Trains.com

ACS-64 technical description

12837 views
42 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 11,919 posts
ACS-64 technical description
Posted by blue streak 1 on Tuesday, May 14, 2013 4:45 AM

Railway age has a good description of the inner workings of the ACS-64 at the bottom of this link.

http://www.railwayage.com/index.php/mechanical/locomotives/siemens-unveils-first-amtrak-cities-sprinters-locomotive.html?channel=35

One item not covered is how much additional weight does the dual frequency equipment carry.?  We know that a 25 Hz / 60 Hz transformer is heavier than a stand alone 60 Hz transformer weighs   The regenerative braking equipment to provide the 25 / 60 Hz inverters would be very interesting.  How fast the motor can go from power to braking is another question ?  Since european frequecy is 50 Hz one has to wonder how much engineering had to be done to make this change ?  Anyone have any information ?    

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, May 14, 2013 7:27 AM

Erikem is admirably suited to answer this in full detail.

It is a bit difficult to extract the sense from what I think is translated German, but it would appear that the setup is heavier because there are separate transformer windings for each of the three powers, presumably stepping down to a standard (unspecified) voltage going to the 'cubicles'.  It woud have been interesting to see a block or circuit diagram showing more precisely how Siemens designed this.

Just what an 'input inverter' that feeds a 'DC link' might be is a mystery.  I suspect they meant 'rectifier' in this context (see here for an example feeding a motor directly) ... but got lost in translation without a proofreader who understood the technology.

Regenerative frequency 'synthesis' is relatively trivial because the frequency 'standard' and any waveform distortion can be read directly off the AC line.  Power-factor and other corrections would then have to be derived to get correct synchronization.  I would expect this to be tracked more or less continuously, and to establish sync within a couple of cycles at most as regenerative power is brought up, so engagement of the brake is essentially 'instantaneous' (modulating it up slowly enough to prevent wheelslide taking longer).  You would not just slam over from full power to full braking because there will be considerable magnetic energy stored up in the inductive components.  However, I would not expect the transition from full power to the initiation of physical regeneration to take very long, and of course the 'excitation' to produce effective dynamic braking can then be made as quick as the rail-wheel interface can handle.

  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 11,919 posts
Posted by blue streak 1 on Tuesday, May 14, 2013 12:56 PM

The following quotation from RA acceleration schedule seems to be overly optomistic ?

""The ACS-64 offers high efficiency, reduced life cycle cost, and better reliability and availability than Amtrak’s existing electrics. In terms of traction capability with the maximum specified trainload, the ACS-64 can accelerate 18 Amfleet coaches with a Head End Power (HEP) load of 1,000kW to 125 mph in just over eight minutes.""

If someone has a tractive effort curve maybe they can figure out how far the train would have to go to achieve 125 MPH.   Do not know what the drawbar pull of 18 caras is to limit the breakaway TE ?.  The acceleration of a standard Regional train of 10 cars would of course be less. 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, May 14, 2013 1:58 PM

blue streak 1
If someone has a tractive effort curve maybe they can figure out how far the train would have to go to achieve 125 MPH.   Do not know what the drawbar pull of 18 caras is to limit the breakaway TE ?. 

I suspect the curve would act as an extrapolation of the one for the AEM7 ... very slow initially, then faster in the middle speed ranges, then lower again with HP limitation at higher speed and resistance.  I would exoect higher effective low-speed acceleration rate from a modern AC drive, however.

Call the speed 185 ft/s; the average acceleration is then about 21.9 ft/s/m (for the eight minutes) or about 0.38 ft/s/s -- well within the 'comfort' zone for acceleration.  (This is about 1/4 of what I recall was the Metroliner peak acceleration rate of about 1.5 ft/s/s).  S (in feet) = 0.5 a t^2, so 0.5 x .38 x 230400 (that's the eight minutes, expressed in seconds, squared) gives you about 8.29 miles.

