Elvas Tower: Diesel Locomotive Performance - Elvas Tower

Jump to content

Posting Rules

All new threads will be started by members of the Open Rails team, Staff, and/or Admins. Existing threads started in other forums may get moved here when it makes sense to do so.

Once a thread is started any member may post replies to it.
  • 20 Pages +
  • « First
  • 9
  • 10
  • 11
  • 12
  • 13
  • Last »
  • You cannot start a new topic
  • You cannot reply to this topic

Diesel Locomotive Performance Rate Topic: -----

#101 User is offline   NickonWheels 

  • Conductor
  • Group: Status: Active Member
  • Posts: 327
  • Joined: 05-December 19
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 04 January 2020 - 01:12 AM

Thanks guys,

indeed I meant THAT starting friction, the Davis values are the easy part but for this kind of friction they are irrelevant (starting friction could be 1000+ newtons even when Davis A says only 1). I dont know ORTSAdhesion, trying it out; until later...

History is considerable when you can learn from it. :ireful2:

#102 User is offline   NickonWheels 

  • Conductor
  • Group: Status: Active Member
  • Posts: 327
  • Joined: 05-December 19
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 04 January 2020 - 02:01 AM

Quote

Empirical static friction forces are used for different classes of mass (under 10 tons, 10 to 100 tons and above 100 tons)


Empirical? Why not let every user figure out the sarting value, it´s more or less the Davis A value and some multiplier, so a difinable curve like traction curves could take such a value for zero speed and the Davis equiations for higher speeds. History should go on... (quite angry)

#103 User is offline   Gehe 

  • Apprentice
  • Group: Status: Active Member
  • Posts: 48
  • Joined: 05-January 14
  • Gender:Male
  • Simulator:MSTS
  • Country:

Posted 04 January 2020 - 08:12 AM

Hi Guys,

being back to the internet I saw a lot of discussion in the last days here on the topic of Diesel Performance as well as on starting resistance to which I would like to add my view:

Regarding the Diesel Performance I am in line with the opinion that the max power values in the .eng files should correspond to the power applied to the rail.
This approach is simple and everybody can do his own efficiency calculation or estimation. To perform this by OR in a satisfying way without adding much complexity is difficult.
E.g. considering things like the type of transmission or to consider the power supply of passenger trains which are often supplied by the prime mover thus reducing the portion which can be used for traction.

Related to starting resistance I have like others the feeling that the values in OR are too high. My personal workaround is to use the friction values of the MSTS despite their "strangeness".
In this case OR calculates the starting friction different (see source code) and they look for me more realistic. To achieve this you have to choose the MSTS (V1 and E1) values accordingly in FCALC2.

In general I experienced the same as many members of the community here that it is difficult to obtain reasonable friction values at all.
A lot of sources are very old (70-90 years). I found some sources dealing with European rolling stock which I use normally and use these as Input for FCALC.
Would be better to use the OR Davis parameters directly but as long as OR has the "starting friction problem" this is difficult.

According to some sources the starting friction disappears after some revolutions of the wheel and is not just a matter of velocity. You can do a lot of research here but i think a simple model allowing to set the starting resistnce which disappears according to velocity or after some distsnce is a good first step. In a second step the results of comprehensive research and investigation can then be introduced.


Regards
Gehe

#104 User is offline   NickonWheels 

  • Conductor
  • Group: Status: Active Member
  • Posts: 327
  • Joined: 05-December 19
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 04 January 2020 - 10:08 AM

Quote

Would be better to use the OR Davis parameters directly but as long as OR has the "starting friction problem" this is difficult.

As soon as you use the Davis lines the whole standstill friction debacle starts as it´s not dependant on Davis A or is directly influenceable, hence I said some kind of curve would be better...

Besides that it generally feels strange not to use ORTS lines (but the old MSTS lines) for more realism as they were originally intended to enhance this, but something had been cut short. It is sad none of the develoment team can really get over this idea as (often insisted) it´s more about sources. How can sources help when everything in only indirect. Traction curves are the most direct way to define the nesseary values, so that´s the missing piece of the puzzle here; of course you need sources to get the traction values right, but I most often also think about fictional locomotives, a whole other section which not wants to be discriminated.

(If you wanna know the fuel consumption you surely would know what the cylinder diameter is, right?)

#105 User is online   steamer_ctn 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 1,879
  • Joined: 24-June 11
  • Gender:Male
  • Country:

Posted 28 January 2020 - 02:14 AM

Given some of the angst being raised about resistance, adhesion, etc and also to satisfy my current work on reviewing the code for diesel locomotive performance I have constructed a further test scenario to prove the accuracy of the diesel locomotive code. Earlier in this thread you will have see reference to test stock that very closely mimics BR test results. These results suggest that OR can be configured for accurate operation of DC locomotives.

In this scenario I wanted to focus on AC locomotives, which are slightly different to DC locomotives. For these locomotives to perform to any test results the resistance, adhesion, etc needs to be accurately portrayed, as the old computer adage says, "Garbage in = Garbage out.

I made an offer to two community members who raised concerns about the accuracy of the OR code, and one of them has accepted the offer, and has been working with me to build and test the scenario.

As testing progresses I hope to share the rolling stock setups and relevant information about the scenario so that other community members can see the results and be assured of the accuracy of OR.

The outcome of this testing will be to confirm the accuracy of the code, and will also identify any potential areas of the code that may require "tweaking" to ensure an accurate outcome against the test scenario.

