Engineering topics

Behrendt, D. R.

Publications and source records attributed to Behrendt, D. R..

At least 19 records

NASA NTRS1995

Reactive melt infiltration of silicon-molybdenum alloys into microporous carbon preforms

Investigations on the reactive melt infiltration of silicon-1.7 and 3.2 at.% molybdenum alloys into microporous carbon preforms have been carried out by modeling, differential thermal analysis (DTA), and melt infiltration experiments. These results indicate that the pore volume fraction of the carbon preform is a very important parameter in determining the final composition of the reaction-formed silicon carbide and the secondary phases. Various undesirable melt infiltration results, e.g. choking-off, specimen cracking, silicon veins, and lake formation, and their correlation with inadequate preform properties are presented. The liquid silicon-carbon reaction exotherm temperatures are influenced by the pore and carbon particle size of the preform and the compositions of infiltrants. Room temperature flexural strength and fracture toughness of materials made by the silicon-3.2 at.% molybdenum alloy infiltration of medium pore size preforms are also discussed.

Singh, M.↗

NASA NTRS1994

Reactive Melt Infiltration of Silicon-Niobium Alloys in Microporous Carbons

Studies of the reactive melt infiltration of silicon-niobium alloys in microporous carbon preforms prepared by the pyrolysis of a polymer precursor have been carried out using modeling, Differential Thermal Analysis (DTA), and melt infiltration. Mercury porosimetry results indicate a very narrow pore size distribution with virtually all the porosity within the carbon preforms open to infiltrants. The morphology and amount of the residual phases (niobium disilicide and silicon) in the infiltrated material can be tailored according to requirements by careful control of the properties (pore size and pore volume) of the porous carbon preforms and alloy composition. The average room temperature four-point flexural strength of a reaction-formed silicon carbide material (made by the infiltration of medium pore size carbon preform with Si - 5 at. % Nb alloy) is 290 +/- 40 MPa (42 +/- 6 ksi) and the fracture toughness is 3.7 +/- 0.3 MPa square root of m. The flexural strength decreases at high temperatures due to relaxation of residual thermal stresses and the presence of free silicon in the material.

Singh, M.↗

NASA NTRS1994

Microstructure and Mechanical Properties of Reaction-Formed Silicon Carbide (RFSC) Ceramics

The microstructure and mechanical properties of reaction-formed silicon carbide (RFSC) ceramics fabricated by silicon infiltration of porous carbon preforms are discussed. The morphological characterization of the carbon preforms indicates a very narrow pore size distribution. Measurements of the preform density by physical methods and by mercury porosimetry agree very well and indicate that virtually all of the porosity in the preforms is open to infiltrating liquids. The average room temperature flexural strength of the RFSC material with approximately 8 at.% free silicon is 369 +/- 28 MPa (53.5 +/- 4 ksi). The Weibull strength distribution data give a characteristic strength value of 381 MPa (55 ksi) and a Weibull modulus of 14.3. The residual silicon content is lower and the strengths are superior to those of most commercially available reaction-bonded silicon carbide materials.

Singh, M.↗

NASA NTRS1993

Theoretical considerations for Reaction-Formed Silicon Carbide (RFSC) formation by molten silicon infiltration into slurry-derived preforms

For reaction-formed silicon carbide (RFSC) ceramics produced by silicon melt infiltration of porous carbon preforms, equations are developed to relate the amount of residual silicon to the initial carbon density. Also, for a slurry derived preform containing both carbon and silicon powder, equations are derived which relate the amount of residual silicon in the RFSC to the relative density of the carbon in the preform and to the amount of silicon powder added to the slurry. For a porous carbon preform that does not have enough porosity to prevent choking-off of the silicon infiltration, these results show that complete silicon infiltration can occur by adding silicon powder to the slurry mixture used to produce these preforms.

Behrendt, D. R.↗

NASA NTRS1992

Studies on the reactive melt infiltration of silicon and silicon-molybdenum alloys in porous carbon

Investigations on the reactive melt infiltration of silicon and silicon-1.7 and 3.2 at percent molybdenum alloys into porous carbon preforms have been carried out by process modeling, differential thermal analysis (DTA) and melt infiltration experiments. These results indicate that the initial pore volume fraction of the porous carbon preform is a critical parameter in determining the final composition of the raction-formed silicon carbide and other residual phases. The pore size of the carbon preform is very detrimental to the exotherm temperatures due to liquid silicon-carbon reactions encountered during the reactive melt infiltration process. A possible mechanism for the liquid silicon-porous (glassy) carbon reaction has been proposed. The composition and microstructure of the reaction-formed silicon carbide has been discussed in terms of carbon preform microstructures, infiltration materials, and temperatures.

