Atomic Systems

Hi all,

Would someone be so kind as to put down some notes about what kind of process they use for building routes to a realistic track plan?

I want to build a route and if it was all flat, I would be completely fine following the layout, using photographs and getting the track just right - however the gradients side of things is really giving me some headache just trying to get it through my thick skull.

It's not so much how to do it, I guess I can figure that out from the tool - it's about how you get the gradient actually right? To what level of accuracy and approach do route builders in the community generally follow?

Do folks just make the track generally follow the DEM terrain?

I used to have a gradient profile for the North Norfolk Railway but that seems to have disappeared now for some reason. That said, the route is dem'd up no problem and I have markers in place for the route of the line. I'm itching to get track down but obviously the third dimension has to play a role in here as it's hilly up in norfolk

Thanks
Matt.

My approach has been the following:

1. Marker files with height data. I use both Series for layout and Named for height data. I also changed the colour for the series marker to yellow to make it stand out more.

2. My height markers are placed at every 1metre height increment. When laying track I lay a section to the next height marker, noting the length of the track section at the bottom of the screen. So for example if the section of track layed to the next marker is 80 metres and I've gained 1metre in height then that's a 1 in 80 gradient (Isn't it?). I then delete the track (or cancel the laying) and lay the section again but with the gradient set to 1:80.

This has given me very good results. On the downside I have many markers.

There are probably better methods but this works for me. Hope it helps.

John

I've never followed DEM'd terrain when laying track. Always gone with the gradient profile. In the marker files I've used in the past for MSTS I always have markers for the mileposts included along the route of the track. That way I can see when I'm getting close to where the next gradient comes in.

Going from level grade to up or down I use about 50m - 100m of track at varied degrees until I want the full gradient. This ensures a smooth run from one grade to another.

MSTS example aiming for a 1 in 110 grade. level, 30m 0.150 degrees, 30m 0.300 degrees, 30m 0.450 degrees, continue track laying at 0.521 degrees until you change to another grade.

Hope you get the idea.

Kerr

Which you can do using the tool in KRS by setting the one length, then splitting it with the angle tool at the positions you want (Though always put a height adjuster at the start of the section as well otherwise you end up lifting the track to the one before it).

KRS is making more and more sense to me as I go along... while work is making less and less sense.
Go off for a year to a different team in the same office, return and everyone that was there before has either got another higher level job related to their PHD, left, been deported or gone off on matty. A contractor has significantly rewritten an important section of the system and for some reason your now the expert.........
hmpf.

Hi Kerr,

Appreciate you were just giving an example of increasing gradient for a smooth gradient transition - however for some people (especially in the UK) it might help to use the 1 in x gradient option in KRS - ie 1/110 by using 1/440, 1/330, 1/220, 1/110 or something similar ( 1/382, 1/190, 1/127 I think in your example) - this just saves converting from 1/x on UK gradient diagrams to degrees.

Hope this helps.

Regards
Paul

Cobos wrote:I've never followed DEM'd terrain when laying track. Always gone with the gradient profile. In the marker files I've used in the past for MSTS I always have markers for the mileposts included along the route of the track. That way I can see when I'm getting close to where the next gradient comes in.

Going from level grade to up or down I use about 50m - 100m of track at varied degrees until I want the full gradient. This ensures a smooth run from one grade to another.

MSTS example aiming for a 1 in 110 grade. level, 30m 0.150 degrees, 30m 0.300 degrees, 30m 0.450 degrees, continue track laying at 0.521 degrees until you change to another grade.

Hope you get the idea.

Kerr

Same as Kerr, I work from the gradient profile, although I use more track sections and gradient changes at gradient transitions. If you work from the terrain how do you know where cuttings and embankments are supposed to go?

As my MSTS Wessex route is very hilly in parts, getting the gradients right was important to to ensure that cuttings and embankments are in the right place. Like most route builders I wasn't blessed with an accurate gradient plan, so I have taken approximate measurements from a small scale gradient profile and relied on cross checking to compensate for errors, as below.

Measuring gradient lengths off the gradient profile, I calculate track distance, milepost position, and elevation at each gradient change along the length of the route. As measurements off the gradient profile are pretty crude, I check to see if this gives me gradient changes that match the observed position of gradient posts on the line. I also check the predicted track elevations to those indicated on the OS map at selected points. If there are material differences, I tweak gradient lengths to match.

I then log each mainline track section as I lay it into a spreadsheet (very useful when I have to relay a section). Based on the track section geometry this calculates route length and elevation as I go, which I can then compare to my outline profile as above.

On most of my route I lay track onto OS map terrtex which shows me trackside benchmark (spot) heights and the location of embankments, cuttings, LC's and bridges; these all give a further check on the accuracy of the track height.

All this usually gives me track height within a meter or two of where it should be everywhere along the route, and cuttings and embankments in the right place.

Now you see why my route isn't finished...

longbow wrote: Now you see why my route isn't finished...

Might not be finished - but your MSTS route still looks fantastic !!!!

Cheers
Paul

Working from Gradient Profiles in the Oakwood Press book for my Nailsworth to Stonehouse route and started laying at Nailsworth exactly on the DEM terrain at the station. The book shows all the gradient changes and gives you a rough idea how long each gradient stretch is relatively. It seemed to keep me pretty close to the DEM Terrain for the first half of the route to Dudbridge but then it's floating probably about 12 feet too high after that Plan to revise that when I get to it, maybe tweak how long a couple of the gradient stetches are to bring it in line. I only laid a single track as a marker guide really initially. I'm then making everything fit and adding sidings when I get to scenerising the area.