Somebody here can probably estimate what an 'average' loaded 18-car consist of modern Amtrak equipment is.  That will give you a figure to calculate the train resistance, if you must.  I doubt a modern AC drive lcomotive will have any problem developing the required "drawbar" TE for any speed below 125 mph, if that's the question; the AC drive will perform its effective function of traction control essentially just as well via Cardan shaft as it would with truck-mounted motors.  And lateral acceleration at the railhead, of course, will be vastly decreased -- the French have known this since the early days of monomoteur bogies, and the high-speed testing in the Fifties.

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 2:42 PM

I backed into those performance numbers.  Here's what it looks like:

after 1/2 mile:  65 mph, 1:05

after 1 mile: 80 mph 1:40

after 2 miles 95 mph 2:30

after 4 miles 115 mph 4:50

I assumed the loco weighs 250,000# and has max adhesion of 40%.  I used the std Davis equation but adjusted for aerodynamics until I backed into 125 mph in roughly 8 miles. Assumed 93% of traction HP makes it to the rails.

Did piece-wise integration on a spread sheet in 5 mph increments.  (I am too stupid or lazy to try calculus anymore!) Dunce

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    February 2005
  • 2,366 posts
Posted by timz on Tuesday, May 14, 2013 2:55 PM

oltmannd
after 1/2 mile:  65 mph, 1:05

after 1 mile: 80 mph 1:40

after 2 miles 95 mph 2:30

after 4 miles 115 mph 4:50

That's with 18 cars weighing what?

Note the 52 mph average (two miles in 2 min 20 sec) while accelerating from 95 to 115 mph.

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 3:13 PM

Max acceleration is from 0 to ~30 mph at 1.5 ft/s/s running at adhesion limit, then tapers off as loco perf hits the HP curve and aero drag starts to build.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 3:14 PM

timz
That's with 18 cars weighing what?

55 tons a pop

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 3:16 PM

timz

oltmannd
after 1/2 mile:  65 mph, 1:05

after 1 mile: 80 mph 1:40

after 2 miles 95 mph 2:30

after 4 miles 115 mph 4:50

That's with 18 cars weighing what?

Note the 52 mph average (two miles in 2 min 20 sec) while accelerating from 95 to 115 mph.

speed distance
5 0.003021
10 0.009079
15 0.018196
20 0.030395
25 0.045705
30 0.06592
35 0.093889
40 0.131119
45 0.179279
50 0.240253
55 0.3162
60 0.409645
65 0.523595
70 0.661705
75 0.828519
80 1.029824
85 1.27319
90 1.568843
95 1.931127
100 2.381126
105 2.951746
110 3.698701
115 4.72808
120 6.283525
125 9.165761

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 3:16 PM

timz

oltmannd
after 1/2 mile:  65 mph, 1:05

after 1 mile: 80 mph 1:40

after 2 miles 95 mph 2:30

after 4 miles 115 mph 4:50

That's with 18 cars weighing what?

Note the 52 mph average (two miles in 2 min 20 sec) while accelerating from 95 to 115 mph.

time speed
3.189835 5
6.388742 10
9.597905 15
12.81853 20
16.05184 25
19.60973 30
23.82909 35
28.74339 40
34.39418 45
40.83299 50
48.12393 55
56.3471 60
65.60333 65
76.02079 70
87.76453 75
101.0506 80
116.1679 85
133.5129 90
153.6483 95
177.4082 100
206.1022 105
241.9561 110
289.218 115
357.6576 120
479.4033 125

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 3:19 PM

speed  TE  trn resist
5       100,000 3615
10       100,000 3888
15       100,000 4195
20       100,000 4536
25       100,000 4911
30          91,733 5319
35          78,629 5761
40          68,800 6237
45          61,156 6747
50          55,040 7290
55          50,036 7867
60          45,867 8478
65          42,338 9123
70          39,314 9801
75          36,693 10513
80          34,400 11259
85          32,376 12039
90          30,578 12852
95          28,968 13699
100          27,520 14580
105          26,210 15495
110          25,018 16443
115          23,930 17425
120          22,933 18441
125          22,016 19491

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Tuesday, May 14, 2013 4:16 PM
Numbers are bad. I'll fix them tomorrow. Mph X seconds <> feet

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Tuesday, May 14, 2013 4:33 PM

oltmannd

timz
That's with 18 cars weighing what?

