Cylindrical roller bearing fatigue life analysis method based on multi-factor coupling
Technical Field
The invention relates to a bearing fatigue life analysis method, in particular to a cylindrical roller bearing fatigue life analysis method based on multi-factor coupling.
Background
The fatigue life analysis of the cylindrical thin-wall roller bearing in the modern high-end equipment generally comprises rolling bodies, outer rings, inner rings and other parts, and generally works under high-speed, heavy-load and high-Wen Keke working conditions such as aerospace, wind power generation, industrial gearboxes and the like, and particularly in the fields, the coupling effects of lubrication, heat and flexible deformation of the thin-wall ferrule under complex and severe working conditions must be considered.
Currently, there are many basic theories for analyzing bearing fatigue life. In the bearing fatigue life analysis theory, the L-P fatigue life analysis method provides important development for bearing fatigue life analysis and is widely applied. However, the L-P theory has certain limitations, such as the inability to cope with the problem of infinite long life. Considering the limitations of the L-P theory, researchers have proposed an I-H theoretical model based on the L-P theory that can take into account stress limit values. Based on the I-H theory, a plurality of documents are provided for researching an analysis method of the fatigue life of the bearing, for example, a plurality of scholars research the influence rule of friction effect on the bearing life, and the scholars also provide a fatigue life analysis method based on the damage accumulation theory.
However, under complex and severe working conditions, the rheological property of the lubricating oil in the bearing, heat generation and deformation of the bearing and flexible deformation of the thin-wall ferrule of the bearing are mutually coupled, and all the current analysis methods cannot analyze the influence of the analysis methods on the fatigue life of the bearing based on the coupling property of the lubricating-heat-flexible deformation. In order to more reasonably analyze the fatigue life of a bearing, a bearing fatigue life analysis method capable of comprehensively considering complex coupling relations among lubrication, heat and flexible deformation is needed.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for analyzing the fatigue life of a cylindrical roller bearing by considering lubrication-heat-flexible deformation coupling.
In order to achieve the aim, the technical scheme of the invention is that a cylindrical roller bearing fatigue life analysis method based on multi-factor coupling,
S1, establishing a bearing heat-flow-solid coupling dynamic model, which comprises the following specific steps:
s101, establishing a ferrule flexible body model;
s102, establishing a bearing lubrication model considering nonlinear viscoelastic characteristics;
s103, establishing a bearing thermal network model taking the temperature of lubricating oil into consideration;
s104, establishing a bearing dynamics model based on interaction relations among the elements of the bearing;
S105, realizing the coupling of the four models in S101-S104 by a parameter coupling mode, and establishing a bearing heat-flow-solid coupling dynamics model;
s2, establishing a contact subsurface stress calculation model;
S3, calculating the fatigue life of the bearing based on the I-H theory.
Further, in the step S101, the ferrule flexible body model is built by using a curved beam unit based on a finite element method.
Further, in the step S102, the constitutive model of the nonlinear viscoelastic fluid property is expressed as follows:
Where τ is the shear stress and, (U s is the sliding speed between the rolling element and the ferrule, h is the central oil film thickness), G is the shear modulus, τ 0 is the yield shear stress, η is the lubricating oil viscosity,
The calculation formula of the center oil film thickness h is as follows:
h=Ry3.05U0.69g0.56W-0.1
where U is a dimensionless speed parameter, u=η 0u/(EeqRy);
W is a dimensionless load, w=q/(E eqRy l);
g is a dimensionless material parameter, g=λe eq (in the above expression, Q is a contact load, η 0 is a reference viscosity, u is a rolling speed, E eq is an equivalent elastic modulus, l is a rolling body length, λ is a viscosity-pressure coefficient of lubricating oil, and R y is a rolling direction equivalent radius).
Further, in the step S102, in view of the influence of the pressure p and the temperature T on the shear modulus G, the yield shear stress τ 0, and the lubricating oil viscosity η, the correction calculation formulas of the shear modulus G, the yield shear stress τ 0, and the lubricating oil viscosity η are as follows:
In the above formula eta 0, Reference viscosity, reference shear modulus and reference yield shear stress, respectively, alpha i、βj、γk、αG、βG、ατ and beta τ are specific constants, T and T 0 are current temperature and reference temperature, respectively, and p is current pressure.
