[steamer_ctn]

During the performance testing of steam engines, the Steam Locomotive Performance Group (SLPG) identified the fact that the resistance of the engine and cars modeled in OR can prevent the locomotive from performing as expected in the "real world".

In summary, the locomotive produces a tractive force, and this is designed to overcome the resistance of the cars/wagons in the train. So for example if the resistance of the train is high compared to the real world values then the locomotive will struggle to pull expected loads, similarly if the value is low then the locomotive will be able to over perform compared to expected loads.

Initially a lot of the testing was based upon UK locomotives as comprehensive test reports existed for a number of locomotives. In studying these reports it was noted that both the locomotive and car resistances values documented in the test reports were less then those calculated with the traditional Davis equations calculated in FCalc. This meant that when FCalc configured stock was used in the testing process, it was difficult to tune the locomotive to achieve the results specified in the test report. However, when the test report resistance values were used it was possible to tune the locomotive to met the measured outcomes in the test report.

The SLPG has spent a reasonable time and effort investigating train resistance in order to try and refine the performance testing of locomotives. Based upon this research they are offering some suggestions as to how resistance might be set "more realistically". Whilst the group have not fully completed this task, it was decided to share the current thinking of the group in order to seek other expert input, and maybe gain access to some currently unobtainable resources on train resistance.

The first step in this piece of work was to review the development of the Davis equations, and the attached paper, based mainly upon information in AREA proceedings, summarises the development of the formulas over time. It will be noted from this paper that the original formulas develop in 1926 were mainly based upon some work that was done in 1917 by Schmidt. Naturally, in the last 100 years railways have changed significantly in terms of their speed of operation, track design and building standards (track work now more rigid) and the design of rolling stock. This has meant that there is general agreement among the railway companies that the original Davis equations are no longer as accurate in these more modern operating environments. In more recent years with the increasing interest in High Speed railways there has been significant testing of rolling stock undertaken at operational speeds of in excess of 200mph. As is demonstrated in the document, the air resistance factor (Value C in the Davis formula) has a huge impact as the speed increases.

Bearing in mind that the original Davis equations are based upon work from the 1917s, when the average speed of a freight train was 40mph, it can be imagined that they didn't really contemplate the more modern speeds of operation, etc. As a consequence the railway industry in America has made modifications in the Davis equations to cater for the changes in railway industry, and thus it has evolved over the years.

The group is therefore offering the following suggestions in regards to setting train resistance:

i) Always use a measured value of resistance from a reliable source as the first preference - a lot of railway companies have done resistance tests on some of their rolling stock. This information may be obtainable from Internet searches, or from railway company archives.

ii) Only use the Davis formula as a default value when it is not possible to source an alternative measured value. Always use the relevant Davis formula most relevant to the era, track type, stock type, etc. See the table in the attached document as a suggested formula selection methodology. Make sure that input values to the formula as as accurate as possible (especially anything impacting the C value).

iii) It appears that Davis may have combined the steam locomotive and tender resistance into a single formula, so it is suggested that the locomotive and tender weight, and axle numbers, etc be combined, and then inserted into the relevant Davis formula as a single unit. Once the output values are obtained, they will need to pro-rata across the relevant ENG and WAG file for the tender and locomotive for correct operation in OR. This suggested change will result in a reasonable decrease in the overall train resistance as calculated by OR.

iv) It appears that Davis equations are not designed for speeds over 125mph, and therefore relevant measured values should always be sought for high speed trains.

At the moment the above are thoughts/suggestions from the group, and we would welcome any feedback, supported by any appropriate documentation in regards to the above suggestions.

The group would also be interested in obtaining copies of the articles by Davis (1926), and Totten (1937) to potentially fill in a few blanks that the group are still trying to understand. In particular if anybody is able to provided worked examples of the application of the Totten formulas, it would be most appreciated.

Attached File(s)

•  Train_Resistance_v3.pdf (728.45K)
  Number of downloads: 9068

[Genma Saotome]

In the .pdf on page 12 you use these qualifiers:

• pre-1950s
  
• post-1950s

That omits the 1950's.

[steamer_ctn]

Hi Dave,

Genma Saotome, on 19 November 2017 - 09:56 PM, said:

In the .pdf on page 12 you use these qualifiers:

• pre-1950s
• post-1950s

That omits the 1950's.

Thanks for that, I don't know how I could have missed such a great era with its music, etc!!!

I will correct it in the document.

By any chance do you have access to the Totten or Davis articles referenced in the paper?

Thanks