How to start Walschaerts Design as a beginner?

This forum is dedicated to the Live Steam Hobbyist Community.

Moderators: Harold_V, WJH, cbrew

Post Reply
Dor_Crank
Posts: 48
Joined: Tue Sep 10, 2013 1:08 pm
Location: LA, CA

How to start Walschaerts Design as a beginner?

Post by Dor_Crank » Mon Feb 24, 2014 1:57 pm

I wish to replace simulated Walschaerts valve gear on my Roundhouse 16mm/ft (1/19th scale, ~4"W x 9"L x 5"H) live steam on 0-gauge
(32mm = 1.25"). E.g., There is no combination lever (but you can put a fake one that directly attached to
the valve stem and the crosshead with no offset on the former). It cannot adjust the cut-off amount. The reason the RndHse
locos runs right out of the box is because their cutoff is 100%!

I know fully working walschaerts gear were made by others in this scale or Ga. 1.

I've searched thru Don Ashton's web and Docstaeder's websites but as about how to use their program and graphs are not
clear.

There is no step by step instruction with a typical set of parameters to go about using their software.

I've read Greenly's and Martin Evan's book and thinking of using the former's geometric (drawing) method 1st with some
simple algebra.

I feel that certain parameter values of the gear must be fixed to start, such as:

1. Cylinder's Stroke
2. Cylinder Valve port separations
3. Crank length
4. Expansion location relative to the cylinder axle

(am I missing any or one too many?)

I couldn't even begin with above websites to plug in the numbers to derive gear part dimensions and locations.

Is Don Ashton's short booklet help much?

Can anyone help? :| This is my 1st attempt.

Upon measurements, I've found the factories cylinder stroke wear marks at front and rear dead centers were
way off (like 1/16" more away from the front cover than the rear mark). I was able to figure that out to
correct this to make the stroke amount symmetrical at least.

Dorian
======================================================
thx gentlemen for the suggestions. i had to put out fires... i'm back on the hobby and this forum finally and
moved this topic to:

Walschaerts All Square Layout Design

i hope you'll still be with me there though i'm not experienced with the valve gear as many of you are.

Dorian
Last edited by Dor_Crank on Wed Jul 02, 2014 2:40 am, edited 1 time in total.

User avatar
AMatthews
Posts: 80
Joined: Mon Jul 27, 2009 8:46 am
Location: Elizabeth Downs South Australia

Re: How to start Walschaerts Design as a beginner?

Post by AMatthews » Mon Feb 24, 2014 5:00 pm

Hi Dorian,

The best thing you can do is get hold of DL Ashtons book and also look at his website

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

his mathematics are far more concise and true than either Evans or Greenly.

the other thing to get hold of is Dr Alan Wallace's Valve gear simulator.

Dockstader's is good also, but I would recommend Wallace's for a beginner.

http://avocetconsulting.com.au/modeleng/

I have used it for the last 15 years and designed 6 sets of Walschaerts and a couple of sets of baker with it.

thoroughly read Ashton's information before you put pencil to paper or click on a screen as you will have a much better understanding of what is required.

Cheers,
Andrew
USRA 2-8-2
NKP S-3
0-8-0 SAR S1

User avatar
VGC
Posts: 66
Joined: Tue Sep 22, 2009 3:17 am
Location: Europe / Earth - yet in freedom ;-)

Re: How to start Walschaerts Design as a beginner?

Post by VGC » Tue Mar 25, 2014 11:42 pm

Why don't you let the program design it for you?

http://www.constructor.valvegear.de/qui ... =us_chaski

(One needs Firefox or Google Chrome browser; Internet Explorer does not work!)
Online program for valve gear design:

ValveGear Constructor [EN]

Image

(New: Website also in English language, now much more user-friendly!)

User avatar
Big Northern
Posts: 42
Joined: Sat Mar 22, 2014 10:04 pm
Location: West-Central Arizona, Colorado River

Valve Gear Duplication?

