1. Summary / before you read / place in the series

• Microgravity is not "no gravity" but "continuous free fall": even at the altitude of the International Space Station (ISS), gravity is roughly 90% of its ground value. The station and everything in it fall around the Earth together, so objects appear to float relative to the floor. [Ref. 1]
• With weak gravity, convection and settling in a melt nearly disappear: crystal and alloy growth can potentially be observed and controlled without being stirred by convection. Some effects remain, such as flows driven by surface tension. [Ref. 6, 9]
• Research has run for over 50 years, but "better products at routine scale" has not been confirmed: ZBLAN fiber, SiGe crystals and alloy property measurements are being studied on the ISS and elsewhere. Most results are at the research and demonstration stage; commercial viability could not be confirmed. [Ref. 13, 14]

2. What microgravity is - and is not

(1) A familiar feeling: inside a falling elevator

If an elevator cable snapped and the whole car fell freely, the person inside would stop pressing on the floor, a bathroom scale would read zero, and they would seem to float. Gravity has not gone away; the car and the person fall with the same acceleration, so no force acts between floor and person. Astronauts appear to float for the same reason. (Never try this in a real elevator; it is only a way of thinking.)

(2) An orbit is "falling continuously while moving sideways fast enough"

Then why does the station not hit the ground? Newton's thought experiment answers it. Fire a cannon horizontally from a high mountain: a slow shell lands near, a fast one lands far. If it is fast enough, the ground curves away as fast as the shell falls, and the shell keeps falling without ever hitting the ground. That is an orbit. The ISS travels at about 28,000 km/h. [Ref. 1]Sourced

(3) Gravity does not weaken quickly with altitude

NASA's page says gravity at about 250 miles (about 400 km) is about 88.8% of its ground value. [Ref. 1] Gravity weakens with the square of distance, and 400 km is small next to the Earth's radius of about 6,371 km. The chart uses the same law, calculated by this article. Even at geostationary altitude (35,786 km) about 2% remains, and at the Moon's distance about 0.03%. Leaving the ground does not switch gravity off.SourcedOur calculation

(4) "Micro" means very small, not zero

NASA defines microgravity as the condition of free fall and puts its size at about one millionth of ground gravity. [Ref. 1] "g" is a unit where ground gravity is 1, so 10^-6 g is one millionth of that. This is only a guide. A report from NASA Glenn's measurement group (PIMS) states that weightlessness is an ideal state that cannot be achieved in practice aboard a spacecraft: crew activity, life-support equipment, drag, gravity gradient, rotation and structural resonances all create small accelerations. [Ref. 2] ISS experiments therefore use accelerometers to record the environment and to interpret results.Sourced

3. What changes inside a melt when gravity is nearly gone

(1) What gravity does in a melt on the ground

Crystals and alloys form as a molten liquid (the melt) solidifies. On the ground gravity does three things inside a melt. (1) Buoyancy convection: warmer regions are less dense and rise, cooler ones sink, so the melt is stirred. (2) Settling and rising: particles, bubbles and immiscible liquids separate by weight. (3) Contact with a container: the melt rests on a container and takes contamination and stress from the wall. [Ref. 8, 9]Sourced

View original SVG, values and sources

Original SVG reference: Conceptual diagram (vector drawing). Melt on the ground (left) and in microgravity (right). Arrows and particles are schematic and do not show real flow sizes.

A paper on the ISS electromagnetic levitator (a device that floats and melts a metal sphere) notes that on the ground the strong electromagnetic fields needed to fight gravity cause turbulent flow that prevents measuring properties such as viscosity. In microgravity the positioning forces are weak, flow becomes close to laminar (non-turbulent), and nearly spherical droplets allow surface tension and viscosity to be measured more precisely. [Ref. 8]Sourced

(2) Checking with numbers: how much slower is settling?

