European Train Control System

Just throwing it out there - if you are starting from the point of "all software is full of bugs" and then looking to apply that thinking to a discussion on a system that is already proven both as a concept and is in use everyday in various parts of the EU and in this country (Thameslink core uses an overlay of ATO over ETCS) then I'm afraid you're not going to get far in the discussion, preferring to get bogged down in examples of why software engineers are supposedly incompetent...

So if I may offer up this - software engineers are human, they err, as we all do. Whatever else happens elsewhere isn't really a good starting point for debate on ETCS being used with steam locomotives. So let's assume for the time being that ETCS works, and that the problem is fitting it and application rather than being fundamentally flawed?

So, problems arising:

1) Fitting to a steam locomotive
2) Its application on a steam locomotive - including emergency brake applications (as discussed above)
3) ATO overlay possibilities
4) Necessity - is there a business need?
5) Funding

I do not believe "power" is an issue as most of the locomotives to be fitted with this will also have a generator of some kind on board powering other things. Stones generators are on all the big locos anyway (Clan Line, Tornado, etc).

Point 4 - necessity - any locomotives going through south London are likely to require ETCS as the signalling upgrade programs take hold. For example, I have been involved at the back end of the Thameslink program, and several relay rooms around the network have been partially upgraded for potential ATO use in future. However - on lines not fitted with ETCS - is there a business case? ECML will be upgraded in time and therefore any steam locomotives looking to use this will need to be fitted.

Point 5 - funding - speaks for itself. There are limited funds available from NWR for some pilot scheme locos but beyond those initial locomotives it is clear there is not a limitless supply of cash to fit out locomotives. Hence my points on the Tornado thread regarding the fitting of the operating systems and where they should be (but also admitting I had forgot about emergency brake applications - an error on my part - and the shutting off of the regulator via the German system seems a reasonable engineering solution there).

 

The idea of an overriding control force being applied to a manually operated lever seems fraught with danger to me and I don't see how you can provide a soft link that would not initially move the regulator handle in preference to the valve itself. If we accepted the dangers of this, it could be made to work on an external linkage but I'm at a loss as to how it could be installed on the more common regulator with a longitudinal rotating rod through the boiler. I do wonder whether a simple butterfly valve in the main steam pipe may be a practical solution? it would be balanced so require relaitively little force to operate. Alternatively, a flap valve would shut automatically if there is a steam flow and the force holding it open was removed. In normal operation, the valve would be open and control of the loco remain with the fitted regulator (whether smokebox or dome mounted) but closing the emergency valve would cut off the steam supply independently of that. I have yet to come across a conventional standard gauge loco that doesn't have a steam pipe within the boiler. The valve wouldn't have to be a perfect seal although that should be the aim. There remains the problem of keeping the valve steam tight against steam in the boiler finding its way into the steam pipe downstream but I don't see this as an insurmountable problem. The rod could be encased in a tube to project outside the barrel with a conventional gland and the operating mechanism outside the boiler. The only external change to the locos aesthetics would be the link and operating cylinder,, plus associated pipework.

 

I don't want to divert too far from the central ETCS discussion but, as I referred to software bugs in my opening post on this thread, and as Forestpines has mentioned testing, I should mention that (some years ago) I was employed by the bit of UK Government responsible for safety of medical devices. My responsibilities included medical equipment containing software and I participated in the development of the first of a series of international standards for safety of such equipment, IEC 62304. The received wisdom in the field of "software safety" is that testing can only ever find some of the bugs, so the only way to get adequate assurance of safety is by a closely controlled design process from specification through implementation and subsequent "maintenance". Unfortunately most software development focusses much more on functionality than on integrity. Not everyone in the field would agree with everything that Martyn Thomas says in the series of lectures that I linked to, but there is not much to disagree with.

 

There is no need for any control over the regulator. I was involved directly with testing on the Cambrian with 44871 when we were looking into a solution for ETCS. We carried out dynamic brake testing at different speeds with the regulator in varying positions. Having the regulator open actually had minimal affect on the overall stopping distance as the brake force of the loco and train easily overcame the applied traction power. There is a test report with all the data that is probably buried in the Network Rail archive somewhere.

 

So...bearing in mind I can't say too much (employer, employee relationship) but I am examining a document giving the details of the class 97 (Cambrian ERTMS fitted class 37s) and there's photographs of the ERTMS cabinets/screen fitted and associated wiring. It doesn't look too onerous to fit into a tender locomotive to be honest, if you accept you're going to put the cabinets somewhere within the confines of the tender (and not necessarily, it has to be said, on the footplate side).

There's another issue - to me anyway - the GSM-R stuff that's fitted to the class 97s. Antennas etc, including axle end speed probes (possibly fit these to the bogies on a 4-6-0, cartazzi on a Pacific?)

Food for thought.

 

I'm not quite willing to bet your favoured king-size Mars bar on this, but I reckon it can be done, although possibly a bit bulky. I need to find time to do some drawing and work it out. An auxiliary regulator as you suggest also seems very feasible. However...

I thought that might be the case, but it's good to hear it from somebody who's been there done that. At least with steam there's not an equivalent problem to burning out the traction motors, as you might with a diesel or electric given the same treatment without cutting the power.

 

Out of interest, were any braking tests with the regulator open carried out whilst running light engine.?

The RSSB standards require the system to automatically cut off the traction effort on locomotives or units in the event of an ETCS activation. See Section 3.2 Traction Power Cut-Off on page 13 of this document:https://www.rssb.co.uk/Library/impr...cification-etcs-onboard-system-baseline-3.pdf

Andy

 

I wonder if the P2 project is looking at incorporating this into their boiler design? It would have thought that it would be easier than retro-fitting it.

Dave

 

And remembering that with Stanier locos if in second valve you have to fully open the regulator first to shut off.

DaveW

 

Not just Stanier's locos.