Update on the Steamers V2

[JohnS]

This is good to here the K4s is getting a mod. I added more steam effects to mine and changed some of the audio. I think I set the brakes up like the old 6 style with "Release, Runninng, Lap, Apply, Emergency". It's the way the brakes on the Corris Railway #7 are for a reference. I like the Loco but it does need a major overhaul in the physics department. Thank you for doing this project.

[dtrainBNSF1]

It's good to hear that there's still some interest in these mods, especially since it's been quite a while since I put out a steam loco mod.

[dtrainBNSF1]

Howdy!

I finally figured out which tender the K4s uses in TS (there were something like 8 different variants in real life!). The one used in-game seems to be most similar to the 110-P-75 tender with 18.5 tons of coal and 11,980 gallons of water.

[dtrainBNSF1]

Hey y'all!

Winter break was long, cold, and I came down with bronchitis, but I'm feeling much better.

Here is a mathematical analysis of the K4s' power and tractive effort from 50rpm to 250rpm. A different analysis will be used for speeds above 250rpm.

Evaporation rate: 51110
Locomotive and tender weight: 258.6125 U.S. tons

50rpm(11.89393939393939mph)=44460; 1410.145454545454hp
15mph=42956.11631509923; 1718.244652603969hp
20mph=38786.86871025149; 2068.632997880079hp
25mph=35341.70814871412; 2356.113876580941hp
30mph=32227.41127729065; 2578.192902183252hp
35mph=29291.83919527842; 2733.904991559319hp
150rpm(35.68181818181818mph)=28900.93986470159; 2749.968217429181hp
40mph=26232.32957802547; 2798.115154989383hp
45mph=23613.75593152866; 2833.650711783439hp
50mph=20995.18228503184; 2799.357638004245hp
55mph=18376.60863853503; 2695.235933651804hp
250rpm(59.46969696969697mph)=16035.7625; 2543.045164141414hp

As you can see, maximum DBHP is achieved at 45mph with 2833.650711783439hp.

[dtrainBNSF1]

Hey guys!

Thought I'd give you guys an update as to how things are going with the K4s as this is probably the longest I've worked start to finish on a single mod.

Things are going a bit slower due to my work as a student and graduate assistant in grad school, but as far as the physical properties (weight, fuel amounts, etc.) the loco and tender are finished. The actual physics and boiler work are still in progress, and here's why:

As you can see by my analysis of the locomotive's tractive effort and horsepower at certain speeds, I'm trying to get down into the nitty-gritty of a locomotive's physics. Looking at some of the more realistic steam locomotives of Train Simulator such as the FEF-3 and the Just Trains version of the Clans you can see that there is a bit of a drop in the total tractive effort in the tevsspeed file, but most of the tractive effort losses are handled in the tevscutoff file. Right now I'm looking at an engine that approaches the losses of tractive effort slightly differently. It's an old engine that I purchased back in 2012 when I was still new to Train Simulator/Railworks and is no longer available in the store although I think you can still find it on Amazon: the LMS Jubilee.

So why am I looking at an old engine that's no longer available on Steam? Because I found a nugget I think is worth exploring: its tevscutoff and tevsspeed files. Of all the steam locomotives in my collection, it's the only one where the tevsspeed file stays constant. No kidding! Every speed entry in the file is accompanied by "1", indicating that full tractive effort is available at all speeds. You might say "The engine's broken", but look at its tevscutoff file. Take a look at what's there:

