GB/T 37306.1-2019 PDF English
Search result: GB/T 37306.1-2019 English: PDF (GB/T37306.1-2019)
Standard ID | Contents [version] | USD | STEP2 | [PDF] delivered in | Name of Chinese Standard | Status |
GB/T 37306.1-2019 | English | 150 |
Add to Cart
|
0-9 seconds. Auto-delivery.
|
Metallic materials -- Fatigue testing -- Variable amplitude fatigue testing -- Part 1: General principles, test method and reporting requirements
| Valid |
BUY with any currencies (Euro, JPY, GBP, KRW etc.): GB/T 37306.1-2019 Related standards: GB/T 37306.1-2019
PDF Preview: GB/T 37306.1-2019
GB/T 37306.1-2019: PDF in English (GBT 37306.1-2019) GB/T 37306.1-2019
Metallic materials--Fatigue testing--Variable amplitude fatigue testing--Part 1. General principles, test method and reporting requirements
ICS 77.040.10
H22
National Standards of People's Republic of China
Metal material fatigue test variable amplitude fatigue test
Part 1. General, test methods and reporting requirements
Part 1. Generalprinciples, testmethodandreportingrequirements
(ISO 12110-1.2013, IDT)
Published on.2019-03-25
2020-02-01 implementation
State market supervision and administration
China National Standardization Administration issued
Content
Foreword I
1 range 1
2 Normative references 1
3 Terms and Definitions 1
4 Test principle 3
4.1 Control signal generation 3
4.2 Test Method Overview 3
5 original load time history 5
5.1 General 5
5.2 Data Filtering 5
6 loading time history 5
6.1 General 5
6.2 Time History Sequence 5
6.3 Cycle Counting Method 6
7 program module 6
8 Conversion matrix and its control signal generation 9
8.1 Matrix establishment 9
8.2 Reconstruction of load signals 9
8.3 Control signal simplification 10
9 Variable amplitude fatigue test 10
10 Test report 11
10.1 General 11
10.2 Raw loading 11
10.3 Test conditions 11
10.4 Initial analysis of test data for individual and serial samples 12
Appendix A (informative) Standard loading time history 13
Appendix B (informative) Randomly extracted loading signal reconstruction example in the transformation matrix
Appendix C (informative) Initial analysis of test data for individual samples 16
Reference 19
Foreword
GB/T 37306 "Metal material fatigue test variable amplitude fatigue test" is divided into two parts.
--- Part 1. General, test methods and reporting requirements;
--- Part 2. Loop count and related data reduction methods.
This part is the first part of GB/T 37306.
This part is drafted in accordance with the rules given in GB/T 1.1-2009.
This part uses the translation method equivalent to ISO 12110-1.2013 "Metal material fatigue test variable amplitude fatigue test Part 1.
General rules, test methods and reporting requirements.
The documents of our country that have a consistent correspondence with the international documents referenced in this part are as follows.
---GB/T 3075-2008 Metal material fatigue test axial force control method (ISO 1099.2006, MOD);
---GB/T 6398-2017 fatigue fatigue crack propagation method for metal materials (ISO 12108.2012, MOD);
---GB/T 24176-2009 Metal material fatigue test data statistical program and analysis method (ISO 12107.2003, IDT);
---GB/T 26077-2010 Metal material fatigue test axial strain control method (ISO 12106.2003, MOD);
---GB/T 34104-2017 Metallic material testing machine for testing the coaxiality of the test (ISO 23788.2012, MOD).
This section has made the following editorial changes.
---Using the national standard GB/T 37306.2 "Metal materials fatigue test variable amplitude fatigue test Part 2. Cycle count and phase
The Data Reduction Method replaces the international standard ISO 12110-2 (see 6.3, Chapter 9).
This part was proposed by the China Iron and Steel Association.
This part is under the jurisdiction of the National Steel Standardization Technical Committee (SAC/TC183).
This section drafted by. Guangzhou University, Shanghai University, Guangdong Entry-Exit Inspection and Quarantine Bureau Inspection and Quarantine Technology Center, Shanghai entry and exit inspection
Quarantine Bureau Industrial Products and Raw Materials Testing Technology Center, Shenzhen Wan Testing Equipment Co., Ltd.
Drafters of this section. Xu Zhonggen, Wu Yiwen, Wang Hongbin, Zhou Qi, Guo Xuan, Huang Xing.
Metal material fatigue test variable amplitude fatigue test
Part 1. General, test methods and reporting requirements
1 Scope
This part of GB/T 37306 specifies the general rules for the cyclical fatigue test of laboratory samples for each cycle of amplitude variation.
