Electronic – BLDC Motor Maximum Locked Rotor Current

There are a few things you have wrong. First, 1789 Nm is the rated ("full load") torque, not the maximum torque. The maximum torque is usually called the "breakdown torque" and according to your data, it is 210% of the full load torque. This puts the maximum torque at 3756.9 Nm. The locked rotor torque also isn't 42.93 Nm, but rather 190% of the full load torque. This puts the locked rotor torque at 3399.1 Nm.

Try fixing those things. Also notice that the graph provided plots the speed, torque, and current as a ratio of the rated values.

The link in case 1 is pretty thorough - and includes the note that a true thermal model of a motor isn't that simple. There are three three losses in motors considered mechanical - iron losses (in the armature core), friction losses (bearings and brushes) and windage. These subtract from the electrical torque generated by the motor - to zero in the case of no load operation, obviously. Iron losses are fairy invariant with temperature, windage falls with temperature, but is probably negligible on a small motor - unless it has a shaft-mounted fan, and the bearing friction will also fall with temperature. Brushes are far from predictable, and their coefficient of friction can vary widely with temperature and surface condition of the commutator. Complicating all that is the temperature coefficient of the magnets. Most materials will reduce strength with temperature, and this affects the Kt value noticeably - so more current will be needed to generate the same loss torque. So will the no-load current rise or fall with temperature? That will all depend on the balance of the losses.