55 tons a pop

If you have an Amcoach filled with 60 guys my size, that is an additional 6 tonsIndifferent

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Tuesday, May 14, 2013 4:35 PM

oltmannd

on a spread sheet in 5 mph increments.  (I am too stupid or lazy to try calculus anymore!) Dunce

Piecewise integration is plenty "good enough" for that very benign and smooth acceleration curve.  You don't need the answer to more than one decimal place anyway.

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    March 2016
  • From: Burbank IL (near Clearing)
  • 13,540 posts
Posted by CSSHEGEWISCH on Wednesday, May 15, 2013 6:33 AM

I will proceed to throw a monkey wrench into the above calculations:  Since the ACS-64 is a straight electric, have short-term ratings been considered in calculating acceleration rates?Whistling

The daily commute is part of everyday life but I get two rides a day out of it. Paul
  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, May 15, 2013 6:43 AM

corrected numbers

time speed distance  TE  trn resist
3.189611 5 0.006645       100,000 3608
6.387621 10 0.017749       100,000 3861
9.59475 15 0.033339       100,000 4134
12.81173 20 0.053446       100,000 4428
16.03931 25 0.078101       100,000 4742
19.58724 30 0.110131          91,733 5076
23.78752 35 0.153884          78,629 5430
28.66812 40 0.211502          68,800 5805
34.26269 45 0.285319          61,156 6200
40.61176 50 0.37791          55,040 6615
47.76417 55 0.492149          50,036 7050
55.77898 60 0.631295          45,867 7506
64.72787 65 0.799087          42,338 7982
74.69838 70 0.999882          39,314 8478
85.79819 75 1.238836          36,693 8994
98.16111 80 1.522153          34,400 9531
111.9553 85 1.857429          32,376 10088
127.3953 90 2.254152          30,578 10665
144.7596 95 2.724435          28,968 11262
164.4178 100 3.284146          27,520 11880
186.8733 105 3.954694          26,210 12518
212.8359 110 4.766024          25,018 13176
243.3492 115 5.761943          23,930 13854
280.0366 120 7.010337          22,933 14553
325.625 125 8.624923          22,016 15272

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Wednesday, May 15, 2013 9:55 AM

Another question:

So this thing has Cardan shaft drive instead of "nose suspended" traction motors for reduced unsprung mass and less pounding of the track at high speeds.

But isn't 120+ tons of total weight on what, four axles, a bit much for the kind of speeds contemplated?

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, May 15, 2013 10:03 AM

CSSHEGEWISCH

I will proceed to throw a monkey wrench into the above calculations:  Since the ACS-64 is a straight electric, have short-term ratings been considered in calculating acceleration rates?Whistling

Nope.  But  I can take a swag at them.  What would be a good guess...+20% for 2 minutes?

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, May 15, 2013 10:04 AM

I have no idea what the thing weighs.  I Mighty Mouse is about 100 tons.  This thing is bigger and stronger, so roughly P42 weight?  

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, May 15, 2013 10:07 AM

What has me scratching my head is this.  Amtrak is just going to ask for a whole bunch of money to upgrade the cat (and speeds) on the south end of the NEC.  Why not buy a loco that can take advantage of the speed?

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    February 2005
  • 2,366 posts
Posted by timz on Wednesday, May 15, 2013 6:10 PM

Still some sort of big error in your calculation-- 100000 lb TE won't accelerate an 1100-ton train at 1.6 mph/sec, even with zero rolling resistance.

If we assume traditional Davis resistance, then train resistance in pounds is A + BV + (C times V squared) where V is speed in miles/hour and

A = 3653.5

B = 33.45

C = 1.0344

That's assuming 125 US tons for the engine-- zat right?

And you're assuming 7338-2/3 rail horsepower at all speeds, except a 100000-lb TE ceiling up to 27.52 mph. Tonnage 1115 actual and I'll guess 1159 effective inertial tons (the increase for rotational inertia).

