luck and referee favored the french.

Muthu

http://www.edison.co.in

 

[Skogsgurra]

OK. Let Bill be the judge.
Muthu: Spline design. I.e. too weak.
Gunnar: Fretting caused by vibration or knocking.

OK? Bill?

Side bets are possible. But no free spins!


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

You guys are putting me on the spot.
I favor both causes.
A combination of marginal spline design that was susceptible to fretting.
If the spline had been much more robust it may have withstood the fretting for years.
If there had been no fretting the spline may not have failed. (I may be subject to correction here, considering the photo of the shaft key badly damaged by fretting.)
Is it acceptable to be guided by the findings of Stockholm Transport?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

Yes Bill. That will, I hope, tell us the truth.
But there may/will be lawyers involved. So the REAL truth may be difficult to reveal.

In the meantime - I think it will take years Before the investigation is finished - all views are welcome. We may even help the lawyers understand that they cannot hide facts when engineers are looking.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[edison123]

LOL. Well done, Bill. Almost Solomonesque.

So, we split 50/50, Gunnar?

Muthu

http://www.edison.co.in

 

[Skogsgurra]

Up early, Muthu?

Up late myself. Very late, actually.

We haven't got the results yet. Bill is just cautious and wants to see what the investigation shows (just learned that DEKRA is doing the investigation). So it is a bit early to split. But if you prefer that - well. Or we drop it, which is the same thing.


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.