The less rigidity of your mill / vise / table ways / part grip....the more I think its advisable to avoid climb milling for bulk of material removal. Save that for the finishing pass (assuming its required). I'd say this applies to a lot of hobby machines. If the teeth can't get a proper bite due to deflection or vibration or whatever, the cutter rides up the work & wants to tractor wheel causing even more deflection & the process kind of feeds on itself in a bad way. Also, the larger the EM diameter, the more contact area, the greater the chance of snatching all things equal. And yes, locking the non moving way, in this case the Y-axis, would be very advisable to help eliminate drifting table and improves finish. Aluminum would probably be even worse.
An alternative strategy might be
- lock the Y
- start on one end, feed down some safe increment, traverse to the other end of slot in conventional milling direction
- back up under power. It is technically climb milling but no real chip load. But you can safely also observe any finishing improvement as feedback
- if the cut went nicely, increment the DOC a bit more & repeat. Or just take your time & repeat if you aren't sure.
- take the finishing pass with say 1-5 thou DOC to remove the Y depth marks. This DOC is hard to judge because every situation is different but again you can increment in & see how it goes.
- now the side is done, that just leaves the over-feed on the floor. This is where I've gotten bit before but reduce the side load to zero is the main culprit. Its really just cutting on the end of the EM. However it can be influenced by the corner type & condition of your EM.
- check that cutter & material isn't getting too hot at any given time because now you introduce thermal expansion
- CI can be fun because its smaller chips (powder sometimes) so clean out the path often with a brush or vacuum. Rogue chips can hang around & trip up your EM even if everything else is going good.