Patent Grant
9702417
This invention relates generally to calibration of a hydraulically actuated clutch of a CVT, and more particularly to calibration by monitoring pressure changes between a hydrostatic pump and motor of the transmission to determine initial engagement of the clutch.
The disclosure of U.S. Provisional Application No. 61/527,523, filed Aug. 25, 2011, is hereby incorporated herein in its entirety by reference.
Transmissions with hydraulically operated clutches need to consistently operate such that the timing of clutch engagement can be controlled precisely.
This can be difficult, since the clutches control multiple plates, and there is variation between springs and other component tolerances, such that the pressure needed to move the piston out to the point at which the plates touch and the clutch starts to transmit torque can vary quite a bit.
Most transmissions up to now in tractors have been power shift. Continuously variable transmissions or CVTs are becoming more popular and have significantly different architectures, often using a hydrostatic pump and motor. The pressure between the pump and motor is a direct indication of the torque through the hydrostatic unit, and in turn an indication of the torque through the clutch. In some designs, the hydraulic pump and motor can be in separate housings, connected with tubes or hoses, but they may be in the same housing (a hydrostatic unit or HSU), which typically increases efficiency.
Typically, known electronically controlled transmissions with such clutches use a calibration routine. Clutches are engaged to create driveline torque to work against the parking brake. A searching method is used to determine the currents needed to provide the pressure to just start to engage the clutch. This point is usually determined by when the engine speed reduces or engine torque increases, indicating that the clutch is transmitting torque. However, a shortcoming with this approach is that engine speed can vary as a result of other reasons, such as when other loads on the engine change, such as hydraulic or PTO loads, and various drags on the engine vary and such. Also, tractor engines have a governor that is quite complex, and may change engine speeds due to complex algorithms to manage emissions, efficiency and other factors. As a result, how much the engine speed dips cannot always be a direct indication of the level of torque on a clutch, which can result in different calibration values. Clutch calibration done using engine torque has similar shortcomings. A torque sensor in connection with a clutch has also been used for calibration, but such torque sensors are expensive and add complexity. Optionally, adding another torque sensor might be one costly option. Also, engine torque could be estimated from fuel, but this would be a slower signal, and again subject to other loads on the engine.
Thus, what is sought is a manner of hydraulic clutch calibration that does not rely on engine speed, torque or separate torque sensors, and overcomes one or more of the shortcomings set forth above.
What is disclosed is a manner of hydraulic clutch calibration that does not rely on engine speed or engine torque, and overcomes one or more of the shortcomings set forth above.
According to a preferred aspect of the invention, the clutch under calibration is held filled in a manner similar to how the clutch is used during a shift. The pressure in the clutch is commanded using a control signal command, e.g., in the form of an electrical current value, translated to a pressure in a pressure reducing valve. This valve is then connected to a rotating clutch which is pressure applied, spring released. It is held at this test current value, and a determination made as to whether the clutch current is too high or too low. This determination is made by checking the pressure in the hydraulic connection between the pump and the motor, or within the HSU. The test is repeated with different control signal values, e.g., test currents, either high or lower, depending on the result. Typically the algorithm starts with a default control signal value, and it is low, so as to not aggressively engage the clutch by accident. Once the pressure is sensed, the clutch is commanded to empty, so the engine does not lug down too much. This is also done to reduce the chance of driving through the parking brake. A separate check is done on the output speed, so the calibration is stopped if the vehicle moves. A baseline pressure is recorded before calibration or before each time the clutch is engaged, since the pressure in the HSU will not be zero when the clutch is not engaged, due to drag, but also the charge pressure in the HSU that keeps the HSU cool.
According to another preferred aspect of the invention, clutch engagement is detected using the difference of the pressure transducers in the HSU as an indication of driveline torque. Rather than actually calculate the clutch pressure or driveline torque, the clutch calibration looks at the HSU pressure signals directly, since at the ideal kiss point, both the driveline torque and clutch torque should be zero. It is expected that there will be noise on the pressure transducers, or other issues with pressure fluctuations. So that this is not a problem or does not result in a false detection, the calibration preferably looks for a change in the difference between the pressures, rather than a set level of the difference.
Use of HSU pressure for clutch calibration is advantageous as it does not depend on other factors, and it is a more direct measure of torque. As noted above, engine speed can vary as other loads on the engine change, such as hydraulic or PTO loads, various drags on the engine that may vary and such. Also, the engine governor is quite complex, and may change engine speeds due to complex algorithms to manage emissions, efficiency and other factors. How much the engine dips cannot always be a direct indication of the level of torque, which can result in different calibration values
Additional sensors may be needed on the HSU (typically two are needed, since the direction of torque through the HSU may change depending on the range), but these are typically present for other reasons. If a separate torque sensor already exists on a vehicle, it may be advantageous to use it, especially if it is used on a powershift (non-CVT version of a transmission).
