I will mail you some investigations I've made on the IORE line.
Basically, within nominal speed limits and with respect to super elevation and radii, speed has a relatively small influence on the curve resistance. The main components, for nominal quasi static conditions, are radii, the alignment of wheelsets and mass/axleload.
However, at low speeds and cant surplus, wheelsets can align themselves against the curvature, thus increasing the frictional forces and the resistance, quite a bit actually. Also when you have cant deficiency or the wheel make contact with the rail on several spots or with its flang. I've made investigations for different types of bogies, superelevations and radii...etc
Most of the formulas are from empircal tests but it is also possible to simulate by means of software...
I can't refer you to specific link or reference, but in our courses for Curve Resistance it is mentioned that this force is not dependent on rolling stock speed and Super-elevation of running rails.
Refer to my documents in railway school courses, there is two type of formula;
1) There is a formula named "Rock Relation" as below:
A complete and clear description is in "Modern Railway track" by C. Esveld (1989). Take a look at attached pages (more specifically paragraphs 4.8, 4.9 and 4.10). I hope they are useful.
most simple models use a running resistance formula for straight and level track, modified davis equations for freight trains, polynomial equation a+bV+C V2 for passengers trains
To take gradients into account, you add a resistance corresponding to the projection of weight in the track direction ( G sin Theta
Usually, curves are taken into account by adding an "equivalent curve/grade" which, for standard gauge, is the Rockl formula given by Mr Firouzian
you can find a lot more information in R Bosquet Doctoral Thesis