<CSVItem>
<cCSVItem d:id="35660848">
<X d:type="sFloat32" d:alt_encoding="0000000000000000" d:precision="string">0.0000</X>
<Y d:type="sFloat32" d:alt_encoding="000000606666C63F" d:precision="string">0.1750</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660860">
<X d:type="sFloat32" d:alt_encoding="00000040E17AA43F" d:precision="string">0.0400</X>
<Y d:type="sFloat32" d:alt_encoding="00000000022BC73F" d:precision="string">0.1810</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660872">
<X d:type="sFloat32" d:alt_encoding="00000040E17AB43F" d:precision="string">0.0800</X>
<Y d:type="sFloat32" d:alt_encoding="00000080EB51C83F" d:precision="string">0.1900</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660884">
<X d:type="sFloat32" d:alt_encoding="000000E051B8BE3F" d:precision="string">0.1200</X>
<Y d:type="sFloat32" d:alt_encoding="000000A0703DCA3F" d:precision="string">0.2050</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660896">
<X d:type="sFloat32" d:alt_encoding="00000040E17AC43F" d:precision="string">0.1600</X>
<Y d:type="sFloat32" d:alt_encoding="000000E0A370CD3F" d:precision="string">0.2300</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660908">
<X d:type="sFloat32" d:alt_encoding="000000A09999C93F" d:precision="string">0.2000</X>
<Y d:type="sFloat32" d:alt_encoding="00000000D7A3D03F" d:precision="string">0.2600</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660920">
<X d:type="sFloat32" d:alt_encoding="000000E051B8CE3F" d:precision="string">0.2400</X>
<Y d:type="sFloat32" d:alt_encoding="000000205C8FD23F" d:precision="string">0.2900</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660932">
<X d:type="sFloat32" d:alt_encoding="0000002085EBD13F" d:precision="string">0.2800</X>
<Y d:type="sFloat32" d:alt_encoding="000000E0A59BD43F" d:precision="string">0.3220</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660944">
<X d:type="sFloat32" d:alt_encoding="00000040E17AD43F" d:precision="string">0.3200</X>
<Y d:type="sFloat32" d:alt_encoding="000000E078E9D63F" d:precision="string">0.3580</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660956">
<X d:type="sFloat32" d:alt_encoding="000000803D0AD73F" d:precision="string">0.3600</X>
<Y d:type="sFloat32" d:alt_encoding="000000208716D93F" d:precision="string">0.3920</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660968">
<X d:type="sFloat32" d:alt_encoding="000000A09999D93F" d:precision="string">0.4000</X>
<Y d:type="sFloat32" d:alt_encoding="000000403333DB3F" d:precision="string">0.4250</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660980">
<X d:type="sFloat32" d:alt_encoding="000000C0F528DC3F" d:precision="string">0.4400</X>
<Y d:type="sFloat32" d:alt_encoding="000000A09999DD3F" d:precision="string">0.4625</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35660992">
<X d:type="sFloat32" d:alt_encoding="000000E051B8DE3F" d:precision="string">0.4800</X>
<Y d:type="sFloat32" d:alt_encoding="000000000000E03F" d:precision="string">0.5000</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661004">
<X d:type="sFloat32" d:alt_encoding="00000000D7A3E03F" d:precision="string">0.5200</X>
<Y d:type="sFloat32" d:alt_encoding="000000403333E13F" d:precision="string">0.5375</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661016">
<X d:type="sFloat32" d:alt_encoding="0000002085EBE13F" d:precision="string">0.5600</X>
<Y d:type="sFloat32" d:alt_encoding="000000606666E23F" d:precision="string">0.5750</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661028">
<X d:type="sFloat32" d:alt_encoding="000000403333E33F" d:precision="string">0.6000</X>
<Y d:type="sFloat32" d:alt_encoding="000000A09999E33F" d:precision="string">0.6125</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661040">
<X d:type="sFloat32" d:alt_encoding="00000040E17AE43F" d:precision="string">0.6400</X>
<Y d:type="sFloat32" d:alt_encoding="000000C0CCCCE43F" d:precision="string">0.6500</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661052">
<X d:type="sFloat32" d:alt_encoding="000000608FC2E53F" d:precision="string">0.6800</X>
<Y d:type="sFloat32" d:alt_encoding="000000000000E63F" d:precision="string">0.6875</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661064">
<X d:type="sFloat32" d:alt_encoding="000000803D0AE73F" d:precision="string">0.7200</X>
<Y d:type="sFloat32" d:alt_encoding="000000403333E73F" d:precision="string">0.7250</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661076">
<X d:type="sFloat32" d:alt_encoding="00000080EB51E83F" d:precision="string">0.7600</X>
<Y d:type="sFloat32" d:alt_encoding="000000606666E83F" d:precision="string">0.7625</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661088">
<X d:type="sFloat32" d:alt_encoding="000000A09999E93F" d:precision="string">0.8000</X>
<Y d:type="sFloat32" d:alt_encoding="000000A09999E93F" d:precision="string">0.8000</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661100">
<X d:type="sFloat32" d:alt_encoding="000000A047E1EA3F" d:precision="string">0.8400</X>
<Y d:type="sFloat32" d:alt_encoding="000000C0CCCCEA3F" d:precision="string">0.8375</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661112">
<X d:type="sFloat32" d:alt_encoding="000000C0F528EC3F" d:precision="string">0.8800</X>
<Y d:type="sFloat32" d:alt_encoding="000000000000EC3F" d:precision="string">0.8750</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661124">
<X d:type="sFloat32" d:alt_encoding="000000E0A370ED3F" d:precision="string">0.9200</X>
<Y d:type="sFloat32" d:alt_encoding="000000403333ED3F" d:precision="string">0.9125</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
<cCSVItem d:id="35661136">
<X d:type="sFloat32" d:alt_encoding="000000E051B8EE3F" d:precision="string">0.9600</X>
<Y d:type="sFloat32" d:alt_encoding="000000606666EE3F" d:precision="string">0.9500</Y>
<Name d:type="cDeltaString"></Name>
</cCSVItem>
</CSVItem>

