Rather ancient carriages I have made, looking for critiques

[AndiS]

The rivet count seems right, but the two on the left definitely have some display issue.
They show as black circles while the others are fine.

[TrabantDeLuxe]

I had hoped noone would notice.

They're too small as well. I'm greatly annoyed now.

[JamesLit]

I had hoped noone would notice.

They're too small as well. I'm greatly annoyed now.

Agreed on both counts. Both things are noticeable but only marginal. Fixing them will be easy and afterwards it will look perfect. Win win.

[AndiS]

I find it hard to judge. Taking a second look, the rods that support the roof seem wider on the photo, but my assumption is that this is caused by the shine on them in the model.

The top row of rivets on the photo looks surprisingly close to the margin while the margins are equal in the model. I would call the photo wrong. This is not as much of a joke as it seems because photos saw lots of retouching in the old days, as part of the artistic/artisan process.

Rivet counting is dead. Long live rivet measuring!

[TrabantDeLuxe]

Nah, there are some issues with the exact shape. I've got to be 100% fair here and say that there's more of these issues with the model. The boiler for instance, is a good inch too wide. The entire model is, last time I checked, a couple of inches short. However, this being intended as a freeware model, I genuinely ask myself whether I'm obliged to anyone to go and correct these issues. And besides, source material is difficult to obtain, and sometimes contradictory. It's the painting versus technical drawing story.

I'm 99% sure that we're looking at an unretouched photograph - I'd love to find out where it has been taken. 1900's photoshop is often quite obvious. the placement of the rivets make sense to me, the wooden boards that form the cab roof are supported below and above by L-profile strips.

[cjbarnes5294]

Nah, there are some issues with the exact shape. I've got to be 100% fair here and say that there's more of these issues with the model. The boiler for instance, is a good inch too wide. The entire model is, last time I checked, a couple of inches short. However, this being intended as a freeware model, I genuinely ask myself whether I'm obliged to anyone to go and correct these issues. And besides, source material is difficult to obtain, and sometimes contradictory. It's the painting versus technical drawing story.

I like this paragraph as I very much agree with the sentiment that you should not be obliged to provide perfection when you are doing something off your own back and for free, hell you're not even obliged to do a good job at all - nobody pays you for your time and effort in providing something for free afterall. I think it is a little bit different for payware - the usual going rate of £12 or the equivalent in euros and US$ is not an amount to be sniffed at and customers should expect to get something that fits the product description and is a pretty faithful representation of the real thing, but perfection? I consider myself a bit of a perfectionist, which is why I sometimes take a bit long to do something well, but I also know that asking for perfection is a futile objective and unreasonable to expect as a customer. The going rate is not to be sniffed at but at the same time it's not an awful lot either when you consider the amount of time that was spent into making the damn thing in the first place.

As for your beautiful loco, well, I think if you hadn't told us all about the model discrepancies none of us would ever have realised! Keep up the inspirational work, it's always a treat coming back to this thread.

Kind regards,
Chris

[malkymackay]

I'll concur with Chris. I'm sure that there are many here that more than appreciate the effort that goes into producing a quality model/route etc, be it free or for financial reward. The wider public may not have the same understanding. You only have to look at the number of posts on forums or social media that continually ask if something is ready/ why isn't it finished yet?

[TrabantDeLuxe]

Okay, I'm doing the steam simulation blueprint. Anyone know how to get the blueprint editor to work in SI units, and if not, explain to me how MaxForce in pound-per-foot is a unit of force? I have a bachelor of science and this stuff makes 0 sense to me ... Maybe I should open a new topic somewhere.

Unrelated note, anyone got an empirical method of estimating boiler and grate performance? I'm going through some issues of the American Engineer and Der Lokomotive, maybe I'll dig something up there .

[AndiS]

Warning: You enter Dark Area here! Many a man has lost his mind, including the writer of these lines.

That said, you could search forums or wait for helpful people to answer where and take that as a starting point of own investigations. I even believe that some of these things did change over time.

I only ever really explored MaxForce. It controls the acceleration of AI trains precisely, together with the total train mass, and you can measure that using a signal object. However, Kariban states that it has no effect on the player train, which sounds very plausible given all the CSV files that want to have a say, too. However, I would not transfer things that are said about Diesels (like there) to steamers without checking.
Source: http://forums.uktrainsim.com/viewtopic. ... 1#p1458631

I found the unit of MaxForce to be klbf at some point in 2011 or so.
1 klbf = 1000 * 0.373 * 9.81 N = 3.66 KN. (Or so I guess.)
Now the sad thing is that I cannot remember all about to change to and from KN. I only remember reporting to RSDL which of their engines used which unit in their blueprints.

Assuming that Kariban is right (and he always was when I could check facts myself), than it only concerns AI which explains why it was low on the priority list for fixes.

When I read the Developer Documentation in 2007, it became obvious that its writer did not have close contact with the creator of the physics logic, to put it mildly. Things like the funny part on air drag are left unexplained through all the years.

I assume that most engine scripters bypass most of the built-in physics by computing tractive effort themselves and setting that via the script. I am not sure that can be done, but something like that will take place. I always stuck with signals as far as scripting is concerned and I gave up deciphering the built-in physics when this ugly bug showed that made wheel slip depend on the initial consist configuration or something.


For the boiler performance, I computed the steam producing surface and multiplied that by some figure for steam production per area that I found suitable - back in the days when I played with MSTS physics.

Of course, this fails for superheating. So leaving that aside, you count heating pipes, take their length in the boiler and their diametre (don't ask which diametre) and multiply that and Pi. Then you estimate the water-covered surface of the firebox and add that. Some sources add some factor for the firebox as it is more efficient in heating the water, but I would leave that aside.

In post WW II Germany the standard maximum for hourly steam production per square metre was 57 kg/m²h, with 70 kg/m²h for engines reconstructed in the 60ies in the East.

A source form 1898 gives these average values:
Mountain engine 34 – 39 kg/m²h
Freight train engine 39 – 47 kg/m²h
Passenger train engine 54 – 58 kg/m²h
Express train engine 58 – 65 kg/m²h

For earlier engines, even less should be assumed as this figure defines the material strain and material was not as enduring in the old days. Of course, all these are maximum figures. Also draft management was not as good back then which reduces specific steam production, too.

Of course, this is a bit simpler than what you find in period journals. If you want a translation from Die Locomotie, just give me the link.


The nice thing is that you don't need to worry about the coal quality at all. The simulation just talks about steam produced and coal consumed, putting a black box around the energy that hides in one kg coal and how that changes the water temperature.

One value for "good" burning capacity before WW II is 450 kg/m²h. The 1898 source quotes 300 – 500, average 400 kg/m²h. So all you need is estimate the grate area with a number in this range.

Now if you ask me where to enter that in the blueprint, I must pass. Despite wild will in 2006 to become a steam guru of Rail Simulator once it was released, the state in which it was released and all the other twists of fate left me without a single engine that was tuned by me. I wonder if that will change for the UE4 incarnation, without much anticipation either way.

[TrabantDeLuxe]

Just a quick reply now, I've taken to playing with particle emitters because of reasons.

I only ever really explored MaxForce. It controls the acceleration of AI trains precisely, together with the total train mass, and you can measure that using a signal object. However, Kariban states that it has no effect on the player train, which sounds very plausible given all the CSV files that want to have a say, too. However, I would not transfer things that are said about Diesels (like there) to steamers without checking.
Source: http://forums.uktrainsim.com/viewtopic. ... 1#p1458631

I found the unit of MaxForce to be klbf at some point in 2011 or so.
1 klbf = 1000 * 0.373 * 9.81 N = 3.66 KN. (Or so I guess.)
Now the sad thing is that I cannot remember all about to change to and from KN. I only remember reporting to RSDL which of their engines used which unit in their blueprints.

I have found the same through experimentation. Perhaps RSC's way of writing klbf is lb/f. I don't know. Values entered in lbf, in the 10e4 range, gave me a rocket. Values entered in klbf seem to give acceptable accelerations. I would really hope the Staats had considered hanging a dynanometer car behind their locos, but alas, they apparently haven't.

Assuming that Kariban is right (and he always was when I could check facts myself), than it only concerns AI which explains why it was low on the priority list for fixes.

For steam at least, MaxForce does impact player locos. I don't know about MaxPower.

When I read the Developer Documentation in 2007, it became obvious that its writer did not have close contact with the creator of the physics logic, to put it mildly. Things like the funny part on air drag are left unexplained through all the years.

The air drag thing makes me wonder if there even is a creator of physics logic. But then I should really shut up...

I assume that most engine scripters bypass most of the built-in physics by computing tractive effort themselves and setting that via the script. I am not sure that can be done, but something like that will take place. I always stuck with signals as far as scripting is concerned and I gave up deciphering the built-in physics when this ugly bug showed that made wheel slip depend on the initial consist configuration or something.

Custom wheelslip is, on the loco's that I have that feature it, seemingly done by calculating when slip should occur, and then slamming the regulator at 0%, so that no tractive effort is applied. Lua wheel animations do the rest. I'm really tempted at having a go at this...

For the boiler performance, I computed the steam producing surface and multiplied that by some figure for steam production per area that I found suitable - back in the days when I played with MSTS physics.

Of course, this fails for superheating. So leaving that aside, you count heating pipes, take their length in the boiler and their diametre (don't ask which diametre) and multiply that and Pi. Then you estimate the water-covered surface of the firebox and add that. Some sources add some factor for the firebox as it is more efficient in heating the water, but I would leave that aside.

In post WW II Germany the standard maximum for hourly steam production per square metre was 57 kg/m²h, with 70 kg/m²h for engines reconstructed in the 60ies in the East.

A source form 1898 gives these average values:
Mountain engine 34 – 39 kg/m²h
Freight train engine 39 – 47 kg/m²h
Passenger train engine 54 – 58 kg/m²h
Express train engine 58 – 65 kg/m²h

For earlier engines, even less should be assumed as this figure defines the material strain and material was not as enduring in the old days. Of course, all these are maximum figures. Also draft management was not as good back then which reduces specific steam production, too.

Of course, this is a bit simpler than what you find in period journals. If you want a translation from Die Locomotie, just give me the link.


The nice thing is that you don't need to worry about the coal quality at all. The simulation just talks about steam produced and coal consumed, putting a black box around the energy that hides in one kg coal and how that changes the water temperature.

One value for "good" burning capacity before WW II is 450 kg/m²h. The 1898 source quotes 300 – 500, average 400 kg/m²h. So all you need is estimate the grate area with a number in this range.

Now if you ask me where to enter that in the blueprint, I must pass. Despite wild will in 2006 to become a steam guru of Rail Simulator once it was released, the state in which it was released and all the other twists of fate left me without a single engine that was tuned by me. I wonder if that will change for the UE4 incarnation, without much anticipation either way.

Those are some ballpark numbers I can work with for now. A heated surface vs evaporation rate constant seams logical.

The main issue is indeed what value in the blueprint does what, but I think experimentation is the best way to go forward. Being Dutch myself, reading German is not a real issue. It's just a funny archaic version of Dutch anyway. I do hope TS:UE will feature better steam simulation features - or at least, more consistent SI-based physics, but I'm not naïve. Really.