This section specifies and gives general rules for variable amplitude fatigue tests to generate for comparison when considering the typical dispersion of fatigue data.
Consistent results.
This section is theoretically applicable to strain and force control or loading conditions for controlling fatigue crack growth rate, using this part of the loading
Mode, not force - Appropriate protection should be taken when controlling the loading mode.
This section applies to single actuator loading mode for single axis loading.
The variable loading time history referred to in this section is deterministic, which is why this part deals with variable amplitude loading rather than random loading.
The following items are outside the scope of this section.
--- Constant amplitude fatigue test with isolated overload or underload phenomenon;
---Experiment of large parts or structural parts;
--- Environmental impacts such as corrosion, temperature/time related creep leading to effects on frequency and waveforms;
--- Multi-axis loading.
2 Normative references
The following documents are indispensable for the application of this document. For dated references, only dated versions apply to this article.
Pieces. For undated references, the latest edition (including all amendments) applies to this document.
ISO 1099 metal material fatigue test axial force control method (Metalicmaterials-Fatiguetesting-Axial
Force-controledmethod)
ISO 12106 Metal material fatigue test axial strain control method (Metalicmaterials-Fatiguetesting-
Axial-strain-controledmethod)
ISO 12107 Metal material fatigue test data statistical program and analysis method Metalicmaterials-Fatigue
testing-Statisticalplanningandanalysisofdata)
ISO 12108 metal material fatigue test fatigue crack growth method (Metalicmaterials-Fatiguetesting-
Fatiguecrackgrowthmethod)
ISO 23788 Metallic material fatigue testing machine for testing the coaxiality (Metalicmaterials-Verification ofthe
Alignmentoffatiguetestingmachines)
3 Terms and definitions
The following terms and definitions defined by ISO 1099, ISO 12106, ISO 12107 and ISO 12108 apply to this document.
3.1
Cumulative frequency diagram cumulativefrequencydiagram
A histogram representing the accumulation of each cycle since the start of the test.
Note. The cumulative frequency map is also known as the cumulative spectrum or cumulative distribution.
3.2
Cycle cycle
Periodically repeated force-time/stress-time/strain-time under constant amplitude fatigue loading, or in some other signal applied to the specimen
The smallest segment.
Note. In variable amplitude loading, the definition of the loop varies depending on the counting method used.
3.3
Loop counting method cyclecountingmethod
A method of calculating the number of load time history loops for a given length.
3.4
Loading loading
A general term for applying varying forces, strains, or other controlled quantities on the specimen.
Note. This section mainly refers to the force control loading mode.
3.5
Load distribution loadingdistribution
A simple distribution or cumulative distribution of loads as a function of cycle.
Note 1. The load distribution is the result of statistical processing of actual service loading records or a typical distribution unique to industrial fields (eg automotive, aerospace). Load distribution
Suitable for force loading mode and strain loading mode or other loading modes.
Note 2. The load distribution is often referred to as the “load spectrum”. Avoid using frequency domain loading.
3.6
Load histogram loadinghistogram
A simple histogram or cumulative histogram of the load as it changes with the cycle.
Note 1. The loading time history is the result of statistical processing of actual service loading records or a typical distribution unique to industrial fields (eg automotive, aerospace). load
Histograms are available for force loading mode and strain loading mode or other loading modes.
Note 2. The term “force history” should have been used in the force-controlled loading mode, but this is not common in the field of variable amplitude fatigue. Whether the control variable is packaged or not
In addition, always use the term "load time history."
3.7
Loading time history loadingtimehistory
A sequence of load cycles in which the load amplitude changes from one cycle to the next.
Note 1. The loading time history is the result of statistical processing of actual service loading records or a typical distribution unique to industrial fields (eg automotive, aerospace). load
The time history applies to force loading mode and strain loading mode or other loading modes.
Note 2. The term “force history” should have been used in the force-controlled loading mode, but this is not common in the field of variable amplitude fatigue. Whether the control variable is packaged or not
In addition, always use the term "load time history."
3.8
Load power spectrum loadingpowerspectrum
Energy density spectrum energydensityspectrum
The frequency domain description of the random loading time history.
Note. The power spectrum is the Fourier integral of the time signal correlation function.
3.9
Ignore omission
A cycle that removes non-damaging cycles or load amplitudes less than a threshold level is removed.
3.10
Ignore the level of the omissionlevel
The threshold level of the non-damaging cycle is removed.
3.11
Peak peak
The first derivative of the load time history changes from positive to negative.