The 18 cars are 1530 ft long; in that distance the train accelerates to 40.37 mph in 48.5 seconds. The standing-start mile takes 97.75 seconds; speed 61.6 mph at that point.

Five miles takes 273.8 seconds; speed there 96.1 mph; it reaches 100 mph in 31688 ft and 310.5 seconds. After ten miles they're at 109.7 mph.

In reality the train won't accelerate to 40 mph in its own length-- none of us knows what it could do if the engineer didn't bother starting smoothly.

Traditional-Davis says the train needs 2763 rail horsepower to maintain 80 mph. Think it actually needs less than that?

  • Member since
    August 2005
  • From: At the Crossroads of the West
  • 11,013 posts
Posted by Deggesty on Wednesday, May 15, 2013 6:47 PM

Quoting overmod: "Just what an 'input inverter' that feeds a 'DC link' might be is a mystery.  I suspect they meant 'rectifier' in this context (see here for an example feeding a motor directly) ... but got lost in translation without a proofreader who understood the technology."

You really need a copy editor who understands the technology; a copy editor has the authority to make changes in text; a proofreader only ascertains that what is going to be printed is what the author wrote and has no authority to correct the original manuscript/text (my wife worked in both capacities for two book publishers--and I gave her assistance when she was copy editing books that had mention of  railkroads).

 

Johnny

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, May 15, 2013 8:42 PM

Deggesty
You really need a copy editor who understands the technology; 

You are completely right.  I *think* I was attributing fact checking to proofreading, but that's no excuse.  I would note in my own defense that when I proof galleys, fact checking is inherently part of the exercise; I don't just limit it to correcting typographical mistakes.

Writer didn't recognize (or perhaps didn't care about) the difference, and editor didn't or couldn't catch it.  Not a good thing all around for a technical publication...  but then again I'm too much of a perfectionist when it comes to printed media...

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, May 15, 2013 9:24 PM

Paul Milenkovic
But isn't 120+ tons of total weight on what, four axles, a bit much for the kind of speeds contemplated?

No.  (Assuming the period and damping of the secondary suspension is correct).  The low unsprung mass, especially with respect to low inertia of the wheel in all axes of rail interaction, is far more significant for high speed.

Weight distribution across those four axles is of course important, and there *might* be a case made for 'taper loading' (a bit less weight on the outer axles) but at those speeds you will not have a problem with primary deflection, or (with any sensible type of track curvature and transition spiraling) high rates of lateral acceleration of heavier mass requiring sudden high flange forces.  So the actual weight of the carbody is of less significance, up to the point the primary springing has to become too 'hard' to give short-period following of rail-level anomalies.  (I do not think this is the case for 30-ton axle load, per se)

If you look at the Siemens drawing of the truck frame, the secondary suspension is four springs, arranged laterally (two and two) right on the lateral centerline of the truck frame.  I presume the torque link is decoupled from the suspension (as it is for the Flexi-Floats) so there is no particular weight transfer on acceleration, and therefore no weird loading of the primary suspension.

I am presuming that Siemens knows where any critical speeds are, and has designed the truck masses and suspension characteristics accordingly.  

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Wednesday, May 15, 2013 11:52 PM

Overmod

Erikem is admirably suited to answer this in full detail.

It is a bit difficult to extract the sense from what I think is translated German, but it would appear that the setup is heavier because there are separate transformer windings for each of the three powers, presumably stepping down to a standard (unspecified) voltage going to the 'cubicles'.  It woud have been interesting to see a block or circuit diagram showing more precisely how Siemens designed this.

Just what an 'input inverter' that feeds a 'DC link' might be is a mystery.  I suspect they meant 'rectifier' in this context (see here for an example feeding a motor directly) ... but got lost in translation without a proofreader who understood the technology.

I'll give it my best shot...

An H-bridge inverter will gladly transfer power from the AC side to DC side (i.e. acting like a rectifier) as well as transferring power for the DC side to AC side (traditional inverter). In fact, many low output voltage (1V or less) DC-DC converters use FET's operated as synchronous rectifiers instead of diodes, as it is easy to make the I*Rdson voltage drop less than the forward voltage drop of a diode. Having the inverter capability to feed regenerated power back to the catenary also saves in not having to carry the DB resistor grids around.

