Jump to content

camlost

Members
  • Posts

    514
  • Joined

  • Last visited

Everything posted by camlost

  1. First about realism: I'm talking about physical model here. Since the goal is to create a simulation of a piston engine and transmission etc, this is realism by definition. In the end you could have everything overpowered, but that's configurable. 1) For example for the wheel you can use, there's no need to have duplicate models. Much simpler this way +PART[roverwheel] { //... } or rather use @PART, by installing the mod people mean to use it. And using @PART won't break save when people delete the mod. 3) Having a central part or not, calculating power is the easy part. How to distribute torque to each wheels is bit complicated. That's why I recommend having a partmodule in wheels to set torque every tick. For example , torqueCurve { key = 0 80 0 0 key = 55 30 0 0 key = 80 0 0 0 } This is totally arbitrary. No matter how you tweak them they'll be off at some point. What I have in mind is : what happens each tick: a) engine calculates RPM based on wheel diameter, surface speed and gear ratio engine calculates torque based on RPM and throttle/control axis, and sets the fuel consumption c) each wheel calculates torque distributed to itself, multiplied gear ratio d) each wheel sets torqueCurve like (for example the torque is 26.4) torqueCurve { key = 0 26.4 key = 1000 26.4 } Also you can check out Snjo's FSWheels, or TT multiwheels, both have side slip I think 4) See the discussion in the first page of Advanced Jet Engine thread. Seems to me the whole "IntakeAir" thing is like a quick and dirty hack to make jet engine work in a rocket's physical model. However it's physically wrong and brings far more problems than benefits. Now over a year later in 0.23 they make jet engines automatically throttle down when intake is slow -- still wrong
  2. Since a year ago there had been quite a few mods about *realism* wheels, engines and track, but they're all dead because the developer's left. I figure you're ambitious judged from the dev plan. If you want this to be an important "realism vehicle" mod, I think you should make it copyleft so that other developers can work on it. Actually the TweakableWheels is under CC-ANS 4.0 which clearly states "share alike". Having "You may not redistribute and/or modify this mod without permission" is bad manner. Now I'm glad people are interested in developing this, here's some ideas regarding the current release: 1) Use MM. There's absolutely no need to pack redundant parts like jerrycan. 2) I think RollKage has some nice models for piston engines, their developer's more into 3-d art instead of physics. You should talk to them about collaboration 3) There needs to be some regulator code for "engine power". I think the simplest way is to go around the resource system: it's not designed for this kind of purpose. You can make each piston have a "engine" part module, to calculate power and fuel consumption based on throttle probably. Each "wheel" has a "wheel" part module, which sums up the power and distribute it, and set the correct torque each tick. 4) Never use resource air intake. Use code to determine oxygen and atmosphere and turbocharger etc.
  3. It's not necessary to calculate aerodynamics really. A pre-WWII low-altitude diver bombing is useful enough. All it needs to do is select a target on the ground, so the program knows the target altitude, then solve a parabolic equation. If I find time I'll fork it.
  4. That's weird. I've tested many dual engine planes, in no circumstances the inlets' causing unsymmetrical thrust, even if the inlets are unsymmetrical
  5. Is the other electric propellers working? Is the KAX turboprop working?
  6. The exception is caused by the nasa code. Sometimes i run into it while using AJEtester. At what speed do u have the problem? It seems usually happens when prat3<10 Have you noticed any problem while flying J57? I haven't
  7. The turbojet is f100 which is suitable for a fighter plane, the so called "turbofan" is J57 which is best for hypersonic planes. The ramjets are from tv's pack. Any problem with this arrangement?
  8. Ok, what fuel? Meanwhile i'll see what I can do about SABRE with variable CPR feature
  9. @TeeGee, the stock "turbofan" is J57 not ramjet now. The heating problem is the result of intake air being too hot at high mach number, which destroys the engine's compressor. It's not DRE's effect. What you should do is turn off the J57 at Mach 4.5 and use rockets. Also I have no converted the RAPIER: I don't know what it is @epicfacecanada, they should have an offset because usually the helicopter's turboshaft engines are installed behind the rotor's axis
  10. It shouldn't be too hard to make a bombing reticle I suppose? I've been thinking about this
  11. Well you have a set of cockpit parts that's at least 98% complete. Why not push a new release of D12 pack including them? It seems very unlikely they're prone to changes by Snjo, nor being eclipsed by the new B9 completely.
  12. You can open a cfg and replace all "Kerosene" with "LiquidFuel"
  13. It's there already. The J57 for example has CoMOffset=0, 2.5, 0 something like that My dev plan is 1) Inlets 2) Numerical model for SABRE, SCRAMJET and whatever Finally 3) Procedural Engine, the coding shouldn't be that hard, but I need an artist to make some parts: a fan, a compressor, a burner+turbine.
  14. Solar powered-plane is just impossible with KSP, because the lack of ultra-light material. The wings of the same area weighs at least 10 times more than what a real solar plane is. With this much weight a 15kW motor is utterly useless. A 15kW propeller will never have 6kN thrust, that's just another eye-balled 10-fold overpowering number.
  15. First you should always ignore what 'stats' tell you because ksp is full of lies. Second, in my dictionary 'absurd' means unreal and against physics. RF and KSPI's definition is 1 EC = 1 kJ. So you know how absurd it is to fly with 3000 EC. Tesla roadster has 200,000 EC
  16. When people talk about forms of wings they assume the same reference area. On top of that, high AR has lower induced drag, thus better gilde ratio. Low AR has better high-AoA capability and less wave drag. That's why low-speed airplanes usually have high AR wings, while modern fighters have low-AR wings.
  17. There are four, Xichang, Jiuquan, Taiyuan and Wenchang
  18. Here's what IDSkill does: if (this.part.temperature > (this.part.maxTemp * 0.8f)) { this.targetIsBurning = true; } I left a message for ID in his thread about making that '0.8f' configurable. Should be a very easy fix.
  19. You don't need to put anything, the work is on ID's part for example ... [KSPField] public float burningPoint=0.95f; ...
  20. ID, I think you should make "temperature damage threshold" configurable. Probably you can make it a partmodule, then use wildcards to add it to all parts containing ModuleEngines or ModuleEnginesFX, then other mods can adjust the threshold accordingly using MM too. Probably something like this: MODULE { name = targetPart targetMaterial = aluminum2 burningPoint=0.95 // when part.temperature=0.95*part.maxTemp it will burn. } In the dll side you can set its default value=0.95 so the MM patches are backward-compatible Currently, the threshold is somewhere near 0.8 I think, as soon after the overheating bar appears there's black smoke pumping out of an engine. That seems too hard when I play AJE.
  21. ID, I think you should make "temperature damage threshold" configurable. Probably you can make it a partmodule, then use wildcards to add it to all parts containing ModuleEngines or ModuleEnginesFX, then other mods can adjust the threshold accordingly using MM too. Probably something like this: MODULE { name = targetPart targetMaterial = aluminum2 burningPoint=0.95 // when part.temperature=0.95*part.maxTemp it will burn. } In the dll side you can set its default value=0.95 so the MM patches are backward-compatible
  22. That is probably a MM problem. I tested with 2.1.4 version. Do you have RF and FAR latest version installed?
×
×
  • Create New...