Update on the Steamers V2

[dtrainBNSF1]

It turns out QD flips out whenever you have a vehicle with 0 weight because it dumped every time I tried to make a run, so I ended up spreading the weight between the front and rear engine units; now QD works correctly.

I should add a bit of detail about the front unit:
The front unit is now a tank engine like I said. It now has its own coal and water amounts which I set to 99999 each just so I wouldn't need to worry about it in scenarios. The front unit also has its own simulation file. The maxforce line reads 0 since I don't want it doing anything, the steam multipliers for the injectors read 0, and the TenderIsRequired line reads eFalse so now it really is independent of the rear unit.

The tender has an adjusted number file so that it reads from 4000-4019 (4020-4024 had slightly different fuel amounts; I'll adjust the number file for the engine units so that there is a difference between the first 20 and the last 5 as the last 5 were heavier and steamed a bit differently than the original 20). The tender is also weighted as it would be if it were empty with the fuel amounts being in imperial units rather than US units. The boiler needs adjusting because even with the injectors on you can run the boiler dry when working hard up the grade.

[dtrainBNSF1]

A new development with the dead front unit:

In speed trials the engine unit is derailing. This is striking because in my previous mods for the Big Boy the engine never derailed once. Ever. Even at top speed of 80mph. The front unit seems to dip beneath the rails and roll as it goes around the curve on the TestTrack. I also noticed that the guiding wheels on the front of the engine unit sink a bit beneath the rails at the beginning of a scenario. I've never noticed this before and I'm going to take a look at a regular unmodified front engine unit to see if it's always sank beneath the rails or if it's something to do with the process of making it a dead unit in RWTools.

[dtrainBNSF1]

Here is the comparison:

Screenshot_TestTraK_51.11707-6.22412_12-01-32.jpg

Screenshot_TestTraK_51.11714-6.22412_12-01-12.jpg

I'm not sure why this is happening. Could someone more familiar with RWTools explain what exactly the application does when it converts a locomotive from a live unit to a dead unit?

[dtrainBNSF1]

After further testing it appears that the pilot dipping below the rails is NOT related to the pilot being dead. I tested it out today with a few more parameters and my thinking is now that it seems to be linked to weight. I lighten the weight and the pilot dips forward. I leave it at its original weight and it's just fine. A curiosity to say the least because I don't remember this being a thing before but that was the one parameter I found that would make it dip. Cue Twilight Zone theme.

Anyway after getting that settled I spent the afternoon getting my speed parameter set up. Given that each set now produces 67.6875klbf I found that my 10% cutoff rating should be no less than 0.192 (19.2% of the total tractive effort) in order to ensure 80mph light engine. That translates to a total of 25992lbf combined (or 12996lbf per unit), which given that a Dash 9 at 75mph should theoretically be producing around 18310lbf it seems reasonable to me. Around TestTrak it hovered at 80mph and sometimes hit 80.1mph but never higher so I'm satisfied with that. For now. We'll see how it does after boiler tests.

It turns out that Big Boy really does fit on the turntable at Cheyenne even better than it should in real life (Back in the day the pilot and rear coupler hung off the ends of the turntable which still allowed it to turn; in TS20XX both the pilot and rear coupler fit completely on the turntable quite comfortably). I never thought it could. I managed to take it 3/4 of a full rotation around the table before it derailed. I'll tinker with the EaseOfDerailment line to get it around the table without derailing. It can be done already by stopping the table at every track but I want to be able to turn Big Boy all the way around in one continuous turn like back in the day.

I found a spelling error in the simulation file! The FeedRateMultiplier line was misspelled and instead read FeedRateMulitplier (emphasis added). I've gone ahead and corrected it; perhaps this will improve the performance of the injectors? I dunno - I'll check it tomorrow.

[dtrainBNSF1]

I would like to mention that given the confirmation that both loco sets do indeed produce tractive effort in-game that means that all of my previous mods for articulated steam locomotives have acted like 2 locomotives having to compensate for the weight of only 1 locomotive. To say the least hauling heavy trains over hilly territory such as Sherman Hill and Donner Pass will undoubtedly be more difficult and slower than previously experienced in my mods. By how much I don't know. When I do get to haulage tests with the Big Boy specifically after sorting out the boiler and the injectors I'll be satisfied to be making 15mph and maybe slightly less than that over main tracks 1 & 2 with 3250t. and for 5800t. over main track 3 maybe 20mph or slightly less by my own calculations given evaporation, driver diameter, rpm, etc. which actually agrees with accounts about the Big Boys on various steam locomotive forums. I dunno. We'll just have to see.