I would suspect that the there may be separate primary windings for the different voltage levels, though one way of handling 25Hz would be to use the same number of turns for 11kV/25Hz primary as for the 25kV/60Hz primary, but increase the number of secondary turns for 25Hz to maintain the same secondary voltage. This would result in about the same peak magnetic flux in the transformer core.

- Erik

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, May 16, 2013 5:05 AM

timz
Traditional-Davis says the train needs 2763 rail horsepower to maintain 80 mph. Think it actually needs less than that?

Has to be.  C is what I had to vary to back into the performance.  There is no published info I could find on a modified Davis Eq. for Amfleet.

I ignored the mass of the locomotive - which isn't a terrible thing to do in the freight world, but matters a bit more here.  I should add it in...it's easy to do.  Adding in rotational inertia?  Meh.  It's less than passenger load (or my error in actual train weight)  Let's just call this a one and a half significant digit exercise!

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, May 16, 2013 5:07 AM

timz
Still some sort of big error in your calculation-- 100000 lb TE won't accelerate an 1100-ton train at 1.6 mph/sec, even with zero rolling resistance.

Probably units in my accel calc.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, May 16, 2013 5:18 AM

oltmannd

timz
Still some sort of big error in your calculation-- 100000 lb TE won't accelerate an 1100-ton train at 1.6 mph/sec, even with zero rolling resistance.

Probably units in my accel calc.

Looks okay to me.  But units are ft/s^2.  No reason engineer can't start a stretched train at that rate after a few seconds.  It's much slower than the 0.1g rate for some transit equipment.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, May 16, 2013 5:22 AM

timz

If we assume traditional Davis resistance, then train resistance in pounds is A + BV + (C times V squared) where V is speed in miles/hour and

A = 3653.5

B = 33.45

C = 1.0344

Yes, except I modified C to back into 8 miles and 125 mph (or thereabouts)

I didn't the "bearing" and "flange/wheel" parts of Davis would differ much.

R=1.3 + 29/W +0.045V +0.0005kAV^2/WN

in #/ton

where k is my fudge factor.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Thursday, May 16, 2013 7:27 AM

Overmod

Just what an 'input inverter' that feeds a 'DC link' might be is a mystery.  I suspect they meant 'rectifier' in this context (see here for an example feeding a motor directly) ... but got lost in translation without a proofreader who understood the technology.

Siemens are very good at providing technical detail on their website, often in English as well as German and circuit block diagrams are provided. The ACS-64 is basically a standard European locomotive adapted for Amtrak's voltages.

Both input and output inverters are basically the same. The input inverters under power, as you imply, operate as rectifiers and feed their intermediate DC to the output inverters. Under power two of the output inverters feed the motors with variable voltage, variable frequency AC while the third provides 60Hz power at constant voltage to the train.

When in regenerative braking the "output" inverters take three phase power at variable voltage and frequency and convert it to DC which they feed to the "Input" inverters which convert it to single phase AC at the line voltage and frequency.

The input and output inverters are basically the same. The three output inverters are identical, and if the train HEP inverter fails, one of the traction inverters takes its place and the train proceeds on half traction power.

To address one point by the original poster, there are two major overhead line frequencies used in Europe, 50Hz (25kV) and16.66Hz (15kV).  These are equivalent to the two voltages and frequencies used by Amtrak and many locomotives are built for the two AC systems. Generally, the transformer is built to take the lower frequency and will operate correctly at the higher frequency. Amtrak's 25Hz transformer will be smaller and lighter than a German 16.66Hz transformer of the same power which would be installed in the equivalent standard locomotive.

For those new to this, 50Hz and 60Hz are unsuitable for direct use in commutator motors of high power so lower frequency current of 16.66Hz or 25Hz was used for rail traction purposes. The catenary of the Pennnsylvania and New Haven was 25Hz and Amtrak only adopted 60Hz after only rectifier locomotives or inverter locomotives were in use.

M636C

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy