Trouble with route design/ready to make scenes but don't know the altitude

[BraselC5048]

My current route design is somewhat settled in the general part, but there's still some outstanding issues. The basics is valley with coal, ore, lumber, a few towns, and various other traffic, goes over spectacular mountain pass, then to mills, then downhill to a seaport.
The problems mostly come from the desire to have a 12,000 ft summit, the preferred scenery being Colorado high country, the line originating in a seaport, and the desire to have most of the more spectacular mountain features being on the side facing loaded trains (which is heading out of the valley - which is by default less altitude change then the other side.)
I mean, there's no reason you can't use selective compression to fudge the amount climbed (particularly compressing the lower-grade section between the pass and seaport), but there's the basic problem that the scenery doesn't fit with a port.

The other problem is that I don't know how much altitude the line will gain/lose on each side of the pass, and won't until I've made it. (Tell me there's an easier way to make custom mountains these days! Or at least a smooth tool for the terrain mesh, like everything else has, so you don't have to fine-tune every point by hand.) (I also don't know how many features will actually wind up being included, and again most of them I'd prefer facing loaded trains.)

Which means that while I'm more then ready to start work on a couple of the major towns etc., I don't know at what altitude to build them at. I take it if I don't want to wind up redoing it later that they'll have to wait until the mountain section is roughed in? Does T:ANE or trs2019 include a way around that problem? (Offsetting the height when merging routes, for example.)

Advice?

 

[JCitron]

Well to go from sea level to 12,000 feet is a substantial climb and that can't be done in a short distance. Railroads and trains in particular like smooth climbs. Trains like flat terrain which means you will have to spread out the climb to this lofty height over some distance. Since you need to keep the grade small, meaning no more than 2% to be realistic, you'll need to do some maths.

The climb is a slope and you can figure out the distance using the simple Rise / Run ratio. So many feet up divided by how far to get there. This will produce your percentage. Since we deal with meters and each baseboard is 720 x 720 meters, and 3 baseboards is a smidge over 1 mile, you'll need to base your number of baseboards on this. Keep in mind that your 12,000 feet is 3,658 meters.

Now, railroads can't do this realistically with a single straight run which means the line will have to be spiraled up and around. This makes the grades a bit steeper, sorry I can't remember why, when Trainz calculates the grades, and you need to keep your curves fairly wide if you are running modern trains, i.e. auto racks, 53ft container double-stacks, etc. This means you'll most likely need to use tunnels and bridges to traverse the mountains as your route climbs. The grade too will determine the length of your train and the amount, meaning weight, that the freight cars can carry. If you have coal and lumber, these trains will be very short, especially for the coal trains.

I know this sounds esoteric and a bit out there and probably doesn't answer your post directly but these are things that I've read about when the big railroads were built through the Appalachian and Rocky Mountain ranges.

There are a couple of ways to generate this terrain. You can use a displacement map. These are a bit awkward to use but doable. The tool for these is located in the Topology tools. You select the displacement map you want to use, select whole baseboards, I've selected 3 or 4 at a time, adjust the height by slider, and then fill. If you don't like the outcome, undo and adjust the slider again until you get what you want. The results are okay but very fiddly, and there's a lot of smoothing to be done afterwards.

The alternative is to invest in TransDEM. TransDEM allows you to import real world DEM terrain and topo-maps and output a full route with nothing more than topographic maps as textures ready for landscaping. Even though this is real world data you can do whatever you want with it. I've created routes on terrain that never had a railroad before. I sighted and plotted the route in Surveyor by getting down on the ground and looking around, smoothed and graded, and then built the track. When I came to towns, I carefully plotted the route through towns and carefully preserved roads as much as possible. This made for a bit of figuring out how I wanted the route to go through and meant putting in bridges over, under, and crossings as needed. This was both a challenge and fun at the same time.

I'm sure others may have some better thoughts on this than I have.