Further, according to step S102, a friction traction curve is obtained, which specifically includes the following steps:
s1021, the constitutive model of the nonlinear viscoelastic fluid characteristic adopts a Runge-Kutta method to obtain the shearing stress tau of the contact surface of the rolling body and the ferrule;
s1022, integrating the shearing stress tau on the contact area to obtain the friction traction force;
S1023, comparing the friction traction with the contact load Q to obtain a friction traction coefficient;
S1024, friction traction curves under different T and U s are obtained by changing the values of the temperature T and the sliding speed U s.
Further, in the step S1024, a BP neural network is used to perform function fitting on the friction traction curve so as to directly use the friction traction curve in the dynamics model.
Further, in the step S103, a calculation formula for obtaining the thermal deformation of each element of the bearing based on the temperature T of each element of the bearing is as follows:
ut=ΓL(T-Tα)
in the above formula, Γ is the thermal expansion coefficient, T α is the ambient temperature, and L is the characteristic length of each element of the bearing.
Further, in the step S103, a temperature node is set in each element of the bearing, and the following calculation formula is listed based on the law of conservation of energy:
node 1 (T 1-T10)/R1cv10+(T1-T2)/R2cc1 =0)
Node 2 (T 2-T1)/R1cc2+(T2-T3)/R3cc2 =0)
Node 3:(T3-T2)/R2cc3+(T3-T4)/R4cc3+(T3-T0)/R0cv3=Q1
Node 4 (T 4-T5)/R5cc4+(T4-T3)/R4cc3 =0)
Node 5:(T5-T4)/R5cc4+(T5-T6)/R6cc5+(T5-T0)/R0cv5=Q2
Node 6 (T 6-T5)/R6cc5+(T6-T7)/R7cc6 =0)
Node 7:(T7-T6)/R7cc6+(T7-T8)/R8cc7+(T7-T0)/R7cv0=Q3
Node 8 (T 8-T7)/R8cc7+(T8-T9)/R9cc8 =0)
Node 9 (T 9-T8)/R9cc8+(T9-T10)/R10cv9 =0)
In the above, T 1 is an outer ring temperature node, T 2 is an outer raceway temperature node, T 3 is an outer raceway lubricant temperature node, T 4 is a rolling element-outer raceway outer side node, T 5 is a rolling element temperature node, T 6 is a rolling element-inner raceway inner side node, T 7 is an inner raceway lubricant temperature node, T 8 is an inner raceway temperature node, T 9 is an inner ring temperature node, T 10 is an air temperature node, T 0 is a lubricant temperature node, Q 1、Q2、Q3 is a friction heat generation rate at a rolling element-outer raceway contact position, a friction heat generation rate between a rolling element and lubricating oil in a bearing cavity, and a friction heat generation rate at a rolling element-inner raceway contact position, R is a thermal resistance value (in the subscript, cv represents heat convection, cc represents heat conduction, and Arabic represents a temperature node).
Further, in the step S104, a GUPTA bearing model is used to build a dynamic model of the bearing.
Further, the step S105 establishes a bearing heat-fluid-solid coupling dynamics model according to the following steps:
s1051, initializing the temperature, the position and the speed of each element of the bearing;
S1052, calculating the current friction traction coefficient based on the bearing lubrication model in the step S102;
S1053, calculating the geometric approach amount and the contact load and the friction heat generation rate between the elements of the bearing based on the GUPTA dynamic model in the step S104;
S1054, substituting the friction heat generation rate obtained in the step S1053 into the bearing thermal network model in the step S103 to update the temperature of each element of the bearing, and obtaining the thermal deformation of each element of the bearing;
S1055, updating the geometric dimensions of each element of the bearing based on the elastic deformation of each element of the bearing calculated by the ferrule flexible body model in S101 and the thermal deformation of each element of the bearing obtained in the step S1054;
s1056, substituting the updated temperature obtained in the step S1054 into the bearing lubrication model in the step S102 to update the viscosity and traction coefficient of the lubricating oil;
s1057, substituting the updated geometric dimensions and traction coefficients in the steps S1055 and S1056 into GUPTA to update the position and speed of the bearing element;
S1058, iterating the steps of S1051-S1057 above until the termination condition is met, and then exiting, wherein the termination condition can be the reaching of the calculation time.
