CR Classification Society
CHAPTER 7 TYPE TESTING
7.1 General
7.1.1 At least one drive shaft/flexible coupling design shall be tested with respect to properties under static torsional load. Fatigue testing shall be carried out as required in the preceding chapters. If the bending moment in the shaft is significant, testing with bending moments may also be required.
7.2 Test Specimens
7.2.1 At least one test specimen shall be prepared for testing of the static strength. Specimens for fatigue testing shall be prepared as agreed with the Society. The test specimens shall be representative for the normal production. The same materials and fabrication methods as applied in the normal production shall be used when fabricating the specimens.
The nominal torque of the specimen(s) for testing shall be at least equal to 30% of the maximum nominal torque included in the range for which the type approval shall apply.
For shafts the length of the central section between the innermost edges of the end flanges shall be at least equal to 3 times the outside diameter of the central section. For particular designs where the length of the component is less than 3 times the diameter the requirement to the length of the specimen may be waived.
The interface between the central section and the end flanges shall be identical in design to normal production shafts.
Modifications to the metallic flanges for testing purposes, not affecting the performance of the joint are acceptable.
7.3. Test under Static Load
The purpose of the test is to verify that the calculated torsional strength and stiffness of the shaft will be reached in actual production with a certain level of confidence. As a minimum one test shall be carried out.
7.3.1 Instrumentation:
The following instrumentation shall be included:
(a) equipment for continuously measuring the torque with an uncertainty < 4%.
(b) equipment for continuously measuring the twist between the end flanges with an uncertainty to be agreed in each case.
(c) equipment for continuous (or equivalent) logging of torque and twist.
It is recommended that additional equipment such as e.g. strain gauges are included to gain further information regarding the performance of the shaft and to verify the design calculations.
7.3.2 Test environment:
The test shall be carried out in a temperature within the range 22 ± 5°C and with a relative humidity within the range 35% – 90% unless otherwise agreed.
7.3.3 Test procedure:
The specimen shall be loaded in pure torsion. Four load sequences shall be carried out:
Seq. 1-3: the shaft shall be loaded to peak torque and back to zero torque three times
CHAPTER 7 TYPE TESTING 7.4. Full Scale Fatigue Testing
CR Classification Society
Seq. 4: the torque shall be increased to failure of the shaft.
In all sequences the torque shall be increased/decreased with a rate not exceeding the nominal torque/60 per second.
When the torque exceeds three times the nominal torque, sensitive measuring equipment, except the equipment measuring and logging the torque, may be disconnected.
After the test has been completed a graph or graphs over torque vs. twist until failure with adequate resolution and covering all sequences shall be submitted to the Society together with documentation of the location of the failure and the mechanism of the failure.
7.3.4 Acceptance criteria:
The maximum torque recorded during the test, Tfail, shall satisfy the following requirement:
Tfail≥ 1.16 × SFmax× Peak torque
Where SFmax is equal to the maximum of the safety factors SF specified in Table 6-1.
Note :
The requirement is based on the assumption that the standard deviation of torsional strength of samples from normal production does not exceed 7% of the mean. It is required that the Tfail should exceed the expected mean failure torque value (1.16 = 1/[1.0 – 2×0.07]).
SF may have different values for different parts of the component (e.g. for the central section and the joints). To be able to test all relevant parts of the component it may be necessary to make specially designed test specimens (different from the final design of the component) for testing of each part of the component, e.g. the central section and the joints to the end flanges. More than one specimen may have to be tested.
If the test result fails to meet the requirement above an additional specimen shall be tested. The mean value of the maximum torques recorded in the two tests shall exceed:
1.16 × SFmax× Peak torque
No result shall be lower than SFmax×Peak torque
7.4. Full Scale Fatigue Testing
Full scale fatigue testing shall be carried out when required as specified in 7.3.
7.4.1 Purpose:
The purpose of the test(s) is to verify the fatigue strength of the shaft and that it will be reached in actual production with a certain level of confidence.
7.4.2 Fatigue test load condition:
The test condition during the fatigue test(s) shall be based on the fatigue load conditions as specified in 2.3. A table as shown below shall be established:
CHAPTER 7 TYPE TESTING 7.4. Full Scale Fatigue Testing
CR Classification Society Condition Mean Amplitude Range Cycles
1 M1 A1 ΔT1 N1
Ai = torque amplitude for condition “i”.
ΔTi = equivalent torque range for condition “i”.
Ni = number of load cycles for condition “i”.
