Walschaerts' valve gear.

I've seen it stated that by having the radius rod in the upper half of the expansion link (as per American S160 2-8-0s and Southern Rly Maunsell 2-6-0s and 2-6-4 Tank locos), this gives better valve events than if the radius rod is operating in the lower half for that direction.

I can't see that it makes any difference myself, but just out of interest the final batch of S160s were turned out with the normal layout, i.e. radius rod in the lower half of the expansion link for forward travel.

Thoughts, anyone?

 

I have a copy of the valve gear drawing for the N and U classes. The valve events are tabulated on this drawing and are better in fore gear than in back gear. I checked this out using the Alan Wallace valve gear simulation programme and this confirmed the drawing is correct. I think this must be due to there being less angularity errors with this arrangement. This valve gear also has the link tail pin with a negative back set. Playing about with this measurement in the simulator confirms that this arrangement improves the valve events. My conclusion is that there were some very clever valve gear designers about a century ago.

Sent from my SM-A105FN using Tapatalk

 

I can only agree. My trigonometry certainly isn't up to understanding exactly how all the linkages behave.

 

Don Ashton's web site is interesting for such things. http://www.donashton.co.uk/html/walschaerts__gear.html I only wish I understood it...

 

I gather that steam engines of all sorts are generally designed so that the gear is in the 'down' position in the normal direction of rotation, in case the linkage breaks

 

Jimc's link to Don Ashton's site on the Walschaerts page provides the following little understood analysis:-

"It will be found in most cases that given equal depths in gear for forward and reverse working the percentage of cut offs in back gear will exceed those in forward gear by as much as 6%. Although some engineers did not favour using the upper expansion link half for forward running because the drive is then indirect, the position is in fact little different from the many cases of rocker use in valve gears, though the results are not at 180 degrees. In the average inside admission arrangement there is therefore scope for reducing the pitch circle using a shorter leading return crank, such that an adequate foregear starting cut off is maintained. Holcroft was evidently aware of this and promoted it on the Southern Railway"

In the usual arrangement of Walschaerts, with the die block in the top half of the expansion link, you get a peculiar effect of the geometry that results in backgear (reverse) having more travel than the die block being in the bottom half of the expansion link for forward gear.

There is a benefit in reducing the return crank throw - therefore if you set the return crank 180 degrees from normal, you can use the top half of the expansion link for forward gear, and have a shorter return crank throw.

It is quite simple, really.

Holcroft understood this, and also arranged for the lifting arm off the weighshaft to face 'backwards', reducing die block slip.

So long as the extra wear on the expansion link suspension trunnions is dealt with, (beefed up a bit), you get a much better design envelope for the geometry of Walschaerts gear.

Holcroft was indeed a very clever chap, as is my old friend Don Ashton for recognising all this some 60 years later!

Cheers,

Julian

 

Can you or someone please explain that? The motion of the expansion link itself is the same regardless of the setting of the reverser. Is the difference something to do with a different effect of the angularity of the radius rod when the loco is going forwards or backwards? And what about the alternative means of moving the die block up and down, by the valve rod either having a slot that runs to and fro over a moveable block or being suspended by a link?

Edit: corrected to mention the valve rod.

 

It's not easy to explain with just text. Are you familiar with the late Charlie Dockstader's programmes on valve gear? These are most informative if you are interested. They can be downloaded at: http://www.billp.org/Dockstader/ValveGear.html
If you open the Walschaerts simulation you will find that the radius rod is moved up and down by a lifting link (as on a BR class 5 loco). In forward gear with the die block near the bottom of the expansion link, the lifting link supports the radius rod in such a way that, as the expansion link moves back and forth, there is very little relative movement of the die block in the expansion link. However, if you now put the gear into reverse, you will see that the die block moves up and down the expansion link quite significantly, known as die block slip and giving quite different valve events. If you have a radius rod that is supported by a fixed, rather than a swinging point (as in a Black 5) there is still some die block slip but it is more or less the same in forward and reverse. If you want to simulate this in the diagram, first turn off the error beep in 'setup' then go to editdimension - framedimension, and set the radius rod extension and the lifting link length to zero and make the reverse arm pivot vertical equal to the link pivot vertical.
HTH.

 

I had downloaded those some time ago. I have now followed your instructions and I see what you mean. Thank you.

 

I've been mulling over MellishR's post No. 7 this afternoon, trying to find words to describe the situation as admirably set out by Steve, however, without success.

 

That is frequently the case but not necessarily so. As is mentioned above, you can have Walschaerts with forward in the top half of the expansion link. You can also have Stephenson's link such that the link is raised for forward and there are locos like this. In such cases, a broken lifting link will cause the gear to default to reverse.

 

This is one of the better threads and raises some worthwhile issues, so thank you to @Eightpot for initiating it. I have never been totally convinced that the oft given reason for a following return crank is to mitigate the effects of breakage of the lifting links, is correct. "Modern" steam locos on China Rail generally had the return crank following, but some older designs (e.g. the JF 2-8-2s) had examples of both following and leading. Chinese textbooks on steam locos are quite scholarly (they seem to be partly copied from Soviet examples), and the working of Walschaerts valve gear is set out in some detail. I have been translating one of these chapters, and it basically says that modern loco types use a following return crank as the die block wear is less in the lower half of the expansion link, which tallies with the concept of the greater die block slippage in the upper part.

 

I have attached the tables of valve events from the valve gear drawing for the N and U classes and an overall view which shows the slightly unusual setting of the link tail pin. Notice how the cut off and valve opening is reasonably equal between front and back ports for most cut offs in forward gear. The events are not so even for backward gear and are very different for full backward gear. The valve gear is set up to work best in forward gear in the normal range of cut offs, which would probably be between 25 and 55%. Studying this drawing makes me realise how much work was put into this one aspect of the design. The drawing is for the later batches of N and U classes. I do not know how the earlier batches were different.

 

It appears that Harold Holcroft perfected Walschearts valve gear but was ignored!

 

Thanks. I note the N has a leading return crank. The expansion link pin position is not quite as unusual as that on 46200/01 - see attached.

 

In awe of anyone who can visualise the implications of these tables... without a graph or a 3d animation its just numbers to me