Know your horses

[jamespetts]

There follows some useful information for players in the pre-motor-car era of the game where horses are an important form of transport. In the Bridgewater-Brunel server game, I often notice players using unsuitable types of horses for various tasks, causing their transport to be unnecessarily inefficient. This information is intended to help players to pick the right type of horse for the task in hand.

Types of horse

There are two basic types of horse:

(1) the draught horse; and
(2) the coaching horse.

Each have their own strengths, weaknesses and ideal applications.

Draught horses

The draught horses in Pak128.Britain-Ex are:

(1) the Irish Draught horse;
(2) the Shire horse; and
(3) the Clydesdale horse.

Draught horses are stocky and strong, good at hauling heavy loads up difficult terrain. They are not fast, but they can pull a good load at a modest speed over long distances without tiring.

Ideal uses for draught horses

- Hauling goods carts and wagons on the road
- Hauling canal boats
- Hauling heavy freight wagons on waggonways (early railways)
- Hauling heavy local passenger omnibuses that carry a large number of passengers at a modest speed
- Hauling local trams that carry a large number of passengers at a modest speed

Coaching horses

The coaching horses in Pak128.Britain-Ex are:

(1) the Feresian horse;
(2) the Hackney horse;
(3) the Cleveland Bay horse; and
(4) the Yorkshire coach horse.

Coach horses are fast and slender. They are good at hauling modest loads at high speeds, especially over relatively easy terrain. However, because they travel at higher speeds, they tire easily and need to be changed/rested after a time, so their range is limited.

Ideal uses for coaching horses

- Pulling stagecoaches on roads (usually in teams of 4 or 6)
- Pulling short stagecoaches on roads (usually in teams of 4)
- Pulling stagecoaches on waggonways (early railways; usually singly)
- Pulling mail coaches on roads
- Pulling flyboats on canals

Horses and rails

Because vehicles that run on rails have much less rolling resistance than vehicles on a road, horses pulling a carriage or tram on a waggonway/railway can haul it much more easily than the same load on an ordinary road. This means that fewer horses are required for the same load, or a higher load can be pulled with the same number of horses. A rail stage coach, for example, can operate as an ordinary stage coach, but with a single horse instead of a team of 6. A horse tram can carry far more people than a horse bus with the same number of horses and at the same speed.

The pony

This is a miscellaneous category of horse. It is intended to be used to haul small carts of goods or mail where it is not economically viable to use other types of horse. It is neither strong nor fast, but it is cheap. Players might find ponies useful for, e.g., carting small quantities of goods from docks or railway stations to local shops, or carting very small quantities of goods from multiple farms to a factory or market in a nearby town.

[freddyhayward]

I have generally used Irish Draught horses for stagecoaches, because the improved speed of other horses tends to be negated by slower climbing hills and being delayed by slower traffic, although I don't know how large this effect is.

[Ranran]

Thanks for the helpful information.  I've only used horses with the highest traction so far 

[jamespetts]

I have generally used Irish Draught horses for stagecoaches, because the improved speed of other horses tends to be negated by slower climbing hills and being delayed by slower traffic, although I don't know how large this effect is.

You should use a team of 6 Feresian horses for stagecoaches - unless the hill is very steep, these should be able to manage it at a sensible speed, even if not maximum speed. Unless your route is extremely hilly, most time will be spent on the flat, and thus the superior speed of a coach horse on the flat will be much more important than any improved speed of a draught horse on a hill.

Speed is now more important for passenger transport than it was until the most recent changes owing to the reduction of average journey time tolerances.

[Mariculous]

For the same reasons as freddy mentioned, I don't consider coaching horses to be worth it currently.
On intercity lines, you will be stuck behind an 8 km/h private coach anyways and within cities it's often worth more to pull up a slope at reasonably speed, instead of 1 km/h as 6 Fresians do as soon as there is more than a single half-height slope involved.

So there is no reasonable advantage in using fresians at cost of lower slope speeds and higher costs.
The only exception where I actually consider Fresians preferable over Irish draught is on intercity lines with dedicated private roads.

Anyways, I do also use 6 Fresians on intercity lines sharing roads with slow private coaches, but that's rather because James complained about my Draughts early on, than for economical reasons.
Might be the economics changed now, I'll need to reconsider building more dedicated private roads.

[jamespetts]

I have now temporarily modified the speed of the private horse carts to 12km/h. The long-term solution is a better system for overtaking, however.

