PatentDe  


Dokumentenidentifikation EP0616203 18.12.1997
EP-Veröffentlichungsnummer 0616203
Titel Motorzylinderdruckwandler
Anmelder General Motors Corp., Detroit, Mich., US
Erfinder Sellnau, Mark C., Bloomfield Hills, Michigan 48304, US
Vertreter Manitz, Finsterwald & Partner, 80538 München
DE-Aktenzeichen 69406699
Vertragsstaaten DE, GB
Sprache des Dokument En
EP-Anmeldetag 24.02.1994
EP-Aktenzeichen 942004623
EP-Offenlegungsdatum 21.09.1994
EP date of grant 12.11.1997
Veröffentlichungstag im Patentblatt 18.12.1997
IPC-Hauptklasse G01L 23/08

Beschreibung[en]

The present invention relates to a cylinder pressure sensor for an internal combustion engine.

It is known in the field relating to internal combustion engines to utilise cylinder pressure sensors for monitoring and controlling various processes, among them being combustion knock, misfire and associated combustion dilution and combustion phasing.

This invention relates to cylinder pressure sensors and more particularly to non-intrusive pressure sensors having no direct contact with the harsh environment of a cylinder. Non-intrusive pressure sensors measuring combustion pressure remote from direct contact with combustion are known which take various applications. Some applications include trapped force ring arrangements such as a piezoelectric annulus between cylinder head bolts and an engine block or alternatively between a spark plug seat and spark plug. Other non-intrusive pressure sensors include probe type and annular insert type which respond to flexure of a first wall defining in part a cylinder relative to a second rigid wall in an engine head. The latter mentioned types of non-intrusive pressure sensors, embodiments of which are set out in US Patent Nos. 4,601,196 and 4,969,352, improve upon signal quality shortfalls of the former mentioned types and are preferred among alternatives.

Some advantages of the preferred probe type and annular insert type pressure sensors over other non-intrusive pressure sensors include reduced dynamic phase lag, relative insensitivity to extraneous loads, lower operating temperature and other advantages due to their placement in proximity to the cylinder.

While the preferred non-intrusive pressure sensors present substantial improvements to the field of cylinder pressure sensing, further improvements can be made. For example, a certain degree of non-linear response and gain variation from pressure sensor to pressure sensor are present with the preferred related art pressure sensors. These undesirable characteristics tend to be exacerbated when the sensor structural material has a modulus of elasticity and yield strength of limited magnitude -- such as is characteristic of aluminium and alloys thereof. This presents practical problems since modern combustion engines commonly have cylinder heads of aluminium alloys and, for reasons of matching thermal expansion characteristics of the pressure sensor thereto, the sensor structural material is preferably similar.

Further, preferred non-intrusive pressure sensors may tend to be sensitive to variations in installation preload caused for example by piezoelectric sensitivity, thread effects and other interface effects thereby demanding relatively tight control over preload. It is desirable that the preferred pressure sensors be insensitive to installation variance such as is typically experienced when conventional torque monitoring is utilised to control threaded installations in an assembly environment, or alternatively as can be experienced in less controlled servicing situations.

A cylinder pressure sensor in accordance with the present invention is characterised by the features specified in the characterising portion of Claim 1.

The present invention involves the application of non-intrusive cylinder pressure sensors of the types generally functioning as those shown in the above mentioned US patents in an engine component, or more specifically, in an engine cylinder head. Structurally the cylinder head has a first wall defining a portion of a cylinder. The first wall flexes in response to variations in cylinder pressure. A second wall located a distance away from the first wall is relatively static with respect to variations in cylinder pressure. The first wall may provide mounting means for a spark plug with the pressure sensor configured as an annular insert installed within a spark plug access well between the first and second walls. The first wall may alternatively define a portion of a coolant channel in the cylinder head or comprise any other appropriate portion of the cylinder head defining in part a cylinder or having combustion forces translated directly thereto. Where the particular application so dictates, the pressure sensor may be configured substantially as a probe between a first and second wall.