The equivalent torque range is defined for R = 0. ΔTi is calculated according to the following equation:
ΔTi = 2 × Ai 1 −Mi
UT+ Ai UT
UT = ultimate torsional strength of the central section as measured in the static test.
Note :
The equation is based on the assumption that the fatigue strength of the component can be described by straight line in a Haigh-diagram where the line intersects the x-axis (mean load) at UT.
7.4.3 Definition of safety margin:
The safety margin applied in the fatigue test is composed of two elements:
(a) to account for possible sequence effects from the service fatigue load history.
(b) to ensure an adequate reliability of the shaft with respect to fatigue failure.
Note :
A composite shaft should have the same reliability with respect to fatigue failure as the corresponding steel shaft.
To account for the first requirement the factor F1 is set to F1 = 5.
To account for the second requirement the factor F2 is set to F2 = 102log(σ) where log(σ) is equal to the standard deviation of the logarithm of the fatigue life. In lack of more precise information log(σ) can be set equal to 0.4. (Log(x) corresponds to the 10-base logarithm).
F1×F2 shall not be taken smaller than 32.
7.4.4 Definition of minimum required fatigue curves:
For each condition “i” calculate mi and Ci according to the following equation:
mi =[log(Ni) + log(F1) + log(F2)]
[log(UT) − log(∆Ti)]
Ci = UTmi
Note :
It is assumed that the fatigue strength of the component can be represented by the following expression (i.e. a linear representation in a log-log-diagram): N = C×ΔT-m.
CHAPTER 7 TYPE TESTING 7.4. Full Scale Fatigue Testing
CR Classification Society
Determine the required fatigue curve:
m = maxi(mi) C = maxi(Ci)
7.4.5 Fatigue damages:
Calculate the fatigue damage for each condition “i”:
Di=Ni
C × ∆Ti−m
Calculate the total fatigue damage and relative fatigue damages:
Dtotal= ΣiDi total fatigue damage di= relative fatigue damage for condition “i”
Note :
It is assumed that linear damage accumulation (Miner’s Rule) is representative.
7.4.6 Fatigue test condition:
Determine the fatigue test condition ΔTtest and Ntest such that the following two conditions are satisfied:
Ntest=C × ΔTtest−m
maxi(δi)
2×mini (Ai) ≤ ΔTtest ≤ torque at onset of matrix cracking1 Note 1: as determined by the design calculations
7.4.7 Instrumentation:
The following instrumentation shall be included:
(a) equipment for continuously measuring the torque with an uncertainty < 5 %.
(b) equipment for continuously measuring the twist between the end flanges with an uncertainty to b e agreed in each case.
(c) equipment for continuous (or equivalent) logging of torque and twist.
It is recommended that additional equipment such as e.g. strain gauges are included to gain further information regarding the performance of the shaft and to verify the design calculations.
7.4.8 Test environment
The test shall be carried out in a temperature within the range 22 ± 5°C and with a relative humidity within the range 35 – 90% unless otherwise agreed.
7.4.9 Test procedure
The specimen shall be loaded in pure torsion.
The following sequence shall be followed:
CHAPTER 7 TYPE TESTING 7.4. Full Scale Fatigue Testing
CR Classification Society (a) the shaft shall be loaded to extreme torque and the load released three times. The torque shall be
increased/decreased monotonously with a rate not exceeding the nominal torque/60 p er second
(b) the torsional stiffness is measured
(c) fatigue test at the following conditions:
range of torque: ΔTtest
R-ratio: ≤ 0.05
number of load cycles: the larger of Ntest or 5×106 load cycles, or to failure.
(d) the torsional stiffness shall be measured at Ntest
During sequence 3 the equipment for measurement of twist may be disconnected.
7.4.10 Acceptance criteria:
In case the number of load cycles to failure Nfail > Ntest the test result is acceptable.
In case the shaft fails at Nfail < Ntest an additional fatigue test shall be carried out. The mean value of the log(Nfail) for the two tests shall be larger than log(Ntest).
In case the shaft fails at a number of load cycles Nfail < Ntest/102 log(σ) the test result is unacceptable.
No failure signifies that no failures or damages of any kind are observed on the FRP central section or in the bonds between central sections and end flanges after completion of the test. After completion of the test the bonds on the shaft shall be inspected carefully such that it can be ascertained that no damages to the bonds have occurred. Normally this will mean that the bond have to be cut through the thickness at least 4 locations around the circumference of the bond such that the bond line is exposed for inspection.
CHAPTER 8 DOCUMENTATION REQUIRED FOR EACH DELIVERY