[Vladki]

Regarding the comparison of road and rail transport pulled by horses. There is record from opening the horse railway Budweis - Linz. The builders claimed that horse will be able to pull 10x more load on railway than on road. A test was done and the result was even better: 30x more load. (on flat track). Another record - first ride with material for construction on the unfinished railway was 14 tons pulled easily by 2 horses for 30 km.
All around average weight pulled by one horse on rail was 3920 kg vs. 560 kg on road. However on the steepest part of track (2.17 %) two horses were required to pull one wagon with 2.5 tons of payload. Horse changing stations were approx every 20-22 km, horses did 2x20 km per day (both freight and passenger). Each stage took about 2 hours (passengers) or 4 hours (freight). Passenger train consisted of two carriages pulled by two horses.  On the steep part, the train had to be divided, and horses added - so only 1 car with 2 horses. I don't know what breed of horses was used.

BTW: multi-modal transport (Ro-Ro trains) is as old as that. Rich people could have had their own coaches put on flat railway wagon.
Also paying for overweight baggage was required on that railway. The final part of track were laid over the streets of Budweis so it was a tram as well.

[Milko]

Hello

The long-term solution is a better system for overtaking, however.

I agree, actually my convoys are always all in the queue behind very slow carriages and never overtake them. Am I wrong or does overtaking work some time ago?

Giuseppe

[Mariculous]

It does work, but there is a minimum speeed difference which must be exceeded, otherwise vehicles won't attempt to overtake.
If I recall correctly, that threshold is 10 km/h.
Further, the whole way required for overtaking needs to be in sight and cleared.

These two requirements result in overtaking happening veeeeeeeeeery rarely in early years.
Although, it can rarely be observed in stop-and-go situations on oneway roads, as these are not restricted by sighting.

[Ranran]

I increased the number of horses for freight and mail to run at 10km/h, but it seems that horses running at 6km/h often obstruct the running, and the cost for that is almost useless. 

[freddyhayward]

I increased the number of horses for freight and mail to run at 10km/h, but it seems that horses running at 6km/h often obstruct the running, and the cost for that is almost useless. 

I have also found this. Is overtaking intended to work for two-way roads? It currently only works for one-way.

[Vladki]

If I recall correctly, that threshold is 10 km/h.

How about changing the threshold to be in % ? Or a different threshold for speeds < 20 km/h ?

[wlindley]

Perhaps change to: A can overtake B, if A's target speed is greater than B's speed shift-left-by-three (i.e., 1/8th or 12.5%) — a simple calculation instead of the greater-than-B's-speed-plus-10...

[jamespetts]

I have already changed the overtaking calculation to use a greater than plus two formula. I have tested this and it does work in ideal conditions, so it is not correct that overtaking does not work on two way roads.

However, the overtaking code is extremely complex, and takes into account far more than speed: it also takes into account whether there is enough space ahead before reaching an intersection to complete the overtaking move.

I did not write the overtaking code and do not know exactly how it works. Amending it beyond simply reducing the speed differential would be an extremely large project and not one that I can prioritise until after economic balancing has been achieved.

If anyone else would like to look into this, this would be most useful.

[freddyhayward]

In anticipation of railway stagecoaches arriving to bridgewater-brunel, there are a couple of points I should bring up:
1. most railway coaching horses are only available singly, rather than in pairs. Is this intended? This is odd since their road equivalents only come in pairs. A single coaching horse is not enough to pull wagons up single slopes, so there is no real advantage for this arrangement. single horses are also slightly more expensive due to their rounded-up running cost, and less length-efficient (although this is unlikely to be a major issue).
2. some coaching horses have not yet had their costs updated in line with other horses, including the cleveland bay and yorkshire coach horses.

[jamespetts]

Thank you for spotting the issue with the prices not being updated: I have now fixed this.

As to them being singular, this is intended: because of the much lower friction coefficient on rail, a coach that would take six horses to pull on a road could be pulled on rails by only one. This was how horses were used in the brief and not very geographically extensive railway stagecoach era, which is why single horses are available.

[Ranran]

About overtaking slow horses,

1. There is no oncoming vehicle
2. There is enough space in front of the slowing horse to overtake
3. There is  no intersection

It is quite difficult to meet these conditions.
So, I think slow horses are still a big obstacle. If there are two horses with the same speed, I think it can't overtake anymore.

[freddyhayward]

As to them being singular, this is intended: because of the much lower friction coefficient on rail, a coach that would take six horses to pull on a road could be pulled on rails by only one. This was how horses were used in the brief and not very geographically extensive railway stagecoach era, which is why single horses are available.

