Engineering Standards, ex-82045 The way ahead?

One important aspect that is continually overlooked when discussing 'batch' production of 82xxx loco's, or indeed any new builds, is the question of 'the knowledge'. Although most of the BR drawings are available, they are simply inert engineering drawings, the construction procedure, idiosyncrasies, order and new material designations are things that are transitory. The importance of this 'knowledge' should not be underestimated. As in the distant past, it resides in the heads and minds of those involved in the construction process. As 46137 stated above "we're mostly too old"- that's the understated kindness that one expects from a Latin teacher ! It would be useful if there were ever consideration for 'batch production', if it came sooner rather than later, before this knowledge and its implications are lost in the mists of time once again.

 

I think you are being rather dismissive of the younger generation of steam engineers, and what has already been achieved in terms of renewals across a large number of locomotives. After all, where did 'the knowledge' come from in the first place?

Steven

 

A point well made.... and why items like this (from the VoR Facebook site) are heartwarming:
https://m.facebook.com/story.php?story_fbid=1840340509314965&id=205064302842602&__tn__=*s*s-R

 

Not at all. Credit and admiration to all steam engineers, whatever their age.

The point I was making is that it's one thing to dismantle something and put it back together again. Even that is demanding and challenging enough in its own right. It is something completely different to start with blank sheets of paper, some probably missing - drawings, and build something blind from scratch. The idiosyncrasy's of the construction process's were developed and overcome in the locomotive works by those employed on the shop floor. Often without the drawing office being aware of minor modifications being undertaken, let alone the order, process and precision required in assembly. These were skills and practices that were transferred by custom and practice through time, and resided mainly in the workforce's heads.

The 'knowledge' was acquired through hard experience on the job, and was not something to be learned from the classroom or documentation. I can assure you from my own experience on new builds, little has changed in the meantime. We say many times, "if only we had known what we know now, we would have done some things differently". That is the 'knowledge' which has had to be re learnt. And as you will be well aware, time effects the bottom line.

 

In an earlier incarnation, one of my tasks was to write Technical Process Specifications .... when your employer worked to BS or MIL standard specs, there wasn't much choice in the matter. AIUI, newbuild locos require much the same as part of the proving and acceptance process. This level of documentation may be anathema to some, but will inevitably form an increasingly important component of the heritage sector.

Obviously, the sum of all knowledge to reliably produce a successful steam loco design would be tricky (to put it mildly!) to write down in a library's worth of documentation, but for the future of our sector/hobby/obsession, such an effort needs to be made before those with the relevant knowledge have fallen off their perches. Here, we could usefully learn from several 'oral history' projects conducted over the past few decades.

Mercifully, specialist areas such as metallurgy are as critical to modern industry as ever they were. Ditto the basic techniques for casting, forging and machining. Some (whisper it quietly) such as welding, have come on leaps and bounds. Other modern techniques, such as nitrogen freezing, make operations such as fitting crank pins a far smoother process than 'back in the day'. Issues over suitable available grades of ferrous materials have been overcome, with the 'new' material occasionally even proving less of a bitch to work with!

In the design field, the tools available to the modern draughtsman are light years ahead of what their predecessors had to hand. That these systems can be connected to pattern production is another major step forward.

None of the foregoing is intended to belittle the scale of problems that will become evident with the passing of the last generations who worked with steam and accumulated a lifetime's experience in the process. Equally, don't despair that today's apprentices will prove any less adaptable than those who came before them. There's never a substitute for learning from a master, but that mustn't be used as an excuse for inaction when it comes to providing as comprehensive a failsafe repository as humanly possible.

If nothing else, the heritage sector is peopled by imaginative, creative and adaptable folk. Every challenge has thusfar been met, problems considered insurmountable a couple of decades ago are now almost matters of routine. Hell..... the heritage sector has even given the construction industry one of it's hallmark materials in the form of 'Shotcrete'. Adaptation and progress will continue. The sun will rise tomorrow, even if it's pi**ing down! The real enemy in all this is complacency. Same as it ever was.

 

A lot of what you have stated is definitely applicable to Swindon Loco Works. The drawing office assumed that the people on the shop floor knew what was required, especially on tolerances.

On another point, you have the apprentice that has just come out of his time. The boss hands him his papers and says to him " Right son, I've taught you all that you know, but you still do not know all that I know."

 

I beg to differ. You can have all the drawings in the world and be an experienced engineer but there are many things that you need to have experienced or, at least, had told to you in order for you to complete the task successfully. Knowledge generally comes from people making mistakes. How many people have started putting things together only to have to dismantle it again because it needed to go together in a certain order? Once you've done that you get the knowledge so hopefully don't end up doing the same thing again. Drawings are not necessarily correct, as well, and it was not unknown for the shop floor to modify things without that fact getting back to the drawing office. I recently made four pipe clips to a Sentinel drawing dating from the 1920's. I cut the bar to the length stated on the drawing and machined the screw threads on all of them. I then came to bend them and discovered they weren't long enough, even though cut to the stated overall length. Result was four pieces of scrap metal. The drawing was wrong but had been used in the build of Sentinel boilers for thirty or more years. The shop floor must have known this but never had the drawing modified.

