Freelance loco, 3D valve gear drawings

[John LluchParticipant @johnlluch21161]

After reading the book from "The model steam locomotive" by Martin Evans and "Design Procedures for Walschaerts' and Stephensons's valve gears" by Don Ashton, I started to draw my own designed loco in Solid Edge ST6.

That's the first time ever I used a 3D CAD software to draw anything, but I guess my willingness to build a locomotive made me learn how to use the software. My previous skills on (manual) technical drawing were limited to what I learnt during my chemical engineering studies, but I hardly needed them at work. Actually, I never used a professional CAD package before, even a 2D one. So, to be honest I find the Solid Edge software a bit tedious and frustrating at times, though I am sure this is due to my lack of skill than to the software itself.

See below what I achieved so far. My goal is to design a live steam loco that roughly matches the 5" Britannia dimensions but with the following characteristics:

– No importance is attached to the model being an actual replica of a full scale locomotive, but it still should look and be clearly identifiable as a steam locomotive. The loco will have free dimensions.

– Get rid of imperial units in drawings. All dimensions will be in plain millimetres. Parts will have exact measures and will be arranged in ortogonal or horizontal/ vertical planes for as much as possible.

– The assembly will be made of standard industrially available parts where this makes sense and it would deliver an improvement over what is usually used in current designs. For example, I may use lubrication free flat miniature bearings on rods and levers, or lubrication free ceramic bearings on wheel axis. I will use commercially available industrial grade miniature valves, safety valve, regulator, water gauge and so on if possible.

– Stainless steel and aluminium will be extensively and primarily used for all parts, and will be chosen over copper, brass or gunmetal for as much as possible. Technical grade polymer materials will be considered as well. Eventually even the boiler will be evaluated to be built out or AISI 316L steel.

– One objective of the project is that all parts should be carefully 3D drawn in a way that most parts can be automatically build by CNC (computerized numerical control) equipment, or laser/water cut directly out of the drawings, with minimal hand made retooling required if at all.

The video below shows what I came out so far for the valve gear design. I chose a Baker valve gear geometry as opposed to a Walschaerts because it's easier to build and I think it's easier to assemble as well.

In theory getting the correct valve events and timings on a Baker valve gear is a bit tricky, but I guess that after much Solid Edge simulation and the study of the Don Ashton book, I acquired an understanding of the gear so I feel now comfortable with it.

I will come with a ton of questions as my design progresses. So far I would appreciate any comments on what I achieved so far, as well as any visions on my (particular) approach. I hope I am not doing anything totally wrong. I am a total beginner so basically I hope my approach is not totally biased towards the wrong side.

The video shows the loco in forward motion first, then I switch the reverse yoke to backwards, at this time the motion still appears to run forward because the simulator just makes the wheel turn in a specific direction, but in fact the valve events are the correct ones to produce a backward motion. 

Here's the video:

http://youtu.be/hzJr6GkfezY

 

Thanks in advance

John

Edited By John Lluch on 26/09/2014 22:35:53

Edited By John Lluch on 26/09/2014 22:44:37

[John LluchParticipant @johnlluch21161]

My first attempt at drawing a 5″ locomotive

[julian atkinsParticipant @julianatkins58923]

hi john,

i think you are on a bit of a flight of fancy here. there is nothing wrong with the 5"g Brit walschaerts valve gear. use of aluminium on a 5"g steam loco and a steel boiler is really quite bizarre for a 5"g Brit. just build it to the Perrier drawings with a silver soldered copper boiler and proper cast iron cylinders with cast iron liners and cast iron piston valves with cast iron rings. many award winning locos that have also entered IMLEC have been built to same specifications and been incredibly successful efficient locos also capable of long trouble free service hauling many many passengers every weekend on club tracks.

cheers,

julian

[John LluchParticipant @johnlluch21161]

Hi Julian, Thanks for your comments.

It is not my motivation to build a scaled down replica or to buy a "kit" and assemble it, but to produce something different. It is not also my motivation to ride the loco. I just seek an opinion on design choices.

Particularly, I dislike the walschaerts valve gear because it inevitably has "sliding" parts. Of course, it works, is easy to understand, and it was the most used in the UK. I am also aware that the walschaerts was recently chosen for the 5AT locomotive.

On the contrary, the Baker valve gear has just "rotating" parts in it, no sliding parts, which I see as a clear advantage both for making and for running it. I can't see anything inherently wrong in it. Also, I have spent a significant amount of time simulating the valve gear until I got a reasonable understanding of how it works and how to optimise it.

