UP Gila Subdivision 2023

[eric]

(Previous thread and progress can be found in the MSTS Route Design forum Gila Sub thread)


What's fun about a digital route is that you can not work on it for 3-5 years and then rediscover all the reasons you started working on it.

I hadn't worked on the Gila Sub since before the pandemic, and decided it was long overdue to get a few sprints of work.

Seriously out of place detour for the El Capitan...

capture_400338.jpg

Favorite spot to hang out with the kids at Twin Peaks road.... we lived about six miles from here:

capture_400346.jpg

Looking north towards the Tortolita Mountains from Twin Peaks Road...

capture_400348.jpg


Picacho Peak

capture_400351.jpg

[eric]

When I did my first deep dive into TSRE, I'd done some work at the opposite end of the route in Yuma, and considerable amounts of track are already removed and awaiting rebuilding.

Why rip out 200+ miles of track?... The imagery from SRT was in the wrong place compared to TSRE, and TSRE is the new source of truth when it comes to editing and building, so there wasn't much of a choice. Here are two examples -- CP Kino (I think) near Tucson and the wye at Picacho where the Phoenix Sub heads north. In both cases, there's an offset of 10-15 meters which is consistent with what I'd seen in my other routes.




Based off how long it took to re-do the CNW Chicago and portions of Packerland, this will take up a good chunk of the summer.

As I've written before, being a former project manager and IT development manager, I tend to apply quasi-Agile principles and break things up into small chunks of work called sprints that last a few days to a week.

I'll typically do a month or so of trackwork sprints, shift and do a scenery and object building sprint or two... Then it's back to trackwork.

Unlike the original route build which went east from Tucson to Yuma, I'll be starting at the Yuma end. That's the most challenging section, and if I remember there was lots of serious cursing involved when I originally worked on that after months of the easy stuff.

Once I get to Wellton, it'll go faster.

To manage expectations... there won't be a release of this until the route is operable, and there's about 30 miles of missing track east of Yuma right now.

capture_400350.jpgcapture_400349.jpg

[pschlik]

Always felt like the sim community's time in Arizona was focused on Seligman. For understandable reasons, sure, but I always enjoy the reassurance that, yes, there's more out there than just the one route we've seen over and over again for over a decade. (Seligman is particularly egregious since I can think of half a dozen different incarnations of the route in less than a minute, but it's not the only route that steals attention from its neighbors.)

Coming back to a project after years is definitely interesting too...returning to realize how many new things the project is missing out on, how many things have become inexplicably broken (or were just always broken). Got the same stuff going on in my own endeavors, so here's hoping everyone who has old stuff lying around can find the motivation and time to get it out to the public in 2023.

[eric]

Seligman is definitely its own legend given the Santa Fe and being Route 66-ish and all... and I've spent time railfanning it because of that.

The Gila/Sunset Route is a bit more personal -- we lived off the mainline for many years, and our silver minivan shown in two of the screenshots spent a lot of time parked in that spot.

This rebuild I'm doing to the same standard I did with the last rebuild I did with the CNW, using actual track charts for grade. I guessed based on the terrain last time, there are some obvious errors in a few places I want to get right this time around.

[landnrailroader]

Looking good Eric. I rode over this section in Jan. 1999 while on a round-robin, Jacksonville to Jacksville by way of DC, Chicago,
Seattle, LA, New Orleans on several Amtrak trains.

[eric]

A few months back, I ordered a set of track charts for the Gila Sub (amongst others) and thanks to other things, hadn't quite gotten around to starting.

But every good route rebuild starts with one session.... and that was tonight.

There's nothing like some good documentation for building a route....

As mentioned before, I'm starting at the Yuma end and working back towards Tucson, so the most logical starting spot per the charts was CP Yuma, right near where the present day Amtrak station is.... so let's break out the chart:

capture_400570.jpg



I've always suspected I had the wrong elevation. Having driven by the Yuma yard dozens of times, I knew it wasn't elevated....


capture_400569.jpg

The tracks on the left are now at the correct height per the chart. The tracks on the right... not so much.

capture_400568.jpg

As I roamed around the area spot checking elevations on the terrain with the chart, I found that the HGT file generated terrain is within 0.5m of what the chart says they should be.

If you've ever watched "My 600 Lb Life", you know the doctor is famous for saying "the scale never lies" to his patients who claim they followed his diet but are amazed they gained weight between visits.

In my case, the charts never lie. They did, however, point out 10 years of mistakes that I was going to spend the next couple of weeks correcting. Nothing like a good smack upside the head to get things rolling.