Further, in the step S2, the subsurface stress of the contact area is calculated by using a discrete convolution (Discrete Convolution, DC) and a fast fourier transform (Fast Fourier Transformation, FFT) method.
Further, the step S2 obtains the subsurface stress of the contact area according to the following steps:
S201, calculating a contact half-width b, and dispersing the whole contact area into N x、Ny and N z points along the length direction (x direction), the contact half-width direction (y direction) and the depth direction (z direction) of the rolling element;
S202, calculating an influence coefficient sequence { D } N (a symbol { } N represents a sequence with the length of N), and expanding the sequence { D } N into a sequence { D } 2N (a symbol { } 2N represents a sequence with the length of 2N) through zero padding and surrounding processing;
S203, performing fast Fourier transform on the sequence { D } 2N in the step S202 to obtain
S204, calculating a pressure sequence { P } N, and expanding { P } N into { P } 2N by a zero filling method;
S205, adopting Fourier transformation to the sequence { P } 2N in the S204 step to obtain a sequence
S206, the sequence in the step S203And in step S205Multiplying between the units to obtain the frequency
S207, for the frequency sequence in S206Adopting inverse Fourier transform to obtain a sequence { V } 2N;
S208, extracting the nth (n E [0, N-1 ]) element from the sequence { V } 2N obtained in the step S206, and obtaining the stress distribution.
Further, the calculation formula of the influence coefficient in the step S202 is as follows:
,x+=xm-xξ+Δ1/2,x-=xm-xξ-Δ1/2,y+=yn-yη+Δ2/2,y-=yn-yη-Δ2/2,ξ and η in the above formula are local coordinates of discrete units, m, n, k are element numbers of the units in x, y, z directions, q, r=x, y or z, respectively (i.e. qr is a combination of two of x, y, z).
Further, for elastic half-space contact, the calculation formula of the function T Nqr used by the stress field is as follows:
TNyy(x,y,z)=TNxx(y,x,z)
In the above-mentioned method, the step of, When z=0, the number of times,(Sign () is a sign function).
Further, the step S3 is to calculate the fatigue life of the bearing according to the following steps:
S301, calculating the service lives of the rolling body-rotating ferrule raceway contact position and the rolling body-static ferrule raceway contact position based on L-P;
s302, obtaining the I-H life of the contact positions of each rolling body, the rotary ferrule and the static ferrule according to the data obtained in the step S301;
And S303, obtaining the fatigue life of the bearing according to the data obtained in the step S302.
Further, in the step S301, for the rotating cage, the calculation formula of the L-P life at the contact point of the rolling element and the raceway is as follows:
For a stationary cage, the L-P life at the rolling element-raceway contact is calculated as follows:
In the above formula, q μj and q vj are contact loads of the rotating ferrule and the stationary ferrule (subscripts μ and v represent the rotating ferrule and the stationary ferrule, respectively, j represents the j-th rolling element), q cμj and q cvj are rated dynamic loads of the rotating ferrule and the stationary ferrule, and the calculation formulas of q cμj and q cvj are as follows:
In the above-mentioned method, the step of, D is the diameter of the rolling element, D m is the pitch diameter of the bearing, α is the contact angle, l is the length of the rolling element, and Z is the number of rolling elements.
Further, according to the formula in the step S301, the I-H lives L IHμj and L IHvj of the j-th rolling element in the step S302 where the rolling element contacts the rotating ferrule and the stationary ferrule are obtained, respectively, and the calculation formula is as follows:
In the above expression, σ VMμj,max and σ VMvj,max are the maximum Von-Mises stress at the contact points of the jth rolling element with the rotating ferrule and the stationary ferrule, respectively, and σ VM,lim is the limit value of Von-Mises stress. While AndAre fixed values which vary only in response to changes in bearing type.
Further, the calculation formula in the step S303 is as follows:
The fatigue life of the whole bearing is obtained by superposing the I-H life of the contact positions of each rolling body obtained in the step S302 and the rotating ring and the static ring respectively.
