There are some VERY misleading comments here.
--The res does not have to be at the highest point.
--The res does not have to be open.
The air will collect in pockets in the loop. This can be in blocks and rad. Air is prone to collect when flow is reduced (this is where the res comes in)
Typically, the pump is powerful enough to push air along the tubes (even down) as the flow rate is high enough here.
The trick is to tilt the case sufficiently (laying it down on either side) so that the air goes from pocket to pocket (block to block to rad) and eventually to the res.
The trick is NOT to let air get back into the pump from the res. This can be caused by tilting too much in the wrong direction [so the pump sucks air from the res], or by letting too much air accumulate in the res without topping up, [so the pump sucks air from the res].
If the pump sucks air, it has to make it round the loop again. The rad is usuall the biggest hoarder of air as its tanks act like res's, where flow is reduced and the air collects easily. Also (depending on mounting and barb position) the air can be hard to get out.
1 make sure rad outlet barb is able to dispense air (higher than inlet and either horizontal or pointing up)
2 make sure the outlet from res to pump will not suck air (positioning and filling regularly as needed)
3 make sure pump is lower than res (usually point 2 ensures this)
If you can find a position than encompasses all 3 points above, the loop will mostly self bleed with a bit off jiggling. On/off the pump or pinch tubing to create some eratic flow will help too.

/edit/ most of the references to "flow" are velocity and not a volumetric measure (lpm etc) I was trying to make it easy reading. The air will be inclined to collect where the speed of the flow is less, the volumetric flow is constant throughout the loop.