As alternative embodiments of the invention, the method works with various CVT architectures, any time such clutches are used and the HSU indicates the torque through them. As another alternative, rather than using separate pressure sensors or transducers in the fluid lines, a single differential pressure transducer configured to determine a pressure differential between two ports could be used.
Referring now to the drawings, apparatus for use with the method of the invention and illustrations of aspects of the invention are shown. In
Referring to
Hydrostatic power unit 12 of transmission 10 includes a fluid pump 16 coupled by fluid conduits 17 in a closed loop to a fluid motor 18. Motor 18 is coupled to power source 4 via an input gear N6 and having an output gear N10. The power to the hydrostatic power unit 12 is provided by a driven gear N4 mounted on the forward shaft and engaged with gear N6. Output gear N10 is connected to ring gear NR of planetary power unit 30 via gears N11 and N12.
Machine 1 includes a processor based controller 100 in connection with an input device 102 located preferably in operator cab 104 of machine 1, via a suitable communications path 108, to adjust the angle of a swash plate of pump 16, denoted by the term “SPA”, for controlling operation of the transmission. As an exemplary embodiment, pump 16 can be an electronically controlled variable displacement hydraulic pump of well known construction.
Planetary power unit 30 includes a primary sun gear NS1 on a planetary input shaft 32 connectable with power source 4 via a forward directional clutch 54 or a reverse directional clutch 52. Power unit 30 is selectively coupled to the load L, coupled to the hydrostatic power unit 12 and selectively coupled to the power source 4, under automatic control of controller 100. For connection to the load L, the hydro-mechanical transmission 10 includes an output shaft 60 coupled to the load L which carries an input gear N18 engaged with an output gear N17 on a range 1/2 shaft of range gear set 58, and a gear N22 engaged with a gear N19 on a range 3/4 shaft. The range 1/2 shaft can be coupled to planetary power unit 30 via automatic operation of range selectors or clutches R1 and R2 for power flow through gears N13 and N14, or N15 and N16, respectively. The range 3/4 shaft can be coupled to unit 30 via range selectors or clutches R3 and R4 for power flow via gears N13 and N20, or N15 and N21. Range 1/2 shaft and range 3/4 shaft can also be simultaneously coupled to power unit 30, to provide dual power flow.
The control of the various clutches will be automatically controlled by controller 100, using actuators 106 connected to controller 100 via suitable conductive paths 108. Transmission 10 also includes appropriate sensors, including pressure sensors 110 for sensing pressure conditions in conduits 17 connecting pump 16 and motor 18, and speed sensors 112 for sensing speeds of load shaft 60, all connected to controller 100 via conductive paths 108. Controller 100 is also connected to engine 4 for receiving speed and other information therefrom.
In operation, the continuously variable hydromechanical transmission 10 can be operated to have a combined hydrostatic and mechanical power flow by engaging the reverse clutch 52 to power planetary power unit 30 via gears N1, N3, N5 and N7, or engaging forward clutch 54 to power it via gears N1, N8, and N2. It is also possible to operate transmission 10 for a pure hydrostatic power flow by disengaging both clutches 52 and 54, and engaging two range clutches. Typically, the R1 and R2 range clutches, and the R1 and R4 clutches.
As a result, with transmission 10, there is no selection for a work range or road range per se. However, the transmission provides a seamless transition between ranges to provide work/road configurations as desired. Speed change from zero to maximum speed is achieved in a smooth and continuous manner by changing the SPA of the pump 16 under control of controller 100. For each speed range, substantially the full range of travel of the swash plate is used. That is, the swash plate will be at one end of the range its travel for minimum speed within the range, it will be at the other end for maximum speed in that range, and the zero tilt or neutral position of the swash plate will be an intermediate position for the speed range, not the zero speed position as it is for some other transmissions. This presents a challenge for execution of some transmission commands that require a change of state wherein the swash plate will have to be tilted to a position significantly different from the present position. And, the direction of power flow through the hydrostatic power unit will often reverse during these range changes.
Transmission 10 includes a parking brake 114 in connection with load shaft 60, which is utilized according to the invention for enabling shuttle shifts and other operations. Parking brake 114 is connected to controller 100 via a suitable conductive path 108 for automatic operative control thereby, including to proportionally or gradually engage, and release or disengage, under certain conditions. To achieve this latter capability, as a non-limiting example, parking brake 114 can be controlled using a proportional pressure reducing valve operated by an electrical signal from controller 100. For operation when machine 1 is not operating, parking brake 114 can be engaged by a spring or other biasing element or elements, or by mechanical means.
