The Massive Obvious BUG Thread

[davveb]

i asked him how accurate a program like fsx is, he said there ok for a bit of fun,but not to be taken too seriously,after all why would companys pay millions of pounds for a flight simulator,if you could nip to pc world and buy a dell inspiron and a copy of fsx for a few hundred pounds,so as i said before the big word is compromise,these titles are first and foremost created to entertain and entertain as wide an audience as possible.

The FAA, Federal Aviation Administration not Fleet Air Arm!, approve both Microsoft Flight Sim and X-Plane for certain aspects of real world flying training, so they can't be too bad! I also understand that the more fully featured aircraft addons for MSFS are used by some airline pilots as procedural training aids to prepare for check flights in the pukka multi-million pound simulators. Maybe in time RailWorks will catch up in the accuracy stakes.

Andy L

Yep, my own field is aircraft avionic systems design, and I am constantly amazed at the detail and fidelity of many flight simulation games, for a pitance of a purchase price. In contrast, I used to work at a company (I won't say which one, for obvious reasons) where we had a very expensive full cockpit simulator/rig and the flight model was so poor even the real test pilots struggled to fly it!

[AndiS]

However software companys dont do themselves any favours by putting out statements like "acurate this! and highly detailed that!"maybe some of the hype needs toning down a little.

This is something that hurt me, too. We had established peace here and the fact that KRS is a train game with cool graphics when, in the ramp-up to RailWorks, they started the "most this and that" again.

At the same time, it is the only straw for us physics and signalling weirdos -- as long as they claim such things, they ought to listen to us.

[jkxx74]

At the same time, it is the only straw for us physics and signalling weirdos -- as long as they claim such things, they ought to listen to us.

They better.. Yes, there are relatively few people who care about things working as they should but very often these are the same people who later on develop enhancements to the sim to make it more lifelike and engaging. And it's true that others will seek their opinions on matters of whether this or that title is any good in the realism department.

Personally I'm too uninformed about diesel power to notice the inaccuracies but it seems with almost all locomotives they accelerate WAY too quickly, especially when driven by the AI. We are supposed to be simulating trains, not Concorde planes taking off. To anyone who has as much as watched real trains this looks very unrealistic.

And then there's the matter of signals. AndiS has said just about everything I could say about this myself but in a much more coherent language. Obviously wrong things such as signals staying open behind a train that's passed and the dispatcher and a given signal disagreeing between themselves should be fixed or at least looked at. It's extremely annoying to see two signals placed at the same spot but for opposite directions (on single track mainline) to both be showing 'caution' simultaneously; AIs waiting in front of an open signals; trains in general casually SPADding through closed sections, and the other irregularities. Even if they have to come up with some convoluted solution to this it would still be better at this point than what's currently offered.

That said RSDL has been at it and constantly fixing various bugs so I have some hope. [Rant over]

Cheers,

jkxx

[sniper297]

Hee-hee, well, Growler, gotta agree and disagree in the same post here.

]Hi
Interesting thread,but i must say the original poster,whilst i,m sure, making some valid points has to be realistic!
A commercial simulator costs millions of pounds,and needs updating and improving all the time,you cannot expect total accuracy from a £30 software title, true, RW does have errors, that need to be addressed sooner than later,but the holy grail train sim,your looking for does not exist yet,and probably never will! the best you can hope for is that RW,if you wish to stick with it, will in time, meet your standards.
With a software title like Railworks there will always have to be a tradeoff between real,and arcade,to make it appeal to as wide an audience as possible,AI and Signalling,will not be as important to a casual buyer as it is to hardened train simmers,who like me, cut our teeth on the bug infested MSTS all those years ago,so maybe a slight lowering in expectations is needed,i like to concentrate on what is postive,lives too short to dwell on the negative.

Europeans are always talking about accepting limitations and setting realistic goals, concepts which we Americans are too dumb to understand.



However software companys dont do themselves any favours by putting out statements like "acurate this! and highly detailed that!"maybe some of the hype needs toning down a little.

