Valve gear design

I have made a good restart on my 5" gauge Gresley rebuild of a Super Claud with round top box and original slide valves. Having started on the tender this is nearly complete.

I have redrawn the loco frames which depart considerably from the Martin Evans design and I wish to incorporate valves under the cylinders as per the original possibly using the Princess of Wales castings.

I would like to ask if it is a good idea to scale the original full size valve gear design for which I have suitable drawings? The original Clauds were known as free running and powerful engines for their size and it would be good to have this replicated in miniature without having to go back to first principles in gear design - always assuming that the Martin Evans version may have shortcomings.

Any guidance would be greatly appreciated.

John B.

Just to add to my post above - the Martin Evans Claud valve gear design was of course indirect drive.

John B

I'm assuming this is Stephenson's valve gear. If so the definitive work is by Don Ashton, I think Camden Road books seel it, there is some good stuff at

**LINK**

Edited By duncan webster on 10/11/2017 01:16:47

Edited By duncan webster on 10/11/2017 01:16:59

Thank you for responding so promptly, Duncan. Yes, it is Stephenson's.

The question is whether it is feasible to replicate the full size gear in miniature without going back to first principles. I have looked at the Don Ashton material but can't discover whether he makes this point directly though the implication appears to me to be yes.

In other words, if a valve gear is effective in full size - can it scale successfully into our miniature gauges?

Kind regards

John B

In principle yes. You could download one of the simulators and enter the dimensions, then you'd know for sure. There is a school of thought that little negines need bigger than scale port openings. I think this is false, as our piston speeds are a lot lower than full size.

Thanks Duncan. That sounds a better plan than me starting from the beginning and then including some self induced errors. Which is more than likely.

Does anyone else have an opinion?

Best wishes

John B

"Does anyone else have an opinion?" on scaling from full size

LBSC certainly had !

His 5" GWR "Pansy" had Stephenson's and slide valves beneath the cylinders. Since it is a powerful little loco, free running etc, is it worth looking at his valve design, especially the porting ?

Described in Model Engineer Volumes 118–120.

I'm sure I'll be corrected if I'm wrong but I think that model engines tend to use over-scale valve travel to avoid really tiny valve ports and more tricksey adjustment..

Neil

Hi John,

We briefly discussed this earlier this year on one of your other threads.

I think the simple answer is no, you cannot scale down the fullsize Stephensons valve gear but this must be qualified as it depends on the loco and arrangement of Stephensons and much else besides.

One of the classic 'cock ups' was Keith Wilson's 5"g 'Bulldog/Dukedog' valve gear - which is also relevant because it is pretty much the same arrangement you wish to copy. Direct drive Stephensons with loco links and slide valves below the inside cylinders. No way does the gear produce enough travel for the valve events required.

So unless you are vary careful and are aware of the pitfalls you could end up in the same mess Keith Wilson landed himself in.

Neil is absolutely correct that miniature locos require or ought to have greater travel of the valve than what would be scaled down from fullsize.

You start with the cylinder block and what width the steam ports and exhaust ports will be. (Only the width of the steam port is relevant for valve travel). Then how much lap do you want on the valve? Then in fullgear (say 78 or 79% cut off for a 2 cylinder loco) you can calculate the required valve travel required.

Then you have to consider the type of expansion link employed (launch type or loco links) and within reason you can predict the eccentric throw required.

Note with slide valves (outside admission) loco links are preferred with direct drive, and launch type links with indirect drive. In each case there will be a small suspension offset and a small amount of die block slip so the eccentric throw need only be a tad more than calculated.

If you have loco links with indirect drive, or launch type links with direct drive (for outside admission/slide valve) the suspension offset will be considerable to achieve equality of valve events and there will be considerable die block slip which also requires a greater eccentric throw and larger expansion link to compensate/accomodate the excessive die block slip.

Note Martin Evans used direct drive and launch links on Princess of Wales with slide valves and specified a totally inadequate amount of suspension offset.

Fowlers Fury mentioned LBSC's Pansy. I can state quite categorically that Pansy has a very poor version of Stephensons gear, but can easily be rectified.

John, you also have to consider the expansion link suspension and whether centrally suspended or suspended by the top of the link if loco links.

Anyway the first thing is to decide on the cylinder block and whether for a 5"g loco the size of Super Claud you have say 5/32" or 3/16" steam port width.

