EP1446366A2 - Oxide ceramic fibre reinforced material and the use thereof - Google Patents

Abstract

The invention relates to oxide ceramic fibre reinforced material which is used according to the invention in energy conversion installations. The aim of the invention is to create a fundamentally new concept which meets requirements which have been, until now, totally incompatible, i.e. improve brittle fracture behaviour in order to secure thermal shock stability and damage tolerance in the temperature range of < 1000 DEG C and provide mechanical high temperature stability to ensure the long term form stability (creep resistance) of components. This is achieved by means of oxide ceramic fibre reinforced material which contains fibres having a monocrystalline structure in certain areas and an average coherence length of >/= 150 mu m, also having a mesoscopically resistant fibre/matrix interface at high temperatures along the length of the axis of the fibre. Said fibre/matrix interface can be subjected to temperatures above room temperature accompanied by propagations of microfissures exhibiting delaminations which are microspcoically localised in their length along the axis of the fibre and which remain at a value smaller than the average fibre coherence length being limited to a maximum of 200 mu m. The invention also aims at using the inventive fibre reinforced material for thermal and/or long-term stressed components and/or installations.

Description

Oxide ceramic fiber composites and their use

FIELD OF APPLICATION OF THE INVENTION The invention relates to oxide-ceramic fiber composite materials, as are used, for example, according to the invention in plants for energy conversion.

PRIOR ART In the interest of increasing energy yields significantly in the future, increased efforts to provide ceramic materials for plants for energy generation at temperatures> 1400 ° C. can be observed. Consistent assessments assume that for load-bearing components under such extreme conditions

- both a sufficient damage tolerance (i.e. thermal shock stability and high room temperature fracture toughness of the actually brittle ceramics) and the required creep resistance of the components can only be achieved by means of a fiber composite design,

- The problem of thermodynamic high-temperature stability in oxidizing atmospheres can only be mastered permanently with per se oxidation-stable oxide ceramics.

Similarly differentiated claims regarding damage tolerance, temperature and corrosion stability must meet high-performance brake discs. At present, the mechanical and thermophysical requirements are best met by composites of long carbon fibers with reaction-bound SiC matrix, the corrosion resistance of which is limited, however, precisely because the corrosion progresses along the non-oxide fibers into the interior of the panes. It was observed that the use of short carbon fibers improves the corrosion stability, but on the other hand limits the level of mechanical properties (R. Gadow, pp. 15-29 in: Ceram. Eng. Sci. Proc. Vol. 21/3, The Am. Ceram. Soc, Westerville / OH, 2000). Purely oxidic composites could therefore also offer an advantageous material alternative here. In view of the complex mechanical requirements, it is evident from the above assessments for all such applications that the mechanical properties of the fiber / matrix bond play a decisive role in the realization of such oxidic composite materials. However, it is also obvious why it was previously impossible to develop a concept that meets the requirements for thermal shock resistance and creep resistance alike:

- The requirement for transferring the creep resistance of a (preferably multidimensional) fiber structure to the component as a whole leads, at least with a view to long-term use of> 10,000 hours, to the obvious requirement for a firm fiber / matrix bond (IW Donald et al., J. Mater See 11 (1976) 5: 949-972); A.G. Evans et al., Pp. 929-955 in Creep and Fracture of Engineering Materials, London, 1987). One could hope to achieve a certain macroscopic dimensional stability of the parts even without a firm fiber / matrix bond by a high creep resistance of the fiber arrangement by a multi-dimensional fiber design, but under the conditions of flowing atmospheres and high pressure gradients in turbines are very complex stresses exist (also with regard to erosion), which composite materials with generally weak interfaces at temperatures around 1400 ° C can hardly withstand for a long time.

- On the other hand, the demand for ceramics resistant to thermal shock, even for repeated loads, currently appears to be met exclusively by a composite design that consumes (brittle) fracture energy (energy dissipation) after stretching and partial breakage of fibers by means of considerable shear deformation along the fiber / Matrix interfaces realized (crack deflection, "pull-out" or "debonding" effects, made possible by weak fiber / matrix bonding).

In implementing this understanding, a large number of proposals have been made, all of which include the formation of weak fiber / matrix interfaces without exception. This takes place partly through fiber coating, partly through an appropriate design of the matrix.

Pejryd et al. (EP 639 165 A1, US 5,567,518) describe “a ceramic composite, particularly for use at temperatures above 1400 ° C.” with a structure made of oxide Fiber / oxide matrix and a choice of material that is determined by claim and possibly greater coating thickness of the fibers> 2 μm, so that weak interfaces ("weak bond Nable to debonding") are expressly achieved. As in most of the proposals known from the literature However, here, too, the real usability for the declared application is not disclosed: the "detection" of energy consumption during breakage (in the form of non-linear stress-strain effects) is only carried out in the examples by tests at room temperature, which leads to real thermal shock behavior and even more so Creep stability does not allow any conclusions to be drawn in the operations mentioned as the target beyond 1400 ° C.

