With their Stress-Guided Lightweight 3D Designs (SGLDBench) benchmark, the researchers have succeeded in overcoming these serious obstacles.

2.) On Alloys
https://phys.org/news/2025-11-future-ai-deciphers-alloy-microstructures.html

“We are sending these materials into increasingly extreme environments,” Stinville said. “They are exposed to intense environments; for instance, structural materials for space applications must be resistant to mechanical loading under extremely low or high temperatures.

“Conventional alloys don’t do as well in these conditions because their mechanical properties tend to degrade under these extreme environments. We want to find new ways to accelerate the identification of alloy chemistries and microstructures that can withstand these harsh conditions.”

36

Tweet

Share

Email

Home
Chemistry
Analytical Chemistry
Home
Chemistry
Materials Science
November 18, 2025

The GIST
Mapping the future: AI deciphers alloy microstructures to enhance properties prediction and design
by University of Illinois Grainger College of Engineering

edited by Lisa Lock, reviewed by Robert Egan
Editors’ notes
(A) Physics-based approach to extract physical descriptors from the reciprocal space and map microstructure heterogeneity through stereographic projection of physical directions (inverse pole figures). (B) The individual Kikuchi patterns are encoded to comprehensively map and highlight microstructural heterogeneity. Credit: npj Computational Materials (2025). DOI: 10.1038/s41524-025-01770-8
In a world of 8 billion people, there’s one thing that makes each of us unique: our fingerprints. A variety of genetic and environmental factors create tiny variations in the skin’s ridges and whorls, such that no two prints are the same.

The spatial distribution of these subtle features makes fingerprinting a useful tool for biometric identification. With the help of modern technology, we can even unlock our personal devices using digital maps made from our skin’s unique arrangement of ridges, valleys and vascular patterns. These technologies succeed because of their ability to spatially capture the arrangement of super-fine detail.

This evolution of recognition technology is mirrored in the field of materials science, where researchers seek new and efficient ways to fully characterize materials, accelerating the discovery of additional new materials. Much like human fingerprints, the performance of metal mixtures called alloys relies on the intricate spatial arrangement of microstructural features. Traditional methods reduce this complexity into a handful of averaged values, causing each alloy to lose its distinctive “fingerprint.”

In a recent complement of papers from the lab of Jean-Charles Stinville, assistant professor of materials science and engineering, Illinois Grainger engineers have introduced new machine learning approaches for identifying alloy microstructures and predicting their properties rapidly. The Illinois researchers’ method will provide new avenues for faster and more efficient materials design.

Microstructures are tiny structural features of metals that influence their strength and behavior. Scientists look to the microstructural properties of metals to assess their functionality. Metals used in propulsion devices like rockets and airplanes have special requirements.

“We are sending these materials into increasingly extreme environments,” Stinville said. “They are exposed to intense environments; for instance, structural materials for space applications must be resistant to mechanical loading under extremely low or high temperatures.

“Conventional alloys don’t do as well in these conditions because their mechanical properties tend to degrade under these extreme environments. We want to find new ways to accelerate the identification of alloy chemistries and microstructures that can withstand these harsh conditions.”

The complete details of these microstructures, including small-scale influential variances called heterogeneities, cannot be easily captured by existing methods. Instead, Stinville and his colleagues used deep learning to analyze diffraction patterns, or the way electrons interact with metals.

By encoding these interactions through a machine learning method onto a spatial latent representation, the researchers captured the full extent of an alloy’s microstructure and its heterogeneity—an approach Stinville calls Material Spatial Intelligence.

“Traditionally, we have used single descriptors or average values to guide data-based alloy design,” he said. “But spatial information from local measurements over a large field of view allows us to capture the microstructure heterogeneity of the alloy. Using such spatial information in a data-based model provides significant improvement in prediction accuracy and enables alloy and microstructure design.”

Published in npj Computational Materials, the initial model is a machine learning approach that successfully identified microstructures and material heterogeneity in unprecedented detail. In a second paper published in Scripta Materialia, Stinville further progressed the model towards the prediction of mechanical properties using the developed approach of material spatial intelligence. This method accelerates alloy property prediction by orders of magnitude and provides a rapid fundamental understanding of structure properties in metals.

Sketch to model via A.I.
https://techxplore.com/news/2025-11-ai-agent-3d-cad-software.html

Anti-matter storage, woo woo! Now let’s just get us some dilithium crystals and it’s warp drive here we come!

