Frequently Asked Questions (FAQ)

  • Where is the data sourced from?

The data is sourced from a large variety of sources including official manufacturers information, British Railway records,  technical publications, technical books and some unofficial records.

  • How is the data captured?

The original chart is scanned as a high resolution .PNG file. The file is then enlarged and if necessary straightened or repaired in Photoshop.  The enlarged file is then opened in the program "EDFI"(Extract data from image). The data is then copied into Excel for plotting.

  • How accurate is the data?

It is suggested that the derivation of tractive effort be considered to have a spread of +/- 3% across the range. There are 4 potential inaccuracies, which may not be cumulative, in the data as detailed below:

  1. Inaccurate digitising of the original curve

The original curves are scanned at 300 dpi resolution and opened in Photoshop.  The curve is enlarged to around 2000-2500 pixels in size and straightened as necessary.  The program EDFI (Extract data from image) is used to digitally capture the data.  The resolution typically gives 1 pixel = 25-30lbf and 0.05-0.07 mph. This equates to an accuracy of 0.1 mph or 60lbf (5 and 12 rail hp at 30 mph and 75mph

   2. Inaccuracy in enlarging the original curve

The original curve in many cases is only available as a relatively small curve.  The enlargement provides a more accurate curve however there is a degree of interpretation required as the curve thickens with the enlargement and care is needed to select the centre of the line.  A number of verifications have been undertaken where the original curve is available in varying sizes and a very good correlation was found to exist between the large original and the enlarged duplicate.

   3. Inaccuracy in the plotting of the original data

In many cases the original curves were plotted half a century ago or longer.  The original curves were drawn manually from a series of data points often at 10 mph intervals.  This required a considerable skill to develop the curves accurately.  This is particularly so at higher speeds where the tractive effort curve flattens and is one reason for plotting the rail horsepower on the same graph

  4.  Inaccuracy of measuring the original data

The original tractive effort curve was in most cases derived from the drawbar horsepower recorded in a dynometer car. It was then converted to tractive effort. This required measurement of the gradient, an accurate determination of the current weight of the locomotive and by derivation an estimation of the resistance of the locomotive.  In cases where the tractive effort was derived multiple times for the same speed a considerable spread of data was often observed.  It is suggested in the absence of more data that the accuracy or repeatability would be no better than +/- 2%. In some cases the locomotive under test was not producing the rated output of the engine.  In most of these cases the engine was producing a little below the rated output although there were cases where the suspicion was that the output was higher than the engine rating.  (An example would be the prototype locomotive LION where it appears that the engine setting was set at 2 800 bhp rather than 2 750 bhp.  This may have been because the Brush competitor FALCON, was rated at 2 800 bhp!)

  • I have a tractive effort curve which is different to the one on this web site.  Which is correct?

This is an common question and is difficult to answer correctly.  For many types of locomotive there are a number of different tractive effort curves available.  In some cases there are inaccurate curves which have been published in various articles and magazines. Often the plotting of the rail horsepower vs speed will indicate an error. This error may be simply to rectify (an axis is in-correctly numbered) or the data may have been theoretically derived rather than from dynometer data.  There are a few derived tractive effort curves on this site (e.g. Class 23 Baby Deltic) where base data has not been found however these curves are clearly labelled accordingly. There are also cases where either the manufacturer or British Railways used junior engineers or draughtspersons to draw copies of curves before there was easy access to copying facilities. In some cases this introduced errors into the curve.

©2018 by RailTechnical. Proudly created with