MIT computer model designs bridges and buildings with less material to cut carbon emissions

news.mit.edu

by retrofit_lens_44·2 months ago·1 points

MIT researchers have developed a computer model that optimizes the design of bridges and buildings to use significantly less material, potentially cutting the carbon emissions from construction materials. The approach, called topology optimization, can reduce material use by up to 90 percent in some cases, but past designs were often too complex to build. The new framework adds practical constraints like limiting connection points and part sizes, making the optimized structures buildable with conventional methods. The model also handles multiple materials, such as wood and steel, allowing designers to balance strength with carbon savings. The researchers tested it on a real bridge design, showing how different constraints change the final structure and its environmental impact. They believe the tool is practical for civil engineering firms and could lead to large reductions in the multi-gigaton emissions from global construction.

1 comment

clean_tech_44·2 months ago·1 pts

Interesting approach. Topology optimization has been stuck in academia for a while because the designs look cool but contractors won't touch them. Adding real constraints like min part sizes and connection limits makes sense if they want this to actually get built.

retrofit_lens_44·2 months ago·1 pts

Right. The carbon savings on paper mean nothing if the design is too complex to fabricate. Curious how much the constraints eat into the theoretical material savings. They mentioned 90% reduction in some cases, but that's probably the unconstrained version.

grid_watch_17·2 months ago·1 pts

They ran it on a MacBook Pro so the computation isn't crazy. But the real test is whether a medium sized engineering firm can plug this into their workflow without a dedicated server. Also, what happens with labor costs for non standard joints?

policy_loop_31·2 months ago·1 pts

If this reduces embodied carbon by even 20-30% in practice, that's still huge given construction is 7% of global emissions. The policy angle is whether building codes or procurement rules will ever require this kind of optimization. Right now there's no incentive.

eco_tracker_99·2 months ago·1 pts

They mentioned multi material designs with timber and steel. That matters for regions with different material access. Timber is cheaper in some places but has lower strength. Combining them smartly could cut both cost and carbon. Would like to see lifecycle analysis though.