What the heck is "cross-porting"?

[Keith S]

Recently while reading about live steam engines, I ran into a reference to "cross-porting", which seems to be a procedure to mill channels into the valve face of a slide valve cylinder to change the side of the cylinder that each port feeds to. Apparently this was done on the engine iquestion so that the valve rod would be on the (prototypically) correct end of the expansion link for the direction of travel.

I'm wondering why this is necessary. Other than being very small, the slide-valve-equipped engines have cylinders that work the same way as big ones. Why would the valve gear operate backwards in this case? When running forwards, the valve rod on my "Billy" is at the bottom of the expansion link. Is this not the case on full-size locomotives?

[david bailey]

Cross Porting is used where the model is using slide valves but the prototype used Piston Valves, therefor with walchaerts valve gear the combination lever would be connected to the radius rod above the pivot for the valve rod, but a slide valve needs it to be connected under the valve rod pivot, so to get the valve to operate correctly and to have the valve gear looking correct the ports are crossed to acheive this, also most prototype engines with this valve gear have the radius rod in the lower position for forward running, exception is the Maunsell Engines used on the Southern Railway.
I built a batch of D&RG K27 2-8-2 Tender engines where I used cross ports to achieve a correct looking valve gear.
David Bailey

[Keith S]

Thanks, David. That really clears things up. So on a model with slide valves, OF a prototype with slide valves, the gear is correct? But when the prototype has piston valves , the model's slide valves need to be cross-ported.

Why don't they just use piston valves on the models?

[bessytractor]

Thanks, David. That really clears things up. So on a model with slide valves, OF a prototype with slide valves, the gear is correct? But when the  prototype has piston valves , the model's slide valves need to be cross-ported.

Why don't they just use piston valves on the models?

in my experience its because slide valves are easier to make.  My 7 1/4 pacific should have piston valves but I didn't feel confident in my ability to make them so I have used slide valves.

Having said that the Beck and Merlin locos had piston valves.

[Keith S]

Maybe, too, it helps to keep inventory levels smaller, if there is only one type of valve that needs to be in stock.

[Pretoria]

I don't wish (?) to complicate or confuse the thread, but I suggest the answer to the original question is a tad more complicated.

On a lot of 16mm models, the valve gear is not really Walschaerts at all -- it is advertised as "simplified Walschaerts".



In true Walschaerts, the basic valve motion is provided by the combination lever and union (or drag) link (the vertical rod connected to the valve spindle and link connected to the crosshead) -- in mid-gear, these two levers move the valve enough to correctly open the ports by the amount of lead at each end of the stroke -- because in this position (mid gear) the radius rod (the rod that joins the expansion link to the valve rod) does not move, this means that the valve movement is unaffected and (if correctly proportioned) also means that the engine will run in either direction with the small amount of steam admitted by the lead. It will run very poorly because the exhaust events are similarly restricted -- but my 3.5" gauge loco will do this. As the link is put into full gear, the motion and direction of the valve is increased significantly and favours either running forward or backward dependant on which end of the expansion link is in use. David Bailey was quite correct in his description of the two connections at the top of the combination lever to give the opposite directions of motion required for either inside admission piston valves or outside admission slide valves.


Returning to the "simplified Walschaerts" gear -- the combination lever and union link are either missing altogether or are connected to the valve spindle at the end of the radius rod and are provided solely to give the impression of Walschaerts -- they do nothing to affect valve events. The valve motion is thus driven by the the return crank (effectively an eccentric on the main crankpin), and driven through the expansion link -- the two ends of which move in opposite directions, giving forward and reverse gear movements. This will only work with "all square" valves and no expansion cut-off. In fact, the valve gear is really a simple eccentric valve gear.

Now comes another difference -- it is "normal" practice for the radius rod to drop to the bottom of the link to achieve forward gear -- this is done in full-size to provide the nearest to a "straight line" set up to minimise out-of-line forces in the various bearings and pins (accepting that working in reverse does not generally occur so much) -- and the two connections at the top of the combination lever can be interchanged for either slide or piston valves to correct the movements. Unfortunately, for the "simple Walschaerts" case, the "normal" return crank position would give the opposite movement required for outside slide valves. So there are four ways round this -- each of which is necessary to alter the valve movement by 180 degrees:

1) For forward motion, run the gear with the radius rod in the "normally" adopted reverse position (easiest and cheapest solution)

2) Cross port the cylinder -- in other words, connect the front port to the back of the cylinder, and the back port to the front of the cylinder. (possible, but complicates cylinder machining and wastes a lot of steam in the extra length ports, and costs more to achieve commercially)

3) Provide a longer return crank on the main crankpin to put the centre of the "eccentric" on the opposite side of the wheel centre (achievable, cheap, but would look strange and even less prototypical)

4) Use piston valves !




I hope ( ) that has clarified and amplified some of the points and, if it's as clear as mud, I have succeeded !