You said a mouthful there! Dunno what you guys call it, but here it's known as "bait and switch" salesmanship, promise something to get them in the store, then try to sell lower quality at a higher price. Got a new trainsim game here called Kuju Rail Simulator, it's not finished yet but if you buy it, we'll add US content and developer tools so you can create AI traffic later. Hmm, looks good so far! Okay, here's the developer tools. Hey, scenario editor is all screwy, what's the deal? Okay, here's the US version, but it costs extra. Batteries not included, some assembly required, accessories sold separately, that figures. And the dev tools don't work with the US version. Okay, here's patch mark 1, and new dev tools compatible with both versions. Uh, that made it worse, instead of fixing the AI traffic you just locked all the switches including the manual ones! Okay, here's patch mark 2, the dynamic brakes on US locos work now. Uh, yeah, but there's no sound from the dynamic brakes, and when are you gonna fix the AI traffic? Announcing a new game: RAILWORKS! In addition to the features available in Rail Simulator, RailWorks includes but is not limited to:
* The Dispatcher has been improved with added functionality including error reporting for problematic scenarios and further enhanced AI traffic handling. Um, I'm not buying it this time, especially since the reports I'm getting back show the AI traffic is still nothing more than animated background scenery rather than actual "traffic".

[davveb]

You said a mouthful there! Dunno what you guys call it, but here it's known as "bait and switch"

Going in to a scary place there. I think this is known in the UK as part of the generic legalise concept of "invitation to treat". Advertising a product does not constitute an obligation to sell or buy, but if an advertisment is misleading and entices a customer into a more advantageous position to make a sale, then it may be considered fraudulent. I think one example is a misleading advert in the window of a shop, if it entices the customer to enter the shop, but is not actually available inside, the customer has been duped into a more advantageous selling position. I thought of mentioning this regarding the previous promise of a "Steam free" DVD, but that's probably going too far. As long as the current adverts and Web shopping sites don't say this, I'm sure it's all above board.

[paulz6]

* The Dispatcher has been improved with added functionality including error reporting for problematic scenarios and further enhanced AI traffic handling.[/b] Um, I'm not buying it this time, especially since the reports I'm getting back show the AI traffic is still nothing more than animated background scenery rather than actual "traffic".

It's not a false statement. A textual error message is given when confronted with a scenario failed to load, and you can now add scenario specific route markers to help with the AI traffic handling.
It's not the complete re-work though, that some of us were hoping for. It will still '!!!!' if you set traffic that requires a signal hold up to run.

[Klaabu]

Thank You Kromaatikse for opening that thread. Very necessary one. At least somebody is talking about functionality of the railway simulation game.
Somebody mentioned here how dispatching is working in Trainz. But I'd like to call attention to another way to build up traffic in Trainz - manual ... which one I personally prefer to automathic. Manual method is based only on how user is setting junctions. Signalisation is fully dependant on how junctions are set. If You start AI train somewhere and pre-program this train (or loco) to drive to trackmark A then train automatically tries to get there at maximal allowed speed ... but will reach the destination only in case all juntions are set so that it's possible to accomplish the task. Another way train will be stopped by signal elements. Junctions position dictate the overall traffic. This method is not easiest one to use because needs lots of attention when pre-programming junctions state and many scripts and rules have to be used simultaniously ... but at least You can do whatever You want. Possibilities are endless. I have created advanced switching (shunting) scenarios based on real-life Deutsche Bahn (Germany) passanger traffic timetable. All passanger trains are keeping in schedule like in real-life and my shunter is zig-zagging in between there. That can give You real feeling of traffic. However Trainz memory handling problems can set limits how many trains can be envolved into process but usually 25-30 are in use.
What I want to say with that? Such a manual dispatching system based on train tasks, pre-programmed junctions (set by user), triggers (time and location) and variables can give very good results. For developers it's also the easiest way to get functionality working. No need for advanced AI dispatching engine. Just make signalling system to function normally, make junctions to be pre-programmed and add trigger and variable functions. And let the rest for users. That's it. Developers of Rail Sim wanted to create intelligent dispatching system ... but what is achieved is terrible mess. And all this messily created system is not documented at all. You can't find information on how track types and train priorities interact. Also You can't read in manual consist operations are not possible for AI.
We as buyers of RW are like co-investors for the development team. And we have rights to ask for better functionality. I hope topics like this can also open eyes of "toy-boys". So keep going.