[dtrainBNSF1]

I've got the Big Boy's steam raising capabilities tuned to where I want them thanks to adjustments to the BlastExponent, MaxOutput, Drafting Effect, Effectiveness, and Superheater lines. Now to make adjustments to the Cylinder Effectivity and the Boiler ExhaustLimit lines to get the speed and reverser settings tuned up.

[dtrainBNSF1]

Speed trials went okay, but haulage tests went...worse than I thought they would. With a 3250t. train I was hitting 12mph and slowing on the .8% incline just outside Cheyenne yard. I thought "Good grief if I'm having this much trouble with 3250t. before I even get to the 1.52% grade on main tracks 1 & 2 there's no way I'll haul 5800t. on the .82% on main track 3!" Something's up with the power so I think I'll have to do the speed trials all over again tomorrow.

Oh well.

[dtrainBNSF1]

Went ahead and ran the speed trials again, made a couple of slight changes, and tried the haulage test out of Cheyenne again. This time I got to Wycon at 10mph still on the .8% part of the grade and decided that it was still wrong. So I said "screw it" and set each locomotive unit to 135.375, ran the speed trials, made appropriate adjustments to the boiler and the TractiveEffortVsCutoff file (the loco is now down to 9.6% power at 10% cutoff - not quite a 1:1 ratio) and ran the test out of Cheyenne again. As expected things were a bit easier this time. I was hitting anywhere from 24-25mph on average on the 1.52% grade with 50-55% cutoff and the injectors running pretty much all the time.

I have noted that since converting the front unit into a tank engine that water now only flows from the tender to the boiler only when you turn on the injectors and that water does stop flowing when you turn the injector off. I'm much less nervous about my tender's water supply lasting me for a whole trip over Sherman Hill thanks to this discovery. It appears that all the occurrences of water flowing from the tender to the boiler without player input was the front unit draining the tender water supply to replenish itself. That flies in the face of convention, however, as before we've all assumed that a locomotive with "eTrue" listed in its "TenderIsRequired" line needed to be directly connected to a tender in order to access coal and water. According to what's been observed, however, it seems that the front unit can indeed access the tender through the player locomotive. Interesting.

The F4 HUD is now much more accurate with what information it displays now that the front unit is a tank engine. The steam pressure on the F4 HUD drops in time with the safety valves lifting and it displays the correct amount of water in the boiler. The F5 HUD is a bit different, however. When sitting idle the F5 HUD and the F4 HUD are out of sync first by a few tenths but it gets more and more out of sync the more it sits there and the figures are not in time with the sound of the safety valves lifting and closing. In motion working hard, however, the F4 and F5 HUDs line up perfectly, even with the lifting of the safety valves. This needs further investigation.

I still need to do some work with the injectors. The water flow from the tender still cannot replenish the boiler's water supply quickly enough. I wound up having to abandon my last test a few miles down from Granite because I was down to 20% in the boiler and I figured I was not going to have enough water to make it to Buford.

There is progress being made now. BTW if anyone is wondering I have not forgotten about the K4s. This work on the Big Boy right now will help me when I get ready to finish the K4s mod.

So this is how the Big Boy mod is shaping up so far:
-A more realistic output of 125000lbs/hour
-Further refinement in the boiler and cylinders
-A top speed of 80mph (give or take a few tenths)
-A brand new method out power calculation which puts even more emphasis on how well the driver can manipulate the cutoff which is affected by how well the boiler is being fired, ranging from 100% power at full cutoff to 9.6% at 10% cutoff
-Elimination of water mysteriously flowing out of the tender without player input
-Of course the standard fix for the particle emitters and the cylinder cocks
-Once I figure it out, placing the coal in the tender at the top of the tender instead of all the way down near the bottom (may need some assistance on this one)

That's about it for now.

[dtrainBNSF1]

Well fellas, it happened by accident this morning, but I fixed the coal animation in the tender! All I did was change the name of one animation. One animation. Didn't need to mess with the geometry. Didn't need to meddle with blueprints. Just change the name of a single animation from BB_Coal to Challenger_Coal. And now it works. Don't know why, don't know how, but I just know it works now.

Lookit!