Singh, M.↗

NASA NTRS1984

Application of Iosipescu specimen geometry to determination of shear strength in unidirectional composites

The Iosipescu (1967) test method, in conjunction with a finite element analysis of anisotropic, unidirectionally reinforced composite specimens, can be used to derive composite shear strength values despite the small size of the samples used. When such a specimen is loaded parallel to the fiber direction, fracture occurs by shear within a narrow region near the center of the specimen; this is consistent with the maximum shear area determined by finite element analysis. The resulting fracture surface, as observed, is characteristic of shear failure, occurring either within the matrix or at the fiber-matrix interface. Problems of compressive and mixed mode failure which accrue to short beam shear specimens are by the present method precluded.

Hurwitz, F. I.↗

NASA NTRS1980

Calculation of residual principal stresses in CVD boron on carbon filaments

A three-dimensional finite element model of the chemical vapor deposition of boron on a carbon substrate (B/C) is developed. The model includes an expansion of the boron after deposition due to atomic rearrangement and includes creep of the boron and carbon. Curves are presented showing the variation of the principal residual stresses and the filament elongation with the parameters defining deposition strain and creep. The calculated results are compared with experimental axial residual stress and elongation measurements made on B/C filaments. For good agreement between calculated and experimental results, the deposited boron must continue to expand after deposition, and the build up of residual stresses must be limited by significant boron and carbon creep rates.

Behrendt, D. R.↗

NASA NTRS1980

Calculation of residual principal stresses in CVD boron on carbon filaments

A three-dimensional finite element model of the chemical vapor deposition (CVD) of boron on a carbon substrate (B/C) is developed. The model includes an expansion of the boron after deposition due to atomic rearrangement and includes creep of the boron and carbon. Curves are presented to show how the principal residual stresses and the filament elongation vary as the parameters defining deposition strain and creep are varied. The calculated results are compared with experimental axial residual stress and elongation measurements made on B/C filaments. This comparison requires that for good agreement between calculated and experimental results, the deposited boron must continue to expand after deposition, and that the build-up of residual stresses is limited by significant boron and carbon creep rates.

Behrendt, D. R.↗

NASA NTRS1979

Some properties of an advanced boron fiber

An advanced coreless boron fiber exhibits tensile strengths which are more than twice that of the normal CVD B/W fibers. The coreless fiber is made by the chemical removal of the tungsten boride core exposed by splitting the as-grown fiber. The easily splittable fiber is made by the chemical vapor deposition of boron on a larger than usual tungsten substrate. It is expected that the ease of splitting is related to residual stresses in these fibers. Measurements of the axial residual stresses in both the normal and the splittable fibers are presented and the results compared. Differences in these stresses are discussed in connection with the ease of splitting in the splittable fibers.

Behrendt, D. R.↗

NASA NTRS1979

Some properties of an advanced boron fiber

An advanced coreless boron fiber has exhibited tensile strengths which are more than twice that of the normal CVD B/W fibers. The coreless fiber is made by the chemical removal of the tungsten boride core exposed by splitting the as-grown fiber. The easily splittable fiber is made by the chemical vapor deposition of boron on a larger than usual tungsten substrate. It is expected that the ease of splitting is related to residual stresses in these fibers. Measurements of the axial residual stresses in both the normal and the splittable fibers are presented and the results compared. Differences in these stresses are discussed in connection with the ease of splitting in the splittable fibers.

Behrendt, D. R.↗

NASA NTRS1978

Ultra-high-strength boron fibers

Boron-on-tungsten fibers with tensile strength and strain-to-failure values increased by fifty percent over commercial grades are produced by controlled chemical-etching process. Improved fibers have potential applications as lightweight composites in ground vehicles, spacecraft, and rotors for energy storage.

Behrendt, D. R.↗

NASA NTRS1978

Axial residual stresses in boron fibers

A method of measuring axial residual stresses in boron fibers is presented. With this method, the axial residual stress distribution as a function of radius is determined from the fiber surface to the core including the average residual stress in the core. Such measurements on boron on tungsten (B/W) fibers show that the residual stresses for 102, 142, 203, and 366 micron diam fibers are similar, being compressive at the surface and changing monotonically to a region of tensile stress within the boron. At approximately 25% of the original radius, the stress reaches a maximum tensile stress of about 860 MN sq m and then decreases to a compressive stress near the tungsten boride core. Data are presented for 203-micron diam B/W fibers that show annealing above 900 C reduces the residual stresses. A comparison between 102-micron diam B/W and boron on carbon (B/C) show that the residual stresses are similar in the outer regions of the fibers, but that large differences near and in the core are observed. Fracture of boron fibers is discussed.