 

[pware]

Another thing to consider is that if you are using true heights (i.e. the Trainz terrain ground height is the same as the actual topological map height so 1,000m in altitude on the map will also be 1,000m in altitude in Trainz) then the maximum terrain height you can have in Trainz is 3,000m - some 658m short of your goal.

 

[GaryReggae]

I would be inclined to use Transdem and find a part of the world that has terrain similar to what you want to simulate. As JCitron suggested, this could be somewhere that doesn't have a railroad in the real world, which would perhaps make building your own fictional railroad an exciting idea although very time consuming. I am tempted to do something similar myself rather than just manually building my hills and mountains etc.

Kind regards,

Gary

 

[BraselC5048]

Well to go from sea level to 12,000 feet is a substantial climb and that can't be done in a short distance. Railroads and trains in particular like smooth climbs. Trains like flat terrain which means you will have to spread out the climb to this lofty height over some distance. Since you need to keep the grade small, meaning no more than 2% to be realistic, you'll need to do some maths.

The climb is a slope and you can figure out the distance using the simple Rise / Run ratio. So many feet up divided by how far to get there. This will produce your percentage. Since we deal with meters and each baseboard is 720 x 720 meters, and 3 baseboards is a smidge over 1 mile, you'll need to base your number of baseboards on this. Keep in mind that your 12,000 feet is 3,658 meters.

Now, railroads can't do this realistically with a single straight run which means the line will have to be spiraled up and around. This makes the grades a bit steeper, sorry I can't remember why, when Trainz calculates the grades, and you need to keep your curves fairly wide if you are running modern trains, i.e. auto racks, 53ft container double-stacks, etc. This means you'll most likely need to use tunnels and bridges to traverse the mountains as your route climbs. The grade too will determine the length of your train and the amount, meaning weight, that the freight cars can carry. If you have coal and lumber, these trains will be very short, especially for the coal trains.

I know this sounds esoteric and a bit out there and probably doesn't answer your post directly but these are things that I've read about when the big railroads were built through the Appalachian and Rocky Mountain ranges.

There are a couple of ways to generate this terrain. You can use a displacement map. These are a bit awkward to use but doable. The tool for these is located in the Topology tools. You select the displacement map you want to use, select whole baseboards, I've selected 3 or 4 at a time, adjust the height by slider, and then fill. If you don't like the outcome, undo and adjust the slider again until you get what you want. The results are okay but very fiddly, and there's a lot of smoothing to be done afterwards.

The alternative is to invest in TransDEM. TransDEM allows you to import real world DEM terrain and topo-maps and output a full route with nothing more than topographic maps as textures ready for landscaping. Even though this is real world data you can do whatever you want with it. I've created routes on terrain that never had a railroad before. I sighted and plotted the route in Surveyor by getting down on the ground and looking around, smoothed and graded, and then built the track. When I came to towns, I carefully plotted the route through towns and carefully preserved roads as much as possible. This made for a bit of figuring out how I wanted the route to go through and meant putting in bridges over, under, and crossings as needed. This was both a challenge and fun at the same time.

I'm sure others may have some better thoughts on this than I have.

Well familiar with the math. Year is 1912, he 2-8-2's have a tractive effort of 75,000 lb (tested - I wrote the e-spec, after all), the 2-6-6-0 has 116,000 lb. Grade is 4.5% (subject to change upward slightly), curvature is maximum 16 degrees, rolling resistances is 8 lb per train ton base, 0.8 lb per train ton per degree of curvature and 20 lb per train ton per percent of grade, meaning a maximum (with a 2-6-6-0 up front and a 2-8-2 pushing) train weight of 1723 tons, probably best limited to 1400 tons or so in practice. The wooden ore hoppers weigh 35 tons loaded, the 2-6-6-0 250 tons with tender and the 2-8-2's 183 tons with tender, the truss rod caboose just 16 tons, for 27 ore hoppers. 32 hoppers if you bump it up to 1600 tons. Other cars as lighter. Adding a second 2-8-2 behind the lead engine would be it up to something like 40 hoppers easy, 44 pushing it.
Of course lines in the mountains loop back and forth to gain altitude. The mountain pass doesn't go all the way to sea level, much of the grade, perhaps 6,000 ft or more is of something like 2.5? percent downhill over a longer route, which would be 73 kilometers? Likely compressed to about 30-40, with the seaport not actually being at sea level.

