EP0290687B1

EP0290687B1 - Hybrid centrifuge rotor

Info

Publication number: EP0290687B1
Application number: EP19870304160
Authority: EP
Inventors: Alireza Piramoon
Original assignee: Beckman Instruments Inc
Current assignee: Beckman Coulter Inc
Priority date: 1987-05-11
Filing date: 1987-05-11
Publication date: 1993-02-10
Anticipated expiration: 2007-05-11
Also published as: DE3784163T2; DE3784163D1; EP0290687A1

Description

The present invention provides a hybrid centrifuge rotor suitable for use in ultra high speed centrifuges (ultracentrifuges).
Ultracentrifuge rotors encounter extreme stress as increasingly higher speeds are necessary for the application of such rotors for new and increasingly complicated scientific experimentation. High capacity and speed rotors are generally larger fixed angle and vertical tube rotors having high kinetic energies. High stresses are created in these high speed, high energy rotors, which makes material selection the primary design criteria. For this reason titanium is the isotropic material most often chosen for high speed, large volume rotor design.
Conventional high kinetic energy rotors, capable of high speed operation, are typically manufactured of titanium, a isotropic material with a density of approximately 4.43 grammes per cm³ (.16 pounds per inch³). Titanium is a stiff metal which is difficult to machine, due to its density and, in addition, is very expensive. Additionally, since increasing speeds require increasing strength, the kinetic energy that a titanium rotor would endure will increase proportionally to its density and the square of its rotational speed and it should be noted that the strength of the metal material cannot be improved significantly by increasing the rotor size, diameter or thickness.
Heretofore, centrifuge rotors have been made, for low speed application, from laminated fibre-resin-material shells as a means of reliance upon materials other than metal. An example of a centrifuge drum shell made from a laminated material of fibres and resins, having a metal bottom and top forming circular metal plates, is disclosed in United States Patent No. 3 993 243 to Dietzel. The metal plates have flanges, arranged to secure said plates to the centrifuge drum shell by means of fastening screws. The drum shell is cross-sectionally configured in numerous layers (reference may be made to Figure 2 of said Patent) including axially extending wires 114 and the use of a wide mesh backing strainer 16. This design, which includes fasteners, is suitable for lower speed applications, but would not sustain the stress induced by ultra high speeds.
Likewise, layered material rotors made from steel covered resin plastic materials are taught by United States Patent No. 3 997 106 to Baram. Embodiments shown in Figures 3 and 4 of the Baram Patent show the use of piano wire 24 to surround a perforated steel belt covering an inner body made from moulded plastic resin material. This design may be useful in centrifuge rotors used by low speed liquid separators but it would not withstand the stress of high drive speeds.
United States Patent No. 4 160 521 (Lindgren) shows the manufacture of a fibre-reinforced rotor for use in centrifugal separators, impellers, generators and flywheels. The rotor body shell is composed of an outer layer made from a fibre wound filament which is wound about an intermediate sleeve made of a plastic resin or aluminium. The outer layer forms a fibre-laminate sleeve member which prestresses the internal intermediate sleeve. The outer layer is affixed to the inner intermediate sleeve by a pressing-on wind of the outer layer to the inner sleeve. This pressing-on wind prestresses the inner sleeve which decreases the diameter of the inner sleeve slightly. As a result of mechanically pre-stressing the inner sleeve, this Patent teaches that the reinforced rotor can operate at speeds where a metal inner sleeve only rotor would deform or break for reaching higher speeds. The Patent also makes note of a temperature "shrinkage fit" wherein an internal metallic member is kept at a low temperature while an external fibre laminate sleeve is applied to the metal member. This prior art document asserts that only a limited temperature reduction is obtainable and that high surface pressure between the fibre-reinforced sleeve and the metal member cannot be achieved when the inner member returns to room temperature. The Patent also teaches against temperature "shrinkage fit" designs and it fails to recognise the possibility of significant shrinkage using very cold temperatures.
Thus, the 4 160 521 Patent teaches away from "shrinkage fit" between differing material layers of a rotor body and towards the concept of pre-stressing a composite sleeve over an inner body, since it is put forward that shrinkage fit would not be achievable.
United States Patent No. 4 468 269 (Carey) and assigned to the present applicant and United States Patent No. 3 913 828 (corresponding to French Patent Publication No. 2 151 074) both disclose an ultracentrifuge rotor design comprising a plurality of "nested" rings of filament wound layers surrounding the cylindrical wall of a metal body rotor. The rings are "nested" together by a thin coat of epoxy between each ring. The rings are manufactured by winding on mandrels having different diameters, the diameters being chosen so that each ring section will be manufactured within a uniform range of ten to fifteen layers of filaments. By fabricating the rings in thin sections and then nesting the rings together, it is possible to achieve a rotor structure where the density of filaments on the outside diameter of the ring is as high as that on the inside diameter of the ring. The nested telescoped polymer rings are epoxy bonded to a coating wall 13 (Figure 1 of the 269 Patent) and pressed onto the wall 13 using small axial loading pressures. The final structure is temperature cured and it relies upon temperature cured epoxy layers to reinforce the central wall 13 structure, the wall being selected from titanium, aluminium or heat-treated steel. Thus, none of the hybrid rotor designs are directed to a single shell, simplified reinforcement structure. Furthermore, the prior documents do not address a fundamental problem, which is that both static and dynamic loads must be analysed and considered when one wishes to design a multi-material rotor for ultracentrifuge application.