In a preferred exemplary embodiment, the engine has a spark plug access well located at a top portion of each cylinder and substantially defined by an annular wall between the first wall and the second wall. The first wall defines in part the cylinder and has an opening for installation of a spark plug therethrough. The first wall further has a mounting boss of larger diameter than the spark plug. The second wall has an opening substantially defined by the annular wall. The pressure sensor takes the form of an annular insert with a top portion being engaged to the second wall and a lower portion thereof being in compressive engagement with the mounting boss. The load produced by combustion pressure is thus transferred through the first wall and split between the annular wall and the sensor. The portion of load transferred through the pressure sensor terminates locally at the second substantially static wall via the engagement interface therebetween. Linearity of pressure sensor response is determined to a great extent by the character of load transfer through the mounting threads that engage the pressure sensor to the second wall since the threads are relatively compliant compared to the remaining structure. It is therefore one object of the invention to provide substantially complete axial translation of axial sensor loads from the pressure sensor to the second wall along the response axis.

The present invention may also provide a pressure sensor having a wide range of acceptable preload force thereby reducing the need for high precision installation monitoring and making it suitable for conventional torque monitored installation.

The present invention may further provide a pressure sensor that exhibits improved durability and ability to withstand the substantial forces placed upon it while maintaining performance levels over time and cycles.

This invention benefits from and retains all advantages of the related art non-intrusive pressure sensors upon which it intended to improve.

The present invention will now be described, by way of example, with reference to the following detailed description, and the accompanying drawings, in which:-

  • Figures 1a-1c show related art non-intrusive cylinder pressure sensors;
  • Figures 2a-2c are graphs illustrating experimental results demonstrative of various characteristics of related art annular insert type non-intrusive cylinder pressure sensors;
  • Figure 3 shows an exemplary embodiment of a portion of an annular insert type non-intrusive cylinder pressure sensor according to the invention; and
  • Figures 4a-4c are graphs illustrating experimental results demonstrative of characteristics of an exemplary embodiment of an annular insert type non-intrusive cylinder pressure sensor according to the invention and as partially illustrated in Figure 3.

The present invention and detailed description of a preferred embodiment can best be understood when explained in relation to performance improvements over preferred related art pressure sensors. As such, significant description of related art sensor performance is given herein and is not presented as a recitation of any well known information in the art, rather it is presented conversely as illustrative of a significant discovery by the applicants through there research which forms the basis for the inventive aspects of the present invention.

Referring first to Figure 1a, an engine generally designated as numeral 10 has a plurality of engine cylinders 12 only one of which is shown in the diagram. Cylinders are defined in cylinder block 11 and are open at an upper surface 13. Engine cylinder head 16 closes on surface 13 to form each engine cylinder 12. Engine cylinder head 16 includes a lower wall 18 defining a portion of the engine cylinder 12 and a remote wall 34 not directly exposed to the engine cylinder. A coolant passage 35 is partially defined between walls 18 and 34. Wall 18 being exposed directly to the combustion pressure in the engine cylinder 12 flexes in response thereto. Remote wall 34 not being directly exposed to combustion pressure in the cylinder 12 is relatively rigid and does not flex in response to combustion therein.

Related art annular insert type pressure sensors conventionally employ compression mode piezoceramics as sensing elements. In theory, a piezoceramic can function linearly at zero preload. However, in practice small surface irregularities either on the piezoceramic or mating surface may compromise load transfer. Similarly, surface irregularities at any interface in the load path can compromise linearity of response. Surface bonding of components at their interfaces may reduce the effects of these irregularities. Alternately, and desirably for simplified assembly, sensor preload is used to force these surfaces together into substantially uniform contact across the sensing element cross-section.

