[Mike B]

I don't worry about the appearance of the exhaust all that much. Do notice that with some locos the diesel exhaust is in fact discrete puffs - in Real Life those smear out a bit so the exhaust appears mostly continuous unless the exhaust system is straight pipes with one per cylinder (like some Baldwins). And in any case, once the exhaust stream is more than a short distance above the locomotive, it all merges together into a single stream.

MSTS did a fairly decent job of representing the continuous appearance (sometimes pulsating, but not discrete puffs with clear air between) of diesel exhaust. OR is improving - recent builds have been much better than those from back around v.1.0 - but hasn't quite gotten there yet (at least with default settings). In the great scheme of things I'd say it's a good thing to improve, but not at the top of the priority list.

[railguy]

^The discrete "puffs" are a major annoyance in the Diesel Exhaust scheme for diesel locomotives in OR--it is absolutely unprototypical in all but a few unique instances, those unique instances mostly confined to diesel engines other than railroad locomotives. I should know--I've spent much of my life around diesel engines. The reason is the same one that makes a locomotive diesel prime mover run pretty smoothly--multiple cylinders (up to 20 of 'em) firing in sequence. The exhaust appears as a stream for the same reason that a motion picture at 24 frames per second appears to be be continuous motion--the "puffs" are at such a rapid rate that the human eye can't distinguish them as individual events. That's much different than the exhausts from a steam locomotive.

Smoke generation in a diesel engine should be tied to fuel delivery rate and RPM, with the darkest smoke occurring when fuel delivery exceeds the amount that the engine can burn at its RPM rate, and turbo and/or supercharger (Roots blower) boost rate (at that RPM, if the engine has a Roots blower or turbo). Somewhere out there is probably a mathematical algorithm that would predict that in a real world engine.

[Eldorado.Railroad]

ErickC, on 24 October 2017 - 06:23 PM, said:

I can't possibly be the first person to have noticed this.

While I can attest that the analysis is quite interesting as an exhaustive exercise, I am at a loss to see what the desired point is for OR and diesel smoke. Maybe you can clarify what you mean in simple terms.

To distill what I can only suspect from what you have written, yes you are right, a two cycle cylinder from an EDM engine emits exhaust on every revolution of the crankshaft. Likewise a GE 7FDL is a four cycle engine, which means exhaust occurs every second revolution for any given cylinder. Would it be great to have OR match emissions given either a two cycle engine and/or number of cylinders, yes of course it would. But the benefit would have to match the ease of coding and testing such a beast. Likewise, it would be a bit of a stretch to match sound with piston firings. I will be the first to stand up and say BRAVO if that happens!

I think it is easier to do that for a steam engine because of the limited number of cylinders and much lower frequency and complexity of sound pulses, etc. as I am sure you are well aware of.

regards,
Steve

[ErickC]

Eldorado.Railroad, on 25 October 2017 - 09:46 AM, said:

While I can attest that the analysis is quite interesting as an exhaustive exercise, I am at a loss to see what the desired point is for OR and diesel smoke. Maybe you can clarify what you mean in simple terms.

To distill what I can only suspect from what you have written, yes you are right, a two cycle cylinder from an EDM engine emits exhaust on every revolution of the crankshaft. Likewise a GE 7FDL is a four cycle engine, which means exhaust occurs every second revolution for any given cylinder. Would it be great to have OR match emissions given either a two cycle engine and/or number of cylinders, yes of course it would. But the benefit would have to match the ease of coding and testing such a beast. Likewise, it would be a bit of a stretch to match sound with piston firings. I will be the first to stand up and say BRAVO if that happens!