Saruhan-Brings et al. (EP 890 559 A1) describe a “process for coating oxidic fiber materials for the production of failure-tolerant, high-temperature-resistant, oxidation-resistant composite materials”. As an example, a la aluminate coating on polycrystalline Al ₂ O ₃ fiber with a mullite matrix is used No high-temperature tests or their results are disclosed either. The same defect is inherent in the patent of Lange et al. (US Pat. No. 5,856,252), where within the framework of "Damage tolerant ceramic matrix composites by a precursor infiltration" a claim is made of an "all oxide ceramic" composite "composed of uncoated Al ₂ θ ₃ fibers and porous mullite matrix without disclosing information on the high-temperature behavior; main content of US 5,856,252 is the description of oxidic composites with damage-tolerant behavior, achieved by" delaminations ... with extensive regions of cracking normal to the rupture plane ".

A similar approach can be found with Dariol in a "Procede d ^' elaboration d'un materiau d'inerphase, materiau obtenu, procede de traitment d ^' un fiber ceramique de renfort avec ce materiau et materiau thermostructural incluant de telles fibers" ( FR 27 78 655 A1), where the formation of a micro-porous fiber / matrix interface area takes place by adding a porosity former (eg carbon which is at least partially oxidized). Lundberg et al. Come to a microstructurally similar result (EP 946 458 A1 ), which for oxide-fiber / oxide-matrix composites, intended specifically for use in oxidizing atmospheres> 1400 ° C, generate microporous fiber / matrix interfaces by immersing the fibers in slip with carbon and Zr0 ₂ . The proposed solutions overlook the fact that the creep rates of the oxides in question increase by about 4 orders of magnitude between 1300 ° C and 1700 ° C (many non-oxide ceramics are more creep-resistant but susceptible to oxidation), so that sufficient mechanical long-term stability ( > 10,000 h) with polycrystalline fibers in polycrystalline matrices will be quite impossible to achieve.

The invention therefore relies exclusively on composites made of oxide-ceramic materials with at least partially single-crystalline fibers, whose mean coherence lengths of the single-crystalline areas are at least 150 μm, preferably> 400 μm, particularly preferably> 1 mm. Surprisingly, the above-mentioned publications contain only an embodiment in EP 639 165 A1 which meets this requirement in that thick single-crystal sapphire fibers (sapphire, USA; fiber thickness> 100 μm) are coated with Zr0 ₂ and in an Al ₂ θ3 Matrix are embedded. However, EP 639 165 also lacks any information regarding the behavior under temperature stress that is actually of interest as the target variable. The concept of weak interfaces is also adhered to when using or in-situ generation of single-crystalline fibers, with the consequence, among other things, that the fibers tend to lower the strength rather than improve it at low temperatures (AA Kolchin et al., Composite Sei. Technol. 61 (2001) 8, 1079-1082).

What the previously known solution proposals have in common is that even in the case of a target definition for use at temperatures> 1400 ° C., the development of properties is oriented exclusively to the thermal shock stress that is relevant only at low temperatures, and that even for this purpose the material properties are not disclosed. All known developments aim at the goal of microstructural design for "failure tolerant" behavior ("damage tolerant"), which is realized by weak fiber / matrix interfaces. Incomprehensibly, the achievement of this goal is in no way verified by appropriate thermal shock investigations in technically relevant temperature ranges; rather, all of the exemplary embodiments mentioned are limited to pure room temperature fracture tests. If this “weak” fiber / matrix bond is generated by weakening the fiber coatings with suitable materials or by artificially created micro-pores, it is also overlooked that such micro-pores are unstable in the intended temperature range, at least during long-term use and that the weakening effect of a fiber coating at higher temperatures, such as at 1400 ° C, is very different from that in the exclusively examined room temperature range. And of course, "weak" fiber / matrix interfaces cannot even at> 1400 ° C meet the dominant requirement for creep stability.

Thus, the known proposals for oxide-ceramic fiber composites for use at temperatures> 1400 ° C neither for

(1) the thermal shock stresses associated with heating / cooling are still for

(2) the problem of creep deformation, which is serious at such high temperatures, is a practical solution or even just a suggestion.

DESCRIPTION OF THE INVENTION The present invention therefore provides a solution to the two previously incompatible requirements for (1) improved brittle fracture behavior to ensure thermal shock stability and damage tolerance in the temperature range <1000 ° C. and (2) mechanical high-temperature stability Ensuring the long-term dimensional stability (creep resistance) of components made of such oxide-ceramic composite materials is a fundamentally new concept.

The object is achieved by the invention specified in the claims. Further training is the subject of the subclaims.