Post by Big Northern » Thu Mar 27, 2014 9:58 pm

Does a major obstacle pop up if one scales down the various gear parts from prototype photographs? Let's say, Baker valve Gear. Are not the really important things the proportion of lengths of lever arms and their pivot points,, eccentric rotation diameter, eccentric angle setting, and vertical location of those assembled parts with respect to the valve's axis centerline, "copyable" in scale to secure satisfactory valve events?

It seems to me, perhaps incorrectly, that the relative movement of the parts scales down directly, producing similar event occurrence in the model.

Or, am I all wet? Let it all blurt out! BN
The last forum I attended, they kicked me out of!

Harold_V
Posts: 18874
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Valve Gear Duplication?

Post by Harold_V » Fri Mar 28, 2014 1:40 am

Big Northern wrote:It seems to me, perhaps incorrectly, that the relative movement of the parts scales down directly, producing similar event occurrence in the model.

I share your thoughts.
Or, am I all wet? Let it all blurt out! BN
It seems to me that if a person scales a model precisely, it should work. If it does not, I expect it's because some features can't be scaled appropriately, thus variables are introduced.

I expect that precious few folks actually build a faithful reproduction in miniature, so it may be difficult to determine. We already know that's true of the boiler.

Like you, I'd like to hear from those who have an opinion, but, more so from those who can answer from experience, and, perhaps, share what they learned.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.

JJG Koopmans
Posts: 335
Joined: Thu Oct 01, 2009 10:01 am
Location: Netherlands

Re: How to start Walschaerts Design as a beginner?

Post by JJG Koopmans » Fri Mar 28, 2014 7:24 am

Since the boiler pressure is not really scaled, as is the weight of the model locomotive, the steam consumption is not scaled. The consequence is that the valves and ports should be different and the motion gear relatively heavier.
A live steam locomotive should be designed in its own right and differences from its full scale example should occur.
Kind regards
Jos Koopmans

User avatar
Fred_V
Posts: 4363
Joined: Sun Jan 12, 2003 3:26 pm

Re: How to start Walschaerts Design as a beginner?

Post by Fred_V » Fri Mar 28, 2014 8:30 am

You can try scaling it down and then plug all the dimensions into this program and see how it works. As Jos says steam, water and coal don't scale down so some things have to be altered.
http://www.jf2.com/bcwrr/Dockstader-Valve-Gear.html
Fred V
Pensacola, Fl.

James Powell
Posts: 401
Joined: Fri Oct 28, 2011 10:42 pm

Re: How to start Walschaerts Design as a beginner?

Post by James Powell » Fri Mar 28, 2014 9:36 am

In principal, exactly scaling the dimensions should work. That's ALL the critical dimensions though...

Some of them can be more easily adjusted than others- the length of the valve spindle can be adjusted so that the cylinder position isn't as critical as the dimensions on the linkages.

If you start by scaling down the full size dimensions, and plug them into one of the valve gear programs, you are likely going to be something like right. Be aware, that even fairly subtle changes in dimensions (like errors in construction sized, .001+-) can result in strangenesses. It is all going to be a series of compromises.

James Powell

User avatar
gwrdriver
Posts: 3237
Joined: Mon Jan 06, 2003 10:31 am
Location: Nashville Tennessee

Re: How to start Walschaerts Design as a beginner?

Post by gwrdriver » Fri Mar 28, 2014 10:29 am

I was going to say exactly what James has correctly said, so now I'll just confirm it. I've posted before that in the not too distant future I have this same process to go through for a 7.5"ga loco. IIRC there are potentially around 57 variables in a Walschaerts gear and the location (length, etc) of any one of them affects many, if not all, the others. My locomotive is an amalgamation of two existing locos of proven designs, both with Walschaerts gear, but since I've changed almost all the basic dimensions (spacings, etc) of the original designs neither of those gears can be used as-is. A new gear design is required.
GWRdriver
Nashville TN

User avatar
AMatthews
Posts: 80
Joined: Mon Jul 27, 2009 8:46 am
Location: Elizabeth Downs South Australia

Re: How to start Walschaerts Design as a beginner?

Post by AMatthews » Fri Mar 28, 2014 12:34 pm

hi all,

You will find that there are many scaled down engines, at least down to 7 1/4" where the valve gear, ports etc have been scaled directly from full size and the loco will work just fine. improvements in valve gear design for models can always be made though. As you get smaller and smaller obviously physics plays it's part and more modifications need to be made.

