I wanted PETG bodies with TPU gaskets printed in one go, no gluing, no inserts. Multi-material is the obvious answer. The question I actually cared about was whether it pays for itself, and the honest answer is: sometimes, and not the times you'd expect.
I'm running a single-nozzle, multi-input setup rather than a tool changer. Every colour or material change means retracting one filament, loading the next, and purging the old one out of the hot end until the new stuff runs clean. That purge is the whole story. On a small part with frequent changes you can easily spend more plastic in the purge tower than in the model itself.
A worked example from this week. A control box lid, two materials, swapping roughly forty times across the print. The part weighed 31g. The purge tower and wipes came to 58g. So nearly two thirds of the filament went straight in the bin, and the print took about 70% longer than the same geometry in a single material because the machine spends real time priming and wiping on every change.
That sounds damning, and for that part it was. I should have printed it in one material and accepted the seam. But there's a category where it genuinely earns its keep:
- Parts where the second material is structural, not cosmetic. A living hinge in TPU, a soluble support interface that lets you print geometry you simply couldn't otherwise.
- Things you'd otherwise assemble by hand at a rate of dozens. The labour saved swamps the wasted filament.
- Prints where the alternative is a failed part, not a slightly uglier one.
The settings that moved the needle for me, after a lot of stringy failures:
purge_volume_min = 70mm3 # PETG -> TPU, the dirty direction
purge_volume_min = 45mm3 # TPU -> PETG
wipe_into_infill = true # reclaim some of the purge as actual part
wipe_into_object = false # tempting, but it shows on the surface
Wiping the purge into infill was the single biggest win. Instead of building a tower of pure waste, the machine dumps the transition plastic into the inside of the model where nobody will ever see it. It doesn't recover everything, the first millimetre or two of a transition is genuinely contaminated and has to go somewhere, but it turned "two thirds wasted" into something closer to a third on the better-behaved material pairs.
So is it worth it? As a default, no. As a deliberate choice for parts that need it, absolutely, and I'd not give it back. The trick is being honest with yourself about which print you're holding. Most of mine, it turns out, were vanity.