A probe type pressure sensor generally designated as numeral 40 is shown in compressive preload arrangement along response axis 25. Pressure sensor 40 has a main portion 44 threadably engaged into a threaded opening in wall 34. A load transferring probe portion 47 extends from main portion 44 to engage wall 18. Flexure of wall 18 is transmuted along probe 47 to be measured by the pressure sensor 40. Ideally, the pressure sensor 40 is engaged to wall 34 such that no relative movement occurs therebetween. However, experience has shown that conventional threaded engagement may compromise load transfer and signal quality of the pressure sensor.

Referring now to Figure 1b, a similar probe type pressure sensor 52 is shown in tensile preload arrangement along response axis 25' in engine 50 cylinder head 51. Similar engine features between Figures 1a and 1b are designated by similar primed numbers. Wall 60 is directly exposed to combustion forces and wall 64 is remote therefrom. Pressure sensor 52 has main portion 56 passing through wall 64. Wall 60 has threaded opening 59 in boss 62 into which probe portion 58 is threadably engaged at threaded portion 63 thereof. Pressure sensor 52 has mounting head 55 with lower shoulders engaging block 51 such that the pressure sensor 52 is tensibly prestressed. In this arrangement with conventional threads, similar compromises in pressure sensor performance may be encountered.

Referring now to Figure 1c and yet another related art pressure sensor 36 is shown which is responsive to flexure of a first wall defining in part a cylinder relative to a second wall remote from direct exposure to cylinder pressure. Again, similar engine features between Figure 1a, 1b and 1c are designated by similar primed numbers. The pressure sensor 36 is shown as an annular insert type in compressive preload arrangement along response axis 25'' in engine 70 cylinder head 15. The pressure sensor 36 has a lower portion 39 engaged to annular boss 33 extending from wall 21 which is directly exposed to combustion pressure. Annular boss 33 is an annular ledge set back from an inner wall 30 of an access well of a spark plug 24. The access well is defined by annular wall 29 extending between wall 21 and wall 20. Upper portion 37 of the pressure sensor 36 is shown threadably engaged to upper wall 20. Again, in this arrangement the conventional threaded engagement of the pressure sensor 36 to an upper wall proves to be similarly problematic with respect to the performance of the pressure sensor.

Conventional threaded engagements, as commonly employed and previously referred to, have a load transfer surface, hereinafter load flank, and a relief flank making equal angles with a line perpendicular to the threaded member major axis. Examples include a 60° standard V-thread having a load flank angle of 30° and a relief flank angle of 30° and will hereafter be associated with conventional threaded engagement.

When used with sensor bodies that are thin walled, conventional threads may introduce radial loads that compromise linearity of load transfer axially through the pressure sensor. A sensor body of an annular insert type sensor may be considered as a thin-walled tube with 30° load flank that act as inclined planes which cause wedging when axial loads are applied to the sensor body. Conventional threads can operate in a first mode characterised by slip (relative sliding motion) between load flank interface surfaces and a second mode characterised by non-slip therebetween. Depending upon variables at the thread interface such as roughness, lube, hardness and plating alloy, slip behaviour may introduce hysteresis into the load path and ultimately into the sensor output signal. This type of signal error is expected to vary significantly and unpredictably. Non-slip behaviour may be caused by welding or galling of the threads, high friction coefficients or inadequate lubrication. Load transfer in this case would be quite linear since the interface effectively behaves as an integral and continuous structure. However, it may be too precarious a position to expect consistent non-slip behaviour induced by these factors will be obtained and thereafter maintained with any degree of predictability or repeatability. Additionally, galling of thread surfaces would cause irreparable harm and unacceptable damage thereto.

Radial loading associated with wedging at the conventional threaded engagement may also contribute to some loss of sensor preload. This would be especially true where the yield strength (creep resistance) of the sensor body material is relatively low or where the annular wall of the sensor body is narrow. Over many stress cycles at elevated temperatures, the pressure sensor could creep diametrically causing an associated axial unloading of preload leading to erroneous output and/or sensor failure.

