Many thanks to all of you for your tips! See my comments below.
If you can manage to get your hands on Paul Mallery's "Trackwork Handbook for Model Railroads", you will find all the prototype info you could ever want. It's old...my copy is from 1974. It is the ultimate track layers bible.
A second hand copy is $5 only, but postage to Australia costs $47.60, which is bit too pricey just for finding two lucky numbers, especially if there is no guarantee the numbers are there.
Use Google Earth ruler to measure a crossover interlocking ... apply that same measurement to your Trainz turnouts ... do the same ithe curve radius
I'm afraid this method isn't too terribly accurate... It can be used to confirm where different turnouts types may be used, but to build turnouts in Surveyor you need +/- 0.05 m and +/- 0.05 degree accuracy. Say, 5 meters is not much in Google Earth, but it can make a turnout with radius of 300 meters calculated as one with 209 meters, almost 100 meters shorter! 100 meters is a lot, even in Trainz.
I'd suggest grabbing an animated (and therefore fixed) turnout from the DLS or out of the game. Lay regular track over it so the rails line up. Then delete the fixed turn out and use the tools in surveyor to measure the radius and so forth.
And, if you can find Blueprints (they are on the net, I've seen them, though I can't recall where) you can use a ruler on your computer monitor to check ratios since they would be scale drawings.
The problem with this method is that I would need to built a turnout out of a template and then use "show curve radius" tool to find out the radius. The tool is terribly inaccurate. It produces several different readings, varying by as much as 50 meters or more, which can be 25% or more of radius length. There is no tool to measure angle, other than laying fixed track in the diverge direction, as in the template and reading angle by turning the fixed track. Some authors, like Natvander, were thoughtful and gave hints what is the turnout angle, for instance "Turnout 1:9", <kuid2:61119:28995:1>, but there is no mentioning of turnout radius. There is set of US turnouts by martinvk, but no turnout properties in the description. As to the blueprints - some turnouts have radius of 1200 meters (about 4000 ft) and I don't know if they will ever build screen big enough to measure such distance in real scale.
Few days ago I came across this document - "CN Engineering Specification for Industrial Tracks". There, on page 33, there is a plan of right hand turnout, for types #8, #10, #12, #15, #20. I think the angle I am looking for is the value of F angle (FROG). This is approximately 7.2, 5.7, 4.8, 3.8, 2.9 degree respectively for the turnout types. But I can't find the radius of the arc at which the straight and diverged tracks are connected. I guess it can be calculated from other properties, but how?
It's probably not as technical as all that, no offense intended. A reasonable radius would probably work just by eyeballing it.
Cheers
AJ
If I can't build a turnout exactly as they do it in Canadian railways and within the limits of Surveyor, it is not a big drama. But I hate this word "cannot". It is often said that the beauty is in the detail and I believe there is some true in this saying. You can read simplified version of "The Great Gatsby" and find out what the novel is about, but you will miss all the style, color and what great literature is about. As to "reasonable" radius - there are strict guidelines on most railways. You can't use the same turnout for a factory siding, where permitted speed is 20 mph or less and for a main line where a train can diverge at speed of 60 mph. So, how to tell the two apart?