I know, if feels crazy. One of the classic proofs that is the case comes from gyroscopic motion. Check out this good RU-vid video: m.ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-ty9QSiVC2g0.html
Torque is a pseudovector that "transforms like a normal vector under proper rotation but gains an extra sign flip under improper rotation such as reflection." en.wikipedia.org/wiki/Pseudovector
Think of it like a bolt on a car. You want the bolt to come "out" of the page, so the force must be applied along the curl of your fingers. Same the other way if you want the bolt to go INTO the page/car.
@@ClutchZ28 Not to make it more complicated but bolts can be reverse threaded and you'd have to turn counterclockwise to go into the page. This is what's frustrating to me that the direction of torque just seems like an unimportant side product of the calculation
Big Guy yeah, the direction of vectors when looking at rotational motion feels very weird! It’s a consequence of the cross product which I didn’t go over in this video. After I teach my classes the direction of torque and of angular momentum I generally do a “precession” demo (like ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-ty9QSiVC2g0.html ) to help prove those directions are right because without proof it feels crazy.
Big Guy the angular acceleration is along the direction of the torque vector not the tangential acceleration. Which means the wheel will rotate towards the direction of the force applied but the angular acceleration will be towards the plane of the page.