metallurgy is *not* my strong suit by any means, but my study occasionally force me to face my fears. Critellium has been the gold standard of transmutation engine construction for a long time now, and for good reason: the kind of properties required for these things are rare, near-contradictory. incredible strength, resistance to heat and the elements, combined with a fluidity and ability to join perfectly with every glyph and component you could dream of. it's *so* important, in fact, that practically every construction for the thing is a trade secret. the price on the stuff is *insane*. the only formulas you can find for it are based on Critelli's original refinement of metal slag, the complexity of which is enough to push most alchemists to simply buy it even with the markup. for a one-off project, this would be fine, but my work requires a regular supply; with the last of it spent on my last failure, i had to sit down and make a serious attempt at purifying it on my own. now, i'm sure this is easy enough for the ace alchemists reading along, but this quickly got frustrating. just to have *something* to work with, i built a machine that would extract an iron atom from each input, extract *only* the quicksilver from the next few to project, and use this alone to build it. it was incredibly slow and inefficient, and just the thought of relying on that for my supplies made me try again from the ground up. now, let's think about ratios. once again, those star thaumaturges of you can probably spot by eye the perfect ratio, but i was happy to sit at 10:1, the smallest number of inputs per single output (that i could find). the amount of waste i still get tells me it can go lower yet, but i'd already optimized away my ability to care to the minimum. using one arm to take inputs, a hex-arm to handle metals, another arm to aid in purification, and a piston to aid in construction lead me to a passable construction; i know a couple ways it can be improved further, but... ugh. i have things to do. - alchemist luna's blog