 

Where did I suggest that the drawings were always right or that what I was referring to ignored the fact that 'what the drawings say' and ' how you really make the thing' are 2 very different issues?

The point I was making is that 'the knowledge' is being passed on, either by those from 'the past' when teaching younger generations, or by engineers (as opposed to 'fitters') learning themselves the reality of engineering. Yes, that may involve making, and learning from, mistakes, but you could argue that the world of preservation, lacking as it does the size of resource of the old locomotive works, is a faster learning ground for 'the knowledge', because what may have been simple tasks to Doncaster or Derby are less so for a small team (maybe one person) in a small shed, so learning how to 'make it work in practice' is more commonly needed. It would certainly be unfair to assume that youth means a lack of the abilioty to work around engineering difficulties - indeed, it can mean a more open minded approach to how to do so, but any such generalisation is potentially unfair.

None of which removes the fact that the present generation are tackling major jobs in modern ways in the reality of the current world, and actually achieving far greater achievements than was even thought possible a couple of decades ago, when it did seem to be assumed some skills had gone forever.

Steven

 

Steve's post #1058 reflects very clearly the sort of drawing office I remember only too well. There's no tactful way to put this, but what he accurately describes was, to put it bluntly, symptomatic of the sort of sloppy practise which contributed to our once mighty engineering sector being left wide open to annihilation by far more efficient competition.

Computerisation is here to stay, or if it isn't..... we'll have more to occupy us than the finer points of steam age engineering!! That CAD errors exist is evident fron the odd glitch with even new builds. Practise makes perfect, eh? What new tech does give us is repeatability, and integrating updates into properly referenced CAD is a damned sight easier than digging out 101 related hard copy drawings.

Don't get me wrong. Were someone to assemble a posse of blacksmiths skilled in late 18th/early 19th century tools and techniques to reproduce, say, Catch-Me-Who-Can using only period specific materials and techniques, my hat would be off to them. If anyone suggested the same to produce a reliable modern mainline loco.... now that'd be a whole different kettle of fish!

 

This discussion reminded me of this passage, re the manufacture of Merlins for the Mustang in the USA:

"In his book “Not Much of An Engineer”, Rolls-Royce engineer Sir Stanley Hooker recalls his introduction to the matter with Ford:

“One day their Chief Engineer appeared in Lovesey’s office, which I was then sharing, and said, ‘You know, we can’t make the Merlin to these drawings.’

I replied loftily, ‘I suppose that is because the drawing tolerances are too difficult for you, and you can’t achieve the accuracy.’

‘On the contrary’ he replied, ‘the tolerances are far too wide for us.’ We make motor cars far more accurately than this. Every part on our car engines has to be interchangeable with the same part on any other engine, and hence all parts have to be made with extreme accuracy, far closer than you use. That is the only way we can achieve mass-production.’”

Like Ford, Packard was obligated to redraw all of the Merlin blueprints to satisfy their own manufacturing requirements."

I suspect the work being carried out today in preserved railway workshops is of somewhat higher quality than in BR days. I guess we automatically assume that in the good old days the Brits were good at building steam locos but, as an example, Hudson Power (about the Victorian Railways R Class) is less than complimentary about the quality of North British workmanship.

 

Good point - not only are workshop conditions different today, so are operating conditions. It isn't a case of 'if you break this one, there is a spare in steam on shed' any longer! Engineering for reliability is key - an ex-BR engineer said that to me a few years ago and of course the problem is it takes a few years of that approach to see the benefits, which occurred after he had retired and other involved had left, meaning they aren't the obvious ones to get the credit for the foundations they laid!

Steven

 

I do like that story (pedant alert: it was actually Ford UK making Merlins for the UK at that stage, I think - Packard did the Merlins for the Mustangs).

North British were certainly variable in their output, the batch of King Arthurs they did for the Southern basically needed heavy general overhauls before they were as reliable as the Eastleigh batch. NB had underquoted for the order and cut a few corners during the build.

My personal take on replicating steam-era parts now is as follows:

If it was a machined part then, it can be made as well or better now.

If it was a casting then, there is a way of making it now.

Things can be welded now to a quality undreamed of then.

Forgings are trickier, but most railway forgings were arguably done that way as a quick way of getting to near net size, rather than for excellent grain structure. Multi-axis machining is so capable now that arguably a connecting rod will be just as good if water jet cut from good quality plate and then machined. For reference, as recently as 2011 Mercedes F1 were not forging crankshafts, they were just machined from bar, heat treated and surface treated. I don't know what they're doing at the moment.

One exception I do know of is boilersmiths grumbling that modern low carbon steel grades are too pure for their own good and suffer more than the older material in steel fireboxes.

Anything can be done if you have the money. That's the hard bit, arguably.

 

I agree wholeheartedly that thee is no point in producing anything as a forging if it can be machined from a block. It avoids the possibility of forging laps and piping. The idea that axles and coupling rods, etc have to be done this way belongs to the last century.
It's also true about boiler plate. Modern steels corrode far more easily and quickly given half the chance. Boiler treatment was never more important.

