Ин таҳқиқот меомӯзад, ки interfacial reactions occurring during heat treatment дар aluminum-matrix syntactic foams reinforced with hollow glass microspheres mechanical strength, compressive deformation ва energy-absorption capacity-ро чӣ гуна change мекунанд. Researchers same MZ40-type glass microspheres-ро ба ZL101, ZL104 ва ZL102 Al-Si alloys containing approximately %7, %9 ва %12 silicon by mass тавассути vacuum-pressure infiltration ҷойгир карданд. Composites аввал at 400 °C annealed шуданд; як қисми онҳо баъдан at 550 °C for 10 hours solution-treated, water-quenched ва at 170 °C for 14 hours aged шуданд.

During solution treatment aluminum бо SiO₂ in microsphere shell reacting шуда, porous and loose reaction layer consisting predominantly of γ-Al₂O₃ formed кард. Дар HGM/ZL101-STA sample full-reaction appearance дар %70,1 of examined large microspheres observed шуд; this fraction in HGM/ZL104-STA was %37,3 and in higher-silicon HGM/ZL102-STA only %8,2. Results show that as matrix silicon fraction increases, interfacial reaction is kinetically suppressed.

Intense reaction impaired load-bearing capacity of glass microspheres. Дар HGM/ZL101 after solution treatment and aging peak compressive stress decreased from 155,56 MPa to 151,32 MPa and plateau stress from 113,64 MPa to 89,18 MPa. Specific energy absorption fell from 50,30 kJ/kg to 41,04 kJ/kg. HGM/ZL104 similarly showed reduced plateau stress and energy absorption.

By contrast, in HGM/ZL102-STA where high silicon content limited reaction, heat treatment produced positive result. Peak stress increased from 194,32 MPa to 212,12 MPa, plateau stress from 138,78 MPa to 151,87 MPa and specific energy absorption from 57,82 kJ/kg to 63,89 kJ/kg. Thus benefit of matrix hardening was not overwhelmed by reaction-induced weakening of microsphere shell.

Digital image correlation results also showed damage mode changed with reaction. In annealed HGM/ZL101 a single large shear band oriented about 45° rapidly crossed specimen, whereas in heavily reacted HGM/ZL101-STA many small mutually independent shear bands formed successively. Accordingly stress curve showed sawtooth-like gradual decrease rather than single sharp drop.

Аз нигоҳи Туркия: Findings метавонанд ба development of low-density metallic syntactic foams for automotive crash elements, rail-system buffers, lightweight marine and aerospace structures, defense-oriented impact absorbers and machine-protection components in Turkey contribution диҳанд. Барои applying method in Turkey, chemical composition and crush strength of local glass microspheres, actual silicon and magnesium content of Al-Si alloy, infiltration homogeneity, reaction-layer thickness and heat-treatment window must be separately validated. Without dynamic impact, high-strain-rate, fatigue, thermal-cycling and real-component-scale tests, this study cannot directly provide design values for a specific vehicle, train, ship or protective structure.

Hollow glass microsphere/aluminum composite чист?

Hollow glass microsphere/aluminum composites are metal-matrix syntactic foams in which many closed glass spheres are distributed within aluminum or aluminum-alloy matrix. In conventional metal foams voids form as irregular cells, whereas in syntactic foams size and geometry of voids are more controlled by microspheres used.

Because glass microspheres are hollow, material density decreases. At same time sphere shells fracture during compression, surrounding metal matrix deforms plastically and voids progressively close, consuming energy. This mechanism can create long and relatively stable compression plateau.

Average density of composites used in study was approximately 1,34–1,35 g/cm³. Microsphere volume fraction was approximately %69,2–69,6 and total porosity approximately %51,2–51,5. These values show that large portion of material consists of hollow microspheres and their cavities.

Main research question

Solution treatment and aging in aluminum alloys can increase matrix strength. However shells of glass microspheres contain high proportion of SiO₂. At high temperature aluminum can react with SiO₂ and chemically transform microsphere shell.

