[R H Steele]

steamer_ctn, on 02 January 2020 - 10:48 PM, said:

Does it really? I was under the impression that the "375" was a constant to allow the use of imperial values rather then metric values, and as far as I am aware it did not include any allowance for efficiency "loss". I would need to do some more research to confirm this though.

I've always read that is so, here's a couple of references...for what they are worth, still looking for a more technical paper that I read...cannot find the link. That particular formula is often quoted with the factor of 308, 315 and 375 --- and yes it is also a constant to convert UoM.--- my understanding is that it has eff. built into it...I'll need to find that paper. But as you explain it...taking the eff. into account would be pointless if the RailHp was specified. --- I rely upon both Derek Morton and Bob Boudoin for their educated eff. guesses...since my experience is considerably less than theirs.

http://cs.trains.com...1/t/164703.aspx and http://cs.trains.com...11/t/75858.aspx

[steamer_ctn]

Hi Gerry,

R H Steele, on 02 January 2020 - 11:46 PM, said:

I've always read that is so, here's a couple of references...for what they are worth, still looking for a more technical paper that I read...cannot find the link. That particular formula is often quoted with the factor of 308, 315 and 375 --- and yes it is also a constant to convert UoM.--- my understanding is that it has eff. built into it...I'll need to find that paper. But as you explain it...taking the eff. into account would be pointless if the RailHp was specified.

Thanks for the clarification, and I can see where your comments come from.

I believe that my comments are still relevant, as it will depend upon which power you are using for the reference point, as to which value multiplier to use.

Ideally for the traction curves, it should always be the rail power/force where possible and the 375 value. The only question remaining in my mind, is whether these types of figures published on the INTERNET have been calculated using the factors that you have pointed out (less reliable), or whether they are based upon manufacturers specifications (more reliable)?

I hope this helps in some way.

[NickonWheels]

All this discussing about the traction curve stuff makes me wonder why this is not there for vehicle friction, the Davis stuff. Why to make weird traction values but no self-examination of the vehicle friction. Pity, as this would solve this never-ending nightmare...

[darwins]

I have attempted to provide some help with vehicle friction with a calculator spreadsheet.
Getting values for the "Davis stuff" is not so easy though. In fact I would say that it was more of an art than a science.

If there are published values for your train model then you can use them - otherwise you will find that different railways in different countries all have different ways of measuring or estimating friction.

If you look at the documentation included with the spreadsheet some of the differences are described and explained.
Ultimately it comes down to performance and testing to try to make sure your train has characteristics similar to the real thing.

[NickonWheels]

Quote

If there are published values for your train model then you can use them - otherwise you will find that different railways in different countries all have different ways of measuring or estimating friction.

If you look at the documentation included with the spreadsheet some of the differences are described and explained.
Ultimately it comes down to performance and testing to try to make sure your train has characteristics similar to the real thing.

You´re right, but without something like the traction curve for friction getting close to the real thing is not that possible when having only three options (the three bearing types) and no way of a real value like for the traction newtons. Calculating such a curve on account of the Davis stuff is rather easy then because the problem is neither of the Davis lines, but standstill friction. A curve would fix this immediately...

[R H Steele]

NickonWheels, on 03 January 2020 - 07:04 AM, said:

All this discussing about the traction curve stuff makes me wonder why this is not there for vehicle friction, the Davis stuff. Why to make weird traction values but no self-examination of the vehicle friction. Pity, as this would solve this never-ending nightmare...

darwins, on 03 January 2020 - 09:14 AM, said:

I have attempted to provide some help with vehicle friction with a calculator spreadsheet.
Getting values for the "Davis stuff" is not so easy though. In fact I would say that it was more of an art than a science.

If there are published values for your train model then you can use them - otherwise you will find that different railways in different countries all have different ways of measuring or estimating friction.

If you look at the documentation included with the spreadsheet some of the differences are described and explained.
Ultimately it comes down to performance and testing to try to make sure your train has characteristics similar to the real thing.

Re: Friction, I would have thought Joe Realmuto's FCalc2 would be more than adequate to find the correct Davis values, after reading the docs that come with FCalc and his threads, comparing with other established work, I'm very satisfied with FCalc2. I'm puzzled by the statements "getting the values for the 'Davis stuff' is not so easy" and "weird traction values" .... all seems pretty straightforward to me...of course I may need further education on the matter...if so, please enlighten me.

