| <railMLv3> Infra Geometry Terminology [message #2021] |
Fri, 23 November 2018 13:01  |
Fabrizio Cosso
Messages: 12
Registered: September 2017
|
Junior Member |
|
|
Dear all,
can someone address me to some documents where I can find the exact meaning of terminology used in Geometry View?
e.g. the exact meaning of each curveType, etc...
Thanks
BR
Fabrizio
|
|
|
|
| Re: <railMLv3> Infra Geometry Terminology [message #2024 is a reply to message #2021] |
Mon, 26 November 2018 16:24  |
christian.rahmig
Messages: 474
Registered: January 2016
|
Senior Member |
|
|
Dear Fabrizio,
Am 23.11.2018 um 13:01 schrieb Fabrizio Cosso:
> can someone address me to some documents where I can find
> the exact meaning of terminology used in Geometry View?
> e.g. the exact meaning of each curveType, etc...
I suggest to use the UML model documentation (HTML report, available in
[1]) or the XSD schema documentation (HTML, available in [2]). The
railML 3 wiki, where besides the automatically generated schema
documentation detailed definitions, best practices and open issues shall
be adressed, is still under construction.
Regarding your specific question with the curve type:
Generally, railway alignment knows three basic curves:
* straight lines (constant infinite radius)
* arcs (constant radius)
* transition curves (changing radius)
railML introduces in <horizontalCurve>@curveType some more specific
versions of transition curves, e.g. clothoides, cubic parabolas etc.
Some of these curve types are described in more detail e.g. in [3] or [4].
Once, we start with the use case "Track Geometry" [5], we may have a
closer look on these different types.
Best regards
Christian
[1] https://svn.railml.org/railML3/tags/railML-3.1-rc/doc/model
[2] https://svn.railml.org/railML3/tags/railML-3.1-rc/doc/schema
[3]
https://www.brainkart.com/article/Railway-Engineering-Transi tion-Curve_4227/
[4] https://pwayblog.com/2016/07/03/the-clothoid/
[5] https://wiki.railml.org/index.php?title=UC:IS:Track_Geometry
--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Christian Rahmig – Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|
|
|
| Re: [railML3] Infra Geometry Terminology [message #2160 is a reply to message #2153] |
Wed, 27 March 2019 07:19 |
christian.rahmig
Messages: 474
Registered: January 2016
|
Senior Member |
|
|
Dear Adam,
Am 20.02.2019 um 17:58 schrieb Adam Hlubuček:
> in relation to terminology, I would like to ask what does
> the value "mixed" of rail3:tGradientCurveType mean and when
> to use it. The semantics seems quite unclear to me. Thank
> you for your feedback.
sorry for answering so late, but I just recovered from 5 weeks of
illness, so that I missed your posting.
A <gradientCurve> with @curveType="mixed" can be used to model a
gradient curve with an "administrative" gradient value. Imagine the
situation that you have a track from A to B and somewhere on this track
a maximum slope of 25 per mille has to be climbed by the train, but you
don't know exactly, where this gradient appears. In that case, you can
model a <gradientCurve> covering the whole track from A to B with
@gradient="25". By setting @curveType="mixed" it becomes clear, that the
gradient of 25 per mille does not affect the whole track from A to B.
If the community knows further examples where different types of
gradients shall be modelled, please share them here with us in the forum
so that we can discuss them and integrate them in planned Advanced
Example and in the Track Geometry use case.
Thank you very much and best regards
Christian
--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Christian Rahmig – Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|
| Re: [railML3] Infra Geometry Terminology [message #2164 is a reply to message #2160] |
Mon, 01 April 2019 13:24 |
Thomas Nygreen JBD
Messages: 68
Registered: February 2017
|
Member |
|
|
Dear all,
The gradient information that we use in Norway can be generalised into two types:
Actual microscopic gradient: only used in an IM perspective (infrastructure models, construction, maintenance, etc.)
Averaged gradient over a certain length: used for operational and signalling purposes. The signalled gradient is a moving average over 1 km. The schematic signal plans show average gradient between signals in the same direction.
Additionally, I would guess that a maximum gradient could also be useful. If so, we should be able to distinguish between types of "mixed" curves, i.e. the aggregation method (average, maximum and maybe more).
Best regards,
Thomas Nygreen
Railway capacity engineer
Jernbanedirektoratet
[Updated on: Mon, 01 April 2019 13:24] Report message to a moderator |
|
|
|
| Re: [railML3] Infra Geometry Terminology [message #2166 is a reply to message #2164] |
Mon, 08 April 2019 16:52 |
christian.rahmig
Messages: 474
Registered: January 2016
|
Senior Member |
|
|
Dear Thomas,
Am 01.04.2019 um 13:24 schrieb Thomas Nygreen:
> [...]