Thus main question is: while metal matrix is strengthened by heat treatment, if glass microspheres are weakened by interfacial reaction, in which direction do overall mechanical and energy-absorption performances of composite change?

Researchers examined this balance using three Al-Si alloys with different silicon contents. Basic assumption is that as silicon content in matrix increases, chemical driving force for diffusion of reaction-released silicon toward matrix decreases, thereby slowing interfacial reaction.

Al-Si alloys used

ZL101 and ZL104 are hypoeutectic Al-Si alloys. Silicon content of ZL102 is closer to eutectic composition. Magnesium present in ZL101 and ZL104 contributes to precipitation hardening after solution treatment and aging. Since magnesium is very low in ZL102, its strength response to heat treatment is more limited.

Properties of MZ40 glass microspheres

To separate broken microspheres, flotation in deionized water was used and intact microspheres dried at 80 °C.

How was preform prepared?

Microspheres were sintered at 620 °C for 150 minutes without applied pressure. SEM image on page 3 shows spheres bonded to each other by small sintering necks.

These bonds limit displacement of microspheres against drag forces generated during liquid-aluminum infiltration and help obtain more homogeneous structure.

Vacuum-pressure infiltration

Composites were prepared in ZYQ250/250-10.1500 vacuum-pressure infiltration furnace. Main applied conditions:

After furnace evacuated and microsphere preform heated, high-pressure argon pushed molten aluminum into mold and voids between microspheres. Samples were taken from region 50 mm away from infiltration inlet.

At least five parallel samples were prepared for each composite and results averaged. However study did not provide standard deviation or confidence interval for all table results.

Heat-treatment groups

10-hour solution-treatment condition at 550 °C was selected as relatively severe to create a distinct reaction layer. Therefore results may not directly represent industrial heat-treatment schedules with shorter time or lower temperature.

Response of matrix alloys to heat treatment

Yield strengths of ZL101 and ZL104 increased more than threefold with aging. In contrast elongation values dropped markedly. Researchers explain this behavior by β″ and β′-type fine precipitates in Mg-containing alloys hindering dislocation motion.

Because ZL102 lacks Mg, precipitation hardening remained limited; increase in yield strength was approximately %32,8. Spheroidization of eutectic silicon reduced stress concentration and contributed to slight increase in elongation.

Composite density and microsphere fraction

Composite density was measured by Archimedes method:

\[ \rho_{\mathrm{sf}}=\frac{m_1}{m_1-m_0}(\rho_0-\rho_1)+\rho_1 \]

Here m₁ represents composite mass in air, m₀ mass in liquid; ρ₀ density of liquid and ρ₁ density of air.

Microsphere volume fraction:

\[ V_{\mathrm{HGM}}=\frac{\rho_m-\rho_{\mathrm{sf}}}{\rho_m-\rho_{\mathrm{HGM}}} \]

and total porosity:

\[ V_p=P_0V_{\mathrm{HGM}} \]

were calculated by these equations.

Small differences in density and microsphere volume among groups reduce likelihood that mechanical differences arise only from porosity fraction.

Interfacial reaction

Main reaction between aluminum and silica in microsphere shell:

\[ 4\mathrm{Al}+3\mathrm{SiO}_2\rightarrow2\mathrm{Al}_2\mathrm{O}_3+3\mathrm{Si} \]

Reaction shown with separate equations for liquid and solid aluminum, but chemical stoichiometry remains same. Aluminum reduces SiO₂; alumina and elemental silicon capable of dissolving or precipitating in matrix are formed.

Calculated Gibbs free-energy changes for 400 °C and 550 °C are −533,44 and −501,95 kJ/mol respectively. Negative values show reaction thermodynamically possible at both temperatures. However actual reaction amount determined by temperature, time and element diffusion rates.

How was reaction product identified?

TEM image on page 7 shows continuous reaction region about 2,3 µm thick at HGM/ZL101-STA interface.

High-resolution TEM measured lattice spacings 0,198 nm and 0,140 nm; rings in selected-area electron diffraction matched (400) and (440) planes of cubic γ-Al₂O₃ phase.

