[Csantucci]

I would like to try to generate a couple of Train Control System scripts, and if necessary add parameters needed also to the OR code. I hope I've got enough programming knowledge to do that now. I'd like to start with the British AWS (unless someone doesn't tell me that this already exists), because at least the basic functionalities seem to be not too complicated.
So I'm asking if someone from the British world can point me to:
- .wav or .mp3 files for the "warning" and "clear" sounds of the bell;
- a link to a downloadable locomotive having a cab with the two-state sunflower.
Also links to an AWS spec or to some useful videos are welcome.

I foresee to implement only the basic functionalities.
Thank you for any contribution.

[Coolhand101]

Hi Carlo

https://www.railways...BTC_AWS1959.pdf

https://railcar.co.u...warning-system/

https://catalogues.r...2%20Iss%203.pdf

https://www.uktrains...orm_fileid=1750

https://www.uktrains...rm_fileid=30037

https://www.uktrains...orm_fileid=5650

https://www.uktrains...orm_fileid=5692

https://www.uktrains...orm_fileid=4488

https://www.uktrains...orm_fileid=4053

https://www.uktrains...orm_fileid=4054

https://www.uktrains...orm_fileid=2846

Guess it's time to update from XP to W10.

Thanks

[Coolhand101]

A couple of videos for now:

https://www.youtube....h?v=deCvub8gH6w

https://www.youtube....h?v=TNYEoex7Bps

[darwins]

Hi Carlo
Sorry if this is turns into a long message - I want to try to pull up a couple of very important points - so will give you a quick history.
(1) Great Western Railway ATC - this was the first mass installation in UK and the basis of the modern system
If the train approaches a caution signal showing clear - an electric bell rings briefly in the cabIf the train approaches a caution signal showing caution - the vacuum brake is applied and a horn is sounded (by air being sucked into the brake pipe)
The brake application made by AWS should take about 12 to 15 seconds to destroy the vacuum in the train pipe. (As opposed to about 1 to 2 seconds for an emergency application by the driver).
If the AWS warning is acknowledged the brake application is cancelled and the driver can restore the normal vacuum to the brake pipe again.
There is no visual warning in the GWR system.
The system only worked on running lines intended for running in one direction. It was not installed on single track routes. It was not installed at major stations and junctions where lines could be used by trains passing in both directions.Also it was not installed throughout the whole system. Hopefully at some future time in OR there can be a flag to show which signals have ATC or AWS installed.
(2) British Railways AWS
From the drivers point of view this works in an identical way to the GWR system with one addition.As well as a bell and horn there is a visual indication.
When the brake application is made and the horn sounds the black and yellow "sunflower" is shown. After the horn and brake application are cancelled the sunflower remains visible until the train approaches a caution signal with AWS giving a clear indication.(I believe there is also a mechanism to cancel the indication when passing from a route with AWS to a route without AWS.)
As air brakes replaced vacuum brakes the system has changed slightly. Now the horn sounds and the sunflower is displayed when a signal showing caution is approached.There is then a delay of 3 seconds before the brake application begins. If the warning is not acknowledged then a brake application is made.

The times taken for a full brake application to be made on a dual braked diesel locomotive are as follows:Vacuum braked mode 12-15 seconds for vacuum in train pipe to fall to zero.
Vacuum unbraked mode 30-40 seconds for vacuum in train pipe to fall to zero.Air passenger mode / Air goods mode is supposed to take 12-15s for Air Train Pipe pressure to drop from 70 psi to 45 psi (the loco test I have to hand took 22s)
These are the same application times as would be made by the deadman's device (DSD or DVD)
In more recent times coverage has become almost universal including single lines and major stations or junctions with two way working.

[Csantucci]

Thank you both. This is a good starting point.

[Csantucci]

I got a first result, so that the ping (when aspect = Clear_2) or the warning horn (in the other cases) are played at 200 yards before signal, and also the sunflower changes state there as wanted.
Now I have some questions:
1) does the warning horn continue to play until driver's acknowledgement?
2) do these trainsets have also a VACMA-type monitoring (driver must periodically activate something, else locomotive goes into emergency braking?) And if yes, are the VACMA reset command and the warning horn acknowledgement commanded by the same driver input (pushbutton)?
3) What occurs if the train SPADs?
4) If the driver does not acknowledge the warning horn in time the train will stop; to restart the driver has to push a different button; is this correct?

Thank you for any answer.

[darwins]

Csantucci, on 24 January 2020 - 01:33 PM, said:

I got a first result, so that the ping (when aspect = Clear_2) or the warning horn (in the other cases) are played at 200 yards before signal, and also the sunflower changes state there as wanted.
Now I have some questions:
1) does the warning horn continue to play until driver's acknowledgement?
2) do these trainsets have also a VACMA-type monitoring (driver must periodically activate something, else locomotive goes into emergency braking?) And if yes, are the VACMA reset command and the warning horn acknowledgement commanded by the same driver input (pushbutton)?
3) What occurs if the train SPADs?
4) If the driver does not acknowledge the warning horn in time the train will stop; to restart the driver has to push a different button; is this correct?