This advantage does not seem to apply in-game, at least in my testing. A team of six road horses in pairs could pull an empty 1.2t coach up a one-tile slope (which are necessarily common because gentler slopes cannot be simulated) at around 14km/h, costing 15c/km and fitting within a single stop tile. The same number of rail horses could only pull an empty 1.0t (with much lower capacity and comfort) coach at around 7km/h at 18c/km and taking up two station tiles. With fewer horses, the speed drops further towards 1km/h.
Edit: aside from the above, I still don't see why pairs are not available alongside singles.

[Mariculous]

I guess the ingame issues are indeed steepnes of slopes and the fact, that it takes only very few horses to move a rail cart on flat tracks but much more to pull it uphill, which in the real-world was solved by adding additional horses around steep slopes or pulling the "train" up in smaller parts.
This cannot (yet) be done in simutrans, so we do always have to pay the whole amount of horses, where on most parts of the way we would only need one or maybe two of them.

Anyways, this does in fact add the requirement for double horses ingame, at least if we don't want excessively long convoys, including long stations for sure.
It might be cheated around by adding a little more power and force to rail horses or something like that, though this is not a pretty good solution either.

[Ranran]

Even in the real world, it would be difficult to lay a flat railroad in an "inter city" unless it was flat plain from the beginning.

I don't think the horse-drawn rail carriage is a vehicle for intercity transportation. I just thought you guys were laying a railroad for the era of steam locomotives on the server.

[wlindley]

The Stockton and Darlington Railway in 1825 had several passenger coaches, on a route of about 10 miles (±15km), drawn by a single horse, and covering the distance in about 90 minutes.  The line does have some hills.

[Mariculous]

Same for the Budweis–Linz–Gmunden line.
It started operation around 1830, is a little less than 200 km long, built on rather hilly terrain and, apart from a few steep sections pulled by single horse.
On these steep sections of up to 39 ‰, a second horse was used in addition.

Is 200 km enough to call this an intercity route?

[Ranran]

Is 200 km enough to call this an intercity route?

What I would like to say to you is to lay a 200km long flat track in the game.
I don't think the slope in the game of simutrans is such a small hill. Few weeks ago, I have noticed that the horse is extremely slow, even on two slopes of gentle slope.
On steep slopes the carts try to roll down the slopes due to their weight. It's hard to pull it up.
In practice, passengers may even better to get off such a hill once from a carriage.
In particular, the horse-drawn carriage can continue to draw a horse-drawn carriage with less force thanks to the low resistance of the rail. So fewer horses are enough. Conversely, it tends to roll down more on slopes.
If there are many horses with all trip for the slope, it is no longer worth using rails. You are just using a road that has the appearance of a railroad track. Many horses on flat ground are very wasteful.
There is currently no way to add horses only along the way. Impersonation will destroy the game balance.

[Mariculous]

Seems to be the same I said a few posts ago.
Our real-world scenario actually is not at at all realistic when it comes to slopes.

The point is it should be possible to operate such real-world lines. Early steam engines will have the same issues with slopes, if not even worse.

Edit: Is it intended that a single horse cannot even move a single passenger wagon at 18 km/h?
From the above discussion I'd at least expect a single horse to achive the full speed at flat track.

Edit: May I ask according to which data force and power of horses were initially balanced?
Reading a little more about this, I found out that on steep sections of track, horses were still prefered over early steam engines because of their good force/weight ratio.
The power of horses seems to be roughly fine, but the force is not.

Might the force have been set under the assumption that gear does not affect it?
Obviously, the actual force a horse can afford greatly depends on the race and exhaustion.
A plough horse had to afford a force of roughtly 1300 N, for short amounts of time even up to 2500 N.
That's according to this source.
It's from Massing and Finsterau open-air museum. They do have an English version of the page but I could'nt find the linked article there.

Our Shire horses in simutrans can effectively afford 560N
Would you expect them to be that much weaker than plough horses?

[Vladki]

Railway horses cannot run in pairs (side to side) because of the tracks. They have to run on after another. That was the case of Budweis - Linz railway. On the czech side, which was built with very gentle slope (suitable for steam operation), two horses were pulling two coaches. On the austrian side, which was much steeper (they ran out of money), two horses were pulling only one coach. See the pictures here: https://de.wikipedia.org/wiki/Pferdeeisenbahn_Budweis%E2%80%93Linz%E2%80%93Gmunden

[freddyhayward]

Railway horses cannot run in pairs (side to side) because of the tracks

In that case there is an inconsistency since shire and irish draught horses can run in pairs, unless this only applies to coaching horses.

[jamespetts]

Freahk - thank you for your research on this, and likewise thank you to Vladki for the useful information about multiple horses.

