GB2066964A

GB2066964A - Controlling mechanical test apparatus electrically

Info

Publication number: GB2066964A
Application number: GB8041561A
Authority: GB
Original assignee: Individual
Current assignee: Individual
Priority date: 1980-01-07
Filing date: 1980-12-31
Publication date: 1981-07-15
Also published as: GB2066964B

Abstract

A notched specimen 10 is subjected to a cyclically varying load applied by a hydraulic actuator 24. Strain gauges in a load cell 21 feed an applied load signal to amplifier 33, and strain gauge 16 on the back face of the specimen feeds a strain signal to amplifier 33 which adds the signals in a weighted ratio and produces an output representing DELTA K, the alternating stress intensity factor at the tip of the crack. The output of amplifier 33 is compared in servo amplifier 36 with a reference value of DELTA K from 37, and a resultant signal controls servo valve 25 of actuator 24 to reduce the load applied to the specimen as the crack grows, to maintain DELTA K, and hence crack growth rate, constant or to vary crack growth rate in a chosen manner. <IMAGE>

Description

SPECIFICATION Test apparatus and method The present invention relates to a test appa ,ratus and method for use in investigating the formation and/or growth of a crack in a cracked or notched specimen subjected to a load of varying amplitude applied across the crack.

The fatigue testing of a so-called compact tension or the like, the rate of crack growth depends on the alternating stress intensity factor (auk) at the tip of the crack which is proportional to the applied load range and which increases as the crack length increases for a given load range. If the crack is to be made to grow at a constant rate under a cyclically varying load, as is often required, the amplitude of the variation in applied load must be reduced as the crack grows. The difficulty is, however, that there has been no simple way of monitoring the growth of the crack either by direct observation or by electri cal methods and thereby deriving the neces sary feedback control signal for the actuator of the testing machine, and such existing systems as attempt this require a computer to process the resulting signals and are very expensive.

According to the present invention in one aspect there is provided a test apparatus for use in carrying out fatigue testing, crack for mation and growth studies, and the like on a specimen in which a crack is caused to grow from one face of the specimen towards another face thereof under a load of varying amplitude applied across the crack or a pre formed notch in the specimen, the apparatus including: means for deriving a signal denoting the load applied to the specimen; means including a strain gauge bonded to the face of the specimen towards which the crack is growing for producing a signal denot ing the strain induced in that face; means arranged to add the load signal and the induced strain signal in a pre-determined relationship (preferably in a ratio determined by the geometry and modulus of the speci men) and to produce an additive output sig nal; and means arranged to adjust the load applied to the specimen in response to the additive output signal in such a manner that the crack grows at a substantially constant rate over at least a proportion of the crack length or at a rate which changes in some predetermined manner.

Frequently the test procedure requires a cyclically varible load to be applied to the speciment. Means may be provided for deriv ing signals denoting the maximum and mini mum values of the load applied to the speci men, and means may be provided for producing a stress ratio signal denoting Pmin - R Pmax where Pmjn represents the minimum load value, Pmax represents the maximum load value and R represents the stress ratio, the stress ratio signal being supplied to the addition means which can adjust the load applied to the specimen to compensate for cumulative plastic strain.

The line carrying the back face strain signal and the line carrying the load signal may be electrically connected through resistors to the input of an operational amplifier, the ratio of values of the resistors being such that the load signal and the induced strain signal are added in a ratio determined by the geometry and modulus of the specimen.Where there is provision for cumulative plastic strain compensation, a maximum load signal line may be connected to an input of the first of three operational amplifiers connected in series, the output of the first operational amplifier being electrically connected to the input of a second operational amplifier together with a minimum load signal line, the gain of the first operational amplifier being set in conformity with the intended stress ratio, and the output from the second operational amplifier being connected to the input of a third operational amplifier of high DC gain and long response time whose output is arranged to supply a stress ratio signal to the addition means.

In another respect the invention relates to a test method for use in testing a specimen for fatigue, crack formation and growth or the like, which comprises: applying a load of varying amplitude across a crack in one face of the specimen or a preformed notch in one face of the specimen from which a crack grows, to cause the crack to grow towards another face thereof, deriving a signal denoting the load applied to the specimen, deriving from a strain gauge bonded to the face of the specimen towards which the crack is growing a signal denoting the strain induced in that face, adding the load signal and the induced strain signal in a predetermined relationship (preferably in a ratio determined by the geometry and modulus of the specimen), and adjusting the load applied to the specimen in response to the additive output signal in such a manner that the crack grows at a substantially constant rate over at least a proportion of the crack length or at a rate which changes in some predetermined manner.