Screenshot_Sherman Hill, revised_41.13025--104.81363_12-00-46.jpg

Screenshot_Sherman Hill, revised_41.13003--104.81518_12-02-01.jpg

Screenshot_Sherman Hill, revised_41.13045--104.81368_12-02-30.jpg

In other news, I found out something else about the Big Boy. The live injector water control doesn't work because it doesn't exist! You can turn on the live injector steam but there is not control for the water so it doesn't flow! I changed the input mapper so that the only think attached to the L key is the live injector water control and no matter how many time I pressed L no water flowed from the tender. The only injector that exists it the Big Boy is the exhaust injector. No wonder the base version couldn't fill the boiler fast enough.

You know, the deeper I go into this particular mod the more I want to take IHH or whoever built this thing and smack 'em across the face because it is so messed up.

I also fiddled with the numbering scheme of the front unit. The two engine units use the exact same id number in a scenario which in TS20XX is a no-no. The fix: change the numbering scheme of the front unit such that it does not read X40-whatever and instead read as A40-whatever. It renders correctly in the simulator, the rear frame where the player sits is still read as X40-whatever (which it needs because it also controls the number board on the front which needs that X), and the tender always had C in its id number, so now each unit has its own unique id number. Maybe a bit of OCD on my part, but I like it.

So the checklist so far:
-Fix the tender coal animation: done
-Improve the physics even more than before: I think I'm on the right track
-Fix the problem with water flowing out of the tender without player input: done
-Fix the injectors: done since the live injector clearly does not exist
-Fix the numbering scheme: done
-Implement all previous fixes from previous mods: done

I'm still going to run some haulage tests. I've always noticed how hard it is to take a 3250-ton train from Cheyenne to Buford compared to how easy it is to take a 5800-ton train from Laramie to Buford. I have a theory that my loco rating chart could be referring to the line that goes via Red Buttes, the older line, rather than the .82% alignment of main tracks 1&2 from Laramie. I decided to take my 5800-ton test train from Laramie through Red Buttes. It honestly isn't that bad of a run and you can still get to Hermosa faster by way of Red Buttes even with 5800 tons. The only real tricky parts are the two 1.6% sections that are almost right next to each other just outside of Hermosa. If you treat these sections as endurance grade and hit them at a fast enough speed you'll still make it to Hermosa without stalling - just. I crawled over the hump into Hermosa at something like 5.5mph coming off the 1.6% grade. But I made it, and that's what counts. I'm going to repeat the tests to make sure I can consistently make it up to Hermosa. I've found that coming out of Laramie on main tracks 1&2 the Big Boy can handle over 7000 tons before it even breaks a sweat.

[AmericanSteam]

Excellent work. I have been following your progress. I know that there have been several attempts to upgrade the physics in the past but I do not think that anyone has delved as deep as you in the original physics model. I have always enjoyed the Big Boy but was unimpressed with the power that it had. Thank you for your efforts.

[thegevo5k]

I've been following your progress as well. I never thought that the tender load had to do with the animation! I can't wait to see this thing in action.

[dtrainBNSF1]

Thank you AmericanSteam and thegeo5k for your comments. It's always nice to know that I'm not just rambling to myself here on the forum

So this morning I decided to try and recreate life in Train Simulator (sort of). As any big steam fan will recall back in the early 90s American Presidential Lines secretly arranged for 3985 to haul 140 double stacks from Cheyenne to North Platte. Once the water car, tool car, and a caboose were added the car count came to 143 cars and the trailing tonnage was reported to be around 7657 tons and 8900 feet. Not since the 50s had a steam locomotive been at the head of a long freight such as this, and 3985 handled the train in fine style, topping nearby Archer Hill around 30mph and reportedly hitting around 65mph on the main line and was the closest of any train to being on time into North Platte that day only being about 10 minutes late.

This morning I had to do it. Using specially modded double stack cars to get the right weight, I recreated 3985's 143-car train, all except the length which came out to over 10,000 feet

Here it is:

Screenshot_Sherman Hill_41.15858--105.52702_09-19-45.jpg

Screenshot_Sherman Hill_41.21336--105.55335_09-14-00.jpg

Screenshot_Sherman Hill_41.18517--105.54856_09-16-11.jpg

The stack cars are modified versions of the stack cars that came with the Big Boy which are specifically weighted to make this possible. The train was put together in the consist editor for QD so I can verify that the trailing tonnage is indeed very near the benchmark of 7657 (it was more like 7656.99-repeating but you get my point). Including the weight of the engine the train was over 8100 tons. The loco managed to hit line speed of 40mph on the line between Laramie and Hermosa. Here's the neatest thing about this run: the math!