Note the two paths through the transmission, one through the hydrostatic unit and one through the directional clutches. The planetary unit is driven by both paths, effectively allowing the hydrostatic unit to vary the ground speed in a continuous range. In ranges 1 and 3, increasing the swash plate angle results in a higher TRR. In ranges, 2 and 4, this is reversed, and decreasing the swash plate angle results in higher TRR.
Clutch Engagement
An example of a clutch engagement is illustrated in
There is an optional last stage, called “Engaged” where the pressure is maintained at the Engaged Pressure. Although this could be viewed as not part of the clutch modulation, it is included as a state to represent times when this particular clutch may be engaged, but another clutch may not have finished the shift. The shift will generally be considered over when the pressure modulation is done in both the engaging the on-coming and disengaging the off-going clutches.
Clutch Disengagement
An example of clutch disengagement (dumping) is shown in
The first phase of the dumping is simply a delay in dumping. This is often needed, since a clutch will be filling at the same time, and the ramping needs to be delayed so engaging and disengaging clutches can be timed such that the shift is smooth.
The second phase is the dump ramp. The ramp is controlled by a look up table, so that any shape may be used. The ramp is started at the “Engaged Pressure” and ended at the zero pressure. The total time of the ramp is “Dump Ramp Time”.
At the end of the second phase, the clutch still may be partially engaged, due to errors in the pressure control and tolerance. After the pressure is ramped to zero, a full negative pressure is commanded to fully empty the clutch. Zero actual pressure is also maintained the whole while the clutch is disengaged, to avoid any chance of even slight engagement or drag. This is represented by a full negative pressure, since zero pressure is defined as the kiss pressure.
Conversion from Pressure to Current
The pressures up to this point have been in terms of the “kiss point” of the clutch, such that zero pressure represents the pressure needed to overcome the spring force and for the plates to begin to make contact. Pressure higher than zero will engage the clutch.
The print for valve includes a graph of the pressure vs. control signal value (current in amps) for a typical transmission. From this graph, the points in Table 1 were estimated (this is example data and may not be the exact values used). Note that the nominal electrical current value of a control signal required to command any pressure is subtracted off (this is a value of 0.130 A).
Application of Calibration Data
The calibration data is represented as an offset on the clutch current. This offset is the current required to produce the pressure in the clutch that just overcomes the clutch springs and plates just begin to touch.
Note that the calibration will account for variations such as
The calibration algorithm will result in a current offset that accounts for this variation. The offset will be used until the calibration is performed again.
Terms
As noted above, a representative vehicle with which the invention can be used is illustrated in
Referring more particularly to
According to the method of the invention, the vehicle is held stationary, preferably by application of parking brake 114. The clutch under calibration is filled in a manner similar to or the same as how the clutch is used during a shift. The pressure in the clutch is commanded using a control signal which is a current command having a value, translated to a pressure in pressure reducing valve 130, as explained above. The valve 130 is held at this test current (representative of the test pressure), and a determination made as to whether the clutch current (test pressure) is too high or too low.
This determination is made by checking the pressure in the hydraulic connection, e.g., lines 17 between the pump and the motor of power unit 12, using sensors 110, or a differential pressure transducer in connection with ports providing the required pressure information. The test is repeated in a test loop with one or more different test currents, either high or lower, depending on the result of the previous test pressure. In particular, if a previous test pressure failed to engage the clutch the next test pressure used will be incrementally higher. Conversely, if the previous test pressure more than just initially engaged the clutch, the next test pressure will be lower. Typically the algorithm starts with a default value, and it is a low pressure, so as to not aggressively engage the clutch by accident. As a representative definition, aggressively is defined generally as sufficient to cause movement of the vehicle were the parking brake not engaged.
Once a pressure or pressure condition in the power unit 12 indicative of initial engagement of the tested clutch is sensed, the clutch is commanded to empty, so the engine does not lug down too much. This is also done to reduce the chance of driving through the parking brake, resulting in vehicle movement. A separate check is done on the output speed, so the calibration is stopped if the vehicle moves. The value of the clutch current for achieving the pressure or pressure condition indicative of the initial engagement is recorded.
It should also be noted that a baseline pressure value for power unit 12 is recorded at the beginning of the calibration or before the clutch is filled, since the pressure in the power unit 12 will not be zero when the clutch is not engaged, due to drag, and there will be a charge pressure in the unit to keep it cool. As non-limiting examples, representative pressure or pressure conditions that can be used as indicative of initial engagement can be a change in a difference between pressure values outputted by the sensors 110, and a minimum or threshold change in the difference can be required such that signal noise is not a factor.