How does "one millionth of gravity" affect settling? The speed at which a small sphere sinks in a liquid is proportional to gravity (Stokes' law: v = Δρ·g·d²/18η, where Δρ is the density difference, g gravitational acceleration, d the particle diameter and η viscosity). Take a 10 μm particle in a water-like liquid as an example.Our calculation

Assumptions: particle diameter 10 μm, density difference 1,000 kg/m³, viscosity 1 mPa·s, 293 K, g = 9.8 m/s². Brownian spread is the one-direction rms √(2Dt) with D = kT/(3πηd) ≈ 4.3 × 10^-14 m²/s. A rough Our calculation, not a value for a real melt.

Reading: a particle that sinks 20 cm (to the bottom of a container) in an hour on the ground moves only 0.2 μm in microgravity, and its thermal jiggling (about 18 μm) is far larger. Settling effectively ends and particles move only by diffusion. For the same reason bubbles and immiscible liquids separate less readily (our commentary).

More detail: buoyancy convection also weakens in proportion to gravity

The dimensionless number that expresses the strength of buoyancy convection driven by temperature differences (the Rayleigh number) is proportional to gravitational acceleration. If gravity falls to a millionth, so does this number, and the driving force of convection drops by orders of magnitude (a textbook relation; our commentary). A paper on ISS experiments likewise says that in microgravity convection is absent or reduced, so growth is controlled by diffusion alone. %s Flows not driven by buoyancy (next table) remain.

(3) Why this matters for crystal growth

Around the growing face of a crystal (the growth interface), dopants and dissolved components are either taken into the crystal or pushed back into the liquid, so concentration changes. With convection this concentration fluctuates over time and leaves stripe-like compositional variation (striations) in the crystal. In microgravity convection vanishes and matter is carried by diffusion alone, so growth can be compared directly with simple theory. [Ref. 9] What matters scientifically is that gravity becomes a knob that can be turned to isolate the cause of defects.Sourced

(4) What remains

Sources: NASA SP-443 and the PIMS report. The table layout is our commentary.

4. How microgravity is obtained

Microgravity exists only while something is in free fall. A short drop gives short times; reaching orbit gives long ones. The method is chosen to suit the experiment.

Sourced Figures as stated by each source. Levels are measured differently, so do not rank methods by them.

5. A research history from Skylab to the ISS

(1) Skylab (1973-1974): the first extensive experiments

NASA's SP-443 (1980) says interest in materials processing in space began in the late 1950s; after limited tests in drop towers and some of the last Apollo flights, the Skylab solidification and crystal-growth experiments were the first extensive demonstration of materials processing in space. [Ref. 6] Skylab was a US space station launched on 14 May 1973; Skylab 3 (from 28 July 1973) and Skylab 4 (16 November 1973 to 8 February 1974) crews stayed aboard. [Ref. 7]Sourced

A representative case is experiment M562 (Witt and Gatos). Crystals of indium antimonide (InSb) doped with tellurium (Te) or tin (Sn) were partly remelted aboard Skylab 3 and 4, held for 60 minutes to reach thermal equilibrium, and solidified in one direction by controlled power reduction. After sectioning and etching, the part grown on the ground showed pronounced dopant striations, while the part grown in space showed no microscopic compositional fluctuations. SP-443 states this shows that growth-rate fluctuations are gravity-related and can be eliminated in low gravity. [Ref. 6]Sourced

Results were not always dramatic. In experiment M557 on immiscible metals, samples solidified in space had a more homogeneous structure, but not the fine, uniformly dispersed structure obtained for gallium-bismuth in drop-tower tests. [Ref. 6] SP-443 also notes that many experiments gave unexpected results. From the start, cases where the benefit was confirmed and cases where it was not both appeared.Sourced

(2) The ISS era: measurement and testing theory

Sourced Hicari-II is as described by JAXA and the paper (Materialia, May 2026). It concerns growth rate; it does not claim crystals better than ground-grown ones.