Researchers state layer is not pure γ-Al₂O₃ only, but mixture predominantly consisting of this phase and possibly small amounts of other phases.

How did silicon content change reaction?

Reaction amount continuously decreased as matrix silicon content increased. According to researchers, silicon released by reaction diffuses toward low-silicon aluminum matrix under stronger concentration gradient. When matrix already contains high silicon, this gradient and therefore reaction propagation rate decrease.

It should be remembered that “full reaction” fraction represents a specific counting criterion within study. Counting performed at 200× magnification, 630×350 µm image area, on microspheres larger than 20 µm diameter and not invaded by aluminum. Results do not directly show complete three-dimensional reaction of whole microsphere shell.

Growth of reaction layer

EDS line scan on HGM/ZL104 measured reaction layer approximately 60 nm in annealed condition and approximately 1,2 µm in STA condition.

Effect of temperature on diffusion explained by Arrhenius relation:

\[ D=D_0\exp\left(-\frac{Q}{RT}\right) \]

D is diffusion coefficient, D₀ pre-factor, Q activation energy, R gas constant and T absolute temperature. As temperature increases, D grows exponentially.

For diffusion-controlled growth of reaction layer:

\[ \frac{dx}{dt}=\frac{k}{2x} \]

relation is given. x is reaction-layer thickness, t time and k parabolic rate constant. As layer thickens, it becomes harder for Al atoms to reach reaction front and instantaneous growth rate decreases.

According to study interpretation, at 400 °C growth is mainly limited by slowness of interfacial chemical reaction, whereas at 550 °C chemical barrier weakens and element diffusion becomes primary rate-controlling step.

Why do nanopores form?

Pores approximately 30–90 nm in size observed in reaction layer. Researchers explain pore formation by two combined mechanisms:

• Formation of Kirkendall voids because Al atoms diffuse inward and reaction-generated Si atoms outward at different rates,
• Volume contraction during transformation of SiO₂ and Al into reaction products with denser atomic packing.

Considering only formation of γ-Al₂O₃ from SiO₂, molar-volume contraction calculated as %13,6; considering all solid products as %12,1.

Voids reduce density of reaction layer, create local stress concentration and facilitate crack initiation during compression. At same time they may slow further reaction growth by creating physical barriers in front of advancing atoms.

Four stages of compression curves

1. Elastic region: Stress increases approximately linearly up to about %3,2 strain.
2. Initial stress drop: Stress decreases rapidly due to microsphere fracture and shear-band formation.
3. Plateau region: Microspheres fracture progressively, metal matrix deforms plastically and structure carries load over long deformation range.
4. Densification: After approximately %61 strain, voids close and stress rises rapidly.

Densification strain of all composites is approximately %61. Main reason is use of same microspheres, similar microsphere volume fraction and same manufacturing method.

Compression-strength results

Why did heat treatment weaken ZL101 and ZL104 composites?

According to rule of mixtures, if microspheres had not been damaged, precipitation hardening of matrix could have increased strengths of HGM/ZL101 and HGM/ZL104 composites by approximately %120,46 and %113,79 respectively.

In reality peak stress of HGM/ZL101-STA decreased %2,7 and plateau stress %21,5. In HGM/ZL104-STA peak stress decreased %5,9 and plateau stress %24,0.

This result shows weakening of microspheres by interfacial reaction dominated over hardening of metal matrix. Particularly in long plateau region, load-bearing capacity is sensitive to successive fracture of weakest microspheres.

Why did ZL102 improve?

In HGM/ZL102 high silicon content greatly limited reaction. Therefore load-bearing capacity of microsphere shells was preserved and more limited solid-solution hardening of matrix could be reflected in composite performance.

In HGM/ZL102-STA peak stress increased %9,2 and plateau stress %9,4. Theoretical increase calculated for matrix hardening was %11,82. Measured increase approaching this estimate supports interpretation that interfacial damage remained low.