Note. For constant amplitude loading, the peak corresponds to the maximum load. For variable amplitude loading, the peak corresponds to the local maximum load in the loading time history.
3.12
Random randomdraw
A semi-cyclic sequence with different ranges and average values.
3.13
Valley valey
The first derivative of the loading time history changes from negative to positive.
Note 1. The valley value is a relative minimum or "valley".
Note 2. The valley value is the minimum load point for constant amplitude loading.
3.14
Variable amplitude loading variableamplitudeloading
All loading modes where the peak loads are not equal or all valley loads are not equal or all peak and valley loads are not equal.
Note 1. Also known as "irregular loading".
Note 2. “Spectrum loading” cannot replace variable amplitude loading because spectral loading is not a function of load and time.
4 Test principle
4.1 Control signal generation
It is difficult to obtain any number of cycles or failures from the original loading time history and thereby directly and effectively control the fatigue test, and
The number of records of the actual load time history added to the sample does not represent the actual load, and the actual load can only be from the full load signal.
It is calculated from the calculation that these statistical characteristics are obtained from a large number of actual loading measurements; therefore, the original loading time history usually needs to be simplified.
Can be applied to the sample.
The simplification of the original load time history is usually done by loading the signal analysis onto two analog load control signals. These two
The analog control signal is obtained by randomly extracting signals from a program module or a conversion matrix. If the testing machine and related electronic equipment can be simplified
The original loading time history, which is not simplified, can be applied directly to the sample.
Raw signal analysis was performed using the cycle counting method. The data obtained from the loop count is then used to build the cumulative frequency of the program module
Rate map or randomly derived transformation matrix.
Note 1. The main advantage of the program module is that the control signal consists of a series of constant amplitude blocks, each of which has a different amplitude. Therefore, there is no need to pass calculations
The machine generates complex digital control signals.
Note 2. Regardless of the complexity of the reconstruction control signal randomly selected from the transformation matrix, the control signal is still more than the actual loading for the actual loading.
Program modules are more representative. In addition, due to the tremendous advances in digital electronics and computers, it has been generated by random extraction over the past few decades.
Control signals have become easier and easier.
Signals should be filtered under the following conditions. Filtering should be carefully selected to avoid improper filtering and significant removal.
Damage fatigue cycle. Average stress, residual stress, isolated high amplitude overload, etc. should be handled with caution.
a) The raw signal obtained from direct measurement of loaded parts in service usually contains electronic noise or other non-fatigue vibrations. In use
These interferences should be eliminated before the loop counting procedure processes the original signal;
b) Since the non-damage cycle is usually the most (see 8.3.1), when the quick test is required, the slave program module can be removed.
Or control signals obtained in random extraction to eliminate non-injury cycles (shortest cycles) to significantly reduce test duration.
Note 3. These isolated high amplitude overloads actually increase the fatigue life due to the generation of favorable residual stresses when there is an isolated high amplitude overload in the load time history.
4.2 Overview of test methods
The variable amplitude fatigue test is applied to the sample using a cumulative frequency map (program module) or a control signal obtained by random extraction.
The response of the sample is monitored by a load sensor or a force sensor or strain gauge given measurement and these output data are used for closed loop control.
Note. The variable amplitude fatigue test usually uses hydraulic servo test equipment, but other actuators can be used in the case of closed loop control test.
The test results are reported when the specimen breaks into two parts or when another failure criterion fails. Test results may include cycles of failure
Number of times or sequence, crack spread measurement or any other sample damage process data.
The specific test and test principle is shown in Figure 1. The detailed information of the main steps of the test is given in the subsequent part of this standard.
Note. The dashed box is an optional step.
Figure 1 Flow chart for collecting variable amplitude loading data and converting it into laboratory variable amplitude fatigue test input data
5 original loading time history
5.1 General
There are two types of sources of load time history for an original component or structure.
a) The first type of source is a direct measurement of the load condition of an in-service component or structural member, using a strain gage or other sensing device.
Record and store the measurement results using a digital data acquisition system;
Note. Automotive hubs, suspension systems, railway bogies, turbine blades, and aircraft wing beams are typical components that are subject to fatigue loading.
b) The second type of source is the typical standard loading time history in the industrial sector, the meaning of which is usually the majority of the relevant industrial sectors.
Recognized by personnel.
The original load time history is usually composed of a load sequence that completely repeats a given length (time or number of cycles). From a sequence
Only very subtle changes can be observed by listing to the next sequence.
When the original loading time history is obtained by directly measuring the in-service component load, filtering may be required to eliminate electrical or mechanical noise.
sound. Filter parameters should be carefully selected when filtering to avoid removal of significant damage fatigue cycles.