Referring now to Figures 2a through 2c, the types of behaviour encountered with conventional threads used to engage an annular insert type pressure sensor to one of the two walls are demonstrated. The curves represent laboratory measurements obtained from quasi-static axial loading tests on a material test machine. A substantially uniform cylindrical body of aluminium alloy was used for the tests as a preferred material for use in actual pressure sensor applications in common aluminium alloy cylinder heads. Inner diameter of the cylindrical body was approximately 24.6 mm, outer diameter was approximately 31.9 mm (for a wall thickness of approximately 3.65 mm) and 60° standard V-threads were used as the conventional threads. The cylindrical body was secured at the threads to a static reference block and loaded axially. Axial deflection was recorded at two diametrically opposed positions around the cylindrical body using a pair of research grade sensors to measure axial load transfer behaviour. To avoid overlay of data for ease of readability, the figures are representative of data generated at only one of the research grade sensors. The figures are generally representative of the thread behaviour.

Referring specifically to Figure 2a, the vertical axis indicates thread deflection and the horizontal axis indicates axial load applied to the cylindrical body. This load was positioned on the axial centre line of the body and applied uniformly on the body face. A curve corresponding to one of the two diametrically opposed position is shown which indicates that as the cylindrical body is loaded and unloaded, such as in combustion cycles, hysteretic displacement results. This behaviour may be attributable to the various slip and non-slip behaviours of the threads.

Referring now to Figure 2b, the vertical axis indicates thread stiffness (k) and the horizontal axis again indicates load applied axially to the pressure sensor. The stiffness is by definition the reciprocal of the slope of deflection vs. load curves, such as that shown in Figure 2a. Raw data (unfiltered and unsmoothed) was used to generate the hatched curves of Figure 2b over several load and unload cycles. The solid line represents a best fit curve to the most linear and constant stiffness portion of the data. The curves demonstrate how the stiffness for conventional thread engagement varies with load. Stiffness in the case of conventional threads only approaches constancy (linear horizontal curve) at the far right of the curve through a narrow envelope of substantial load. Also noted in the curves of Figure 2b is the dependence that the conventional thread shows upon load and unload portions of a cycle caused by the hysteretic behaviour as previously described with reference to Figure 2a. Since the stiffness of the sensor threads has a direct influence upon gain and sensitivity, the operation of the pressure sensor with conventional threads would realistically be limited substantially to the narrow, large magnitude preload range exhibiting flat stiffness. This limits bandwidth of sensor operation and demands tight preload monitoring. The steeply sloping load-unload dependent portion of the curve delineates an undesirable range of operating preloads since even more precise preloading would be required in order to ensure that the sensor gain is within a band compatible with any interfaced control electronics. Ideally, one would like to see no slope or minimally sloping stiffness curve between a wide range of preload forces so that installation or reinstallation of a pressure sensor would not need to be closely monitored as far as preload force is concerned. Precise monitoring and/or inspection of preload force is generally quite expensive and cumbersome.

The curve in Figure 2c is derived from data obtained during an accelerated creep simulation test, the vertical axis indicating thread deflection and the horizontal axis indicating load. A cylindrical body having an annular wall of reduced thickness (approximately 50% that used in the hysteresis and stiffness tests of Figures 2a and 2b) was used in the accelerated creep testing. Identical thread forms were used (60° standard V-thread) The curve exhibits rapid deformation of the threaded interface within the first decade of load cycles and suggests that conventional threads in conjunction with a full thickness annular wall may exhibit similar deformation over a pressure sensor's lifetime.

Referring now to Figure 3, a cross sectional illustration of a threaded portion of an annular type pressure sensor shows a preferred thread form according to the invention. To avoid unnecessary duplication of illustrations, the circled portion of Figure 1c referenced as numeral 3 indicates the general area for application of the preferred thread form shown in Figure 3. Figure 3 is also consistent with the thread form tested in comparison to the conventional threads previously described. The thread arrangement is generically referred to as a buttress thread and has a load flank substantially normal to the direction of the preload and compressive forces and a relief flank triangulated therefrom at a relatively wide angle. Buttress thread (ANSI B1.9) is one exemplary type of buttress thread suited for application hereto and has a load flank angle of 7° and a relief angle of 45°. Other flank angle geometry's are fully within the contemplated scope of the present invention, ANSI B1.9 being mentioned herein as exemplary and not limiting. A preferred thread geometry, and the one illustrated in Figure 3 and tested in comparison to the 60° standard V-thread has a load flank angle of 5° and a relief angle of approximately 56°, both angles being measured relative to a line perpendicular to the threaded member's major axis. A portion of the sensor body wall is illustrated as 102.

Having an axial force introduced on the substantially normal surface limits the transfer of that force to a radial direction due to wedging at the threads. With conventional threads, the load bearing flank has a much greater angle offsetting it from a normal position relative the sensor axis and any force introduced on that load bearing flank would introduce radial compression of the pressure sensor, and deformation of the sensor threads and body sufficient to cause undesirable hysteretic, gain and creep behaviour heretofore explained and illustrated with reference to Figures 2a-2c. Preferably, for ease of manufacture using conventional high-speed cutting tools, the buttress type thread is machined with the load bearing flank not quite normal to the direction of the force. However, the load bearing flank is substantially normal to the load so as to greatly limit undesirable slip and non-slip behaviours.

Aluminium alloys are commonly used for cylinder heads because of mass and corrosion advantages over cast iron, but stiffness, creep and fatigue properties are somewhat poorer with respect thereto. For similar mass and corrosion advantages, aluminium alloys are an advantageous choice for the sensor body. Also, an aluminium alloy for the sensor body is preferred where the cylinder head to which it is matched is an aluminium alloy because of similar thermal expansion properties. Such matching of thermal expansion properties will reduce the effects that an expansion differential has upon sensor preload thereby allowing use of lower preloads and component stresses. Both improved durability and preload maintenance are expected. Without such a buttress thread for an aluminium alloy cylinder head and mating sensor body, conventional threads thereon will tend to introduce undesirable deformation upon the sensor body more so than they would if they were on a material with greater modulus of elasticity and yield strength. Therefore, the import of the present invention is heightened in view of modern combustion engines which commonly employ aluminium alloy heads. It is likewise desirable to prevent galling of the threads which is quite likely with raw aluminium components. This can be accomplished by anodic deposition of a file-hard coating or alternatively by establishing a nickel coating such as by electrolysis over the threads of the pressure sensor.

Figures 4a through 4c show improvements to the types of behaviour encountered with conventional threads and can be compared directly to corresponding Figures 2a through 2c. Again, the curves represent laboratory measurements obtained from quasi-static axial loading tests on a material test machine. The material tested and methods of testing correspond substantially identically to those which generated the data used in Figures 2a through 2c; the only difference being in the thread form used for the test. Here, the aluminium alloy body sports a buttress thread as heretofore described and illustrated in Figure 3 having a load flank angle of substantially 5° and a relief angle of substantially 56°. In directly comparing Figure 4a with Figure 2a, one can see the significant improvement in hysteretic behaviour. In the buttress thread arrangement, as load is applied and relaxed along the horizontal axis, one can see that the load and relax paths are not significantly separated from each other. The curve in Figure 4b, illustrative of the stiffness of the buttress thread arrangement and derived from the data used to generate the curves in Figure 4a, shows improved stiffness characteristics over that shown in Figure 2b. The range of preload forces corresponding to relatively constant stiffness is not only significantly wider, but also extends through lower preload forces. Also noted here is the lack of load-unload dependency shown in Figure 2b as caused by the hysteretic performance of conventional threads. Thus the broadened range of preload force through which stiffness is constant allows for a much larger tolerance in installation preload about a lesser nominal magnitude of preload, thereby eliminating precise preload force monitoring during original installation or reinstallation after service.

