Collision and derailment near Salisbury (Fisherton Tunnel) 31/10/21

[bahnause]

I am not sure that ‘throw it all in’ is the correct approach. Figure 34 in the report shows an extended recovery time for the WSP in step 3 compared with step 2.

Both braking actions took place in different conditions. We can therefore not draw any conclusions about the braking behavior at different braking steps.

 

[GC class B1]

Both braking actions took place in different conditions. We can therefore not draw any conclusions about the braking behavior at different braking steps.

While it is correct that the railhead conditions for 1L53 were likely to have resulted in a lower available adhesion than the previous two trains, it is clear from the OTDR trace that the time taken for the brake to release and the wheelset to commence rotating again was significantly longer for 1L53 than the previous trains. It is therefore possible that the deceleration rate for the train was lower than would have been the case with the shorter WSP response time with the lower brake cylinder pressures in steps 1 or 2. The longer response time for the speedometer wheelset in step 3 would also have applied on the other wheelsets on the train. A shorter WSP response time along with the effect of sanding and railhead conditioning may have allowed a higher deceleration rate with better adhesion.
Another factor that may apply is that the static wheel/rail coefficient of friction is higher than the dynamic coefficient. This means that a wheel that is not sliding will have better adhesion than a sliding wheel. The better adhesion available to following wheelsets may allow their WSP to perform better or avoid those wheelsets sliding.

 

[D6130]

I believe it may be different with SWR, but not sure if it was at the time of this incident, but our step 1 on a 158 (GWR) is usually terrible

The big difference between the three step braking systems is that on the class 150/153/155/156 units it operates tread brakes whereby you can normally make a perfectly satisfactory initial application in step 1....but with disc braked units - such as class 158/158/321 etc. step 1 is little more than a holding brake to keep the train stationary while stopped at stations. For any meaningful degree of deceleration,
you must make the initial application in step 2.

 

[GC class B1]

The big difference between the three step braking systems is that on the class 150/153/155/156 units it operates tread brakes whereby you can normally make a perfectly satisfactory initial application in step 1....but with disc braked units - such as class 158/158/321 etc. step 1 is little more than a holding brake to keep the train stationary while stopped at stations. For any meaningful degree of deceleration,
you must make the initial application in step 2.

The driver of 1L43 braked early in step 1 and deceleration was steady without wheel slip. The driver of 1L43 brought his train to a controlled stop at signal SY31 using a combination of step 1 and release. I have calculated the step 1 deceleration rate to be about 3% g from figure 34.

 

[Stigy]

The big difference between the three step braking systems is that on the class 150/153/155/156 units it operates tread brakes whereby you can normally make a perfectly satisfactory initial application in step 1....but with disc braked units - such as class 158/158/321 etc. step 1 is little more than a holding brake to keep the train stationary while stopped at stations. For any meaningful degree of deceleration,
you must make the initial application in step 2.

I understand tread braked units have a sharper step one, but disc braked trains do often have sufficient step 1 to use as a main braking step (class 16x for example) and there’s no reason why other traction can’t have increased pressure in that brake step (it’s not that the disc brake set up isn’t capable of it).

 

[GC class B1]

I understand tread braked units have a sharper step one, but disc braked trains do often have sufficient step 1 to use as a main braking step (class 16x for example) and there’s no reason why other traction can’t have increased pressure in that brake step (it’s not that the disc brake set up isn’t capable of it).

My understanding is that the brake cylinder pressures for both disc and tread braked units are different and are set to achieve 3%, 6% and 9% g deceleration in steps 1, 2, and 3 respectively. Tread braked vehicles often have a higher brake cylinder pressure for the same retardation rate as disc braked vehicles because of the design differences. This does not mean tread braked vehicles have a higher deceleration rate than disc braked vehicles in the same brake step.
Some sprinters (they all have tread brakes) have cast iron brake blocks. Cast iron has a significantly higher coefficient of friction at lower brake cylinder pressures than equivalent composition brake blocks and this may explain your experience of sharper step one braking.

