Patent Application
20100297360
The present invention relates to a method and apparatus for densifying porous articles with a desirably high rate of production, particularly, but not necessarily only, with respect to the field of friction braking articles, such as aircraft brakes.
In the field of friction materials, it is generally known to use porous materials to manufacture friction members, such as friction brake disks.
The manufacture of such friction members generally begins with the construction of a porous preform. For example, in many friction brake applications, annular preforms are used.
The annular preforms can be constructed using several different known methods. For example, carbon fiber fabric plies can be needled together and annular preforms can be cut from the stacked material.
Also, near net shape preforms can be formed, for example, by weaving carbon fibers, or by braiding the carbon fiber into a desired shape. Certain carbon fiber fabrics are known having a weave that facilitates laying the fabric in a spiral form. In this context, “near-net” refers to forming structures having a form close to a desired shape of the final article, such as an annular brake disk.
Oxidized polyacrylonitride (“PAN”) fibers or pitch-based fibers are common examples of starting fibers used in this type of application. Subsequently, these fibers may be carbonized in a high temperature treatment step. In another conventional approach, the starting fibers are formed using a resin or pitch, and the resultant mass is later cured with a reactive gas, such as nitrogen gas. The thusly cured mass is then carbonized to obtain a semi-rigid preform.
In any event, it is desirable to further densify the resulting porous preform (especially, but not necessarily only, with a carbonaceous material) so as to obtain desired friction and mechanical properties.
Chemical vapor infiltration (“CVI”) is a widely used conventional technique in this regard for obtaining carbon/carbon composite materials. CVI uses a hydrocarbon-containing gas to infiltrate a porous preform. The CVI gas is then cracked under high temperatures so as to leave a carbon coating on the fiber structure of the preform.
Conventional CVI typically requires several hundred hours of processing in order to obtain a carbon/carbon (“C/C”) structure having a desired density and mechanical properties. By way of example, a typical conventional CVI process includes a first infiltration cycle performed, for example, over approximately 300-500 hours or more.
However, conventional CVI frequently causes rapid blockage of the surface porosity of the preform before interior portions of the preform are adequately densified. In order to “reopen” the surface porosity to permit further densification, an intermediate machining step becomes necessary. In general, this intermediate machining (using a known method, such as milling) removes surface layers of the preform having carbon-blocked pores to expose open pores of the preform so that the hydrocarbon gas can again infiltrate the preform structure. Taking into account that several hundred preforms are densified in a typical densification process, the intermediate machining step can add as much as 48 hours to the overall conventional CVI densification process.
Once the intermediate machining of the partially densified articles is completed, a second CVI process is performed to make use of the reopened surface porosity of the preforms. This second CVI process step can last, for example, another 300-500 hours or more. This generally completes the conventional densification process using CVI.
Another approach to densifying porous preforms uses a liquid instead of gaseous hydrocarbon precursor. This method of densification is sometimes referred to in the art as “film boiling” or “rapid densification.”
Film boiling densification generally involves immersing a porous preform in a liquid hydrocarbon so that the liquid substantially completely infiltrates the pores and interstices of the preform. Thereafter, the immersed preform is inductively heated by appropriately placed electrical elements, such as induction coils, to a temperature above the decomposition temperature of liquid hydrocarbon (typically 1000° C. or more). More particularly, the liquid hydrocarbon adjacent to the inductively heated preform structure dissociates into various gas phase species within the preform porosity. Further thermal decomposition of the gas phase species results in the formation of pyrolitic carbon on interior surfaces in the open regions of the porous material. The use of liquid precursors for densification is discussed in, for example, U.S. Pat. Nos. 4,472,454, 5,389,152, 5,397,595, 5,733,611, 5,547,717, 5,981,002, and 6,726,962. Each and every one of these documents is incorporated herein by reference in its entirety.
The concept of inductive heating in this field is generally known, including as described in the aforementioned references. However, heating a preform to high temperatures (at least 1000° C. and as much as 1400° C.) while it is literally immersed in highly volatile hydrocarbon liquids (such as, for example, cyclohexane) raises very important safety issues.
The present invention relates to a process of densifying porous substrates using a liquid precursor, in which the volume of “new” or “fresh” liquid precursor used is usefully reduced by maintaining the liquid precursor used for densification at a level of purity that is less than pure, but still chemically suitable for the densification process. In effect, the present invention uses an artificially “aged” (in terms of the impurities present therein) liquid precursor.
The present invention will be even more clearly understood with reference to the drawings appended hereto, of which:
The feature and details of the method and apparatus of the invention with the reference to the accompanying drawings are shown by way of illustration and not as limitations of the invention.