 

[BraselC5048]

Another thing to consider is that if you are using true heights (i.e. the Trainz terrain ground height is the same as the actual topological map height so 1,000m in altitude on the map will also be 1,000m in altitude in Trainz) then the maximum terrain height you can have in Trainz is 3,000m - some 658m short of your goal.

Well aware. 15,000 ft peaks would be 4570 meters. Probably set the valley at -1500? meters, peaks at 2000 meters, summit at 1000 meters, sea level would be -2650 meters. With selective compression non-valley side of the pass could be at -500 meters, and with 30 kilometers of line sea level as high as --1250 meters. Far end of the valley would be -1100 to -900 meters.

 

[JCitron]

I would still use a route generated by TransDEM. That would give you the realistic terrain you are looking for without needing to smooth everything all over the place. You can cut, trim, and slice a TransDEM route just as you do any route you've manually created without needing to manually create hills that never come out the way you want them.

I'm glad you have the traction end of things under control. Writing e-specs is beyond my paygrade. 30 to 40 km is still a long route. That's about 50 to 60 miles and building that will take years to get it right and I'm sure you want to get the route done the right way.

 

[BraselC5048]

Might also start the line at the fictional port of 'Denver' (different city with the same name).

 

[JCitron]

Well aware. 15,000 ft peaks would be 4570 meters. Probably set the valley at -1500? meters, peaks at 2000 meters, summit at 1000 meters, sea level would be -2650 meters. With selective compression non-valley side of the pass could be at -500 meters, and with 30 kilometers of line sea level as high as --1250 meters. Far end of the valley would be -1100 to -900 meters.

That makes sense to me. You can do that in TransDEM before you export the route. This can no longer be done after the fact due to file changes in the current version of Trainz.

 

[JCitron]

You can get TransDEM here. It's not free but relatively inexpensive. It's $26 USD cost, roughly these days, is worth more than you paid for it. It's got a bit of a learning curve but once you go through them, it's easy to work out a workflow. Like most program, there's a lot more to it than you'll need at any given time and the support is excellent. There's a dedicated forum on the software website and there's support available here. Dr. Ziegler is a member here in the forums as well and will provide support plus make announcements when available.

TransDEM

 

[BraselC5048]

I would still use a route generated by TransDEM. That would give you the realistic terrain you are looking for without needing to smooth everything all over the place. You can cut, trim, and slice a TransDEM route just as you do any route you've manually created without needing to manually create hills that never come out the way you want them.

I'm glad you have the traction end of things under control. Writing e-specs is beyond my paygrade. 30 to 40 km is still a long route. That's about 50 to 60 miles and building that will take years to get it right and I'm sure you want to get the route done the right way.

You can get TransDEM here. It's not free but relatively inexpensive. It's $26 USD cost, roughly these days, is worth more than you paid for it. It's got a bit of a learning curve but once you go through them, it's easy to work out a workflow. Like most program, there's a lot more to it than you'll need at any given time and the support is excellent. There's a dedicated forum on the software website and there's support available here. Dr. Ziegler is a member here in the forums as well and will provide support plus make announcements when available.

TransDEM

Can you scale DEM terrain, both it terms of east/west/north/south and vertically? Or cut and paste at non-baseboard boundaries?
That would be pretty important.

 

[BraselC5048]

Or generate displacement maps?

 

[JCitron]

Can you scale DEM terrain, both it terms of east/west/north/south and vertically? Or cut and paste at non-baseboard boundaries?
That would be pretty important.