6. ZBLAN optical fiber: the flagship of microgravity manufacturing research

(1) What ZBLAN is and what is expected of it

ZBLAN is a heavy-metal fluoride glass made of fluorides of zirconium, barium, lanthanum, aluminum and sodium (ZrF4-BaF2-LaF3-AlF3-NaF); the name strings together the element initials. [Ref. 12] It transmits longer infrared wavelengths (mid-infrared, 2-4 μm) than ordinary silica fiber, so it has long been eyed for long-distance links and for infrared sensors and lasers. The 2026 paper states that the theoretical attenuation of ZBLAN in the mid-infrared is up to 20 times lower than the minimum loss of silica (at about 1.55 μm) and that intrinsic loss is below 0.01 dB/km in the 1.5-2.5 μm region. [Ref. 14] A 1994 NASA technical memorandum gives the theoretical loss coefficient as 0.001 dB/km at 2 μm. [Ref. 12] These are theoretical values; as the table below shows, real ZBLAN fibers are still far from them.Sourced

(2) Why it is hard to make: crystals form

According to the 2026 paper, ZBLAN has a higher critical cooling rate (the minimum cooling speed needed to solidify without crystallizing) than silica, so crystals form readily both when the glass is made and when the fiber is drawn. The cooling rate during drawing is set by draw speed and tension, but theory and experiment indicate that avoiding crystallization would need impractical speeds and tensions. Impurities in the raw materials (transition metals, hydroxyl groups, oxides) not only absorb light but can also seed crystals. NASA's page says crystals scatter light and make the fiber brittle. [Ref. 14][Ref. 13]Sourced

(3) Why microgravity slows crystallization

NASA's page says only that crystals grow more slowly in microgravity, but the 2026 paper gives a model-based explanation. Crystals need matter to be transported, either by molecular diffusion or by buoyancy-driven convection. At the drawing temperature ZBLAN has a high viscosity of about 10^5 Pa·s and diffusion coefficients of about 10^-11 to 10^-10 m²/s. Transport over a millimetre by diffusion alone takes hours to a day (the paper's statement). On the ground (1 g), buoyancy convection therefore contributes substantially to transport and speeds crystal growth. In microgravity the convective contribution disappears, transport becomes diffusion-only, and crystallization is suppressed. The paper also says that at higher temperatures viscosity drops, diffusion dominates, and the effect of convection, and hence of microgravity, is negligible. [Ref. 14] This is the "no convection" effect of Chapter 3 at work in a specific case.Sourced

View original SVG, values and sources

Original SVG reference: Conceptual diagram (vector drawing). Left: schematic of fiber drawing. Upper right: schematic of crystallization. Lower right: timeline of key events (as stated on the NASA page and in the paper). Number and size of crystals are illustrative.

(4) What has been tried, from the 1990s to the ISS in 2024

1990s (ground-based aircraft and similar): Heating tests in short microgravity (mostly under two minutes) showed reduced crystallization. A 1994 NASA technical memorandum reports that ZBLAN fiber in evacuated quartz ampoules was heated to the crystallization temperature on the KC-135 aircraft; the 1 g and 2 g samples slumped and crystallized, while the 0 g samples showed no evidence of crystallization. This was a heating test to look for crystallization, not fiber drawing. [Ref. 12][Ref. 14]Sourced

ISS: According to NASA's page, ISS ZBLAN development began in 2014, and work included the first privately funded ZBLAN draw (Made In Space Fiber Optics), FOP-2 (which repeated 20 m draws) and Space Fibers from FOMS Inc. From mid-February to mid-March 2024, Flawless Space Fibers-1 (Flawless Photonics, supported by the ISS National Laboratory) is reported, as a preliminary result, to have made more than seven miles (11.9 km) of optical fiber, run eight draws of more than 700 m and drawn about 1,141 m in one day (the previous space record being 25 m). Three of the four goals of NASA's InSPA program are described as achieved, while the fourth, "ten times better quality than Earth-made fiber," depended on analysis after return and was undetermined at the time of the page. [Ref. 13]SourcedNot yet confirmed

The July 2026 paper (results): From about 12 km, a total of about 2.1 km in the sections with the most uniform diameter was compared with fiber drawn on the ground. Ground and ISS used the same draw-tower design and the same preforms, and on the ISS the fiber was drawn inside the Microgravity Science Glovebox with minimal crew intervention. Loss was measured by several institutions with the cut-back method, in which a fiber is shortened repeatedly. The table gives results at a wavelength of 1,550 nm. [Ref. 14]Sourced

Source: Fig. 1 and Table 1 of Arat et al. (2026), at 1,550 nm. Sourced Units are dB/m (1 dB/m = 1,000 dB/km).

Strength (tensile testing) was compared too. For 421 segments (230 from space, 191 from ground), median breaking strength was +25% for single-index and +80% for step-index fibers, and the Weibull modulus (a number expressing how small the scatter in strength is) was +39% for step-index fibers (-5% for single-index). [Ref. 14]Sourced

(5) The limits the paper itself states, and the company claim

The paper splits loss into five parts: (1) intrinsic loss, (2) absorption by transition-metal and rare-earth impurities, (3) scattering from bubbles and interface imperfections, (4) scattering from crystals formed during drawing and (5) loss in the polymer coating. Of these, only (4) is affected by microgravity. (2) (about 60 dB/km = 0.06 dB/m at 1.5 μm for the raw-material purity of this study) and (3) do not depend on gravity. The paper concludes that microgravity "substantially reduces crystallization-driven loss" but "yields no benefit when attenuation is dominated by impurity or noncrystalline defect contributions," and stresses that improving raw-material purity is important. [Ref. 14]Sourced

External benchmarks are also given. The best ZBLAN value reported in 1992 was 0.65 dB/km at 2.56 μm (about 0.3 dB/km of it scattering); even if all the scattering came from crystals, microgravity could remove at most about 0.3 dB/km, a reduction of total loss by about half. Commercial multimode ZBLAN fibers are said to show more than ten times lower loss than the single-index fibers of this study at 1.5 μm. [Ref. 14] Converted by this article, the mean 0.75 dB/m of the space-drawn step-index fiber is 750 dB/km, more than 1,000 times the value above (a reference only, because the wavelengths differ; our calculation). So this study showed that microgravity reduces crystal-related loss; it did not show that low-loss ZBLAN fiber has been made in space.SourcedOur calculation

The company claim: On its own website, Flawless Photonics states that it is "the only company to successfully manufacture commercially viable lengths of optical fiber in micro-gravity, pulling over 11 kilometers at the International Space Station and returning it to Earth," and lists low attenuation and wide bandwidth as features of its product "SpaceFiber." [Ref. 15] This is a company claim. Note that the five institutions involved in the paper's evaluation include Flawless Photonics itself, which made the fiber. [Ref. 14] Shipments, prices, mass-production plans and customer evaluations could not be confirmed in this check. The paper says future missions are planned to reduce bubbles and crystallization through improved ground casting and in-space glass melting; this is Not yet confirmed. [Ref. 14]SourcedNot yet confirmed

7. Semiconductor crystals, alloys and company demonstrations

(1) Semiconductor crystals and alloys: research aims mainly at understanding

For semiconductor crystals, Skylab InSb (where dopant striations disappeared) and current SiGe growth-rate work are representative. [Ref. 6, 10] For alloys, ISS-EML property measurements and observations of transparent model alloys are representative. [Ref. 8, 9] What these studies mainly aim at is to understand growth and solidification without convection and to use the knowledge for ground-process design and for testing simulations. This check found no evidence that semiconductors or alloys made in space are used in ground products.Sourced

(2) Space Forge ForgeStar-1: a plasma demonstration on a free-flying craft

The official page of Space Forge (UK), dated 25 March 2026, says its satellite ForgeStar-1 was launched to low Earth orbit in June 2025 and that, using its in-house growth chamber, it has generated plasma in orbit more than 100 times. The company says this shows that the extreme conditions needed for gas-phase crystal growth, a core building block of semiconductor production, can be created and controlled on an autonomous platform in orbit, and claims it is the "first free-flying commercial semiconductor manufacturing tool" ever operated in space. [Ref. 16] This is a company claim.SourcedNot yet confirmed

What the page states goes as far as "plasma was generated." Growing a crystal in orbit, returning it to Earth and measuring its quality were not stated within what we checked. The stage is therefore "part of the conditions needed for growth were created on a small free-flying craft," not established manufacturing.

8. A materials engineer's view, and what could not be confirmed

What could not be confirmed

• The cost, yield and repeatability of ISS ZBLAN drawing (the paper reports one mission; repeated reproduction could not be confirmed)
• Whether the paper's measurements follow IEC 60793-1-40 (the paper says "cutback method" but no standard number appeared in the text; the reference list was not checked)
• Shipments, prices and customer evaluations of ZBLAN fiber; crystal growth, return and quality for Space Forge
• The microgravity level of parabolic flights (not on the ESA page)
• Any case of semiconductors or alloys made in space being used in ground products

Summary

Microgravity is not an absence of gravity but a state of free fall, and it nearly removes convection and settling in a melt. Experiments that turn this "gravity knob" have been used for over 50 years, from Skylab onward, to understand crystals and alloys. Work on ZBLAN fiber, SiGe crystals and alloy properties continues on the ISS, but within what could be confirmed it is at the research and demonstration stage. The value of "making it in space" has to be judged by comparison with controls, by quantity and repeatability, and by comparison with ground methods.

9. How to read the yardsticks and standards

Two kinds of yardstick appear in space manufacturing. One is the quality of the microgravity environment (how small, how long and how quiet); the other is the quality of the material produced (loss, uniformity, growth rate, properties). The environmental yardstick looks at time, level and vibration (frequency). Material yardsticks can use the same measurements as on the ground, but it is essential to compare space and ground under the same conditions. Checked on 17 October 2026.

The arrangement of the table is Our calculation (organization). No applicable ASTM standard was found.

Reading notes (for materials engineers)

• Ground tests and orbit differ: the 1994 aircraft heating test did not draw fiber and is a different stage from the 2024 ISS drawing.
• Company statements and independent verification differ: the 2026 paper is a joint evaluation by five organizations, but they include Flawless Photonics itself, which made the fiber. "Only" and "commercially viable" are the company's own statements.
• Match units and conditions: ZBLAN loss changes greatly with dB/m versus dB/km, wavelength (1.5 μm versus 2.56 μm) and fiber type (single-index versus step-index). The paper's comparison used the sections of most uniform diameter, not the whole-length average.
• Plans and achievements differ: InSPA's "ten times better quality" was undetermined at the time of the NASA page. Read business plans for manufacturing in space as Not yet confirmed.

