"chugging"

[dieselsmoke]

Hi there

I hope I can find someone who knows EMD engines by heart, but of course any input or ideas are welcome, so thanks in advance

Here is the thing. I have noticed something on roots blown engines, both 567 and 645.

Some of them have "even" exhaust notes, so you can actually hear all cylinders firing, but some of them, and this is not a rare thing, have
"dominant" cylinders and make a serious chugging sound.

I've heard this problem on both 567 and 645 engines.

Here is how a "healthy" one sounds:

http://free-os.t-com.hr/redmist/2062kretanje.mp3


and here is one that "chugs" :

http://www.yardoffice.com/RR/DieselSoun ... Sample.mp3

both are 16-cylinder EMD's


someone toled me that the problem is in the uneven adjustment of injectors/pumps during injector replacement , because some people adjust it "bye ear".

But I would like to hear more opinion.
What do you think makes some EMD engines sound like that, is it something that happens inside them after some time, or is it a mis-tune and mechanics fault, or something else


thanks a lot

[pennsy]

Hi There,

Not much of a difference, so I will agree with the parameters you mentioned and add one more that could make the difference. That would be either different fuel or a different supplier of the fuel.

[dieselsmoke]

Well the "difference" is a little subjective, but there is an objective part to it too. If you slow down the second recording (the chugging one), you can hear that every 8th beat there is a much louder explosion that overwhelms the other 7.

On the first recording, you can clearly hear the repeating 8-beat pattern with no beats sticking out too much.

Can you explain how could fuel cause a cylinder to fire differently (the sound suggests its the same cylinder because the loud beats are strictly periodic)

[Typewriters]

I can tell you right now that you CANNOT hear every cylinder firing/exhausting. It's way too fast for us humans to hear and process.

Think about it -- if the engine were rotating, say, at just one revolution per second or 60 RPM then for a 16-cylinder engine there would be 16 firing sounds PER SECOND. Two-stroke engine, one firing event for every cylinder on every crank revolution. Double that, to 120 RPM; now you'd have to hear 32 events PER SECOND. We're not anywhere near idle speed yet for an EMD 567 or 645 either. 300 RPM? I get 80 exhausts per second. 800 RPM? I get 213 exhausts per second. So, you see - it's impossible! Now, I don't know how much you're slowing down the recordings...

What you have to think of then is that what we hear is a combination of sounds, and that there are resonant harmonics to consider. Resonance, ie in the exhaust passages/manifolds themselves. This differs quite a bit between 16-cylinder and 12-cylinder engines. Find a video that has actual sound, and compare say a GP-9 and an SW-7 or SW-9 or SW-1200 and you'll see what I mean. Now think of the varied exhaust setups fitted to a given engine (shape and size, number of actual exhausts) and you will begin to see that you can't just determine that a different exhaust sound indicates a problem per se in this context. (For instance -- one E-8 and two SW-7's both contain two 1200 HP 12-cylinder 567's but the exhausts don't sound the same!)

Looking at the sound links, I see that track 2 you linked is a four-stack rebuilt GP-7 on C&NW; I don't know what the other one is, or any details since I can't get to where it's from. The four-stack rebuilt 567's do have a different sound; maybe that's what you're hearing and picking up on.

-Will Davis

[dieselsmoke]

I can tell you right now that you CANNOT hear every cylinder firing/exhausting. It's way too fast for us humans to hear and process.

I said if you slow down the recording, which is what I did, I slowed it down about 10 times. The quality is low, but you can hear individual beats. They are still fast, but if you listen to them a couple of times, you can write them down and count them.

Though I think you might be right about one part. I hear a 8-beat repeating pattern, whereas it should be a 16-beat repeating pattern.
I'm guessing 2 beats get mixed together somehow. But still, if that is so, a pair is standing out.

I used a different recording to check if the 8-beat patterns indeed represents a full revolution of the shaft, and I think it does, because measuring how many patterns I hear in one second I got about 480 rpm, which is about right for the notch the engine was in on the recording (it was a third recording, not these two) . This would mean that indeed you can't hear all 16 beats but for some reason two beats get mixed together.


It's a little difficult on 16-cylinder engines, but it's much easier on straight6 there you can hear the individual beats at idle at maybe 25% speed of the recording.


Anyway, sorry for not saying anything about the source of the first recording. It's a 16-645E also with 4 stacks (export model), same kind of manifolding as the GP7 has

[MEC407]

Just to add some fuel to the fire:

A few years ago when I was railfanning on the St. Lawrence & Atlantic, I noticed that their GP11s sounded very gruff and had a lot of "bark" in the exhaust sound, compared to their GP9s which sounded smoother and more mellow. The GP11s had four exhaust stacks and the GP9s had two, so I just assumed that was the reason for the difference in sound, since both units had 16-567s under the hood.

[trainiac]

Regarding the "8-beat" rhythm of the EMD units, I've noticed a similar thing with 4-cycle engines. MLW 12-cylinder engines have three beats every 2 revolutions (or for every main "chug") while 16-cylinders have 4 beats per chug. I guess it depends on the firing order... I'm wondering if the 8 beats per revolution on the EMDs has something to do with the fact that there are 2 banks of 8 cylinders.

A few years ago when I was railfanning on the St. Lawrence & Atlantic, I noticed that their GP11s sounded very gruff and had a lot of "bark" in the exhaust sound, compared to their GP9s which sounded smoother and more mellow. The GP11s had four exhaust stacks and the GP9s had two, so I just assumed that was the reason for the difference in sound, since both units had 16-567s under the hood.

I heard the GP11's all the time! They were the loudest units on the SLR and my absolute FAVOURITE in terms of engine sound. (They were also old and apparently weak pullers, but I was looking at them as a railfan... ) I think the four exhaust stacks played a major role in the chugging sound, in their case--like a car with a free-flow muffler.

The freshly rebuilt LLPX GP38-2's used by the SLR at the same time (2001) were the smoothest, most consistent and healthiest-sounding units I have ever heard. They had a pronounced tonic-dominant sound (120 Hz--8 beats per rev--and 90 Hz) that I haven't heard as much in other GP38-2's. (I might add that turbocharged EMDs often have three very strong frequencies: 120 Hz, 135 Hz and 150 Hz)

Among identical engines, I think the difference in sound comes down to age and maintenance.

A few SLR recordings thrown in for good measure:

Sets of 4 LLPX GP38-2's
http://trainiax.net/recordings/crec-slr ... px2246.wav
http://trainiax.net/recordings/crec-slr ... px2246.wav

A pair chugging GP11's (the only time I caught both SLR units together)
http://trainiax.net/recordings/crec-slr ... 8-8748.wav

The best chugging I have on tape... A GP11 third among four units, overpowering all other sounds.
http://trainiax.net/recordings/crec-slr ... 9-3512.wav

[dieselsmoke]

Regarding the "8-beat" rhythm of the EMD units, I've noticed a similar thing with 4-cycle engines. MLW 12-cylinder engines have three beats every 2 revolutions (or for every main "chug") while 16-cylinders have 4 beats per chug. I guess it depends on the firing order... I'm wondering if the 8 beats per revolution on the EMDs has something to do with the fact that there are 2 banks of 8 cylinders.

Firing order doesn't change the fact that all cylinders have to fire eventually. So, except for engine types where two cylinders fire at the same time, you have to have the number of beats equal to the number of cylinders. The number of cycles determines if you are going to have that pattern played over one or two revolutions.

I'm confused about the sound of EMD 16-cyl because I have used the same methods on 4-cycle straight-6 engines and the math adds up, you can hear a 6-beat repeating pattern over 2 revolutions.

Using the same method, an EMD should show a 16-beat pattern.

On EMD 16-cyl engines, every cylinder fires alone, so you have an explosion at ever 22 1/2 degrees of the crankshaft rotation, as is described in the EMD engine manual.

It is possible however, that exhaust notes get mixed in pairs because the speed of sound.
For example, in order for sound in air to travel one meter (which is sort of close to distance between ever other cylinder though I'm talking of the top of my head) is about 0,003 seconds, and the firing interval at say 500 rpm is 0,0078 seconds. These are similar values, at least of the same order
So I guess by the time the last cylinder in the row sends the pulse through the pipe, it gets mixed with the first cylinder's pulse.

I wish I could get a chance to record the exhaust in idle, maybe it would reveal all 16 beats because there is more difference between the speed of sound and speed of firing.

[trainiac]

This is getting really interesting!

Here's the sound of an idling GP40:

http://trainiax.net/recordings/crec-slr ... 8-3004.wav

I analysed the sound with a sound editing program, and the idle speed is just under 300 rpm. In the frequency spectrum, there is a huge peak at 80 Hz, which corresponds to 16 beats per revolution.

Here's another recording of 5 low-revving GP40s:

http://trainiax.net/recordings/crec-slr ... 7-3203.wav

-At the beginning, the units are in run-1 and there is a peak at 78 Hz, or 16 beats per rev.
-Around the first whistle blast, the units are in run-2 and the peak is at 98 Hz, again 16 beats per rev.
-Around the second and third whistle blasts, the units are in run-3 (peak at 67 Hz) and run-4 (peak at 74 Hz) which correspond to 8 beats per rev.