Cheers,

Julian

Thank you Julian. Yes we did touch on this earlier.

I am bound to be seeking further advice on this. But for now, it occurs to me that most of the designers we know, LBSC, Martin Evans, Don Young and even Keith Wilson, a professional model engineer if there was one, weren't necessarily that clever at this very heart of loco engineering.

So where does this leave the rest of us?

I have been looking again at the Stratford Great Eastern drawings and I can see now why the question of valve travel is an issue for model scales.

With best wishes

John B

Hi John,

There are a significant number of designs where Martin Evans made a hash of the Stephensons valve gear (and also Walschaerts). Don Young always made a number of errors on his Stephensons valve gear when using loco links as he drew the gear out incorrectly and didn't understand what he was doing. Keith Wilson was ok doing the classic Churchward GWR arrangement of Willie Pearce's gear for Stephensons which is less indifferent to scaling down, but when Keith tried other arrangements he was seriously out of his depth. I dont think that Keith ever understood valve gears properly.

LBSC made serious mistakes on all his 5"g valve gear designs except his Joy valve gear Minx and Maid of Kent.

What we know now thanks to H S Gowan and Don Ashton really only was made public from circa 1974/5 onwards.

Don't worry. What you decide to do can easily be checked on a valve gear simulator program these days before you cut metal.

I have been involved in re-designing a number of well known ME loco valve gears in recent years for various builders. John Baguley has done the same. The understanding of the niceties of Stephensons valve gear is a fascinating topic and can become an obsession. If correctly applied and understood the superlative performance on the track is well worth the time and study involved.

Cheers,

Julian

Thanks again Julian.

What a fascinating subject this is. And no doubt similar errors had also been made historically by many loco works drawing offices and the trains ran despite it all; not sparing the fireman's back!

Despite this our little engines have been happily rattling round the tracks seemingly for ever.

Don Ashton is very critical of traction engine design. While much is justified it is recorded that the finest of them all, the Fowler Superba ploughing engine of 1934 with Firth single eccentric gear, produced an efficiency approaching the contemporaneous GW Castle Class.

With best wishes

John

Edited By John Billard 1 on 10/11/2017 21:58:58

I have learned much on this site since I started this thread but think that I have some way to go.

To recap I am looking at Stephenson's gear, direct drive with loco links. Taking the original Martin Evans valve dimensions as a start we are looking at valve travel of 5/8", lap of 5/32", port opening in full gear 5/32".

The original loco had a cylinder incline of 1 to 16 to horizontal centre line through the axle, and valve rod incline 1 to 8 to centre line of cylinders; valves below the cylinders

I have looked at the on line data from Don Ashton and his references but I have yet to find a simulation that would tale me further. (Though I have yet to have Don's book).

Advice on my next design steps would be very much appreciated. (I remain a pen and ink person).

Kind regards

John B

Hi John​
what version of Windows are you running? Some (most) of the older modelling software can be difficult but not impossible to run on W7 and W10. Once I know which you've got an appropriate package will be sent. However, without Don's book you will struggle. All the models do is tell you what valve events you will get, not how to get good events if you see what I mean. It wouldn't be fair to him for me to send you an electronic copy as it is copyright (I assume)

Thank you Duncan.

I am on Windows 10.

I'm quite happy to buy a copy of the book; just haven't had an opportunity yet.

Best wishes

John B

John, I've sent you a PM

Hi John,

Don Ashton's book is available from Camden Miniature Steam Services.

On pages 28 and 29 you will find a worked example that would fit your requirements. Lap is 0.120" and note it is not necessary for the steam port to open fully to incoming steam so valve travel is a bit less than your 5/8"

In Don's example the eccentric throw is 0.460"

Cheers,

Julian

Thanks, Julian, very helpful.

John

Thanks again Julian

Looking at the Ashton worked example the difference is that I have a much shorter connecting rod at 7.44" compared to the worked example of 10.25" (Presumably the worked example refers to a six coupled engine or similar unlike my 4-4-0).

So far I am getting wildly out valve events so far on the simulation and I have to admit that my ancient O Level geometry is struggling a bit!

Don Ashton's work is superb but not particularly user friendly for my level of thought. Any pointers would be appreciated.

With kind regards

John

Hi John,

I have sent you a PM.

Cheers,

Julian