Thank you for any answer.

( 1 ) Yes. The horn continues to play until acknowledged by the driver.
(a) Steam Locos - brake is being applied as soon as horn begins to sound - the application is not "emergency" as it is slower taking 12-15s for vacuum to drop to zero.(B) Diesel and Electric locos - horn plays for 3 seconds before brake application begins - brake is then applied as for steam.
( 2 )(a) Steam Locos - No. There is no DV device. AWS works alone.
(B) Diesel and Electric - there is a DV device - this takes the form of either
(i) a dead man's handle which must be held down continuously or
(ii) a dead man's treadle which must be depressed periodically (in response to an alert) if no other control is moved during that time interval.
The AWS cancel button is separate to the DV device, it is not the same input.
The DV device however causes the same type of brake response as the AWS. There is a three second pause, then the brakes are applied over 12-15 seconds (for vacuum to drop to zero or ATP to fall from 70 to 45 psi).
( 3 ) Nothing happens if the trains SPADS. AWS is operated only by caution signals and only in the way mentioned above. ( a ) For semaphore stop signals and two aspect stop signals - that is only a stop signal - that means nothing at all happens. ( b ) For multiple aspect signals then i believe that as any stop signal is also a restrictive aspect and will operate the AWS in the same way as a caution signal - perhaps coolhand can confirm.
( 4 ) No. The driver only has to cancel the AWS using the acknowledge button. Once this is done the driver once again has full control. This can be done at any time - the train does not need to stop.

In some files I have used a crude copy of the driver vigilance device from MSTS: VigilanceMonitor(
MonitoringDeviceMonitorTimeLimit( 45 )
MonitoringDeviceAlarmTimeLimit( 37 )
MonitoringDevicePenaltyTimeLimit( 0 )
MonitoringDeviceCriticalLevel( 0 )
MonitoringDeviceResetLevel( 0 )
MonitoringDeviceAppliesFullBrake( 1 )
MonitoringDeviceAppliesEmergencyBrake( 0 )
MonitoringDeviceAppliesCutsPower( 0 )
MonitoringDeviceAppliesShutsDownEngine( 0 )
)
Although I am not sure it has the braking time correct it performs more or less the function required. I have used - MonitoringDeviceAppliesCutsPower( 0 ) - because the monitoring device does not cut the power, but a drop in the vacuum or pressure in the train brake pipe below a certain level will cut the power.
On steam locos the full brake application was considered sufficient to stop the train even if the regulator was still open.
Thanks
Darwin

[Coolhand101]

For UK diesel locos, the Class 56 was the first to have a Driver's Vigilance Device D.V.D. It had a longer time interval than the AWS, usually around 5 - 7 seconds before the brakes applied. This only applied if the driving controls had not been move for 60 seconds. I'm unsure about the first A.C locos which had this fitted( Probably the Class 87 and upwards ).

All electric/diesel stock, had a 'dead mans' device, as Darwins mentioned, either on the power handle or a treadle( foot ).With EMUs/DEMU, the brake application was instant, if the handle was released. For DMMU, there was around a 5 second delay so the driver can change hands. For diesel and electric locos, I'm unsure if there was a delay or not!

I used this for locos/units that have a D.V.D:

MonitoringDeviceMonitorTimeLimit( 67 )
MonitoringDeviceAlarmTimeLimit( 60 )
MonitoringDevicePenaltyTimeLimit( 60 ) ---- timeout period after the train was brought to a stand.
MonitoringDeviceCriticalLevel( 0 )
MonitoringDeviceResetLevel( 0 )
MonitoringDeviceAppliesFullBrake( 0 )
MonitoringDeviceAppliesEmergencyBrake( 1 )
MonitoringDeviceAppliesCutsPower( 1 )
MonitoringDeviceAppliesShutsDownEngine( 0 )
)

Back to AWS:

Semaphore Distant, Colour light distant signals, 75% of British routes had AWS for these signals by the 1970s, the A.W.S was slowly introduce from the 1960s*(* citation needed ).

Semaphore stop signals and 2-aspect colour stop signals, had no AWS for absolute block(AB) lines.

2-Aspect colour light stop and Colour light repeater signals for Track Circuit Block( TCB )had AWS.

To recognize 2-aspect colour light stop signals on AB lines, perhaps the script can check if a distant signal was used prior to the CL stop signal? or by a form of distance, finding a distant signal which applies to the colour light stop signal?.

There was no AWS leaving a terminus station or siding*.( Perhaps start/end of track nodes could use this method ?). Also, no AWS at any form of shunt signals. *Today's railway, some depots/sidings do have a AWS magnet permanently set to a warning, to test the AWS before entering service.