The average stress modulation filter can be used (see 8.3).
5.2 Data Filtering
5.2.1 General requirements
Efficient data filtering can greatly reduce the amount of data used to generate the variable amplitude fatigue control signal.
Filtering typically contains a threshold or threshold for the set load or load amplitude.
The choice of threshold or threshold should take into account the existing knowledge and experience of the fatigue process studied, and in particular to avoid ignoring parts or
The true damage cycle that plays a major role in the fatigue damage process of the component.
5.2.2 Noise Filtering
High-frequency low-amplitude peaks superimposed on the fatigue loading signal that do not contribute to fatigue are usually ignored, as by data recording systems (strain gauges)
The resulting electronic noise shall be indicated in the test report for each individual sample.
6 loading time history
6.1 General
The load time history can be described in any of the following three ways.
---Time history sequence;
--- Cycle count;
--- Power or energy density spectrum.
6.2 Time history sequence
Variable amplitude loading can be divided into discontinuous or partially continuous random processes. These stochastic processes can be calculated by in-service measurements or time step calculations.
determine.
The short range can be represented by a continuous force-time signal, and the long range needs to be represented by a series of consecutive short ranges (see Figure 2), where each short range
Generally continuous. The long distance is obtained by connecting the short distances in the correct order.
Description.
X --- time;
Y --- load;
1 --- Load order.
Figure 2 Load-time history
6.3 Cycle counting method
The load time history can be represented by a series of loop numbers. In this case, the order in which the loop occurs is lost.
The raw load time history is processed by a loop counting program that defines the number of cycles and cycles by any time of fatigue life.
To describe the original signal. This program allows the number of fail cycles for a component to be defined in the same way as a constant-load condition.
Different cycle counting methods are available. All methods are to divide the entire load range (between the minimum and maximum values) or
level. The 32 loading levels are basically sufficient (see Figure 2). Refer to GB/T 37306.2 for the counting method.
Note 1. General industrial practice uses 64 loading levels.
Note 2. The most common cycle counting methods are.
--- Cross level counting;
--- Peak count;
--- Simple range count;
--- Range pairs count;
--- "Rain flow" count.
When the original load time history is completely processed by the correlation counting method, it is performed in one of two ways.
a) determine the cumulative frequency map and establish the program module;
b) Determine the transformation matrix by random extraction and reconstruct the load time history.
The fatigue life of the sequence of program modules is sometimes quite different from the fatigue life under random loading sequences and should be avoided as much as possible.
7 program module
The program module is derived from a cumulative frequency map corresponding to the cumulative probability expressed by the relationship between the loaded excess and the number of cycles.
It is a smooth continuous curve (see Figure 3).
Description.
1---Gauss normal distribution;
X---loop;
Y---normalized stress range.
Figure 3 Example of cumulative load graph
Simplify the cumulative frequency map into discrete box plots. Load division should be carried out in the form of equivalent damage. The first box is the largest discrete load level
Box diagram, other discrete box diagrams take the average of each load. Such a constant amplitude loading box is also called a module (see Figure 4).
Description.
X --- cumulative cycle;
Y --- normalized stress range.
Note. Such modules allow Palmgren-Miner to sum.
Figure 4 Modularization of cumulative load maps
These modules are extracted from the cumulative frequency map to create a load sequence as shown in FIG. 8~10 modules in the general sequence are enough
A cumulative frequency map with a cumulative number of no more than 106 times is well described.
Description.
a---stress;
b---a time series.
Figure 5. 500000 cycle Gassner eight-step design program sequence
The cumulative frequency plot is also a typical representation of the correct loading pattern for a real part or structure. The load distribution shown in Figure 6 is non-Gaussian
An example of a distribution in which a high relative frequency corresponds to a high load. This is a typical high load structure such as cranes, lifting bridges and the like. Load distribution
"a" represents the load that is loaded at a constant amplitude under the maximum load. The load distribution "d" represents the Gaussian normal distribution of the load (expected to be overloaded at the maximum load)
For example, zero load at 106 o'clock).
Description.
X --- cumulative cycle, Nc;
Y --- normalized stress range.
X=(Ntot)1-Y
X = cumulative cycle;
Ntot = total cumulative cycle;
Y = uniform load amplitude (maximum load corresponds to 1).
Figure 6 Example of different load distribution functions
The process of designing a load sequence by building a module is often referred to as block programming. Since each module is included in constant amplitude loading,
These program module blocks are easily generated by the test machine controller without the need for a complex real-time computer control system to run the program modules.