It does help having the real route on my doorstep as I can quickly jump on the bike and go and check it out and how it fits into the surrounding landscape in a particular location. Most of the route is now a Cycle Path.

I will go with the gradient profile if available but I am not a slave to it. Certainly in MSTS using DEM I've followed the profile and it's not come out right so I've tinkered it. Would be a sad person who downloaded a route and actually went through checking every gradient is accurate to start/location and angle!

More recent efforts such as my European routes have not had the luxury of a gradient profile. Generally I will survey the course of the line about 500m ahead though perhaps taking into account any sudden dips or bumps that might require a climb starting in the plains to surmount. Look at the map for clues - is the line on an embankment, does it cross any roads on a bridge (in which case needs to be @15 feet above the terrain) or a level crossing?

That's the approach I used for my Swedish route in MSTS and the same route I'm now doing in TRS. TRS has the advantage of a terrain spot height tool and "get" height of track vertex, not sure if EARS offers this.

I use a curve and gradient diagram and lay the gradients accurately which both MSTS and RS have no problem with. I also transition my gradients so that the change is in most cases imperceptible (except for the drop or increase in AMPS). I find hill crests/troughs with steep gradients either side can be difficult to smoothly transition over with out that ‘stepping’ effect. That’s what external view is for, diversion.

I use google earth, GPS and other aerial surveys to create marker files for lat/long position of track work and use SRTM for terrain elevation. If there is a discrepancy I normally modify terrain and markers to fit the curve and gradient profile. My reasoning for this is that matching simulator train performance to real world performance requires the track work to be accurate. If a simulated gradient is steeper/shallower or starts earlier/later than the real world gradient you will not be able to match the real world performance with what you are experiencing in the simulator.

I may seem a little anal about this but I have a driver testing my MSTS route and he can detect the variations in the route where I’ve had to squeeze or stretch track work to fit terrain or marker files.

Route and content building requires compromises to varying degrees. The more accurately you can model a route the closer the simulation can come to emulating real world performance, within the limits of the physics coding in the simulator.

If you primarily use RS (or any simulator) for taking screen shots of models or watching trains drive by then none of the above really matters.

Having put down literally hundreds of miles of track - I agree with others that it's pointless and not necessary to follow gradients slavishly.

I do take Gradient profiles and hand draw them, with mile markers (roughly) where they are in real-life (per the profile) I then use the gradient converter, and note the MSTS gradient where necessary. When I'm on each section, as per Kerr and others, I 'smooth' the gradient in and out.

My personal preference is to work on a terrraformed route, and make my start point one where I know the level is spot-on - even if it means some electronic excavation...! I will then use certain Key points on each line as a guide to the overall accuracy - if, for instance, Falahill summit was not quite right, that is more noticeable than some obscure cutting or embankment.....

I also tend to make all pointwork (IE stations etc) on the level, as it's easier, and faster to do than fiddling with small sections of track in yards to get the correct angles...

I also prefer not to cut beds/cuttings beforehand, rather, find a 'known' level, start there and if you follow the gradient profiles fairly closely, then it all fits into place........ that is the critical part IMO - if you are off by several meters at the start, (even if the sim says otherwise) - the whole route will look wrong.

i think the option of a top down view with the abillity to turn of ground textures would be a great help when track laying

cheers

Gav

basically, the only real way to create a perfect route layout is to convert the distances covered on the gradient profiles into meters, thenfind a place where you already know the altitude (I think Bristol TM is 50 feet above sea level(brunel designed it that way)) then start laying track, and BACK UP FREQUENTLY.

thanks

chaddockdk

Hi

I have raised the terrain and loaded in the route markers for The Spa Valley Railway with data taken from Google Earth and you can see from this shot that without any additional work, the marked route is remarkably close to the undulations of the terrain. The marked route even disappears into the terrain where cuttings are known to exist so on the face of it, gradient profiles are not strictly necessary if you take enough marker points from Google Earth.



The chimney is just so that I can see the start of the route from a distance.

As this is a relatively short route I will be trying to match the gradient profiles with the track but Google Earth has done much of the work for me.

Geoff

Using terrain as a guide to gradients might give good results in Kent, but it doesn't in hillier parts of the country!

1) Obviously, track in hilly areas tends not to follow terrain contours very closely.

2) Less obviously, the DEM used in RS isn't very accurate. It's quite low resolution, which means the terrain is heavily smoothed. This isn't a big issue in flat country, but in hilly areas ridges are likely to be too low and steep valley floors too high. My Ilfracombe route has both, and there are areas where the DEM terrain is easily 10m out from where it should be. If I'd used the terrain as a guide the 1 in 36 out of Ilfracombe would have been nearer a 1in 20!

The RS SeriesMarker simply produces a straight line between each point in the SeriesMarker file. If it runs underground it may indicate a cutting, but equally it may just indicate that the terrain contours are wrong, or that your marker points are too far apart. If you want a reliable guide, use NameMarkers to indicate the start and end of cuttings and embankments.

I'd also endorse the comments above about establishing benchmark track heights at points on the route, and adjusting gradients to suit. I do this by putting my start height and gradient profile into a spreadsheet, calculating elevations at key points along the route, and comparing these to my spot height benchmarks and to the DEM terrain at these points. If things don't match I can then tweak gradients and/or distances in the spreadsheet to suit.