[gillespiejr]

Diesel-Electrics

The first, and most relevant at a standstill, is the internal resistance of the motor's windings. At a standstill, all of the power of the engine and generator is lost as heat through this resistance, as described by Ohm's Law.

If that were true then there would be no power available to move the train! Some power is lost as heat in both motors and generator at all speeds.

But it is true. Power doesn't move the train from standstill, in fact power never moves anything. Force moves the train. Refer to the equations of a previous post.

Diesel-Electrics

The equations that apply to a diesel-electric are:

Power = Torque * Speed

Power = Force * Speed
...

So, at stanstill, Speed = 0 giving Power = 0 .
The equations quoted don't just apply to diesel-elctric locomotives, they apply to everything.

[Kromaatikse]

Okay, I did forget to mention field-weakening. This effectively changes the characteristics of the motor, rather than the motors' linkage to the generator, and would require additional pre-calculated tables with the scheme I suggested. The effect as seen by the driver is qualitatively similar.

Looking at the fault manual for the 37 http://locodocs.co.uk/brmanuals/37refur ... de1985.htm it does indeed appear that there is no transition away from the hybrid series-parallel mode possible - the only switches in the traction circuit are for isolation, reversing and field-weakening. That's something I had previously overlooked.

The books I read stated quite clearly that the typical locomotive-type load regulator measured and controlled for fuel flow, as a proxy for power. I suppose that a governor that provided a fixed fuel flow (reacting only to gross over and under speeds as in modern automotive practice), combined with a load regulator that controlled for engine speed, would also be a sensible design - I just haven't found reference to such a design in the literature.

Finally, there is no way that power can be transmitted to the rail while at a standstill, because (as mentioned in the Fundamental Equations) power = torque * speed. Therefore all the power input to the motor must be dissipated as heat. However this does not at all preclude torque being transmitted to the rail, and torque depends mostly on the motor current.

[AndyUK]

The books I read stated quite clearly that the typical locomotive-type load regulator measured and controlled for fuel flow, as a proxy for power. I suppose that a governor that provided a fixed fuel flow (reacting only to gross over and under speeds as in modern automotive practice), combined with a load regulator that controlled for engine speed, would also be a sensible design - I just haven't found reference to such a design in the literature.

The example I quoted was from the book Diesel Traction - Manual for Enginemen that I referred to. Whilst I'm not familiar with electronic governors and electronically controlled fuel injection systems I've never come across a diesel engine control system that measures fuel flow, but that of course doesn't mean they don't exist.

Finally, there is no way that power can be transmitted to the rail while at a standstill, because (as mentioned in the Fundamental Equations) power = torque * speed. Therefore all the power input to the motor must be dissipated as heat. However this does not at all preclude torque being transmitted to the rail, and torque depends mostly on the motor current.

You, and gillespiejr, are of course both correct in correcting me on that. Apologies. Sack cloth and ashes time for me . I've just enrolled on a remedial physics course!

Andy L

[davveb]

This thread has spawned a complimentary thread over on TS.com:

http://forums.flightsim.com/vbts/showth ... p?t=282887

This contains some interesting additional observations from a North American perspective.

[TheTazman]

RW is now slapping me in the face with branches over hanging the lines. I seem to recall that these were removed in RS MK 2 patch but now reappeared.

Simon

[paulz6]

RW is now slapping me in the face with branches over hanging the lines. I seem to recall that these were removed in RS MK 2 patch but now reappeared.

Simon

Is that happening in the default routes or non-default routes? The 3D trees have been re-worked and are now much larger than their predecessors.