Using a calculator and the particular way I'm treating power in this particular mod, with the throttle all the way open and the F5 HUD showing that I was using 38% cutoff to maintain speed at 40mph, I determined that I was putting out 51813.88274336285lbf to maintain track speed. Take that figure and multiply it by 40mph then divide by 375 and we can see that the loco is producing 5526.814159292037hp. I was still building steam by this point, so this loco wasn't even working at full capacity.

In short: she's a beast. Just a few more tweaks and she should be better than ever.

[AmericanSteam]

Whoo Hoo!

[dtrainBNSF1]

Alright, if anybody has a calculator handy you'd see that something was amiss with that 143-car test run I made out of Laramie. With the tractive effort the Big Boy was putting out you'd see that the numbers don't line up. The loco was acting like I was towing 3159.383094107491 tons. I honestly don't know how to account for that. Maybe a blueprint didn't load properly when the QD scenario started? I don't know.

The reason I bring this up is because I ran the test again, this time as a free roam scenario coming out of Cheyenne on the Harriman line. With the exact same train I was struggling to make anywhere from 16 to 18mph. This troubled me seeing as it stood in stark contrast to what had just happened coming out of Laramie, so I checked the numbers, reset, and the same thing happened. I ran the test out of Laramie as a free roam instead of a QD and once again I was struggling to get anywhere from 16 to 18mph. I still cannot account for why I was so easily making 40mph up the grade on one test and suddenly dragging my feet in the next test. The only explanation I can give is that perhaps the blueprints did not load correctly.

I took some time off to familiarize myself with the new Raton Pass route. It has plenty of good stuff, but I'm still conflicted about how I feel about it (nice range of freight cars but none of them bear any ATSF marks which looks a bit off and makes it look generic, the locos have way too much horsepower so they accelerate much too quickly, the semaphores have only 2 animation frames which change instantly, despite the signals clearly being 3-aspect they only display either green or red, the superliner Trans-dorm car is NOT a Trans-dorm car but looks exactly like a regular sleeper, the scenarios aren't exactly the most thrilling things to play, the dynamic brakes are too powerful which can result in a lot of BS drive quality penalties, and sure modelling 45 miles of 3.5% grades in the mountains shouldn't be anything to sneeze at but it would have been nice to go all the way to La Junta to recreate a full shift and take more advantage of the so-called "high-speed" aspect of this route but hey we got some nice ATSF Dash 8s, articulated trailer flats and Amtrak MHCs at least) and also I needed a break from running the Big Boy.

Work continued today. The freight stock has been re-weighted to the following specifications:
The boxcar now has an empty and a loaded variant, weights 49t. and 131.5t. respectively (based on Union Pacific's specs)
The hopper also has an empty and a loaded variant, weights 35.1t. and 131.5t. respectively (based on what's printed on the side of the car in-game)
The double stacks also have 2 variants, the first variant replicating two empty 53-foot containers and the second variant replicating 2 loaded 53-foot containers, weights 35.85t. and 92.45t. respectively (based on intermodal specs for 53-foot containers and the tare weight of a husky stack car)
The caboose now weighs 27t. (based on a museum article centered on an extended-vision caboose)

I made just the most minute of changes to the Big Boy's physics and I can get the 7657-ton intermodal up the grade somewhere in the 20s now which I'm happy to accept.

My next step with this mod is to now replicate the 4884-2 class of Big Boys. Union Pacific had 2 classes of Big Boys, the 4884-1 and the 4884-2. Both classes were pretty similar but the 4884-2 class, being built in the middle of WWII, were heavier due to restrictions in building materials and steamed differently than the earlier 4884-1 class and their tenders were slightly modified to carry more water and slightly less coal. These will present a different driving experience due to their different power-to-weight ratio and different steaming characteristics. How much different? I won't know until I try it

Again I assure you guys that the K4s mod will be released before this latest Big Boy mod is released. This latest mod with the Big Boy serves as a proof-of-concept mod for some of the ideas I've been tossing around for the last year or so which will serve as a guide to what needs to be done with the K4s (I left off during the speed trials I think...). I'm thinking as a K4s never actually hit 100mph with a train and the fastest speed recorded was 91mph with a 1000-ton test train on a descending .51% grade that setting an absolute top speed of 100mph for the K4s light engine should be appropriate based on the loco's 80-inch driving wheels (later eras have to take into account better balancing, streamlining, and the application of roller bearings all of which yielded higher top speed). We'll see how that translates to practical every-day performance once I get there.

Oh yeah and then I read somewhere that the SP AC10-12 classes actually had a top speed of 63? I'll look more into that.