Elvas Tower: Curtius Kniffler Adhesion parameter - Elvas Tower

Jump to content

  • 5 Pages +
  • 1
  • 2
  • 3
  • 4
  • 5
  • You cannot start a new topic
  • You cannot reply to this topic

Curtius Kniffler Adhesion parameter Rate Topic: -----

#21 User is online   Weter 

  • Member, Board of Directors
  • PipPipPipPipPipPipPipPipPipPip
  • Group: ET Admin
  • Posts: 6,806
  • Joined: 01-June 20
  • Gender:Not Telling
  • Simulator:ORTS
  • Country:

Posted 27 July 2021 - 07:13 AM

ErikC, thanks a lot for that report-very interesting.
I just would add, that AC generators are smaller and much more reliable, also demands less time for service.
More vulnerable part of DC machines-the collector, so AC-to-AC transmission is more efficient in this sense too.
And, of course, solid state control long system makes possible modulation of tractive effort, as it was told here.

#22 User is offline   steamer_ctn 

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

Posted 27 July 2021 - 08:51 PM

 darwins, on 27 July 2021 - 01:09 AM, said:

Does this have any value? In this case I adjusted both A and C values, but found that to do this for some locos the B value would need to be changed as well.
In most publications that I have seen, the adhesion coefficient is a "family" of parallel curves, hence the recommendation to just vary the C value.


 darwins, on 27 July 2021 - 01:09 AM, said:

Is this an alternative to the use Antislip ( ) in OpenRails?

Would it be more realistic to get Antislip ( ) working correctly for such locos where you would see the power being reduced (without the throttle moving!) and then increased again until the throttle setting could be achieved?

In terms of the antislip technology based on wheel rotation speeds I was reading yesterday of an example where this was applied to individual axles rather than the whole locomotive.
Improvements in adhesion can be due to a number of different factors.

For example, this paper describes the early attempts to improve adhesion by designing high adhesion bogies that shifted weight around.

This paper describes some methods that could also improve adhesion.

In short to model anti-slip (adhesion) improvement accurately in OR would be extremely complex, and would require significant recoding to build a track model, wheel model and different anti-slip algorithms.

Antislip ( ) probably only represents one factor potentially improving adhesion.

 darwins, on 27 July 2021 - 01:09 AM, said:

Is adjusting the Curtius-Kniffler adhesion the right way to model these?

Nearly all the publications that I have studied indicate that the final outcome of ALL anti-slip control measures is an improved adhesion value. Hence I think that adjusting the adhesion value to suit the locomotive is a valid way of representing improved slip control.

#23 User is offline   darwins 

  • Superintendant
  • Group: Status: Elite Member
  • Posts: 1,222
  • Joined: 25-September 17
  • Gender:Male
  • Simulator:Open Rails
  • Country:

Posted 28 July 2021 - 12:50 AM

I have just tried some driving in slippery conditions and although it may well be true to say that

Quote

the final outcome of ALL anti-slip control measures is an improved adhesion value

this does not give the most realistic driving experience, I think that adhesion is one aspect of simulating this, but I also think that we need to get Antislip() working properly.


What changing the ORTSadhesion ( ) does


Improvements in bogie design and control technology on modern locomotives lead to greater adhesion, allowing greater tractive effort in the same rail conditions.

This means that you can push the throttle further before you get first a wheelslip warning and then wheelslip, but keep pushing the throttle and you will get them.


On the other hand what should Antislip ( ) do

In the above case even if you continue pushing the throttle up, Antislip ( ) should reduce the power to the wheels as soon as a slip warning is given - making actual slip difficult or impossible.

For older locos that have older bogies but have an antislip system - then we would not expect an improvement in adhesion, we certainly would not expect to get more tractive effort. What should happen is simply that power to the traction motors will be reduced (at the normal rate of reduction) as soon as a slip warning is given or will not increase to the throttle position until adhesion is improved and the slip warning has gone.

#24 User is offline   ATW 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 638
  • Joined: 07-January 13
  • Gender:Male
  • Simulator:MSTS Open Rails
  • Country:

Posted 28 July 2021 - 10:23 AM

I can tell you with all the flash floods an random down poors in my working areas these modern AC motors climbing long heavy grades have showed its powers over coming from stalling by managing slippage by Sand, Rail Cleaner, automatic temporary power reduction for some seconds. My last trip I had was almost a scare of doubt we were making it up the hills curves of Beaumont Hill with our 3 mile train because we dropped from 13mph to 9MPH when rain started that we felt too much slack shaking action for quite the miles I thought we may break a knuckle even with a single loner cut in DPU. But with power of automatic sanding an the rail cleaner turning on like a hand blow dryer everytime speedometer dropped to 9mph in rain. It almost feels as if AC motors are managing its own adhesion factor an random change when it doesn't want to allow constant slipping an huge power reduction drops to stall.