“Strongest” and “weakest” microsphere control

Study proposes that peak and plateau stresses are controlled by different mechanisms.

Peak stress: strongest-microsphere control

Before first major shear band forms, low-strength microspheres can fracture locally. However macroscopic shear band cannot propagate until microsphere with highest load-bearing capacity in stress-concentration region fractures. Therefore peak stress is proposed to be determined by “strongest microspheres”.

Plateau stress: weakest-microsphere control

During plateau, weak microspheres near existing shear band fracture successively, forming low-resistance path along which band can propagate. Thus plateau stress partly depends on load-bearing capacity of “weakest microspheres”.

This interpretation is used to explain why plateau stress decreases much more than peak stress in heavily reacted ZL101 and ZL104 composites. Mechanism is based not on direct measurement of individual microsphere forces, but on combined interpretation of SEM, DIC and macroscopic curves.

What did DIC images show about damage mode?

Digital image correlation results on page 15 compared HGM/ZL101-O and HGM/ZL101-STA.

Annealed HGM/ZL101-O

• Initially strain was distributed relatively uniformly over surface.
• At approximately %3 strain, single localization band about 45° to compression direction formed.
• After peak, band rapidly crossed specimen.
• Stress curve showed sharp step-like drop.

HGM/ZL101-STA

• From %1,8 strain onward, multiple local localization regions appeared near lower surface.
• After peak at approximately %2,4 these regions developed into independent small shear bands.
• Instead of one band suddenly crossing specimen, distributed progressive damage formed.
• Stress curve showed slower sawtooth-like decline.

Researchers attribute this change to reaction not occurring to same extent in all microspheres, causing their load capacities to diverge. Microspheres with different degrees of weakening fracture at different load levels, so damage initiates in many small regions rather than concentrating in single plane.

Fracture surfaces

In annealed HGM/ZL101, tearing ridges indicating plastic deformation were observed on aluminum matrix. Broken microsphere edges were smoother and more continuous, while some shells carried multiple cracks together with matrix.

In STA condition thick reaction layer became crack-initiation region. Microspheres fractured at lower stresses and successive damage developed in neighboring spheres. At same time ductility of hardened matrix decreased; number of tearing ridges dropped and flat cleavage surfaces larger than 10 µm appeared in some regions.

How was energy absorption calculated?

Energy absorption per unit volume:

\[ W=\int_0^{\varepsilon_m}\sigma\,d\varepsilon \]

was calculated by equation. Here σ is compressive stress, ε strain and ε_m densification strain.

Specific energy absorption per mass:

\[ SEA=\frac{W}{\rho} \]

was obtained. ρ is composite density.

Specific energy absorption of HGM/ZL101-STA decreased approximately %18,4 and HGM/ZL104-STA approximately %19,9. In HGM/ZL102-STA, approximately %10,5 increase occurred.

Highest specific energy absorption was 63,89 kJ/kg for HGM/ZL102-STA. Annealed HGM/ZL104-O also showed high performance at 61,76 kJ/kg.

Why is ZL104 additionally important?

Although ZL104 matrix responded strongly to precipitation hardening, interfacial reaction increased substantially during long solution treatment at 550 °C. Therefore theoretical strengthening potential of matrix was not reflected in composite.

Researchers think optimized heat treatment such as lower temperature or shorter duration might exploit ZL104 matrix hardening while limiting microsphere reaction. This proposal was not experimentally tested within study and is presented as future-study possibility.

Strengths of study

• Three different Al-Si matrices compared using same microsphere and manufacturing method.
• At least five parallel samples reported for each composite.
• Microsphere fraction and density are close among groups.
• Reaction product studied not only by EDS but also TEM, HRTEM and SAED.
• Reaction-layer thickness and Al penetration distance quantitatively evaluated.
• Interfacial-reaction fraction quantified by image analysis.
• Mechanical properties of matrix alloys measured separately.
• DIC strain maps used together with compression curves.
• Damage mechanism supported by SEM fracture surfaces.
• Volumetric and specific energy absorption reported together.