The order in which the modules appear may have a large impact on the fatigue life of a few pieces of fatigue testing. For example, in only two modules
In the fatigue test, the fatigue life of the module that first acts on the large value is smaller than that of the module that first acts on the small amplitude. In order to reduce the appearance of the module
The effect of the sequence on fatigue life is recommended to repeat at least 20 tests.
8 Conversion matrix and its control signal generation
8.1 Matrix establishment
Another method of implementing a variable amplitude fatigue test is to reconstruct the control signal by random extraction using a transformation matrix.
The transformation matrix is built from the level or level of loading defined in the loop count process. From a relative extremum to the next relative extremum
The number of conversions is recorded in the matrix. For example, the number of conversions aij from the i level to the j level is recorded in the elements of the matrix i rows and j columns. If you turn
The change is from the relative maximum in the i class to the relative minimum in the j class (i>j) (peak valley), then the conversion aij is recorded below the diagonal of the matrix
(See Figure 7). The value aij in the diagonal should be zero or very close to zero because the transition from i level to the same level i rarely occurs. in case
This conversion occurs more frequently at the same level and the original signal should be re-divided into more narrower levels. This will improve signal modeling
Precision. When using computer technology to analyze the original load time history, the loop count and the transformation matrix can be established.
Time to proceed.
The setup is complete when the conversion matrix is completely filled with the number of conversions. The control signal used to load the sample is then passed through the matrix
A random extraction is performed for reconstruction.
Description.
1---stress level i;
2---stress level j;
3---from minimum to maximum conversion, i \u003cj;
4---The conversion from largest to smallest, i>j.
Figure 7 transformation matrix
8.2 Reconstruction of load signals
8.2.1 Conversion matrix method
The load signal is reconstructed by random decimation of the transformations in the matrix using the following procedure. Actually only some and not all extractions are
Really random, because discontinuities should be avoided in the reconstructed signal, and each up-conversion should be followed by a down-conversion and vice versa (up
And the downward conversion should be alternated).
An example of random extraction is given in Appendix B. The value of i is random, such as i=α. j should satisfy α \u003cj≤n。在区间
The first conversion randomly selected in [j, β] is ααβ. In order to avoid discontinuities and produce a downward turn after the first transition
In other words, the next conversion should be αβj and j should satisfy 1≤j< β. Randomly extracted within this interval, j = λ, and the second conversion is
(λ< δ).
Note. In special cases, the constraint can be fixed at the maximum acceptable difference between two subsequent conversions (1/2 cycle) to prevent a non-representative cycle.
8.2.2 Energy density spectroscopy
Energy density spectroscopy can be applied to reconstruct the load signal using a variety of programs. For each program, the key is to verify the time letter generated.
The energy of the number is equal to the spectral energy. For example, the following program can be applied.
S is the total area (total energy) of E(ω) in the energy density spectrum, and a series of uniformly distributed ωi is selected to cover the full range of spectrum ω;
The number is N; the interval between two consecutive ωi is Δω.
Load the signal as shown in equation (1).
F(t)=α∑
Aisin(iΔωt ψi) (1)
In the formula.
Ai---component amplitude, equal to [ΔωE(ωi)]1/2;
Ψi---the random phase angle randomly extracted from the uniform distribution;
α --- total energy adjustment factor, see equation (2).
∑2∑
2A
i=S (2)
8.3 Control signal simplification
8.3.1 Overview
Although the reconstructed control signal is a simplified model of the average true loading experienced by the component, it is usually still a complex
signal.
In many cases, as long as the load threshold level is defined, the loop of the smallest magnitude can be ignored. Can then eliminate the amplitude less than plus
Any cycle that carries a threshold level. This involves filtering the original signal or the reconstructed signal or both.
Studies have shown that a variable amplitude cycle with amplitudes less than the loading threshold level can cause damage to certain materials. Therefore, the test results should be clearly remembered
Record when and how to eliminate low load width cycles from the load signal.
Since the non-damaging small cycle is the most (low-load amplitude vibration), this filtering makes the control signal simpler, thus making the fatigue test
The process is easier.
If not properly filtered, an error may occur. the higher the threshold level, the simpler the resulting control signal, but the actual damage cycle is ignored.
The greater the probability of overestimating the fatigue life of the tested parts.
8.3.2 Ignore small amplitude loops
A small amplitude cycle with an amplitude less than a fixed threshold is considered non-invasive and has little effect on the fatigue test results and can be ignored.
The average stress modulation filter can......
...... Source: Above contents are excerpted from the PDF -- translated/reviewed by: www.chinesestandard.net / Wayne Zheng et al.
|