In this post I would like to describe the test scenario (subsequent posts will describe some of the other elements of the test scenario and results).

Rather then making up a test scenario based upon my perception of how a train should perform, I prefer to use well documented test results from the real world. This removes any doubt about the outcome and its accuracy.

With this in mind I have found a Norfolk Southern document which is titled "Locomotive Tractive Effort Tests on Norfolk Southern Railway", and describes tests that the railway undertook to determine the tonnage rating for various diesel locomotives.

In simple terms, this document gives us a tonnage rating, speed, and gradient that the locomotive is tested on. Hence we should be able to duplicate these tests, and if OR is configured (and coded correctly) we should get the same results (within an acceptable degree of accuracy) as published in the document.

When reading the document the following "invisible information" is relevant:

Locomotive Group E1-E8 appear to be the following, respectively: C39-8, SD70, D9-40CW, SD70M, ES40DC, SD60, D8-40C, SD70M-2 (? on that last one)
Locomotive Group F1-F6 appear to be the following, respectively: 45AC (not sure exactly what this is in NS lingo), ES44AC, SD70ACe, AC4400CW, 90MAC, 80MAC

For this test scenario, the SD70ACe (F3) was chosen.

A test gradient equivalent to 1.4% has been added to the CTN test route (not published yet), and it will be deemed a successful test if the train consist(s) described in the document can maintain a balancing speed up the gradient as suggested in the test report. (note: over-speeding is considered just as big an issue as under-speed)

#106 User is offline   superheatedsteam 

  • Conductor
  • Group: Status: Contributing Member
  • Posts: 496
  • Joined: 28-June 08
  • Location:Perth, WA
  • Country:

Posted 28 January 2020 - 03:33 AM

Hi Peter,

Thank you for your work on this. The NS pdf is an interesting read.

As a matter of interest, do the curves and gradients of your test track match the track profile of the 'TEST SEGMENT' on page 3?

Cheers,

Marek.

#107 User is online   steamer_ctn 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 1,879
  • Joined: 24-June 11
  • Gender:Male
  • Country:

Posted 28 January 2020 - 01:26 PM

View Postsuperheatedsteam, on 28 January 2020 - 03:33 AM, said:

Thank you for your work on this. The NS pdf is an interesting read.
Thanks for that.

View Postsuperheatedsteam, on 28 January 2020 - 03:33 AM, said:

As a matter of interest, do the curves and gradients of your test track match the track profile of the 'TEST SEGMENT' on page 3?

The test track that I have created is a straight track section at 1 .4% gradient. I don't believe that matching the profile will change the results significantly (as hopefully will become obvious with my further posts), however always happy for somebody to build an exact replica of the test track on pg 3 to see the difference in the results.

#108 User is online   steamer_ctn 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 1,879
  • Joined: 24-June 11
  • Gender:Male
  • Country:

Posted 28 January 2020 - 05:58 PM

Before I describe the test setup in more detail, I want to make sure that we are on the "same playing field" and have the same foundational understanding of the different elements at play as we undertake these tests (So a little Train Physics 101).

So take a few moments to read Principles for Train Movement, and pay particular attention to the diagram shown in this section of the page.

In short, the diagram shows that we have opposing forces at play when we try to move a train. Whenever the tractive effort is greater then the resistive forces of the train, then we should see movement and acceleration in the train. When the resistive forces are greater we will see the train slow down, and in some instances stall (stop) completely. From the diagram, we can see that the resistance forces intersect the tractive forces curve, and this is the point where we see the train speed remain constant. I will refer to this as the "Balancing Speed Point" (BSP).

In effect the NS document has achieved this balancing speed point for each of the trains as they ascend the 1.38% gradient. As the resistance or tractive forces change, so will the BSP.

For example, different locomotives have different BSPs, as the tractive force curves for each of the locomotives is different. Similarly if we vary the load, then the train resistance curve will vary, and hence so will the BSP.

Of interest to note was the need for NS, as a result of the testing, to change the load ratings for Group F locomotives under wet conditions.

In my following posts I will describe how the test stock was configured in regard to these two forces, ie Tractive and Resistive

#109 User is online   R H Steele 

  • Executive Vice President
  • PipPipPipPipPipPipPipPipPip
  • Group: ET Admin
  • Posts: 3,404
  • Joined: 14-March 13
  • Gender:Male
  • Location:known universe
  • Simulator:Open Rails
  • Country:

Posted 29 January 2020 - 12:10 PM

Peter, thank you for the NS test docs and also for the link to your physics/resistance page --- for some reason I had not read this page...thanks, I'm reading it. Are you contemplating adding any specific ORTS parameters .... for instance ORTSRailHP to the current work you are pursuing?

#110 User is online   steamer_ctn 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 1,879
  • Joined: 24-June 11
  • Gender:Male
  • Country:

Posted 29 January 2020 - 07:19 PM

View PostR H Steele, on 29 January 2020 - 12:10 PM, said:

Are you contemplating adding any specific ORTS parameters .... for instance ORTSRailHP to the current work you are pursuing?

At this stage I don't think that their is a need for any addition parameters, the ones already in existance should do the trick (I think).

However we will see what happens as the testing progresses.

  • 20 Pages +
  • « First
  • 9
  • 10
  • 11
  • 12
  • 13
  • Last »
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users