Main limitations of study

• Study is a preprint not peer reviewed.
• 10-hour solution treatment at 550 °C was selected specifically to generate strong reaction and may not represent typical production conditions.
• Interfacial-reaction fraction calculated only from microspheres larger than 20 µm in diameter; median microsphere diameter is 13,6 µm.
• Two-dimensional cross-section image does not directly show three-dimensional reaction extent of whole microsphere shell.
• No standard deviation or total number of microspheres counted for reaction fraction.
• Most composite mechanical results lack error bars and statistical-significance analysis.
• DIC comparison performed only on HGM/ZL101.
• Study examined only quasi-static loading at 0,001 s⁻¹.
• Impact, blast and high-strain-rate energy absorption not measured.
• Fatigue, repeated compression and cyclic damage not examined.
• Samples taken from only specific billet location; axial homogeneity not mapped in detail.
• Individual crush strength of microsphere shells not directly measured after heat treatment.
• “Strongest” and “weakest” microsphere mechanisms not verified by direct single-sphere tests.
• Minority phases in reaction layer not definitively identified.
• Actual residual stress and interfacial bond strength after heat treatment not measured.
• Long-term temperature, humidity and corrosion resistance not studied.
• No testing on actual automotive, rail-system or protective-structure component.
• Production cost, mass producibility and recycling not evaluated.

Important reporting issues in source

• Conclusion states HGM/ZL101 peak stress decreased %6; Table 6 values 155,56 and 151,32 MPa correspond to approximately %2,7 decrease.
• Peak-stress reduction of about %5,9–6 applies to HGM/ZL104.
• Figure 6 caption labels samples (e) and (f) again as HGM/ZL101; experiment sequence and main text indicate they are HGM/ZL102.
• Figure 14 called “tensile fracture morphology”; main text states samples are fragments detached from shear band during compression.
• In Table 7 composite names changed from HGM/ZL to MZ40/ZL; they appear to represent same material groups.
• Expression “full interfacial reaction” is based on visible cross-sectional morphology and should not be interpreted as full volumetric transformation of whole sphere.
• Number of microspheres evaluated for interfacial-reaction amount and between-sample distribution not provided.
• Order of matching thermal-expansion-coefficient values to materials requires separate checking in source.

Which conclusions are supported?

• HGM/Al syntactic foam with density about 1,34–1,35 g/cm³ can be produced by vacuum-pressure infiltration.
• Long solution treatment at 550 °C markedly increased reaction between Al and SiO₂ in glass shell.
• Main component of reaction layer identified as metastable γ-Al₂O₃.
• Pores 30–90 nm in size formed in reaction layer.
• Fraction of microspheres showing full-reaction appearance decreased as matrix silicon content increased.
• Intense reaction reduced plateau stress and energy absorption of HGM/ZL101 and HGM/ZL104.
• High-silicon ZL102 matrix largely limited interfacial reaction.
• HGM/ZL102-STA reached peak stress 212,12 MPa and specific energy absorption 63,89 kJ/kg.
• Intense reaction associated with transition from one large shear band to many small shear bands.
• Plateau stress is more sensitive than peak stress to partial microsphere weakening.

Which conclusions are not proven?

• Does not prove %12 silicon is universal optimum for all glass-microsphere/aluminum systems.
• Does not show 550 °C for 10 hours is industrially optimal heat treatment.
• Does not prove HGM/ZL102 will be best composite under real impact and blast loads.
• Does not show quasi-static energy-absorption values remain same in high-speed collision.
• Does not show interfacial reaction can be completely prevented.
• Does not prove all phases in reaction layer are γ-Al₂O₃.
• Does not measure load-bearing capacity of each microsphere individually.
• Does not demonstrate fatigue and long-term durability of composites.
• Does not prove material is ready for direct use as automotive or defense component.
• Does not evaluate whether production is economical, scalable or environmentally superior.