The Lindgren Patent and the Carey Patent addressed the static load problem, but not the dynamic loading results of their respective designs. Static loading is the stress between the core body and the outer material ring when the rotor is at rest. These static loads are important to hold together any hybrid rotor assembly at rest; but, when rotation of the rotor occurs, a tangential outward stress arises. The core bodies of the designs shown in the aforesaid Patents expand at a slower rate than the outer ring, so that the outer ring may not remain stress loaded on the core body during dynamic rotor operation. The core body expands at a slower rate than the outer ring during conditions of dynamic loading, therefore, the ring may slip away from the core body and they no longer maintain a pre-stress load on the body.
Heretofore, aluminium and titanium have been the materials selected to comprise an isotropic material rotor. Aluminium alone may be used for high speed applications, only where the stresses created by a rotor filled with fluid specimen are considerably less than the stress failures point of the aluminium body. Where high capacity rotors require greater speed, the rotors have been designed using titanium, a dense material capable of withstanding higher stresses and capacities of sample fluid than that of a comparable size high capacity aluminium rotor. Heretofore, only titanium has been able to withstand the dynamic stresses of high speed, high capacity centrifugation and any lower density design, such as an aluminium and composite material ring design must be able to withstand the same dynamic loads that the titanium rotor encounters.
The problem which the present applications seeks to overcome, is the provision of a hybrid rotor design which takes into account the need for addressing both the static and dynamic load characteristics of such a multi-material design. While an outer fibre-reinforced shell may pre-stress the core during static conditions, this inward stress causes undue and unwanted non-uniform stress to arise within the core body during rotational conditions of dynamic loads.
In accordance with the present invention, there is provided a hybrid centrifuge rotor, comprising : a rotor core body,; and a reinforcement ring surrounding said core body and made from an anisotropic material; wherein the reinforcement ring is prestressed by the core body by an interference fit between the core body and the ring, characterised in that said core body comprises a body structure of an axially symmetric block of isotropic material in which a plurality of cavities are provided for receiving sample tubes; wherein said core body exerts an outward pressure in excess of 3.45 x 10⁸ N/m² against the interior surface of said ring.
French Patent Publication No. 2 151 074 discloses the use of thermal shrinkage for fitting a reinforcing ring onto a ball type rotor in order to achieve a zero-clearance fit of the ring on the rotor as compared with the present invention when cryogenic cooling is used to fit a reinforcing ring onto a solid-body rotor such that the rotor core exerts an extremely high outward pressure on the ring. The hybrid centrifuge rotor body disclosed herein is made from two primary portions, a rotor core body and a reinforcement ring. The rotor core body is an isotropic material such as aluminium metal. The reinforcement ring which surrounds the rotor core body is made from an anisotropic material such as a graphite fiber and an epoxy resin filament wound material. (As used in this application, the term "anisotropic" shall mean a material having properties, such as bulk modulus, strength, and stiffness, oriented in a particular direction). The Resin in the reinforcement ring may have thermoplastic or thermoset properties. The fiber of the filament-wound ring may be a material such as glass or the organic resin fibrous material Kevlar^R, made by Du Pont, which are both useful substitutes for graphite.
The isotropic core body of the hybrid centrifuge rotor prestresses the filament-wound anisotropic reinforcement ring. The core body is designed to originally be nominally of greater diameter than the reinforcement ring. The core body is then uniformly temperature shrunk in a cryogenic environment and fit for positioning within the inner diameter of the reinforcement ring upon final assembly of the rotor, telescoped within the reinforcement ring, so that the core body pre-stresses the reinforcement ring. The core body pre-stresses the ring as the body returns to ambient temperatures after removal from the cryogenic environment. The pre-stressing arises as the core body attempts to restore its original dimensions. The cryogenic environment may be achieved by bathing the isotropic core body in a liquid nitrogen bath. The core body may be uniformly shrunk to a maximum degree calculated according to the equation " $δ = αLΔT$
, where δ is the shrinkage; α is the coefficient of thermal expansion; L is the diameter of the metal core body; and, ΔT is the temperature difference between the ambient room temperature and the cryogenic environment. The core body is then bathed in a cryogenic environment for an optimum time period to achieve a shrinkage of the core body within the computed range defined by the above equation. The core body then re-expands at ambient temperature after being positioned and telescoped within the reinforcement ring at a uniform rate.
The anisotropic reinforcement ring which surrounds the core body may be filament wound in a conventional circumferential manner where the orientation of the fiber wind is zero degrees from a horizontal plane parallel to the base of the rotor core body. In an alternative embodiment, the filament-wound ring may be wound in a criss-cross fashion at an orientation angle other than the horizontal plane, in order to further enhance and strengthen the reinforcement ring.
By pre-stressing the reinforcement ring, due to the interference fit of the core body and ring, a strong static load is achieved. The outward expanding core body pre-stresses the reinforcement ring by achieving a high static load of the core body against the ring. During rotation and conditions of dynamic loading, the high modulus present in the reinforcement ring tend to contain the core body as centrifugal forces tend to build within the core body. In this manner, a hybrid rotor design is achieved which compensates for both static and dynamic loading.