The oxide ceramic fiber composite materials according to the invention, which contain fibers with an at least partially monocrystalline structure (average length of the monocrystalline "coherent" areas of the fiber at least 150 μm, preferably> 400 μm, particularly preferably> 1 mm), are characterized in that the fiber / matrix -Interfaces at high temperature stress along the fiber axis one in mesoscopic length ranges over the middle fiber- Show coherence length tight bond, whereby this bond with stress types above room temperature, which are associated with the propagation of macro cracks (fracture, thermal shock), only microscopic, locally limited to areas smaller than the mentioned coherence length and at most 200 μm delaminations (debonding) and thus prevents a quasi-macroscopic detachment of the fibers upon break (pull-out) as well as extensive creeping of the interface in the high temperature range.

According to the invention, the term a fiber / matrix interface that is mesoscopically fixed along the fiber axis over the fiber coherence length means that under thermomechanical loading there is no continuous tearing of the matrix from the fiber (“pull-out” in the axial direction) there is a crack dimension exceeding the mean coherence length of the fiber and thereby at most 200 μm. Thus, while "delaminations ... with extensive ... cracking" (US 5,856,252) are avoided here, microscopic areas that are smaller than the mean coherence length of the Fibers and at most 200 μm, localized detachments. This microscopically local delamination cannot, however, lead to longer pull-out effects of the fibers in the fracture surface due to the binding of the interfaces mesoscopically in the axial direction of the fibers; the pull-out lengths that can be observed on fracture areas remain small compared to the stated minimum coherence length of 150 μm and in no case exceed an average value of 200 μm.

This illustration also shows that a blanket qualification of the fiber / matrix interfaces as "firm" or "weak" makes little sense due to the different temperature ranges and without specification of the length dimensions considered. The fiber / matrix bond referred to above as mesoscopically fixed could, on the one hand, be addressed as relatively weak in this temperature range in view of its brittle material behavior (<1000 ° C) in the temperature range compared to the possibly firmer character of the same interface in the high temperature range, which enables creep resistance above 1000 ° C; on the other hand, however, this fiber / matrix interface is undoubtedly relatively “solid” in the range <1000 ° C. compared to other interfaces, which, as described according to the prior art, break at the separation lengths of the variety of the fiber diameter. Surprisingly, the energy dissipation associated with microscopic local fiber detachment of composites according to the invention is sufficient to significantly influence the brittle fracture processes, as they determine strength and thermal shock resistance in the temperature range <1000 ° C. The hitherto insurmountable contrast between the requirements for mastering the mechanical behavior in this low temperature range and for shape stability (creep resistance) and structure cohesion (for example with a view to erosion stability) at very high temperatures is achieved with the design according to the invention for purely oxidic composite materials solved for the first time.

The choice of materials and the production technology are of secondary importance compared to the basic design of the composite concept according to the invention. The use of materials known as creep-resistant, such as mullite or Y-Al garnet for matrix and fibers, generally offers advantages. Additional measures to check the structure under long-term high-temperature stresses can also make sense (such as the use of special doping, duplex or generally multi-phase matrix materials in order to prevent or limit grain growth). The respective selection of the base material for matrix and fibers and / or additives will, however, always be determined by the respective intended use and is not subject to any restrictions in the context of the composite design according to the invention described here. The material selection of a possible fiber coating can also be carried out within a wide framework within the scope of the design concept disclosed here, because the material of the fiber / matrix interface only defines the first prerequisites for fulfilling the feature of the new composite design: the actual development of an inventive one Mesoscopically solid fiber / matrix interface, which allows types of stress above room temperature, which are associated with the propagation of macro cracks (fracture, thermal shock), microscopically-localized delaminations (debonding), only takes place after coordination of this material selection with others for known process steps (such as the generation of an optimized degree of sintering of the matrix by varying the sintering temperature). If coated fibers are used, the thickness of the coating must be kept <2 μm in order to implement the mesoscopically firm bond according to the invention.

Of course, it is also irrelevant whether the composite is manufactured using a prefabricated single-crystalline oxide ceramic fiber with the specified coherence lengths or from polycrystalline fibers, fabrics or other fiber-like precursors that are only in situ in the course of the sintering of the composite (e.g. produced by impregnation) the single-crystalline state described according to the invention have been converted.

The present invention can be used in a special way for components or systems that are subject to high thermal loads and / or long-term high loads. They can also advantageously be used for components which are used in particular under changing thermal and / or long-term loads. These composites can advantageously be used in particular for components and / or systems for energy conversion or for high-performance brake discs.

BEST WAY OF IMPLEMENTING THE INVENTION The invention is explained in more detail using an exemplary embodiment.