[Keith S]

It all sounds very clever, but I'm afraid I am still confused about the relationship between how the combination lever moves the valves as opposed to how the motion from the radius rod moves the valves. I always understood the combination lever was there to contribute some lead or lag to valve events, but I think I would need to see a real engine taken apart and explained by someone on the spot to really understand. I once was aboard a stern-wheel steamship and noticed that the valve gear was actuated by nothing more than a combination lever, with no input from eccentric motion of any kind. This made me consider that the combination lever was more important than I had previously thought. (Other stern-wheel boats had simple eccentric driven valves, and some even had Walschaerts' valve gear.)

I mean: the original explanation of "cross-porting" is still correct, is it not? That piston valves are usually "inside admission" and slide valves can only be "outside admission", and that if you want to use slide valves in a replica of a piston-valve machine you need to cross-port it?

Also, rather than "cross-porting", wouldn't it work to simply reverse valve events by turning the reverse cranks by 90 degrees and leaving the rest of it alone? I'm afraid the geometry of the various events is difficult for me to visualise. I shall have to get my locomotive from the shelf and look at it. I haven't "pinned" the cranks so I can turn them and see what happens if necessary.  :oops:

*edit: By the way, I this is very interesting and thank you guys for participating in this thread. Please continue to respond: I will eventually understand all this. In my job I deal with aircraft turbine engines, and they are much easier to understand than reciprocating engines with valve events and so on. Turbines are regulated by nothing more than very complicated versions of the good old "watt's governor", and my brain is no longer configured to be able to easily visualise reciprocating events!

[TommyDodd]

I once was aboard a stern-wheel steamship and noticed that the valve gear was actuated by nothing more than a combination lever, with no input from eccentric motion of any kind. This made me consider that the combination lever was more important than I had previously thought. (Other stern-wheel boats had simple eccentric driven valves, and some even had Walschaerts' valve gear.)

It gets better, there are side-wheel boats still in service in Europe that have 2-cylinder oscillating engines! Though they do have proper Stephensonesque valve gear and slide valves and don't rely on fixed ports for their timing and events (hardly practical with a cylinder the size of a bin!

http://www.saechsische-dampfschiffahrt. ... esbar.html

http://www.saechsische-dampfschiffahrt.de/index.html

[GTB]

Keith,

If you are Windoze based, download Charles Dockstader's valve gear simulator and have a play around. As well as the common or garden loco types, it also has valve gears you've never heard of.....

http://www.billp.org/Dockstader/ValveGear.html

It seems to have been the ambition of every late 19th/early 20th century engineer to climb the path to fame and fortune by designing the 'perfect' reciprocating steam engine valve gear.............

Charles has a lot of them in there, but it's probably still a drop in the bucket.

Graeme

[Pretoria]

Keith

I found this video of a close up of real Walschaerts in action. This may help you understand just how much the combination lever and union link actually do.

http://www.youtube.com/watch?v=TSNfJqiT ... el&list=UL

[Keith S]

That's a good video. You can see, in the first bit, that the valve motion seems to start with the action of the combination lever, and is followed through by the motion of the expansion link. Later in the video, you see the driver change the cut-off and you see the valve stroke change as a result. It's difficult to see, however, the reason he chose that particular time to decrease the cutoff, after having started with it in an increased position.

However, I think I may have figured out what is going on. When the engine is "notched up", the valves move through less distance. If the combination lever wasn't there, as is the case in "simplified" walschaert's gear, this would simply admit less steam in general and also restrict the exhaust opening, and the engine would simply run poorly.

When the engine is in "full gear", the valve events seem to take place when the piston is stationary at either extremity of its travel on simplified walschaert's valve gear. This would appear to be true on a full walschaert's valve gear as well, but with the exception that there is a small amount of valve movement opposite to the piston's movement due to the combination lever, which would seem to cause valve events to happen slightly earlier than they would without the combination lever.

The eccentric seems to contribute valve motion when the piston is at the stationary "dead center" of its travel, while the combination lever seems to contribute valve motion while the piston is in the middle of its travel.

As the radius rod is moved closer to the mid-gear position, it would seem that the valves not only move LESS, but the "early" motion of the valves is a more significant portion of their movement, because a greater component of the total motion is derived from the combination lever, and less from the eccentric motion of the return crank.

I'm presently pushing my "billy" back and forth on the counter, trying to picture how the addition of a combination lever would help it to run nicely in a "notched up" position. I'm afraid I'm still a little confused. In the "maximum cut-off" position there must be a point where neither port is open, and wouldn't that restrict exhaust or admit steam into the other side to oppose the "expansion" of the first side? It's all very confusing. I shall probably never understand until I have a chance to peer into the open valve chest of a locomotive with the full Walschaert's gear.

I think we may have discovered the true limit of my intelligence.

[Pretoria]

Discovered this site which should help better:

http://www.donashton.co.uk/html/walschaerts__gear.html

[Keith S]

Thanks! There is some good stuff on there, although technically a little above my head in some spots, I can picture what is happening a little better now that he compared it to the Stevenson's valve gear. The combination lever provides the same modification to the eccentric motion of the crank, as does the opposite-direction eccentric in a Stevenson's valve gear.

It is complicated compared to a simple eccentric-driven engine, isn't it?