AWS was also used for advance speed restriction boards( Morpeth boards ). Before the 1990s, this was for line speeds above 75mph( This was sometimes lowered to 70mph line speeds ) where a 25mph+ speed reduction was needed. On today's railway, it can be any line speed, when a 25mph+ reduction was needed. ie 60 mph to 35 mph.( At the moment, I use a 'warning' track sound source 200 yards before these boards below):
http://www.atomic-album.com/showPic.php/178898/Open Rails 2020-01-25 12-27-01B.jpg

Regarding the AWS indicator in the cab. When the 'sunflower' indicator was shown when approaching a signal aspect other than green, the sunflower when momentary go to black, and once the driver cancel the AWS warning horn, the sunflower indicator would re-appear. Also there was a noticeable sound, ( low frequency 'click') when the sunflower changes from caution to black.

Also can the script allow shapes files to be added at signals which have AWS?. An AWS ramp shape file could then be automatically added on the track, 200 yards before the signal?

On 2nd generation trainsets with sliding doors and later, the train had to stop before the brakes release, if the driver missed the AWS reset button ( westcode brake ). For older stock, the brake pipe started to charge once the AWS reset button was pressed!


Thanks

[darwins]

Quote

Also can the script allow shapes files to be added at signals which have AWS?. An AWS ramp shape file could then be automatically added on the track, 200 yards before the signal?

That might be drifting into route editor territory, but it would be brilliant if the AWS ramp and signal were linked. Then if there is an AWS ramp the AWS is activated in the cab.
If there is no AWS ramp then either that signal does not have AWS ( semaphore or two aspect AB home ) or the route does not have AWS - and so no AWS activity in the train.
It would be a big job to apply for the future though or to apply to pre-existing routes.
On locos built until at least the 1980s neither the Deadman's device or AWS would give an "emergency" application. In fact the application time was around 3 to 4 times slower than if the driver had moved the brake handle directly to the "full service" position.
Nor would either Deadman's or AWS cut power directly. They both applied the brake and the drop in brake pipe pressure would cut power (if the loco was under power) controlled by the vacuum or air governor.

Vacuum brake

Cuts power when vacuum falls below approx 12.5 inches ( in train pipe ) and power can not be restored until vacuum rises above 16 inches.


Air brake

Cuts power when train pipe pressure falls below approx 46.5 psi and power can not be restored until ATP pressure rises above 60 psi.

Looking at loco docs then it seems class 87 had the same deadman's as 81-7 (with a 5 second delay). So perhaps class 90 would be the first electric with an "alerter".

So clearly there are some differences between older and newer diesel and electric trains, as well as differences with steam.

[Coolhand101]

darwins, on 24 January 2020 - 11:58 PM, said:

That might be drifting into route editor territory, but it would be brilliant if the AWS ramp and signal were linked. Then if there is an AWS ramp the AWS is activated in the cab.
If there is no AWS ramp then either that signal does not have AWS ( semaphore or two aspect AB home ) or the route does not have AWS - and so no AWS activity in the train.
It would be a big job to apply for the future though or to apply to pre-existing routes.
On locos built until at least the 1980s neither the Deadman's device or AWS would give an "emergency" application. In fact the application time was around 3 to 4 times slower than if the driver had moved the brake handle directly to the "full service" position.
Nor would either Deadman's or AWS cut power directly. They both applied the brake and the drop in brake pipe pressure would cut power (if the loco was under power) controlled by the vacuum or air governor.

Vacuum brake

Cuts power when vacuum falls below approx 12.5 inches ( in train pipe ) and power can not be restored until vacuum rises above 16 inches.


Air brake

Cuts power when train pipe pressure falls below approx 46.5 psi and power can not be restored until ATP pressure rises above 60 psi.

Looking at loco docs then it seems class 87 had the same deadman's as 81-7 (with a 5 second delay). So perhaps class 90 would be the first electric with an "alerter".

So clearly there are some differences between older and newer diesel and electric trains, as well as differences with steam.

You be amazed what a script can do. It may be able to look in the OR directory of a route, find the shape and automatically add it 200 yards from the signal in that route for signals that use AWS.

The deadman/AWS on locos, did indeed apply the rate of the full service application( Some locos used an emergency rate for the AWS). I did this for the alerter in MSTS using the BIN patch for the make shift AWS.
On the older slam door EMUs, the deadman release was the quickest way to apply the Westinghouse brake, and I'm 99% sure the AWS application used the emergency rate!

Yes the brake pipe governor causes the power to be cut. If you multiply the full service pressure drop by 2.5x, you can set the 'brake cuts power for brake cylinder' token between 50 to 65 to simulated this in OR. It would be realistic if the brake pipe did do this in OR. However, for M.Us with the Westcode brake, it was the brake cylinder that caused the power to be cut above Brake Step 1( 1.2 bar).

Thanks