So with today's modern slip control technology an options in the computers to turn on automatic sanding, automatic rail cleaner or manually I say we need a number of options to be automatic instead of always manual like sanding. I look forward to the option of anti slipping technology to reduce traction power for some chosen seconds an rising back up slowly to throttle setting instead of ORTS reducing the Throttle handle an you having to power back up by yourself.

#25 User is offline   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 28 July 2021 - 12:02 PM

 ATW, on 28 July 2021 - 10:23 AM, said:

... I look forward to the option of anti slipping technology to reduce traction power for some chosen seconds an rising back up slowly to throttle setting instead of ORTS reducing the Throttle handle an you having to power back up by yourself.

Yes agreed, a computerized system has the ability to respond more rapidly and with more control than even the most experienced operator. I'm sure though that this is not an easy option to code and implement within the existing OR framework.


#26 User is offline   darwins 

  • Superintendant
  • Group: Status: Elite Member
  • Posts: 1,222
  • Joined: 25-September 17
  • Gender:Male
  • Simulator:Open Rails
  • Country:

Posted 29 July 2021 - 02:07 AM

After some further explanation from Peter, I tried to calculate the Curtius-Kniffler 'C' parameter for British diesel electric locomotives. Here are the results:
https://i.imgur.com/giuc4wd.png

The ORTS default value is 0.131. All of the older diesel types come out with lower adhesion that that, generally around 0.11 to 0.12. Whilst improvements in starting adhesion in USA began in the early 1970s, they do not seem to have arrived in UK until we imported our first US built heavy freight locos in 1985 (Class 59). Recent heavy freight locos (except class 67) all have a clear improvement in adhesion due to new technology. Most notable amongst these is the class 68 - an 85 tonne Bo-Bo, geared for a top speed of 160 km/h.

An interesting observation is the seemingly very low adhesion of the two locomotives geared for 200 km/h (Class 43 HST and Class 67). Do they really have lower adhesion than locos geared for lower speeds? Or is the tractive effort of these locos limited by control of traction current rather than by adhesion?

#27 User is offline   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 July 2021 - 05:23 AM

Interesting, I've learned quite a bit from the thread.

#28 User is offline   Aldarion 

  • Engineer
  • PipPipPipPipPip
  • Group: ET Owner
  • Posts: 619
  • Joined: 11-February 13
  • Gender:Male
  • Location:Lisbon, Portugal
  • Simulator:Open Rails
  • Country:

Posted 29 July 2021 - 01:52 PM

I'm going to give my point of view about adehrence in rail vehicles. I am seeing much discussion in what i find a very theoretical point of view, and I think I have a more practical point of view from my practice in calculating the movement of trains. I apologise in advance if I do not use terminology in a correct manner... and also by trying to xplain things very basicaly so many others can follow.

when you applied a momentum (M) to a wheel supporting a certain weight (P), making it spin around it's center (O), the wheel will only progress over it's suport basis ( the infrastructure, in this case, the rails ), and having into account the friction coeficient (f) between the two surfaces ( in our particular case, steel wheel and steel rail), when the periferal force (F) originated by the momentum obeys the followinf condition

F<=f*P

This is the fundamental adherence equation, translating the necessary conditions for any vehicle to react on the infra-structure it runs on, in controled conditions. (loss of adherence means an uncontroled condition).
Ence, F is the maximum tangent effort that must be delivered to the wheel, before it slips, turning the progress of the vehicle unstable and uncharacteristic of normal movement.
https://i.ibb.co/8rwmr3f/roda.jpg
R- wheel radius,
P- weight on the axis, applied at the point of contact between wheel and rail
N- normal reaction force to P - Newton's 3rd law.
M- rotation momentum
F- tractive force aplied at the point of contact between wheel and rail

So, to a wheel with R radius, supporting P weight, should not be aplied a momentum M greater than f*R*P or else it will slip. This has consequences when esigning the locomotive and establishing the nominal power availabe for traction, and of course it is importante to distinguish between tractive wheels and non-tractive wheels. On that account I have seen ORST values for it to be defined in eng files. moving ahead...

Being E the sum of all F force contributions of all wheels of a locomotive (the tractive effort), then the fundamental adherence formula is

E<=f*P

where P is the adherent weight of the locomotive. If all axis are motor axis, then P equals the entire weight of the locomotive.
But what happens in practical termns is that the progression of a locomotive is only compromissed if all wheels enter in a slip state at the same time, when E is very high. That is not very frequent to happen, meaning that we have the following:

E/P=@>f

where @ is the adherence coeficient, whith no sliping. This is essencially an experimental based coeficient, I mean, you define it experimentally.