About materials, I mentioned aluminium as I mentioned polymers (a.k.a plastic). All these are materials are extensively used in the transport industry (automotive, rail, and so on).

For example, you can find both lineal and rotative bearings made out of a special high strength polymer that will run maintenance free and no wear for a very long period of time even at temperatures above 200ºC.

I also do not see the point on using cast iron except for the sake of mimicking the full size design, which is not what I intend to do. For such relatively small pieces on a model locomotive I think we can just use blocks of extruded or laminated stainless steel and machine or bore them to the appropriate dimensions, no?. Other than that you must be a bit more careful with the tooling and the machining speed, what is the problem with that?, In turn, stainless steel will last forever, no?

John

 

Edited By John Lluch on 27/09/2014 01:20:18

[MuzzerParticipant @muzzer]

John

That looks like interesting work you are doing there. You are taking the same approach I would if I were to make some form of steam locomotive. I have to admit I'm not a model maker myself but I prefer to do some of the design work rather than just making to a drawing.

I used SE ST6 a fair bit last year and got to quite appreciate it in the end. There are frustrating aspects to it like any CAD system but it has some pretty nice features. The optional Synchronous Technology takes some getting used to but I can see the concept now and it makes sense for some stuff. I've been using Solidworks 2014 for most of this year and it's also pretty nice to use, with its own little annoyances. I'll probably go back to SE when I return to the UK in a few weeks and lose access to the SW license – quite looking forward to it in some ways.

Keep us updated on progress – I'd love to know how you get on. I'm sure you will get plenty of advice, hopefully much of it constructive!

Murray

[julian atkinsParticipant @julianatkins58923]

hi john,

believe it or not i will incorporate any 'modern' design methods/materials to achieve a successful efficient miniature loco in 5"g. however a steel boiler for a 5"g Brit type loco is a definite 'no!'

stainless steel is a very strange material for working parts in a steam locomotive. it has very peculiar wearing characteristics. it has no advantage over traditional materials, and lots of disadvantages! i was warned of same some 30 years ago by a very experienced marine engineer!

i have known some very eminent model engineers who have built all sorts of miniature locos over the years who have experimented with all sorts of different materials, and in every case they have ended up reverting to traditional materials as used in fullsize loco construction, for very sound reasons.

ok, you might want to employ one of the modern plastics as a bearing material for coupling rods etc, and PTFE piston valve rings if gunmetal cylinders. but IMHO that's as far as it goes employing 'new' materials!

the baker valve gear is completely non prototypical for a UK fullsize loco in miniature. i thought you were originally hankering after such a prototype? LBSC was the first person to build a loco in the UK with Baker valve gear albeit in miniature. however in fullsize it didnt 'catch on' in the UK, and in the USA fell out of favour to traditional walschaerts valve gear for very good reasons.

anyway good luck with your design!

cheers,

julian

[JasonBModerator @jasonb]

John, you say you won't "ride" the loco which I assume means you won't be running it on a track so why are you worried about the lonevity of stainless steel and wear rates of bearing surfaces if its going to be a static model?

If you are actually intending to run it then the use of aluminium and lightweight materials will affect the tractive weight so you could end up justy spinning your wheels. . Also think of electrolitic action between the aluminium and other metals when it gets steam on it and in the joints.

I don't see why the stainless boiler you  mention is not a possibility though beyond what a club inspector can deal with so a bit more costly to get tested, the extra weight wil help compensate for the lighter materials used elsewhere

Edited By JasonB on 27/09/2014 08:18:23

[John LluchParticipant @johnlluch21161]

Murray, yes I suppose all 3D CAD programs tend to converge over time in some form or another as users request features only available on one of them. My choice of Solid Edge was merely based on what was being used by a friend of mine. He taught me the basic aspects (mostly conceptual) and then I began to improve by looking at tutorials or youTube videos showing other people's use of it. I definitely will keep posting my progresses

Jason, Oh no!, I didn't mean I won't be running the loco on a track. I definitely do not intend it to be a static piece of work. I will move next year to a location (Fornells de la Selva) that has a medium sized track for 5" and 7.5" locos and a active model club I have already joined. They already own a 2-6-2 live steam loco and a bigger one is currently being build by one of the old members. They also have a bunch of battery powered engines for both the 5" and 7." track. Every two weeks they meet to run their locos for the public, so I will have plenty of occasions to do so as well. It is just that my primary motivation is to design the loco, not particularly riding it.