The anti-hydrogen front
https://phys.org/news/2025-11-physicists-antihydrogen-breakthrough-cern-technique.html

Wiring
https://phys.org/news/2025-11-cable-mitigates-flaws-superconducting-wires.html

I’m simply dubious that math sims are enough to validate the new Orion EDL profile sufficiently to entrust the lives of crew to it. An unmanned test would settle that issue without crew risk.

I think the new Avcoat formula for Artemis 3’s Orion most likely is better than the current one. But it needn’t be tested at full-scale with an actual SLS-Orion launch to verify that. A sub-scale test article could likely be flown by a Falcon 9 on a trajectory that uses that powerful 2nd stage to drive it back into Earth’s atmosphere at lunar return velocity. Aim it at Point Nemo and also give it the same dunking it would get on an actual Artemis return. If the first such test proves disappointing, there would be time to run a second with a still different formula or to revert to a heat shield based on the current formula assuming the notional unmanned Artemis 2 test was adequately successful. If Isaacman de-crews Artemis 2, he should also authorize the fabrication of a current-formula heat shield for Artemis 3’s Orion as a contingency.

Jeff Wright,

Not every project will have a Casey Handmer. Only the projects that deserve one.

The basis of that 7:1 claim for Apollo has always seemed questionable to me. Still, given how much of what Apollo developed along the way was actually new, the ratio likely was positive.

But SLS-Orion is not Apollo. There is nothing new coming from it. From the standpoint of the general economy, the project is a dead loss. You are simply unable to see NASA as the stooped and wrinkled crone it has become, maintaining an unsupportable fantasy that it is still the gorgeous ingenue it was in the ’60s.

Apollo gave Americans a seven to one return.

Well, again: Apollo wasn’t averaging a launch once every four years.

That NASA simply no longer exists.

On SETI
https://phys.org/news/2025-11-machine-framework-scan-extraterrestrial-life.html

https://x.com/CJHandmer/status/1990547737956528218

I just want to cry.

Not worth wasting my readers’ time on it.

No, you’re right, it really isn’t.

Just amused that Lockheed actually thought it was necessary to pay for this thing right now in the first place., and I thought you might have the same sense of amusement.

I hope Jared I. is quickly confirmed as NASA Administrator and also hope one of his first acts is to de-crew Artemis 2 and reassign them to Artemis 3. An extra 2.5 – 3 years before flying, but with much greater assurance of returning alive and with a walk on the lunar surface as a bonus for two or three of them – or all four if Elon’s SLS-Orion replacement is ready to go at the same time as Starship HLS. Launch them both at the same time from Starbase pads 1 & 2, rendezvous and refill from two different depot ships in LEO, then travel to the Moon in formation.

The difficulty is that time is running out. Even if the Senate expedites his confirmation process, I think he’d be lucky to be in place by Christmas. Meanwhile, Artemis II is scheduled for its first attempt at launch in….11 weeks.

Isaacman doesn’t know the details of the heat shield situation, so he’d have to commission some new effort to look at it again (hopefully including Charlie Camarda!), and involve himself in that to get up to speed. Not much time for that. I suppose he could delay the mission until this work is finished, though….

But here is the problem I have. This heat shield configuration is only going to be used once, right? They’re changing it for Artemis III’s Orion. So what point to flying an uncrewed test of a heat shield version that is now obsolete? What you really want is to test out the *new* heat shield configuration. If that is the case, then I cannot help but think that you just cancel Artemis II as it exists right now, outright, pull the Orion, remove its heat shield entirely, and then once one is ready, install the new version. Then you fly the whole thing uncrewed on a test flight, repeating Artemis I unless you can find a way to generate lunar return velocities on a high earth orbit trajectory.

But this creates other problems, the most obvious of which is that it delays the revised Artemis II mission by 2-3 years (as you say and maybe this is what you meant by this), and in turn that delays Artemis III siimilarly. Which denies Trump any PR triumph on his watch. Also, the SRB’s are stacked, and that’s too much time for them to be sitting stacked. So you basically have to scrap those and stack new ones when the time comes.

Of course, the simple, best solution is just to cancel SLS and Orion outright. But Ted Cruz and friends ain’t having none of that.

I’m sure Lock-Mart paid handsomely for that SpaceNews placement. Perhaps the paywall is not exactly proving to be a goldmine?

I have no idea of what their financials look like, but if they are like other trade papers, they have to be getting most of their revenue from advertising anyway.

“I would laugh if it weren’t so sad.”

“If you have no tragedy, you have no comedy. Crying and laughing are the same emotion. If you laugh too hard, you cry. And vice versa.”

Sid Caesar

But, I could be wrong.