The method has the beneficial effects that the method is different from the single factor in the prior art for calculating the fatigue life of the bearing, is based on the actual working condition of the bearing, and considers the influence of lubrication, heat and flexible deformation of the ferrule on the service life of the bearing under the actual working condition, so that the fatigue life obtained by the method is more fit with the actual working condition, and provides sufficient theoretical support for designing a production end;
Secondly, the lubrication, the heat and the flexible deformation of the ferrule are coupled through parameters, so that the operation difficulty can be reduced, the calculation force cost required by the operation is reduced in a mode of reducing the operation difficulty, and the operation efficiency is improved; and the influence of multiple factors on the service life of the bearing is comprehensively considered, and the interaction among the multiple factors is also considered as correction of each independent factor, so that the fatigue life data obtained by the method is further close to the actual working condition.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of the present invention;
FIG. 2 is a schematic diagram of the overall composition of a dynamic model of thermal-fluid-solid coupling of a bearing according to an embodiment of the present invention;
FIG. 3 is a friction drag curve based on step S102 according to an embodiment of the present invention;
FIG. 4 is a BP neural network model for fitting a friction traction curve according to an embodiment of the present invention;
FIG. 5 shows rolling element-inner ring (rotating collar) contact loads at different temperatures in an embodiment of the present invention;
FIG. 6 shows rolling element-outer ring (stationary race) contact loads at different temperatures in an embodiment of the invention;
fig. 7 shows the bearing life at a temperature of t=300K in an embodiment of the invention;
Fig. 8 shows the bearing life at a temperature of t=310K in an embodiment of the invention.
Detailed Description
In the embodiment of the invention, the cylindrical roller bearing is selected as the bearing, the inner ring of the cylindrical roller bearing is a rotary ferrule, and the outer ring of the cylindrical roller bearing is a static ferrule. The following describes the calculation of the fatigue life of the cylindrical roller bearing according to the analysis method in the present embodiment.
S1, establishing a bearing thermal-fluid-solid coupling dynamic model, wherein the fixing steps are as follows:
s101, establishing a ferrule flexible body model based on a finite element method;
The flexible ferrule is modeled using a two-dimensional curved beam unit.
S102, establishing a bearing lubrication model capable of considering nonlinear viscoelastic characteristics;
the constitutive model of the nonlinear viscoelastic fluid employed can be expressed as:
Where τ is the shear stress and, (U s is the sliding speed between the rolling element and the ferrule, h is the central oil film thickness), G is the shear modulus, τ 0 is the yield shear stress, η is the lubricating oil viscosity,
The calculation formula of the center oil film thickness h is as follows:
h=Ry3.05U0.69g0.56W-0.1
where U is a dimensionless speed parameter, u=η 0u/(EeqRy);
W is a dimensionless load, w=q/(E eqRy l);
g is a dimensionless material parameter, g=λe eq (in the above expression, Q is a contact load, η 0 is a reference viscosity, u is a rolling speed, E eq is an equivalent elastic modulus, l is a rolling body length, λ is a viscosity-pressure coefficient of lubricating oil, and R y is a rolling direction equivalent radius). Here, the influence of pressure and temperature on fluid parameters is obvious, so that the shear modulus G, the yield shear stress τ 0 and the lubricating oil viscosity η need to be corrected according to the pressure p and the temperature T, and the calculation formula is as follows:
In the above formula eta 0, Reference viscosity, reference shear modulus and reference yield shear stress, respectively, alpha i、βj、γk、αG、βG、ατ and beta τ are specific constants, T and T 0 are current temperature and reference temperature, respectively, and p is current pressure.
According to step S102, a friction traction curve can be obtained, which comprises the following specific steps:
S1021, carrying out numerical integration on the constitutive model of the nonlinear viscoelastic fluid characteristic by adopting a Runge-Kutta method, and obtaining shearing stress tau of the contact surface of the rolling body and the ferrule after integration;
s1022, integrating the shearing stress tau on the contact area to obtain the friction traction force;
S1023, comparing the friction traction with the contact load Q to obtain a friction traction coefficient;
s1024, obtaining friction traction curves under different T and U s by changing the temperature T and the sliding speed U s;
s1025, the friction traction curve can be directly used in the dynamics model after the BP neural network is adopted to carry out function fitting on the friction traction curve.