Happy Steaming
Cheers,
Andrew
USRA 2-8-2
NKP S-3
0-8-0 SAR S1

User avatar
Curtis_F
Posts: 830
Joined: Mon Jan 06, 2003 10:14 am
Location: Reedley, Ca. 93274
Contact:

Re: How to start Walschaerts Design as a beginner?

Post by Curtis_F » Thu Apr 03, 2014 10:44 am

This forums own LSGOD has built several 1-1/2" scale engines with the valve gear 100% scaled and they work fine. Generally they're over-bored due to scale being linear and weight being cubic, so he runs them at reduced pressure (70-80 psi), but they run great!

Though in general I tend to agree that a non-scaled engine should have all the calculations performed to give it a proper valve gear motion, which isn't easy, but considering the engine will be operated for decades or centuries to come, I would think it's worth the time to "do it right" the first time.
JJG Koopmans wrote:Since the boiler pressure is not really scaled, as is the weight of the model locomotive....and the motion gear relatively heavier."
Due to the weight being cubic and strength of material being linear you can build the valve motion lighter than scale and still be good. There's no physics related need to make it heavier.

1.5" : 12" = 1/8th Scale
1.5^2 : 12^2 = 1/64th Cross Section / Area
1.5^3 : 12^3 = 1/512th Weight

--EDIT--
Here's an Example:
Full Size Piston Dia.: 9"
Full Size Piston Rod Dia.: 1.25"
Steam Pressure: 175 psi.
Piston Load: 11,133 lbs.
Cross Section of Piston Rod: 1.23 sq. in.
Load on Rod Per Square Inch: 9,072 lbs.

Same engine @ 1-1/2" (1/8th) Scale:

Scale Piston Dia.: 1.125"
Scale Piston Rod Dia.: 0.15625"
Steam Pressure: 175 psi.
Piston Load: 174 lbs.
Cross Section of Piston Rod: 0.0192 sq. in.
Load on Rod Per Square Inch: 9,072 lbs.

They both have the same load Per Square Inch, so if both are made from the same material, they will both work the same.

Now lower the Boiler pressure to common hobby standards and the PSI stress on the Piston Rod drops even further meaning you can use weaker materials or a lighter design on the model:
@125 PSI. Load on Rod Per Square Inch: 6,480 lbs.
@100 PSI. Load on Rod Per Square Inch: 5,184 lbs.

--EDIT--


Cheers,

Curtis F.
16 Years of SolidWorks experience...now I feel old.
"You can lead a horse to water, but you can't make him drink."
"Winning an argument on the internet is like winning the Special Olympics."
www.facebook.com/HillcrestShops

JJG Koopmans
Posts: 335
Joined: Thu Oct 01, 2009 10:01 am
Location: Netherlands

Re: How to start Walschaerts Design as a beginner?

Post by JJG Koopmans » Fri Apr 04, 2014 3:46 am

Curtis_F wrote:.......
Due to the weight being cubic and strength of material being linear you can build the valve motion lighter than scale and still be good. There's no physics related need to make it heavier.
Curtis F.
There is a pitfall in oversimplification! A locomotive on an 1/8 scale may have its volume reduced to 1/512 but not its weight! For instance a 1:1 boiler plate thickness of say 3/4 inch is not scaled to 3/32 but more, adding weight.
Regarding the rods in a locomotive, the above reasoning disregards the fact that rods are pulled and pushed 50% of their working time. Since you do not want your eccentric rod to deform it should be stiffer to avoid buckling. (for an explanation of buckling see the Wikipedia page). Since the buckling force formula has the rods Moment of inertia divided by the square of the rods lengths, this goes wrong if the diameter and length are scaled to 1/8 of the original. The inertia moment of a round rod is pi*d^4/64 and since 1/8^4 is only 0.000244 it should be clear that a scaled rod cannot withstand a scaled pushing force! So there is very much physical evidence for a heavier model valve motion.
Kind regards
Jos Koopmans

Post Reply