Brief Description of the Drawings

Figure 1 is an elevational cross-sectional view of a hybrid centrifuge rotor of this invention taken along the diameter of a vertical tube rotor body manufactured according to this invention.
Figure 2 is a cross-sectional elevational view of a fixed-angle rotor manufactured as a hybrid centrifuge rotor, taken along the diameter of said rotor.

Detailed Description of the Preferred Embodiment

The hybrid rotor of this invention comprises two major components, namely, an isotropic rotor core and an anisotropic reinforcement ring. With reference to Figure 1, a cross-sectional view of the preferred embodiment of a hybrid centrifuge rotor is shown generally. The isotropic rotor core 12 is preferably made from aluminum, which has a density of approximately 3.04 g/cm³ (0.11 lb/in³). This metal rotor core body 12 has a density substantially less than the conventional isotropic unibody rotor made from titanium, which has a density of 4.43 g/cm³ (0.16 lb/in³).
Circumferentially surrounding the rotor core body is an anisotropic reinforcement ring 14. The reinforced ring 14 is a cylindrical graphite fiber and epoxy resin filament-wound ring. The density of the composite material filament-wound reinforcement ring 14 is approximately 1.67 g/cm³ (0.06 lb/in³). Thus, for a vertical tube rotor as shown in Figure 1, the density of the entire rotor assembly of the preferred embodiment is substantially lower than a conventional titanium rotor. The cut-away view of the hybrid rotor of Figure 1 shows a vertical test tube cavity 16 formed within the rotor core body 12. Figure 1 shows the hybrid rotor assembly having a drive shaft hole 18 for receiving the drive shaft (not shown).
The design of a large capacity centrifuge rotor body which may be driven at high speeds must meet various criteria. Among these criteria is a consideration of the density or weight of the rotor. The lower the density, the lower the kinetic energy that will be generated at a given rated speed in the event of a mishap. A lighter weight rotor, such as disclosed in this application, will require substantially less acceleration and deceleration time, and the choice of aluminum as a core body represents a substantial reduction in manufacturing cost over the conventional titanium rotor. Heretofore, even with the disadvantages of cost and a difficulty of machining, titanium was chosen as the conventional design for a high density, high energy centrifuge rotor because only titanium could withstand very high speeds for large capacity applications that were desirable for the application and performance of various research biological separations of organic materials. Although a titanium rotor offers strength, its higher density generates higher kinetic energy as the rotor speed increases, which requires a corresponding increase in the strength and weight of the centrifuge barrier ring needed to contain this higher energy rotor in the event of a rotor failure. The hybrid rotor disclosed herein is of lighter weight and density than the conventional titanium centrifuge rotor, and thus a lower kinetic energy than a conventional titanium rotor requiring no change in centrifuge barrier ring design, yet capable of withstanding high speeds without undue stress to the hybrid rotor core body.
The hybrid rotor of this invention accomplishes the goal of providing a lower density centrifuge rotor assembly by the nature in which the rotor core body and reinforcement ring interface. Heretofore, conventional interface in prior multi-layer material centrifuge rotor body designs have relied on prestressing of the rotor body core. Heretofore, the outer ring was designed to prestress the rotor core body such that a strong interface would be accomplished. However, prestressing the rotor core body by pressure-fit methods in a conventional manner may lead to faults, cracks, and strains within the rotor core body 12 which, at the ultra high speeds attainable by a centrifuge motor drive, might lead to destruction of the rotor body and a potential mishap.
The disclosure of this invention chooses to prestress the reinforcement ring. This is accomplished by an interference fit resulting in an interface outward stress in excess of 3.45 x 10⁸ N/m² (50,000 psi) static loading between the reinforcement ring 14 and the rotor core body 12. (As used in this application, the term "interference fit" means that where two parts are mated to form an assembly, like the core body 12 and the ring 14, one of the parts (in this case the core body) is forced to occupy a space smaller than its normal size, so that internal stresses within that first part to exert a uniform stress against the other second part wich the first part abuts). The rotor core body 12 is designed to initially have a diameter which is nominally larger than the inner diameter of the reinforcement ring 14. The rotor core body 12 is then placed into a cryogenic environment where it is uniformly temperature shrunk so that the diameter of the rotor core body 12 will be just under the inner diameter of the reinforcement ring 14, so that the reinforcement ring 14 surrounds the core body 12 and the rotor core body 12 is telescoped within the reinforcement ring 14. The rotor may, alternatively, be assembled by placing the core body 12 within the inner diameter of the reinforcement ring 14. The rotor core body 12, in the preferred embodiment, is placed in a liquid nitrogen bath at a temperature of -179°C (-290°F) for a period of .25 hours. By bathing the rotor core body 12 in a liquid nitrogen bath, uniformity of shrinkage is achieved. The maximum amount of shrinkage possible may be calculated according to the formula $δ = α · LΔT$

:
where,

δ =: shrinkage (measured in inches);
α =: the coefficient of thermal expansion; (measured in inches per degree Fahrenheit);
L =: the diameter of the metal core measured in inches; and,
ΔT =: the temperature difference between the temperature of the cryogenic environment liquid nitrogen bath (approximately -179°C (-290°F) and an ambient temperature of approximately +25°C (+77°F).