The curve in Figure 4c is, similar to that in Figure 2c, derived from data obtained during an accelerated creep simulation test, the vertical axis indicating thread deflection and the horizontal axis indicating load. An annular wall of reduced thickness was used in the accelerated testing. The curve shows no measured creep or deformation of buttress threads or increased hysteresis. These results suggest that buttress threads in conjunction with a full thickness annular wall may exhibit similar desirable performance properties over the pressure sensor's lifetime.

In addition to the preferred embodiment described in reference to Figure 3, other alternative arrangements would place buttress threads on a probe type pressure sensor similar to those shown in Figures 1a and 1b. Again, axial forces are not transmuted radially due to the substantially normal load bearing flank of the buttress thread. The threads in an arrangement such as that shown in Figure 1b are disposed on the lower probe member with the load bearing flank of the buttress thread supporting a tensile preload. Improvements and advantages heretofore described are equally applicable to these two probe type pressure sensor embodiments. Other embodiments envisioned within the scope and spirit of the present invention include engagement means other than threads such as keyed flanges provided that substantially all loads introduced axially are not transmuted to a radial direction.


Anspruch[en]
  1. A cylinder pressure sensor of the type being responsive to force along a response axis thereof and disposable between a first wall and a second wall of an engine component housing, the first wall substantially defining at least one side of a cylinder of an internal combustion engine and flexing in response to varying pressure therein and the second wall being relatively rigid, the cylinder pressure sensor comprising a sensor assembly including a first end and a second end substantially aligned with the response axis; the first end of the sensor assembly being engageable with one of the walls; and engaging means for engaging the second end of the sensor assembly with the other of the walls so as to pre-stress the cylinder pressure sensor along the response axis when disposed in the internal combustion engine; characterised by preventing means (100-102) for preventing loads generated along the response axis from being transmuted to radial loads upon the sensor assembly.
  2. A cylinder pressure sensor as claimed in Claim 1, wherein the preventing means comprises means for translating substantially all loads introduced along the prestressed response axis at one of the ends to the other of the ends without transmutation outside of the response axis.
  3. A cylinder pressure sensor as claimed in Claim 1 or Claim 2, wherein the engaging means has at least one load transfer interface surface substantially normal to the response axis.
  4. A cylinder pressure sensor as claimed in Claim 3, wherein the second end is substantially cylindrical and each load transfer interface surface extends radially outward therefrom.
  5. A cylinder pressure sensor as claimed in Claim 4, wherein each load transfer interface surface is delimited circumferentially around the response axis to define a respective arcuate flange.
  6. A cylinder pressure sensor as claimed in Claim 3, wherein the second end is substantially cylindrical and each load transfer interface surface comprises one surface of a respective buttress thread.
  7. A cylinder pressure sensor as claimed in Claim 3, wherein the second end is substantially cylindrical and each load transfer interface surface comprises one surface of a respective trapezoidal thread.
  8. A cylinder pressure sensor of the type being responsive to force along a response axis thereof and disposable between a first wall and a second wall of an engine component housing, the first wall substantially defining at least one side of a cylinder in an internal combustion engine and flexing in response to varying pressure therein and the second wall being relatively rigid, the cylinder pressure sensor comprising a sensor assembly including a first end and a second end, the first end having threads for engagement with one of the walls, the second end being engageable with the other of the walls, characterised in that the threads (100,101) provide substantially linear axial load transfer between the first end of the sensor assembly and the wall threadably engaged thereto when the cylinder pressure sensor is disposed in the internal combustion engine.






IPC
A Täglicher Lebensbedarf
B Arbeitsverfahren; Transportieren
C Chemie; Hüttenwesen
D Textilien; Papier
E Bauwesen; Erdbohren; Bergbau
F Maschinenbau; Beleuchtung; Heizung; Waffen; Sprengen
G Physik
H Elektrotechnik

Anmelder
Datum

Patentrecherche

Patent Zeichnungen (PDF)

Copyright © 2008 Patent-De Alle Rechte vorbehalten. eMail: info@patent-de.com