> ## Content Index > Fetch the complete content index at: https://clever-new-stuff.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # Cold-rolling makes cheap steel as strong as door-beam steel - URL: https://clever-new-stuff.ghost.io/cold-rolling-makes-cheap-steel-as-strong-as-door-beam-steel/ - Published: 2026-10-14T02:10:00.000Z - Updated: 2026-10-14T02:10:00.000Z - Author: CNS Staff *Not for you? There's a one-click unsubscribe at the bottom.* ###### MATERIALS AND MANUFACTURING ###### ### Cold-rolling makes cheap steel as strong as door-beam steel University of Science and Technology Beijing researchers report they made a cheap, low-carbon steel about as strong as the hot-stamped steel used in car door beams, by cold-rolling it – rolling it at room temperature. Until now, steel like this reached that strength only by warm-rolling: rolling at about 300 °C, give or take 20 degrees, and reheating between each run through the mill. If the method works at scale, mills could make very strong steel on the equipment they already have, without that tight temperature control. The steel is a dual-phase steel: a mix of a soft and a hard form of iron. Cold-rolled, the soft part usually does all the stretching and the hard part cracks. The authors first heated the steel to 900 °C for three minutes and plunged it into water, three times over. This left thin alternating layers of hard and soft steel and narrowed the difference in hardness between them by more than half. They then cold-rolled it, cutting its thickness by 85 percent. The layers thinned to about 37 nanometers (millionths of a millimeter) without the hard layers cracking. The rolled steel reached a yield strength – the load at which it bends for good – of 1.98 gigapascals (GPa), a measure of force per unit of area. Its tensile strength, the load at which it breaks, was 2.15 GPa. The same steel, warm-rolled, reached a yield strength of 2.04 GPa, and an earlier cold-rolled dual-phase steel 1.58 GPa. The recipe is cheap: 0.15 percent carbon, 1.8 percent manganese, 1.2 percent silicon and 0.4 percent aluminum, with no costly nickel or cobalt. **The limits.** The rolled steel stretched only 2–3 percent before breaking, which the authors call typical of steel rolled this hard. Sheet steel is usually pressed or bent into shape after rolling. The study did not test whether this steel survives that, or any treatment to make it less likely to crack. The heat cycles before rolling are still needed, and have not been tried on a production line. The work was done in the lab, with a 30 kg batch, sheets 1.2 mm thick and at least three test pieces for each result. The authors pulled the steel only in the direction it was rolled; they expect it to behave differently when pulled at right angles to that, and plan to check. **[Read the paper](https://doi.org/10.1016/j.jmrt.2026.09.200?ref=clever-new-stuff.ghost.io):** *Niu et al., Journal of Materials Research and Technology 45 (2026) 1718–1725\. Free to read.* --- ###### ENERGY ### Polymer gel slows storage losses in silicon-rich batteries Researchers at Huazhong University of Science and Technology, with co-authors from the carmakers FAW and Geely, report that lithium-ion batteries with silicon in the anode – the electrode that holds lithium while a battery is charged – wear out in storage far faster than those with the usual graphite anode. Makers add silicon to pack more energy into the same space. Kept fully charged at 40 °C for four weeks, cells with graphite anodes could still hold 75 percent of their original charge, their capacity. Cells whose anodes were half silicon oxide could hold about 10 percent. The authors watched the damage with an ultrasound scanner, which showed gas building up inside the sealed cells. After three weeks, gas had spread across 91 percent of the scanned area in the cells with the most silicon, against 34 percent in graphite cells. They put it down to a chain of reactions. Lithium drifts on its own from the graphite to the silicon particles, which swell and crack the thin protective layer on their surface. The exposed silicon then reacts with the electrolyte – the liquid that carries lithium inside the battery – using up lithium and giving off gas, mostly carbon dioxide. The electrolyte also breaks down into an acid that eats away the protective layer. Their fix was to turn the electrolyte into a stretchy polymer gel after the cell was sealed, by warming it to 50 °C for an hour. The gel coats the silicon particles and holds them together as they swell. In cells whose anodes were 20 percent silicon oxide, capacity after storage rose from 64 to 77 percent, and the cells barely swelled. Over 180 charges and discharges, the gel cells kept 84 percent of their capacity, against 74 percent without it. **The limits.** The cells were pouch cells – flat, foil-wrapped batteries – of about 3 amp-hours, roughly a phone battery's capacity. The methods describe building them in the lab, though the paper also calls them commercial. Their cathodes, the other electrode, were lithium cobalt oxide, used mainly in phones; most electric cars use other types. The storage test ran four weeks, fully charged, at 40 °C; results at 50 and 60 °C are in a supplementary file. The gel was tried at only one silicon level, and it made current flow less easily in new cells. The paper does not say how many cells were tested for each result. **[Read the paper](https://doi.org/10.26599/NR.2026.94908634?ref=clever-new-stuff.ghost.io):** *Huang et al., Nano Research 19 (2026) 94908634\. Free to read.* --- ###### MATERIALS AND MANUFACTURING ### 2-hour hold strengthens 3D-printed jet-engine alloy Researchers at Beijing University of Technology report a heat treatment that makes 3D-printed IN718 about 45 percent stronger than as printed. IN718 is a nickel superalloy – a metal that keeps its strength when hot – used in jet engines and gas turbines. Holding the printed metal at 900 °C for two hours, before the standard final heat treatment known as aging, raised its tensile strength – the load at which it breaks – to about 1,480 megapascals (MPa), a measure of force per unit of area. As printed, it broke at about 1,010 MPa. Earlier heat treatments reached 1,120–1,320 MPa for IN718 printed the same way, and 1,370 MPa for a laser-printed version. If the result holds, makers printing large IN718 parts could get stronger parts by changing only how long they heat them. The parts were made by arc printing, which builds metal up by welding wire, layer on layer. For large parts it is cheaper and faster than printing from metal powder, but heat treatments designed for cast and forged IN718 suit the printed metal poorly. The authors first heated the printed metal to 1,150 °C for two hours to even out its make-up. They then held it at 900, 930 or 960 °C for up to 16 hours, and aged it. Strength rose and fell twice as the holding time grew. At 900 °C, the authors put the first peak, at 20 minutes, down to small particles of a nickel-niobium compound called the delta phase. These form along the borders between the metal's crystals and hold them in place. They put the second, higher peak, at two hours, down to delta phase that has grown into thin needles. Under load, the needles snap again and again into small pieces, which strengthen the metal as the small particles do. They report that the cracks stayed in the needles and did not spread into the metal around them. Beyond about eight hours, strength fell, as the delta phase used up niobium the alloy needs for its main strengthening. **The limits.** All tests were at room temperature, but IN718 is chosen for parts that run hot. The paper reports no tests at high temperature, of fatigue (wear from repeated loading) or of creep (slow stretching under a steady load when hot). The heat-treated metal stretched about 6–8 percent before breaking, against about 12 percent as printed. The test pieces were small, three or four for each result, all cut in the same direction from a single printed wall. The authors do not share their data, citing ongoing work. **[Read the paper](https://doi.org/10.1016/j.jmrt.2026.09.165?ref=clever-new-stuff.ghost.io):** *Zhou et al., Journal of Materials Research and Technology 45 (2026) 1603–1623\. Free to read.* ***Reply with corrections, tips or leads. If this was useful, forward it to someone who’d want it. Subscribe at clevernewstuff.com.***