About aluminium, I just mentioned it for the sake of it. You are right, it should not be used for structural parts and its light weight would affect tractive efficiency if used extensively. I suppose it could be considered if we were talking about rolling stock, just maybe. Or it can be used for example as the boiler skin to keep the isolating material in place. I like very much the look of aluminium, so I guess it can be used that way if I decide not to paint parts of the loco.

A stainless steel boiler is however definitely possible in my opinion. Maybe even one that has *no* or very few soldered parts in it, as bizarre as it may sound. This is something I want to investigate. I've designed industrial heat exchangers for some time in the past, and copper was never an option on that field. So, I think there may be some room for improvement over what it is currently being used on model steam locos. You can call me a fool, but I don't think that cooper is used on a steam powered electric power plants. Said that, I may end to the conclusion that copper is after all the right material to use on the smaller scales, specially given all the advice I am getting against not using it, which I really appreciate. I am not at this stage yet though, so more will come !!

Julian, I would appreciate some more advice or opinion on whether there may be an actual issue or concern on using the Baker valve gear as opposed to the walschaerts. As said, I like the former over the latter because it has no 'sliding' parts and thus it is easier to build and possibly it ends with less overall wear after use. To my understanding the one that produces the most accurate time events is the Stephenson one, possibly followed by the walschaerts.

The Baker produces a somewhat less lineal movement on each side of the piston valve (i.e not fully symmetrical), but this can can be minimised if the levers are relatively long and properly orientated. I was barely there during the steam powered times, as I was just a very small child when steam locos were discontinued, but for what I found on the internet the Baker was only used in the USA, and it seems to me that its significant rate of adoption was more based on aggressive commercial actions by the manufacturer than on real benefits on full sized locos. However, my point is that on the smaller model locomotives the Baker may have its strengths specially if lubricant free polymer based bearings are used. What do you think?

Thanks

 

 

 

Edited By John Lluch on 27/09/2014 09:29:23

[Neil WyattModerator @neilwyatt]

Hi John,

You've opened a can of worms! You will inevitably get much conflicting advice, some good, some just prejudice.

I take it your fundamental aim is to see what different and new options you can apply. As Julian says some of the traditional solutions continue simply because they work best, but as new things come along how are we really to judge them if no-one tries them out? Perhaps you could try out some ideas in electrically driven test jigs?

Not following a full-size prototype will make it easier for some people to accept these new things as a 'test bed' – there have been some fascinating models made in the past plastered in excess gauges, pipes and 'features' that would have been considered impractical monstrosities in full size, but they were fascinating exercises in themselves. Freed from a prototype you will not be tempted to compromise!

I would encourage you to aim for a proper working loco as the end point, if only because some of your ideas would benefit from being proven by extended use and hard work.

Neil

P.S.; My personal view is that the best valve events of all would be provided by solenoid poppet valves under microprocessor control, followed by corliss valve gear, but LBSC put the kybosh on using poppet valves for locos half a century ago…

[JasonBModerator @jasonb]

Sounds like you are in mainland Europe in which case there are quite a few stainless boilers running about so you should be fine there

Yes as cladding and trim the aluminium should be fine, could even use it for the tender and bond it together no need for rivits if its a freelance people won't be able to count them Funny enough you would be surprised how much aluminium and JB Weld were in a certain steam roller that got a Gold at the ME show last year or the one before!!

Keep us posted of your progress

J

[John LluchParticipant @johnlluch21161]

Thanks Neil. Your points are excellent, and yes you fully got what I intend to be at. My stronger motivation is to try different things even at the risk of some of them not working, and of course I will be sharing all my experiences -good or bad- during the process. And of course my ultimate aim will be to produce a long working, reliable loco. The boiler is the one so far that has spurred the more inflammatory comments, but this is something I leave for a later stage.

Initially, my intention is to build and test some moving parts on compressed air, possibly under stress. This of course does not anticipate how they will work on hot steam but I guess some useful data could be extracted from that.

As you said, implementing popped valves managed by a programmable controller (possibly using a digital rotary encoder on the driving wheel as the input device) you could get the absolute perfection at all times, also considering the actual speed and load of the locomotive to dynamically modify not only cut-off but also lead, and considering as well different timings for intake and outtake events. However, I think that attempting this would be going too far because one objective is to design a loco that actually looks like a steam loco and one that can be made by everyone out of standard commercially available parts if proven successful. This is why I am proposing the Baker valve gear design on this case.

John.

