Doon Revisited

[tree68]

VOLKER LANDWEHR

The soil is encased in a box, therefore the water has no way to exit to the side accelerating the liquefaction. The missing foundation don't have any influence on the process.

Wouldn't that more or less have been the case if the water level of the river were uncharacteristically high, completely submerging the roadbed?  The effect would have been to place the roadbed in suspension, with the passing train simply stirring the pot.

We've noted that four foot surge - that may have placed the water level well above the levels to which the MOW forces were accustomed.  

[Anonymous]

tree68

Wouldn't that more or less have been the case if the water level of the river were uncharacteristically high, completely submerging the roadbed?

I don't think so. The walls of the box confine the energy, which is brought into the system, to the area of the box. Without the walls or a box with a larger area the same energy would have activate a larger area slowing the process. In the field the wall would need to be the surrounding soil. It will never be as rigid as a metal wall.

What happens in reality is a question of the local situation. Is the soil condition, allowing liquefaction, only a local problem on a number of sq ft or a general problem along the fill. So the results of this laboratory test cannot be transfered one-to-one to the field.
Regards, Volker

[Euclid]

Overmod

Now, in order to produce liquefaction, you need a motion of water sufficient to fluidize an appreciable volume of soil for a considerable time under a train's weight. A very likely component of force to produce such motion would be vibration, and a good engineer can calculate frequency ranges for a given grade or subgrade that would be most efficient at producing it and compare those to actual vibration characteristics produced by an oil-train consist. The argument, however, is slightly different: would liquefaction producing purely a vertical displacement of the track result in a derailment? I find this relatively unlikely based in part of my knowledge of the dynamic response of three-piece trucks and the experience of the Army demolitions people investigating the best ways to derail trains in WWII. This is not to dismiss the idea that liquefaction is related in some way to the incident, only to note that concluding liquefaction IS either the expedient cause or a major contributor to the incident is at best premature. Possibilities do, I think, include the idea of liquefaction-related settlement under one 'side' of the train, or (as noted) lateral slip of some part of the grade/subgrade resulting in impairment of track geometry. I would note that neither of these, if they were present, seem impossible of remediation within the timeframe and apparent resource allocation BNSF provided after the incident;

When I opine that liquefaction caused the derailment, I have suggested that the support gave way under the weight of the train.  However, I do not mean to suggest that the center of the liquefaction zone was necessarily located dead center under the track.  Indeed, it might have been located off center to the extent that some of the bank slope may have dropped away due its own weight under the influence of liquefaction as the train vibrated the soil during its passage.  And then that dropping soil may have reduced the support of the train.  Actually, liquefaction may have spread to the entire fill from side-to-side and top-to-bottom. In cases where the liquefaction occurred without causing bank slope to slide, it could nevertheless liquefy soil that did not require cohesiveness to remain stationary.  In that case, the added weight of the train would merely drop into the liquefied soil without any sliding away of soil having occurred. 

I have seen those videos (that you mention) showing tests of how much track damage it took to derail trains.  And it was surprising how much it took to derail a train.  However, I would not conclude that it indicates that liquefaction would be unlikely to cause enough loss of support to derail a train. 

This is because the video tests were on solid roadbed with ballasted ties, and the only derailing factor was the removal of various lengths of rail.  And also, that equipment was relatively light compared to a loaded oil train of today.  So, I think liquefaction would have the potential to very effectively cause a derailment.

If the entire width of track bed lost its ability to support the train, the track would suddenly drop under the weight of the train to the point of being submerged in the muck.  This reduction of support would likely be irregularly distributed so that the cars would dip and rise as conditions vary in direction of travel as well as side to side on the track.  This would place very high stresses on the track which might break rail and ties or cause the rail to tip over or separate from the ties.  Or the irregular movement might just directly derail cars as their wheels momentarily cleared the rail head.   

It would be interesting to know whether the crew recalls feeling any sort of instability in the track as they were passing through the flood zone where the derailment occurred.  It is possible that liquefaction began to develop from the passage of previous trains.