So, we can write, for any motorized vehicle: E <= @*P
this is the basis and fundamental equation relating to locomotion mechanics in surface transport.

In the case of road vahicles f depends on the surface roughness of the pavement, hmidity of said surface and the carachteristics of tyres.
In a modern dry road values can vary between
0,5-0,9 depending of vehicle speed, but in a wet surface thay can be as low as 0,25-0,35.

On conventional railways, f is in the order of 0,20-025 but today's standards can easely raise them to 0,33-0,35. In wet rails, however it may be 0,10 or less.

Concerning the study of adherence and friction in the field of physics, I have to point out that inicially, it was thought it was due to the roughness ( at a molecular level ) of surfaces. Today it is well known that polished surfaces du present friction between each other, due to molecular atractions forces. These are variable depending on the distance beteen the contacting surfaces and the state in wich they are. Acording to studies by Hertz and Boussinesq ( aplying the theroy of elasticity ) they verified that the contact surface between wheel and rail is an elips and that the pressure, being maximum at the center can be as much as 90daN/mm2. This is a value way above the elasticity limit of steel however there is no permenent deformity and it is believed that there is molecular interpenetration between the two surface creating elevated molecular atraction forces ( Verbeek, Rail Internacional, May 1973). of course the area depends on the mean pressure and the wheel radius.
So in essence it is all about the relation between E (on the wheel-rail contact) and the adherent weigh of the locomotive, dependent, as we all know, the state of the rails, roughness of rail head surface and also and curves not passed at the speed for with superelevation was established, as far as the infrastructre is concerned, but also change in the weigh on the wheels by applying tractive effort, oscilations of the vehicle due to suspensions, lateral movements that provoce minor laetral slippage of wheels, irregularities of torque or sudden changes in tractive effort s far as the vehicle is concerned.

Because all of these it is practical to adopt a friction coeficient obtained experimentaly, observed and confrimed in test trials under different conditions.

For a train to initiate movement, the E tractive effort at the wheel must be suficient to win oever resintences to movement.
If R>P*@, the wheels will spin but they will not progress, just rolling over the rail. At that point the reaction of rails and wheel is defined by P*f' , f' being the friction coeficient between moving surfaces- yes - a rolling wheel on a suface has it's elements static when in contact wyth the rail surface, hence a static friction coeficient.

https://i.ibb.co/tKLzwQ2/coeficient.jpg

_______ static friction coeficient (no slipage)
_ _ _ _ dinamic friction coeficient (sliping wheel)

It is aklso noticable, that as long as the wheel starts moving over the rails surface the friction coeficient (and there fore the adherence coeficient) drops. I wont get over the causes of that but I want to share the empirical formula to calculate the variation of f due to velocity V

f(V)= f / ( 1+0,01V)

f comming from the friction coeficient at nill velocity. It is commonly accepted that the adherence coeficient @ follows the almost exact rule.
This variation is visible in this graph:
https://i.ibb.co/74PSK7Z/variacao.jpg
As you can see at 100kph the coeficient is down by about 50%

The maximum effort to apply to the wheel can be defined as:

F= f*P = P*f / 1+0,01V

and the available power at the wheel rim is then

W= (100/27) * (P*f*V / 1+0,01V) in horsepower, bein P in metric ttons and V in kph.

So, the tractive effort can be defined by this:

E = Px@ / 1+0,01V

If at any moment E>P*@ (either because E was raised or @ diminished ) there will be wheel slip defined by P*f', until it becomes inferior to P*f. there fore, E must be reduced, and @ must be raised, by using sanders.

Hope these equations are usefull and the explanation is clear. I really tried to shorten it...

#29 User is offline   steamer_ctn 

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

Posted 29 July 2021 - 07:47 PM

 Laci1959, on 26 July 2021 - 03:22 AM, said:

Thanks for the table. Very useful. Perhaps it would be a useful addition to an imperial-metric shift formula.

I have updated the spreadsheet to cater for imperial or metric values, and incorporated some other feedback that I was given.

One other point that I would make based upon the description in paper that I referenced above.

Fig 4 in this paper indicates that a locomotive can be either horsepower or adhesion limited. So if a locomotive is calculated as having an adhesion value less then the default, it will not always be a given that the adhesion should be reduced. It merely means that the locomotive may not be able to take advantage of the available adhesion due to a low power output.

EDIT: Spreadsheet has another minor tweak.

#30 User is offline   darwins 

  • Superintendant
  • Group: Status: Elite Member
  • Posts: 1,222
  • Joined: 25-September 17
  • Gender:Male
  • Simulator:Open Rails
  • Country:

Posted 29 July 2021 - 10:05 PM

Thanks Aldarion for the explanation.
Thanks Peter for the updated calculator.


  • 5 Pages +
  • 1
  • 2
  • 3
  • 4
  • 5
  • 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