(Liftoff has other boo-boos, so it's not an entirely fair comparison to make. I really dislike making comparisons, but take a look at the behavior of Liftoff in forward flight - it's not perfect either, but closer to how a real quad handles. In forward flying with the nose down, it's still necessary to use close to 100% throttle to keep the quad flying straight, which is not ideal and not at all similar to how a freestyle quad with a cam at 30 deg actually flies. At 0 degree tilt (hovering), the behavior seems better now, though I would say the quad still takes off with way too litle throttle. Things improved a bit, but the "getting sucked into the ground" effect is still there. Were you experienceing this issue with any other quads besides the ones that you mentioned by name? Thanks for the write-up, I'll make sure that our physics guys get a chance to look it over. Originally posted by DRL_Wyatt:Hi seattleite09, I tried upping the thrust, decreasing the gravity, etc - all hacks, but in the end none of them truly fix the underlying bad equations. You can repro these problems with "stock" Chameleon or Alien model settings. Something is not right currently with the throttle + thrust, and that makes handling in DRL really really bad compared to real life aircraft. I would take another look at the physics / equations for buoyancy close to the ground, both for #1 and #2 above. When diving in DRL, I need to apply maximum throttle for a couple of seconds just to stop the quad, and often that's insufficient and I crash anyway. In DRL, this ground effect either doesn't exist, or seems to be broken. One throttle blip and the quad comes to a stop in the air. This creates extra air pressure, which greatly helps with the quad recovery from that dive/split-S/etc. Moreover, when the quad drops (from a split-S, dive, or some other maneuver), the cusion of air underneath "compresses" for a fraction of a second while the air is forced sideways between the props and the ground. Basically the quad can float easily with very litle throttle. In real life quads, the props spinning close to the ground create a cusion of air that provides extra buoyancy to the aircraft. Where in real life I can recover from drops and fly straight, in DRL I end up smashing the model into the ground most of the time. I end up using 30%, then 40%, then 50%, etc all the way to 100% throttle just to keep it going straight parallel to the ground, and in the end 100% becomes insuficient and the quad crashes. It's easy to reproduce: start flying straight and parallel with the ground, and pay attention to the throttle stick. In DRL it takes a non-constant (*increasingly high*) amount of throttle to fly horizontally under the same circumstances. I can put the nose down and fly faster with more throtle, but it's always a constant amount of throttle per angle. In real life quads, it takes a constant amount of throttle - usually 40-50% on my 600g 5" quads - to fly in a straight horizonal line with camera at 30 deg. Especially close to the ground, there are two problems:ฤก. Since I used it last time (around February) the physics engine seems to have been changed quite a bit.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |