Trains.com

Modern diesel efficiency compared to first generation..

12273 views
34 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    March 2003
  • From: Central Iowa
  • 6,901 posts
Posted by jeffhergert on Wednesday, January 5, 2022 10:03 PM

Overmod

 

 
CSSHEGEWISCH

It would be interesting to see the overall fuel usage and efficiency figures on the variable horsepower Cv40-9i's (NR class) introduced by National Rail in Australia.

 

It appears that most of the recent work in Australia/New Zealand follows the theory that 'managed' trains best operate at sustained speed, directed via a 'smart' system like LEADER or TO -- that implies reasonably fast acceleration subject to pollution control, then modulation of power as needed to control speed.

 

What this does NOT do is control the engine governors in a consist to achieve the equivalent of restricted notch or excitation.  Instead it acts to keep some locomotives in the consist at high notch while near-idling the rest (see SmartConsist or Smart HPT) which is not what the three-stage derating system on the NDs apparently does.

 

First time I've heard LEADER or Trip Optimizer referred to as "smart" systems.  Definitely something you won't hear from anyone that uses them.

Jeff

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Thursday, January 6, 2022 10:08 PM

Note that I put the word in quotes.  The current train management systems are 'smart' in the sense smart houses and smart bombs are.

And remember that the smart in 'whip-smart' has two meanings... Surprise

  • Member since
    January 2019
  • 1,690 posts
Posted by Erik_Mag on Friday, January 7, 2022 12:13 AM

With respect to improving efficiency, I was reading a couple of articles about a company that is using a "commutating pole" to drastically reduce switching loss in inverters. They claim their prototype board is capable of achieving a peak of 99.5% efficiency in converting 600 to 800 VDC to three phase AC.

I've bolded the "switching loss" as the active devices (e.g. IGBT's, FET's, SiCFET's, GaNFET's, GTO's, etc) experience two kinds of loss. One is conduction loss, caused either by the voltage drop due to Rds_on in the various FET's or the VCE_ON for IGBT's and GTO's. The other loss is switching, with much of that loss involved in charging or discharging device capacitances and it what the new technology promises to drastically reduce with switched capacitance (this is not snake oil).

Note that the real benefit from increasing efficiency from 99% to 99.5% is not the savings in energy, but instead is in the reduction in heat that must removed by coolants.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, January 7, 2022 3:26 PM
  • Member since
    January 2019
  • 1,690 posts
Posted by Erik_Mag on Sunday, January 9, 2022 10:46 PM

Overmod

Coming up with a control algorithm that will adjust timing to accomodate changes in the circuit parameters. Examples include juncton temperature affecting rise and fall times, with corresponding changes required dead time. One difference between then and now is that real time processing power to do a cycle by cycle adjustment in timing is a lot more affordable now that back then. Another difference is using SiC or GaN FET's, nether of which have the turn off current tail from IGBT. The 300kW demo inverter uses USiC (now part of Qorvo) cascode JFET's.

I came across a similar situation with an H-bridge using SiC FETs, looking at the cicuit response while adjusting deadtime - saw that increasing deadtime a bit allowed the slightly inductive load to commutate the H-bridge - the "inductive current" would charge the output capacitance of the device that was turned off, and the other device in the totem pole could be turned on with zero voltage across it.

The reason why eliminating switching loss is a big deal is that otherwise the design would need to find a compromise between switching and conduction losses. Conduction losses can be reduced by going to a larger die, but that comes at the expense of increased Drain - Source capacitance, which then entails more energy each time a device switches. Eliminating switching losses also allows operation at a higher frequency, which makes fitering easier.

I'll have to pass the link to my son at Purdue as one of his housemates is working on Purdue's DC House. Yes, I did see the purdue.edu in the link...

The story of the commutating pole for inverters reminds of a story about high power low distortion class A + AB audio amplifier stage. When the engineer asked why he didn't patent it, he said the stage was prone to self-destruction without the VI limiter that he did patent.

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