Preferably the step of deriving the induced strain signal comprises deriving a signal denoting the compressive strain at the centre of the said face of the specimen. The test method may involve applying a cyclically variable load to the specimen. It may be desirable to compensate for cumulative plastic strain by deriving signals denoting the maximum and minimum values of the load applied to the specimen, producing a stress ratio signal denoting Pmjn - RPmax, where Pmjn and PmaX have the meanings previously given, and adjusting the load applied to the specimen in response to the stress ratio signal.

An embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings in which: Figure 1 is an oblique view of a compact tension specimen suitable for used in fatigue crack growth testing; Figure 2 is a block diagram of a hydraulically operated testing machine and associated control circuit for applying a cyclically variable load to the specimen shown in Fig. 1; Figure 3 is a diagram of a load and strain operational amplifier circuit forming part of the control circuit shown in Fig. 2; and Figure 4 is a diagram of a plastic strain compensation circuit which also forms part of the control circuit shown in Fig. 2.

In Fig. 1 a metal or other specimen 10 in the form of a generally rectangular block has a horizontal notch 1 3 in its front face and loading points such as through holes 11 and 1 2 above and below the notch to receive upper and lower fixing bolts or other clamping means of fixed and movable membrs of a closed loop servo-hydraulic testing machine.

The machine is arranged to apply an alternating load to the specimen which may be of a variety of types intended to simulate the service conditions in which articles made of the metal are intended to be used. For example, the specimen may be subjected to a sinusoidally changing load having a constant or variable mean value and having a constant or variable frequency. The amplitude of the load variation may be constant or may itself vary, for example by introducing dwell at peak load or by increasing the amplitude to simulate random overload conditions.Strain in the specimen is concentrated by the notch 1 3 and after a time a crack 14 grows from the notch 1 3 towards the back face 1 5. The distance between the loading points and tip of the crack is denoted by the letter a, and the distance between the loading points line and the back face of the specimen is denoted by the letter W, the dimensionless ratio a/w denoting the fractional crack length irrespective of the actual dimensions of the specimen.

A strain gauge 1 6 is bonded to the back face of the specimen and as pointed out by Deans etal(Strain, October 1977, pages 152-154) the strain perceived by the strain gauge increases linearly with the applied load. It has been found that this linear relationship applies over a range of different values of a/w although the proportionality constant varies depending on the particular value of a/w and also on the nature and geometry of the specimen.

The stress intensity factor produced by a load P at the crack tip is represented by a parameter K, and if a cyclically varying load is applied to the specimen there is produced a cyclic stress intensity factor AK. Both K and AK increase with the crack length. It has been found that the crack growth rate in a cyclically loaded notched specimen depends on the value of INK induced by the testing machine, and if the crack is to grow at a constant rate AK must be kept constant and the load P must be reduced in conformity with the in crease in crack length. A variety of optical, mechanical and electrical methods have been used to measure the crack length, but they all suffer from the disadvantages of inconvenience and expense.It has recently been found, however that the measured value of the strain using a strain gauge on the back face of the specimen is approximately linearly related to (a/w) as well as being proportional to K so that the information derivable from a strain gauge in this position is of great utility.

It has now been realized that as a fatigue crack grows in a specimen the valve of AK can be kept constant by supplying a first input signal significant of the applied load and a second input signal significant of the back face strain induced in the specimen in an appropriate ratio to a summing circuit which produces an output signal proportional to the linear sum of the first and second signals, and using the output signal to control the amplitude of the variable load which the testing machine applies to the specimen.

In Fig. 2 the specimen 10 is bolted at the hole 11 to one end of a connecting rod 20 whose other end is connected via a load cell 21 to a crosshead 22 or other fixed member.

The hole 1 2 is bolted to the upper end of a rod 23 which is extended or retracted by a hydraulic actuator 24 controlled by a servo valve 25 to apply the required cyclically varying load across the notch.

Strain gauges in the load cell 21 produce an indication of the load P applied to the specimen 10 and signals are passed through line 30 to a strain gauge amplifier 31 whose output is passed through line 32 to the input of an operational amplifier unit 33. Signals from the back face strain gauge 1 6 on the specimen 10 are supplied via line 17, together with signals from dummy gauges 16' via line 1 71 to provide temperature compensation as known in the strain gauge art, to a second strain gauge amplifier 34 whose output signal is also supplied via line 39 to the input of the operational amplifier unit 33 The output from the operational amplifier unit is fed via line 35 to a servo amplifier 36 where it is compared with a command signal significant of the required AK value fed in from an input device 37. The resulting control signal which denotes the difference between the measured and intended values of AK is fed via line 38 to the servo valve 25 of the hydraulic actuator.

The actuator 24 is normally intended to induce a fixed value of AK throughout the test but cumulative plastic strain causes the means value of K to drift as the test proceeds. In order to minimise or prevent this drift, the strain gauge amplifier 31 supplies signals to a maximum and minimum load sensor 40 (often fitted as a standard unit on servohydraulic testing machines) whose output signals Vmax and Vmjn denoting maximum and minimum load states are supplied through lines 41 and 42 to the input of a stress ratio discrepancy unit 43 whose output signal is fed via line 44 as a third input to the operational amplifier unit 33.

In Fig. 3 in which a representative circuit diagram is shown for the conventional amplifier unit, the back face strain signal from line 39, the load signal from line 32 and the stress ratio signal from line 44 are respec tively fed via variable resistor VR, and input resistors R, and R2 to the summing junction of an operational amplifier OA, whose gain is controlled by variable resistor VR2. The added and inverted output signal from operational amplifier OA, is supplied through input resis tor R3 to an operational amplifier OA2 whose gain is controlled by the ratio of resistor R4 to R3. The doubly inverted output signal from OA2 is supplied through line 35 to the servo amplifier 36.For a given K value a linear summation in the appropriate proportions of the applied load signal and the back face strain has been found to be constant to + 1% over a relatively wide range of values of a/w typically from 0.3 to 0.6. Appropriate proportions can be determined by measuring back face strain as a function of load and crack length and consulting the relevant ASTM compliance formula for the specimen. Once tables have been computed for any given specimen type (e.g. C.T.S.) it is necessary only to discover the relationship between back face strain and load for the initial notch or crack length for a particular material in order to determine the required proportions of the two signals. The proportions of the two sig nals can then be adjusted appropriately using VR,.If the summed signal is used as the feedback signal for e.g. a servo-hydraulic test ing machine, the load applied to the specimen will automatically reduce as the crack grows and a constant growth rate is obtained.

By way of example, the following is a table to enable VR, to be set correctly (for a particu lar situation), showing the strain signal for unit load signal at given values of a/W:- a/W P .30 .5253 .31 .5554 .32 .5870 .33 .6202 .34 .6550 .35 .6915 .36 .7298 .37 .7698 .38 .8117 .39 .8555 .40 .9014 .41 .9493 .42 .9995 .43 1.0521 .44 1.1071 .45 1.1649 .46 1.2255 .47 1.2891 .48 1.3562 .49 1.4269 .50 1.5015 .51 1.5804 .52 1.6640 .53 1.7528 .54 1.8473 .55 1.9497 .56 2.0554 .57 2.1704 .58 2.2935 .59 2.4257 .60 2.5677 .61 2.7204 .62 2.8850 .63 3.0625 .64 3.2540 .65 3.4609 .66 3.6844 .67 3.9261 .68 4.1875 .69 4.4703 .70 4.7762 An example of the way in which this table might by used is shown by the following steps. Firstly, the value of a/W for the specimen is determined, and the corresponding value of P is noted.The specimen is mounted in the test machine. With connections to elements 39 and 44 removed, a small load (insufficient to cause irreversible deformation) is applied to the specimen, and the voltage appearing at element 35, V,, is measured.

The connections to element 32 is then removed, and that to element 39 remade. VR, is adjusted until the measured voltage at element 35 equals P V,. Next connections to elements 32 and 44 are remade, and the signals will now be in the correct proportions for a crack growth test to be carried out.

In many test procedures the maximum and minimum loads P applied to the specimen are required to be in a constant ratio R = Vmin/Vmax as determined from the output signals from the load sensor 40. The signal from line 41 is supplied through an input resistor R5 (Fig. 4) to a first stage operational amplifier OA3 whose variable gain control resistor VR3 is adjusted so that the operational amplifier OA3 supplies a signal - RVmax to an input Resistor RB of a second operational amplifier OA4. Also supplied to the amplifying junction of amplifier-OA4 is a V,, signal from line 42 supplied through a second input resistor R7.Amplifier OA4 delivers an output signal proportional to Vm,n - RxVma,, and if the testing machine is already running at the required load levels, VR3 can be adjusted so that the output from OA4 into a third stage input resistor R8 is zero. Gain in amplifier OA4 can be adjusted by means of a gain control variable resistor VR4. The time constant of the load sensor 40 is long compared to the load cycle time, and in order for the control to be stable it is desirable that the stress ratio adjustment unit should have a time constant far exceeding that of the sensor 40.Accord ingly the signals from OA4 are fed through RB (which is typically of resistance 100 M52) into a field-effect transistor operational amplifier OA5 across which is connected a capacitor C (typically 10 yf capacity) in parallel with a normally open switch S,. Feedback through capacitor C defines an appropriate gain for OA5 and causes its output to drift slowly in response to changes in the input signal. For setting up the circuit S, is closed and VR3 is adjusted to give the required zero output from OA4 after which S, is opened.The signals from OA5 are fed to the third input resistor R2 of OA, and a substantially constant stress ratio R is thereby maintained despite alterations in the plastic strain induced on the back face of the specimen.

Various modifications can, of course, be made to the described embodiment without departing from the invention. For example the operational amplifier unit and the stress ratio discriminator unit can be replaced by a microprocessor which may be more versatile in the calculations it can perform, in which case the method may be extended to use test specimens of arbitrary shape where the back face strain is a more complex function of crack length. Although the invention has been described with reference to a servo-hydraulic testing machine, other types of testing machine can be used, for example an electromagnetic resonant testing machine such as an Amsler Vibrophore test machine.

Claims

1. Test apparatus for use in carrying outfatigue testing, crack formation and growth studies, and the like on a specimen in which a crack is caused to grow from one face of the specimen towards another face thereof under a load applied across the crack or across a preformed notch in the specimen, the apparatus comprising:: means for deriving a signal denoting the load applied to the specimen; means including strain gauge means bonded to the face of the specimen towards which the crack is growing for producing a signal denoting the strain induced in that face; means arranged to combine the load signal and the induced strain signal together in a predetermined relationship and to produce an output signal representative of the combined signals; means arranged to adjust the load applied to the specimen in response to the combined output signal in such a manner that the crack grows at a substantially constant rate over at least a proportion of the crack length or at a rate which changes in a predetermined manner.

2. Apparatus according to Claim 1 in which the combining means combines the load signal and the induced strain signal by adding the signals together in a predetermined weighted ratio.

3. Apparatus according to Claim 2 in which a line carrying the induced strain signal and a line carrying the load signal are electri cal ly connected through respective resistor means to an input of a operational amplifier, the ratio of values of the respective resistor means being such that the load signal and the induced strain being such that the load signal and the induced strain signal are added together in the said predetermined ratio.

4. Apparatus according to Claim 2 or 3 in which the said predetermined ratio is determined by the geometry and modulus of the specimen.

5. Apparatus according to any preceding claim including means for deriving signals denoting the maximum and minimum values of a cyclically varying load applied to the specimen, and means for producing a stress ratio signal denoting Pm - R PmaX where Pmjn in - max represents the minimum load value, and Pmax represents the maximum load value, and R represents the stress ratio, the stress ratio signal being supplied to the said combining means for adjusting the load applied to the specimen to compensate for cumulative plastic strain.

6. Apparatus according to Claim 5 in which a maximum load signal line is connected to an input of the first of three operational amplifiers connected in series, the output of the first operational amplifier being electrically connected to the input of a second operation & amplifier together with a minimum load signal line, the gain of the first operational amplifier being set in conformity with the intended stress ratio, and the output of the second operational amplifier being con nected to the input of a third operational amplifier of high DC gain and long response time whose output is arranged to supply a stress ratio signal to the combining means.

7. A test method for use in testing a specimen for fatigue, crack formation and growth, or the like comprising: applying a load across a crack in one face of the specimen or across a preformed notch in one face of the specimen from which a crack grows, to cause the crack to grow towards another face of the specimen, deriving a signal denoting the load applied to the specimen, deriving from strain gauge means bonded to the face of the specimen towards which the crack is growing a signal denoting the strain induced in that face, combining the load signal and the induced strain signal in a predetermined relationship to produce an output signal representative of the combined signals, and adjusting the load applied to the specimen in response to the combined output signal in such a manner that the crack grows at a substantially constant rate over at least a proportion of the crack length or at a rate which changes in a predetermined manner.

8. A method according to Claim 7 in which the step of combining the load signal and the induced strain signal comprises add ing the signals together in a predetermined weighted ratio.

9. A method according to Claim 8 in which the said combining step includes deter mining the said predetermined weighted ratio by reference to the geometry and modulus of the specimen.

.

10. A method according to Claim 7, 8 or 9 in which the step of deriving the induced strain signal comprises deriving a signal de noting the compressive strain at the centre of the said face of the specimen.

11. A method according to any of claims 7 to 10 in which the method includes the steps of applying a cyclically varying load to the specimen, and compensating for cumula tive plastic strain by deriving signals denoting the maximum and minimum values of the load applied to the specimen, producing a stress ratio signal denoting Pmin - R PmaX, where Pmm represents the minimum load va lue, Pmax represents the maximum load value, and R represents the stess ratio, and adjusting the load applied to the specimen in response to the stress ratio signal.

1 2. Test apparatus for use in carrying out fatigue testing, crack formation and growth studies, or the like, substantially as hereinbe fore described with reference to the accompa nying drawings.

1 3. A test method for use in testing a specimen for fatigue, crack formation and growth, or the like, substantially as hereinbe fore described with reference to the accompanying drawings.