Walschaerts Help ***UPDATED 3-28-2010***

[Fuelrush]

I'm making a Walschaerts Valve Gear for a narrow gauge 1:20.3 model. I have took a lot of time to try to get the working proportions and keep it as close to prototype as possible. For example using a drop link, and locating the reversing arm in the correct location. I've had some success but I'm still a long ways off where I need to be. I've read Model Locomotive Valve gears by Martin Evans which has been the biggest help, but also have studied prototype books by William Wallace Wood and also Yoder and Wharen.

Here are some details about the valve gear. All dimensions are in inches.
Driver Dia: 2.364
Cylinder Bore: .788
Cylinder Stroke: 1.182

Using Evans Calculations and Suggested Sizes I came up with:
Piston Valve Dia: .450
Steam Port: .093
Lap: .069
Lead: None
Cutoff: 81.85% or 82%
Valve Trave in Full Gear: .324

I believe I have everything correct or close to what makes the valve and piston move when the radius rod is centered in the expansion link. I think this is correct because I followed Evans calculations and it appears I'm getting correct valve movement of the valve and equal swinging motion of the combination lever. I did however have to come up with the drop link on my own because Evans book didn't elaborate. I just knew I could lengthen the combination lever as long as it was in proportion. Also The Radius Rod and the expansion link should be correct. This was calculated using Evans calculation assuming full gear travel was .324.

My biggest confusion was the backset of the expansion link. The book barely mentioned this and didn't elaborate on how it is determined exactly. I got it close but I don't think its correct. The book made it seem that the Eccentric Crank circle was determined by the backset, and the backset was determined by the the Eccentric Crank circle, leading me in a circle.

I am using VGC to test my valve gear data. I first scaled down the default valve gear to my scale (only to temporarily fill in the gaps as I entered and calculated my data) So in my file the reversing arm data isn't exactly what I've come up with just that I haven't been able to get far enough with the rest of the gear to calculated how that will be laid out. As I said before all the eccentric parts and backset on the expansion link could be way off.

I don't understand fully on how the the steam diagrams and graphs are suppose to look when you have a good design. So I'm lost in trying to tweak the gear to run as its should.

I feel like I'm asking to much for help, but I've exhausted myself. I started this project in November. I realized real quick I wasn't getting anywhere without the help of a book on the subject. When I got Martin Evans book it helped tremendously, but I'm still struggling. Its keeping me awake at night!!!

Even if you can't solve my problem, any help or thoughts on how to get going again would be a great help. I have attached my VGC file of my progress so far. I hope it isn't an atrocity. Again Thanks!

[baggo]

Have a look at Don Ashton's website.


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


He's quite an authority on Walschaert's gear (and Stephenson) and explains the details of both as well as having links to spreadsheets for designing them. The Walschaert's one will design the complete gear for you from just the few basic dimensions.

You can also download a couple of simulators to test the finished design and 'tweak' some of the dimensions to give virtually perfect valve events.

The most important thing is to get equal cut-offs at each end of the cylinder so that each stroke of the cylinder is doing the same work. Ideally, the cut-offs should differ by less than 3%. Note that the actual valve openings may not be equal when the cut-offs are equal.

John

[Fred_V]

also look at this site and download the program you need. this will simulate your engine and show up errors.
http://www.tcsn.net/charlied/

i couldn't open the file you provide.

[Krank Turner]

Hi and Greetings;

It is not my intention to pirate this thread, but it seemed to be a good spot for this question.

Would the piston rod displacement have any effect on power balance on a cylinder, or is there an offset of the stroke or valve position to counter this? Or perhaps would this be too insignificant to address?

Very curious

[baggo]

Don suggests that the effect of the piston rod can be offset by increasing the cut-off of the rear half of the cylinder by 1 to 1.5% in relation to the front i.e if the front cutoff is 50%, the rear would be 51 to 51.5%. In smaller gauges the effect is probably insignificant but I don't know if this was done in full size?

John

[alanstepney]

John,
It was on GWR locos, dont know about others though.