Sapphire fibers with a diameter of approx. 160 μm (Advanced Crystal Products, USA) were coated with 0.15 - 1.2 μm thick SrO _* 6AI ₂ 0 ₃ layers using the sol / gel process and then together with a high-purity corundum powder (> 99.99% Al ₂ 0 ₃ ) and 0.2 μm average grain size (TM-DAR, Boehringer Ingelheim Chemicals, Japan) in a hot press to flat plates of about 5 mm thick. By comparison, the properties of fiber-free samples were also examined. The influence of the fiber thickness is not shown here, but thin fibers are generally advantageous to use.

A relatively large mutual fiber spacing of approx. 1 - 1.5 mm was determined for the experiments, the dimensions of the samples were 3.6 x 6.8 x 60 mm ³ . The fibers were aligned parallel to the longitudinal axis of the bending fracture bars and thus oriented perpendicular to the (macroscopic) fracture surfaces that were created. The grain size of the matrix of the dense oxide-ceramic fiber composites produced was changed by different hot pressing temperatures; Such an adjustment of the grain size of the matrix is important for the mechanical stability at very high temperature: 1330 ° C / 2 h - 0.6 μm,

1550 ° C / 2 h - 6.0 μm,

1800 ° C / 2 h - 15.5 μm.

All mechanical tests in the range from room temperature to 1400 ° C were carried out in a 3-point bend with a loading rate of 0.5 mm / min; In preliminary tests between 0.1 and 1 mm / min, no influence of the loading rate, the packing density of the fiber or an additional hot isostatic post-compression of the composites was found.

The results of the examination of the fracture surfaces and the mechanical data confirm the character of the inventive oxide-ceramic composite materials.

As an example, the figures show fracture surfaces of composites whose sapphire fibers were provided with Sr and aluminate layers with a thickness of 0.6 and 1.2 μm:

Fig. 1 No longer fiber detachment from the matrix when broken at room temperature, the fibers (coating thickness 1.2 μm) break in almost the same plane as the surrounding matrix (minimum pull-out lengths <100 μm).

Fig.2a / b The greater magnification of those generated at room temperature

Fracture surface shows: the strength-increasing effect of the coated fibers (coating thickness 1.2 μm) in the temperature range of brittle material behavior (see table below) goes hand in hand with (i) a local, microscopic juxtaposition of (i) areas of the fiber that are still firmly sintered after breaking / Matrix interface (which prevents longer fiber detachment from the matrix) and (ü) microscopically localized crack formation (energy dissipative effects) flow of the brittle fracture with effect, for example in the sense of increased strength or improved thermal shock resistance).

Fig. 3 Even in the upper temperature range of brittle material behavior there is no significant pull-out effect (here example 900 ° C; coating thickness 1.2 μm).

Fig. 4a / b fracture surface 1200 ° C (similar at 1400 ° C): Even at very high temperatures, fiber coatings that lead to oxide-ceramic composite materials according to the invention (in the example with a 0.6 thick Sr-aluminate coating) are mesoscopic solid fiber / matrix bond is realized, which survives the macroscopic break without longer fiber detachment from the matrix (similar result also with a 1.2 μm thick fiber coating).

Mechanical tests were used to investigate the behavior of the oxide ceramic fiber composite materials under the influence of temperature.

The desired positive effect of the fibers with coating and with the sintering matched to this is shown in the following table with regard to the brittle fracture behavior (which is addressed, for example, when subjected to thermal shock in the temperature range between 20 and 1000 ° C); the data refer to structures with an average matrix grain size of 0.6 μm:

At very high temperatures, on the other hand, brittle fracture behavior and thermal shock resistance play a subordinate role. What is important here is above all a firm fiber / matrix bond as a prerequisite for sufficient long-term dimensional stability of the component. While the microscopic strength of the interfaces and the absence of pull-out mechanisms are shown in Fig. 4a / b for 1200 - 1400 ° C, the table below shows the constancy of the macroscopic for the example of composite materials with a matrix grain size of 6 μm Strength between 1200 and at least 1400 ° C regardless of the thickness of the fiber coating in the area examined (similar data also for matrix with 14.5 μm grain size):

Claims

claims 1. Oxide-ceramic fiber composite materials, containing fibers of a regionally single-crystalline structure with an average coherence length of> 150 μm, with a fiber / matrix interface which is mesoscopically fixed at high temperature along the fiber axis and which, under stresses above room temperature, which are associated with the propagation of macro cracks, Has delaminations, which are microscopically locally limited in their extension along the fiber axis to values smaller than the average fiber coherence length and thereby limited to a maximum of 200 μm. 2. oxide ceramic fiber composite materials according to claim 1, containing coated fibers with a thickness of the interface material of less than 2 microns. 3. Use of oxide ceramic fiber composite materials according to claim 1 for thermally and / or long-term highly stressed components and / or systems. 4. Use according to claim 3, for components and / or systems under changing thermal and / or long-term loads. 5. Use according to claim 3, for components and / or systems for energy conversion or for high-performance brake discs.