S103, establishing a bearing thermal network model capable of considering the temperature of the lubricating oil, calculating the internal temperature distribution of the bearing and the thermal deformation of each element of the bearing,
The internal temperature calculation process of the bearing is that temperature nodes are arranged at each key part of the rolling bearing, and if the air temperature and the internal temperature of the bearing cavity are known, 9 heat balance equations can be listed according to the law of conservation of energy (the heat flowing in and out of each node is equal), namely, node 1 is (T 1-T10)/R1cv10+(T1-T2)/R2cc1 =0)
Node 2 (T 2-T1)/R1cc2+(T2-T3)/R3cc2 =0)
Node 3:(T3-T2)/R2cc3+(T3-T4)/R4cc3+(T3-T0)/R0cv3=Q1 node 4 (T 4-T5)/R5cc4+(T4-T3)/R4cc3 =0)
Node 5:(T5-T4)/R5cc4+(T5-T6)/R6cc5+(T5-T0)/R0cv5=Q2 node 6 (T 6-T5)/R6cc5+(T6-T7)/R7cc6 =0)
Node 7:(T7-T6)/R7cc6+(T7-T8)/R8cc7+(T7-T0)/R7cv0=Q3 node 8 (T 8-T7)/R8cc7+(T8-T9)/R9cc8 =0)
Node 9 (T 9-T8)/R9cc8+(T9-T10)/R10cv9 =0)
In the above, T 1 is an outer ring temperature node, T 2 is an outer raceway temperature node, T 3 is an outer raceway lubricant temperature node, T 4 is a rolling element-outer raceway outer side node, T 5 is a rolling element temperature node, T 6 is a rolling element-inner raceway inner side node, T 7 is an inner raceway lubricant temperature node, T 8 is an inner raceway temperature node, T 9 is an inner ring temperature node, T 10 is an air temperature node, T 0 is a lubricant temperature node, Q 1、Q2、Q3 is a friction heat generation rate at a rolling element-outer raceway contact position, a friction heat generation rate between a rolling element and lubricating oil in a bearing cavity, and a friction heat generation rate at a rolling element-inner raceway contact position, R is a thermal resistance value (in the subscript, cv represents heat convection, cc represents heat conduction, and Arabic represents a temperature node).
After the temperature T of each element of the bearing is calculated, the thermal deformation of the bearing element is calculated by the following formula:
ut=ΓL(T-Tα)
in the above formula, Γ is the thermal expansion coefficient, T α is the ambient temperature, and L is the characteristic length of each element of the bearing.
S104, establishing a dynamic model of the cylindrical roller bearing by considering the interaction relation among the elements of the bearing.
And establishing a dynamic model of the cylindrical roller bearing based on the GUPTA bearing model.
S105, realizing the coupling of the models in the four steps S101-S104 by a parameter coupling mode, establishing a thermal-flow-solid coupling dynamic model of the cylindrical roller bearing, and respectively calculating the contact loads between the rotating ferrule and the static ferrule and the rolling body, wherein the specific steps are as follows:
s1051, initializing the temperature, the position and the speed of each element of the bearing;
S1052, calculating the current friction traction coefficient based on the bearing lubrication model in the step S102;
S1053, calculating the geometric approach amount and the contact load and the friction heat generation rate between the elements of the bearing based on the GUPTA dynamic model in the step S104;
S1054, substituting the friction heat generation rate obtained in the step S1053 into the bearing thermal network model in the step S103 to update the temperature of each element of the bearing, and obtaining the thermal deformation of each element of the bearing;
S1055, updating the geometric dimensions of each element of the bearing based on the elastic deformation of each element of the bearing calculated by the ferrule flexible body model in S101 and the thermal deformation of each element of the bearing obtained in the step S1054;
s1056, substituting the updated temperature obtained in the step S1054 into the bearing lubrication model in the step S102 to update the viscosity and traction coefficient of the lubricating oil;
s1057, substituting the updated geometric dimensions and traction coefficients in the steps S1055 and S1056 into GUPTA to update the position and speed of the bearing element;
S1058, iterating the steps S1051 to S1057 above until the termination condition is satisfied (the calculation time is reached).