[Kromaatikse]

Since it looks like the most serious Coupling problems are being addressed by the devs, I'll leave a discussion of that to a later post, if it is still necessary at all. It's very heartening to see that simple improvements are being made relatively quickly, and taking advantage of the (now appropriately named!) Steam distribution platform.

So instead... Steam Locomotives

Anyone who has driven, or even stood next to, a steam locomotive will agree: it appears to be alive. That is a large part of it's charm. They have many little quirks, and no two locomotives are ever quite the same. Driving or firing one is also a real test of skill, especially if any kind of tight timetable is called for. But to use those skills effectively, the engine must behave realistically, and by far the most effective way to do that is to model the functional parts in quite some detail.

A steam locomotive has three main parts: The fuel storage, the boiler, and the engine. Fortunately, the fuel storage is quite easy to model, so I'll focus on the other two parts, which are far more subtle. The world's best engineers spent 120 years refining Stephenson's design, and subsequent history has shown that they still hadn't got even close to the best possible performance from it - but we now have a pretty good idea of what actually goes on in those tubes, pipes and valves.

Put very simply, the boiler (which includes the firebox, superheater and smokebox, for our purposes) generates power from water and coal (I know, other fuels can be used too), and the engine uses that power to generate force. The links between the two are high-pressure steam as input, and low-pressure steam as output to the blastpipe.

To be fair, I don't see anything astonishingly wrong with the boiler model presently in use - the basic behaviour is okay. But there are still significant improvements that could be made:

- Take account of a pulsed exhaust, whose effects are visible in the brightness of the flames in the firebox, not just in the volume of steam from the chimney. Anecdotal accounts describe a "white flash" corresponding to each beat of the exhaust when pulling away, interrupting the usual dull red glow. When running fast under power, the flashes would gradually merge until there was a relatively steady yellow colour to the fire.

- The uneven firebed that a poor fireman can cause. Holes in the firebed cause cold air jets in the firebox, which reduces efficiency. Thick spots don't burn as efficiently as there isn't enough air coming through, so they would smoke and appear dark, and again provide less heat to the tubes. Some way for players to take more detailed control of the fireman's role would doubtless be appreciated by some users, but I suppose it's something we can live without.

- The poor combustion that results for a short time after coal has been shovelled in, since it takes time to heat up to burning temperature, which would be visible as extra black smoke from the chimney and a redder glow in the box, and also provide less heating for a time. (See below.)

- A sharp exhaust blast combined with very hard working (or slipping!) can actually lift the fuel from the firebed, ripping holes in it and sending unburned fuel out the chimney! The fireman would be giving you a very dirty look (figuratively and literally).

- Air is provided from two places - through the firebed via the dampers, and into the upper firebox via the firehole (or, on larger engines, also by a dedicated vent). These two air sources essentially burn the two different components of coal, and the heated "combustion gases" - for it is no longer air - heat the water via the firebox and tube walls (and thermic syphons if present). Air from the firebox door has very little effect on the coal on the grate.

- Coal doesn't all burn on the grate - only the solid carbon part does. Much of the energy content is in so-called volatiles, which require plenty of extra air and a high firebox temperature to burn. As a rough approximation, it is the volatiles failing to burn that causes black smoke, and it is also the main cause of lower efficiency when the fireman puts too much coal on at once.

- Different types of coal have different proportions of carbon and volatiles, and these different types were typically available in different areas. The GWR used exceptionally good coal, from it's own coal mines in South Wales, until the end of WW2 - and GWR engines always lost a good deal of their performance when forced to burn the more typical coal used by other British railways.

And it would definitely be nice if the AI fireman would try to avoid blowing the safety valves every couple of minutes! But maybe that's just an encouragement to fire manually. Use of the damper and blower (in anything other than the "open" positions) would probably do the trick well enough. Of course the blower is required in tunnels...