After this maximum shrinkage is calculated, an experiment plot of shrinkage vs. time is undertaken to discover the optimum length of time that the core body 12 should be bathed in liquid nitrogen to achieve a desirable shrinkage within the calculated range.
After the rotor core body 12 is bathed in the liquid nitrogen cryogenic environment for an optimum period of time, the reinforcement ring may be assembled to surround the rotor core body. Along the interface 22 between the rotor core body 12 and the reinforcement ring 14 there is an interference fit. This interference fit between the rotor core and the reinforcement ring is accomplished by thermally shrink fitting the rotor core body 12 to a small enough diameter that the reinforcement ring 14 will surround the core body rather than mechanically forcing the rotor core body 12 to fit inside the reinforcement ring 14. The interference fit is achieved by taking advantage of the higher coefficient of thermal expansion of the metal core, when compared with the coefficient of thermal expansion of the reinforcement ring. Once the cryogenically treated core body is placed into a telescoped position within the ring, the ring is rapidly cooled, initially, by the core body due to the residual thermal effect which the extremely cold core body imparts on the ambient temperature ring 14; but, the ring expands slower than the core body due to its lower thermal coefficient of expansion. In other words, the cryogenic environment more rapidly shrinks the rotor core body 12 than any residual thermal effect the assembly procedure might have on the reinforcement ring. Thus, upon assembly, as the rotor core body 12 returns to ambient room temperature, it presses outward and prestresses the reinforcement ring 14, without undue stress on the rotor core body 12. This is so because the rotor core body 12 expands outward uniformly so any stress on the core body 12 is distributed throughout this body so that it will be less likely for cracks and faults to arise.
The reinforcement ring 14 is manufactured as a filament-wound composite of a graphite fiber and epoxy resin. In the preferred embodiment, the graphite fiber used to form the ring 14 is wound at a zero degree angle to the horizontal plane of the base of the rotor body 12. This form of construction induces hoop stress within the reinforcement ring which balances and opposes the outward stress of the rotor core body 12. The core body 12 and ring 14 are held tightly together due to their interface 22 by means of an interference fit. The reinforcement ring 14, due to its uniform windings, is an anisotropic material which orients stresses which may arise when the centrifuge assembly is rotating in a preferred direction, utilizing the unidirectional stresses which arise in the reinforcement ring 14 to hold the aluminum core body 12 together at speeds in excess of that which an aluminum body of unitary construction could possibly sustain. By having the reinforcement ring 14 be one continuous cylindrical filament-wound ring, the invention is able to take advantage of the relative difference between the coefficient of thermal expansion of the metal rotor core body 12 and the reinforcement ring 14, to reduce stresses arising in the rotor core. The reinforcement ring 14 acts to reinforce the strength of the rotor core body 12 without placing substantial additional stress on the rotor core body, since it is the core body which pushes outward against the inner diameter of the reinforcement ring 14 such that, along the interface 22 between the rotor core body 12 and the reinforcement ring 14, stress is displaced out to the reinforcement ring and away from the rotor core body.
Alternative materials may be used to construct the reinforcement ring 14. Instead of graphite, one may select a fiber for the filament-wound ring made from glass or the organic resin Kevlar®, made by Du Pont. In place of the epoxy resin, one may use any conventionally acceptable thermoplastic or thermoset resin.