First of all, I have implemented Vladki's suggestion of removing all the double rail horses, as these do not appear to be correct for rail (as opposed to road or tram) horses: for rail, one would have to use multiple horses in line, rather than side by side for the reason given by Vladki. Loading a saved game with old double horses will automatically convert them to single horses of the same type.

As to the tractive effort of horses, this is complex. The comment in the .dat file for the road shire horse explains the process for and problems in calibrating this, done in December 2016:

Code Select Expand


# These values are not fine enough to capture
# the differing power of different horses.
# See http://gluedideas.com/content-collection/Textbook-on-Roads-and-Pavements-1908/Tractive-Power-of-Horses.html
# (also reproduced at https://archive.org/stream/textbookonroadsp00spalrich/textbookonroadsp00spalrich_djvu.txt)
# for information: an average horse will have a tractive effort
# of 80lb/f, or 0.33kN at approx. 5km/h and an average power of 0.45kW.


This source seems to differ considerably from the source that you gave of a plough horse having a tractive effort of 1.3kN for "many hours" (I used Google Translate to read the page).

The original notes did state that the gear did not affect the tractive force, but, even putting that aside, the tractive force for the shire horse is 1kN * 0.57 = 0.57kN, greater than that of an "average horse", as one would expect from a Shire. The Fresian horse, meanwhile, has a tractive effort of 1 and a gear of 0.32, giving a tractive effort after the gear of 0.32kN, very slightly below the average given in the reference work that I was using, which, since this will be the horse with the last tractive effort aside from the pony, seems reasonable.

Thus, we need to get to the bottom of the inconsistency between the sources. One possible explanation is as follows, taken from the same source I have used above,

The tractive force that may be exerted by a horse, at moderate speeds, varies approximately inversely as the rate of speed ; or, in other words, the power that a horse can exert through any considerable time is nearly constant for varying velocities.

This suggests that a horse, like an electric motor, is a constant power machine, and therefore that there is considerable variation in horse tractive efforts at lower speed. Since ploughing is likely to be done at an extremely low speed (perhaps 2-3km/h at most), this would explain how it is possible for horses to do so with such high tractive effort. The speed given in the source on which I rely is 250ft/minute, which equates to 4.57km/h, which is faster than a likely ploughing speed, and about the same as average human walking pace of circa 5km/h.

Indeed, the source goes on to describe a formula for calculating the tractive force from the speed:

Thus it may be assumed, as an average value, that a horse working regularly ten hours per day can put forth a tractive effort of 80 pounds at a speed of 250 feet per minute on an ordinary level road surface.

For the power of the horse we then have

Power force X velocity = 80 X 250 = 20000 foot-lbs. per minute.

For any other rate of speed, as 200 feet per minute, we would have 20000 -f- 200 = 100 pounds as the tractive force exerted by the horse.

Converting the latter to metric, we get 200ft/minute = 3.66km/h and 100lbf = 0.44kN.

Perhaps the apparently excessive weakness of horses is caused by using the tractive effort at speed rather than the starting tractive effort. I recall that Bernd Gabriel, who wrote the physics code many years ago, explained at some point that the tractive effort figure was intended to be the starting, not the continuous tractive effort. However, I cannot find any tractive effort curves for horses, so it is likely to be difficult to unpick this. Further thought/research will be required. Perhaps the ploughing tractive effort might be the right place to start after all if it is closer to the starting tractive effort?

Edit: If we assume for present purposes that a horse's tractive effort curve is level until 2km/h, then, using the above calculation at 2km/h, or 55ft/minute, we get 20000 / 55 = 363.63lbf, which converts to 1.6kN. This would suggest a tractive effort value of 2.8 for an "average" horse to be multiplied by the gear of 0.33 to get to this figure.

Edit 2: I have tripled the tractive effort of the game's various horses in order to deal with this: I suspect that the apparently excessively weak character of horses was caused by using the continuous rather than the starting tractive effort and tripling the tractive effort appears to be a sensible place to start for calibrating to a probable starting tractive effort.

[Mariculous]

Perhaps the ploughing tractive effort might be the right place to start after all if it is closer to the starting tractive effort?

Yes, that was my thoughts. If it can pull that much a low speed, it must have a starting force of at least that force.
Given the source is correct, for sure.

I suspect that the apparently excessively weak character of horses was caused by using the continuous rather than the starting tractive effort and tripling the tractive effort appears to be a sensible place to start for calibrating to a probable starting tractive effort.

That indeed sounds like a very good start.
More research data won't be easy to find, I'll see what I can do.