Notice the cutoff values (the X values). In most locomotives the lowest value is .1, or 10%, but here in the Jubilee we see that it goes all the way down to .04, or 4%. All of the losses in tractive effort are handled in this file, and that's actually closer to my understanding of a steam locomotive's tractive effort IRL.

Here's what I mean: If you look at any locomotive's performance chart you'll see that tractive effort drops off as speed increases. Why is that? Consider a manual transmission vehicle. When it reaches a certain speed, you have to down shift in order to ensure that the vehicle operates smoothly and efficiently and to allow the vehicle to move faster. This drops the vehicle's torque but you don't need a whole lot of torque to move a single automobile quickly down the road. It would be perfectly happy to provide more pulling power, but it can't physically move quickly unless you shift.

Steam locomotive's operate with much the same principle. When you put the locomotive into full forward gear, you're maximizing the amount of space of the steam chest will be filled, thus allowing the locomotive to use its full tractive effort. At a certain speed if you haven't wound back the gear you'll notice that the locomotive either ceases to accelerate or the locomotive starts to buck/bounce. This is the locomotive telling you in the most unsubtle way possible that it's time to wind back the gear, allowing it to move faster but at the cost of tractive effort. When you wind back the gear, you're reducing the amount of space in the steam chest that can be filled with steam, thus reducing tractive effort. Theoretically a steam locomotive will be more than happy to keep producing full tractive effort at any speed if physics could be ignored to allow you to have full forward gear at any speed, but the amount of steam allowed by the cutoff limits the ability of the locomotive to move quickly in real life, and that's where the tractive effort curves come from: the reduction of the cutoff reducing the available tractive effort as the locomotive increases its speed.

The Jubilee is the only engine in my collection that hits on this principle. The distribution for each value could be better, but if you fire the loco correctly and keep the evaporation rate up you can get this loco moving down the line at a good speed with a decently heavy load.

What I'm doing now is trying to figure out how to apply the power curve to the tevscutoff file, and that's some math that I need time to work through and figure out because it's not as simple as plunking a value down and calling it done, and I'm just trying to find time to work it all out, then run the locomotive through tests for speed and haulage capabilities and make any neccesary changes to the boiler behavior before I can call this done.

Whoo that's a long post.

In short, it's coming, but I just need the time to sort it all out.

[mrennie]

You still need the slight drop in t.e. at higher speed to account for the wire-drawing effect that occurs in the valves.

There's one other thing that's very important to understand about why you have to shorten the cutoff as speed increases - it's about the limit on the exhaust valves' capacity to expel steam from the cylinders. As the speed of the pistons increases, there's less time for the steam to go through the exhaust port. There's also less time for the live steam to expand (doing work through expansion and lowering the pressure). Consequently, if you keep a long cutoff as you go faster and faster, there's more and more high-pressure (live) steam still in the cylinder when the piston changes direction and pushes against that steam, i.e. there's higher back pressure. That's what stops you from going faster (and can actually slow you down). TS does a pretty good job of simulating that all on its own, as long as you put the proper values for Exhaust Limit in the Boiler section of the engine sim blueprint (it ought to be in the Cylinder section, but there are a lot of mistakes like that in the blueprints). You have to put a high enough value to allow the cutoff to be increased far enough before the F5 HUD "steam chest pressure" (which isn't steam chest pressure - it's really the mean pressure in the cylinders on each stroke) gets as high as the boiler pressure. If you set it too low, you'll soon hit the limit in the exhaust ports' capacity for expelling steam, before you've been able to advance the cutoff very far.

The fact that the FEF-3 (and now the Connie too) has a back pressure gauge is a real boon in understanding what's going on in the cylinders, and especially on the exhaust side of the piston strokes.

[dtrainBNSF1]

You still need the slight drop in t.e. at higher speed to account for the wire-drawing effect that occurs in the valves.

There's one other thing that's very important to understand about why you have to shorten the cutoff as speed increases - it's about the limit on the exhaust valves' capacity to expel steam from the cylinders. As the speed of the pistons increases, there's less time for the steam to go through the exhaust port. There's also less time for the live steam to expand (doing work through expansion and lowering the pressure). Consequently, if you keep a long cutoff as you go faster and faster, there's more and more high-pressure (live) steam still in the cylinder when the piston changes direction and pushes against that steam, i.e. there's higher back pressure. That's what stops you from going faster (and can actually slow you down). TS does a pretty good job of simulating that all on its own, as long as you put the proper values for Exhaust Limit in the Boiler section of the engine sim blueprint (it ought to be in the Cylinder section, but there are a lot of mistakes like that in the blueprints). You have to put a high enough value to allow the cutoff to be increased far enough before the F5 HUD "steam chest pressure" (which isn't steam chest pressure - it's really the mean pressure in the cylinders on each stroke) gets as high as the boiler pressure. If you set it too low, you'll soon hit the limit in the exhaust ports' capacity for expelling steam, before you've been able to advance the cutoff very far.

The fact that the FEF-3 (and now the Connie too) has a back pressure gauge is a real boon in understanding what's going on in the cylinders, and especially on the exhaust side of the piston strokes.

Thanks for your comment, Mrennie. I admittedly haven't done much with the exhaust limit as I wasn't really sure what that was in relation to the MaxOutput line. Now that you mentioned that it's related to back pressure it makes a lot more sense. Some lingering questions I've had about some of my previous mods are making sense, which means I have even more things to tinker with.

I keep being impressed by how complex modding these steam locomotives really is in comparison to diesels. With diesels you just need to figure out how efficient the transmission is, plug and chug some numbers, and BAM! it's done, but with steamers you have to take into account the size of the cylinders, diameter of the driving wheels, heating surface of the firebox and the superheater (if it even has a superheater), then whether or not it has roller bearings, the boiler's evaporation rate, the maximum drawbar horsepower the locomotive actually achieved in service, speed, etc.

No wonder IHH messed up SO hard with the Big Boy, Challenger, the Clan, the 9F, and so on.

I'll keep at it, though

[JohnS]

That's a lot of info to take in and make sense of. Like you said it's no surprise now to see why more steam isn't made for RW or the ones that were made before SmokeBox behaved so poorly. I think you taking this on is great and will much improve what is already available.

[dtrainBNSF1]

It will no doubt. I just need the TIME to finish one project.

By the way Mrennie, what exactly is the wire-draw effect?

[AmericanSteam]

I found this article: http://chaski.org/homemachinist/viewtopic.php?t=77553

[mrennie]

I found this article: http://chaski.org/homemachinist/viewtopic.php?t=77553

There's a lot of really good information throughout that thread. Incidentally, I read that treatise, the one by Prof. Ludy from Purdue University, several years ago and I highly recommend it. You can find the entire document, in PDF, online (Google search).

Basically, wire-drawing is damage to the valve seats that leads to narrow channels through which steam can still pass when the valve ports are supposed to be shut. That makes the cylinders less efficient and the drop in efficiency becomes more noticeable at higher speeds and shorter cutoffs.

[dtrainBNSF1]

Hey y'all. Been a bit because, well, it's been a bit since I've been able to do anything mod-wise on TS2018.

Here's what happened:
-End of the year stuff at school seemed to happen all at once.
-My grandpa on my father's side in Idaho passed away so my dad and I made the trip from Illinois to Idaho in his pickup
-Finals right after the funeral
-My mom had knee surgery and since by then I was beginning Summer vacation I was the only one with free time during the day to be at home and look after her

So yeah, I've been a bit busy

Good news: Mom's getting around better and I now have some time to mod. I even posted a proposal in the DovetailLive forum asking that the Big Boy and Challenger be completely rebuilt for TS2018 a little while back and made a few refinements to the Dash 9 mod. A while back I found a workbook for Norfolk Southern which gave me a better idea how extended range dynamic brakes work (max dynamic brake power is attained from 25-24mph, then 18, 12, and 6mph before trailing away at 2mph) and setting tractive effort values not just for every 5mph but for every single mile per hour. As a consequence the loco's low-speed performance is much improved. I've gone ahead and done similar improvements to the SD70Ace and MAC and a mod for the CSX ES44AC in the Marketplace is in the works.