I think it is easier to do that for a steam engine because of the limited number of cylinders and much lower frequency and complexity of sound pulses, etc. as I am sure you are well aware of.

regards,
Steve

I don't think it would necessarily be too difficult. You'd just need to specify how many puffs per revolution, the sim already calculates RPM. The numbers of cylinders is irrelevant, as the number of exhaust pulses is tied to the cycle, not to the number of cylinders. It's a subtle thing, so it's hard for me to explain. I'll try to explain again. There is a single, big pulse of exhaust noise per revolution in a two-stroke engine, and a single, big pulse of exhaust noise every two revolutions in a four-stroke engine. Listen to my clips again. The first clip is an actual recording of an EMD 567, albeit with noise above the exhaust range filtered out. Notice that each time you hear the beat of the engine (the woof-woof-woof noise), it occurs once per revolution. If you were to copy and paste my clip so that it was a minute long, you'd hear roughly 275 pulses because the engine is running at about 275 RPM. The pulses are spaced about .218 seconds apart. But a 16-567 running at 275 RPM doesn't fire a cylinder every .218 seconds. It fires a cylinder every .014 seconds. You cannot hear the individual cylinders. The second clip is what an idling EMD 567 would sound like if you could. Notice that it doesn't sound anything like the real engine does. You don't hear a pulse for each cylinder, you hear a single pulse for every revolution of the engine. If it were a GE, you'd hear a single pulse every other revolution of the engine. I suspect it's related to the engine's ability to exhaust the total flow per cycle.

As to the visible exhaust - I have spent a fair amount of time observing diesel exhaust as well, and I am not kidding when I say that the individual smoke pulses are not only visible, but tied to the pulse of the engine, which is once per cycle, not once per cylinder. Watch this clip. This is typical of what I have observed. At around 1:36, the power is increased. You can see discrete puffs of exhaust with each audible pulse of the engine, increasing in frequency as the engine speed increases. At 2:00, you can see it again, but the rate is much higher. Here's a GE example. Notice that each time you hear the clack-clack-clack of the engine, you can see a discrete puff of smoke from the exhaust stack. You can tell it's not a continuous flow, because if you look very closely at the smoke, you can see definite sharp boundaries between puffs. It's most visible at the very bottom of the stream, before it diffuses out. It's exceptionally visible at 0:56 or so. Look at the dark spots in the stream. Notice how there is a dark spot emitted with each cycle, which diffuses into a more continuous cloud as it ascends. It's a bit harder to tell at this particular engine speed, but notice how the bottom of exhaust stream - where the stream velocity is greatest - looks like it has a regular "lumpy" structure. It is not random, as you would expect from a continuous flow. It has rhythm. Now, listen to the rhythm of the engine, and you will see the individual pulses emanating from the top of the stack every time you hear the engine's "heartbeat," before the stream coalesces into a more diffuse whole as the velocity decreases. Visible here again. Look at the bottom of the stream. Here's a different clip. ALCo set the gold standard for smoke. Super obvious here. you'd have to be blind not to see it at 0:57. Same locomotive, better view. There is a somewhat continuous stream, yes, but notice that it has a definite pulse, and this pulse occurs once per cycle (in the case of the EMD, once per revolution, in the case of the ALCo or GE, once every other revolution). It's not a smooth flow by any means. It's subtle, but it's there, visible in real life, and definitely not an illusion.

You can actually put your hand behind your car's exhaust when it's idling and feel it. You'll feel that it's not like a smooth flow of air, it's bumpy. It's a regular, fast pulse, like a massaging shower head.

To make this work right, though, we'd definitely need a way of controlling exhaust velocity as it exits the stack. It's got to shoot upward, not meander out.

[ErickC]

Hmm, I upgraded from an older test version last night, and, suddenly, my diesel exhaust in engines with OR parameters was pure white, no matter what I did to the parameters. This is interesting, because my personal exhaust texture, as previously uploaded to this thread, uses a white base instead of a black base. So I quickly edited my dieselsmoke.ace, and concluded that OR is not multiplying the colour value by the base colour; it's ignoring the colour value set in the diesel engine parameters entirely, instead using the colour in the RGB channel of dieselsmoke.ace.