Next week, the fun starts in earnest by ripping out five miles of track, including a bunch of curves and the yard, which is also on a curve. And I really hate curved yards... but fortunately the alignments are right, and it's just a matter of resetting the elevations and not the track geometry.

With luck, the 40 miles to Wellton should go quickly.

[eric]

I had some downtime tonight and spent around five hours digging into and wrapping up most of the Yuma yard. Felt good to do a long editing session for a change vs. chasing down database bugs and forum permissions....

This week I'd been working on MP732 to MP 738, which doesn't sound like a lot until you start working on it.... Earlier in the week I'd managed to get the mainline graded between MP 738 and MP 734.5. A lot of that time was re-learning TSRE's quirks, as it's been way too many months since I did any serious editing.

Tonight, I had some additional downtime, and was able to wrap up MP 732.5 (Amtrak Depot) to MP734.5 (Pacific Avenue) , which is a curved yard and an elevation change of about 75 feet between MP 733 to MP 735. Fun stuff.

Fortunately, when Melanie created Scalerail over 15 years ago, she'd added yard curves with increments that fit together like Legos...

I even got to use my switch ladder graphic aid that someone had asked about two weeks ago.... the 03d ladder is a little more complicated, and I'd made some custom pieces long ago using Dynatrax that fit the gaps between the turnouts, versus having to use several fractional pieces.

capture_400573.jpg capture_400576.jpg

Once the track was in place, I ran WorldFileHacker to add my berms as well as the highway overpasses I was ready to test.

Using the activity editor, I did some fit testing with my Fallen Flag Double Stack consist. I used to have to fire up ORTS and create a path for where I was working, but by creating a throw-away activity, I can drop a consist into my editing session without having to close things down.

Pacific Avenue is perfect:

capture_400577.jpg

Not so much with Avenue 36: Another 2m of height is clearly needed for the roadway... and apparently I didn't have the road sections used paired up with bridges in WFH, so I'll have to fix that next time I do a deep dive.

capture_400578.jpg

As I was wrapping up, something caught my eye on the two leftmost yard tracks.... TSRE Zigs.

For whatever reason, this happens randomly.... Your track will look fine when you edit and save, but upon coming back into the editor or running in ORTS, split vectors will appear usually in areas with multi-track yard sections. This is how they appear:

capture_400575.jpg

The other random problem are flipped factors at turnouts. That manifests itself with the lack of a yellow line between the red pole of your turnout and the piece attached to the red pole.

It's easy enough to correct either of those situations with a pair of Z keystrokes (one to remove the track from the TDB, and the second to re-add it to the TDB). Usually it sticks unless something happens to the vector.

Gotta say it felt good to get back into an editing session and see some progress, even if it is only 5 miles out of the 250 miles of track I'm rebuilding. Certainly more fun than chasing down database errors or forum permissions...

[landnrailroader]

What is "world file hacker". That is a new one to me. I have seen those two random problems, and the double Z worked for me too.
J. H. Sullivan

[eric]

WFH is like CatMaster - it reads the static detail level of the track/road/static/signal objects in the World files, and then will add other items using the same XYZ coordinates and qDirection using a cross reference file.

For example, roads with SDL of 2 will get a bridge deck, SDL 1 will get a solid filled base, SDL 3 might get a turf embankment/berm, and so on.

With track - by default (SDL 0), all track gets a matching ballasted berm underneath, 03d and 05d switches get ballast plus a CTC cabinet, yard 06d switches get an overhead light pole, and my concrete track sections will get either third rail or catenery depending on the route.

Signals get a ballasted base with a cabinet with tie retaining walls to give a working area around the cabinet and signal pole.

Again, that's all configurable to fit what your railroad and era would use for trackside accessories or roadway.

[eric]

Wrapping up sprint 2 just few miles east of Yuma... when working on the double track mainline near Fortuna, this happened... track running right thru the intersection of four tiles.

For those who are new to route building... this is one of the biggest danger zones you can find. Never let a single piece of track span more than two tiles. Bad things happen with TDB calculations, and in MSTS this was a guaranteed "slag" moment as Vince would put it.

capture_400598.jpg

In this case, a 2T 10m section (highlighted in blue) fit just fine and worked as the simplest solution. Sometimes you'll have to use combinations of straights with different lengths to get multiple tracks to avoid the third tile.

capture_400599.jpg

A little further down, the Yuma Proving Grounds make a brief appearance:

capture_400606.jpg

Between Fortuna and and Blaisdell is a 250m long trestle over a wash, which is pretty common out west. This was something I'd built in Sketchup in 10m and 50m sections.

capture_400605.jpg

Rolling stock: Diesels West SP Pack

[eric]

capture_400616.jpg No screnshots tonight, but sprint 4 came to a close - track is reset from MP735 to MP767, just shy of Wellton, AZ. While 32 miles is a bit more than I expected to cover, getting to Wellton is a mental victory as that's where things get easy.