K4s is making progress. I just finished a few test runs this morning. The boiler works well but I have to wind the cutoff back a bit more quickly than I like. That could be because I'm running the engine light engine - haven't made any loaded tests yet.

I had a thought this morning while running the K4s around the TestTrack:
Typically a steam locomotive with Walshaerts valve gear hits its stride down the mainline somewhere around 20-25% cutoff. Running down the line at about 5-10% is typical of a locomotive with caprotti valve gear if I recall correctly. So wouldn't it make sense to model a caprotti-valved locomotive with its top speed around the 10% mark (which we don't have in TS2018 yet I think) while a locomotive with Stephenson or Walshaerts valve gear with its top speed around 20%?

[dtrainBNSF1]

I HAD AN EPIPHANY WHEN I WOKE UP THIS MORNING!!!

A month or so ago I posted a proposal on the DovetailLive forum about articulated steam locomotives entitled "What Silly Things These Articulateds..." wherein I outline some flaws in the implementation of articulated steam locomotives in Train Simulator and one of the things I pointed out was that the loco is essentially a double-headed steam locomotive which Train Simulator does not like because it displays conflicting information in the HUDs and can mislead players as to what's actually happening in the locomotive boiler at the time which can lead to unexpected errors ending the scenario such as the boiler running dry when the F4 HUD shows you have +90% of the boiler filled.

I've been busy lately (big surprise, right?), but I've still been thinking about what can be done with the Big Boy and other articulated locomotives in order to fix this particular issue since I'm pretty sure DTG is not going to approve of my proposal to completely rebuild the Challenger and the Big Boy from the ground up as one engine unit instead of 2 with better physics etc. etc. no matter how many people vote that they either strongly agree of simply agree with me. It hit me when I woke up this morning: RWTools has a function that turns any engine unit into a dead unit. I decided to give it a whack with the front engine unit with the Big Boy to see if it fixes anything. So I took a Bob-standard unedited Big Boy and converted its front engine unit into a dead unit and took it for a short spin on Quick Drive after making a light engine consisting of the dead unit up front.

The initial results: the steam pressure gauge not works properly! For the first time ever the steam pressure gauge on the F4 HUD works! Usually unedited the F4 HUD looks like it hovers at 300psi even after you get going down the track at speed for a bit. Right away with the dead front unit the gauge no longer hovers at 300psi, but regularly drops to 295psi every few seconds as the steam-happy boiler produces steam and the safety valve vents the excess steam out of the boiler as it should.

This means that I WAS RIGHT and now we have some solid clues as to how to REALLY fix our articulateds rather than my previous attempts which I consider more band-aids rather than real fixes. Next I'm going to do some loaded test runs and tinker with the input mapper to try and get the injectors working correctly.

By the way, if you haven't already drop by the DovetailLive forum and check out my proposal "What Silly Things These Articulateds..." Give it a read, give it some thought, leave a comment if you want, and cast your vote at the top of the page.

[AmericanSteam]

If you dead head the front engine do you see a loss of power. Can this be corrected in the engine physics by increasing the power in the rear engine?

[dtrainBNSF1]

If you dead head the front engine do you see a loss of power. Can this be corrected in the engine physics by increasing the power in the rear engine?

I noticed this a bit ago when I did a loaded test. Yes this can be corrected by increasing the power of the rear engine to compensate, but you might want to hold off on that.

As of posting my last post, I've gone ahead and converted the front engine into a tank engine because I found in my first test that somehow water was draining from my tender before I could even fill the boiler (easier than I thought it'd be - just copy/paste some lines of code from the rear unit of the Cab Forward since it's already set up as a tank engine), dropped its weight to 0, adjusted fuel amounts, then made MOD versions of the rear unit and the tender, adjusted more weights and fuel amounts, and after this next test I'll start tinkering a bit with the power.

The reason you might want to hold off on adjusting the power for now is that up until now the front engine unit and the rear engine unit have been working off of the same simulation file (for some reason I thought that there were two simulation files. Oh well.) So we really were getting 2 engines for the price of 1. Each engine was set to have a starting tractive effort of 190,000lbf, comparable to a modern AC diesel-electric locomotive, so in reality when driving in Train Simulator with both units unmodified you would essentially have 380,000lbf (no wonder it always seemed over powered!). Just this second I'm trying to balance the power.