 

Duncan Ballard as the SVR's Boiler Shop manager told me that nearly all steel now being replaced in a boiler overhaul is steel that had already been replaced in preservation.

 

Comments regarding (especially) grades of metals for boilers and fireboxes set me thinking. Clearly, a full size boiler, even for a small industrial tank, is an expensive and safety critical item.

Question: When it comes to applying techniques or materials new to steam loco construction and maintainance, could model engineering disciplines usefully and cost effectively provide some of the initial R&D effort formerly undertaken by in-house and independant manufacturing facilities? Are there any instances where this has already happened?

Obviously there are areas where simply 'scaling up' an idea, material or component isn't either straightforward, applicable or perhaps desireable, but properly applied scientific rigour might permit work in smaller, i.e. much more manageable, scales to be usefully undertaken, with a high degree of confidence, with a view to inclusion in documented procedures for full sized locos.

 

Beachy Head's cylinders were fabricated: before making them, a small-scale model was made of each component. I believe the primary reason was to work out the assembly technique of the various components and ensure that they could be assembled from individual components before the full size parts were ordered. Move on 10 years or so, and with developments of 3D printing: 1:8 resin models of No. 27's cylinders have been made, again primarily to check the drawings before the full-size items are cast. (Also a useful eye-catcher on sales stands!)

On forgings: Beachy Head's motion was all cut from slab and machined to profile, rather than forged.

On materials specifications: old ones are often not the same as what is available today, which can make repairs difficult. P class No. 27 will be getting complete frame plates, in part because it proved difficult to guarantee a high-quality weld repair by cutting out wasted material and welding in new steel of possibly different grade.

On drawings: I seem to recall a statistic that the original LB&SCR drawing set for Beachy Head comprised of 200 drawings, but the new loco will require about 1,000. Some of that difference is because the 200 doesn't include standard fittings that would come from stores; but some of it is because workshop practice was just "make to fit" in some areas that are now drawn: the layout of pipework between the frames (there is air, vacuum, steam and lubrication to fit) was probably done by hand originally, whereas has now been done according to a drawing.

Tom

 

You mean like Martin Evans' Simplex which caused a stir at the time as it has only girders supporting the inner firebox roof, no crown stays? One gets the impression that there is a great reluctance to deviate from the original design, witness the expense of replacing syphons on Bulleid boxes.

 

Couldn't agree more.

To take nothing away from the Engineers of the past, they could only dream of what we can accomplish now with the aid of technology. To turn out 100,000 parts within 2 thou of each other, would have been a formidable challenge then, and is often routine nowadays.

God only knows what they would have thought of 11 Axis live tool CNC machining and fully meshed components being analysed on 1million node plus FEA solvers!

3D fully parametric modelling, with integrated drawing tools, has moved design on leap years. That you can determine, what fits and how to make it fit, without metal ever being cut, or models ever being made, is a revolution.

It is my belief that the new builds are, on tolerance and fitment terms, far improved, not mention the improvements in manufacture and design allowing more flexibility and assurance.

I suppose I better add a caveat to this. The opportunity for new builds to be better engineered and manufactured from the outset exists today, due to the technical advances mentioned. However, it is, as always, up to the engineering leadership how well engineered and made any locomotive is.

 

Testing design practicality is certainly one useful application of model engineering, which it's great to see is already happening.

I wonder how much computerised machine control interconnected with CAD has skewed the production of components (especially for small runs/one offs) from forging to machining? The percentage of waste material is clearly far less of an issue for heritage/new build applications than for BR sized production runs. Whether the costs thus incurred are counted against raw materials or machined components is a question Steven (@Bean-counter) is infinitely more qualified to answer than me.

Could this route be productively extended to establish suitability of novel material applications? Thinking here of the comments in earlier posts concerning fireboxes, especially the issue of resistance to corrosion with currently available grades of steel. If it proved possible to marry small scale proving with computer modelling to establish reliable baseline data, it would be a significant benefit to all future new builds, perhaps even to getting an entirely new design, with few or no 'grandfather rights' accepted. It might even settle the perennial 'Leader' debate once and for all!

This is pretty much bang on the sort of potential for design features which I had in mind. Isn't the new Blodge double fairilie boiler design adopting something pretty close to this feature?
https://www.festipedia.org.uk/w/images/d/d4/JSBoiler.jpg
Some improvements to Bulleid thermic syphons wouldn't exactly go amiss either.

I suspect there will be more sympathy for future proofing steam technology than you might expect Martin!

The key issue is in the confidence of vehicle acceptance bodies in computer modelling. The balancing of wheelsets by A1SLT is a case in point. T'would appear that experience gained with 'Tornado' was used to produce a computer model of the balancing for the P2:
https://www.p2steam.com/2016/10/31/getting-balance-right/
With the wealth of knowledge being gained up by the various new builds, the potential for a comprehensive reference database exists.... for the first time since 92220 left Swindon all those years ago, and this time, it's bang up to date with the latest CAD developments.