And the establishment of the bearing thermal-fluid-solid coupling dynamic model is completed through the steps.
S2, establishing a contact subsurface stress calculation model, and calculating the contact subsurface stress according to the contact load obtained by calculation in the S1, wherein the specific steps are as follows:
S201, calculating a contact half-width b, and dispersing the whole contact area into N x、Ny and N z points along the length direction (x direction), the contact half-width direction (y direction) and the depth direction (z direction) of the rolling element;
S202, calculating an influence coefficient sequence { D } N (a symbol { } N represents a sequence with the length of N), and expanding the sequence { D } N into a sequence { D } 2N (a symbol { } 2N represents a sequence with the length of 2N) through zero padding and surrounding processing;
S203, performing fast Fourier transform on the sequence { D } 2N in the step S202 to obtain
S204, calculating a pressure sequence { P } N, and expanding { P } N into { P } 2N by a zero filling method;
S205, adopting Fourier transformation to the sequence { P } 2N in the S204 step to obtain a sequence
S206, the sequence in the step S203And in step S205Multiplying between the units to obtain the frequency
S207, for the frequency sequence in S206Adopting inverse Fourier transform to obtain a sequence { V } 2N;
S208, extracting the n (n E [0, N-1 ]) element sequence number from the sequence { V } 2N obtained in the step S206, and obtaining the stress distribution.
The calculation formula of the influence coefficient in step S202 is as follows:
,x+=xm-xξ+Δ1/2,x-=xm-xξ-Δ1/2,y+=yn-yη+Δ2/2,y-=yn-yη-Δ2/2,ξ and η in the above formula are local coordinates of discrete units, m, n, k are element numbers of the units in x, y, z directions, q, r=x, y or z, respectively (i.e. qr is a combination of two of x, y, z). For elastic half-space contact, the function T Nqr used for stress field is calculated as follows:
TNyy(x,y,z)=TNxx}y,x,z)
In the above-mentioned method, the step of, When z=0, the number of times,(Sign () is a sign function).
The contact subsurface stress calculation model is completed through the steps.
S3, calculating the fatigue life of the bearing based on the I-H theory, wherein the specific steps are as follows:
S301, calculating the service lives of the rolling body-rotating ferrule raceway contact position and the rolling body-static ferrule raceway contact position based on L-P;
for a rotating cage, the calculation formula for the L-P life at the rolling element-raceway contact is as follows:
For a stationary cage, the L-P life at the rolling element-raceway contact is calculated as follows:
In the above formula, q μj and q vj are contact loads of the rotating ferrule and the stationary ferrule (subscripts μ and v represent the rotating ferrule and the stationary ferrule, respectively, j represents the j-th rolling element), q cμj and q cvj are rated dynamic loads of the rotating ferrule and the stationary ferrule, and the calculation formulas of q cμj and q cvj are as follows:
In the above-mentioned method, the step of, D is the diameter of the rolling element, D m is the pitch diameter of the bearing, α is the contact angle, l is the length of the rolling element, and Z is the number of rolling elements.
S302, according to the data obtained in the step S301, the I-H life L IHμj and L IHvj of the contact positions of the jth rolling element and the rotary ferrule and the static ferrule are obtained, and the calculation formula is as follows:
In the above expression, σ VMμj,max and σ VMvj,max are the maximum Von-Mises stress at the contact points of the jth rolling element with the rotating ferrule and the stationary ferrule, respectively, and σ VM,lim is the limit value of Von-Mises stress. While AndAre fixed values which vary only in accordance with the variation of the bearing type, in this embodiment, that is, the coefficients of the cylindrical roller bearing are as follows:
S303, obtaining the fatigue life of the bearing according to the data obtained in the step S302, wherein the calculation is shown as follows:
The fatigue life of the whole bearing is obtained by superposing the I-H life of the contact positions of each rolling body obtained in the step S302 and the rotating ring and the static ring respectively.
And the fatigue life calculation of the bearing is completed through the steps.
The above embodiment is only one of the preferred embodiments of the present invention, and common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention are included in the scope of the present invention.