A useful technique for the AI fireman to adopt, though, is to maintain the boiler water level according to the boiler pressure - so when full pressure is available, that's a good time to make sure the water is near maximum (which also staves off the safety valves' lifting), while if the pressure is dropping through hard working, it's often better to "use some water" to keep as much pressure as possible for the driver to use, and refill the boiler again when the pressure recovers (or if the level drops dangerously low).

But I have a lot more to say about the Engine.

Straight away, I'm going to avoid a lot of complexity by not talking about Compound engines. Since RailWorks concentrates on British steam, that seems appropriate. Compounding will be much more important to modellers working outside British practice, but we have to start somewhere.

Before we even get to the cylinders, I'm going to talk about pipes and reservoirs.

The boiler, of course, acts as a huge reservoir containing pressurised, saturated steam. Saturated steam has a fixed relationship between temperature, density and pressure, which makes for a very convenient starting point.

The first thing this steam encounters on it's way out of the boiler is a valve, the regulator. A valve is very much like an electrical resistor - the flow through it depends on the pressure difference across it and the resistance the valve has to steam flow, and you can use a repurposed version of Ohm's Law to model it. This resistance can be set up in a simple lookup table provided by the blueprint author, removing any need for strange and incorrect "dual porting" semantics.

Corollary: if there is nothing consuming steam at the other end of the pipe, the pipe will slowly fill up to full boiler pressure even if the regulator is only a crack open. As such, small regulator openings are a great way to get a steam locomotive moving reliably without slipping, especially if the cylinders are already warm and therefore the drain cocks can be closed. (Which may be why some drivers seem to delight in filling the station with steam before moving...)

As an aside, real regulator valves come in two main types: the "pilot valve" type, and the "double seat" type. The pilot valve is used to start equalising the pressure on the two sides of the valve, to make it easier to open the regulator the rest of the way. The double-seat valve is designed to balance the forces naturally, removing the need for a pilot valve. As far as simulation is concerned, the two types are basically the same, just with different resistance-versus-position curves. The only deviation from this practice I've come across is with certain compound engines, which I'm not considering here.

While a valve only has (variable) resistance, a pipe has both resistance and volume. In short, a pipe can be filled with steam (like charge on a capacitor), and it presents a (hopefully small) resistance to steam flowing through it. It's probably okay to consider a pipe to be a monolithic reservoir with a half-resistance at each end.

The main steam pipe leading from the regulator valve tends to be on the large side, therefore with low resistance and relatively high volume. It leads to the superheater, if one is fitted - and virtually all "modern" designs do have one.

The superheater's effect is as another pipe, but with the special effect of transferring even more heat from the combustion gases into the steam inside it. This means that, in all probability, the steam stops being saturated, and gains a higher temperature and lower density than normal for it's pressure. Because it's still connected to the main steam pipe and, via the regulator, to the boiler, it is unlikely that the pressure will go up significantly, but the extra energy in the downstream steam has other benefits. Obviously the effectiveness of the superheater depends on how big it is, and how hot the combustion gases are in that part of the tubes.

The presence of the superheater means that for accuracy, rather than attempting to track pressure, temperature and density directly, you must consider the steam as having a mass for quantity and an enthalpy for energy. Both of these quantities can be considered as fluids that flow in the same way down the pipes and are affected to the same extent by valves, except that enthalpy can also be conducted through walls. The density and pressure are calculated from the volume of the container and the specific enthalpy, that is, the enthalpy per mass unit, and formulae to do so can be found in the literature. The relative pressures thus calculated govern flow between any two elements in the steam circuit.

Either before or after the superheater, there is a "snifting" or "anti-vacuum" valve which lets air into the pipe but not steam out. This basically ensures that the pipe never goes below atmospheric pressure. While we're on the subject, it's much easier to keep track of absolute pressure, where the atmosphere is taken as 1013 mbar at sea level (and varies with weather and altitude), rather than constantly correcting for gauge pressure.

Another pipe leads from the superheater to the steam chest, which is a reservoir immediately on the live side of the valve gear. This pipe has been the subject of significant improvements in performance through reducing it's resistance, so it is important. In some locomotive types, the steam chest may be of minimal size, in which case this pipe leading to it effectively does the same job.