[NickonWheels]

Quote

I'm puzzled by the statements "getting the values for the 'Davis stuff' is not so easy" and "weird traction values" .... all seems pretty straightforward to me...of course I may need further education on the matter...if so, please enlighten me.

Of course.
with traction/dynamic brake curves you can define the force even at zero speed but for friction at those speed ORTS just uses the vehicle weight and inserted bearing type and mixing it up with some quite unfair working equation (notice slow spped friction is always in order between 0-5 mph). When a similar curve to the traction list would work it´s easy to determine this crucial value yourself. Then you have the right to choose indefinately; that´s why I said weird traction values because you can them pretty unrealistic by inserting whatever you want, but in the end this is what makes the traction curves so useful. So a friction curve would be useful too; it´s basicly similar with speed on the left and newtons at right (speed dependacy just like Davis). I´m not really getting behind why ORTS could not have implemented such thing in the past, but I would say it´s nessesary for some time soon.

[darwins]

R H Steele, on 03 January 2020 - 09:50 AM, said:

Re: Friction, I would have thought Joe Realmuto's FCalc2 would be more than adequate to find the correct Davis values, after reading the docs that come with FCalc and his threads, comparing with other established work, I'm very satisfied with FCalc2. I'm puzzled by the statements "getting the values for the 'Davis stuff' is not so easy" and "weird traction values" .... all seems pretty straightforward to me...of course I may need further education on the matter...if so, please enlighten me.

FCalc2 gives you the modern estimates of Davis A, B and C based on the original Davis equation and further Canadian work on various types of modern freight vehicles.These tend to be fairly good for most North American stock and particularly for heavy freight trains.
On the other hand FCalc2 values for European passenger stock, including British stock tend to be somewhat higher than those measured by various railway companies.There is also considerable variation when it comes to modern high speed trains and these in particular need to be based on the nearest available railway data.
Some explanation for this together with references are given in the documentation.  Friction_Calculator.zip (394.79K)
Number of downloads: 434
(All that I have researched is concerned with the resistance of moving trains. Low speed resistance starting from a standstill is a different ball game again.)

[R H Steele]

darwins, on 03 January 2020 - 11:40 AM, said:

FCalc2 gives you the modern estimates of Davis A, B and C based on the original Davis equation and further Canadian work on various types of modern freight vehicles.These tend to be fairly good for most North American stock and particularly for heavy freight trains.
On the other hand FCalc2 values for European passenger stock, including British stock tend to be somewhat higher than those measured by various railway companies.There is also considerable variation when it comes to modern high speed trains and these in particular need to be based on the nearest available railway data.
Some explanation for this together with references are given in the documentation. Friction_Calculator.zip
(All that I have researched is concerned with the resistance of moving trains. Low speed resistance starting from a standstill is a different ball game again.)

Downloaded, Thank you darwins...very kind of you...will read/study. Please ignore PM about getting copy of friction calculator. Your explanation has similarities to the OR development differences between DC and AC traction models for diesel locomotives. --- in that the code was written by someone with DC traction expertise, or so it has been explained to me. Again, thank you.

[R H Steele]

NickonWheels, on 03 January 2020 - 10:02 AM, said:

Of course.
with traction/dynamic brake curves you can define the force even at zero speed but for friction at those speed ORTS just uses the vehicle weight and inserted bearing type and mixing it up with some quite unfair working equation (notice slow spped friction is always in order between 0-5 mph). When a similar curve to the traction list would work it´s easy to determine this crucial value yourself. Then you have the right to choose indefinately; that´s why I said weird traction values because you can them pretty unrealistic by inserting whatever you want, but in the end this is what makes the traction curves so useful. So a friction curve would be useful too; it´s basicly similar with speed on the left and newtons at right (speed dependacy just like Davis). I´m not really getting behind why ORTS could not have implemented such thing in the past, but I would say it´s nessesary for some time soon.

For your consideration...a little of the development history of OR... https://open-rails.r...st/physics.html --- manual is also loaded with information. ( maybe more up-to-date )