Reminds one a bit of that old saying about lawyers – “When you’ve got the law, pound on the law. When you’ve got the facts, pound on the facts. When you’ve got neither, pound on the table.”

Richard M: I saw that PR puff piece and laughed. Very embarrassing. Not worth wasting my readers’ time on it.

I hope Jared I. is quickly confirmed as NASA Administrator and also hope one of his first acts is to de-crew Artemis 2 and reassign them to Artemis 3. An extra 2.5 – 3 years before flying, but with much greater assurance of returning alive and with a walk on the lunar surface as a bonus for two or three of them – or all four if Elon’s SLS-Orion replacement is ready to go at the same time as Starship HLS. Launch them both at the same time from Starbase pads 1 & 2, rendezvous and refill from two different depot ships in LEO, then travel to the Moon in formation.

Thanks for the Astrolab video. In addition to offloading from HLS Starship, the vid also shows FLEX doing some ops with Blue Moon Mk. 1.

I’m sure Lock-Mart paid handsomely for that SpaceNews placement. Perhaps the paywall is not exactly proving to be a goldmine?

Jeff Wright,

If you were on Arty II, the price could easily be your life. If you died, you would single-handedly cut my comment output here roughly in half.

Re: Gallium and Starship TPS,

Congratulations – you have finally figured out a way to make Starship even more expensive than SLS. Gallium costs about $7.30 per ounce. That is, of course, without considering the sharp price increase such a notional increase in Gallium demand would induce.

I would give anything to be on Arty II

A supercomputer was charged with etching a quantum Chip recently–a video of which qualifies as art
https://phys.org/news/2025-11-supercomputer-simulates-quantum-chip-unprecedented.html

Which brings me back to this:
https://techxplore.com/news/2025-11-ultra-strong-lightweight-metal-composite.html

I wonder if you could have a lobe of RCC emerge from this aluminum.

Moreover…what if low melting point gallium could come into play?

Why a low melt metal?

The idea would be to assemble a heat shield on Starship, piece by piece…melt out the metal, and fill the void with something better as the low melt metal flows out.

This could unitize Starship ‘s TPS into a single piece.

It goes on as tiles–but cannot fall off as tiles.

https://spacenews.com/orion-safeguarding-humanitys-return-to-the-moon-and-the-journey-beyond/

I would laugh if it weren’t so sad.

One assumes they have some insight into the design of the cargo version of Starship HLS, and if that is the case, I guess we are seeing what the doors and deployment mechanism look like. At least, as they cuyrrently exist in SpaceX’s design process. The level of detail is rather low, however, and that is probably just as well.

https://x.com/Astrolab_Space/status/1990506391467930050

Sent Orion around the Moon and back to Earth, and found its heat shield had serious technical flaws. So of course NASA is going to fly astronauts around the Moon in Orion early next year, without fixing the problem. Par for the course for this incompetent agency.

To give them their due, NASA’s position seems to be that they DID fix the problem – first, by keeping the shield as is but altering the reentry profile on Artemis II, and then modifying the shield on Artemis III going forward by making the Avcoat material more permeable and modifying the manufacturing process. They insist that after exhaustive testing and modeling, they have total buy-in from the IRT (independent review team) for this resolution.

Of course, it’s obvious what the problems with NASA’s “fix” are. If the ultimate answer is a redesigned heat shield, why not hold off and do it on the Artemis II mission, too? Well, we all know the answer why – it would push the mission back *years*, and NASA management has decided that this is not institutionally/politically acceptable. The same difficulty applies to the other alternative, which is to fly the next mission uncrewed to get actual flight data on the new reentry profile before risking human lives on it.

Meanwhile, Charlie Camarda insists that two members of the IRT *do* have reservations, and NASA is not being forthcoming. He insists that NASA cannot be trusted, as it is, and that its assertions can no longer be taken at face value. “NASA did not post the results of the IRT,. Why wouldn’t they post the results of what the IRT said? If this isn’t raising red flags out there, I don’t know what will.” The IRT’s chairman, Paul Hill, insists that Camarda is wrong about there being dissenting views on the team, but agrees that it was wrong for NASA not to publish the results.

None of this should fill any of us with confidence about this mission. I hope we’re all wrong, because NASA seems determined to fly it, and if we are right, four astronauts may die.

Meanwhile, we reflect on the fact that we are on the three year anniversary of Artemis I, and no subsequent Artemis mission has flown. NASA went only six months between its final uncrewed test flight of Saturn V/Apollo and the first crewed mission in 1968. It’s astounding how agonizingly drawn out this schedule has become. But SpaceX has done 11 uncrewed development test flights of Starship since Artemis I flew.