MIT built a concrete beam in a fabric mold with a third of the concrete and half the steel

A 3-meter reinforced concrete beam cast at MIT carried the same load as a conventional beam while using about a third of the concrete and half the steel reinforcement. The beam was poured into a flexible mold sewn from fabric pieces, and the shape it took was the one a computer optimization said the load actually needed. Mark Hellrich, who completed his master’s degree at the institute in 2026, led the work.

The tech news worth your time: 400,000+ professionals from Google, Apple and OpenAI read Techpresso, a daily email with the most interesting tech news and insights.

Subscribe to Techpresso →

We may earn a commission if you sign up.

Standard beams are rectangular because rectangular wooden or steel forms are cheap to build, which means much of the concrete sits where it does little structural work. Cement production accounts for roughly 8 percent of human-caused carbon emissions, so every cubic meter left out is a cut in a very large pile.

Sewing the mold from flat fabric panels

Fabric lets wet concrete bulge into a curved form under its own weight, but predicting that bulge is hard. The team wrote an automated pattern-making process that converts a three-dimensional optimized design into flat fabric pieces, accounting for how the cloth stretches and deforms as it fills. MIT News describes three molds tested, built from 36, 56 and 76 fabric pieces. The 56-piece version struck the best balance between accuracy and sewing effort.

A simple timber frame holds the fabric in place. The concrete mix was moderately wet and sand-rich with small stones, and workers packed it by hand without a vibrator. Hellrich reported that all three beams came out free of wrinkles and creases.

Topology optimization and graphic statics as the design engine

Two computational methods decide the beam’s geometry. Topology optimization determines where concrete is needed and where it can be removed. Graphic statics then maps the path of the steel so the reinforcement carries an equal share of the weight along its length, which keeps the bars simple to bend and place. Josephine Carstensen, the Gilbert W. Winslow (1937) Career Development Professor in Civil Engineering and senior author, said the mold’s flexibility is not just a limitation to manage but something the design can exploit.

Co-authors include Jackson Jewett, who earned a PhD at MIT in 2025, and Abhinit Sati, Edward Chien and Emily Whiting from Boston University’s computer science department. The research was partly supported by the U.S. National Science Foundation. A previous project in the same line of work produced a 1.2-meter beam, so the new prototype is a jump in scale.

The result is documented in a Journal of Building Engineering paper on affordable construction of topology-optimized reinforced concrete through fabric formwork. In load testing, 36 water-filled buckets stacked on the beam brought it to about 30 percent of its design load, and the measured deflection at the center landed within about 2 percent of the computer model’s prediction.

Costs that flip between the Global South and the Global North

Material savings do not automatically mean cheaper buildings. The researchers compared costs in two settings. In the Global South, where sewing labor is inexpensive, fabric versions came out about 10 to 20 percent cheaper than conventional beams. In the Global North, higher labor costs made them about 16 to 55 percent more expensive.

That split shapes where the method fits. Fabric formwork suits custom, one-off or small-batch components, while conventional reusable molds remain more economical for long runs of identical rectangular beams. Jewett joked that the group is developing the H&M of the construction industry, a nod to the idea of cutting patterns from cloth and sewing them quickly.

The timing matters because of how much building is coming. The number of buildings in the world is expected to nearly double by 2050, with about 85 percent of that growth in developing countries, according to the MIT account. Cement is already a major industrial emitter, as the International Energy Agency’s cement overview lays out, and a design that cuts concrete volume by two-thirds in a beam is aimed squarely at that growth.

Testing to failure and the limits of scaling up

The prototype was loaded to only about 30 percent of its design capacity, so the harder claim, that it performs safely at full load and beyond, is still untested. The next phase is testing multiple beams to failure. Larger beams will need new designs too, because their weight and the pressure they put on the fabric change as length grows, and a pattern that works at 3 meters will not simply stretch to 10.

No commercial partner or building-code pathway has been announced, and the release does not say how long certification of a fabric-formed beam for real buildings might take, since engineers would need failure data and repeat tests before any design standard could cite it. Hellrich and Carstensen’s group has framed the current result as a proof of concept for accuracy: the physical beam behaved the way the model said it would, which is the precondition for everything that follows. Whether sewing labor can be automated enough to erase the cost penalty in wealthy markets is the question the team has left open.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

The tech news worth your time: Subscribe to Techpresso →

We may earn a commission if you sign up.