Solely by way of example and/or illustration, mention is made hereinbelow of porous preforms, such as preforms for manufacturing friction brake disks. It is expressly noted, however, that the present invention is more generally applicable to densifying other kinds of porous substrates in the manner described.
A highly schematic representation of a facility for performing densification using liquid precursors is illustrated in
The facility may optionally include a relatively smaller local storage tank 105 for keeping a relatively small quantity of new precursor liquid close to the processing equipment, if desired.
The piping system (including pumps and the like) used to interconnect various parts of the facility is conventional and may be of any construction and arrangement appropriate for the transport of liquid precursor being used, particularly, but not necessarily only, liquid hydrocarbons. The fluid transfer system is preferably but not necessarily computer-controlled. Commercially available computer-controlled systems (for example and without limitation, those commercially available from the company OPTO 22) can be used for monitoring and controlling this type of fluid transfer system, including loading of new liquid precursor from an outside supplier.
The liquid precursor is supplied to one or more reaction chambers (collectively indicated at 110) from the local precursor storage tank 105. Preferably, sufficient liquid precursor is provided to substantially immerse the one or more preforms being densified therein, as well as the induction heating coils associated therewith.
As mentioned before, the film boiling process creates gaseous species that in part causes the formation of pyrolitic carbon on interior surfaces of the preform porosity. Precursor vapor is captured to the extent possible and condensed at the conventional condenser unit 115 for possible recycling in the process. A commercial cooling tower 140 is available to maintain an adequate water temperature for the cooling of the condenser unit 115.
Effluent gas that is still left over is preferably conveyed to a thermal oxidizer 120 of a known configuration to burn off residual hydrocarbons in the effluent gas.
The power from the power supply 125 is transferred to the induction coils 25 by metal bus bars 30 constructed according to a given arrangement of elements in a facility and according to appropriate desired dimensional considerations. The bus bars may be made from copper, for example, and are preferably but not necessarily water-cooled by water cooling networks 50. (See
Each power supply 125 may have remote PID loop control capability and can be monitored and controlled from a computer control terminal. Power density control, voltage control, current control, frequency control, and/or temperature control of the densification process by known methods is also within the scope of the contemplated arrangement.
In an example of loading and unloading coil/reactor chamber 110, a top cover panel 15 is provided with a conventional locking mechanism for sealing the chamber 110. Each reactor chamber 110 (as many as are provided) is provided with a common liquid precursor supply line connection 20, and a common exhaust line 10 operably connected to the condenser 115 and thermal oxidizer 120.
Each reactor chamber 110 can be desirably filled, drained, and monitored from the computer control system. The exhaust liquid precursor vapor from the densification process is condensed and fed back to the reactor chambers 110, whereas residual effluent gas is then taken to the thermal oxidizer 120 and burned.
Because volatile liquid hydrocarbons are a particular example of the liquid precursor used in the present invention, it is desirable (but not obligatory) to provide a nitrogen (N2) gas supply system (not shown) to, for example, flush out the piping systems and generally fill voids in the system with an inert gas (instead of oxygen-containing air) so as to decrease the risk of combustion. In a particular example, empty spaces in both the remote and local liquid precursor storage tanks are maintained at a slight, continuously supplied, overpressure of nitrogen (or other conventionally known inert) gas so as to prevent potentially dangerous accumulations of volatile vapors. Hydrocarbon species mixed with the exhausted nitrogen gas are sent to the thermal oxidizer 120 so that the hydrocarbons can be burned off before the gas is exhausted to the exterior.
Also, because the system uses a “wet” process, it is useful to provide a drying oven 130 in the system to dry off densified preforms following densification. The exhaust from such a drying oven 130 is preferably also connected to the thermal oxidizer 120 in order to process heavy and light aromatics entrained in the resultant effluent gas. With respect to safety considerations, it is useful to use an oven structure that is structurally resistant to failure in the event of an explosion therein, given the presence of volatile gases in the oven during drying. The drying process can be, for example, computer controlled in order to simplify process control.
Table 1 provides the analysis of aged C6H12 and the major impurities produced from the densification process. In this study, eight densification cycles in three reactors were performed with carbonized preforms using a dedicated power curve. In particular, an objective was to reduce the usage of new or “fresh” liquid precursor (which can be relatively costly) to replenish the precursor being used during production.
Sixteen aged C6H12 samples from eight consecutive runs were collected and sent for analysis. The goal was to correlate individual contaminant peak value to the aged C6H12. Samples were collected before and after each densification cycle and identified as A and B.
