The Last 15%: What to Do When Topology Optimization Runs Out

Every lightweighting program ends the same way. The part is ribbed. It's cored. The topology run has carved it into something that looks grown rather than designed. And it is still forty grams over target.

At that point, the geometry is finished. Every gram left in the part is there because physics wants it there - and no further mesh refinement will argue it out. You have exactly two moves left.

Move one: thinner walls

The problem is that your alloy votes no. A380/ADC12 fills reliably at 2.0–2.5 mm. Push below that and you trade mass for porosity, cold shuts, and a scrap rate your program manager will notice before your mass tracker does. The wall thickness floor isn't a design choice - it's a property of molten aluminium in a die.

Move two: a different alloy

Magnesium changes both variables at once. AZ91D is 1.74 g/cc against A380's 2.70 - and it casts reliably at 1.0–1.5 mm walls, below aluminium's floor. The same geometry gets lighter twice: once from density, again from walls the aluminium version was never allowed to have.

The honest number, after compensating wall and rib depth for magnesium's lower stiffness, is 20–30% out on a real housing — not the 35% the density ratio promises. If a supplier quotes you the flat 35%, they haven't run the correction.

Run the number before the next review

Whether the swap pays depends on your part's mass, its governing load case, and how thin its walls already are. That's a thirty-second calculation, not a meeting.

Run your part through the Mg Swap Calculator - the honest number, stiffness compensation applied, no signup. https://www.exclusivemagnesium.com/learnmore/magnesium-weight-savings-calculator