> Averaged gradient over a certain length: used for
> operational and signalling purposes. The signalled gradient
> is a moving average over 1 km. The schematic signal plans
> show average gradient between signals in the same
> direction.
>
> Additionally, I would guess that a maximum gradient could
> also be useful. If so, we should be able to distinguish
> between types of "mixed" curves, i.e. the aggregation method
> (average, maximum and maybe more).
thank you for your valuable input from a Norwegian point of view!
From technical side, two possible solutions are possible:
1) extending the value list for <gradientCurve>@curveType replacing the
value "mixed" with new values "maximum" and "average".
2) adding a new attribute <gradientCurve>@aggregationReference with
values "maximum" and "average", which is only set in case of
<gradientCurve>@curveType="mixed".
I prefer solution 1), but the more interesting question for me is if
there are further types of "mixed" gradients apart of maximum and average.
Any feedback is highly appreciated...
Best regards
Christian
--
Christian Rahmig - Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Phone Coordinator: +49 173 2714509; railML.org: +49 351 47582911
Altplauen 19h; 01187 Dresden; Germany www.railml.org
Christian Rahmig – Infrastructure scheme coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|
| Re: [railML3] Infra Geometry Terminology [message #3339 is a reply to message #2166] |
Wed, 16 October 2024 17:16 |
Torben Brand
Messages: 169
Registered: March 2016
|
Senior Member |
|
|
In railML3 the element <gradientCurve> is meant for the modelling of the vertical alignment. [1]
The vertical alignment can be modelled in different ways. This forum post will suggest how to model them unambiguously in railML3. Thus this forum post consist of three items:
1. a guide of how to use <gradientCurve>
2. Semantic constraint to be added
3. Definition of @length
The consensus of this post should then be posted on the <gradientCurve> wiki page [1].
Gradient curves can be vertical alignments of type: slope, vertical curve, compositions curves, vertical transition curve. As the alignment is a rather visual item. I have made an illustration that is attached.
The vertical alignment can be modelled in (at least) 4 ways:
1. Full modelling of the vertical alignment with all slopes and vertical curves
2. Simplified modelling of the vertical alignment with all slopes, but no vertical curves.
a) The slopes are extended to a theoretical crest (hilltop) or sag (bottom of valley)
b) The slops are not extended. The gaps use curveType="mixed"
3. short fixed length discrete slopes (usually from measurement; example value 20m)
4. average gradient over a longer length of track.
The question is:
1. when to use which value for @curveType
2. which attributes (@gradient, @radius and @deltaGradient) to use in combination with the chosen @curveType and if there should be semantic constraints for those
3. Should there be additional attributes to describe the type of vertical model?
For question 1 see attached pdf with illustrations and table for modelling types and suggested usage of @curveType in combination with attributes @gradient, @radius and @deltaGradient
The alternative usage of @curveType is for defining @curveType to describe the curve in the representation and not in reality. The suggested usage of @curveType is to describe reality. The usage of @gradient (in combination with curveType="mixed") then shows that the representation is a straight line.
PS. This means the current semantic constraint in the wiki [1] has to be removed. I then also suggest adding new semantic constraints based on the table above to make the usage unambiguous:
If curveType="arc" then use @radius
If curveType="straight" then use @gradient
If curveType="mixed" then use @deltaGradient or @gradient
The usage of "mixed" with @gradient requires a change in the definition of @gradient? Either change to:
a) gradient:"constant or average gradient of the gradient curve in per mille; ..."
b) gradient: "constant gradient at the start of the gradient curve in per mille; ..."
In b) you then get the gradient at the end of the section by adding @deltaGradient, or the average by adding half @deltaGradient.
I would recommend option a)
It would be useful to be able to describe the type of modell (as discussed previously/above in this forum post). We suggest either a simple approach with usage of new attribute of type boolean @aggregated (alternative"linearized")=true/false or of a type enumeration value: @modelType= "design"/"approximation"/"linearized"/"averaged"
The quality of the modell is a different topic.
The attribute gradientCurve@length needs to be defined more precise. Which length are we referring to? Is it the element length, the crest or the tangent? I would assume the element length. I was discussed in SCTP working group 11.10.2024 and we concluded it should be the element length.
It was also discussed if the length then would be redundant as the element length of the <gradientCurve> is given in the sub-element <associatedNetElement> under <linearLocation> of the <gradientCurve>. But this would not always contain the necessary information and mileage changes would have to be considered. Thus we concluded to keep gradientCurve@length (which is optional)
A further question that I got after the meeting is what if a track and subsequent gradientCurve branches out. Is the length then for all tracks or only the main track?
If I forgot to note down comments from the SCTP meeting discussion 11.october or if you have some other comments please feel free to comment here.
[1] https://wiki3.railml.org/wiki/IS:gradientCurve
|
|
|
|
Members
Search
Help
Login
Home
