GB2531542A - Methods and systems for torque tool calibration - Google Patents

Abstract

A torque calibration apparatus 10 such as a test rig includes a shaft 14 with one end arranged to be driven by a torque tool 12. An angle monitoring device (24, Figure 5) is arranged to detect the degree of rotation of the shaft. A brake 16 is arranged to act on the shaft. A controller (30, Figure 5) is arranged to control the application of the brake based on the detected degree of rotation of the shaft. The brake may be a compressed air brake. A static torque sensor may be arranged to detect the torque applied to the shaft.

Description

Methods and systems for torque tool calibration The invention relates to methods and systems for calibrating torque tools.

In many applications, fasteners such as nuts and bolts must be tightened to a specific tension. In practice, the tension of such fasteners is difficult to measure directly, but a fastener may be tightened to an adequate tension by applying a specified torque.

Torque tools, including torque wrenches and powered torque tools are used for applying a specific torque to such fasteners. Traditional mechanical torque wrenches comprise an internal mechanism (e.g. a lever mechanism) that causes the wrench to signal mechanically when the set torque has been reached.

More recently electronic torque wrenches have been produced. These use a strain gauge to measure the torque that is being applied during use. A digital display allows a user to set the desired nominal torque and a visible or audible signal is given when this is reached. Additionally, in the case of motor driven torque tools, the motor can be stopped when the desired nominal torque is reached.

In many applications, it is necessary from a safety point of view to apply an accurate torque and it is therefore important that torque wrenches are accurately adjusted and the results recorded by means of a calibration with low uncertainty of 25 measurement.

Torque wrenches are usually calibrated by placing them horizontally on a rig which applies an increasing torque, which is measured accurately by a transducer. The wrench is set to a particular nominal setting and the torque at which the wrench signals is then measured and compared with the nominal setting. If there is too great a disagreement (usually expressed as a percentage) the wrench can be adjusted. Multiple measurements of the torque at signal may be taken to ensure that the torque at signal is within a desired tolerance. -2 -

Rotational, motor driven torque tools may be calibrated in a similar manner. The tool is set to drive a test shaft while a resistive torque is applied to the shaft by a brake. When the tool indicates that the correct torque has been reached, the tool's measurement is compared with the resistive torque applied by the brake. If there is too great a disagreement, the tool can be adjusted. Again, multiple measurements are usually taken to ensure that the tool is within a desired tolerance.

One problem with existing rigs and calibration methods is that the applied torque (e.g. applied by the brake) is ramped up essentially linearly with time. However that does not provide a good simulation of the torque profile of an actual joint. In a typical joint, as the joint is tightened and the resistive torque increases, the tool slows down and therefore the rate of change of torque decreases. As the rig continues to simulate the joint by ramping up torque linearly with time, the simulation is not accurate and this can lead to errors in the calibration.

According to a first aspect, the invention provides a torque calibration apparatus comprising: a shaft with one end arranged to be driven by a torque tool; an angle monitoring device arranged to detect the degree of rotation of the shaft; a brake arranged to act on the shaft; and a controller arranged to control the application of the brake based on the detected degree of rotation of the shaft.

Controlling the brake based on the detected degree of rotation allows the apparatus to simulate the torque of a real joint more accurately by setting the resistive torque applied by the brake to be almost exactly the torque that would be expected at a certain angle of rotation of the simulated joint. This technique takes into account any variations in speed of the torque tool and ensures that the torque profile experienced by the tool during calibration is as close as possible to the torque profile that would be experienced on a real joint.

Traditional nut and bolt based calibration rigs rely on tightening threaded components, e.g. tightening an internally threaded nut on an externally threaded bolt. Such systems need to undo the joint by winding the nut back along the bolt between each test. This process is both time and energy consuming. A particular advantage to using a brake to apply the torque is that the brake can be released almost instantaneously, thus resetting the calibration rig for the next test. This -3 -allows rapid repeat testing, i.e. life cycle testing for the torque tool, allowing a large number of tests to be performed in a reduced time. The time taken to reset the rig between tests becomes very important during extended life cycle testing where a large number of tests, e.g. tens of thousands, hundreds of thousands or even a million repetitions may be performed.

Another advantage to using the brake and controller is that any joint can be simulated by varying the torque profile that is applied by the brake to the shaft. For example, the rate of increase of torque with angle can be varied to simulate nuts/bolts with different thread pitches. Also the rate of increase can be varied so as to simulate different materials and the different friction associated with different rubbing surfaces. The relationships simulated by the brake and controller may be linear in torque versus angle (as is the case for many joints) or they may be nonlinear. This technique is very versatile. The controller can also adjust the torque profile being applied to compensate for changes in environmental factors such as temperature and humidity which can affect the calibration.

The brake may be a compressed air brake in which the input air pressure determines the output braking strength. This arrangement is advantageous as the air pressure applied to the brake has a well-defined relationship to the torque applied by the brake to the shaft. Also, the air pressure may be applied by an electronic air regulator that has a well-defined relationship between its input voltage and its output air pressure. This means that the applied voltage to the air regulator can be used to determine the input air pressure accurately and that in turn provides an accurate determination of the torque being applied by the brake.

Accordingly, the torque calibration apparatus preferably further comprises an electronic air regulator for supplying compressed air to the brake, and the controller may be arranged to calculate the torque applied to the shaft by the brake based on an input signal supplied to the electronic air regulator.

The torque calibration apparatus may further comprise a torque sensor arranged to detect the torque that has been applied to the shaft. The torque sensor may be a static torque transducer and the brake may be situated on the shaft between the torque sensor and the tool. Alternatively, the torque sensor may be a rotary torque -4 -transducer and the torque sensor may be situated on the shaft between the brake and the tool. The rotary torque sensor is preferred in some implementations because it takes better account of inertia, e.g. in the case of fast application of the brake. However, in most implementations the static torque sensor is preferred due to its lower cost and greater reliability.

Although the torque applied by the brake can be detected (either directly on the shaft or indirectly, e.g. based on the applied air pressure), at low applied torques (at low applied air pressure), the brake does not engage with the shaft. This is due to the need to overcome various frictions such as within the sealing components before brake pad or disc movement occurs. The result of this is that as the controller ramps up the signal from zero torque, the actual applied torque by the brake remains zero up to a certain point and then engages suddenly with a step change to applying a positive torque value to the shaft. Thereafter, the torque applied by the brake follows a predictable and smooth relationship. The point at which the brake engages varies and is different with every application of the brake which means it cannot simply be factored in in the calculations. When simulating large torque joints this step change in torque at low values has a negligible effect on the overall calibration. However when trying to calibrate for low torque values it becomes necessary to compensate for it. As the point at which the brake engages varies with every application it is preferred to measure or determine that point on every calibration, i.e. on every application of the brake. Therefore preferably the controller is arranged to continually monitor torque on the shaft while the controller instructs the brake to ramp up its braking force and the controller is preferably arranged to detect the point in time at which torque is first detected. Preferably this point (the torque value and the point in time at which it is reached) are stored so that they can be used later if desired.

Until the brake engages, the torque tool is not experiencing any resistive torque as would be expected from a real joint. Therefore the angle through which it has rotated up to the point at which the brake engages is preferably ignored. The controller may thus be arranged to monitor the degree of rotation of the shaft from the point in time at which torque is first detected on the shaft. This technique monitors the angular duration for which the tool experiences resistive torque. -5 -

When the brake engages, it jumps in at a certain torque level. In a real joint that torque level would have been reached after a predictable angular duration (i.e. after rotation through a predictable angle). The controller may calculate that angle from the detected initial torque value and control the brake so as to follow the correct torque profile from that point on. This provides the correct torque profile to the tool under test from the point at which the brake engages.

It will be appreciated that the controller may reset the angle monitor to zero and count up to a lower total angular duration, or it may set the angle monitor to the angle that corresponds to the initial torque value on the torque versus angle profile and proceed to count up to the total expected angular duration. Either way, although the torque tool being calibrated ends up experiencing the right final torque, the tool does not experience torque throughout the full turning duration of the joint being simulated. This can result in erroneous calibration.

Preferably the torque tool is allowed to stabilise after the sudden onset of braking so that any transient oscillations introduced by that sudden onset are allowed to die down before the brake ramps up the torque. Preferably therefore the controller is arranged such that after first detecting torque on the shaft, the controller instructs the brake to hold its braking force constant until the shaft has rotated through a determined angle. This ensures that the tool under test is operating stably against resistive torque when the controller instructs the brake to increase the applied torque. This provides more consistent calibration results.

Preferably the torque tool being calibrated should experience resistive torque throughout the whole angular (turning) duration of the joint being simulated. Therefore preferably the determined angle through which the shaft rotates after torque is first detected is an angle that a joint would normally rotate through to reach the torque level being applied by the brake. This technique ensures that the tool has applied torque against a resistive load for the full angular duration of the simulated joint and that it ends up at the correct simulated torque at the end of that angular duration. The tool under test has then experienced as close as possible the correct duration and the correct torque profile for the simulated joint. -6 -

The simulated joint may of course be varied for different thread pitches, different materials, etc. as discussed above. The determined angle through which the tool rotates (before ramping up of the brake begins after torque is initially detected) will vary depending upon the joint being simulated.

According to a second aspect, the invention provides a method of calibrating a torque tool, comprising: driving one end of a shaft with the torque tool; monitoring the degree of rotation of the shaft; and applying a brake to the shaft, wherein a controller controls the application of the brake based on the detected degree of rotation of the shaft.

The preferred features described above in relation to the apparatus also apply equally to the method of calibrating. Therefore preferably the brake is a compressed air brake in which the input air pressure determines the output braking strength. The method may further comprise: determining the torque applied to the shaft by the brake based on an input signal supplied to an air regulator that provides compressed air to the brake. The method may further comprise: detecting the torque that has been applied to the shaft. The torque may be detected by a static torque transducer and the brake may be situated on the shaft between the torque transducer and the torque tool. The torque may be detected by a rotary torque transducer and the torque transducer may be situated on the shaft between the brake and the torque tool. The controller may continually monitor torque on the shaft while the controller instructs the brake to ramp up its braking force and the controller may detect the point in time at which torque is first detected on the shaft.

The controller may monitor the degree of rotation of the shaft from the point in time at which torque is first detected on the shaft. After the controller first detects torque on the shaft, the controller may instruct the brake to hold its braking force constant until the shaft has rotated through a determined angle. The determined angle may be an angle that a joint would normally rotate through to reach the torque level being applied by the brake.

Certain preferred embodiments of the invention will now be described, by way of example only, and with reference to the accompanying drawings in which: Fig. 1 shows an embodiment of a test rig for torque tool calibration; -7 -Fig. 2 shows another embodiment; Fig. 3 shows a graph of torque versus brake input signal; Fig. 4 shows a graph of torque versus angle for one calibration method; and Fig. 5 schematically illustrates a test rig.

Fig. 1 shows a test rig 10 for testing a torque tool 12. The torque tool 12 is coupled to one end of an input shaft 14 of the test rig 10. A brake unit 16 is configured to apply a braking torque to the input shaft 14 so as to counter any torque applied by the tool 12. The brake 16 may apply torque via a disc brake mechanism (not shown) or any other suitable braking mechanism. A static torque transducer 18 is located behind the brake unit 16 (i.e. the brake unit 16 is positioned on the shaft between the static transducer 18 and the torque tool 12). The static torque transducer 18 detects the torque that the brake 16 is applying to the input shaft 14. A rotary encoder (angle measurer) is provided on the end of the input shaft 14, just behind the static torque transducer 18. However, it will be appreciated that the rotary encoder could be placed anywhere on the shaft 14.

An alternative test rig 20 is shown in Fig. 2. This test rig is broadly the same as that of Fig. 1 and like reference numerals indicate like features. The difference is that a rotary torque transducer 22 is used to measure the torque being applied to the input shaft 14 by brake unit 16. The rotary torque transducer 22 is positioned between the brake unit 16 and the torque tool 12 that is under test. A rotary encoder is again positioned at the end of the shaft 14, just behind the brake unit 16.

As discussed previously, the rotary torque transducer is preferred in some implementations because it takes better account of inertia, e.g. in cases where the brake is applied very rapidly. However, in the majority of implementations where this is not a major consideration a static torque sensor is likely to be used due to its lower cost and greater reliability.

In use, the test rig (10 or 20) is operated by starting the torque tool 12 so that it starts rotating the input shaft 14. A controller is used to start the tool 12 and control the brake 16 and also to receive data from the rotary encoder and the torque transducer (18 or 22). The controller instructs the brake 16 to ramp up the torque in accordance with a pre-defined torque profile based on the angle that the input shaft -8 - 14 has rotated through. Therefore the torque applied by the brake 16 is a function of the angle detected by the rotary encoder. The torque profile simulates a particular joint by increasing the torque over a predefined angle up to a predefined maximum torque. As the brake 16 is controlled based on the sensed angle, if the torque tool 12 slows down towards the end of the joint (as the resistive torque becomes high), this slowing will be compensated by a slowing in the rate of increase of the torque applied by the brake 16.

When the torque tool 12 indicates that it has reached the desired torque level (preset by the rig 10 or 20 according to the joint that is being simulated), the tool stops and the torque measured by the tool 12 is compared to the torque being applied by the brake 16 (and measured by the torque transducer 18 or 22). If there is a discrepancy beyond the acceptable tolerance, the tool 12 can be adjusted.

Fig. 3 shows a graph of input signal versus output torque for a compressed air brake that can be used for the brake unit 16 in the systems of Figs. 1 and 2.

The graph in Fig. 3 shows that until a minimum operating signal is applied, the brake 16 does not respond. This is due to forces that must be overcome in the brake system 16. For example, the air cylinder that supplies compressed air to the brake has gas seals to prevent loss of air. The friction of these seals needs to be overcome before the brake will move. Similarly, in order to prevent the brake from sticking in the braking position when no force is applied, the brake is biased towards the open (non-braking or freewheeling) position by coil springs. The bias force of these springs needs to be overcome before the brake will move.

At the point at which the minimum operating signal is reached (in this case 10 units), the brake 16 performs a step change to some output level (in this case Nm). This means that the torque tool 12 does not experience a steady ramping torque between 0 Nm and 80 Nm. This step change does not matter in the calibration at high torque levels (e.g. greater than 500 or 1000 Nm final torque) as the torque tool 12 experiences the correct resistive torque throughout the majority of the test process. However for calibration at lower torques (e.g. less than 500 Nm), this step change can be problematic as the torque tool 12 misses a significant portion of the ideal (realistic) torque profile. -9 -

Having appreciated the step change response at the lowest output condition, it is important also to note that the step change is variable. For example the actual torque step could be anywhere in a range from 60-110 Nm with each discrete operation of the brake. In other words the height of this torque step (in Nm) varies with every test joint performed. These points should ideally be addressed for the most accurate results to be obtained.

Firstly it is desirable to address the issue of achieving accurate durations (i.e. the torque tool 12 rotating the tests joint through the correct angle) given the issue of a step change response of the brake 16.

To yield the correct angle, the method illustrated in Fig. 4 is used. This method accounts for the inherent step change response by pausing the increasing pressure of the brake 16 until the actual torque profile intersects with the "ideal" torque profile. In Fig. 4, the dotted line 41 shows the ideal torque profile, the solid line 42 shows the actual torque profile. This approach ensures that the final angular duration is accurate (i.e. that the torque tool 12 drives against resistance for the correct angular duration) and also that the later stages of joint simulation follow the ideal profile. The resulting tool calibration is as close to ideal as possible.

Secondly, it is desirable to address the varying torque conditions when step change occurs. In order to tackle the issue of when to start measuring the joint it is necessary to know when torque is first generated on the input shaft 14 by the brake 16 and how much torque is applied at that point. Thus the system needs to actively adapt to the natural variations in the initial torque conditions provided by the brake 16. To do that, a torque transducer (18 or 22) is provided. By setting up internal detection thresholds it is possible to provide a torque detection signal. The controller monitors this torque detection signal to establish when torque started. By referencing the corresponding signal (e.g. from the controller) being applied to the brake 16 combined with the preprogramed brake characteristics it is possible to resolve the effective torque being applied at the initial torque conditions. The control system can then calculate the angle at which the simulated joint would have reached that initial torque value and can therefore establish how much angular dwell should be allowed before commencing with a torque profile that matches the -10 -ideal. During the angular dwell the torque applied by the brake 16 remains constant at the initial input value so that the tool 12 experiences a resistance throughout that period. Only once the torque tool 12 has rotated against resistance for the correct angle is the brake 16 ramped up by the controller to increase the torque in accordance with the simulated joint profile.

Fig. 5 schematically illustrates a test rig 20. A controller 30 is shown. The controller 30 receives data from torque transducer 22 and from rotary encoder 24 which sense the torque and angular position respectively of input shaft 14.

Controller 30 also has access to memory 31 (note that although this is illustrated separately for clarity, the memory 31 may be located on the controller 30 itself). Memory 31 includes a joint profile lookup table 32 which defines the torque that should be applied at any given angle so as to simulate a desired joint. Memory 31 also includes a brake characteristics profile 33 which stores the brake's expected output torque value for any given input signal value.

It will be appreciated that the schematic of Fig. 5 can readily be adapted for the system 10 of Fig 1.

The performance characteristic of each brake 16 is measured during commissioning of the test rig 10 or 20. This includes generating a profile of the applied input signal versus the applied output torque. This data may be stored in the form of a function or in the form of a lookup table 33 that can be referred to by the controller 30 based on the detected level of input signal. The system 10 or 20 provides the facility to store the characteristic in the controller 30. This means that for any given torque level the controller 30 has a reference to the brake 16 performance at that torque. This allows to account for any subtle non linearity in the performance and also account for changes in performance due to wear / temperature etc. The system 10 or 20 may also provide a run-in mode. This enables the tool 12 to be run for a pre-set time under static load conditions. This enables the tool 12 to be settled prior to calibration and produces faster and more consistent calibrations.

The controller may also include cyclic modes of operation. This enables repetitive testing to be performed (i.e. life-cycle testing of a torque tool 12). The joint control is as described above, but the following additional parameters are used: Target number of cycles (determines when the rig 10, 20 will shut off due to completion of the desired test cycles).

Tool delay (used for electric torque tools 12 to prevent overloading. This provides a delay to allow a chance for cooling of the tool 12 between cycles.

The following functions may also be provided: * Data log of the joints performed during testing, i.e. an output of angle versus measured torque on the brake and optionally applied torque as measured by the tool 12.

* Tool failure detection system to detect if the tool 12 has broken or underperformed. The test may be aborted in these circumstances.

* Run down speed measurement system to measure the rotation speed of the tool while it is running at maximum speed with no brake applied (e.g. to compare against expected values for known tools or to measure the speed of an unknown tool) * Tool lubrication system control to provide efficient tool lubrication during testing, especially extended lifecycle testing.

Claims (20)

1. -12 -Claims A torque calibration apparatus comprising: a shaft with one end arranged to be driven by a torque tool; an angle monitoring device arranged to detect the degree of rotation of the shaft; a brake arranged to act on the shaft; and a controller arranged to control the application of the brake based on the detected degree of rotation of the shaft.
2. 2. A torque calibration apparatus as claimed in claim 1, wherein the brake is a compressed air brake in which the input air pressure determines the output braking strength.
3. 3. A torque calibration apparatus as claimed in claim 2, further comprising an electronic air regulator for supplying compressed air to the brake, and wherein the controller is arranged to calculate the torque applied to the shaft by the brake based on an input signal supplied to the electronic air regulator.
4. 4. A torque calibration apparatus as claimed in any preceding claim, further comprising a torque sensor arranged to detect the torque that has been applied to the shaft.
5. 5. A torque calibration apparatus as claimed in claim 4, wherein the torque sensor is a static torque transducer and wherein the brake is situated on the shaft between the torque sensor and the tool.
6. 6. A torque calibration apparatus as claimed in claim 4, wherein the torque sensor is a rotary torque transducer and wherein the torque sensor is situated on the shaft between the brake and the tool.
7. 7. A torque calibration apparatus as claimed in any preceding claim, wherein the controller is arranged to continually monitor torque on the shaft while the controller instructs the brake to ramp up its braking force and wherein the controller is arranged to detect the point in time at which torque is first detected.

   -13 -
8. 8. A torque calibration apparatus as claimed in claim 7, wherein the controller is arranged to monitor the degree of rotation of the shaft from the point in time at which torque is first detected on the shaft.
9. 9. A torque calibration apparatus as claimed in claim 7 or 8, wherein the controller is arranged such that after first detecting torque on the shaft, the controller instructs the brake to hold its braking force constant until the shaft has rotated through a determined angle.
10. 10. A torque calibration apparatus as claimed in claim 9, wherein the determined angle is an angle that a joint would normally rotate through to reach the torque level being applied by the brake.
11. 11. A method of calibrating a torque tool, comprising: driving one end of a shaft with the torque tool; monitoring the degree of rotation of the shaft; and applying a brake to the shaft, wherein a controller controls the application of the brake based on the detected degree of rotation of the shaft.
12. 12. A method as claimed in claim 11, wherein the brake is a compressed air brake in which the input air pressure determines the output braking strength.
13. 13. A method as claimed in claim 12, further comprising: determining the torque applied to the shaft by the brake based on an input signal supplied to an air regulator that provides compressed air to the brake.
14. 14. A method as claimed in claim 11, 12 or 13, further comprising: detecting the torque that has been applied to the shaft.
15. 15. A method as claimed in claim 14, wherein the torque is detected by a static torque transducer and wherein the brake is situated on the shaft between the torque transducer and the torque tool.

    -14 -
16. 16. A method as claimed in claim 14, wherein the torque is detected by a rotary torque transducer and wherein the torque transducer is situated on the shaft between the brake and the torque tool.
17. 17. A method as claimed in any of claims 11 to 16, wherein the controller continually monitors torque on the shaft while the controller instructs the brake to ramp up its braking force and wherein the controller detects the point in time at which torque is first detected on the shaft.
18. 18. A method as claimed in claim 17, wherein the controller monitors the degree of rotation of the shaft from the point in time at which torque is first detected on the shaft.
19. 19. A method as claimed in claim 17 or 18, wherein after the controller first detects torque on the shaft, the controller instructs the brake to hold its braking force constant until the shaft has rotated through a determined angle.
20. 20. A method as claimed in claim 19, wherein the determined angle is an angle that a joint would normally rotate through to reach the torque level being applied by the brake.