Усул ва бозёфтҳои таҳқиқот

Summary of experimental method

Matrix and interface relationship

Main quantitative findings

Performance change before and after heat treatment

Separation of evidence levels

Key application implication

Heat-treatment selection cannot be based only on hardening of aluminum matrix. Chemical stability of microsphere shell must be evaluated simultaneously. Even if matrix hardens strongly, composite performance can decrease if reaction layer weakens microspheres.

Study data show higher matrix silicon can be an effective compositional tool to limit interfacial reaction. However effects of high silicon on ductility, machinability, casting behavior and other service properties were not evaluated within this study.

Ёддошти манбаъ ва усул

Full original title: Influence of Interfacial Reaction on Mechanical Properties and Energy Absorption Performance of HGM/Al Composites

Authors and correct order: Fangzhou Zhang, Yangwei Wang, Daiyuan Li, Jiawei Bao, Rui An, Changle Zhang

Co-first author or equal contribution: No co-first-authorship or equal-contribution statement in uploaded version.

Corresponding author: Jiawei Bao

Corresponding author email: baojiawei993@163.com

Author emails:

• Fangzhou Zhang: 15262996767@163.com
• Yangwei Wang: wangyangwei@bit.edu.cn
• Daiyuan Li: ldy010914@163.com
• Jiawei Bao: baojiawei993@163.com
• Rui An: bitanrui@163.com
• Changle Zhang: zhangchangle1209@163.com

Institutional affiliations:

• School of Materials Science and Engineering, Beijing Institute of Technology, Beijing 100081, China
• Tangshan Research Institute, Beijing Institute of Technology, Tangshan 063000, China
• Luoyang Ship Material Research Institute, Luoyang 471023, China

Author-institution matching:

• Fangzhou Zhang: Beijing Institute of Technology, School of Materials Science and Engineering
• Yangwei Wang: Beijing Institute of Technology, School of Materials Science and Engineering
• Daiyuan Li: Beijing Institute of Technology, School of Materials Science and Engineering
• Jiawei Bao: Beijing Institute of Technology, School of Materials Science and Engineering
• Rui An: Tangshan Research Institute, Beijing Institute of Technology
• Changle Zhang: Luoyang Ship Material Research Institute

ORCID: No ORCID information in uploaded version.

DOI:10.2139/ssrn.7195560

Journal: Not published in peer-reviewed journal.

Original journal publisher: None.

Publication platform: SSRN

Publication date: Appears as 28 July 2026.

Page count: 25

Source type: Preprint research study including aluminum-matrix syntactic-foam production, heat treatment, interface characterization, quasi-static compression and energy-absorption experiments

Peer-review status: Study has not undergone peer review. All pages of uploaded version contain “Preprint not peer reviewed” warning.

Official link:Official SSRN preprint page

Funding: Authors state research received no specific support from public, commercial or nonprofit funding organizations.

Conflict of interest: Authors declare no known financial interest or personal relationship that could affect study.

Data access: Data will be shared on request. No open data-repository link is provided for raw microscopy, DIC and mechanical-test data.

Author contributions:

• Fangzhou Zhang: Investigation, original draft, review and editing
• Yangwei Wang: Conceptualization, project administration, review and editing
• Daiyuan Li: Investigation
• Jiawei Bao: Supervision, review and editing
• Rui An: Supervision, review and editing
• Changle Zhang: Investigation and supervision

Ин шарҳи тоҷикӣ бо examining text of uploaded study, its nine equations, seven tables, fifteen main figures, figures repeated on pages 20–25, microsphere-preform image, alloy microstructures, tensile and compression curves, SEM-EDS maps, TEM-HRTEM-SAED analysis, reaction-fraction graph, DIC strain maps, fracture morphologies and energy-absorption results омода шудааст.

External sources were used only to verify bibliographic identity of SSRN record, title, DOI and publication timing. No experimental result from external source was added to scientific content of study.

Main limitations are lack of peer review, severe solution-treatment condition selected specifically to make reaction visible, reaction counting based only on large microspheres, limited reporting of statistical uncertainties, DIC analysis applied to only one alloy and all energy-absorption results derived from quasi-static compression tests.