 

[Stigy]

My understanding is that the brake cylinder pressures for both disc and tread braked units are different and are set to achieve 3%, 6% and 9% g deceleration in steps 1, 2, and 3 respectively. Tread braked vehicles often have a higher brake cylinder pressure for the same retardation rate as disc braked vehicles because of the design differences. This does not mean tread braked vehicles have a higher deceleration rate than disc braked vehicles in the same brake step.
Some sprinters (they all have tread brakes) have cast iron brake blocks. Cast iron has a significantly higher coefficient of friction at lower brake cylinder pressures than equivalent composition brake blocks and this may explain your experience of sharper step one braking.

I assume by sprinter you’re referring to all sprinter class trains other than 158s, as they have discs.

The point I was making, was that 158s have a particularly poor step one brake, whereas that’s not true of all disc braked trains.

 

[D6130]

I assume by sprinter you’re referring to all sprinter class trains other than 158s, as they have discs.

The point I was making, was that 158s have a particularly poor step one brake, whereas that’s not true of all disc braked trains.

In my experience, class 321 and 322 units also had a particularly poor step one brake....thus we were instructed that the initial application should always be in step 2 for those units as well as 158s. However I once drove an SWT 159 on the Settle & Carlisle line - during its brief period of loan to Northern about twenty years ago - and noticed that the step 1 braking was noticeably superior to that of our 158s.

 

[GC class B1]

I assume by sprinter you’re referring to all sprinter class trains other than 158s, as they have discs.

The point I was making, was that 158s have a particularly poor step one brake, whereas that’s not true of all disc braked trains.

Yes. Class 158/9 aren’t sprinters they were called express.
There is no reason a correctly maintained class 158/9 shouldn’t brake at a nominal 3% g in step 1.

In my experience, class 321 and 322 units also had a particularly poor step one brake....thus we were instructed that the initial application should always be in step 2 for those units as well as 158s. However I once drove an SWT 159 on the Settle & Carlisle line - during its brief period of loan to Northern about twenty years ago - and noticed that the step 1 braking was noticeably superior to that of our 158s.

There is no reason a correctly maintained class 158 shouldn’t brake at 3% g in step 1 the same rate as a class 159.

 

[Stigy]

Yes. Class 158/9 aren’t sprinters they were called express.
There is no reason a correctly maintained class 158/9 shouldn’t brake at a nominal 3% g in step 1.


There is no reason a correctly maintained class 158 shouldn’t brake at 3% g in step 1 the same rate as a class 159.

They’re a class of sprinter nonetheless. Express Sprinters.

I guess this is what I was initially talking about, the SWR 159s do indeed have a better step one brake than 158s, and the differing brake setups don’t directly mean disc brakes are naturally any worse than tread brakes in lower steps.

I’ll be honest, I don’t understand all this stuff about ‘g’ or the science behind it.

 

[GC class B1]

They’re a class of sprinter nonetheless. Express Sprinters.

I guess this is what I was initially talking about, the SWR 159s do indeed have a better step one brake than 158s, and the differing brake setups don’t directly mean disc brakes are naturally any worse than tread brakes in lower steps.

I’ll be honest, I don’t understand all this stuff about ‘g’ or the science behind it.

In simple terms, a deceleration rate of 1 g is 9.81 m/s/s. G is the gravitational force of the earth at sea level and an object in free fall will accelerate at 9.81 m/s/s = 1G. Expressing an acceleration or deceleration rate in %g is normal practice. It is also used as a term for the force experienced when going round a corner fast - i.e. if the g force is one, your body is subject to a force that prevents you from flying off at an acceleration rate of 9.81 m/s/s. This force would accelerate you at 9.81 m/s/s if you were stationary.
In round numbers, 3%g is a deceleration rate of about 0.3m/s/s. As a reference, a 3% g brake application should reduce the train speed by 10 MPH in about 16 to 17 seconds from selecting step 1, including brake cylinder pressure build up time. Hope this helps.
i am interested in why you have found class 158 braking in step 1 to be noticeably less than designed. Did you also find steps 2 and 3 to be less than designed?

 

[Brush 4]

Had this reply from NR about trees in general and specifically this bit about Salisbury. Not sure what 'in accordance with our standards ' means though. Just a few trimmed back next to the line? No return to the 70's scene with the Warship then?

With the trees at Salisbury, they have recently been surveyed and work Is planned at the tunnel entrance based on risk to the operational railway and in accordance with our standards. This work will be subject to a suitable possession of the railway (when trains are not running) and at present we’re unable to confirm the date of the work.