Figure 2 shows a fixed angle rotor that may be assembled as a hybrid centrifuge rotor according to the method of manufacture for assembling the vertical tube rotor of Figure 1. The fixed angle rotor of Figure 2 has a central tapered rotor core body 20 surrounded by a frustoconically shaped reinforcement ring 24. The frustoconical reinforcement ring 24 may be manufactured according to the criteria heretofore discussed for making the reinforcement ring 14. Such a fixed angle design, providing test tube inserts 26 within the rotor core body 20 is possible, because when the tapered rotor core body 20 is bathed in liquid nitrogen, its shrinkage along its diameter is linear along each of a plurality of horizontal planes; i.e., the relationship between diameter shrinkage and the coefficient of thermal expansion of the tapered rotor core body 20 is linear and uniform along any horizontal plane cutting through the rotor body 20. This might not be the case if a conventional pressure or mechanical fit were attempted to assembly tapered rotor core body 20 into the frustoconical reinforcement ring 24.
In manufacturing the assembled centrifuge rotor of a hybrid design, it is important to recognize that a distinctive feature of the design is the interference fit. An experimentally derived plot of time vs. amount of shrinkage in a cryogenic environment indicated empirically that .25 hours was the optimum time for the liquid nitrogen bath to surround the metal rotor core body 12 or 20. In the preferred embodiment, a bath duration beyond .25 hours produces no significant further shrinkage, as the metal core body 12 has reached equilibrium with the nitrogen bath beyond a 15 minute period; however, using the shrinkage equation and the empirically derived plot of shrinkage against time is useful in deriving the optimum cryogenic bathing conditions when the shrinkage criteria differs from the preferred embodiment. After removing the rotor body core 12 from the liquid nitrogen bath, three to four hours are required after the reinforcement ring 14 is placed around the rotor core body 12 for the full effect of the interference fit to take place. Upon reaching the ambient temperatures, the core body 12, in attempting to restore its original ambient diameter exerts in excess of 3.45 x 10⁸ N/m² (50,000 psi) outward pressure against the interface 22 with the inner diameter of the reinforcement ring 14.
The interference fit, resulting in an outward directed uniform stress by the core body 12 against the reinforcement ring 14, accommodates the design criteria for static and dynamic loads. While a hybrid rotor, designed to have an isotropic core body surrounded by an antisotropic reinforcement ring, is of greater static load strength than an all-aluminum rotor, it is the interference fit which provides a design which is also of superior dynamic load characteristics. By having the core body pre-stressing the ring, rather than the conventional approach of having the outer ring stress the core body, a load characteristic may arise which is effective in handling dynamic tangential centrifugal forces which arise during ultracentrifugation. The hoop stress present in the reinforcement ring of this invention, which counteracts the outward static load of the rotor core, also may contain the dynamic tengential forces so that the ring remains affixed to the core body during rotation. The interference fit, by placing primary static and dynamic loads on the ring rather than the core, as well as less load on the rotor drive bearings, accounts for the ability for the hybrid rotor of this invention to sustain rotor speeds equivalent or in excess of the speeds achievable by conventional isotropic titanium rotors.
A fully assembled large capacity hybrid centrifuge rotor, manufactured in accordance with the teachings of this disclosure, will allow an aluminum core rotor to safely reach an upper limit equivalent to or exceeding a conventional titanium, rotor as opposed to a conventional large capacity aluminum unibody construction rotor having a lower maximum speed. The hybrid design, by prestressing the reinforcement ring 14, allows the reinforcement ring 14 to remain secure against the surface of the rotor body core 12 as higher centrifugal speeds are attained by the hybrid body rotor. The reinforcement ring 14 acts to safely hold the rotor core body 12 in place at speeds above the conventional upper limit of an aluminum body rotor. Because the aluminum core body stresses are within design criteria, and since aluminum is easier to machine, a core body having more capacity for a vertical tube rotor, like Fig. 1, can be made. This capacity is greater than the conventional eight cavity titanium rotor.