I've noticed that with run-8 units (particularly turbocharged EMDs) the "chugging" of the engine affects the frequencies to a great extent. 900 rpm translates to a frequency of 15 Hz, and run-8 EMDs have a whole bunch of frequencies all separated by 15 Hz, with the three most dominant being 120 Hz (8 beats per rev) 135 Hz and 150 Hz.

Here's a recording of 5 run-8 GP40s:
http://trainiax.net/recordings/crec-slr ... 1-3004.wav

Here's part of the frequency spectrum taken as they approach:

[Jtgshu]

Even though I don't really have a clue about much of what is being discussed here, im learning and its very interesting!!!

Ive noticed that a GP40 in Notch 2 is the "loudest" - at Notch 8 its loud, but it doesn't go "through you" like it does at Notch 2. In the cab, at Notch 2, you just can't hear anything else except the engine. At Notch 8, the engine is loud, but you can hear other things. Anyone following me here???

The only time ive noticed that you can hear each cylinder firing is when the engine is starting up - its pretty quick (about 2 seconds), but you can count each cylinder firing (so long as the engine is in good condition and not misfiring, etc)

[dieselsmoke]

trainiac , that's really interesting. I see your methods are a little more sophisticated than mine. I used finger counting and slow-mo listening, but the results are the same

It's difficult to keep track of beats in turbo units, because they are naturally muffled by the turbo. Roots blown engines have sharp bursts coming from the exhaust, which makes keeping track of beats a lot easier.

Tomorrow I'll go to the station and see if I can record a louder one (every one of them seems to be unique in sound) in idle.

I think, what you have said confirms that at mid to high runs, the beats get paired up.


Just today I was walking on street, and I saw an old bus, which made a really strong chugging sound, and it wasn't a manifold leak, but the engine was really vibrating and you could hear low booms from the exhaust hole on every chug beat. Must be the same problem as with some EMD's.

[trainiac]

trainiac , that's really interesting. I see your methods are a little more sophisticated than mine. I used finger counting and slow-mo listening, but the results are the same Very Happy

This is something I've really gotten into during the last year or so--ever since I noticed that I couldn't pick out any single dominant frequency in run-8 turbocharged EMDs just by listening.

Ive noticed that a GP40 in Notch 2 is the "loudest" - at Notch 8 its loud, but it doesn't go "through you" like it does at Notch 2. In the cab, at Notch 2, you just can't hear anything else except the engine. At Notch 8, the engine is loud, but you can hear other things. Anyone following me here??? Smile

I've noticed the same thing with GP40's in run-6--the drone goes through you even from trackside! And, after doing a little fiddling, I've found out that the dominant frequency in run-2 is around 100 Hz and the dominant frequency in run-6 is... around 100 Hz. I think the engine is entering in resonance with the chassis, which may explain why it varies between 8 and 16 beats per rev. That would also mean that it probably wouldn't be the same in an SD40-2 with a longer frame.

[dieselsmoke]

trainiac, here is something to play around with:

http://free-os.t-com.hr/redmist/ler.mp3

This is idle recorded from a little wooden bridge about 2 meters above the exhaust.


It's a 12-645 turbocharged engine

[ExEMDLOCOTester]

trainiac , that's really interesting. I see your methods are a little more sophisticated than mine. I used finger counting and slow-mo listening, but the results are the same Very Happy

This is something I've really gotten into during the last year or so--ever since I noticed that I couldn't pick out any single dominant frequency in run-8 turbocharged EMDs just by listening.

Ive noticed that a GP40 in Notch 2 is the "loudest" - at Notch 8 its loud, but it doesn't go "through you" like it does at Notch 2. In the cab, at Notch 2, you just can't hear anything else except the engine. At Notch 8, the engine is loud, but you can hear other things. Anyone following me here??? Smile

I've noticed the same thing with GP40's in run-6--the drone goes through you even from trackside! And, after doing a little fiddling, I've found out that the dominant frequency in run-2 is around 100 Hz and the dominant frequency in run-6 is... around 100 Hz. I think the engine is entering in resonance with the chassis, which may explain why it varies between 8 and 16 beats per rev. That would also mean that it probably wouldn't be the same in an SD40-2 with a longer frame.

Run 6 is where the Turbo begins to freewheel

[ExEMDLOCOTester]

When all the data from the above threads was recorded, what was the status of the air compressor? When the air compressor loads, a very distinct chug is created..... The Air Compressor just happens to be at the same rotational velocity as the engine crankshaft....