[8.2] Going uphill seems a bit unrealistic.

Topic: [8.2] Going uphill seems a bit unrealistic. (Read 3795 times) previous topic - next topic

[8.2] Going uphill seems a bit unrealistic.

August 05, 2010, 07:56:01 am

I've noticed that, while the cornering is done very well, it would seem that going uphill is not well refined yet.

After taking a trip through the Canadian Rockies, I've noted that loaded semi-trucks go very slow up steep inclines, 3~8% gradient and even unladen semi-trucks struggle to climb. These loaded trucks are probably going at 50~60km/h up the 5% inclines, while the unladen trucks seem to top out at 60~75km/h.

There were also 300~400hp caravans and RVs, which must have around a gross weight of 5 to 14t going up these slopes at around 65~90km/h.

In game, the automobiles and trucks go up 1 in 4 grades much too effortlessly and they climb 4 heights without too much loss in speed at all and only for the vehicles that are already underpowered going in a straight line. There would be no point trying to find the gentlest route through mountainous terrain or creating switchbacks, because you would run out of map height before a 'proper' mountain range can be scaled.

In the game, going up hill for trains seems to be a piece of cake, even for heavy trains and rather steep grades. A steep slope for a train would be about 1 in 23 (4.5%) , while a gentler slope is about 1 in 46 (2.2%) and they climb up hills too effortlessly in the game. For the game, even a 1 in 6 rise for 18 tiles is done at full speed, 120km/h for a train that has a gross weight of 2533t, being pulled by 5860kw engine.

At this sort of length, it's not too hard to plan a railway that goes up a mountain at full speed and defeats the purpose of having high powered banking engines with high tractive effort that are built for this purpose.

I've also noticed that it seems like only the first car in the convoy is affected by the slopes and all following cars have no weight at all. This is most apparent when a convoy is forced to stop just before the incline, compared to having just the first car sitting on top of the incline.

I was using pak britain ex, BR-90 hauling 24 HTA hoppers.
At 26 hoppers, this configuration starts to lose top speed.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #1 – August 07, 2010, 08:27:46 am

Why are locos too strong going uphill?

In simutrans there is only 1 slope inclination. This inclination is 2.8% (calculated from the standard simutrans slope "friction" penalty of 50). Usually the inclination should not exceed 2.0%.

Being realistic, early steamer loco driven trains could not climb up that 2.8% hills without ****istance by a second or third engine.
Thus for playability the friction impact has been reduced.

- If it is intended, that early locos don't use slopes, then we could increase the friction impact again.
- If there is a way to introduce a mix of half height and full height slopes, I'll support that.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #2 – August 07, 2010, 09:06:17 am

Quote from: Bernd Gabriel – on August 07, 2010, 08:27:46 am

If there is a way to introduce a mix of half height and full height slopes, I'll support that.

We discussed this a while back (here), albeit full height and double height in that instance. Summary: it's technically possible, and was apparently included in (very) early versions of the game. There was some interest/enthusiasm, but as is often the way, this did not (so far as I know) translate into the idea being taken forward, due to a number of technical reasons I think.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #3 – August 07, 2010, 10:22:44 am

here is what I mean.

retire the train at the top blocking the path so the coal trains all take off at the same time.
http://files.[ simutrans [dot] us (site down, do not visit) ]/files/get/hXn0RPYsWI/hill-test.sve

you'll notice that the slowest climber is, surprisingly, the train with the gentlest slope and that the speed and momentum carried by the trains with steeper slopes gets to the station faster.

The results would be different if all trains were allowed to hit top speed, but I just wanted to demonstrate how building a longer, gentler slope doesn't matter if you're faced with a steep cliff to scale in limited space.

I'm not entirely sure why the results are inconsistent, but, generally, the trains with faster speed entering the slopes will make it to the top quicker than the trains with a slower entry speed.

results would only be exaggerated with weaker engines.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #4 – August 07, 2010, 01:44:19 pm

Quote from: AEO – on August 07, 2010, 10:22:44 am

I'm not entirely sure why the results are inconsistent, but, generally, the trains with faster speed entering the slopes will make it to the top quicker than the trains with a slower entry speed.

That is intended and physically correct. With some inrun you run up the hill faster than from standing still. That's a feature of the new physics model: There are no (regular) abrupt speed changes any more.

I'll have a look into your save game... (pure Britain 128 EX pak?)

Impressive test track!
Fantastic idea with the Cl**** 66 stopper!

You don't see much impact of the effective slope inclination as the slope is too short. The trains have so much energy, that it pushes them uphill. You can see in the convoy windows, that no train would reach the top of the hill without inrun (max. speed is 0..120km/h, 0 with current load, 120, if empty).

If you start the trains at the bottom of the slope, you can see different behaviours until too many hoppers reach the slopes and the trains creep with the playability owed 4 km/h.

{ I have merged your double post. As a reminder, when you are the last one to post in a thread, and you want to add more, and it has been less than 24 hours since you posted: Please modify your existing post instead of making a new post. Thanks! -Isaac }

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #5 – August 07, 2010, 05:38:51 pm

yes, the save file is pure pak brit ex.

The slope is too short, but then again the game only allows for -9 to 14 for the map terrain levels and I don't think it is possible to create a map that features both somewhat deep oceans and mountainous terrain where a huge mountain between two plains would prove to be an obstacle.

There's really no incentive to build a gentle slope up a mountain or bridge over a gorge after the 1900's, when more faster locomotives start to become available as long as you pay attention to the run-up length to the slope.

As long as the train can hit 120km/h, it can climb the slope without hitting the playability 4km/h by the time it reaches the top. I tried it with the LMS garratt, that also has 5000kw, and in this case the steeper slopes are penalized. There is also the LNER P2, which was built for hilly terrain p****enger services and all the extra power it has also allows it to pull more carriages, rather than only allow it to negotiate slopes better.

Trains going through hilly terrain should require more power with additional banking engines or be shorter and lighter. It's not too unusual for banking engines to be added to trains going through a hilly section, especially for the steam age. I think there's too much weight for the momentum value carried by the trains. Pardon the pun.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #6 – August 07, 2010, 06:30:28 pm

that's a clear disadvantage of the 2d game engine, we only have discrete inclinations.

i think however the system works quite well at the moment. just ad a few signals or bends to your mountainous setup. as soon as a train has to break and move again you really notice the slopes.

if you want you can see the 4 km/h 'playability' speed as a complexity reduction for adding a banking engine and pulling the train over the slope.

is there a reason to keep the total height difference limitation (-9 to 14) with the greatly enlarged maps we use nowadays?

by the way i exploited your physics engine in suburban train setups in the steam age, by building my stations on viaducts. saving the kinetic energy as potential energy to have them start faster.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #7 – August 08, 2010, 09:09:55 am

further testing...

I made a new test bed for the slopes with full pakbrit ex.
http://files.[ simutrans [dot] us (site down, do not visit) ]/files/get/iOHG-m3Em4/hill-test-v3.sve

It would seem speed counts for a lot more than power and tractive effort.
Consecutive slope tiles don't penalize the climbing ability too much so there's not much need to "find" or create the smoothest route. In fact it's better to just allow the convoy some extra run up room to build up speed so it can climb 3 consecutive slope tiles instead of allowing it to climb a 1 in 2 or 1 in 3 slope that eats more room.

Re: [8.2] Going uphill seems a bit unrealistic.

Reply #8 – August 08, 2010, 12:06:57 pm

AEO,

thank you for your detailed research. Bernd; do you think that you could have a look into this, and perhaps reduce the extent to which momentum has an impact on hill-climbing ability?