Once again thanks for getting in touch with your concerns. I hope this information is helpful.
Kind regards

 

[6Gman]

Had this reply from NR about trees in general and specifically this bit about Salisbury. Not sure what 'in accordance with our standards ' means though. Just a few trimmed back next to the line? No return to the 70's scene with the Warship then?

I think that's all you can realistically expect. In the same way that highway authorities have standardised approaches to things like speed limits, potholes, road improvements (balancing safety, cost and the efficient use of resources) the railway will have a procedure for tree works (balancing safety, environment, cost etc.).

 

[stuving]

I think that's all you can realistically expect. In the same way that highway authorities have standardised approaches to things like speed limits, potholes, road improvements (balancing safety, cost and the efficient use of resources) the railway will have a procedure for tree works (balancing safety, environment, cost etc.).

As you say, a letter like that can't repeat the whole of their management guidance for the subject: there's too much of it. And it's not just about how to do work, but whether, when, and where. The main document I can see is the Lineside Vegetation Management Manual (NR/L2/OTK/5201). That is public, but is very high level. All the detail is in four modules:
MOD01 Lineside vegetation inspection and risk assessment
MOD02 Lineside vegetation management requirements.
MOD03 Route vegetation management plans
MOD04 Tree Management

Those modules do not appear to be public.

 

[norbitonflyer]

The trees at the tunnel mouth were not the problem in this case. The train started to slide over a mile away.

 

[PG]

The trees at the tunnel mouth were not the problem in this case. The train started to slide over a mile away.

Good point. Does anyone have pictures of that area showing how the vegetation looks/looked over time?

 

[3141]

Good point. Does anyone have pictures of that area showing how the vegetation looks/looked over time?

Google Earth shows trees lining the railway for some distance towards London. Perhaps the illustrations included by RAIB of the area near the tunnel weren't strictly relevant. Or perhaps they couldn't find any showing how things looked fifty years ago on the section where the train was sliding.

 

[david1212]

The trees at the tunnel mouth were not the problem in this case. The train started to slide over a mile away.

While true what state were the rails approaching the tunnel mouth in? Had the area above the junction and tunnel approach been kept like ~50 years ago and hence the rails clean(er) might the incident have been less serious?

Good point. Does anyone have pictures of that area showing how the vegetation looks/looked over time?

The 1970 photograph is linked back on post 780
Post 780

70 096 290470 Salisbury Tunnel Junction D819

Seen at Salisbury Tunnel Junction on the 29th of April 1970 Warship D819 Goliath is on the 1O12 (I think) 13.10 Exeter SD to Waterloo service. The loco was withdrawn in October 1971.

www.flickr.com

Major work now to get back to this plus as well as what is visible there are all the roots and associated risks of earthworks movement. If only the growth had been controlled, a perfect example of ' a stitch in time saves nine '.

 

[PG]

The 1970 photograph is linked back on post 780

Thanks. I was more interested if any pictures exist of the area in which the SWR train first started to slide rather than immediately in the vicinity of the accident.

You do make valid points (no pun intended) that the slide may not have been as serious if the junction area vegetation had been maintained as per that photograph.

 

[30907]

While true what state were the rails approaching the tunnel mouth in? Had the area above the junction and tunnel approach been kept like ~50 years ago and hence the rails clean(er) might the incident have been less serious?

Possibly, but by the time the train reached the junction the state of the track would have made very little difference either way. Is it even mentioned in the report?

 

[Peter Sarf]

While true what state were the rails approaching the tunnel mouth in? Had the area above the junction and tunnel approach been kept like ~50 years ago and hence the rails clean(er) might the incident have been less serious?


The 1970 photograph is linked back on post 780
Post 780


Major work now to get back to this plus as well as what is visible there are all the roots and associated risks of earthworks movement. If only the growth had been controlled, a perfect example of ' a stitch in time saves nine '.

My bold.

Hello - gardner here.
yes.

And I look in alarm at all the Budlea growing out of brickwork that had been lasting well since Victorian times. Some of the vegetation is so established you can see the distortion of courses of brickwork. Ask someone to put a value on the long term damage being done to our assets and the answer will be stakeholders do not need to worry about things so far ahead (what is meant is I will have had my bonus by then).