The valve gear is effectively the heart of the engine. It's job is to regulate the flow of steam into and out of each end of each power cylinder. In all but the very earliest machines, there is a way to change the time the inlet ports remain open during each stroke - this is effectively done by altering the travel and phasing of the valve, for the common slide and piston valve gears. A very few engines - albeit sometimes very important ones due to their superior performance - have poppet valves operated by cam arrangements.

While slide/piston valves and poppet valves operate in very different ways, effectively we can build a two-dimensional table giving the resistance of each flow path (from input to each end, and from each end to output) for each reverser setting and wheel position. This copes perfectly well with every practical valve gear yet encountered (except turbines, which are a whole other kettle of fish but very rare indeed).

The power cylinders themselves are effectively pairs of reservoirs with variable volume, which depends on the crank position. Each half has a minimum volume, and the cylinder as a whole has a swept volume, which is added to the minimum volumes to give the maximum volumes. Once you've calculated the pressure in each half, the difference between them translates to a force on the piston rod, which itself produces a torque on the driving wheels coupled to it, depending on the crank position.

The drain cocks are just small valves which are attached to each end of each cylinder, and have a fairly short but rather thin pipe between them and the atmosphere. The pressures in the exit pipes can be used to drive steam animations.

But note that, of all the places in the steam circuit, the cylinders are the most likely for condensation to occur - the phase change from steam to water, caused by the steam passing down through the saturation line. That's why the drain cocks are there, to stop the condensation collecting so much that it fills the cylinders' minimum volumes. The expansion itself can cause enough cooling for condensation, but heat loss to the outside doesn't help either. This heat loss is significant and should be available as a parameter in the blueprint.

The main outputs of the valve gear are into the blastpipe, which has been another substantial area of optimisation for locomotive engineers. The same mass of steam must traverse this pipe as came in through the steam chest, but at vastly greater volumes because the pressure and density are much lower. Thus the flow in this pipe tends to be very fast, and the resistance of the pipe is very significant to performance and efficiency.

At the top end of the blastpipe, which is in the smokebox, the exhaust steam interacts with the (now fairly cool) combustion gases from the fire (remember that?). The type and dimensions of the chimney arrangement determine how effective the coupling between the steam exhaust and the fire exhaust is, and how much backpressure is exerted on the end of the blastpipe. In the 1950s, the Kylchap exhaust was just beginning to become available, and has revolutionised efficiency for the relatively small number of locomotives (including many of the Gresley A4s) that it was fitted to. Most locomotives have plain single or double blastpipes, which merely suffice. Clearly, another table relating blastpipe velocity to backpressure and smokebox vacuum induction is necessary, since this was another critical factor in performance, as well as how easily the blast could "lift the fire" (remember I mentioned that too?).

Finally, the smokebox vacuum is related back to the firebox and the firebed and the firehole and the dampers, by, you guessed it, more pipes. The firebox is a reservoir for gases. The tubes aggregate as a pipe with a defined resistance. The firehole is a valve, the firebed is rather like the superheater, and the ashpan and damper make another reservoir and valve.

It's all rather simple and elegant when you think of it that way.

One last thing. The shape of the boiler had a distinct effect on whether the steam would be lifted up and away, or whether it would stick around and block the driver's view. The "blinkers" on the sides of many later locomotives were usually quite effective at deflecting exhaust upwards, as was the unique streamlining on later Gresley designs, but the boxy "air smoothing" on the Bulleid Pacifics had precisely the opposite effect (having been designed to allow easy cleaning, not aerodynamics). That's something to think about.

[GavNormandale]

hi, i understand all this discussion is needed but i feel it has taken over this bug thread leaving it hard to find if this has been mentioned already but i feel the class 37 brakes are far too good, had a run on bath to templecombe and found going down a 1 in 50 slope, 30-31% brought the train to a stop so quickly i thought id hit the emergency brakes, is this happening for everyone?

Gav