maybe thats why they don't have pre-tensioning....
the bridge would sagg and tear the cables right from the steel
Dont forget: adding tension reduces the amount of ´real´ load the cable can hold!
I agree with kvinge that easing into gravity is not a solution to making suspension bridges work, pretensing the cables is needed, and I hope to see this thread address some of the ways this can be implemented.
Based on the limited knowledge I have, this is not something that can be implemented with the current bridge format. I think this is something that can be added, but will require a change in the way the cables are stored, and will result in a new version of the pxb file.
It should also be possible to convert version 1 pxb file to version 2, without causing some radical change in the P* engine. After reading the header, P* can determine if it is dealing with the newer file format, or the old format, and add a pretension of 0 to all the old bridges.
One of the main reasons for wanting pretensioning is so that you can control the initial sag that occurs when the simulation starts, to avoid the excessive stresses that this puts on the structure. A simple solution to this would be for the physics engine to gradually ramp up gravity for the first couple of seconds, so that it can settle into a stable position more easily. This wouldn't be totally realistic - but then bridges don't spontaneously pop into existence either.
And it would help the game a lot
This might also provide the opportunity for CL to fix some of the physics bugs and/or modify physical properties without destroying the old game somewhat.
It would be ok if we can get flexible nodes (discussed elsewhere), too. But I would prefer 'real', adjustable pretension :)
Although pre-tensioning would work, you would need to do this for EVERY section of cable, steel, internal strut etc. until you approximated all the stresses on a stable bridge. Think of trying to replicate the various shades of read/blue you see now on a bridge that has 'stabilised'. That isn't going to be an easy task.
Pretensioned cables, while a great idea, and something I want to see implemented, are not the end all. I think that the ease into gravity is more improtant to more than just cables, and should be implemented first.
The latter is practicable right now. I solved one of my problems with my lame contest bridge design that way.
I used to have it so the cable tower was built on the lower anchor and had the first deck joint integrated in its steel structure. This resulted in the second deck piece taking a lot of strain, because one joint was fixed in the heavy-steel-supported tower, the other was cabled and connected to sagging rest of the bridge. The stress was noticeably reduced when I built the tower past the first deck joint (removing the angle in the tower, too!) and suspended the first deck joint from that same or an auxiliary tower, I don't remember which. What I do remember is that this setup made the first joint sag a little, too, and that allowed it to distribute some of that stress farther to the outside, resulting in a more even stress distribution and a better bridge.
(Actually, this is a "current version" idea, so I'm a little bit off-topic here :-).
What you could do would be to build the deck at an angle, so it's allowed to sag a little, thus stressing the cables. The problem with this is that the deck gets stressed, too - and that should not be happening with a sus bridge...
If the members were still pin-jointed as they were in BB, the sag would not induce much extra stress, so instead of demanding prestressed cables, we could instead demand a no-diagonals beam (demanded elsewhere on this forum as well, if I recall). I dub this "swinging bar".
Swinging bars could be easily implemented as a new material type without changing the current file format, and it could even use the same colors now used for light and heavy steel.
Until this post, I never gave much thought as to why it works. I also find myself using that technique subconsiously now. I've been intentionally building in a way so as not to physically support the deck. In order to create stronger support towers, I've been sagging the deck by 10m to allow me to build taller towers. (See the short links bug)
Anyway, just wanted to give mendel some feedaback on his post, and I'll try to make more of an appearence when I get back in town.
Please impliment it, CL!
At some point the cable strength just wouldn't be enough. You can only pull it so tight before it still breaks.
I don't think that situation would be an issue if there was a good way to handle cable stregth. Even with that God mode grab thing you can whip the train around so fast that even 'the hand of God" can't hold it...
I think pre-tensioning would be good.
sopwath
Please impliment it, CL!Has my vote... even so I think the pre-tensioned cables would be needed for some structures. I can't think of any easy way to implement it though.
There are only four possible states for a mechanical system. It can be overdamped, underdamped, undamped and critically damped. The bridges in P* are an underdamped system. If the bridges were critically damped, they would not have any overshoot, and they would reach steady state in a minimal amount of time. If they were overdamped, they wouldn't overshoot (and bounce) either, but the time to reach ss would be longer. If the coefficients of damping were too great, it may take hours to reach ss. The opposite of course would be an undamped system. In this system, the bridge would continue to bounce forever.
P* is an underdamped system because the bridge will bounce, and it will reach ss at sometime.
You are correct that as an underdamped system, slowly introducing gravity will not elliminate a problem with bounce. However, if the bridge is allowed to reach ss before gravity is increased additionally, then the impact of that bounce on the overall design will be minimal.
If the current system causes the bridge to overshoot its ss position by 1m when the level is first tested, then by easing into gravity, it may only overshoot by .01m. This makes a tremendous difference when waiting for the bridge to settle, and an even greater difference when you start adding mass (in the form of a train).
Of course, a poorly designed bridge is still going to collapse, but perhaps the deflection in the bridge will be less from the train than by the start of the simulation. A level that only requires 2 cars or so would gain from such a change.
Quote: from beaujob on 10:10 pm on Oct. 31, 2001
My vote is for having a settle button in the editorI think this is a tremendous idea. It is certainly needed if there is going to be a way to tense the cables. In such a mode, I think CL needs to change the damping coefficient to be critical. This way, you won't need to wait for a settle period before you tense the cables. An underdamped system is still reasonable for the actual simulation.
Might make for some funny slingshot effects also ;)
Just take a look at your bridge for Map Pack 1 Level 16, and think about what made it difficult. It took me over an hour, even though I already had plenty of practice with arches. It's difficult because the bridge is so large that it sways and must be very carefully cabled and reinforced. I had one verson of MPACK116 that actually stood up with no broken links after four passes 80% of the time (it depended on when you started the train), but it would start swaying and tear itself apart if you just let it stand there for ten minutes without even running the train.
If you could make the tension on cables tighter, it would make things ridiculously easy. After I completed all 48 levels that come with Pontifex, I started going back and making them cheaper. When you're cutting costs, even smaller bridges can use a lot of cable, and the challenge is to make stable designs that don't oscillate and tear themselves apart--a challenge imposed mostly by cable slack.
If you could increase the cable tension, this aspect of the game would be totally removed, and almost every level would be much less challenging. Without slack, the bridge doesn't oscillate, and it quickly stabilizes itself after the train passes, so there would really be no point in levels requiring four train passes--you would know immediately as soon as the train made it across once that every other run would be successful because the bridge wouldn't be swaying.
OTOH, a good addition might be high-tension cable at a higher price. It would have to be at least three times as much as regular cable, though.
That said, with some level redesigning, the game would be much improved.
From a coding standpoint, however, pre-tensioning would be a bitch to code. But then again, phizicks engines are bitches, and ChronicLogic seems pretty adept at writing those...
In conclusion, OG WANT PRE-TENSIONING!