[Fuelrush]

UPDATE

Baggo, the link to Dan Ashtons has helped so much thank you!

I believe I have 95% of the valve gear correct now.

My current problem is in the valve movement itself.

With the spread sheet from Don Ashtons site shows correct calculations for what I came up with.

Cutoff: 81.85% or 82%
Valve Travel in Full Gear: .324

The VGC simulator by CharlieD shows that my Lap is correct at .069" and lead is 0. Same as shown in Don Ashtons spread sheet. VGC confirms Lap and Lead with Ashtons gear dimensions. So I would think all is good.

The problem is under "DrawGraphs" then "SteamDiagram" it shows cuttoff at 74% / 78% which is not equal nor is it correct. It should be closer to 82%

Also with the driver angle at 180 Degrees it shows slight steam admission to the piston. (with steam display on, where it shows live steam as red into the cylinder) With the driver angle at 0 it doesn't show admission to the cylinder. Not enough to actually see anyway. I suppose this problem is caused by the wrong and unequal cuttoff.

Anyone able to help again? Again thanks so much!

Below is the file for VGC. I can't attach a .dwi file in this forum so I renamed it to .doc. Download file and rename it to .dwi to open it in CharlieD's VGC.

[baggo]

You're not getting the maximum cut-off because the reversing arm is not moving enough in the simulator and your not getting enough valve travel in full gear. Try increasing the 'Reverse arm to centre normal' length in the Frame Dimensions to say 0.37".

Increasing the length of the eccentric rod to 2.665" will improve the balance of the cut-offs for forward gear. Try to get them balanced at about 50% cut-off which is where you will probably be doing most of your running. Unless you intend to be running backwards a lot, don't worry too much about the figures for reverse. It's difficult to get good figures for both forward and reverse when using link suspension for the radius rod as this introduces a lot of die slip which affects the valve events, especially if the suspension link is very short. Try and keep the link as long as possible and it can help to try moving the suspension point on the radius rod a bit to see if that makes a difference.

John

[Fuelrush]

Update

I think its finished unless I overlooked something. I had to tweak some things. Like you suggested Baggo. Also I tweaked some of my own. So as it stand now VGC shows correct Lap and Lead. A cuttoff of 81.6 to 81.8 at full forward gear. (Goal was 81.85.) And a cuttoff at .1/.1 at center. I think thats it?

Once again thanks for the help!

[VGC]

Hello Fuelrush,

could you tell me what the shorthand "VGC" means in full words? When I was looking for a proper name for my program, I could not find anything like that!

Stephan

By the way: Do you have Firefox?

[Fuelrush]

On Charles Dockstader's website he calls his Calculator/Computer "Valve Gear on the Computer". Hence, Charlesd VGC.

[Metal]

Hi all,

I’m also wrestling with the Walschaerts valve gear. The plan: building a narrow gauge 1/20 scale live steam engine, using Roundhouse like cylinders. However Roundhouse uses "semi-Walschaerts", while I want full Walschaerts.

Some details (mm):

Driver Dia: 35
Cylinder Bore: 14
Cylinder Stroke: 16
Steam Port: 1.5
Exhaust port: 3
Lap: 1
Lead: None
Intended Cutoff: 84%

After reading some books and visiting a lot of websites and boards I started designing the gear, using books, the spreadsheet from Ashton and several simulators. However, my designs often have the same flaws:
Although the cutoff in full forward and reverse gear are almost equal, the angle of the swing and combination lever is fully symmetrical and the displacement of the valve is symmetrical, is I encounter the following situations:

The cutoffs of the gear are very equal (fore/after piston) in for example forward, but they differ in reverse (wrong angle of eccentric crank?)

The cutoffs of the gear are very equal but when decreasing the amount of steam entering the cylinder (so, decreasing the cutoff) there is a big difference (up to 10 %) around 50%.

So, can anybody explain to me how this is possible, and more important, how to stop this from happening? I’ve added one of my “less successful” designs as example (to open, change .doc in .dwo)