I'm not sure about horizontally but you can vertically because the DEM data is divided into UTM coordinates. You can trim the DEM within TransDEM using the built-in tools but you have to keep in mind that the route-export is still constrained by the baseboard dimensions. The height changes are done on first export. You export the route then use the route height adjustment feature to make your height adjustments. Once done, you import your route into Trainz. TRS19 and up has made file changes which means once you've edited and saved, you can't adjust the height. N3V changed the data format to accommodate their new sliced route data model used to increase performance and they have not released any technical information on this to any third-party which is unfortunate.

Once the route is exported from TransDEM, regular Trainz editing applies just the same as any route.

 

[JCitron]

Or generate displacement maps?

I don't think that can be done through TransDEM. I wonder if Microdem might work for that... I haven't used that in a decade or more.

 

[BraselC5048]

Wait - by unable to change elevation once you’ve edited and saved do you mean specifically DEM, or any height adjustment? Since there’s going to be at least some manual mountains, and those take longer then a work session to get the height right.

 

[wreeder]

As someone that has attempted to build the RGS to scale 48 times, I can share some of the issues you are going to run into when working with big mountains. First you are talking about a huge layout even if you scale down the height differences. The bigger the route the slower it is to move around in Surveyor. Dropping the draw distance down helps but then you can't see the top of some mountains. There is a DEM map of the RGS route on the DLS so I suggest you download it to see how your PC handles moving around and doing normal stuff like laying track and texturing the ground. The lowest point on that DEM is Delores which is 66oo ft and it takes around 60 miles to climb to Lizard Head Pass at 10,200 ft. Also the Trainz grid system tends to round off mountains into giant hills rather than steep rocky mountains so I have never been able to get the right look to the terrain. Frankly, I would be amazed if you can do what you want and still have it fit into a .CDP file. I haven't tried with HD terrain mostly due to I know the route file would be gigantic. Good luck.

 

[HiBaller]

The Darjeeling Himalayan Railway (on the DLS) took 57 miles of VERY steep grades, switchbacks, and loops to go from 450ft to 7408ft. To do this and not have those grades would take probably an additional 75 to 100 miles. This is a LOT of baseboards.

Bill

 

[JCitron]

Wait - by unable to change elevation once you’ve edited and saved do you mean specifically DEM, or any height adjustment? Since there’s going to be at least some manual mountains, and those take longer then a work session to get the height right.

Height adjustment. Sorry for being clear as mud on that. There is always some manual editing required no matter what but what the DEM will do is get you at least the majority of the way.

Definitely, think hard about this project because this is a big one and it may be more than you can chew. Heed Bill's and wreeder's advice on this. They have both worked on mountain routes. Being a northeasterner and a coastal one at that, I tend to model routes in the 37- to 80-meter height range.

 

[BraselC5048]

The thing is, about 800 or so meters of the valley side of the pass is already accounted for in just 3-4 features, and another may account for as much as 600 more, another feature 140 (and there's more features I'm looking for. Just remembered one that's a short loop up a side valley for 155 meters. Many of them something like 6 x 4 baseboards or much less. Just about any feature is at least a hundred meters, and 4 baseboards of track (which is closer to 3 baseboards straight line) between them is 140 meters on it's own. That's a single loop for 2 baseboards up a side valley. Looking at train length in-game, the grade is certainly going up to 5%, which helps. 5% gets you uphill pretty fast, and I know that from experience. (It's also constant 5%, not just the steepest sections.) Still will probably reduce the elevation change by a few hundred meters. Straight line distance is likely to be about 30 baseboards. Might actually start at the summit and work backwards. That it's probably going to be a 10m grid for most of it reduces workload.

Summit-outbound side elevation chance of 1500 meters equates to 42 baseboards of track, which is going to have a whole lot of curves, and may be more like 20-30 actual baseboards straight line distance. Probably will reduce it to about 1100 meters or so.

750 meters to the port at 2.5% is something like 42 baseboards as well, mostly in a line, so I'm certainly reducing the elevation change - and I don't particularly care if the port never gets connected to the first 8-12 or so baseboards after the pass, and instead is its own separate route.

The valley is only going to be about 20 baseboards long. At 2% grade that's about 250 meters of elevation. Probably will plan for 300 and if the pass comes up short reduce it.