Conditions for Clutch Calibration
If calibration is entered and the vehicle is moving, parking brake state is not changed (if it is off when entering calibration mode, it will remain off, if it was on when calibration mode is entered, it will remain on). Vehicle moving message is displayed. Operator must move the FNRP selector to P (Park) before and re-entering calibration mode.
Clutch Calibration Search Technique
Each calibration “step” will consist of the quickfill pulse and additional fill time. The additional fill may be extended slightly to give time to detect whether the clutch engaged or not. Ramping is not performed.
After each setup, the clutch will be fully dumped, and the next step will not proceed until a set time has elapsed that is long enough to ensure the clutch is fully empty.
The calibration will use a search technique to find the clutch fill current that results in any level of clutch engagement. The preferred search technique uses a “divide and conquer” type approach (depending on the tuning, the step size is limited and may not always “divide” on the second trial). First, a guess is made at the fill current and the test is performed to see if the pressure is too high or too low after this fill. Next the fill current is adjusted by an increment, either up or down, depending on whether it was too high or too low. For the next guess, if the fill current indicated in the opposite direction, the increment is cut in half (otherwise it is not).
During each calibration step, the Clutch Current under test will be displayed to the operator.
In order to detect clutch engagement, the remainder of the driveline must be completely engaged. Since it is desired to perform the calibration in a controlled, repeatable way, vehicle movement is not desirable, so the parking brake is on during calibration, and the driveline will work against it. To complete the torque path in this transmission, two clutches need to be engaged. The table below indicates the second clutch to be engaged for the given clutch under test.
In addition to engaging a second clutch, the SPA needs to be adjusted so the clutch under test will have a set speed difference across it. If the SPA were adjusted for zero speed across it (as if engaging powered zero), either no or little torque would develop. Illustrative values for SPA for each clutch is shown in the table below, and picked to create a large speed difference across the clutch.
If the engine speed falls too low during the calibration step, the clutch is immediately dumped to prevent engine stall. The minimum fill pressure shall be reduced for the next step.
Detecting Clutch Engagement
Clutch engagement is preferably detected by sensing rise of the difference between the values outputted by the pressure sensors 110 in the HSU. Generally, the pressure difference will start near zero when the clutches are not engaged, and rise in a linear fashion as the clutch torque increases. This is advantageous as it provides a very direct measure of only driveline torque, and does not vary with other factors such as other loads on the engine.
Rather than actually calculate the clutch pressure or driveline torque, the clutch calibration looks at the HSU pressure signals directly, since at the ideal kiss point, both the driveline torque and clutch torque should be zero. It is expected that there will be noise on the outputted signals of pressure sensors 110 and/or other issues with pressure fluctuations. The calibration looks for a change in the difference between the outputted pressure values of sensors 110, rather than a set level of the difference.
As noted above, the HSU pressure difference will not be zero when the driveline is completely disengaged, so the baseline pressure must be recorded prior to the quickfill pulse.
Use of hydrostatic power unit pressure for clutch calibration is advantageous as it does not depend on other factors, and it is a more direct measure of torque. As noted above, engine speed can vary as other loads on the engine change, such as hydraulic or PTO loads, various drags on the engine that may vary and such. Also, the engine governor is quite complex, and may change engine speeds due to complex algorithms to manage emissions, efficiency and other factors. How much the engine dips cannot always be a direct indication of the level of torque, which can result in different calibration values.
In light of all the foregoing, it should thus be apparent to those skilled in the art that there has been shown and described a novel METHOD OF CALIBRATING A HYDRAULICALLY OPERATED CLUTCH OF A CONTINUOUSLY VARIABLE TRANSMISSION USING PRESSURE BETWEEN A HYDROSTATIC PUMP AND MOTOR. However, it should also be apparent that, within the principles and scope of the invention, many changes are possible and contemplated, including in the details, materials, and arrangements of parts which have been described and illustrated to explain the nature of the invention. Thus, while the foregoing description and discussion addresses certain preferred embodiments or elements of the invention, it should further be understood that concepts of the invention, as based upon the foregoing description and discussion, may be readily incorporated into or employed in other embodiments and constructions without departing from the scope of the invention. Accordingly, the following claims are intended to protect the invention broadly as well as in the specific form shown, and all changes, modifications, variations, and other uses and applications which do not depart from the spirit and scope of the invention are deemed to be covered by the invention, which is limited only by the claims which follow.