Claims

A hybrid centrifuge rotor, comprising:
a rotor core body (12); and
a reinforcement ring (14) surrounding said core body (12) and made from an anisotropic material; wherein the reinforcement ring (14) is prestressed by the core body (12) by an interference fit between the core body (12) and the ring (14);
characterised in that said core body (12) comprises a body structure of an axially symmetric block of isotropic material in which a plurality of cavities (16) are provided for receiving sample tubes; wherein said core body (12) exerts an outward pressure in excess of 3.45 x 10⁸ N/m² against the interior surface of said ring (14).
A centrifuge rotor according to claim 1, characterised in that the rotor core body is a light-weight aluminum material.
A centrifuge rotor according to claim 1 or claim 2, characterised in that the reinforcement ring is a cylindrical graphite fibre and epoxy resin filament-wound ring.
A centrifuge rotor according to claim 3, characterised in that the resin has thermoplastic properties.
A centrifuge rotor according to claim 3, characterised in that the resin has thermoset properties.
A centrifuge rotor according to claim 3, characterised in that the fibre of the filament-wound ring is an organic resin material.
A centrifuge rotor according to any of claims 1 to 6, characterised in that said anisotropic reinforcement ring surrounding said core body is filament wound circumferentially at an orientation angle from the horizontal plane of the rotor core body equal to zero degrees.
A centrifuge rotor according to any of claims 1 to 6, characterised in that the fibre is wound in a criss-cross fashion at an orientation angle from the horizontal plane other than 0 degrees.