The C6H12% concentration at the end of 8th cycle was still above 94%. The major resulting impurity peaks were Benzene, Naphthalene, Toluene, Styrene, Cyclohexene, Cyclopentadiene, and Indene. Six out of the seven measured impurities followed a steady, generally increasing trend throughout the eight cycles. A gas chromatography (GC) analysis made from the aging study indicates all major contaminants and Cyclohexane concentrations are very predictable. All impurity peaks are consistently repeated from one cycle to the next cycle.
Simulation of aged liquid precursors by doping a liquid precursor with various impurities is also contemplated. In this manner, it may be possible to beneficially extend the usable life of a liquid precursor and/or reduce the need for highly pure (and, therefore, relatively expensive) liquid precursors. In other words, on the one hand, a “minimum” level of precursor purity can be identified, such that the precursor can be used at purity levels above that minimum (before replacement with highly pure precursor). On the other hand, a lower acceptable level of precursor purity can be identified such that the contemplated densification process would use that lower purity precursor on an ongoing basis. Either way, the usage of high purity precursor liquid is desirably reduced.
Aged liquid precursors were simulated by doping a liquid precursor with various impurities mentioned above in order to approximate the gradual “deterioration” of a highly pure liquid precursor after several process cycles. A model was developed to predict Cyclohexane concentrations at various cycles. Extrapolation was performed based on experimental data from the eight consecutive runs previously described.
Nine chemicals were purchased from Alfa Aesar to dope the cyclohexane precursor. Table 2 provides the information about the chemicals used for the doping. The purity of the purchased chemicals ranged from 99.6% to 80%.
Table 2 provides the quantity used for the individual chemicals to obtain a 90% Cyclohexane concentration. A 90% Cyclohexane concentration was used for the initial mix in order to simulate the model, although cyclopentadiene was not used in the mix for safety reasons because this gel-like material is extremely flammable in air. The effect on the test is believed however to be minimal. Six consecutive densification cycles were performed using the initial 90% mix. Samples of the liquid precursor were collected before and after each cycle and sent for GC analysis.
A total of twelve aged precursors from six consecutive cycles were sent to a laboratory for gas chromatography analysis. Aged precursor samples were collected before and after the densification cycle from the 500 gallon tank. Table 3 provides the GC analysis for the simulated 90% mix precursor. It was observed in Table 3 that the reported concentrations from all chemicals closely followed the mass % used in the 90% mixture with the exception of Cyclopentadiene, which was not actually used due to its high flammability in air. It is however believed that the omission of cyclopentadiene from the actual samples has a minimal overall effect on the test.
Table 4 illustrates a progression of the major contaminants from the six consecutive cycles, or twelve precursor samples. The purity generally started at 90.3% and ended at 88.4% after six densification cycles. In most of the cases, contaminants reached steady state or slowly declined with the exception of Benzene. (See, for example, Cyclopentadiene, Cyclohexene, Toluene, Ethylbenzene, Phenylacetylene, Styrene, Indene, Naphthalene, Methylnaphthalene, Acenaphthalene, and Fluorene in Table 4) This indicates that continuation of chemical breakdown from Cyclohexane (C6H12) to Benzene (C6H6) is most likely to continue at lower C6H12 concentration levels. (See, for example, the benzene concentration in
In practice, several methods can be used to manage the precursor. Some of them are described hereafter.
Several consecutive cycles could be run with the same liquid precursor bath, followed by periodically replacing the whole storage tank with new cyclohexane. In that manner, the cyclohexane concentration will vary between fresh cyclohexane concentration and a low concentration depending on the number of consecutive cycles run.
Another way to manage the precursor is to periodically replace only a part of liquid precursor in the storage tank. This approach reduces the variations in concentrations and therefore permits better control of the pyrocarbon deposited from this precursor (microstructure, thermal-mechanical or friction properties). In this regard,
According to these two methods, the storage tank is never totally drained. Multiple runs were performed according to the approaches corresponding with
The experimental data confirm the calculation is an accurate method to forecast the cyclohexane concentration versus the runs number (see
Moreover, the analyses show that the concentration of every chemical component remains substantially steady after several runs.
Although the present invention has been described above with reference to certain particular examples for the purpose of illustrating and explaining the invention, it is to be understood that the invention is not limited solely by reference to the specific details of those examples. More specifically, a person skilled in the art will readily appreciate that modifications and developments can be made in the preferred embodiments without departing from the scope of the invention as defined in the accompanying claims.
This application claims priority from U.S. Provisional Application No. 60/821,596, filed on Aug. 7, 2006, the entire content of which is incorporated herein by reference.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP07/58193 | 8/7/2007 | WO | 00 | 4/10/2009 |
| Number | Date | Country | |
|---|---|---|---|
| 60821596 | Aug 2006 | US |