Behrendt, D. R.↗

NASA NTRS1978

Axial residual stresses in boron fibers

The axial residual stress distribution as a function of radius was determined from the fiber surface to the core including the average residual stress in the core. Such measurements on boron on tungsten (B/W) fibers show that the residual stresses for 102, 142, 203, and 366 micron diameter fibers were similar, being compressive at the surface and changing monotonically to a region of tensile within the boron. At approximately 25 percent of the original radius, the stress reaches a maximum tensile stress of about 860 mn/sq.m and then decreases to a compressive stress near the tungsten boride core. Data were presented for 203 micron diameter B/W fibers that show annealing above 900 C reduces the residual stresses. A comparison between 102 micron diameter B/W and boron on carbon (b/C) shows that the residual stresses were similar in the outer regions of the fibers, but that large differences near and in the core were observed. The effects of these residual stresses on the fracture of boron fibers were discussed.

Behrendt, D. R.↗

NASA NTRS1977

Residual stresses in boron/tungsten and boron/carbon fibers

Longitudinal residual stress distribution is determined for 102-micron diam B/W and B/C fibers. The 102-micron diam B/W fibers are deposited on a 12.7-micron diam tungsten wire resistively heated in a BCl3-H2 reactor. The 102-micron diam B/C fibers are made by deposition of boron on a pyrolytic graphite-coated carbon fiber. The longitudinal residual stress distribution is calculated from measurements of the change in length of the fiber produced by removal of the surface through electropolishing. It is found that for both types of fibers, the residual stress vary from a compressive stress at the surface to a tensile stress in the boron near the core. Closer to the core and in the core, significant differences in the residual stresses are observed for the B/W and B/C fibers.

Behrendt, D. R.↗

NASA NTRS1977

Residual stresses in boron/tungsten and boron/carbon fibers

By measuring the change in fracture stress of 203 micrometer diameter fibers of boron on tungsten (B/W) as a function of fiber diameter as reduced by chemical etching, it is shown that the flaws which limit B/W fiber strength are located at the surface and in the tungsten boride core. After etching to a diameter of 188 micrometers m virtually all fiber fractures were caused by core flaws, the average strength being 4.50 GN/sq m. If both the surface and core flaws are removed, the fracture strength, limited by flaws in the boron itself, is approximately 6.89 GN/sq m. This was measured on B/W fibers which were split longitudinally and had their cores removed by chemical etching. The longitudinal residual stress distribution was determined for 102 micrometer diameter B/W and B/C fibers.

Behrendt, D. R.↗

NASA NTRS1976

Longitudinal residual stresses in boron fibers

A technique is proposed for measuring the longitudinal residual stress distribution in commercial CVD (Chemical Vapor Deposition) boron on tungsten fibers of 102, 142, and 203 microns in diameter. The experimental apparatus is so designed that continuous measurements are made of the length changes of a boron fiber specimen as the surface of the fiber is removed by electropolishing. The effects of surface removal on core residual stress and core-initiated fracture are discussed. The three sizes of boron fibers investigated show similar residual stress distributions, i.e., compressive at the surface, tensile near the core, and for the 102-micron fiber compressive again in the core. It is shown that an increase in UTS is due to the increase in the compressive stress at the core produced by fiber contraction during surface removal. An expression is derived for calculating the longitudinal residual stress at a given radius for an as-received fiber of a certain radius from measurements of the axial strain produced by removal of the surface by electropolishing.

Behrendt, D. R.↗

NASA NTRS1976

Longitudinal residual stresses in boron fibers

A method of measuring the longitudinal residual stress distribution in boron fibers is presented. The residual stresses in commercial CVD boron on tungsten fibers of 102, 142, and 203 microns (4, 5.6, and 8 mil) diameters were determined. Results for the three sizes show a compressive stress at the surface 800 to -1400 MN/sq m 120 to -200 ksi), changing monotonically to a region of tensile stress within the boron. At approximately 25 percent of the original radius, the stress reaches a maximum tensile 600 to 1000 MN/sq m(90 to 150 ksi) and then decreases to compressive near the tungsten boride core. The core itself is under a compressive stress of approximately -1300 MN/sq m (-190 ksi). The effects of surface removal on core residual stress and core-initiated fracture are discussed.

Behrendt, D. R.↗

NASA NTRS1973

Effect of diffusional creep on particle morphology of polycrystalline alloys strengthened by second phase particles

Diffusional creep in a polycrystalline alloy containing second-phase particles can disrupt the particle morphology. For alloys which depend on the particle distribution for strength, changes in the particle morphology can affect the mechanical properties. Recent observations of diffusional creep in alloys containing soluble particles (gamma-prime strengthened Ni base alloys) and inert particles have been reexamined in light of the basic mechanisms of diffusional creep, and a generalized model of this effect is proposed. The model indicates that diffusional creep will generally result in particle-free regions in the vicinity of grain boundaries serving as net vacancy sources. The factors which control the changes in second-phase morphology have been identified, and methods of reducing the effects of diffusional creep are suggested.

Wittenberger, J. D.↗