This application claims the benefit of U.S. Provisional Application No. 61/527,523, filed Aug. 25, 2011.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/US2012/052573 | 8/27/2012 | WO | 00 | 2/24/2014 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2013/029058 | 2/28/2013 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 2097436 | Bennetch | Nov 1937 | A |
| 3810531 | Edmunds | May 1974 | A |
| 4055047 | Hara | Oct 1977 | A |
| 4102222 | Miller et al. | Jul 1978 | A |
| 4167855 | Knapp | Sep 1979 | A |
| 4310078 | Shore | Jan 1982 | A |
| 4489552 | Watanabe et al. | Dec 1984 | A |
| 4530416 | Kassai | Jul 1985 | A |
| 4543786 | Shuler | Oct 1985 | A |
| 4653350 | Downs et al. | Mar 1987 | A |
| 4759185 | McConnell et al. | Jul 1988 | A |
| 4811225 | Petzold et al. | Mar 1989 | A |
| 5184466 | Schniederjan et al. | Feb 1993 | A |
| 5224577 | Falck | Jul 1993 | A |
| 5337871 | Testerman | Aug 1994 | A |
| 5406793 | Maruyama et al. | Apr 1995 | A |
| 5449329 | Brandon et al. | Sep 1995 | A |
| 5467854 | Creger et al. | Nov 1995 | A |
| 5468198 | Holbrook et al. | Nov 1995 | A |
| 5531304 | Ishino et al. | Jul 1996 | A |
| 5540051 | Maruyama et al. | Jul 1996 | A |
| 5573473 | Asayama et al. | Nov 1996 | A |
| 5580332 | Mitchell et al. | Dec 1996 | A |
| 5671137 | Ishino et al. | Sep 1997 | A |
| 5684694 | Ishino et al. | Nov 1997 | A |
| 5737979 | McKenzie | Apr 1998 | A |
| 5853076 | McKee et al. | Dec 1998 | A |
| 5980411 | Wontner | Nov 1999 | A |
| 6080074 | Ulbrich et al. | Jun 2000 | A |
| 6088645 | Kawasaki et al. | Jul 2000 | A |
| 6115661 | Hosseini et al. | Sep 2000 | A |
| 6250077 | Iino et al. | Jun 2001 | B1 |
| 6285942 | Steinmetz et al. | Sep 2001 | B1 |
| 6292732 | Steinmetz et al. | Sep 2001 | B1 |
| 6295497 | Kuras | Sep 2001 | B1 |
| 6332860 | Hubbard et al. | Dec 2001 | B1 |
| 6442934 | Okuda et al. | Sep 2002 | B1 |
| 6481314 | Nemoto et al. | Nov 2002 | B2 |
| 6524205 | Irikura et al. | Feb 2003 | B1 |
| 6616559 | Hori et al. | Sep 2003 | B1 |
| 6672990 | Netzer | Jan 2004 | B2 |
| 6832978 | Buchanan et al. | Dec 2004 | B2 |
| 7037236 | Ishibashi et al. | May 2006 | B2 |
| 7063638 | Weeramantry | Jun 2006 | B2 |
| 7082757 | Teslak et al. | Aug 2006 | B2 |
| 7147239 | Teslak et al. | Dec 2006 | B2 |
| 7278953 | Meyer et al. | Oct 2007 | B2 |
| 7549287 | Foster et al. | Jun 2009 | B2 |
| 9115772 | Dix | Aug 2015 | B2 |
| 20030097874 | Milender et al. | May 2003 | A1 |
| 20050032605 | Booth et al. | Feb 2005 | A1 |
| 20080139363 | Williams | Jun 2008 | A1 |
| 20080194384 | League | Aug 2008 | A1 |
| 20080242464 | Kumazaki et al. | Oct 2008 | A1 |
| 20100312443 | Long | Dec 2010 | A1 |
| 20140200114 | Dix | Jul 2014 | A1 |
| Number | Date | Country |
|---|---|---|
| 102062206 | May 2011 | CN |
| 19739215 | Mar 1998 | DE |
| 1150031 | Oct 2001 | EP |
| 2053262 | Apr 2009 | EP |
| 2105626 | Sep 2009 | EP |
| Entry |
|---|
| Supplementary European Search Report in EP Application No. 12 82 5630.2, Nov. 25, 2016, 8 pages. |
| European Search Report in EP Application No. 12 82 5630, Aug. 11, 2016, 5 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20140207350 A1 | Jul 2014 | US |
| Number | Date | Country | |
|---|---|---|---|
| 61527523 | Aug 2011 | US |