10. Glossary

11. References

1. NASA What Is Microgravity? (ISS Research: free fall, about one millionth, gravity versus altitude, orbital speed) — www.nasa.gov/missions/station/iss-research/what-is-microgravity/
2. NASA Glenn (PIMS) Principal Investigator Microgravity Services, ISS Increment-2 Quick Look Report (May to June 2001) (accelerometers SAMS and MAMS, quasi-steady and vibratory regimes, "weightlessness is an ideal state") — gipoc.grc.nasa.gov/pims/pimsdocs/public/PIMS%20Report/PIMSOuickLook.pdf
3. ZARM (University of Bremen) Drop Tower Bremen: General information (drop mode 4.7 s, catapult mode up to 9.3 s, below 10^-6 g) — www.zarm.uni-bremen.de/en/drop-tower/general-information.html
4. ESA Airbus A310 Zero-G (parabolic flights: about 31 parabolas per flight, about 20 s of free fall) — www.esa.int/Our_Activities/Human_Spaceflight/Research/Airbus_A310_Zero-G
5. Airbus Airbus-built TEXUS sounding rockets take to the skies (TEXUS 59 and 60, 2024: about 6 min, 362 s, about 10^-4 g) — www.airbus.com/en/newsroom/news/2024-03-airbus-built-texus-sounding-rockets-take-to-
6. NASA NASA SP-443 "Materials Processing in Space: Early Experiments" (Naumann and Herring, 1980: Skylab M562, M557 and others) — ntrs.nasa.gov/citations/19810007559
7. NASA Skylab (launched 14 May 1973; Skylab 3 and 4 dates) — www.nasa.gov/history/skylab/
8. npj Microgravity Mohr M. et al. "Electromagnetic levitation containerless processing of metallic materials in microgravity: thermophysical properties" (2023, 9:34, DOI 10.1038/s41526-023-00281-4) — doi.org/10.1038/s41526-023-00281-4
9. npj Microgravity Akamatsu S. et al. "Microgravity studies of solidification patterns in model transparent alloys onboard the International Space Station" (2023, DOI 10.1038/s41526-023-00326-8) — doi.org/10.1038/s41526-023-00326-8
10. Materialia Arai Y. et al. "Increasing SiGe crystal growth rate utilizing microgravity" (May 2026, DOI 10.1016/j.mtla.2026.102709) — doi.org/10.1016/j.mtla.2026.102709
11. JAXA JAXA news item on the first Hicari-II paper (12 June 2026: 0.15 mm/h, about 20%, 0.12 mm/h) — humans-in-space.jaxa.jp/en/biz-lab/news/detail/005790.html
12. NASA Tucker D. S. "Effect of Microgravity on Crystallization of ZBLAN Fibers", NASA TM-108461 (June 1994) — ntrs.nasa.gov/citations/19950003911
13. NASA Optical Fiber Production (ISS Research: Flawless Space Fibers-1, February to March 2024, preliminary results) — www.nasa.gov/humans-in-space/optical-fiber-production/
14. Space: Science & Technology Arat et al. "Space-Based Fabrication of ZBLAN Optical Fibers: A Breakthrough in Microgravity Manufacturing" (23 July 2026, Vol. 6, 0572, DOI 10.34133/space.0572) — spj.science.org/doi/10.34133/space.0572
15. Flawless Photonics Company website (SpaceFiber; the company's own claim to be the only one to make commercially viable lengths in microgravity) — www.flawlessphotonics.com/
16. Space Forge ForgeStar-1 mission page (dated 25 March 2026: plasma generated more than 100 times; company claims) — www.spaceforge.com/missions/forgestar-1
17. IEC IEC 60793-1-40 "Optical fibres - Part 1-40: Attenuation measurement methods" (IEC webstore page for the 2019 edition: four methods including cut-back; the 2019 edition is withdrawn) — webstore.iec.ch/en/publication/60732
18. EVS (Estonian Centre for Standardisation) EVS-EN IEC 60793-1-40:2025 (base document IEC 60793-1-40:2024; valid from 3 February 2025) — www.evs.ee/en/evs-en-iec-60793-1-40-2025

12. Claim-to-source audit

Last updated: 17 October 2026. Sources are limited to primary materials (NASA, ESA, JAXA, ZARM, Airbus, NASA SP-443, peer-reviewed papers, IEC and EVS, company pages). The cost, yield and repeatability of ISS ZBLAN drawing, crystal growth, return and quality for Space Forge, a common standard for microgravity levels, and ASTM standards could not be confirmed and are not stated. Company claims of "first" and "only" are treated as company claims, and all figures are conceptual diagrams or figures including our own calculations. This is not medical or investment advice.