Distance is fairly misleading on the Gila - there are many long stretches of straight track with gradual elevation climbs, and there are some short spans with many, many curves and rapid elevation changes. The track between Yuma and Wellton is the latter.

I did something outside the norm and added a sprint into the existing plan (as Scrummaster and the PMO, it's allowed). As I alluded to in the Dynatrax thread, I spent last week experimenting with Blender, specifically building some track sections using a script developed by Doug Jones.

While most of my routes use 4M track spacing, the Gila is built to a more modern 5M spacing.

I've long had a need for some gap filling double track sections that Melanie opted not to include with Scalerail - specifically 750r01d and 750r05d, 1000r01d, 1500r01d and 2000r01d. I've made do with single curve pieces but have wanted to have the simplicity of the 2T pieces for a while.

Thanks to the script, those five 2Tcrv sections are now built and in use on the section of track I just finished up. I will probably make 2Ycrv equivalents before I start back up on Packerland again (which is not imminent) or I extend the CNW Chicago north to include the New Line and Kenosha Subs (also not imminent but more likely to get started this year).

I've also been working on new replacement m03d switches, also to be generated by Doug's script. While the 06d Scalerail switches have frogs, the basic 03d switches do not because of their modular nature i.e. the basic swt piece doesn't have the complete diverging track and straight track, which is where the frog would be. The m03d_ext variants do have a gap in the rails at the crossing, but no guard rails. The only possible gap I have to fix there is that I need to build a model of a low profile and throw switch stand. Doug's script has a high profile lever style switch, but my preference and what's more prototypically accurate are the low profiles that can essentially be moved with your boot...

While track is track to some people, building sections has been a nice diversion from laying track, and having those little victories are what keep us route builders moving.

The next sprint 5 will be a combination of testing out the track and looking for flipped vectors and zags, plus swapping out some of the modular 03d switches in the Yuma yard and at the passing sidings between MP738 and MP767. Since the TDB errors manifest as track vectors are modified, the swaps might happen first.

And yes, you'll have screenshots.

[Curt C]

Really looking forward to this route! I've been there, and the scenery is amazing!

[eric]

It truly is amazing sccenery. Reminds me how much we loved living there every time I run the route....

[eric]

OK, I just realized that we're going into Week 13 of the NFL season, and I haven't touched this once since the season started.

Break time is officially over, and it's time to start up on the track rebuilding...

capture_500105.jpg


Tonight I was able to knock out about 15 miles from Welton (MP 771) to Colfred (MP 885).

While Colfred really is in the middle of nowhere, it's 50 miles from Yuma, and 200 miles from Tucson, so that's 20% complete for the track rebuild.

Looking at the rest of the route..... that remaining 80% really is nowhere near as challenging as the first 20% was.

MP785-792: No curves for 7 miles
MP792-795: Mohawk Pass, with six curves and a passing siding
MP795-809: No curves for 14 miles
MP810-822: No curves for 12 miles
MP822-825: Eight curves east of Stanwix
MP826-830: No curves for 8 miles
MP830-833: Three curves at Sentinel
MP833-852: Slight curves every 4-5 miles
MP852-867: No curves for 15 miles (Gila Bend)
MP867-879: Thirteen curves
MP879-888: No curves for 9 miles
MP888-892: No curves for 4 miles
MP892-897: Five mile curve....
MP898-944: No curves for 46 miles (Maricopa, Casa Grande, Eloy, Picacho)
MP945-947: Two curves
MP947-959: No curves for 8 miles
MP959-968: No curves for 9 miles
MP969-971: Two curves
MP975-985: Fourteen curves (Marana to Tucson)

Each of those straight chunks does have grade changes averaging between 0.5% and 1%, so while it's not a slam dunk, it's relatively fast going once you're in a rhythm.

Broken up into straights and curves, there are really only about 20-30 "days" of work left. Maybe not by year end, but entirely possible by mid-January.

[jbtower]

You make it sound so simple Eric and I imagine it is once you learn the tools. Looking good!