[MuzzerParticipant @muzzer]

I must admit, I love the idea of an ECU for controlling a steam engine's valves. Electronic steam injection! Even scale model engines are turning pretty slowly compared to IC engines or electric motors, so you should be able to control the process quite easily in any number of current platforms, possibly including Arduino / Raspberry Pi. You'd need to calibrate it on a dyno of course…

Murray

[John LluchParticipant @johnlluch21161]

Variable valve timing is definitely used on high performance internal combustion engines, such as sports street cars and racing cars.

On street sports cars (say BMW M3-M4) valves are still mechanically driven from the crankshaft but an electro-hidraylic system is continuously adjusting the "cam-phasing" on both the intake and outake valves to achieve the best possible performance and low emissions at any time.

On track racing cars (say F1 cars) a more complex system is used that involves not only cam-phasing adjustments but also duration. In such case I believe (not sure though) that the actual valves are directly driven by the system by means of electric solenoids, i.e not from the crankshaft through mechanical means. This of course requires the system to be fast enough to cope with the rotation speed of the engine.

My guess is that although all this is possible and definitely much easier on steam engines -simply because they turn very slowly, as you point out-, I wonder if this would bring any benefit at all. Maybe from the point of view of efficiency we could squeeze one or two drops of water, but I believe that hardly more power could be achieved from the engine.

Efficiency on a rolling steam engine is very poor for a multitude of reasons starting by the fact that we need to heat water before we can use it and we need to keep it hot and build pressure at all times regardless of the actual load conditions, so a lot is lost through the relief valve alone!. A fully controllable cut-off should achieve some efficiency improvements on the cylinders though.

From the point of view of raw power I think that no improvement could be achieved. Max power and torque is obtained at high steam pressure and long cut-offs (at the expense of efficiency). In this scenario I believe that variable timing can not deliver any extra power to the system, essentially because existing mechanical systems (valve gears) are already doing what it is supposed to do.

However, this does not preclude us to implementing if for the sake of it, of course . In such case I think that even a small sized, off-the-shelf, PLC (Programmable Logic Controller) has enough processing power to *fully* control a steam engine.

These devices are readily available and I have plenty of experience with them, so that part should not be a major issue. However, I am not aware of any electrically driven, fast switching, small valves that could be used reliably on hot steam. Could you point the community to some kind of solenoid valve that could be used to this purpose?

John.

[nigel jones 5Participant @nigeljones5]

F1 valve control is universally by pneumatic valve actuation…very fast!

[Neil WyattModerator @neilwyatt]

I think Corliss-style valves might be the way to go. Some sort of dual actuator that uses a stepper to set the lift and a solenoid to actually operate the valve. very bulky and impractical for a loco. Balanced double drop valves could be operated by small solenoids. Locos would probably demand poppet valves, but bear in mind that the spring rates would be much lower than for an IC engine.

I like the idea of a steam engine that could be controlled from flat out to a crawl, and even jiggled back and forth like a stepper motor. I imagine this would need a lot of 'fuzzy logic' to get the fine control. Fizzy logic?

Neil

[nigel jones 5Participant @nigeljones5]

no such thing as Fizzy logic…..ask my wife

[MuzzerParticipant @muzzer]

Could a gasoline fuel injector be made to work as a servo control valve? They only seem to be rated up to 3-4 Bar but I expect they will be automotive temp rating ie probably 125C. A high flow-rate injector from a large engine would be easy to drive – but would a lower pressure servo circuit be an issue? Not being a steam operator, I'm not familiar with the sort of steam pressures used in scale engines but it looks as if 120psi may be typical for the primary circuit.

Another option might be the electromechanical "unit injector" (bypass) type diesel fuel injectors used on commercial vehicles. Presumably these must be rated for very high injection pressures, high flow rates and automotive temperatures. The injection is done by the camshaft but excess fuel is released by opening a valve with the solenoid part way through injection, allowing electronic control of the fuel quantity. That solenoid must be quite suitable, although obviously it requires more power to operate at 3000rpm (25 injection events per second) or so. I think the drive voltage is around 100V.

Current electronic diesel injection systems used in cars operate up to 2000 bar(!) and you would struggle to drive any of the injectors.

Murray

[julian atkinsParticipant @julianatkins58923]

i think that some of you are missing the point here!

ive built quite a few locos to don ashton's valve gear specifications with greatly improved performance and efficiency. effective superheaters and a good steam and exhaust circuit are also important, as are good springing, and draughting, plus attention to free gas flow through the boiler and grate and ashpan design.

when you've driven a miniature loco with valve gear that can be well 'notched up' (with decent superheaters and the draughting etc is ok) nothing else compares! it's pure joy and a wonderful experience!

cheers,

julian

[Author]

Posts